Tobacco Documents Online

REC'D JUL 8 128S = American Health Foundation ~ Preventive Medicine An Intermational Journal Devoted to Practice and Theory ACADEMIC PRESS, INC. San Diego Orlando New York London Tc*onto Montreal Sydney Tokyo TI BU 31549

REC'D JUL 8 128S = American Health Foundation ~ Preventive Medicine An Intermational Journal Devoted to Practice and Theory ACADEMIC PRESS, INC. San Diego Orlando New York London Tc*onto Montreal Sydney Tokyo TI BU 31549

Preventive Medicine EDITOR-IN-CHIEF SENIOR ASSOCIATE EDITOR CONSULTING EDITOR MANAGING EDITOR ASSISTANT EDITOR Ernst L. Wynder Charles B. Arnold Lesley A. Schurmann Candyce Kornblum Jeannette M. Sims American Health Foundation. 320 East 43rd Street. New York. New York 10017 U.S.A. ASSOCIATE EDITORS Jerome D. Cohen Steven Jonas Ovide F. Pomerleau Walter J. Rogan David Schottenfeld Moyses Szklo St. Louis Universi7y School of Medicine State University of New York at Stony Brook School of Medicine University of Connecticut School of Medicine National Institute of Environmental Health Sciences Memorial Sloan-Kettering Cancer Center The Johns Hopkins University School of Hygiene and Public Health CONSULTANTS Kunio Aoki Nagoya. Japan Keith Ball London, England E. E. Baulieu Bicetre. France Erik Bjelke Norway Gerald S. Berenson New Orleans, LA Lester Breslow Los Angeles. CA Jacob A. Brody Joseph Fraumeni, Jr. Bethesda, MD Silvio Garattini Milan, Italy Leon Gordis Baltimore, MD Merwyn R. Greenlick Portland, OR Peter Greenwald Bethesdd, MD John H-C. Ho Kowloon, Hong Kong Stephen B. Hulley San Francisco, CA Anthony B. Miller Toronto, Ontario. Canada Baruch Modan Tel Aviv. Israel Pekka Puska Helsinki. Finland Abraham Rivenson New York, NY Claude Robyn Brussels, Belgium Itving J. Selikoff New York, NY K. Shanmugaratnam 0 Bethesda, MD Donald B. Louria Newark, NJ Singapore ... Joseph W. Cullen Bethesda, MD Frits de Waard Bilthoven, The Netherlands Russell V. Luepker Minneapolis, MN Joseph L. Lyon Salt Lake City, UT Robert D. Sparks Battle Creek, MI Jerzy Staszewski Gliwice, Poland M ~ ~ Lennart Domellof Umed, Sweden Joseph D. Matarazzo Portland, OR Karl Uberla Berlin, West Germany Preventive Medicine is the offscial journal of the American Socie+of Preventive Oncology. I

Volume 13, Number 6, November 1984 Copyright ® 1984 by Academic Press, Inc. All'rights reserved No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the copyright owner. The appearance of the code at the bottom of the first page of an article in this journal in- dicates the copyright owner's consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copy- right Clearance Center. Inc. 127 Congress Street. Salem. Massachusetts 019701. for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-19it4 articles are as shown on the article title pages; if no fee code appears on the title page, the copy fee is the same as for current articles. 0091-7435/84 33.00 MADE IN THE UNITED STATES OF AMERICA PREVENTIVB MEDICINE Published bimonthly by Academic Press. lnc.. I11 Fifth Avenue. New York. New York 10003 1984: Volume 13. Price: S120.00 U.S.A. and Canada; i138.011 all other cuuntries 1985: Volume 14. Price: i129.50 U.S.A. and C:mada: 5148.SU all uther cuuntnes All prices include postage and handling. Information concerning personal subscription rates may be obtained by writing to the Publisher..For the special student rate, please see the [nformation for Authors. All correspondence and subscription orders should be addressed to Ihe Office of the Publishers at t I I Fitth Avenue, New York. N.Y. 10003. Send notices of change of address to the Office of the Publishers at least 6 to 8 weeks in advance. Please include both old and new addresses. POSTMASTER: Send changes of address to Preventive Medicine. I l l Fifth Avenue. New York. New York 10003. Second class postage paid at New Yortt. N.Y., and additional mailing offices. Copyright 0 1984 by Academic Press. Inc. TI BU 31551 T scic: o rcll mFlr tiun sive eng. the spe: iln y It YCle invi. thei H mur emc tant t:h4 Zac. etfu tlist cert usir Sllrt turG CUo n T tssu imp con T brinm iiasu e:1JC CnLc

Preventive Medicine EDITOR-IN-CHIEF SENIOR ASSOCIATE EDITOR CONSULTING EDITOR MANAGING EDITOR ASSISTANT EDITOR Ernst L. Wynder Charles B. Arnold Lesley A. Schurmann Candyce Kornblum Jeannette M. Sims American Health Foundation. 320 East 43rd Street. New York. New York 10017 U.S.A. ASSOCIATE EDITORS Jerome D. Cohen Steven Jonas Ovide F. Pomerleau Walter J. Rogan David Schottenfeld Moyses Szklo St. Louis Universi7y School of Medicine State University of New York at Stony Brook School of Medicine University of Connecticut School of Medicine National Institute of Environmental Health Sciences Memorial Sloan-Kettering Cancer Center The Johns Hopkins University School of Hygiene and Public Health CONSULTANTS Kunio Aoki Nagoya. Japan Keith Ball London, England E. E. Baulieu Bicetre. France Erik Bjelke Norway Gerald S. Berenson New Orleans, LA Lester Breslow Los Angeles. CA Jacob A. Brody Joseph Fraumeni, Jr. Bethesda, MD Silvio Garattini Milan, Italy Leon Gordis Baltimore, MD Merwyn R. Greenlick Portland, OR Peter Greenwald Bethesdd, MD John H-C. Ho Kowloon, Hong Kong Stephen B. Hulley San Francisco, CA Anthony B. Miller Toronto, Ontario. Canada Baruch Modan Tel Aviv. Israel Pekka Puska Helsinki. Finland Abraham Rivenson New York, NY Claude Robyn Brussels, Belgium Itving J. Selikoff New York, NY K. Shanmugaratnam 0 Bethesda, MD Donald B. Louria Newark, NJ Singapore ... Joseph W. Cullen Bethesda, MD Frits de Waard Bilthoven, The Netherlands Russell V. Luepker Minneapolis, MN Joseph L. Lyon Salt Lake City, UT Robert D. Sparks Battle Creek, MI Jerzy Staszewski Gliwice, Poland M ~ ~ Lennart Domellof Umed, Sweden Joseph D. Matarazzo Portland, OR Karl Uberla Berlin, West Germany Preventive Medicine is the offscial journal of the American Socie+of Preventive Oncology. I

PUtVENTIVE NEllIC1NE 13, 557 (1984) Medical Perspectives on Passive Smoking: Foreword to Passive Inhalation . The well-being of the public is the foremost responsibility of the physician and scientist alike. Considerations of cost, convenience, and custom must all be sub- urdinated to safeguarding health. ln an increasingly complex society, however, many public health issues involve low-order associations and, as such, are emo- tionalljr charged and prone to elude definitive scientitic/medical resolution. Pas- sive inhalation of tobacco smoke presents a case in point. The lay community is engaged in intense debate over laws regarding smoking in public places and in the workplace. At the same time, the scientific community is engaged in a broad spectrum of research to clarify the issue as to the breadth, or even existence, of any possible public health hazard. . it is precisely because this question is so intensely debated by the public and scientists alike that it is appropriate to convene meetings of those most intimately involved with the problem in order to review the state of the art of research from their respective points of view. History has, time and again, shown that where an association is not a de- monstrably strong one, different investigators employing different techniques will emerge with differing conclusions. As a consequence, it is fundamentally impor- tant in such cases to marshal investigators with distinct areas of expertise- chemistry, biology, physiology, epidemiology, and public health. Recognizing that each of these disciplines has limitations; their totality is such that the resultiilg effort brings to a scientific issue the maximum feasible clarity with the minimum distortion. This then permits us to determine whether the data at hand have a certain consistency and to recommend appropriate courses df action accordingly, using the process clearly described in the tirst Surgeon General's Report on Smoking and Health in a chapter entitled "Criteria of Judgment." It is the uni- formity and degree of various associations that makes the case for an associa- tion-or fails to do so. The papers presented here represent the state of the art for data regarding the issue of passive inhalation as it exists today. Undoubtedly, it will be refined and improved upon in the future. 'Thus, the materials contained herein should be considered a catalyst for further inquiry into this important public health issue. These papers provide a valuable addition to the body of scientific literature and bring us closer to achieving a consensus of opinion as to the existence of an association between passive inhalation and lung cancer and other chronic dis- eases. The public deserves a definitive answer. We trust that the Vienna confer- ence will hasten the day when that answer is at hand. GERHARD LEHNERT Direktur, Zetrtrulinstitut Jitr Arbeiesnredizin .adupk-Schunf'rlder-Sn•usse J D-?OU Huutburg 76, Fetlerul Republic ul'Gerutuiry ERNST L. WYNDER Presidertt. Autrrirun Hculth Furcndutiua 320 East 43rd Strert, Nrw York, 1Yrw Yurk 1(1017 557 0091-7435184 53.00 l'upynghl G tYWt by .icaJemic Pre». lnc. All nt{hla ui repruducuue in rey turm reaerved. TI BU 31552

Volume 13, Number 6, November 1984 Copyright ® 1984 by Academic Press, Inc. All'rights reserved No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the copyright owner. The appearance of the code at the bottom of the first page of an article in this journal in- dicates the copyright owner's consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copy- right Clearance Center. Inc. 127 Congress Street. Salem. Massachusetts 019701. for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-19it4 articles are as shown on the article title pages; if no fee code appears on the title page, the copy fee is the same as for current articles. 0091-7435/84 33.00 MADE IN THE UNITED STATES OF AMERICA PREVENTIVB MEDICINE Published bimonthly by Academic Press. lnc.. I11 Fifth Avenue. New York. New York 10003 1984: Volume 13. Price: S120.00 U.S.A. and Canada; i138.011 all other cuuntries 1985: Volume 14. Price: i129.50 U.S.A. and C:mada: 5148.SU all uther cuuntnes All prices include postage and handling. Information concerning personal subscription rates may be obtained by writing to the Publisher..For the special student rate, please see the [nformation for Authors. All correspondence and subscription orders should be addressed to Ihe Office of the Publishers at t I I Fitth Avenue, New York. N.Y. 10003. Send notices of change of address to the Office of the Publishers at least 6 to 8 weeks in advance. Please include both old and new addresses. POSTMASTER: Send changes of address to Preventive Medicine. I l l Fifth Avenue. New York. New York 10003. Second class postage paid at New Yortt. N.Y., and additional mailing offices. Copyright 0 1984 by Academic Press. Inc. TI BU 31551 T scic: o rcll mFlr tiun sive eng. the spe: iln y It YCle invi. thei H mur emc tant t:h4 Zac. etfu tlist cert usir Sllrt turG CUo n T tssu imp con T brinm iiasu e:1JC CnLc

PREVENTIVE MEDICINE 13, 558 (1984) Proceedings of the International Symposium on Medical Perspectives on Passive Smoking' Opening Address As president of the Austrian Society for Occupational Medicine, I welcome you all most cordially to the symposium "Medical Perspectives on Passive Smoking." I would like to extend a special welcome to our guest from Bavaria, Landesminister of Labour and Social Affairs, Dr. Pirkl. I would also like 'to welcome our own Federal Minister of Health and the Environment, Dr. Kurt Steyrer, and the initiator-or at least one of the initiators-of this event, the president of the German Society for Occupational Medicine, Professor Valentin. Finally, as a representative of our guests from abroad, I welcome the president of the American Health Foundation, Dr. Ernst Wynder. Ladies and gentlemen, we are very proud that you have assembled here in Austria for this symposium on passive smoking, not only because the state has had a monopoly on tobacco for the past 200 years, but also because we are probably one of the first countries, within the framework of legislation aimed at providing protection for employees, to give consideration to the problem of pas- sive smoking. The situation with respect to the implementation of the relevant law is, perhaps, another matter, but at least we have created a possibility aimed at ensuring the,"rights" of the nonsmoker vis-a-vis the smoker. For your infor- mation, 30% of the overall Austrian population are smokers. Broken down into men and women, the most recent available statistics show that from the age of 16, 42% of Austrian men and 28% of the women smoke. The remaining percentage of the population are, therefore, passive smokers. I believe we have a heavy program before us: I now officially declare this meeting open, and call upon the next speaker, Professor Valentin, to address a few words totyou. . t April 9-12. 1984. Vienna. Auatria. fx)91-7435184 33.00 CupynyUl s: IYtl4 by Ac;Wemrc Preaa. lne. All nl{hta u/ «pruJualua in any furm rexcvtyl. E. BAUMGARTNER Austrian Society for Occupational Medicine Leiter des Arbeitsmedizinschen Zentrums A-6060 Hall Tirol Tirol, Austria 558 TI BU 31553 vW, reF

PREVENTIVE MEDICINE 13, 558 (1984) Proceedings of the International Symposium on Medical Perspectives on Passive Smoking' Opening Address As president of the Austrian Society for Occupational Medicine, I welcome you all most cordially to the symposium "Medical Perspectives on Passive Smoking." I would like to extend a special welcome to our guest from Bavaria, Landesminister of Labour and Social Affairs, Dr. Pirkl. I would also like 'to welcome our own Federal Minister of Health and the Environment, Dr. Kurt Steyrer, and the initiator-or at least one of the initiators-of this event, the president of the German Society for Occupational Medicine, Professor Valentin. Finally, as a representative of our guests from abroad, I welcome the president of the American Health Foundation, Dr. Ernst Wynder. Ladies and gentlemen, we are very proud that you have assembled here in Austria for this symposium on passive smoking, not only because the state has had a monopoly on tobacco for the past 200 years, but also because we are probably one of the first countries, within the framework of legislation aimed at providing protection for employees, to give consideration to the problem of pas- sive smoking. The situation with respect to the implementation of the relevant law is, perhaps, another matter, but at least we have created a possibility aimed at ensuring the,"rights" of the nonsmoker vis-a-vis the smoker. For your infor- mation, 30% of the overall Austrian population are smokers. Broken down into men and women, the most recent available statistics show that from the age of 16, 42% of Austrian men and 28% of the women smoke. The remaining percentage of the population are, therefore, passive smokers. I believe we have a heavy program before us: I now officially declare this meeting open, and call upon the next speaker, Professor Valentin, to address a few words totyou. . t April 9-12. 1984. Vienna. Auatria. fx)91-7435184 33.00 CupynyUl s: IYtl4 by Ac;Wemrc Preaa. lne. All nl{hta u/ «pruJualua in any furm rexcvtyl. E. BAUMGARTNER Austrian Society for Occupational Medicine Leiter des Arbeitsmedizinschen Zentrums A-6060 Hall Tirol Tirol, Austria 558 TI BU 31553 vW, reF

something of the spirit of this city, which has something warm and unifying about it. Austria has traditionally exercised a sort of bridging function; we see ourselves as mediators between ideological principles and scientific evidence. I am there- fore very happy that you have selected Vienna as your venue, and wish you a pleasant stay here. DR. KURT STEYRER Ministry of Health and Environmental Protection Stubenring 1 A-1010 Vienna, Austria ,

something of the spirit of this city, which has something warm and unifying about it. Austria has traditionally exercised a sort of bridging function; we see ourselves as mediators between ideological principles and scientific evidence. I am there- fore very happy that you have selected Vienna as your venue, and wish you a pleasant stay here. DR. KURT STEYRER Ministry of Health and Environmental Protection Stubenring 1 A-1010 Vienna, Austria ,

eRL•VtNTtvE MEllIC1Nli 13, 55y-5611 (1984) :al !lcome 'assive avaria, like to -. Kurt nt, the dentin. -sident lere in ate has we are med at of pas- :levant aimed r infor- vn into age of 0 : entage .n: this 1ress a RTN ER -clicine ntrums It Tirul Uustria Proceedings of the International Symposium on Medical Perspectives on Passive Smoking' Opening Address Mr. Federal Minister, Mr. Landesminister, ladies, and gentlemen: On behalf of the executive committee of the German Society of Occupational Medicine, and also on behalf of the standing committee, the supervisory com- mittee (Kuratorium), and the managing directors of the Bavarian Academy for Occupational and Social Medicine, I should like to welcome you most cordially to our convention of experts. My special thanks go to the Austrian Federal Min- ister of Health and the Environment, Dr. Kurt Steyrer, who has undertaken the patronage of this event, and to the Bavarian Minister of Labor and Social Affairs, Dr. Fritz Pirkl, who willingly agreed to take on the sponsorship. Mention must also be made of the World Health Organization in Geneva and the International Green Cross, also in Geneva, whom I should also like to thank for their ready cooperation. My special greetings to the active participants from all countries represented, who deserve our thanks for their readiness to come here and report on their results, insights, and experience. Our topic, "Medical Perspectives on Passive Smoking," has, over the last 14 years, attracted increasi,ng attention in numerous areas of public life and has acquired considerable importance. This applies not only to the working world, but also to the environment. In this connection, I should like to remind you that as long ago as the spring of 1977, the Bavarian Academy for Occupational and Social Medicine organized a convention, "Passive Smoking at the Workplace," in Munich. Together with numerous specialists from the German-speaking countries, an interdisciplinary scientific stock-taking of the phenomenon of passive smoking was then undertaken, and an analysis of social, medical, and legal aspects was made. At that time, the impression was gained that the topie had been dealt with objectively, appropriately, and exhaustively. Since then, however, the phenom- enon of passive smoking has acquired new international dimensions and topicality through the publication of further epidemiological, toxicological, and function- analytical studies. As a result, discussions on the dangers to the health of passive smoking have been both expanded and intensified, and have also become increas- ingly polarized. This has been the stimulus for the convening from across national borders of a number of scientific organizations, together with representatives of international research and science, with the aim of discussing the present situation with respect to the passive smoking problem. The participation of high-ranking representatives of the.Warid Health Organization. and the adoption by this or- 1 Aprit 9-12. 1984, VVienna. Austria. 559 TI BU 31554 0091-74351h4 i3.110 Cupyngtx ti IYlH by AcuJnmc Pre». Inc. .111 nghls ui ropruJUCuuo in sny iurm rcxrvcJ.

560 H. VALENTIN ganization of the scientific patronage of this symposium serve as evidence of the world-wide interest in an objective scientific clarification of the questions in- voived. In the spring of 1983, the societies and institutions represented on the organizing committee decided to hold a further convention with the aim of re- viewing the present state of our knowledge and detining the importance of the issue from the medical point of view. We all shared the opinion that the time was ripe for a new scientific symposium. It might be emphasized: here that, as I see it, the statutes and regulations of the organizing societies and institutions give them not only the right but also the duty to declare their position on current problems of occupational and social medicine, and on the protection of health and the assessment of risks. In general, this is realized in the form of special events in which experts in.the field concerned give a lecture which is then fol- lowed by discussions among a•wider circle of participants. The scientific program over the next two days is compact. An attempt has been made to give suitable consideration to all aspects of the problem. The objective of our joint discussions, free of all emotion, ideology, or even utopian ideas, is_ to arrive at scientifically immaculate and verifiable results. In this manner, it is to be hoped that science can contribute toward ensuring that the politicians re- ceirve reliable, consistent data, with an eye toward enabling them to enact positive legislative measures. Before concluding, I would like to say a word about our venue and our host country. lt was no accident that Vienna was selected as the site of our meeting. Since April of last year, Austria has become the first European country to pass a workers' protection act aimed at protecting the employee from the effects of tobacco smoke. In my view, this new regulation incorporates both the interna- tionally recognized penchant of the Austrians for due diplomatic procedure, and alsu the possibility of keeping open the door to the latest scientific insights. While recognizing the need to protect the nonsmoker from tobacco smoke in the work- place, the act nevertheless avoids demanding a blanket prohibitiQn on smoking. Rather, it is flexible enough to take into account existing conditions in the work- place by requiring a regulation adapted to the prevailing situation. The spectrum of possible measures extends from improved room ventilation to the physical separation of smokers and nonsmokers and localized no-smoking areas. lf leg- islative control is considered necessary at all, the Austrian regulations would seem to be both appropriate and moderate. It is, however, of interest, and this should perhaps provide us with food for thought. that, in its protection program for nonsmokers, the Federal Republic of Germany, which is otherwise known for the perfectionism of its legislation, merely makes recommendations with no legal character, putting its main hope in the ability of smokers and nonsmokers to reach an amicable agreement. I end my address of welcome with the hope that, in addition to the scientific discussions, you will also find time to take a look at some of the sights that this venerable, cultured, and tradition-rich city of Vienna has to offer. Thank you. H. VALENTIN Institute for Occupational and Social Medicine University of Erlungen-Niirnberg Schillerstrasse 25-29 D-8520 Erlangen, Federal Republic of Germany TI BU 31555 rKiV Mr. 4 mu: Mir ia 11 etfe LI&L mir thia furt wu yea itut Sal unl Jut sut an cur clu t Un in shc ~s. thu: the

560 H. VALENTIN ganization of the scientific patronage of this symposium serve as evidence of the world-wide interest in an objective scientific clarification of the questions in- voived. In the spring of 1983, the societies and institutions represented on the organizing committee decided to hold a further convention with the aim of re- viewing the present state of our knowledge and detining the importance of the issue from the medical point of view. We all shared the opinion that the time was ripe for a new scientific symposium. It might be emphasized: here that, as I see it, the statutes and regulations of the organizing societies and institutions give them not only the right but also the duty to declare their position on current problems of occupational and social medicine, and on the protection of health and the assessment of risks. In general, this is realized in the form of special events in which experts in.the field concerned give a lecture which is then fol- lowed by discussions among a•wider circle of participants. The scientific program over the next two days is compact. An attempt has been made to give suitable consideration to all aspects of the problem. The objective of our joint discussions, free of all emotion, ideology, or even utopian ideas, is_ to arrive at scientifically immaculate and verifiable results. In this manner, it is to be hoped that science can contribute toward ensuring that the politicians re- ceirve reliable, consistent data, with an eye toward enabling them to enact positive legislative measures. Before concluding, I would like to say a word about our venue and our host country. lt was no accident that Vienna was selected as the site of our meeting. Since April of last year, Austria has become the first European country to pass a workers' protection act aimed at protecting the employee from the effects of tobacco smoke. In my view, this new regulation incorporates both the interna- tionally recognized penchant of the Austrians for due diplomatic procedure, and alsu the possibility of keeping open the door to the latest scientific insights. While recognizing the need to protect the nonsmoker from tobacco smoke in the work- place, the act nevertheless avoids demanding a blanket prohibitiQn on smoking. Rather, it is flexible enough to take into account existing conditions in the work- place by requiring a regulation adapted to the prevailing situation. The spectrum of possible measures extends from improved room ventilation to the physical separation of smokers and nonsmokers and localized no-smoking areas. lf leg- islative control is considered necessary at all, the Austrian regulations would seem to be both appropriate and moderate. It is, however, of interest, and this should perhaps provide us with food for thought. that, in its protection program for nonsmokers, the Federal Republic of Germany, which is otherwise known for the perfectionism of its legislation, merely makes recommendations with no legal character, putting its main hope in the ability of smokers and nonsmokers to reach an amicable agreement. I end my address of welcome with the hope that, in addition to the scientific discussions, you will also find time to take a look at some of the sights that this venerable, cultured, and tradition-rich city of Vienna has to offer. Thank you. H. VALENTIN Institute for Occupational and Social Medicine University of Erlungen-Niirnberg Schillerstrasse 25-29 D-8520 Erlangen, Federal Republic of Germany TI BU 31555 rKiV Mr. 4 mu: Mir ia 11 etfe LI&L mir thia furt wu yea itut Sal unl Jut sut an cur clu t Un in shc ~s. thu: the

PUtVENTIVE NEllIC1NE 13, 557 (1984) Medical Perspectives on Passive Smoking: Foreword to Passive Inhalation . The well-being of the public is the foremost responsibility of the physician and scientist alike. Considerations of cost, convenience, and custom must all be sub- urdinated to safeguarding health. ln an increasingly complex society, however, many public health issues involve low-order associations and, as such, are emo- tionalljr charged and prone to elude definitive scientitic/medical resolution. Pas- sive inhalation of tobacco smoke presents a case in point. The lay community is engaged in intense debate over laws regarding smoking in public places and in the workplace. At the same time, the scientific community is engaged in a broad spectrum of research to clarify the issue as to the breadth, or even existence, of any possible public health hazard. . it is precisely because this question is so intensely debated by the public and scientists alike that it is appropriate to convene meetings of those most intimately involved with the problem in order to review the state of the art of research from their respective points of view. History has, time and again, shown that where an association is not a de- monstrably strong one, different investigators employing different techniques will emerge with differing conclusions. As a consequence, it is fundamentally impor- tant in such cases to marshal investigators with distinct areas of expertise- chemistry, biology, physiology, epidemiology, and public health. Recognizing that each of these disciplines has limitations; their totality is such that the resultiilg effort brings to a scientific issue the maximum feasible clarity with the minimum distortion. This then permits us to determine whether the data at hand have a certain consistency and to recommend appropriate courses df action accordingly, using the process clearly described in the tirst Surgeon General's Report on Smoking and Health in a chapter entitled "Criteria of Judgment." It is the uni- formity and degree of various associations that makes the case for an associa- tion-or fails to do so. The papers presented here represent the state of the art for data regarding the issue of passive inhalation as it exists today. Undoubtedly, it will be refined and improved upon in the future. 'Thus, the materials contained herein should be considered a catalyst for further inquiry into this important public health issue. These papers provide a valuable addition to the body of scientific literature and bring us closer to achieving a consensus of opinion as to the existence of an association between passive inhalation and lung cancer and other chronic dis- eases. The public deserves a definitive answer. We trust that the Vienna confer- ence will hasten the day when that answer is at hand. GERHARD LEHNERT Direktur, Zetrtrulinstitut Jitr Arbeiesnredizin .adupk-Schunf'rlder-Sn•usse J D-?OU Huutburg 76, Fetlerul Republic ul'Gerutuiry ERNST L. WYNDER Presidertt. Autrrirun Hculth Furcndutiua 320 East 43rd Strert, Nrw York, 1Yrw Yurk 1(1017 557 0091-7435184 53.00 l'upynghl G tYWt by .icaJemic Pre». lnc. All nt{hla ui repruducuue in rey turm reaerved. TI BU 31552

PRBVENTIVE MEDICINE 13, 561-562 (1984) the ~ in- the t re- the was see give rent :alth -cial tol- ieen :tive ;, is it is s re- itive host ting. pass ts of : rna- and /hile ork- :ing. c)rk- :rum sical leg- ould this ;rarn lfur egal -s to ;tific this )u. vTIN rine berg '5-?9 runy Proceedings of the International Symposium on Medical Perspectives on Passive Smoking' Opening Address Mr. L.andesminister, committee members, ladies, and gentlemen: As Federal Minister of Health and Environment, I should like to welcome you most cordially to Vienna and to Austria. For me, both as a physician and as the Minister of Health, this symposium is extremely important. We know-and there is little controversy on this point-that active smoking has extremely injurious effects on the respiratory and cardiovascular systems. But, time and again, the cigarette manufacturing industry and the highly expensive advertising in this area minimize the effects of passive smoking. For this reason, I hope very much that this symposium will provide data that will help the ministry make decisions on further action. On the basis of my experience as a doctor of medicine (I have worked both as a doctor at a large firm and privately as a dermatologist for 30 years and am well acquainted with much of the damage caused by smoking), we initiated a large-scale campaign against smoking. My predecessor in office, Dr. Saicher, rode into battle with the slogan "Ohne Rauch geht's auch" (don't smoke unless you have to). I have, however, become pessimistic about the ability of the smoker to change. Therefore, I would like to ask you to produce scientific evi- dence for the dangers of passive smoking, so that our injunction can be medically tiubstantiated. An important action instituted by my ministry was the founding of an organization for patient information and education. I am sure that we shall continue with these measures, although, as I have already mentioned, they are clouded for me by a certain degree of pessimism. If the scientist can "set the tiignals." then, I believe, the politicians must act. One of the most important priorities is the fight against smoking in the schools. Unfortunateiy,.Austria has permitted the introduction of special smoking rooms in tchools. I am of one opinion with the Minister of Education that these rooms ihould be abolished, as they represent a clear sign that the state accepts smoking as a normal social phenomenon. For this reason, we must oppose these school smoking rooms. On the other hand, the workers' protection law is a useful mea- !iure that can provide protection for nonsmokers at their places of work. We in Austria will continue along this diplomatic path-and, Mr. President, I thank you for your recognition of our efforts-aimed at providing protection for the nonsmoker through a flexible application of legislation. I hope, ladies and gentlemen, that despite the inclement weather you will feel . ' April 9-12. 1984. Vienna, Austria. 561 001-7435/84 S3.00 CopYnglu i 1%4 by Ac;udemic Yreaa- lne. Ali nghts ul repruJucUUn in any turm rcaarveJ. TI BU 31556

PlttVl:NfIVE NtillIC1NE 13, 563-56411984) about ielves there- you a :YRER t'Ctlolt ring I ustria 0 Proceedings of the International Symposium on Medical Perspectives on Passive Smoking' Opening Address Mr. Federal Minister, Mr. President, ladies, and gentlemen: First of all, I should like to thank the organizers for making this meeting pos- ~.ible and for all,the effort they have-put into the preparations necessary for such a meeting. I would also like to express my sincere thanks to you, Mr. Federal Minister, for your hospitality and your friendly words of welcome. Finally, I thank all the participants for taking the trouble to attend and contribute. When I agreed to the plans of the Bavarian Academy for Occupational and Social Medicine to hold a working meeting on "Passive Smoking at the Work- place" in the spring of 1977 and took over its patronage, 1 felt the meeting fulfilled an urgent need for sound scientific information. It had been preceded by nu- merous complaints presented by nonsmoker initiatives, questions asked by mem- bers of parliament in the House, court decisibns that resulted in newspaper head- lines, and, above all, by the repeated demands of a pugnacious Bavarian doctor and university professor, who called for the prohibition of smoking at the work- place as the "dictate of the hour." Prior to the meeting, a joint notification was issued by the Bavarian Chancellery and the nine Bavarian Ministries, aimed at protecting the passive smoker working for the public authorities. With this noti- fication, the Bavarian State Government wished to set an example with a regu- lation based on consideration for others and tolerance. A notice wa,S signed by the 10 heads of the Bavarian ministerial offices, 7 of them nonsmokers and 3 smokers, that is to say, a mixed body. The 1977 meeting, in brief, concluded that: (a) Nonsmokers may be inconvenienced, and might pos- sibly be hampered in their work, but their health is not endangered; (b) it is the task of a health policy to prevent-as far as this is possible-damage to health. In accordance with the state of our knowledge at that time, as expressed in the discussions, no necessity was therefore seen for legislation: and (c) in the absence of tolerance, it is not possible to achieve a satisfactory state of balance between the opposing interests of smokers and nonsmokers. Response to this event was mixed. In general, reporting by the press was objective, and the discussions and the papers published in a brochure were pos- itively assessed by the public at large as a welcome clarification of a diffiCult problem. However, the opponents of tobacco, and in particular, the proponents of the theory that passive smoking has a detrimental effect on health, expressed dissatisfaction,. sometimes of a polemic nature,. and wme made personal attacks ' April 9-12, 1984. VVienna, Austria. 563 0091-7435/84 33.110 Cupynght il 1994 by .{o:ulemta P'eas. Inc. All nghta o/ repruJuctwn tn .ny turm reaGrved. TI BU 31558 0

PRBVENTIVE MEDICINE 13, 561-562 (1984) the ~ in- the t re- the was see give rent :alth -cial tol- ieen :tive ;, is it is s re- itive host ting. pass ts of : rna- and /hile ork- :ing. c)rk- :rum sical leg- ould this ;rarn lfur egal -s to ;tific this )u. vTIN rine berg '5-?9 runy Proceedings of the International Symposium on Medical Perspectives on Passive Smoking' Opening Address Mr. L.andesminister, committee members, ladies, and gentlemen: As Federal Minister of Health and Environment, I should like to welcome you most cordially to Vienna and to Austria. For me, both as a physician and as the Minister of Health, this symposium is extremely important. We know-and there is little controversy on this point-that active smoking has extremely injurious effects on the respiratory and cardiovascular systems. But, time and again, the cigarette manufacturing industry and the highly expensive advertising in this area minimize the effects of passive smoking. For this reason, I hope very much that this symposium will provide data that will help the ministry make decisions on further action. On the basis of my experience as a doctor of medicine (I have worked both as a doctor at a large firm and privately as a dermatologist for 30 years and am well acquainted with much of the damage caused by smoking), we initiated a large-scale campaign against smoking. My predecessor in office, Dr. Saicher, rode into battle with the slogan "Ohne Rauch geht's auch" (don't smoke unless you have to). I have, however, become pessimistic about the ability of the smoker to change. Therefore, I would like to ask you to produce scientific evi- dence for the dangers of passive smoking, so that our injunction can be medically tiubstantiated. An important action instituted by my ministry was the founding of an organization for patient information and education. I am sure that we shall continue with these measures, although, as I have already mentioned, they are clouded for me by a certain degree of pessimism. If the scientist can "set the tiignals." then, I believe, the politicians must act. One of the most important priorities is the fight against smoking in the schools. Unfortunateiy,.Austria has permitted the introduction of special smoking rooms in tchools. I am of one opinion with the Minister of Education that these rooms ihould be abolished, as they represent a clear sign that the state accepts smoking as a normal social phenomenon. For this reason, we must oppose these school smoking rooms. On the other hand, the workers' protection law is a useful mea- !iure that can provide protection for nonsmokers at their places of work. We in Austria will continue along this diplomatic path-and, Mr. President, I thank you for your recognition of our efforts-aimed at providing protection for the nonsmoker through a flexible application of legislation. I hope, ladies and gentlemen, that despite the inclement weather you will feel . ' April 9-12. 1984. Vienna, Austria. 561 001-7435/84 S3.00 CopYnglu i 1%4 by Ac;udemic Yreaa- lne. Ali nghts ul repruJucUUn in any turm rcaarveJ. TI BU 31556

eRL•VtNTtvE MEllIC1Nli 13, 55y-5611 (1984) :al !lcome 'assive avaria, like to -. Kurt nt, the dentin. -sident lere in ate has we are med at of pas- :levant aimed r infor- vn into age of 0 : entage .n: this 1ress a RTN ER -clicine ntrums It Tirul Uustria Proceedings of the International Symposium on Medical Perspectives on Passive Smoking' Opening Address Mr. Federal Minister, Mr. Landesminister, ladies, and gentlemen: On behalf of the executive committee of the German Society of Occupational Medicine, and also on behalf of the standing committee, the supervisory com- mittee (Kuratorium), and the managing directors of the Bavarian Academy for Occupational and Social Medicine, I should like to welcome you most cordially to our convention of experts. My special thanks go to the Austrian Federal Min- ister of Health and the Environment, Dr. Kurt Steyrer, who has undertaken the patronage of this event, and to the Bavarian Minister of Labor and Social Affairs, Dr. Fritz Pirkl, who willingly agreed to take on the sponsorship. Mention must also be made of the World Health Organization in Geneva and the International Green Cross, also in Geneva, whom I should also like to thank for their ready cooperation. My special greetings to the active participants from all countries represented, who deserve our thanks for their readiness to come here and report on their results, insights, and experience. Our topic, "Medical Perspectives on Passive Smoking," has, over the last 14 years, attracted increasi,ng attention in numerous areas of public life and has acquired considerable importance. This applies not only to the working world, but also to the environment. In this connection, I should like to remind you that as long ago as the spring of 1977, the Bavarian Academy for Occupational and Social Medicine organized a convention, "Passive Smoking at the Workplace," in Munich. Together with numerous specialists from the German-speaking countries, an interdisciplinary scientific stock-taking of the phenomenon of passive smoking was then undertaken, and an analysis of social, medical, and legal aspects was made. At that time, the impression was gained that the topie had been dealt with objectively, appropriately, and exhaustively. Since then, however, the phenom- enon of passive smoking has acquired new international dimensions and topicality through the publication of further epidemiological, toxicological, and function- analytical studies. As a result, discussions on the dangers to the health of passive smoking have been both expanded and intensified, and have also become increas- ingly polarized. This has been the stimulus for the convening from across national borders of a number of scientific organizations, together with representatives of international research and science, with the aim of discussing the present situation with respect to the passive smoking problem. The participation of high-ranking representatives of the.Warid Health Organization. and the adoption by this or- 1 Aprit 9-12. 1984, VVienna. Austria. 559 TI BU 31554 0091-74351h4 i3.110 Cupyngtx ti IYlH by AcuJnmc Pre». Inc. .111 nghls ui ropruJUCuuo in sny iurm rcxrvcJ.

SAA FRITZ PIRKL against the organizers and the presenters of scientific papers. In responding to their numerous letters, I pointed out the many legal modes already available to provide protection from annoyance by smokers, both for the private individual and for employees. This, however, had no effect on the legislative calls for action either in Germany or in other countries. Since then, the discussion gradually has spread into other areas where smokers and nonsmokers come together. I must admit, ladies and gentlemen, that I also experience a certain unease. Was that university professor right, after all, when, he listed 40 carcinogenic substances in tobacco smoke and declared that not only the active smoker, but also the passive smoker, was at risk of contracting cancer'' Nor am I sure that ministerial appeals to smokers to show consideration and to nonsmokers to practice tolerance suffice to calm the fears of so many people. These considerations are in themseiv~s proof of the urgent need for a sym- posium such as this. We simply cannot do without further objective clarification of the problem. And thus, along with you, Mr. Federal Minister, I consider this meeting truly an aid for decision-making for those of us who are forced, in the areas of legislation and administration, to confront the problems at issue. You may therefore be assured, ladies and gentlemen, that we will take careful note of the results of this meeting, and of the scientific declarations made here, with the aim of translating them into a sound basis for action in the administrative and legislative sectors. With this in mind, I wish this symposium every success. I hope, too, that each of you will personally experience a meeting of minds that might provide further stimuli not only in this particular area, but also in others. DR. FRITZ PIRKL Minister of Labor and Social Affairs Schellingstraj3e 9 D-8000 Munchen 40, West Germany TI BU 31559 191t. Ik a !b a r

PlttVl:NfIVE NtillIC1NE 13, 563-56411984) about ielves there- you a :YRER t'Ctlolt ring I ustria 0 Proceedings of the International Symposium on Medical Perspectives on Passive Smoking' Opening Address Mr. Federal Minister, Mr. President, ladies, and gentlemen: First of all, I should like to thank the organizers for making this meeting pos- ~.ible and for all,the effort they have-put into the preparations necessary for such a meeting. I would also like to express my sincere thanks to you, Mr. Federal Minister, for your hospitality and your friendly words of welcome. Finally, I thank all the participants for taking the trouble to attend and contribute. When I agreed to the plans of the Bavarian Academy for Occupational and Social Medicine to hold a working meeting on "Passive Smoking at the Work- place" in the spring of 1977 and took over its patronage, 1 felt the meeting fulfilled an urgent need for sound scientific information. It had been preceded by nu- merous complaints presented by nonsmoker initiatives, questions asked by mem- bers of parliament in the House, court decisibns that resulted in newspaper head- lines, and, above all, by the repeated demands of a pugnacious Bavarian doctor and university professor, who called for the prohibition of smoking at the work- place as the "dictate of the hour." Prior to the meeting, a joint notification was issued by the Bavarian Chancellery and the nine Bavarian Ministries, aimed at protecting the passive smoker working for the public authorities. With this noti- fication, the Bavarian State Government wished to set an example with a regu- lation based on consideration for others and tolerance. A notice wa,S signed by the 10 heads of the Bavarian ministerial offices, 7 of them nonsmokers and 3 smokers, that is to say, a mixed body. The 1977 meeting, in brief, concluded that: (a) Nonsmokers may be inconvenienced, and might pos- sibly be hampered in their work, but their health is not endangered; (b) it is the task of a health policy to prevent-as far as this is possible-damage to health. In accordance with the state of our knowledge at that time, as expressed in the discussions, no necessity was therefore seen for legislation: and (c) in the absence of tolerance, it is not possible to achieve a satisfactory state of balance between the opposing interests of smokers and nonsmokers. Response to this event was mixed. In general, reporting by the press was objective, and the discussions and the papers published in a brochure were pos- itively assessed by the public at large as a welcome clarification of a diffiCult problem. However, the opponents of tobacco, and in particular, the proponents of the theory that passive smoking has a detrimental effect on health, expressed dissatisfaction,. sometimes of a polemic nature,. and wme made personal attacks ' April 9-12, 1984. VVienna, Austria. 563 0091-7435/84 33.110 Cupynght il 1994 by .{o:ulemta P'eas. Inc. All nghta o/ repruJuctwn tn .ny turm reaGrved. TI BU 31558 0

SAA FRITZ PIRKL against the organizers and the presenters of scientific papers. In responding to their numerous letters, I pointed out the many legal modes already available to provide protection from annoyance by smokers, both for the private individual and for employees. This, however, had no effect on the legislative calls for action either in Germany or in other countries. Since then, the discussion gradually has spread into other areas where smokers and nonsmokers come together. I must admit, ladies and gentlemen, that I also experience a certain unease. Was that university professor right, after all, when, he listed 40 carcinogenic substances in tobacco smoke and declared that not only the active smoker, but also the passive smoker, was at risk of contracting cancer'' Nor am I sure that ministerial appeals to smokers to show consideration and to nonsmokers to practice tolerance suffice to calm the fears of so many people. These considerations are in themseiv~s proof of the urgent need for a sym- posium such as this. We simply cannot do without further objective clarification of the problem. And thus, along with you, Mr. Federal Minister, I consider this meeting truly an aid for decision-making for those of us who are forced, in the areas of legislation and administration, to confront the problems at issue. You may therefore be assured, ladies and gentlemen, that we will take careful note of the results of this meeting, and of the scientific declarations made here, with the aim of translating them into a sound basis for action in the administrative and legislative sectors. With this in mind, I wish this symposium every success. I hope, too, that each of you will personally experience a meeting of minds that might provide further stimuli not only in this particular area, but also in others. DR. FRITZ PIRKL Minister of Labor and Social Affairs Schellingstraj3e 9 D-8000 Munchen 40, West Germany TI BU 31559 191t. Ik a !b a r

ing to ole to v idual 3ction ly has mu5t s that ces in assive )peals utfice sym- :ation r this :n the . You ote of .h the and ,. ninds ISo ln 'tRKL Jtiurs aJ.te 9 many 0 19ll:vtNT1VE MEUICINE 13, 565-569 (1984) Passive Smoking: A Scientific Consideration' H. VALENTIN lnstitute for Occupational and Social Medicine, University Erlttngen-Nurnberg. Schillerstrasse 25-29. D-8320 Erlangen. Federal Republic of Germany With respect to the scientific discussion of passive smoking, the following points have to be considered: 1. Science as understood in Anglo-American usage is a late result of cultural evolution. Defined operationally, present-day scientific research is humanity's conscious endeavor, in compliance with the rules of logic, to obtain knowledge of the real world, including humans themselves. At first, the results of such efforts are individual bits of intormation (facts, data), which are used inductively to form a hypothesis which, in turn, is either rejected or confirmed by further deduced information or by observation. If confirmed, a hypothesis can then "crystallize" into a theory. Whether a scientific proposition is true or whether-to use a term employed by K. Popper-it has "proved useful," depends exclusively on its compliance with the following two principles: (a) the principle of objectivity: that is, the proposition must be intersubjec- tively understandable, reproducible, and independent of the method employed; and (b) the principle of positive prediction: that is, the conclusions derived from x:ientitic knowledge in accordance with the principles of logic must produce a salf-consistent system, and deduced predictions must be fulfilled. If either of these two principles is not met, the proposition is either scientifically untrue or it cannot be ascribed to the area of science. In the field of medicine, a permanent dialogue is required between experiment and theory on the one hand and the practical and clinical experience on the other. This dialogue must be conducted on the basis of arguments and should be based on logic (i.e., free from emotional content). Only in this way is it possible to check the truth of results, intormation and experience, hypotheses, theories, computer projections or fore- ca!jtb, "gray zone" figures, etc. It is here that7 for numerous occupational-med- ical problems as well as for passive smoking and its attendant risks to health, we can decide whether we are dealing with "science fiction" or with the "real world." 2. In the discussion of cause and effect, our starting point is the term validated scienti/ic knowledge. In this subject matter, a piece of information may be con- ,idcred to have been validated only when: I (a) it has been established by methodical research: that is, plausible and relevant with respect to the proposition; ' Nrewnted at the Sympo.ium "Medical Perspectives on Passive Smoking," April 9-12. 1984, Vicnna, Auatria. 565 0091-7435184 S3.00 Cupynght+D ty1M by Academw Press.lnc. AII ngbts ui repnWucuun in any fam rexerved. TI BU 31560

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 569 ave it a Bh- hial ars, an vittt ac- un- to t ia :ms far idi- git- ,ive I of the on. bly ge, ion ted the liv- ect ;ci- zc- ied at, or lth l/n ts, >n- nd quantification of risks, their acceptance, and analyses of their regulation must s,ull take place. This applies in particular to the problem of any low-risk associ- rtwn. Furthermore. we must also be sure to check the plausibility and relevance uf any suspected association. In conclusion, I am of the opinion that, from the scientific viewpoint, the rclcv;utt points mentioned here should receive more attention than has so far been the case in any discussion of the medical perspectives on passive smoking. The question of whether long-term passive smoking represents a danger to health, ur whether a higher incidence of health damage occurs, is again under consid- eratiun. It is to be hoped that the papers, propositions, and discussions will contribute to a clarific;ttion of the true situation and to the introduction of greater ubjectivity into the debate. The arguments and results presented at this meeting ~,houid be confronted with an open mind and tolerance in our endeavors to, de- termine scientific truths in the real world. In the meantime, as in numerous other areas of life, we can be guided by the maxim that our aim is the preservation of and increasee in morally justified freedom-the freedom not only of the group as a whole, but of the individual. This, briefly, might also be stated as follows: "As much government as necessary, as much freedom as possible." The discussion of problems of the workplace and of environmental questions ~,huuld be based on reason and logic. Medical evaluation pointing up dangers and damage to health should not be overworked. After all, to quote Kant, the ideal uf maturity and freedom is "to make public use of one's reason in all spheres."

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 569 ave it a Bh- hial ars, an vittt ac- un- to t ia :ms far idi- git- ,ive I of the on. bly ge, ion ted the liv- ect ;ci- zc- ied at, or lth l/n ts, >n- nd quantification of risks, their acceptance, and analyses of their regulation must s,ull take place. This applies in particular to the problem of any low-risk associ- rtwn. Furthermore. we must also be sure to check the plausibility and relevance uf any suspected association. In conclusion, I am of the opinion that, from the scientific viewpoint, the rclcv;utt points mentioned here should receive more attention than has so far been the case in any discussion of the medical perspectives on passive smoking. The question of whether long-term passive smoking represents a danger to health, ur whether a higher incidence of health damage occurs, is again under consid- eratiun. It is to be hoped that the papers, propositions, and discussions will contribute to a clarific;ttion of the true situation and to the introduction of greater ubjectivity into the debate. The arguments and results presented at this meeting ~,houid be confronted with an open mind and tolerance in our endeavors to, de- termine scientific truths in the real world. In the meantime, as in numerous other areas of life, we can be guided by the maxim that our aim is the preservation of and increasee in morally justified freedom-the freedom not only of the group as a whole, but of the individual. This, briefly, might also be stated as follows: "As much government as necessary, as much freedom as possible." The discussion of problems of the workplace and of environmental questions ~,huuld be based on reason and logic. Medical evaluation pointing up dangers and damage to health should not be overworked. After all, to quote Kant, the ideal uf maturity and freedom is "to make public use of one's reason in all spheres."

568 H. VALENTIN on the basis of numerous new scientific publications, a number of questions have to be examined. These may be expressed as follows: (a) Does passive smoking present a nuisance for a normal population, is it a danger to health, or is it known to damage health'? In this connection, both high- risk groups within the population (such as people with a hypersensitive bronchial system, children, the aged, or the sick) and specific worksites (nightclubs, bars, restaurants, offices, etc.) must be taken into account. (b) Given the present-day worksite and environmental situation, can it be shown, with any degree of probability, that passive smoking represents aq, jntox- ication due to exposure to exogenous agents? (c) Is the consequence of many years of exposure to passive smoking an increased carcinogenic risk potential to be expected within our population'? In finding answers to these questions, the scientific view is concerned with arguments, evidence, and proofs. Each of these topics has been taken into ac- count, comprehensively and relevantly, in the program of our meeting. 7. In the permanent dialogue between politicians and scientists in many coun- tries of the world, considerable attention has been devoted in recent times to health dangers posed by exogenous agents, particularly at the worksite and in the environment. In this political discussion on health, the social security systems of the various countries should be given more attention than they have so far received. Comparative considerations of the measures instituted by each indi- vidual country without consideration of these points, are not unreservedly legit- imate. This also applies to the problems of possible dangers to health from passive smoking. To illustrate the need for political action or regulation in the tield of health, the following two models might serve for discussion: (a) On the basis of individual observations, suspicions, or hypotheses, the aim is to achieve a political regulation under the general aspect of prevention. Although this would then mean being on the so-called safe side, it might possibly be that, as a result, the freedom of the individual, or of our population at large, would be unduly curtailed, and the effort required might be out of all proportion to the effect achieved. (b) In the case of the second model, we must first prove causality, as indicated above. In the Federal Republic of Germany, for example, in the case of the recognized occupational diseases, it is necessary to demonstrate, either unequiv- ocally or with a high degree of probability, that there actually is a cause and effect relationship. Furthermore, such a relationship must also represent accepted sci- entific opinion. The possibility of such a relationship is not sufficient for its rec- ognition. With respect to both these models, which have so far been arbitrarily applied and interpreted with a certain degree of imprecision, it must be remembered that, despite all political or governmental measures-whether with preventive or causal justification-it is not possible to secure absolute protection for the health of our populations. Life itself, with all its numerous facets, represents a certain risk. The task of clarifying, through dialogue between politicians and scientists, the residual risk that should or must be accepted and borne by the groups con- cerned remains. Discussions of risk assessment encompassing the estimation and TI BU 31563 i ; L rct bcI Th or cr: w ub aht (Cr m: ftti Tt 41

566 H. VALENTIN (b) it has been investigated with the aid of validated procedures: that is, the results can be repeated at any time; and (c) it has found general acceptance among the specialists concerned. Alone, none of the three criteria suffices to permit us to employ the term "validated" in the sense of this definition. 3. In etiological research (research into causes), various methodological ap- proaches can be distinguished. They may be brietly described as follows: (a) the empirical, causal observation of the accumulation of diseases or in- juries to health in response to certain exogenous exposures (challenges); (b) the analytical approach employing the most up-to-date scientific methods; (c) the study of the toxicity of the various inorganic and organic substances in animal experiments; (d) the use of modern short-term tests for investigating toxic, carcinogenic, or mutagenic potential; (e) the sociological approach of questioning individual groups of patients or populations on their well-being, symptomatology, and health disorders; and (f) the use of modern epidemiological-statistical methods aimed at detecting late sequelae of a wide range of exogenous exposures. In the case of exogenous exposures and relevant effects, each of these meth- odological approaches produces results and provides certain information. Untor- tunately, however, these are often introduced into public discussion as absolute truths and are used to postulate dangers, relationships, and causalities. Efforts to make a practical-clinical assessment, or to evaluate the findings, frequently do not take place. Rarely, is any attempt made to verify or rebut the basic sci- entific data obtained with different approaches, in groups under relevant expo- sure, or in groups of relevant patients. Nor have risk assessments to date been carried out on any appreciable scale. Frequently, on the basis of fundamental scientific information, a degree of certainty that cannot be confirmed in the prac- tical or clinical setting is claimed. It is sometimes neither realized nor accepted that all areas of life are associated with a certain risk. As a result, in recent years, a progressive discrepancy has developed between basic scientific data and prac- tical-clinical experience w4th respect to exogenous agents and their effects on human health. 4. Furthermore, the area between health and disease must be reexamined. So far, it has not been unequivocally established, with respect to exogenous expo- sures, which "adverse effects" are actually represented. In this connection, at- tention is drawn to the international symposium held in Luxemburg in December 1980, which was organized by the EC, OSHA, and NIOSH. These symposium proceedings have since been published. Here, we can read about the problems of cause and effect in the human organism with respect to short-term or chronic exposures, sub-chronic, or long-term effects. The question of adverse effects, and its attendant problems, has been with us for decades but is particularly pertinent to the workplace and the environment if moditied in the following way: -With respect to cause and effect in the presence of exogenous exposures, where is the boundary between health and disease?" Elsewhere, we have indicated numerous practical-clinical examples that illus- trate the difficulties of the problems involved and their interpretations. 'TI BU 31561 !

566 H. VALENTIN (b) it has been investigated with the aid of validated procedures: that is, the results can be repeated at any time; and (c) it has found general acceptance among the specialists concerned. Alone, none of the three criteria suffices to permit us to employ the term "validated" in the sense of this definition. 3. In etiological research (research into causes), various methodological ap- proaches can be distinguished. They may be brietly described as follows: (a) the empirical, causal observation of the accumulation of diseases or in- juries to health in response to certain exogenous exposures (challenges); (b) the analytical approach employing the most up-to-date scientific methods; (c) the study of the toxicity of the various inorganic and organic substances in animal experiments; (d) the use of modern short-term tests for investigating toxic, carcinogenic, or mutagenic potential; (e) the sociological approach of questioning individual groups of patients or populations on their well-being, symptomatology, and health disorders; and (f) the use of modern epidemiological-statistical methods aimed at detecting late sequelae of a wide range of exogenous exposures. In the case of exogenous exposures and relevant effects, each of these meth- odological approaches produces results and provides certain information. Untor- tunately, however, these are often introduced into public discussion as absolute truths and are used to postulate dangers, relationships, and causalities. Efforts to make a practical-clinical assessment, or to evaluate the findings, frequently do not take place. Rarely, is any attempt made to verify or rebut the basic sci- entific data obtained with different approaches, in groups under relevant expo- sure, or in groups of relevant patients. Nor have risk assessments to date been carried out on any appreciable scale. Frequently, on the basis of fundamental scientific information, a degree of certainty that cannot be confirmed in the prac- tical or clinical setting is claimed. It is sometimes neither realized nor accepted that all areas of life are associated with a certain risk. As a result, in recent years, a progressive discrepancy has developed between basic scientific data and prac- tical-clinical experience w4th respect to exogenous agents and their effects on human health. 4. Furthermore, the area between health and disease must be reexamined. So far, it has not been unequivocally established, with respect to exogenous expo- sures, which "adverse effects" are actually represented. In this connection, at- tention is drawn to the international symposium held in Luxemburg in December 1980, which was organized by the EC, OSHA, and NIOSH. These symposium proceedings have since been published. Here, we can read about the problems of cause and effect in the human organism with respect to short-term or chronic exposures, sub-chronic, or long-term effects. The question of adverse effects, and its attendant problems, has been with us for decades but is particularly pertinent to the workplace and the environment if moditied in the following way: -With respect to cause and effect in the presence of exogenous exposures, where is the boundary between health and disease?" Elsewhere, we have indicated numerous practical-clinical examples that illus- trate the difficulties of the problems involved and their interpretations. 'TI BU 31561 !

ing to ole to v idual 3ction ly has mu5t s that ces in assive )peals utfice sym- :ation r this :n the . You ote of .h the and ,. ninds ISo ln 'tRKL Jtiurs aJ.te 9 many 0 19ll:vtNT1VE MEUICINE 13, 565-569 (1984) Passive Smoking: A Scientific Consideration' H. VALENTIN lnstitute for Occupational and Social Medicine, University Erlttngen-Nurnberg. Schillerstrasse 25-29. D-8320 Erlangen. Federal Republic of Germany With respect to the scientific discussion of passive smoking, the following points have to be considered: 1. Science as understood in Anglo-American usage is a late result of cultural evolution. Defined operationally, present-day scientific research is humanity's conscious endeavor, in compliance with the rules of logic, to obtain knowledge of the real world, including humans themselves. At first, the results of such efforts are individual bits of intormation (facts, data), which are used inductively to form a hypothesis which, in turn, is either rejected or confirmed by further deduced information or by observation. If confirmed, a hypothesis can then "crystallize" into a theory. Whether a scientific proposition is true or whether-to use a term employed by K. Popper-it has "proved useful," depends exclusively on its compliance with the following two principles: (a) the principle of objectivity: that is, the proposition must be intersubjec- tively understandable, reproducible, and independent of the method employed; and (b) the principle of positive prediction: that is, the conclusions derived from x:ientitic knowledge in accordance with the principles of logic must produce a salf-consistent system, and deduced predictions must be fulfilled. If either of these two principles is not met, the proposition is either scientifically untrue or it cannot be ascribed to the area of science. In the field of medicine, a permanent dialogue is required between experiment and theory on the one hand and the practical and clinical experience on the other. This dialogue must be conducted on the basis of arguments and should be based on logic (i.e., free from emotional content). Only in this way is it possible to check the truth of results, intormation and experience, hypotheses, theories, computer projections or fore- ca!jtb, "gray zone" figures, etc. It is here that7 for numerous occupational-med- ical problems as well as for passive smoking and its attendant risks to health, we can decide whether we are dealing with "science fiction" or with the "real world." 2. In the discussion of cause and effect, our starting point is the term validated scienti/ic knowledge. In this subject matter, a piece of information may be con- ,idcred to have been validated only when: I (a) it has been established by methodical research: that is, plausible and relevant with respect to the proposition; ' Nrewnted at the Sympo.ium "Medical Perspectives on Passive Smoking," April 9-12. 1984, Vicnna, Auatria. 565 0091-7435184 S3.00 Cupynght+D ty1M by Academw Press.lnc. AII ngbts ui repnWucuun in any fam rexerved. TI BU 31560

568 H. VALENTIN on the basis of numerous new scientific publications, a number of questions have to be examined. These may be expressed as follows: (a) Does passive smoking present a nuisance for a normal population, is it a danger to health, or is it known to damage health'? In this connection, both high- risk groups within the population (such as people with a hypersensitive bronchial system, children, the aged, or the sick) and specific worksites (nightclubs, bars, restaurants, offices, etc.) must be taken into account. (b) Given the present-day worksite and environmental situation, can it be shown, with any degree of probability, that passive smoking represents aq, jntox- ication due to exposure to exogenous agents? (c) Is the consequence of many years of exposure to passive smoking an increased carcinogenic risk potential to be expected within our population'? In finding answers to these questions, the scientific view is concerned with arguments, evidence, and proofs. Each of these topics has been taken into ac- count, comprehensively and relevantly, in the program of our meeting. 7. In the permanent dialogue between politicians and scientists in many coun- tries of the world, considerable attention has been devoted in recent times to health dangers posed by exogenous agents, particularly at the worksite and in the environment. In this political discussion on health, the social security systems of the various countries should be given more attention than they have so far received. Comparative considerations of the measures instituted by each indi- vidual country without consideration of these points, are not unreservedly legit- imate. This also applies to the problems of possible dangers to health from passive smoking. To illustrate the need for political action or regulation in the tield of health, the following two models might serve for discussion: (a) On the basis of individual observations, suspicions, or hypotheses, the aim is to achieve a political regulation under the general aspect of prevention. Although this would then mean being on the so-called safe side, it might possibly be that, as a result, the freedom of the individual, or of our population at large, would be unduly curtailed, and the effort required might be out of all proportion to the effect achieved. (b) In the case of the second model, we must first prove causality, as indicated above. In the Federal Republic of Germany, for example, in the case of the recognized occupational diseases, it is necessary to demonstrate, either unequiv- ocally or with a high degree of probability, that there actually is a cause and effect relationship. Furthermore, such a relationship must also represent accepted sci- entific opinion. The possibility of such a relationship is not sufficient for its rec- ognition. With respect to both these models, which have so far been arbitrarily applied and interpreted with a certain degree of imprecision, it must be remembered that, despite all political or governmental measures-whether with preventive or causal justification-it is not possible to secure absolute protection for the health of our populations. Life itself, with all its numerous facets, represents a certain risk. The task of clarifying, through dialogue between politicians and scientists, the residual risk that should or must be accepted and borne by the groups con- cerned remains. Discussions of risk assessment encompassing the estimation and TI BU 31563 i ; L rct bcI Th or cr: w ub aht (Cr m: ftti Tt 41

574 TRIEBIG AND ZOBER Nitrogen Oxides Tables 4 and 5 provide an overview of the NO and nitrogen dioxide (NO,) concentrations in indoor areas under realistic conditions, as found in the litera- ture. In addition, mean concentrations of the outdoor air are also indicated. Al- though NO itself is not an irritant substance, it should also be included in our considerations, since, in the presence of oxygen, it can react to form NO,. The NO values measured in offices are of the same order of magnitude as the outdoor concentrations and are, overall, below the MIK value. Under realistic smoking conditions, NO, concentrations reach the MIK value, and may even exceed it at higher levels of tobacco smoke contamination (CO above 10 ppm). According to Weber and Fischer (49), however, the NO, values are unrelated to CO and NO; thus other sources of NO, must also be considered, [for example, gas or kerosene-tired heating systems (14, 21, 46)]. Aldehydes (Acrolein, Formaldehyde) In Table 6, the indoor air concentrations of acrolein under various realistic conditions are listed. A striking feature is the relatively large variation in the levels measured. According to Fischer et a!. (11), these levels were usually at the lower limit of detection by their analytical method. Formaldehyde concentrations are, as far as we could establish, available from only a single experimental investigation (see Table 6). Under conditions of high pollution with sidestream smoke in a climatic chamber, figures of up to 640 ppb have been measured for formaldehyde (for CO, up to 43 ppm). In connection with the evaluation of formaldehyde in indoor air, in addition to the MIK value of 20 ppb, the upper limit of 100 ppb recommended by the Bun- desgesundheitsamt Berlin, should also be mentioned. In the case of formaldt:- hyde, smoking-unrelated sources are also of importance for indoor air quality. Table 7 lists the results of a number of more recent studies on this subject. For example. urea formaldehyde foam insulation or certain wood materials can emit formal- dehyde and give rise to mean air concentrations of 100-400 ppb. In summary, at the present time, the lack of appropriate investigations does Ref. Level tppb) Location Vcntil.uiun Range Mean 2 Restaurants Natural 14-218 80 and 1'_U Tavern N atur.tl 66-414 195 Cafetaria Air wnUitiuning :-38 5 and 9 Outdoors 4- [15 44 Working rooms .3ir conUitiuning ur window venulutiun tSU-.i5U 84 Outdoors 63 Nutr. 41IK vulue (24 hr): alN) ppb. TABLE 4 NITROGEN MONOXIDE LEVELS MEASUKI:D UNDER REALISTIC SMOKING CUNDITtONS AND IN THE GI:NtKAL ENVIRUNMkNT TI BU 31569 I t: . i .

572 TRIEBIG AND ZOBER TABLE I SURVEY OF TOBACCO SMOKE COMPONENTS RELEVANT ro HUMAN HEALTH Gas phase Particulate phase Carbon monoxide Respirabie particulates Aldehydes (acrolein. formaldehyde) Nicotine Nitrogen oxides Polycyclic aromatic hydrocarbons Ammottia Nitrosamines Phenol Metals (cadmium, nickel) Hydrogen cyanide Sulfur dioxide As a basis for our further evaluation we have made use of both the maximum permissible concentration (MAK)3 and the maximum emission concentration (MIK)' values. For a number of substances, the non-tobacco-smoke-related con- centrations in the air are also indicated. The MAK value is defined as the max- imum permissible concentration of a chemical compound present in the air within a working area that, according to current knowledge, generally does not impair the health of the employee nor cause undue annoyance. Under these conditions, exposure can be repeated and of long duration ranging from an average of 8 hr per day up to 42 hr weekly (9). In accordance with the Verein Deutscher Ingenieure (VDI) regulations, the MIK value represents the concentration of an air-polluting substance, in atmo- spheric layers close to the grotind, that may be considered harmless to humans, animals, or plants-even when effective over a given period ('h- hr, 24 hr, I yeur) and present at a given incidence. The MAK and MIK values are recommendations only, with no legal authority. In addition, these maximum permissible values are established only for air pol- lution by a single substance. As defined, they cannot be applied to mixtures of substances, as is the case with tobacco smoke. A comparison of the single-substance MAK and MIK values with maximum permissible values, however, appears to us to be both permissible and plausible, provided that the following points are taken into account: (a) The purpose of these "maximum permissible values" is to secure the pro- tection of the health of the population, or of defined groups within the population. For this reason, the guideline concentrations were established with consideration of safety margins. (b) On the basis of this concept, even if excessive values occur, they are not readily correlated with any established or assumed injury to health in the indi- vidual case. Further criteria of causality must be taken into account before any conclusions can be drawn. (c) If other parameters or evaluation criteria are used, it is, of course, possible to draw other conclusions. Carbon Monoxide In Table 3, the CO levels as obtained from studies of passive smoking are.listed. s MAK = Maximak Arbetsplatzkonzentration. 4M1K = Maxintale lrrmissioeskoezentrAtion. TI BU 31567

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 567 the ; Iixlay, the following classification of exogenously caused injuries to health may ~ be made: (a) minimal lesions with no appreciable impairment to the state of health; :erm 4 (b) allergic reactions with temporary nuisance value or impairment; i (c) intoxications with reversible or irreversible damage to health; and ( ap- ~ (a) malignant diseases following exposure to agents possessing a carcinogenic potential in man. r in- 5. Over the last few decades, considerable advances have been made in diag- nuaib and therapy in the various sections and branches of medicine. In addition, ods; the diverse causal factors have increasingly received more attention. This remark rices applies not only to disposition and constitution, destiny and age, lifestyle and i hcalth-impairing behavior, etc., but also to exogenous noxious factors at the :nic, j workplace and in the environment. As indicated above, in the research into l causes, various methodological approaches can be differentiated. Particular im- .s or ~ purtance attaches to the use of modern epidemiological-statistical methods for f• the dctection of late sequelae in a range of exogenous exposures. In the meantime, aing ~ "criteria catalogs" for the assessment of causal relationships have been worked i uut with the aim of achieving an objective and professional interpretation and of eth- ~ rvuiding the drawing of erroneous conclusions. In this connection in particular, ifor- i the following points should be considered: ,lute ~ (a) the evaluation of every individual epidemiological study in accordance orts i with modern principles; ntly j (b) consistency of the association in various investigations; sci- (c) both necessary and adequate specificity of the association; <po- . (d) intensity of the association in the sense of a high relative risk; )een (e) temporal coherence of cause and et'fect; !ntal (t) epidemiological plausibility through increased incidence; rac- i (g) biological plausibility through agreement of hypothesis and the state of )ted ' ` wicntitic knowledge; ars, ! (h) demonstration of a dose-response or a dose-incidence relationship; rac- (i) statistical predictability and signiticance for the identification of risk fac- un turs; i ( j) reduction of incidence of damage to health through the use of preventive So d measures; po- (k) agreement with knowledge and experience in human medicine; and at- ~ (I) demonstration of a patholugical effect after appropriate exposure of ex- i ber I lx;rimental animals, or demonstration of a mutagenic potential in so-called short- ium j term tests. :ms With my collaborators, Kentner and Zober. I have prepared a short explanation )nic iur each factor, which, cannot, however, be-discussed here. The step from the and ,tatistical association to the assumption of a causal relationship is. in view of the .znt multicausality of various health impairments or chronic diseases, not only a sci- ith cntitic-theoretical, but also a practical-clinical, problem. For contirmation, a the number of the points listed here are required. Under these conditions, this criteria cutalug for the assessment and evaluation of the numerous epidemiological studies 'us- published to date has proved of value to our working group at Erlangen. 6. In the scientific discussion ut• passive smoking and its dangers to health, and TI BU 31562

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 571 ery, dour air ,tandard studies de. The nmental an mun- e. sulfur _ hydro- (eraure he max- . u 19" z! following of several olysis); (b) :e; and (c) .hat smoke :am smoke wffs (23). z literature ene, Bibli- Dukumen- il reports). years: Ry- riftenreihe HO Docu- Report of lyg2 (39); :ealing with `1- 1=. 1984, d uid Sucial .erstra,be ?5l . I thc topic, we cannot make any claims as to completeness. Nevertheless, we have June our best to ensure that this review takes into account all the major facts presently known. It should first be noted that the relevant results refer almost caclusivcly to investigations into cigarette smoking, and not to other forms of tuh:,t;cu consumption. This "restriction" is, however, realistic, since, statisti- .~,,lly. cigarette smoking represents the major form of tobacco consumption. Fur- thcrmure, the demonstrable chemical differences between cigarette smoke on the one hand and cigar smoke, for example, on the other have little relevance to our rtwin interest: sidestream smoke. Finally, we mainly considered results of studies obtained under "natural" and "realistic" conditions or results of tield studies. RESULTS AND DISCUSSION '1'ubaccu smoke consists of a large number of different substances, which are cla,aitied into a gas phase and a particulate phase. Two to several thousand Jitlcrcnt substances have been reported. To date, several hundred substances hase been identified, and our consideration of them must be differentiated with rc,lxct both to the concentrations at which they occur in the air and to the rc,ulting potential health effects. In this connection, the substances may be clas- %iticd into irritants, toxins, and carcinogens. A critical analysis of the literature shows that the substances, or groups of substances, listed in Table 1 are of importance. With respect to their biological ctl'ccts in humans, carbon monoxide (CO), nicotine, and hydrogen cyanide (HCN) bclung to the resurptive toxins. Respirable particles, certain aldehydes, phenol, rntnwnium, and sulfur dioxide (SO_) are mainly irritant substances. Cadmium tCd) and nickel (Ni) compounds are also resorptive, or local, toxins. More than 40 tobacco smoke constituents have been referred to as possible human carcin- ugcna. On the basis of our present knowledge, the polycyclic aromatic hydro- carbons-of which a major representative is benzo[uJpyrene-and certain vol- alile nitrosamines must be included among the tobacco smoke constituents with potential carcinogenicity in humans. In addition to the different biological effects of the tobacco smoke constituents, the qualitative and quantitative differences between MSS, and SSS have to be considered [see Table 3'and Ref. (34)]. Since, as is well known, up to 85% of the indoor air pollution by tobacco smoke by-products is caused by SSS, the signif- icance of its constituents in connection with environmental and health-related problems is under current discussion. However, one must consider that these cuncentration ratios have been calculated from the results of experiments. For in,tance, the quantitative results of SSS analysis in the "Neurath chamber" may be influenced by the specific climatic conditions within the chamber, and thus provide values that are high in comparison with those found in a more normal cnvirunment (24). This applies in particular to the nitrosamines, for which the largcst ratios have been calculated. Apart from these methodological intluences, the abaulute concentrations of the tobacco smoke constituents are, for obvious rca.,uns, indispensable for any toxicological evaluation. Thus, it may be con- cluded that the ratios have only an orientating character and are not suitable to serve as a basis for well-founded conclusions. „ TI BU 31566

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 567 the ; Iixlay, the following classification of exogenously caused injuries to health may ~ be made: (a) minimal lesions with no appreciable impairment to the state of health; :erm 4 (b) allergic reactions with temporary nuisance value or impairment; i (c) intoxications with reversible or irreversible damage to health; and ( ap- ~ (a) malignant diseases following exposure to agents possessing a carcinogenic potential in man. r in- 5. Over the last few decades, considerable advances have been made in diag- nuaib and therapy in the various sections and branches of medicine. In addition, ods; the diverse causal factors have increasingly received more attention. This remark rices applies not only to disposition and constitution, destiny and age, lifestyle and i hcalth-impairing behavior, etc., but also to exogenous noxious factors at the :nic, j workplace and in the environment. As indicated above, in the research into l causes, various methodological approaches can be differentiated. Particular im- .s or ~ purtance attaches to the use of modern epidemiological-statistical methods for f• the dctection of late sequelae in a range of exogenous exposures. In the meantime, aing ~ "criteria catalogs" for the assessment of causal relationships have been worked i uut with the aim of achieving an objective and professional interpretation and of eth- ~ rvuiding the drawing of erroneous conclusions. In this connection in particular, ifor- i the following points should be considered: ,lute ~ (a) the evaluation of every individual epidemiological study in accordance orts i with modern principles; ntly j (b) consistency of the association in various investigations; sci- (c) both necessary and adequate specificity of the association; <po- . (d) intensity of the association in the sense of a high relative risk; )een (e) temporal coherence of cause and et'fect; !ntal (t) epidemiological plausibility through increased incidence; rac- i (g) biological plausibility through agreement of hypothesis and the state of )ted ' ` wicntitic knowledge; ars, ! (h) demonstration of a dose-response or a dose-incidence relationship; rac- (i) statistical predictability and signiticance for the identification of risk fac- un turs; i ( j) reduction of incidence of damage to health through the use of preventive So d measures; po- (k) agreement with knowledge and experience in human medicine; and at- ~ (I) demonstration of a patholugical effect after appropriate exposure of ex- i ber I lx;rimental animals, or demonstration of a mutagenic potential in so-called short- ium j term tests. :ms With my collaborators, Kentner and Zober. I have prepared a short explanation )nic iur each factor, which, cannot, however, be-discussed here. The step from the and ,tatistical association to the assumption of a causal relationship is. in view of the .znt multicausality of various health impairments or chronic diseases, not only a sci- ith cntitic-theoretical, but also a practical-clinical, problem. For contirmation, a the number of the points listed here are required. Under these conditions, this criteria cutalug for the assessment and evaluation of the numerous epidemiological studies 'us- published to date has proved of value to our working group at Erlangen. 6. In the scientific discussion ut• passive smoking and its dangers to health, and TI BU 31562

Jns laximum ntration .ted con- he max- ir within )t impair nditions, : of 8 hr ons, the in atmo- humans, ., I year) uthority. . air pol- ;tures of a laximum lausible, the pro- wlation. deration are not :he indi- :ore any possible 'e listed. SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 573 TABLE 2 U15'tR1tlUT1ON OF LMPORTANT TOBACCO SMOKE COMPONENTS IN MAINSTREAM (MS) aND SIDESTItEAM (SS) SMOKE Smoke component Ratio SS/MS N•Nitrosodimethylamine 1 ! -437 Ammonia 73-IS0 Formaldehyde 50 Nickel 0.2-31 Acrolein 12 Cadmium 4-7 Carbon monoxide 2-5 Nicotine 3 aenzo(ulpyrene 3 I Hydrogen cyanide 0.05-0.4 Nutr. Ad•,pted from Ref. (24). Ai can be seen, air contamination by tobacco smoke-related CO lies mainly bctween I and 10 ppm. In the presence of only moderate smoke pollution, for cxample, in offices, CO concentrations under 5 ppm generally occur, thus not excctaling the MIK value. In the case of more marked tobacco smoke pollution, values in excess of 10 ppm can be achieved. Even higher concentrations, namely 2t1 to 50 ppm, are found in bars and taverns, for example. Here, however, other wurct:s of CO may also be involved-for example, gas-fired heaters emit con- uderabie quantities of CO (2). More recent studies of the non-tobacco-smoke- rclated CO burden have revealed CO values of 3 to 3.5 ppm in indoor areas tuttices, restaurants, stores) (28). TABLE 3 CARBON MONOXIDE LEVELS MEASURED UNDER REALISTIC SMOKING CONDITIONS AND IN THE GENERAL ENVIRONMENT Level (ppm) Rcf. Location Ventilation Range 4fean 1181 Offices Natural 2.5-10 - 142) 49 Offices Natural 2.7 • 171 2 Taverns 6 and 2 Changes/hr 10- 12 11.5 and 17 and 3-22 151 10 Ottices - 1.5-4.5 2.5 15 Restaurants - 1.0-9.5 . 4.0 14 Bars and taverns - 3.0-29 13.0 111) 2 Restaurants Natural 1.4-9.9 2.6 and 5.1 luvern Natural 2.4-9.4 4.8 Cateteria Air conditioning 0.3-1.7 0.5 and l.: Outdoors 0.4-4.8 (491 44 Working rooms Air conditioning or Maximal 4.9 2.8 Outdoors window ventilation 2.0 .Yute. MAK value (l98i): 311 ppm; M1K value (24 hr): 8.7 ppm. TI BU 31568 .

574 TRIEBIG AND ZOBER Nitrogen Oxides Tables 4 and 5 provide an overview of the NO and nitrogen dioxide (NO,) concentrations in indoor areas under realistic conditions, as found in the litera- ture. In addition, mean concentrations of the outdoor air are also indicated. Al- though NO itself is not an irritant substance, it should also be included in our considerations, since, in the presence of oxygen, it can react to form NO,. The NO values measured in offices are of the same order of magnitude as the outdoor concentrations and are, overall, below the MIK value. Under realistic smoking conditions, NO, concentrations reach the MIK value, and may even exceed it at higher levels of tobacco smoke contamination (CO above 10 ppm). According to Weber and Fischer (49), however, the NO, values are unrelated to CO and NO; thus other sources of NO, must also be considered, [for example, gas or kerosene-tired heating systems (14, 21, 46)]. Aldehydes (Acrolein, Formaldehyde) In Table 6, the indoor air concentrations of acrolein under various realistic conditions are listed. A striking feature is the relatively large variation in the levels measured. According to Fischer et a!. (11), these levels were usually at the lower limit of detection by their analytical method. Formaldehyde concentrations are, as far as we could establish, available from only a single experimental investigation (see Table 6). Under conditions of high pollution with sidestream smoke in a climatic chamber, figures of up to 640 ppb have been measured for formaldehyde (for CO, up to 43 ppm). In connection with the evaluation of formaldehyde in indoor air, in addition to the MIK value of 20 ppb, the upper limit of 100 ppb recommended by the Bun- desgesundheitsamt Berlin, should also be mentioned. In the case of formaldt:- hyde, smoking-unrelated sources are also of importance for indoor air quality. Table 7 lists the results of a number of more recent studies on this subject. For example. urea formaldehyde foam insulation or certain wood materials can emit formal- dehyde and give rise to mean air concentrations of 100-400 ppb. In summary, at the present time, the lack of appropriate investigations does Ref. Level tppb) Location Vcntil.uiun Range Mean 2 Restaurants Natural 14-218 80 and 1'_U Tavern N atur.tl 66-414 195 Cafetaria Air wnUitiuning :-38 5 and 9 Outdoors 4- [15 44 Working rooms .3ir conUitiuning ur window venulutiun tSU-.i5U 84 Outdoors 63 Nutr. 41IK vulue (24 hr): alN) ppb. TABLE 4 NITROGEN MONOXIDE LEVELS MEASUKI:D UNDER REALISTIC SMOKING CUNDITtONS AND IN THE GI:NtKAL ENVIRUNMkNT TI BU 31569 I t: . i .

PREVENTIVE MEDICINE 13, 570-581 (1984) Indoor Air Pollution by Smoke Constituents-A Survey' GERHARD TRIEBIG2 AND tYIAX ANDREAS ZOBER Institute of Occupational unJ Social Merlicine, University of Erlungen-Nurnbers, Erlwegen, Federal Republic uf Germuny A review ut' the literature on the present state of knowledge about the topic "indoor air pollution by smoke constituents" has been made. While consideration is given to standard bibliographic works and more recent reviews, an evaluation ut ihe results of tiekd studies and experimental investigations carried out under realistic conditions is also made. Thn selection of the tobacco smoke constituents discussed here was bused on environmental and toxicological aspects. The following substances are considered in detail: carbon mon- oxide, nitrogen oxides, formaldehyde, acrolein, ammonia, phenol, hydrogen cyanide. sulfur dioxide, cadmium, nickel, nicotine, some volatile nitrosamines, polycyclic aromatic hydro- carbons, and respirable particulates. The smoke constituent concentrations in the literature are presented in tables. The results are evaluated and discussed with respect to the max- imum allowable concentrations presently valid in the Federal Republic of Germany. o ivsw AcaJamtc Press. Inc. INTRODUCTION Regarding tht: topic of indoor air pollution by smoke constituents, the following physicochemical facts are relevant: (a) tobacco smoke is a product of several complex chemical processes (such as distillation, sublimation, and pyrolysis); (b) to date, thousands of substances have been detected in tobacco smoke; and (c) mainstream and sidestream smoke usually differ in their composition. According to the Coresta definition (6), mainstream smoke (MSS) is that smoke issuing from the end of a cigarette when a smoker draws on it. Sidestream smoke (SSS) is that smoke given off by the cigarette in the pauses between puffs (23). The basis for this review was an evaluation of information given in the literature published between 1970 and 1983 (CIS card index, Abstracts of Hygiene, Bibli- ography on Smoking and Health, Current Contents-Life Sciences, Dokumen- tation Soziafttledizin, Public Health Service and Occupational Medical reports). We also included reviews of this subject matter published in recent years: Ry- lander, 1974 and 1984 (32, 33); Klosterkotter and Gono, 1976 (23); Schriftenreihe Bayerischen Akademie f. Arbeits- und Sozialmedizin, 1977 (36); WHO Docu- ments, 1977 (51, 52); Badre et al., 1978 (1); Valentin et al., 1978 (44): Report of the U.S. Surgeon General, 1979 (40); Repace, 1982 (31); Sterling et al., 1982 (39); Weiss et al., 1983 (50); and Rahmede, 1983 (29). Because of the extensive amount of information in the literatt2re dealing with ' Presented at thC Symposium "Medical Perspectives on Passive Smuking," April 9-12. 1984, Vienna, Austria. = To whom requests for reprints should be addressed at the (nstitute of Occupational and Social Medicine (Director: Prof. Di.med. H. Valentin). University Eriangen-Nurnbcrg, Schillerstrasse 25/ 29. tl520-Eriangen, Federal Republic of Germany. 570 0091-7435/84 33.00 Copynghl L PpN by Ac+demk Preaa. Inc. AU nylua ui repruJUawn in any Curm reaerveJ. TI BU 31565 i t I t

576 TRIEStG AND ZOBER TABLE 7 FORMALDEHYDE LEVELS MEASURED UNDER REALISTIC CONDITIONS IN DOMESTtC ENVIRONMENTS Levei (ppb) Ref. Location Source Range Mean (21) Indoor - 60-ISO (8) Indoor office Resins 10-1090 100 buildings 380-490 440 (27) DwellinQs Resins 10-930 290 offices. schools 5-770 240 (34) 6 Houses Urea foam insulation 2-230 (43) Offices - 33-47 41 Outside air 20-21 (25) Working room Wood material Maximal 240 are only minimally elevated. In the mean, the MIK value is reached, or slightly exceeded. With respect to HCN, ammonia, Cd, and Ni, we have so far found no information obtained under realistic conditions. Although data on the levels of Cd and Ni in smoke or cigdrettes are available, no air concentrations have been reported in studies on passive smoking (24, 35). The mean total smoke concen- trations are, according to Szadkowski et al. (41), 0.19 µg Cd/cigarette and 0.23 µg Ni/cigarette. Other authors have also reported similar levels (35). Respirable Particulates In Table 9, the particulate concentrations resulting from tobacco smoke as measured in field studies are shown. Particulates are detined as those substances that are retained by a Cambridge filter in excess of 50% under standardized smoking conditions. Cigarette smoke particulates are both respirable and small enough to-enter the alveolar system. The measured mean concentrations are pre- dominantly in the range of 0.2 mg/m3 and are thus of the order of magnitude of the MIK value for a 24-hr exposure. TABLE 8 PHENOL AND SULFUR DIOxIDE LEVEL! MEASURED UNDER REALISTIC SMOKING CONDITIONS AND IN THE GENERAL ENVIRONMENT Level (ppb) Ref. Location Ventilation Range Mean (22) 4 Cafes -- Phenol: 0.07-0.012 (11) 2 Restaurants Natural SO,: 5-32 13 and 20 Tavern N atural SO,: 5-18 30 Cafeteria Air conditioning (lI chanbesrhr) SO,: 15- l03 7,,nd 15 Outside _ SO,c 6-12 Note. Phenol MAK value (1984): 5 ppm. Sulfur dioxide MAK value (1984): 2,000 ppb; MIK value (24 hr): 14 ppb. TI BU 31571 4 4 ., I

O,) : ra- Al- our I'he oor : ue, c0 ues ed. ,om tigh ppb n to l un- ltle- .ible ple. nal- :oeS? '_U SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 575 TABLE 5 N/rRtX',EN DlOX1DE LEVELS MEASURED UNUER REALISTIC SMOKING CONDITLONS ANU lN THE GENERAL ENVIRONMtNT Level(ppb) Ref. Location Ventilation Range Mean 111.48) 2 Restaurunts Natural ?4-105 63 and 76 Tavern Natural 1-61 21 Cafeteria Air conditioning 15-1U3 27'and 58 Outdoors 34-63 1491 44 Working rooms Outdoors Air conditioning or window ventilation 10-200 60 -Nuta. `tAK value (1984): 5.UUU ppb: MIK value l34 hr): 55 ppb. nut lxrmit any definitive conclusions to be drawn as to the formaldehyde levels %;:,uhctl by tobacco smoke under realistic smoking conditions. .Ilisc•rlluitruns Under the term "miscellaneous," the smoke constituents ammonium, phenol, hytlrugen cyanide (HCN). sulfur dioxide (SO,), and the metals nickel (Ni) and wdmium (Cd) are considered. In Table 8, the concentrations of phenol and sulfur dtuxide measured under realistic conditions are listed. According to these figures. phenol was found only in the "ppb range." The levels of SO, in res- I.,urants and a tavern were, on ave-ragz, between 10 and 20 ppb. Thus, in com- p.trisun with the outdoor levels of 6-1? ppb, the indoor concentrations of SO, TABLE 6 a ACRULI:IN AND FOKMALDEHYUE LEVELS MEASURED. RESPECTIVELY, UNDER REALISTIC AND EXPEKIMENTAL CONDITIONS LCvel(ppb) i Rcf. i Location Range Mean i tI) Acrolein" 5 Cafes 3U-100 Train =U-12U Car with 3 ~,mukers (open) 30 Car with 3 smokers (closedl 300 111 i 2 Restaurants 3-13 7and$ Tavern 5-18 t0 I iaxt t.t7t Cafeteria Formalclehyde' Climatic chamber 10. 1 changerhrl I-IU 30-640 5 and 6 (CO: 1-43 ppm) • NAK value t l9l41: 100 ppb: MIKU value: 5 ppb. ' MAK value (1984): 1000 ppb; M1Ku value: 20 ppb. TI BU 31570

576 TRIEStG AND ZOBER TABLE 7 FORMALDEHYDE LEVELS MEASURED UNDER REALISTIC CONDITIONS IN DOMESTtC ENVIRONMENTS Levei (ppb) Ref. Location Source Range Mean (21) Indoor - 60-ISO (8) Indoor office Resins 10-1090 100 buildings 380-490 440 (27) DwellinQs Resins 10-930 290 offices. schools 5-770 240 (34) 6 Houses Urea foam insulation 2-230 (43) Offices - 33-47 41 Outside air 20-21 (25) Working room Wood material Maximal 240 are only minimally elevated. In the mean, the MIK value is reached, or slightly exceeded. With respect to HCN, ammonia, Cd, and Ni, we have so far found no information obtained under realistic conditions. Although data on the levels of Cd and Ni in smoke or cigdrettes are available, no air concentrations have been reported in studies on passive smoking (24, 35). The mean total smoke concen- trations are, according to Szadkowski et al. (41), 0.19 µg Cd/cigarette and 0.23 µg Ni/cigarette. Other authors have also reported similar levels (35). Respirable Particulates In Table 9, the particulate concentrations resulting from tobacco smoke as measured in field studies are shown. Particulates are detined as those substances that are retained by a Cambridge filter in excess of 50% under standardized smoking conditions. Cigarette smoke particulates are both respirable and small enough to-enter the alveolar system. The measured mean concentrations are pre- dominantly in the range of 0.2 mg/m3 and are thus of the order of magnitude of the MIK value for a 24-hr exposure. TABLE 8 PHENOL AND SULFUR DIOxIDE LEVEL! MEASURED UNDER REALISTIC SMOKING CONDITIONS AND IN THE GENERAL ENVIRONMENT Level (ppb) Ref. Location Ventilation Range Mean (22) 4 Cafes -- Phenol: 0.07-0.012 (11) 2 Restaurants Natural SO,: 5-32 13 and 20 Tavern N atural SO,: 5-18 30 Cafeteria Air conditioning (lI chanbesrhr) SO,: 15- l03 7,,nd 15 Outside _ SO,c 6-12 Note. Phenol MAK value (1984): 5 ppm. Sulfur dioxide MAK value (1984): 2,000 ppb; MIK value (24 hr): 14 ppb. TI BU 31571 4 4 ., I

572 TRIEBIG AND ZOBER TABLE I SURVEY OF TOBACCO SMOKE COMPONENTS RELEVANT ro HUMAN HEALTH Gas phase Particulate phase Carbon monoxide Respirabie particulates Aldehydes (acrolein. formaldehyde) Nicotine Nitrogen oxides Polycyclic aromatic hydrocarbons Ammottia Nitrosamines Phenol Metals (cadmium, nickel) Hydrogen cyanide Sulfur dioxide As a basis for our further evaluation we have made use of both the maximum permissible concentration (MAK)3 and the maximum emission concentration (MIK)' values. For a number of substances, the non-tobacco-smoke-related con- centrations in the air are also indicated. The MAK value is defined as the max- imum permissible concentration of a chemical compound present in the air within a working area that, according to current knowledge, generally does not impair the health of the employee nor cause undue annoyance. Under these conditions, exposure can be repeated and of long duration ranging from an average of 8 hr per day up to 42 hr weekly (9). In accordance with the Verein Deutscher Ingenieure (VDI) regulations, the MIK value represents the concentration of an air-polluting substance, in atmo- spheric layers close to the grotind, that may be considered harmless to humans, animals, or plants-even when effective over a given period ('h- hr, 24 hr, I yeur) and present at a given incidence. The MAK and MIK values are recommendations only, with no legal authority. In addition, these maximum permissible values are established only for air pol- lution by a single substance. As defined, they cannot be applied to mixtures of substances, as is the case with tobacco smoke. A comparison of the single-substance MAK and MIK values with maximum permissible values, however, appears to us to be both permissible and plausible, provided that the following points are taken into account: (a) The purpose of these "maximum permissible values" is to secure the pro- tection of the health of the population, or of defined groups within the population. For this reason, the guideline concentrations were established with consideration of safety margins. (b) On the basis of this concept, even if excessive values occur, they are not readily correlated with any established or assumed injury to health in the indi- vidual case. Further criteria of causality must be taken into account before any conclusions can be drawn. (c) If other parameters or evaluation criteria are used, it is, of course, possible to draw other conclusions. Carbon Monoxide In Table 3, the CO levels as obtained from studies of passive smoking are.listed. s MAK = Maximak Arbetsplatzkonzentration. 4M1K = Maxintale lrrmissioeskoezentrAtion. TI BU 31567

SYMPOSIUM: MEDICAL PEHSPECCIVES ON PASSIVE SMOKING 577 ; as ices ized nail A p re- e of 3 cN .5 TAHLE9 COMA J•V1RAf1UNS OF RLiSPIRABLE PAKfICULATIiS Mti.1SUHED UNDER RSALIST/C SMOKING CONDITIONS ANU IN THF GLNIi1tA1. ENVIRONMENT Level (mg/ms) Nef. Location Ventilation Range Mean 12.1) a Cates - 1.2 Outdoors 0.6 t:u1' Restaurant - 0.05-0.45 0.2 Bar - 0.17-0.64 0.4 Hus - 0.14-0.37 0.2 (11 Tavern 6 Changes/hr 0.23-0,35 13U1 4 Eteataurants -- 0.08-0.11 14 Public places 0.09-0.70 t19/ 44 Wurking rooms Air conditioning or window ventilation 0.01-1.13 0.17 Nuse. MAK value (19tl4): 6 mgJm}; MIK value (24 hr): 0.2 mglm3. ,Vicuiinr Table 10 shows the concentrations of nicotine in air measured under realistic cunditions in differing tobacco smoke atmospheres. With the exception of ex- Ireme concentrations of smoke-e.g., three smokers in a car, carbon monoxide in excess of 50 ppm-the nicotine levels under natural conditions are on the order uf up to 50 µg/m3. At present, no MIK value is available for nicotine. The MAK value was fixed to protect tobacco industry employees from developing an rtnblyopia (19). Nirrusurnines The concentrations of the volatile nitrosamines N-nitrosodimethylamine ( YDMA) and N-nitrosodiethylamine (NDEA) quoted in reports on passive imuking studies are shown in Table 11. According to these tigures, nitrosamine `uncentrations in air vary from less than 10 to a maximum of 240 nglm3. Control TABLE 10 NICOTINE t.EVELS MEASURED uNDER REALISTIC SMOKlNG CONDITIONS Level (µg/ma) Itef. Location Ventilation Range Mean t:ul Restaurants - 5.2 Bar - 10.3 Bus - 6.3 111 6 Ca/es - 25-52 Train 36-50 Car with 3 smokers (open) 65 Car with 3 smokers (closed) 1010 1;y1 44 Wcxking rooms Maaitnai: 13.8 1.1 Nure. MAK value 119!l;1: 500 µglm'.

SYMPOSIUM: MEDICAL PEHSPECCIVES ON PASSIVE SMOKING 577 ; as ices ized nail A p re- e of 3 cN .5 TAHLE9 COMA J•V1RAf1UNS OF RLiSPIRABLE PAKfICULATIiS Mti.1SUHED UNDER RSALIST/C SMOKING CONDITIONS ANU IN THF GLNIi1tA1. ENVIRONMENT Level (mg/ms) Nef. Location Ventilation Range Mean 12.1) a Cates - 1.2 Outdoors 0.6 t:u1' Restaurant - 0.05-0.45 0.2 Bar - 0.17-0.64 0.4 Hus - 0.14-0.37 0.2 (11 Tavern 6 Changes/hr 0.23-0,35 13U1 4 Eteataurants -- 0.08-0.11 14 Public places 0.09-0.70 t19/ 44 Wurking rooms Air conditioning or window ventilation 0.01-1.13 0.17 Nuse. MAK value (19tl4): 6 mgJm}; MIK value (24 hr): 0.2 mglm3. ,Vicuiinr Table 10 shows the concentrations of nicotine in air measured under realistic cunditions in differing tobacco smoke atmospheres. With the exception of ex- Ireme concentrations of smoke-e.g., three smokers in a car, carbon monoxide in excess of 50 ppm-the nicotine levels under natural conditions are on the order uf up to 50 µg/m3. At present, no MIK value is available for nicotine. The MAK value was fixed to protect tobacco industry employees from developing an rtnblyopia (19). Nirrusurnines The concentrations of the volatile nitrosamines N-nitrosodimethylamine ( YDMA) and N-nitrosodiethylamine (NDEA) quoted in reports on passive imuking studies are shown in Table 11. According to these tigures, nitrosamine `uncentrations in air vary from less than 10 to a maximum of 240 nglm3. Control TABLE 10 NICOTINE t.EVELS MEASURED uNDER REALISTIC SMOKlNG CONDITIONS Level (µg/ma) Itef. Location Ventilation Range Mean t:ul Restaurants - 5.2 Bar - 10.3 Bus - 6.3 111 6 Ca/es - 25-52 Train 36-50 Car with 3 smokers (open) 65 Car with 3 smokers (closed) 1010 1;y1 44 Wcxking rooms Maaitnai: 13.8 1.1 Nure. MAK value 119!l;1: 500 µglm'.

PREVENTIVE MEDICINE 13, 570-581 (1984) Indoor Air Pollution by Smoke Constituents-A Survey' GERHARD TRIEBIG2 AND tYIAX ANDREAS ZOBER Institute of Occupational unJ Social Merlicine, University of Erlungen-Nurnbers, Erlwegen, Federal Republic uf Germuny A review ut' the literature on the present state of knowledge about the topic "indoor air pollution by smoke constituents" has been made. While consideration is given to standard bibliographic works and more recent reviews, an evaluation ut ihe results of tiekd studies and experimental investigations carried out under realistic conditions is also made. Thn selection of the tobacco smoke constituents discussed here was bused on environmental and toxicological aspects. The following substances are considered in detail: carbon mon- oxide, nitrogen oxides, formaldehyde, acrolein, ammonia, phenol, hydrogen cyanide. sulfur dioxide, cadmium, nickel, nicotine, some volatile nitrosamines, polycyclic aromatic hydro- carbons, and respirable particulates. The smoke constituent concentrations in the literature are presented in tables. The results are evaluated and discussed with respect to the max- imum allowable concentrations presently valid in the Federal Republic of Germany. o ivsw AcaJamtc Press. Inc. INTRODUCTION Regarding tht: topic of indoor air pollution by smoke constituents, the following physicochemical facts are relevant: (a) tobacco smoke is a product of several complex chemical processes (such as distillation, sublimation, and pyrolysis); (b) to date, thousands of substances have been detected in tobacco smoke; and (c) mainstream and sidestream smoke usually differ in their composition. According to the Coresta definition (6), mainstream smoke (MSS) is that smoke issuing from the end of a cigarette when a smoker draws on it. Sidestream smoke (SSS) is that smoke given off by the cigarette in the pauses between puffs (23). The basis for this review was an evaluation of information given in the literature published between 1970 and 1983 (CIS card index, Abstracts of Hygiene, Bibli- ography on Smoking and Health, Current Contents-Life Sciences, Dokumen- tation Soziafttledizin, Public Health Service and Occupational Medical reports). We also included reviews of this subject matter published in recent years: Ry- lander, 1974 and 1984 (32, 33); Klosterkotter and Gono, 1976 (23); Schriftenreihe Bayerischen Akademie f. Arbeits- und Sozialmedizin, 1977 (36); WHO Docu- ments, 1977 (51, 52); Badre et al., 1978 (1); Valentin et al., 1978 (44): Report of the U.S. Surgeon General, 1979 (40); Repace, 1982 (31); Sterling et al., 1982 (39); Weiss et al., 1983 (50); and Rahmede, 1983 (29). Because of the extensive amount of information in the literatt2re dealing with ' Presented at thC Symposium "Medical Perspectives on Passive Smuking," April 9-12. 1984, Vienna, Austria. = To whom requests for reprints should be addressed at the (nstitute of Occupational and Social Medicine (Director: Prof. Di.med. H. Valentin). University Eriangen-Nurnbcrg, Schillerstrasse 25/ 29. tl520-Eriangen, Federal Republic of Germany. 570 0091-7435/84 33.00 Copynghl L PpN by Ac+demk Preaa. Inc. AU nylua ui repruJUawn in any Curm reaerveJ. TI BU 31565 i t I t

Jns laximum ntration .ted con- he max- ir within )t impair nditions, : of 8 hr ons, the in atmo- humans, ., I year) uthority. . air pol- ;tures of a laximum lausible, the pro- wlation. deration are not :he indi- :ore any possible 'e listed. SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 573 TABLE 2 U15'tR1tlUT1ON OF LMPORTANT TOBACCO SMOKE COMPONENTS IN MAINSTREAM (MS) aND SIDESTItEAM (SS) SMOKE Smoke component Ratio SS/MS N•Nitrosodimethylamine 1 ! -437 Ammonia 73-IS0 Formaldehyde 50 Nickel 0.2-31 Acrolein 12 Cadmium 4-7 Carbon monoxide 2-5 Nicotine 3 aenzo(ulpyrene 3 I Hydrogen cyanide 0.05-0.4 Nutr. Ad•,pted from Ref. (24). Ai can be seen, air contamination by tobacco smoke-related CO lies mainly bctween I and 10 ppm. In the presence of only moderate smoke pollution, for cxample, in offices, CO concentrations under 5 ppm generally occur, thus not excctaling the MIK value. In the case of more marked tobacco smoke pollution, values in excess of 10 ppm can be achieved. Even higher concentrations, namely 2t1 to 50 ppm, are found in bars and taverns, for example. Here, however, other wurct:s of CO may also be involved-for example, gas-fired heaters emit con- uderabie quantities of CO (2). More recent studies of the non-tobacco-smoke- rclated CO burden have revealed CO values of 3 to 3.5 ppm in indoor areas tuttices, restaurants, stores) (28). TABLE 3 CARBON MONOXIDE LEVELS MEASURED UNDER REALISTIC SMOKING CONDITIONS AND IN THE GENERAL ENVIRONMENT Level (ppm) Rcf. Location Ventilation Range 4fean 1181 Offices Natural 2.5-10 - 142) 49 Offices Natural 2.7 • 171 2 Taverns 6 and 2 Changes/hr 10- 12 11.5 and 17 and 3-22 151 10 Ottices - 1.5-4.5 2.5 15 Restaurants - 1.0-9.5 . 4.0 14 Bars and taverns - 3.0-29 13.0 111) 2 Restaurants Natural 1.4-9.9 2.6 and 5.1 luvern Natural 2.4-9.4 4.8 Cateteria Air conditioning 0.3-1.7 0.5 and l.: Outdoors 0.4-4.8 (491 44 Working rooms Air conditioning or Maximal 4.9 2.8 Outdoors window ventilation 2.0 .Yute. MAK value (l98i): 311 ppm; M1K value (24 hr): 8.7 ppm. TI BU 31568 .

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 581 i•'-13- w.1jurycun Gener.tl'~, Report. "Smoking and Health." U.S. Department of Health, Education, and aLions- 41 :h dem ~ (1977). ~ ae von 1 4; ialtiger ~ +1 zeugen 44 oncen- ?assiv- places. af .r ~ i7 Law- coffee Purusi- nd Ne- )llutiun Tokdy u w w iI ~: cncr to 1 . Beitr. S cund. nd out- I Wellare. Public Health Service. Washington. D.C.. 1979. St,ulLuw%ki. D.. Schultze. H.. Schaller, K. H., and Lehnert, G. Zur okologischen Bedeutung ,lcs SchwermeLallgChaILCS von Zigaretten- lJtei-, Cadmium- und Nickelunalysen des Tabaks wwie der Gaa- und Partikelphase. Arch. Hyg. Buktertal• 153, l-8 (1969). Si,,,tluwtkt. D.. Harke. H.•P.• and Angerer. J. Kohlrnmonoxidbelastung durch Passivrauchen in liururaumen. Inn. Wed. Prax. Klin. 5, 103 (1976). iunel• 1.. Hulluwell• C. D.. 4liksch. R. R•, Rudy, J. V., and Young. R. A. The effects of reduced ~cnulatwn on indoor air quality in an office building. Aunos. Environ. 17, 51-64 (1983). Valcnun, H.. kiost. H.-P., and Wawra, E. Das Passivrauchen am Arbeitspiatz-cine Gesund- hutahchadigung2 Zbl. Bukt. Hyg. 1. Abt. Orig. B 167, 405-434 (1978). VDI-Kummission. M1K-WertelVDI Richtlinien 2306 und 2310 VDI-Handbuch Reinhaltung der Luft. VDI-Veriag Gmbh, Dusseldorf. 1974. WoJe. W. A.. Ill, Cote, W. A., and Yocom. J. E. A study of indoor air quality. l. Air Pollut. ('unuu! A.isuc. 25, 933-939 (1975). %ebor-Tachupp. A., Fischer, T., and Grandjean. E. Objektive und subjektive physiologisehe Wirkungen des Passivrau~hens. Int. Arch. Occup. Environ. Health 37, 277-288 (1976). weher. A.. Fischer, T.. and Grandjean. E. Passive smoking in experimental and field conditions. h.nvrrun. Res. 20, 205-216 (1979). Weher. A.. and Fischer. T. Passive smoking at work. Int. Arch. Oc•c•up. Envirurt. Health. 47, 2I19-221 11980). 11'eub. S. T.. Tager, 1. B.. Schenker, M., and Speizer. F. E. The health effects of involuntary %muking. Ainer. Rev. Respir. Dix. 128, 933-942 (1983). w'urld fle•rlth Organization. "Nitrates. Nitrites and N-Nitroso Compounds." Environmental HcriLh Critcria 5. WHO. Geneva. 1977. N'urld (iealth Organization. •'Carbon Monoxide." Environmental Health Criteria 13. WHO. Ge- nev;t, 1979. , .

578 TRIEBIG AND ZOBER TABLE 11 NrTROSAMINE LEVELS MEASURED UNUI:R REALiJrIC SMUKJNG CONDITIUNS Range or vdlue Ref. Location (ng/ml) (3,4) House, sports hall, dibcothcyue NDMA: <3-90 Bar N DMA: 240 Train NDMA: 110 and 130 (37) Ot'fice N DMA: 30 NDEA: 30 Restaurant N DMA: 10-50 NDEA: <10 Conference room NDMA: 29-33 NDEA: <10 Note. NDMA. N-nitrosodimethylamine: NDEA. N-nitrosodirthylamine. measurements made in the households of nonsmokers revealed NDMA concen- trations of less than 5 ngJm3, according to Brunnemann and Hoffmann (4). No MAK or MIK values are now available for nitrosamines. Polycyclic• Aromatic Hydrocarbons Table 12 presents the results of determinations of polycyclic aromatic hydro- carbons under various smoking conditions. For these complex compounds, benzo[aJpyrene (BAP) may serve as a prime example. With a few exceptions, the BAP levels are of the order of magnitude of 10 to 20 ng/m3 and are thus in the same range as outdoor air concentrations. In comparison, in highly indus- trialized areas, annual mean levels of 6-68 ng BAP/m3 were found. In residential areas during heating periods, extreme values of 350 ng/m3 have been measured. For example, the Ruhr district in 1971 had an estimated annual mean value of I 11 ng BAP/ms. TABLE 12 POLYCYCLIC AROMATIC HYDROCARliON LEVtLS MEASURED UNDER REALISTIC SMUKING CONDITIUNS Level (ngtm3) Ref. Location ' Range Mean 113) Restaurant BAP: 28-144 6.2 (2'-) Cafes BAP: 0.25-10.1 BEP: 3.3-23.4 ANT: 0.5-1.9 (10) Arena BAP: 7.1 BEP: 9.9 ANT: 21.7 (15, 16) Working room fiAP: <3-25 2: - BEP: <2-21 - Iis ANT: <3-3 3 Note. BA P. benzo[ulpyre ne; BEP. benzolr lpyrene: ANT, anthanthrene. TI BU 31573 .4 h. It u ~ n n t t4 I c t t

O,) : ra- Al- our I'he oor : ue, c0 ues ed. ,om tigh ppb n to l un- ltle- .ible ple. nal- :oeS? '_U SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 575 TABLE 5 N/rRtX',EN DlOX1DE LEVELS MEASURED UNUER REALISTIC SMOKING CONDITLONS ANU lN THE GENERAL ENVIRONMtNT Level(ppb) Ref. Location Ventilation Range Mean 111.48) 2 Restaurunts Natural ?4-105 63 and 76 Tavern Natural 1-61 21 Cafeteria Air conditioning 15-1U3 27'and 58 Outdoors 34-63 1491 44 Working rooms Outdoors Air conditioning or window ventilation 10-200 60 -Nuta. `tAK value (1984): 5.UUU ppb: MIK value l34 hr): 55 ppb. nut lxrmit any definitive conclusions to be drawn as to the formaldehyde levels %;:,uhctl by tobacco smoke under realistic smoking conditions. .Ilisc•rlluitruns Under the term "miscellaneous," the smoke constituents ammonium, phenol, hytlrugen cyanide (HCN). sulfur dioxide (SO,), and the metals nickel (Ni) and wdmium (Cd) are considered. In Table 8, the concentrations of phenol and sulfur dtuxide measured under realistic conditions are listed. According to these figures. phenol was found only in the "ppb range." The levels of SO, in res- I.,urants and a tavern were, on ave-ragz, between 10 and 20 ppb. Thus, in com- p.trisun with the outdoor levels of 6-1? ppb, the indoor concentrations of SO, TABLE 6 a ACRULI:IN AND FOKMALDEHYUE LEVELS MEASURED. RESPECTIVELY, UNDER REALISTIC AND EXPEKIMENTAL CONDITIONS LCvel(ppb) i Rcf. i Location Range Mean i tI) Acrolein" 5 Cafes 3U-100 Train =U-12U Car with 3 ~,mukers (open) 30 Car with 3 smokers (closedl 300 111 i 2 Restaurants 3-13 7and$ Tavern 5-18 t0 I iaxt t.t7t Cafeteria Formalclehyde' Climatic chamber 10. 1 changerhrl I-IU 30-640 5 and 6 (CO: 1-43 ppm) • NAK value t l9l41: 100 ppb: MIKU value: 5 ppb. ' MAK value (1984): 1000 ppb; M1Ku value: 20 ppb. TI BU 31570

578 TRIEBIG AND ZOBER TABLE 11 NrTROSAMINE LEVELS MEASURED UNUI:R REALiJrIC SMUKJNG CONDITIUNS Range or vdlue Ref. Location (ng/ml) (3,4) House, sports hall, dibcothcyue NDMA: <3-90 Bar N DMA: 240 Train NDMA: 110 and 130 (37) Ot'fice N DMA: 30 NDEA: 30 Restaurant N DMA: 10-50 NDEA: <10 Conference room NDMA: 29-33 NDEA: <10 Note. NDMA. N-nitrosodimethylamine: NDEA. N-nitrosodirthylamine. measurements made in the households of nonsmokers revealed NDMA concen- trations of less than 5 ngJm3, according to Brunnemann and Hoffmann (4). No MAK or MIK values are now available for nitrosamines. Polycyclic• Aromatic Hydrocarbons Table 12 presents the results of determinations of polycyclic aromatic hydro- carbons under various smoking conditions. For these complex compounds, benzo[aJpyrene (BAP) may serve as a prime example. With a few exceptions, the BAP levels are of the order of magnitude of 10 to 20 ng/m3 and are thus in the same range as outdoor air concentrations. In comparison, in highly indus- trialized areas, annual mean levels of 6-68 ng BAP/m3 were found. In residential areas during heating periods, extreme values of 350 ng/m3 have been measured. For example, the Ruhr district in 1971 had an estimated annual mean value of I 11 ng BAP/ms. TABLE 12 POLYCYCLIC AROMATIC HYDROCARliON LEVtLS MEASURED UNDER REALISTIC SMUKING CONDITIUNS Level (ngtm3) Ref. Location ' Range Mean 113) Restaurant BAP: 28-144 6.2 (2'-) Cafes BAP: 0.25-10.1 BEP: 3.3-23.4 ANT: 0.5-1.9 (10) Arena BAP: 7.1 BEP: 9.9 ANT: 21.7 (15, 16) Working room fiAP: <3-25 2: - BEP: <2-21 - Iis ANT: <3-3 3 Note. BA P. benzo[ulpyre ne; BEP. benzolr lpyrene: ANT, anthanthrene. TI BU 31573 .4 h. It u ~ n n t t4 I c t t

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 581 i•'-13- w.1jurycun Gener.tl'~, Report. "Smoking and Health." U.S. Department of Health, Education, and aLions- 41 :h dem ~ (1977). ~ ae von 1 4; ialtiger ~ +1 zeugen 44 oncen- ?assiv- places. af .r ~ i7 Law- coffee Purusi- nd Ne- )llutiun Tokdy u w w iI ~: cncr to 1 . Beitr. S cund. nd out- I Wellare. Public Health Service. Washington. D.C.. 1979. St,ulLuw%ki. D.. Schultze. H.. Schaller, K. H., and Lehnert, G. Zur okologischen Bedeutung ,lcs SchwermeLallgChaILCS von Zigaretten- lJtei-, Cadmium- und Nickelunalysen des Tabaks wwie der Gaa- und Partikelphase. Arch. Hyg. Buktertal• 153, l-8 (1969). Si,,,tluwtkt. D.. Harke. H.•P.• and Angerer. J. Kohlrnmonoxidbelastung durch Passivrauchen in liururaumen. Inn. Wed. Prax. Klin. 5, 103 (1976). iunel• 1.. Hulluwell• C. D.. 4liksch. R. R•, Rudy, J. V., and Young. R. A. The effects of reduced ~cnulatwn on indoor air quality in an office building. Aunos. Environ. 17, 51-64 (1983). Valcnun, H.. kiost. H.-P., and Wawra, E. Das Passivrauchen am Arbeitspiatz-cine Gesund- hutahchadigung2 Zbl. Bukt. Hyg. 1. Abt. Orig. B 167, 405-434 (1978). VDI-Kummission. M1K-WertelVDI Richtlinien 2306 und 2310 VDI-Handbuch Reinhaltung der Luft. VDI-Veriag Gmbh, Dusseldorf. 1974. WoJe. W. A.. Ill, Cote, W. A., and Yocom. J. E. A study of indoor air quality. l. Air Pollut. ('unuu! A.isuc. 25, 933-939 (1975). %ebor-Tachupp. A., Fischer, T., and Grandjean. E. Objektive und subjektive physiologisehe Wirkungen des Passivrau~hens. Int. Arch. Occup. Environ. Health 37, 277-288 (1976). weher. A.. Fischer, T.. and Grandjean. E. Passive smoking in experimental and field conditions. h.nvrrun. Res. 20, 205-216 (1979). Weher. A.. and Fischer. T. Passive smoking at work. Int. Arch. Oc•c•up. Envirurt. Health. 47, 2I19-221 11980). 11'eub. S. T.. Tager, 1. B.. Schenker, M., and Speizer. F. E. The health effects of involuntary %muking. Ainer. Rev. Respir. Dix. 128, 933-942 (1983). w'urld fle•rlth Organization. "Nitrates. Nitrites and N-Nitroso Compounds." Environmental HcriLh Critcria 5. WHO. Geneva. 1977. N'urld (iealth Organization. •'Carbon Monoxide." Environmental Health Criteria 13. WHO. Ge- nev;t, 1979. , .

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 571 ery, dour air ,tandard studies de. The nmental an mun- e. sulfur _ hydro- (eraure he max- . u 19" z! following of several olysis); (b) :e; and (c) .hat smoke :am smoke wffs (23). z literature ene, Bibli- Dukumen- il reports). years: Ry- riftenreihe HO Docu- Report of lyg2 (39); :ealing with `1- 1=. 1984, d uid Sucial .erstra,be ?5l . I thc topic, we cannot make any claims as to completeness. Nevertheless, we have June our best to ensure that this review takes into account all the major facts presently known. It should first be noted that the relevant results refer almost caclusivcly to investigations into cigarette smoking, and not to other forms of tuh:,t;cu consumption. This "restriction" is, however, realistic, since, statisti- .~,,lly. cigarette smoking represents the major form of tobacco consumption. Fur- thcrmure, the demonstrable chemical differences between cigarette smoke on the one hand and cigar smoke, for example, on the other have little relevance to our rtwin interest: sidestream smoke. Finally, we mainly considered results of studies obtained under "natural" and "realistic" conditions or results of tield studies. RESULTS AND DISCUSSION '1'ubaccu smoke consists of a large number of different substances, which are cla,aitied into a gas phase and a particulate phase. Two to several thousand Jitlcrcnt substances have been reported. To date, several hundred substances hase been identified, and our consideration of them must be differentiated with rc,lxct both to the concentrations at which they occur in the air and to the rc,ulting potential health effects. In this connection, the substances may be clas- %iticd into irritants, toxins, and carcinogens. A critical analysis of the literature shows that the substances, or groups of substances, listed in Table 1 are of importance. With respect to their biological ctl'ccts in humans, carbon monoxide (CO), nicotine, and hydrogen cyanide (HCN) bclung to the resurptive toxins. Respirable particles, certain aldehydes, phenol, rntnwnium, and sulfur dioxide (SO_) are mainly irritant substances. Cadmium tCd) and nickel (Ni) compounds are also resorptive, or local, toxins. More than 40 tobacco smoke constituents have been referred to as possible human carcin- ugcna. On the basis of our present knowledge, the polycyclic aromatic hydro- carbons-of which a major representative is benzo[uJpyrene-and certain vol- alile nitrosamines must be included among the tobacco smoke constituents with potential carcinogenicity in humans. In addition to the different biological effects of the tobacco smoke constituents, the qualitative and quantitative differences between MSS, and SSS have to be considered [see Table 3'and Ref. (34)]. Since, as is well known, up to 85% of the indoor air pollution by tobacco smoke by-products is caused by SSS, the signif- icance of its constituents in connection with environmental and health-related problems is under current discussion. However, one must consider that these cuncentration ratios have been calculated from the results of experiments. For in,tance, the quantitative results of SSS analysis in the "Neurath chamber" may be influenced by the specific climatic conditions within the chamber, and thus provide values that are high in comparison with those found in a more normal cnvirunment (24). This applies in particular to the nitrosamines, for which the largcst ratios have been calculated. Apart from these methodological intluences, the abaulute concentrations of the tobacco smoke constituents are, for obvious rca.,uns, indispensable for any toxicological evaluation. Thus, it may be con- cluded that the ratios have only an orientating character and are not suitable to serve as a basis for well-founded conclusions. „ TI BU 31566

584 Mo P.P.A. 30 CARL HUGOD N 150 Fte. 2. Air concentrations of NO during Experiments l(0). 2 (x), and 4(O). accelerated declines in concentrations of NO, HCN, acrolein, and aldehydes are probably caused by condensation of these components onto skin, hair, and clothes, and probably also by binding to smoke particles. Cigarette equivalents/hour was used as an expression of the amount of each single tobacco smoke constituent inhaled by the passive smoker as compared with the amount of the same component inhaled by a smoker with the mainstream smoke of one cigarette. We prefer to use the reciprocal value-the cigarette equivalent time (CET)-which expresses the length of time (in hours) one vul- unteer would need to breathe the smoke-polluted atmosphere in order to inhale the same amount of each single smoke constituent as inhaled by one smoker from the mainstream smoke of one cigarette (Table 1). The vast differences in CET values between the different smoke constituents found in Experiment 2 can be 0.03 0.01 0.01 30 90 150 Ftc. 3. Air concentrations of NO, during Experiment 4. (!n Experiments I and 2. NO, wa, nut detected in the stmosphere.) TI BU 31579 i ~. _.._ ~.. ~ r ~ .wa

586 CARL HUGOD koi..as WJwj 9~+ 2.0 1.5 . ~ 1.0 E • O.S . ~ 7l1 90 [SO Fia. 6. Air concentrations of aldehydes in Experiments I t0l, :( x), and 4 i0). Aldehydes were estimated culorimetricully by the method of Harke et al. (4) as described by Rothwell and Grant iyl. bronchitis, sinusitis, and conjunctivitis and of cardiovascular disease 'with Stt:n- ocardia (1, 2). Also, persons allergic to tobacco smoke constitute groups at special risk (11). The special type of passive smoking experienced by babies born to mothers who smoke during pregnancy comprise a third special risk group. General conclusions about lasting adverse health effects cannot be drawn from quantitative short-term studies. However, it is possible, with reference to thc literature and to existing medical knowledge, to identify groups of persons who should be protected against passive smoking or be encouraged to abstain from smoking. There are sufficient data to (a) advise pregnant women to avoid smoking ToTAt. PaaTICUU4TE MATTER (TPM) N1COTINE MG /M3 1413 /1413 8.0 x 14 0 x . 1Z 0 6.0 0 . - 0.10 4.0 2.0 FiG. 7. Air concentrations of total particulate matter (TPM) and nicotine in Experiments I*t. : ( x i, and 4 (C 1. Niwtine was estimated atter passing a total of 2,540 liters of air through a Csmbnugc tllter over a S•hr period. Nicotine wab extracted and determined by gab chrumutugruphy. TN'+A was determined gravimetrically on the Ntlme filter. Concentrations of nicotine and TPM in relation to llnle could therefore not be estimated. TI BU 31581 r F

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 579 iuN5 "omatic hydro- c cumpounds, w exceptiuns, .nd are thus in t highly indus- . In residential ten measured. mean value ot' ctuG Curvurnons 6.2 CONCLUSIONS ncc major objective of our contribution is to provide an actual review on the .ubjcct uf air pollution by smoke constituents with respect to the exposure of p.n„ive smokers. A synoptic consideration of the available tabular data results in r nonuniform picture. If the individual substances are evaluated on the base uf their MIK values-under the given restrictions-CO and SO, levels are gen- crally not exceeded when usual worksites (e.g., offices) are considered and indoor rrcab with particularly high concentrations of tobacco smoke pollution (e.g., ntlhtclubs and taverns) are excluded. t'unt;tcntratiuns within the range of the MIK values or in excess of these levels crn bc f'uund for respirable particulates, NO,, and acrolein. At the present time, hccaua: of a lack of MIK values and/or of data obtained under realistic conditions, no relevant evaluation is possible for nicotine, HCN, phenol, nitrosamines, puly`yt:lic aromatic hydrocarbons, or the metals Cd and Ni. Vtnally, it must be stated that, at the present time, no quantitative estimation uf individual personal body burden, possible adverse effects, or manifest impair- mcnl tu health is possible on the basis of the concentrations of tobacco smoke cunatitutents in indoor air. REFERENCES I U,Wre, R., Guillerm. R., Abran, N., Bourding, M., and Dumas, C. Pollution atmospherique par la lumec de tabac. Ann. Phurni. Fr. 36, 443 (1978). : Urunckreef, B.. Smit, H. A., Biersteker, K., Buleij, J. S. M., and Lebret, E. Indoor carbon ntunuxtde pollutiun in the Netherlands. Environ. Inr. 8, 193-196 (L982). 1. Urunnemann, K. D., Adams. 1. D.. Ho. D. P. S., and Hotfmann, D. The intluence of tobacco smuke on indoor atmospheres. It. Volatile and tobacco specific nitrosamines in main- and ~tdcstraun smoke and their contribution to indoor pollution, in "American Chemical Society 4th Joint Conference on Sensing of Environmental Pollutants," 1977. 4. Urunncman. K. D.. and Hoffmann, D. Chemical studies on tobacco smoke. LIX. Analysis of volatile nitrosamincs in tobacco smoke and polluted indoor environments, in "Environmental Aspects of N-Nitroso-Compounds" (E. A. Walker, L, Griciute, M. Castegnaro, and R. E. Lyle. Eds.), pp. 343-355. lnternationai Agency for Research on Cancer, Lyon, 1978. t Chappell. S. B., and Parker. R. Smoking and carbon monoxide levels in enclosed public places in New Brunswiek. Cunud. J. Public Health 68, 159 (1977). 60+Kt.,tA. Cuuperation Centre for Scientific Research relative to Tobacco. Paris, Kunferenz der Arbcttsgruppe Rauch, Budapest, Ungarn: 1962. 7 Cuddeb;uk, J. E., Donovan, J. R.. and Burg, W. E. Occupational aspects of passive smoking. Amrr. Ind. Hyg. Ase•oc. J. 37, 263 11976). a tkJly, K. A., Hunrahan. L. P., Woodbury, !Gt. A., and Kanarek, M. S. Formaldehyde exposure in nunoccupatiunal environments. Arc•!t. Environ. HraUh 36, ?77-?64 (1y81). v heut,che Furschungsgcmeinschal't. "Maximale Arbeitsplatzkunzentationen und 13iulugische Ar- bettaatufftulcranzwcrte 1984." Verlag Chemie. Weinheim, 1984. gu. Elliott. 1. f?. and Rowe. D. R. Air quality during public gatherings. J. Air Pollur. CvnrroJ Ae•soc. 25, 635 (1975). . 11 fi.cher. T.. Weber. A., and Grandjean. E. Air pollution by tobacco smoke in restaurants and bars. Jrtt. Arch. Orc•up. Enmirun. Heu!!h 41, 267 (1978). 12. Fnedrichs, K. H.. Stuke, J., Brockhaus. A.. and Steiger. H. Luftverunreinigung im Ruhrgebitt- ticurtedung cines Messprogramms unter lutihygicnischen Gesichtspunkten. Sruab 31, 323-326 11971). 11, ti.rluakinova. V. 3-4 Benzotu)pyrene determination in the smokey atmosphere of social meeting rooms and restaurants. Neupfusrn 2. 465-468 (1964). 14 Gorinan,1. R.. Apte. M. G.. Traynor. G. W., Allen, J. R., and Hollowell. C. D. Pollutant emission TI BU 31574

SYMIK)SIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 585 s. are and each )aued ream trette . -vol- nhale from CET tn be 0.10 0.06 0.06 0.0% 0.02 . . 3' S0 90 t50 ~ - f t .. e Air cuncentrations of HCN in Experiments 1(0), 2( x), and 4(O). HCN was measured 1 ti emo*nt{ rn air ~,ample of 45 liters (500 mUmin for 90 min) through IW ml of 0. 1 molelliter NaOH 5 ~ .rn wu.cyucnt estiination of cyanide by the method described by Elkins (3). t +ccn in this table. Further it is obvious that estimating CO exposure gives no I pcncr;J tmpresaion of exposure to other tobacco smoke constituents. `uw we turn to the subjective discomfort of passive smoking Figure 8(general r . annu)rncel illustrates that a maximum value is reached at'ter approximately I hr arW that irritation remains at this level for the remainder of the experiment. It is am,wMurthy that "general annoyance," as experienced by exposure to the gas l+fiaw uf tubacco smoke, is at the same level as that derived from total sidestream ,mu6c antl, further, that exposure to acrolein (in concentrations three times those 0 ui l:xpcriments I and 2) causes only slight discomfort. The curves in Fig. 8 are almust ttlt:ntical to those obtained for eye irritation. It t% unyuestionable that passive smoking is connected with immediate discom- 1furt. which may be accentuated considerably by the concomitant presence of i ACp.L`. ;5 W ( A r - - 0 2.0 - 1-2 - - 1.5 f : I 1.0 0.5 . ln nut Fr. 5. Air cuncentrations of acrolein in Experiments I l0). 2 (x ), and 4 (0). Acrolein was esti- wuJ culurimetric:dly (4). TI BU 31580

584 Mo P.P.A. 30 CARL HUGOD N 150 Fte. 2. Air concentrations of NO during Experiments l(0). 2 (x), and 4(O). accelerated declines in concentrations of NO, HCN, acrolein, and aldehydes are probably caused by condensation of these components onto skin, hair, and clothes, and probably also by binding to smoke particles. Cigarette equivalents/hour was used as an expression of the amount of each single tobacco smoke constituent inhaled by the passive smoker as compared with the amount of the same component inhaled by a smoker with the mainstream smoke of one cigarette. We prefer to use the reciprocal value-the cigarette equivalent time (CET)-which expresses the length of time (in hours) one vul- unteer would need to breathe the smoke-polluted atmosphere in order to inhale the same amount of each single smoke constituent as inhaled by one smoker from the mainstream smoke of one cigarette (Table 1). The vast differences in CET values between the different smoke constituents found in Experiment 2 can be 0.03 0.01 0.01 30 90 150 Ftc. 3. Air concentrations of NO, during Experiment 4. (!n Experiments I and 2. NO, wa, nut detected in the stmosphere.) TI BU 31579 i ~. _.._ ~.. ~ r ~ .wa

580 TRIEBIG AND ZOBER rates from indoor combustion appliances and sidestream cigarette smoke. Environ. lnr. 8, 213- 221 (1982). 15. Grimmer, G., Bohnke, H., and Harke. H. P. Zum Problem des Passivrauchens: Konzentratiuns- mesaungen von polyzyklischen aromatischen Kohlenwasserstoffen in Innenraumen nach dem maschincllen Abrauchen von Zigaretten. Int. Arch. Occup. Environ. Health 40, 83-92 (1977). 16. Grimmer. G.. Bohnke, H., and Harke, H. P. Zuni Problem des Passivrauchens: Aufnuhme von polyzyklischen aromatischen Kohlenwasserstoffen durch Einatmen von zigaretttnhaluger Luft. Int. Arch. Occrrp. Environ. Heulth 40, 93-99 (1977). ~ 17. Grimmer, G. Bilanzierung der krebserzeugenden Wirkung von Emissionen aus Kraftfahrzeugen und Kuhleofen mit carcinogenspezitischen Tesren. Funkt. Blul. 1?ed. 1, 29-38 11982). 18. Harke, H.-P. The problem of passive smoking. 1. The influence of smoking on the CO concen- tration of office rooms. Int. Arch. Arbritsmed. 33, 99 (1974). 19. Henschler, D. Aufnahme, Metabolismus und Wirkung von Nikocin beim Aktiv- und Passiv- Rauchen. Arbeitsmed., Soziulmed.. Priiventivnred. 12. 237 (1975). ~ 20. Hinds, W. C., and First. M. W. Concentrations of nicotine and tobacco smoke in public places. New. Engl. J. Med. 292. B44 (1975). 21. Hollowell, C. D. "lmpact of Energy Conservation in Buildings on Health." Report 9379, Law- rence Berkeley Laboratory, Berkeley, Calif. 22. Just. J., Borkowska, M., and Maziarka, S. Air pollution by tobacco smoke in Warsaw coffee house. Rocz. Panrstw. Zakf. Hig. 23. 129 (1972). 23. Klosterkdtter, W., and Gono, E. Zum Problem des Passivrauchens. Zentrlbl. Bukteriul. Purua•i- i tenkd. lnjekriunskr. H,vg. Abr. 1: Orig. Relfre B 162, 51 (1976). 24. Klus. H., and Kuhn, H. Verteilung verschiedener Tabakrauchbestandteile auf Haupt- und Ne- ~ benseromrauch (Eine Ubersicht). Beitruge Tabuk}'orsch. 11, 229 (1982). 25. Klus, H., Begutter, H., Novak, A., Pinterits, G., Ultsch, I., and Wihlidal. H. Indoor air pulluttun to tobacco smoke under real conditions: Preliminary results. Paper presented at the Tukay University Medical Symposium on "Effacts of Indoor Air Pollution with Special Rttercnce tu Nitrogen Oxides and Smoking," Lyngby, Denmark. 1984. 26. Lipp, G. Zur Definition der Stlektivitut und der verschiedenen Rauchstrome der Zigarette. Bear. Tubakjursch, 3, 220 ( l%5). 27. Niemeia. R., and Vainio, H. Formaldehyde exposure in work and the general environment. Scund. J. Work Environ. Heulth 7, 95-l00 (1981). ' 28. Ott, W., and Flachsban. P. Measurement of carbon monoxide concentrations in indoor and uut- door locations using personal exposure monitors. E,rvirun. Int. 8, 295-304 (1982). 29. Rahmede, J. "Passivrauchen. Gesundheitliche Wirkungen und rechtliche Konsequenzen. Juru- tische Schriften." Band 27: Recht und Medizin. Mannhold Verlag, Gelscnkirchen 1983. 30. Repace, J. L., and Lowrey, A. H. Indoor air pollution. tobacco smoke, and public health. Science (Was/ritrgtua, D.C.) 208, 464 (1980). 31. Repace, J. L. Indoor air pollution. Environ. Int. 8, 21-36 (1982). 32. Rylander, R. (Ed.). "Environmental Tobacco Smoke Effects on the Ion-Smoker. Report from a Workshop." University of Geneva. Geneva. 1974. 33. Rylander, R., Peterson. Y., and Snella, M.-C. "ETS-Environmental Tobacco Smoke. Report from a Workshop on Effects and Exposure Levels." University of Geneva. Genev:,, 1983. 34. Schenker, M. B., Weiss, S. T., and Mur.twaski, B. 1. Health effects of residence in homes with urea formaldehyde foam insulation: A pilot study. Envirun. lnt. 8, 359-363 111)82). 35. Scherer. G., and Barkemeyer, H. Cadmium concentration in tobacco and tobacco smoke. Ecu- ruzicut. Environ. Safety 7, 71-78 (1983). 36. Schriftenreihe der Bayerischen Akademie fur Arbeits- und Sozialmedizin, Munchen. "Passtv- rauchen am Arbeitsplatz." A. W. Gentner Verlag, Stuttgart 1977. 37. Stehlik. G.. 0. Richter. and H. Altmann. Concentration of dime(hylnitrosamine in the air oi smoke-811Cd rooms. Ecuru.ricul. Envirun. Sujerv 6, 495-500 (1982). 38. Sterling, T. D., and Kobayashi. D. M. Exposure to pollutants in enclosed "living spaces." En- viron Res. 13, t-35 (1977). 39. Sterling, T. D., Dimich, H., and Kobayashi, D. Indoor byproduct levels of tobacco Smuke: A critical review of the literature. J. Air Pullat. Contrul Assuc. 32, 250-259 (1982). TI BU 31575 ,

580 TRIEBIG AND ZOBER rates from indoor combustion appliances and sidestream cigarette smoke. Environ. lnr. 8, 213- 221 (1982). 15. Grimmer, G., Bohnke, H., and Harke. H. P. Zum Problem des Passivrauchens: Konzentratiuns- mesaungen von polyzyklischen aromatischen Kohlenwasserstoffen in Innenraumen nach dem maschincllen Abrauchen von Zigaretten. Int. Arch. Occup. Environ. Health 40, 83-92 (1977). 16. Grimmer. G.. Bohnke, H., and Harke, H. P. Zuni Problem des Passivrauchens: Aufnuhme von polyzyklischen aromatischen Kohlenwasserstoffen durch Einatmen von zigaretttnhaluger Luft. Int. Arch. Occrrp. Environ. Heulth 40, 93-99 (1977). ~ 17. Grimmer, G. Bilanzierung der krebserzeugenden Wirkung von Emissionen aus Kraftfahrzeugen und Kuhleofen mit carcinogenspezitischen Tesren. Funkt. Blul. 1?ed. 1, 29-38 11982). 18. Harke, H.-P. The problem of passive smoking. 1. The influence of smoking on the CO concen- tration of office rooms. Int. Arch. Arbritsmed. 33, 99 (1974). 19. Henschler, D. Aufnahme, Metabolismus und Wirkung von Nikocin beim Aktiv- und Passiv- Rauchen. Arbeitsmed., Soziulmed.. Priiventivnred. 12. 237 (1975). ~ 20. Hinds, W. C., and First. M. W. Concentrations of nicotine and tobacco smoke in public places. New. Engl. J. Med. 292. B44 (1975). 21. Hollowell, C. D. "lmpact of Energy Conservation in Buildings on Health." Report 9379, Law- rence Berkeley Laboratory, Berkeley, Calif. 22. Just. J., Borkowska, M., and Maziarka, S. Air pollution by tobacco smoke in Warsaw coffee house. Rocz. Panrstw. Zakf. Hig. 23. 129 (1972). 23. Klosterkdtter, W., and Gono, E. Zum Problem des Passivrauchens. Zentrlbl. Bukteriul. Purua•i- i tenkd. lnjekriunskr. H,vg. Abr. 1: Orig. Relfre B 162, 51 (1976). 24. Klus. H., and Kuhn, H. Verteilung verschiedener Tabakrauchbestandteile auf Haupt- und Ne- ~ benseromrauch (Eine Ubersicht). Beitruge Tabuk}'orsch. 11, 229 (1982). 25. Klus, H., Begutter, H., Novak, A., Pinterits, G., Ultsch, I., and Wihlidal. H. Indoor air pulluttun to tobacco smoke under real conditions: Preliminary results. Paper presented at the Tukay University Medical Symposium on "Effacts of Indoor Air Pollution with Special Rttercnce tu Nitrogen Oxides and Smoking," Lyngby, Denmark. 1984. 26. Lipp, G. Zur Definition der Stlektivitut und der verschiedenen Rauchstrome der Zigarette. Bear. Tubakjursch, 3, 220 ( l%5). 27. Niemeia. R., and Vainio, H. Formaldehyde exposure in work and the general environment. Scund. J. Work Environ. Heulth 7, 95-l00 (1981). ' 28. Ott, W., and Flachsban. P. Measurement of carbon monoxide concentrations in indoor and uut- door locations using personal exposure monitors. E,rvirun. Int. 8, 295-304 (1982). 29. Rahmede, J. "Passivrauchen. Gesundheitliche Wirkungen und rechtliche Konsequenzen. Juru- tische Schriften." Band 27: Recht und Medizin. Mannhold Verlag, Gelscnkirchen 1983. 30. Repace, J. L., and Lowrey, A. H. Indoor air pollution. tobacco smoke, and public health. Science (Was/ritrgtua, D.C.) 208, 464 (1980). 31. Repace, J. L. Indoor air pollution. Environ. Int. 8, 21-36 (1982). 32. Rylander, R. (Ed.). "Environmental Tobacco Smoke Effects on the Ion-Smoker. Report from a Workshop." University of Geneva. Geneva. 1974. 33. Rylander, R., Peterson. Y., and Snella, M.-C. "ETS-Environmental Tobacco Smoke. Report from a Workshop on Effects and Exposure Levels." University of Geneva. Genev:,, 1983. 34. Schenker, M. B., Weiss, S. T., and Mur.twaski, B. 1. Health effects of residence in homes with urea formaldehyde foam insulation: A pilot study. Envirun. lnt. 8, 359-363 111)82). 35. Scherer. G., and Barkemeyer, H. Cadmium concentration in tobacco and tobacco smoke. Ecu- ruzicut. Environ. Safety 7, 71-78 (1983). 36. Schriftenreihe der Bayerischen Akademie fur Arbeits- und Sozialmedizin, Munchen. "Passtv- rauchen am Arbeitsplatz." A. W. Gentner Verlag, Stuttgart 1977. 37. Stehlik. G.. 0. Richter. and H. Altmann. Concentration of dime(hylnitrosamine in the air oi smoke-811Cd rooms. Ecuru.ricul. Envirun. Sujerv 6, 495-500 (1982). 38. Sterling, T. D., and Kobayashi. D. M. Exposure to pollutants in enclosed "living spaces." En- viron Res. 13, t-35 (1977). 39. Sterling, T. D., Dimich, H., and Kobayashi, D. Indoor byproduct levels of tobacco Smuke: A critical review of the literature. J. Air Pullat. Contrul Assuc. 32, 250-259 (1982). TI BU 31575 ,

sYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 587 :dehydes were and Grant (9). - with stcn• ?s at speciul ,ies born to ;roup. drawn from ence to the ,zrsons who .bstain from oid smoking TABLE l SrwL CuNSfITUENTS INHALED BY PASSIVE SMUKING EXYRES5ED AS CIGARETTE EQUIVALEN'rS PER ftUUR ICE/hr) AND CIGARETTL• EIlUIVALENT TIME (CET) (IN hr) CunSuwent CL. hr CET thr) NO 0.60 1.6 l'O 0.50 2.0 Aldehyde 0.26 3.9 Acrulein 0.14 7. 1 , rPM 0.09 11.0 N icutine 0.02 50.0 Cyanide 0.02 50.0 %..+. Fur catcutatiun of CET, mean CO yield from mainstream smoke was taken from Hoegg (61. •4.,w,ne And TPM yields were based an analytical data from cigarettes used in the present study. 4.a." yicthi ul uther compounds are taken from "The Health Consequences of Smoking" (1972) ttt) F.x crlcutattuns refer to (7). w urdcr to tiecure optimal growth of the fetus; (b) advise parents of babies and .nt.all chtldren to abstain from smoking in the presence of their children; and (c) jwutcct patients with chronic heart or vascular disease and patients with chronic pulmonary tlisease, including bronchial asthma, against passive smoking. Further, a t, thi.: opinion ot' the Danish National Board of Health that smokers should be en.uurn6ed to be conscious of the immediate discomfort experienced by people tn thctr t;nvironment who are forced to become passive smokers. • at .esitn.er t UN/r1 ments l (0), 2 h ;i C:rmbndgc ohy. TP.'N waa C1:111Un to ttnte a 30 60 90 120 750 180 wINUis! Fw.- tl. General annuyunce of passive smukers in Expcriments I (A). 4 t0). 5(A). and 3 (C). TI BU 31582 > i

588 CARL HUGOD REFERENCES I. Anderson, E. W., Andelman. R. J.. Str.tuch. J. yl.. Furtuin, N. J., and Knelson. 1. H. Effect of tow-level carbon monoxide on onset anJ duration of angina pectoris. Ann. Inrern. .tileJ. 79, ~ 46-50 (1973). 2. Aronow. W. S.. Stemmer, E. A.. and Isbell, M. W. Effect of carbon monoxide exposure on ~ intermittent claudication. Circuluriun 49, 415--t17 (1974). 3. Elkins, H. B. "The Chemistry of Industrial Toxtcology," 2nd zd., p. 345. Wiley, New York. 1959. 4. Harke, H. P., Buars, A.. Frahm. B., Peters. H., and Schultz. C. Zum Problem des Passivrauchens. lnr. Arch. ArbeitsmeJ. 29, 323-339 (1972). : 5. Hinds. W. C., and First. M. W. Concentrations of nicotine and tobacco smoke in public places. New Engf. J. Ntrd. 292, 8•i4-845 (1975). 6. Hoegg, U. R. Cigarette smoke in closed spaces. Environ. Health Perspect. 117-128 (1972). 7. Hugod. C., Hawkins, L. H., and Astrup. P. Exposure of passive smokers to tobacco !,mukc constitutents. lrtt. Arc'It. Occup. Environ. Health 42, 21-29 (1978). 8. Jones. R. D., Commins, B. T., and Cernik, A. A. Blood lead and carboxyhaemogiubin leveh in London taxi drivers. Luncet 2, 302-303 (197:). Fwr I t , 9. Rothwell. M., and Grant. C. C. "Standard Methods for the Analysis of Tobacco Smokt," 2nd ed., Research Paper 2. Toba.co Research Council. London, 1974. 10. Russell. M. A. H., Cole, P. V., and Brown, E. Absorption by non-smokers of carbon monoxide from room air polluted by tobacco smoke. Lunc•et 1, 576-579 (1973). ~ l t. U.S. Department of Health. Education, and Welfare. "The Health Consequences of Smoking. A 3+ I Report of the Surgeon General." U.S. Govt. Printing Office, Washington, D.C., 1972. ~ a f 0. TI BU 31583

588 CARL HUGOD REFERENCES I. Anderson, E. W., Andelman. R. J.. Str.tuch. J. yl.. Furtuin, N. J., and Knelson. 1. H. Effect of tow-level carbon monoxide on onset anJ duration of angina pectoris. Ann. Inrern. .tileJ. 79, ~ 46-50 (1973). 2. Aronow. W. S.. Stemmer, E. A.. and Isbell, M. W. Effect of carbon monoxide exposure on ~ intermittent claudication. Circuluriun 49, 415--t17 (1974). 3. Elkins, H. B. "The Chemistry of Industrial Toxtcology," 2nd zd., p. 345. Wiley, New York. 1959. 4. Harke, H. P., Buars, A.. Frahm. B., Peters. H., and Schultz. C. Zum Problem des Passivrauchens. lnr. Arch. ArbeitsmeJ. 29, 323-339 (1972). : 5. Hinds. W. C., and First. M. W. Concentrations of nicotine and tobacco smoke in public places. New Engf. J. Ntrd. 292, 8•i4-845 (1975). 6. Hoegg, U. R. Cigarette smoke in closed spaces. Environ. Health Perspect. 117-128 (1972). 7. Hugod. C., Hawkins, L. H., and Astrup. P. Exposure of passive smokers to tobacco !,mukc constitutents. lrtt. Arc'It. Occup. Environ. Health 42, 21-29 (1978). 8. Jones. R. D., Commins, B. T., and Cernik, A. A. Blood lead and carboxyhaemogiubin leveh in London taxi drivers. Luncet 2, 302-303 (197:). Fwr I t , 9. Rothwell. M., and Grant. C. C. "Standard Methods for the Analysis of Tobacco Smokt," 2nd ed., Research Paper 2. Toba.co Research Council. London, 1974. 10. Russell. M. A. H., Cole, P. V., and Brown, E. Absorption by non-smokers of carbon monoxide from room air polluted by tobacco smoke. Lunc•et 1, 576-579 (1973). ~ l t. U.S. Department of Health. Education, and Welfare. "The Health Consequences of Smoking. A 3+ I Report of the Surgeon General." U.S. Govt. Printing Office, Washington, D.C., 1972. ~ a f 0. TI BU 31583

SY.MPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 583 .civerse Cted to JbaccU rations ,y con- evatc:d : ported hy- at was ersons f ~ )assive .z con- )assive tion. I room '_. 1984. E:.%pcriments 1, 2, and 4 were started by burning 20 cigarettes in the zxperi- tnent.J ruum. thus creating a CO concentration of 20 ppm. This concentration kticl was ,ubsequently maintained throughout the 3-hr experimental period by .,xnt+(cmentary burning of cigarettes (about 10 cigarettes/hr). tn Experiment 4, thc p.,rticlc phase of the smoke was retained in a high-voltage filter, and in Ex- p<rtntcnt 5 a calculated amount of acrolein was evaporat.ai into the atmosphere pxtur tu the ntart of the experimental period. Experiment 3 served as a control c+pcnntcnt with no intended air pollution. In Experiments 1, 4, and 5 the vol- uotccr% were not admitted to the room until the air pollution had reached the jo.,trrt] level. Hgurt; I demonstrates that air concentrations of CO were at the intended levels in tiatx:riment!i 1, ?, and 4. The mean increase in COHb concentrations ranged trum 0.6 to 1.7%. Figure 2 shows air concentrations of NO. It can be seen that luwcr valucs were measured when people were present than when people were ah+cnt. Nitrogen dioxide could be detected in the room (Fig. 3) only when the .mukc was passed through the high-voltage filter. Lower mean values were also rA+.crvcd t'ur HCN in sidestream smoke when people were present than when tttcy Ncro absent (Fig. 4). In +pite of the presence of people, higher values were found with gas phase p,Ilutiun than with sidestream smoke, possibly because of liberation of HCN hunt cyanugenic compounds which pass through the high-energy field in the filter. I-ur acrulein and aldehydes the same situatiun holds true: lower values were t.x,nd with people present than without (Figs. 5 and 6). Consistently higher values Mcrc fuuntd when gas phase alone was supplied. . I•ur total particulate matter (TPM) and nicotine, again lower concentrations »crc found with people present than when absent (Fig. 7). In the gas phase ctipcritnent. only traces were found of these compounds-demonstrating the ef- fiactwy of the filter. A physiological respiratory uptake by the volunteers could pu.ubly have accounted for the disappearance of TPM and nicotine, but the w 20 10 q qp lS0- .. Fw. i. Air eunaentrations of CO during Experimcnts I t*). 2 (x ). and 4 IC1.

SY.MPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 583 .civerse Cted to JbaccU rations ,y con- evatc:d : ported hy- at was ersons f ~ )assive .z con- )assive tion. I room '_. 1984. E:.%pcriments 1, 2, and 4 were started by burning 20 cigarettes in the zxperi- tnent.J ruum. thus creating a CO concentration of 20 ppm. This concentration kticl was ,ubsequently maintained throughout the 3-hr experimental period by .,xnt+(cmentary burning of cigarettes (about 10 cigarettes/hr). tn Experiment 4, thc p.,rticlc phase of the smoke was retained in a high-voltage filter, and in Ex- p<rtntcnt 5 a calculated amount of acrolein was evaporat.ai into the atmosphere pxtur tu the ntart of the experimental period. Experiment 3 served as a control c+pcnntcnt with no intended air pollution. In Experiments 1, 4, and 5 the vol- uotccr% were not admitted to the room until the air pollution had reached the jo.,trrt] level. Hgurt; I demonstrates that air concentrations of CO were at the intended levels in tiatx:riment!i 1, ?, and 4. The mean increase in COHb concentrations ranged trum 0.6 to 1.7%. Figure 2 shows air concentrations of NO. It can be seen that luwcr valucs were measured when people were present than when people were ah+cnt. Nitrogen dioxide could be detected in the room (Fig. 3) only when the .mukc was passed through the high-voltage filter. Lower mean values were also rA+.crvcd t'ur HCN in sidestream smoke when people were present than when tttcy Ncro absent (Fig. 4). In +pite of the presence of people, higher values were found with gas phase p,Ilutiun than with sidestream smoke, possibly because of liberation of HCN hunt cyanugenic compounds which pass through the high-energy field in the filter. I-ur acrulein and aldehydes the same situatiun holds true: lower values were t.x,nd with people present than without (Figs. 5 and 6). Consistently higher values Mcrc fuuntd when gas phase alone was supplied. . I•ur total particulate matter (TPM) and nicotine, again lower concentrations »crc found with people present than when absent (Fig. 7). In the gas phase ctipcritnent. only traces were found of these compounds-demonstrating the ef- fiactwy of the filter. A physiological respiratory uptake by the volunteers could pu.ubly have accounted for the disappearance of TPM and nicotine, but the w 20 10 q qp lS0- .. Fw. i. Air eunaentrations of CO during Experimcnts I t*). 2 (x ). and 4 IC1.

586 CARL HUGOD koi..as WJwj 9~+ 2.0 1.5 . ~ 1.0 E • O.S . ~ 7l1 90 [SO Fia. 6. Air concentrations of aldehydes in Experiments I t0l, :( x), and 4 i0). Aldehydes were estimated culorimetricully by the method of Harke et al. (4) as described by Rothwell and Grant iyl. bronchitis, sinusitis, and conjunctivitis and of cardiovascular disease 'with Stt:n- ocardia (1, 2). Also, persons allergic to tobacco smoke constitute groups at special risk (11). The special type of passive smoking experienced by babies born to mothers who smoke during pregnancy comprise a third special risk group. General conclusions about lasting adverse health effects cannot be drawn from quantitative short-term studies. However, it is possible, with reference to thc literature and to existing medical knowledge, to identify groups of persons who should be protected against passive smoking or be encouraged to abstain from smoking. There are sufficient data to (a) advise pregnant women to avoid smoking ToTAt. PaaTICUU4TE MATTER (TPM) N1COTINE MG /M3 1413 /1413 8.0 x 14 0 x . 1Z 0 6.0 0 . - 0.10 4.0 2.0 FiG. 7. Air concentrations of total particulate matter (TPM) and nicotine in Experiments I*t. : ( x i, and 4 (C 1. Niwtine was estimated atter passing a total of 2,540 liters of air through a Csmbnugc tllter over a S•hr period. Nicotine wab extracted and determined by gab chrumutugruphy. TN'+A was determined gravimetrically on the Ntlme filter. Concentrations of nicotine and TPM in relation to llnle could therefore not be estimated. TI BU 31581 r F

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 579 iuN5 "omatic hydro- c cumpounds, w exceptiuns, .nd are thus in t highly indus- . In residential ten measured. mean value ot' ctuG Curvurnons 6.2 CONCLUSIONS ncc major objective of our contribution is to provide an actual review on the .ubjcct uf air pollution by smoke constituents with respect to the exposure of p.n„ive smokers. A synoptic consideration of the available tabular data results in r nonuniform picture. If the individual substances are evaluated on the base uf their MIK values-under the given restrictions-CO and SO, levels are gen- crally not exceeded when usual worksites (e.g., offices) are considered and indoor rrcab with particularly high concentrations of tobacco smoke pollution (e.g., ntlhtclubs and taverns) are excluded. t'unt;tcntratiuns within the range of the MIK values or in excess of these levels crn bc f'uund for respirable particulates, NO,, and acrolein. At the present time, hccaua: of a lack of MIK values and/or of data obtained under realistic conditions, no relevant evaluation is possible for nicotine, HCN, phenol, nitrosamines, puly`yt:lic aromatic hydrocarbons, or the metals Cd and Ni. Vtnally, it must be stated that, at the present time, no quantitative estimation uf individual personal body burden, possible adverse effects, or manifest impair- mcnl tu health is possible on the basis of the concentrations of tobacco smoke cunatitutents in indoor air. REFERENCES I U,Wre, R., Guillerm. R., Abran, N., Bourding, M., and Dumas, C. Pollution atmospherique par la lumec de tabac. Ann. Phurni. Fr. 36, 443 (1978). : Urunckreef, B.. Smit, H. A., Biersteker, K., Buleij, J. S. M., and Lebret, E. Indoor carbon ntunuxtde pollutiun in the Netherlands. Environ. Inr. 8, 193-196 (L982). 1. Urunnemann, K. D., Adams. 1. D.. Ho. D. P. S., and Hotfmann, D. The intluence of tobacco smuke on indoor atmospheres. It. Volatile and tobacco specific nitrosamines in main- and ~tdcstraun smoke and their contribution to indoor pollution, in "American Chemical Society 4th Joint Conference on Sensing of Environmental Pollutants," 1977. 4. Urunncman. K. D.. and Hoffmann, D. Chemical studies on tobacco smoke. LIX. Analysis of volatile nitrosamincs in tobacco smoke and polluted indoor environments, in "Environmental Aspects of N-Nitroso-Compounds" (E. A. Walker, L, Griciute, M. Castegnaro, and R. E. Lyle. Eds.), pp. 343-355. lnternationai Agency for Research on Cancer, Lyon, 1978. t Chappell. S. B., and Parker. R. Smoking and carbon monoxide levels in enclosed public places in New Brunswiek. Cunud. J. Public Health 68, 159 (1977). 60+Kt.,tA. Cuuperation Centre for Scientific Research relative to Tobacco. Paris, Kunferenz der Arbcttsgruppe Rauch, Budapest, Ungarn: 1962. 7 Cuddeb;uk, J. E., Donovan, J. R.. and Burg, W. E. Occupational aspects of passive smoking. Amrr. Ind. Hyg. Ase•oc. J. 37, 263 11976). a tkJly, K. A., Hunrahan. L. P., Woodbury, !Gt. A., and Kanarek, M. S. Formaldehyde exposure in nunoccupatiunal environments. Arc•!t. Environ. HraUh 36, ?77-?64 (1y81). v heut,che Furschungsgcmeinschal't. "Maximale Arbeitsplatzkunzentationen und 13iulugische Ar- bettaatufftulcranzwcrte 1984." Verlag Chemie. Weinheim, 1984. gu. Elliott. 1. f?. and Rowe. D. R. Air quality during public gatherings. J. Air Pollur. CvnrroJ Ae•soc. 25, 635 (1975). . 11 fi.cher. T.. Weber. A., and Grandjean. E. Air pollution by tobacco smoke in restaurants and bars. Jrtt. Arch. Orc•up. Enmirun. Heu!!h 41, 267 (1978). 12. Fnedrichs, K. H.. Stuke, J., Brockhaus. A.. and Steiger. H. Luftverunreinigung im Ruhrgebitt- ticurtedung cines Messprogramms unter lutihygicnischen Gesichtspunkten. Sruab 31, 323-326 11971). 11, ti.rluakinova. V. 3-4 Benzotu)pyrene determination in the smokey atmosphere of social meeting rooms and restaurants. Neupfusrn 2. 465-468 (1964). 14 Gorinan,1. R.. Apte. M. G.. Traynor. G. W., Allen, J. R., and Hollowell. C. D. Pollutant emission TI BU 31574

sYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 587 :dehydes were and Grant (9). - with stcn• ?s at speciul ,ies born to ;roup. drawn from ence to the ,zrsons who .bstain from oid smoking TABLE l SrwL CuNSfITUENTS INHALED BY PASSIVE SMUKING EXYRES5ED AS CIGARETTE EQUIVALEN'rS PER ftUUR ICE/hr) AND CIGARETTL• EIlUIVALENT TIME (CET) (IN hr) CunSuwent CL. hr CET thr) NO 0.60 1.6 l'O 0.50 2.0 Aldehyde 0.26 3.9 Acrulein 0.14 7. 1 , rPM 0.09 11.0 N icutine 0.02 50.0 Cyanide 0.02 50.0 %..+. Fur catcutatiun of CET, mean CO yield from mainstream smoke was taken from Hoegg (61. •4.,w,ne And TPM yields were based an analytical data from cigarettes used in the present study. 4.a." yicthi ul uther compounds are taken from "The Health Consequences of Smoking" (1972) ttt) F.x crlcutattuns refer to (7). w urdcr to tiecure optimal growth of the fetus; (b) advise parents of babies and .nt.all chtldren to abstain from smoking in the presence of their children; and (c) jwutcct patients with chronic heart or vascular disease and patients with chronic pulmonary tlisease, including bronchial asthma, against passive smoking. Further, a t, thi.: opinion ot' the Danish National Board of Health that smokers should be en.uurn6ed to be conscious of the immediate discomfort experienced by people tn thctr t;nvironment who are forced to become passive smokers. • at .esitn.er t UN/r1 ments l (0), 2 h ;i C:rmbndgc ohy. TP.'N waa C1:111Un to ttnte a 30 60 90 120 750 180 wINUis! Fw.- tl. General annuyunce of passive smukers in Expcriments I (A). 4 t0). 5(A). and 3 (C). TI BU 31582 > i

PREVENTIVE NEUICINE 13, 5a2-5$$ (19134) Indoor Air Pollution with Smoke Constituents- An Experimental Investigation' CARL HUGOD National Board uf Health. Department uf Hygiene, 1. St. Kungensgude. DK-126i Cupenhugen K. Denmark In a sealed room of approximately 60 m3. sidestreum cigarette smoke was maintained at a wnstant level of approximately 20 ppm during a 3-hr experimental period. Air concen- trations of carbon monoxide, nitrogen dioxide, hydrogen cyanide, acrolein, and other al- dehydes were measured in situations with and without persons present. It was demonstr,ttcd that these smoke constituents were tncountered in lower concentrations with than without persons present. (t was further demonstrated that measurement of carbozyhemoglobin cun- centrations was not a good general indicator for exposure to various tobacco smoke cun- stituents because of the vast differences in cigarette equivalent time values between the dilTerent smoke constituents. Subjective discomfort was estimated by means of question- nsures distributed every 30 min during the experimental period and was found to be almost identical irrespective of exposure to whole Nideatream smoke or to only its gas phase. Exposure of volunteers to acrolein caused considerably less discomfort than exposure to 5ldeatream smoke or to the ph phaJe alone. t: IWG1 .icalerm: Pre». Inc. E\{k RNntJI k.cl w• %~%tutpIc, t hc p.u + [Vrtlllli pt1Ur t.' c.lWtttt untccrs Jc%trct! 1'tt;ut tn Ii~p• tnz•m O. hlNCr v 'Alwc n t. .uluhe tllcy' wl at' In lxlllutic lrtttll C} Fur . t~tuns ~ Fur wcrc t~ c\lx:rit Gacnc ptlaibl , Apart from epidemiologicat evidence scientific documentation for an adverse , ~ health effect resulting from passive smoking until recently has bec:n restricted to measurements of carbon monoxide (CO) and nicotine concentrations in tobacco ~ smoke-polluted air. More recently, carboxyhemoglobin (COHb) concentrations ~ r in exposed populations have been determined. This has been followed by con- s siderations of possible adverse health effects in people with slightly elevated ~ COHb concentrations (5, 6, 8, 10). The purposes of the experiments reported 1 : here are (a) to measure air concentrations of CO, nitrogen dioxide (NO,), hy- drogen cyanide (HCN), acrolein and other aldehydes in an atmosphere that was heavily polluted with sidestream smoke in situations with and without persons r present; (b) to investigate whether an increase in COHb concentrations in passive L smokers is a good general indicator for exposure to various tobacco smoke cun- stituents; and (c) to estimate the subjective discomfort experienced by passive ~ ; smokers. Five experiments were conducted: ~ 1. Measurement of sidestream smoke constituents with persons present. 2. Measurement of sidestream smoke constituents with persons absent. ~ 3. A control experiment with persons present, but without smoke pollution. ; 4. Measurement of gas phase constituents with persons present. 5. Exposure of persons to a known acrolein concentration. t "Persons present" always refers to lU volunteers. The same unventilated rounl of approximately 60 ms was used in all cases. ' Presented at the Symposium "Medical Perspectives on Passive Smoking," April y-1'_. lyri•i, Vienna. Auatria. 0091-7435/84 S3.00 C~pynghl ti I•Ra by AcaJem/c Pre». hpc. .ali ngma ut reprwWcnun in any turm reaervet. 582 TI BU 31577

PREVENTIVE NEUICINE 13, 5a2-5$$ (19134) Indoor Air Pollution with Smoke Constituents- An Experimental Investigation' CARL HUGOD National Board uf Health. Department uf Hygiene, 1. St. Kungensgude. DK-126i Cupenhugen K. Denmark In a sealed room of approximately 60 m3. sidestreum cigarette smoke was maintained at a wnstant level of approximately 20 ppm during a 3-hr experimental period. Air concen- trations of carbon monoxide, nitrogen dioxide, hydrogen cyanide, acrolein, and other al- dehydes were measured in situations with and without persons present. It was demonstr,ttcd that these smoke constituents were tncountered in lower concentrations with than without persons present. (t was further demonstrated that measurement of carbozyhemoglobin cun- centrations was not a good general indicator for exposure to various tobacco smoke cun- stituents because of the vast differences in cigarette equivalent time values between the dilTerent smoke constituents. Subjective discomfort was estimated by means of question- nsures distributed every 30 min during the experimental period and was found to be almost identical irrespective of exposure to whole Nideatream smoke or to only its gas phase. Exposure of volunteers to acrolein caused considerably less discomfort than exposure to 5ldeatream smoke or to the ph phaJe alone. t: IWG1 .icalerm: Pre». Inc. E\{k RNntJI k.cl w• %~%tutpIc, t hc p.u + [Vrtlllli pt1Ur t.' c.lWtttt untccrs Jc%trct! 1'tt;ut tn Ii~p• tnz•m O. hlNCr v 'Alwc n t. .uluhe tllcy' wl at' In lxlllutic lrtttll C} Fur . t~tuns ~ Fur wcrc t~ c\lx:rit Gacnc ptlaibl , Apart from epidemiologicat evidence scientific documentation for an adverse , ~ health effect resulting from passive smoking until recently has bec:n restricted to measurements of carbon monoxide (CO) and nicotine concentrations in tobacco ~ smoke-polluted air. More recently, carboxyhemoglobin (COHb) concentrations ~ r in exposed populations have been determined. This has been followed by con- s siderations of possible adverse health effects in people with slightly elevated ~ COHb concentrations (5, 6, 8, 10). The purposes of the experiments reported 1 : here are (a) to measure air concentrations of CO, nitrogen dioxide (NO,), hy- drogen cyanide (HCN), acrolein and other aldehydes in an atmosphere that was heavily polluted with sidestream smoke in situations with and without persons r present; (b) to investigate whether an increase in COHb concentrations in passive L smokers is a good general indicator for exposure to various tobacco smoke cun- stituents; and (c) to estimate the subjective discomfort experienced by passive ~ ; smokers. Five experiments were conducted: ~ 1. Measurement of sidestream smoke constituents with persons present. 2. Measurement of sidestream smoke constituents with persons absent. ~ 3. A control experiment with persons present, but without smoke pollution. ; 4. Measurement of gas phase constituents with persons present. 5. Exposure of persons to a known acrolein concentration. t "Persons present" always refers to lU volunteers. The same unventilated rounl of approximately 60 ms was used in all cases. ' Presented at the Symposium "Medical Perspectives on Passive Smoking," April y-1'_. lyri•i, Vienna. Auatria. 0091-7435/84 S3.00 C~pynghl ti I•Ra by AcaJem/c Pre». hpc. .ali ngma ut reprwWcnun in any turm reaervet. 582 TI BU 31577

SYMIK)SIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 585 s. are and each )aued ream trette . -vol- nhale from CET tn be 0.10 0.06 0.06 0.0% 0.02 . . 3' S0 90 t50 ~ - f t .. e Air cuncentrations of HCN in Experiments 1(0), 2( x), and 4(O). HCN was measured 1 ti emo*nt{ rn air ~,ample of 45 liters (500 mUmin for 90 min) through IW ml of 0. 1 molelliter NaOH 5 ~ .rn wu.cyucnt estiination of cyanide by the method described by Elkins (3). t +ccn in this table. Further it is obvious that estimating CO exposure gives no I pcncr;J tmpresaion of exposure to other tobacco smoke constituents. `uw we turn to the subjective discomfort of passive smoking Figure 8(general r . annu)rncel illustrates that a maximum value is reached at'ter approximately I hr arW that irritation remains at this level for the remainder of the experiment. It is am,wMurthy that "general annoyance," as experienced by exposure to the gas l+fiaw uf tubacco smoke, is at the same level as that derived from total sidestream ,mu6c antl, further, that exposure to acrolein (in concentrations three times those 0 ui l:xpcriments I and 2) causes only slight discomfort. The curves in Fig. 8 are almust ttlt:ntical to those obtained for eye irritation. It t% unyuestionable that passive smoking is connected with immediate discom- 1furt. which may be accentuated considerably by the concomitant presence of i ACp.L`. ;5 W ( A r - - 0 2.0 - 1-2 - - 1.5 f : I 1.0 0.5 . ln nut Fr. 5. Air cuncentrations of acrolein in Experiments I l0). 2 (x ), and 4 (0). Acrolein was esti- wuJ culurimetric:dly (4). TI BU 31580

590 W ERN ER STOBER oxide, can be disregarded here, because they do not influence the particulate matter composition of the smoke. While the multicomponent aerosol system is moving toward a chemical zqui- librium, the smoke kinetics involve an additional unidirectional physical procesb. which occurs simultaneously: contrary to hydrosols, the particulate phase uf an aerosol is never a truly stable dispersion. Instead, coagulation constantly reduct:a the degree of dispersity. Random particle collisions and subsequent coalescence of the colliding droplets diminish the number concentration of the droplets and increase the droplet sizes. This process is highly significant in the early stages ui' smoke formation when high number concentrations of very small droplets are generated. However, in aged smoke as it occurs in "smoke-tilled rooms," the process proceeds at an insignificantly slow rate. In practice, there are two types of tobacco smoke aerosols: one type is gen- erated by active smoking; i.e., it is created when the smoker takes a puff. Thc other type is more complex; it emerges through different routes from the burning zone of the tobacco and mixes directly with the surrounding air. The former type is known as mainstream smoke and differs in composition and physical charac- teristics from the latter type, appropriately called sidestream smoke. Presumably, the different ways of oxygen supply for combustion, i.e., active transport vs diffusion, and the pronounced occurrence of distillation and condensation within the tobacco strands during pufting cause qualitative and quantitative ditterenceh between mainstream and sidestream smoke. Although the inhalation of mainstream smoke is a feature of smoking, it must be noted that, in cigarette smoking, about 80% of the tobacco is consumed be- tween puffs by way of sidestream smoke formation. Thus, bn a per cigarette basis, more organic smoke components are released with the sidestream smuke than with the puffs inhaled by the smoker. Tables I and 2, based on the data uf Patriakanos and Hotfimann (25) and Pyriki and co-workers (28-30), respectively. TABLE I AROMATIC AMINES tN IVtAINSTREAM (MS) AND SIDESTREAM ($S) SMOKE OF A 70-mm NONFIt-71:R U.S. BLENU C1GAItETTE Compound (µging) (µgJcig) SSIMS Aniline 10.8 0.364 29.7 2-Toluidine 3.03 0.162 18.7 3-Toluidine . 2.0!! 0.0304 68.4 4-Tuluidine 1.73 0.0338 51.2 2-Ethylaniline + 2.6-dimethyianitine 1.24 0.0542 22.9 2.5-Dimethytaniline 2.37 0.0872 27.2 3-Ethylaniline + 2.4-dimethylaniline 1.20 0.0567 21.2 4-Ethylaniiine + 2,3-dimethytaniline 0.494 0.0273 18.1 1-Naphthylsunine U.067 0.0017 39.4 2-Aminubiphenyl 0.110 U.W3 36.7 3-Aminobiphenyl 0.132 0.005 26.4 4-Aminobiphenyl U. (43 0.0046 31.1 2-Methyl-l-naphthylaminc 0.117 0.0036 32.5 Source. Patrialwnos and Hoffmann (25). TI BU 31585 0 . ~ ~ ~ .. i tr.+ a,l nt.~ rc.~ th.~ nl. .ut FW; F1K '1t lt~~ uit t i ~ la~~ m.. t+~ t j~a t ~ {. ~ f.>t ~ /lt.u 1.>t• ~ ~ ul ~ ¢N1 ptt.

,mrticulatc nical equi- al process. hase of an ly reduces jalescencc iplets and r stages of oplets are oms." the pe is gen- putf. The ie burning ,rmer type al charac- esumably. .aspurt vs ion within . ifferenca:s, ig, it must ,umed be- - cigarette lm smoke he data uf pectively, 0 iUNPILTER f t t s1 sfPOStUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 591 TABLE 2 P1-ll4 Yl'L1C ABOMATIC HYDNOCARBONS IN MAINSTREAM (MS) AND SIDESTRfiAM (SS) SMOKE SS (ng/cig) MS (ng/cig) SSIMS lteniu(ulpyrene 25 12 :.1 lku(u Iatthracene 81 30 2.7 h rcnc 57 30 1.9 tficnrnrhrcne 60 29 2.1 S.wre. Pynki and co-workers (28-30). p,c t)ptcal examples. The mass ratio between sidestream and mainstream smoke t..umtstcntly well above I. In cuntraat to mainstream smoke, it is difficult to determine the actual concen- twuun% uf sitiestream smoke at the point of generation. Mainstream smoke has a Jctinctl air tluw past the combustion zone with a confined puff volume, so puff aw,% or druplet number measurements can be easily translated into original con- ccntrrtiuns. Sidestream smoke, however, is divided into a glow stream that mixes rc.Wtly with ambient air by convection, a dit'fusion stream of condensable vapor thrt txnetrates the cigarette paper, and a smoldering stream that exits from the muuthpict:c of the cigarette between puffs. Therefure, the emerging smoke con- wrtucnts can be determined accurately only by their mass or droplet number per pcttu,l ut' sampling time. They actually merge with the ambient air in differing 6...uuna and contribute in tiitterent ways to the total particulate matter in side- Wrc.un smoke. Thus, the smoke formation process of the three sidestream frac- twn, Jiffcra markedly from the smoke formation mechanism of mainstream .mukc and, 'therefore, produces smoke droplets of difterent composition and, pu+,tbly, nize. hnfurtunately, the current literature neither proves nor disproves the assump- trwt that mainstream and sidestream smoke droplets may be of different cum- la„tttun and site. Measurements on dry condensates indicate that the particulate aL,% ratio per cigarette between sidestream and mainstream smoke (SS/MS) is hctwccn 1.5 and 2.5 (34). Neurath-and Ehmke (23) quote respective total masses t+cr cigarette as 5?.U and 31.4 mg, thus yielding a ratio of 1.66, which is close to 1.7 ,u found by Brunnemann et u!. (5). Relating the Neurath and Ehmke figures tu r numinal tobacco consumption ratio of about 4.0 between sidestream and autnstrcam peritxls, the prorated condensate production per mass of consumed tuh"-cu indicates that sidestream is a less effective smoke generator by a factor ut ,,Ix,ut 2.5. This also indicates that gaseous combustion products are more prumtncnt in sidestream than in mainstream smoke. There is a sufficient body of c•ldcnt;c to support the existence of differences in the composition of the gaseous ui' sitiestream and mainstream smoke (see Table I). However, no cunsir tcnt ;ulnpu,itiun protiles of the dry condensates of sidestream and mainstream wwhc are a,•ailable to contirm the differences in composition. With regard to the potential differences in the size distribution and the mean ul'thc smoke particles, the situation in the literature is similar. Papers dealing 31586 ~

592 WERNER JTOHEtt with this problem are of varying quality, but there is a trend toward assigning somewhat smaller size values to siclestream smoke particles (16, 19. 24). The fact that sidestream smoke has a bluish tint, in contrast to the more whitish appear- ance of mainstream smoke, may point toward this probability. According to Baker (2), this color difference reflects the fact that the organic vapors in the two smoke streams experience ditferent rates of cooling and mixing with ambient air. Con- szquently, a change in particle composition is indicated. Actual size measurements have been reported by various authors. The results for mainstream particles range from 0.18 µm as count median diameter to 0.70 µm as mass median aerodynamic diameter. Sidestream particles, however, are in a somewhat lower range, from 0.10 to 0.54 µm. Porstendorfer and Schraub (27) found no significant difference in the two streams and reported values uf count medians of 0.10 to 0.12 µm in diameter. The generation rate for sidestream par- ticles was given as 6 x 109 sec' 1, which implies that sidestream smoke is subject to immediate coagulation, as long as ambient air dilution does not diminish its sidestream concentration too rapidly. In view of this, the nominal value of 3.6 x 1012 sidestream particles per cigarette as calculated for smoking one cigarette in 10 min is somewhat artificial, although it closely matches the data of Keith as quoted by Klus and Kuhn (17) and compares to a similarly artificial value uf 1.05 x 1012 particles per cigarette for mainstream smoke generation. With regard to the size distribution of the smpke droplets, investigators have measured or postulated log-normal size distributions and have given geometric standard dcs- viatiuns between 1.3 and 1.6. The current literature does not give any good reliable data confirming the as- sumptiun that differences exist in the chemical composition and physical size Ut primary mainstream and sidestream smoke particles. Aging cigarette smoke in room air will, on one hand, probably wipe out chemical differences by random coagulation, but may also change original composition by losing the more vola- tile components to the vapor phase. This is another field where the lack of ex- perimental data is prominent. Therefore, it is imperative to acquire more empirical data for all types of smoke and to investigate, in particular, the lung deposition of sidestream smoke inhaled from smoke-filled rooms. The suspicion is justified that these deposits are different from deposits of freshly generated mainstream smoke as inhaled by a smoker. In the context of this survey, primary impurtance is given to the actual depu- sition of sidestream smoke aerosul'particles in the human respiratory tract. be- cause this aerosol deposit and its bioavailability to the body by dissolution and resorption will determine the actual dose imparted to "passive smoktrs" as the body burden that may or may not cause a chronic health effect. To assess such burden, it is necessary to recall the physiological mechanics of lung deposition of inhaled particulate matter. The deposition of aerosol particles in the human respiratory tract has been a subject of theoretical and experimental research tur almost 50 years. The first theoretical model was published in 1935 by Findeisen (7) in Germany. !n subsequent years. in particular after World War 11, a variety of authors (3. 18, 36) added many theoretical and semiempirical refinements to TI BU 31587 k !; ~. tita h. it.6ht c .lt,.+a F. wn.~. Ira~t. e~ct) (.m:tui cIcr. M t11i Il .b.itk .kctct i t...! n 1 . upjW1 hrunc aa J.1t Jit:4 . a1J ). httKc N' i t and tt ,uui. i thL: k:.& JclxI. In I Utl 12.1 tnq~c,t the hc ut hui N,itc tn .kl- o( tlte modci t.4l • c%tcn.A tu'oo tI tht hrunc t cpithc. p.~r t lL: t~t rc~pir. l~lbi rc.3cnt

..... •.: te c xt.uta tNt: 13, 589-60I 1 1984) Effect of .tlerl. 79, posure un "ork. 1959. irauchens. 7IIC places. 1972). .co Nmuke Lung Dynamics and Uptake of Smoke Constituents by Nonsmokers-A Survey' WERNER STOBER IF .,,,,,..,rlrr lrtaatuie of Toxicology and Aerosol Research, Hannover, Federal Republic uj Germuny Mrxlcl, ui %muke kinetics and lung dynamics of inhaled particles are discussed and wm• t+wcJ wrth the availuble literature on mainstream and sidestreum smoke particles. The wc++turo sc:uch reveals a dearth of reliable informatiun on the deposition of inhaled par- t.ut.+te tuba.;cu !~moke components in the human lung. Scanty results on mainstream smuke r.nge Irunt unexpectedly high deposits to values in line'with predictions of conventionui Fwtncnwiw:,l Jcpusitiun models confirmed in tes(s with stable aerosols. Fur sidestreum .a.rac. unly one well-described experimental result is avuilable. It is in agreement with s.t.t)lnhcrl Jepunitiun probabilitics. Experimental and theoretical estimates of relative par- Lk Jepu+tuun in the human lung range from wmc lu'~7o for mdestream smukc particles to .wrc thrn Ku':F fur mainstream aerosol. This indicates a need for more, and better. cxptr- wtentrl J4W. c iv14t Ac:uternn; Preaa. Inc. (1k.untbu,tiun of tobacco in pipes, cigars, and cigarettes produces a relatively ,kn.c ,,nd. in most cases, visible smoke. From a physicuchemical point of view, th,. unukc is actually a multicomponent aerosol system which is in the process .4 ..ppru,,;hing a multiphase equilibrium between the vaporized smoke constit- r:,tt. tntacd into the air and a highly dispersed condensate phase made up of tnrny uny droplets of inhalable size. Each droplet contains at least one very small "tet curo uf carbonaceous and/or inorganic combustion residues. The cores serve ...uaJcnsatiun nuclei for the tinal droplet. They are enveloped by a multitude d uiyutic smoke constituents which adhere to the nuclei by adsorption and ..NtJ~n,atiun. These compounds may be products of either pyrolysis or incom- pktc cumbustiun of tobacco ingredients, or they may be original tobacco cum- t.,twnt, distiiled and recondensed near the combustion zone. Due to the various .,44ultuc, ut• the droplet constituents, the ensuing vapor phase is part of a system u)tnb Iu move toward an equilibrium state of stable partial pressures and droplet •+4c,. Lvtdcntly, all processes involving a decrease in the vapor concentration, 4" upcn air smoke dissipation or active dilution by clean air, will also diminish thc Iin..l ,mukc: particle size. Conversely, the addition of compounds occurring ,n tr,th the vapor phase and the condensed particulate state of the smoke will ..Au,c sume growth of the dispersed droplets. It is obvious that, for instance, .atcr r, a condensable combustion product will be a participant in the equili- tw.atng changes of the smoke aerosol system. Thus, tht; humidity of the gas phase *rll Intlucnct: the water content of the smoke droplets. On the other hand. cum- hv.ttun pru+lucts occurring exclusively in the gas phase, such as carbon mun- Nre+cntcJ at the Symposium -Medical Perspectives on Passive Smoking." April y-I:. 1984, Is.cnn.. Au.trta. 589 IW91-7345/H4 53.UQ C..pynght 6. IYril by icaJcmrc Prce.. In.. .VI nghn or repruJw;tuun in .my tnrm rexrvcJ. TI BU 31584

594 WERNER STOBER FIG. 1. Schematic diagratn of the pulmonary structure of the lung. tAl Terminal bronchiole ltra- cheobronchial compartment): (B. C) respiratory bronchioles; (U) alvrular duct; lEl atrium to the alveolar sacs, shown at moderate intlation. and the inhaled air takes place. In.many cases, the deposition in this region is of special importance, because an interference of the deposit with the gas exchange could impair the vital function of the lung. Furthermore, there is no ciliated epithelium in this compartment, so lung clearance is not supported by active directional transport. Instead, clearance depends either upon the solubility of the deposit or upon a very~slow phagocytic and lymphatic removal of the particles. Thus, insoluble deposits may persist and accumulate in the alveolar region for a long time. If, however, the same material is deposited in the other two compart- ments, it can .be cleared relatively quickly by ciliary transport of the mucus layer. Figure 2 gives the total relative deposition in the respiratory tract upon nose breathing. The curve represents the prediction by the Task Group model in com- parison with a variety of experimental data points. The agreement is of a quali- AERODYNAMIC DIAMETER Fu;. 2. Tutal depu,i(iun in the human respirutury iract upon nose breathing: Theoretical curve. Taak Group model 132): +: 13.3 roapiruttun per minutC. 750 cm} tidal volume I61; 0: 15.0 respirutiuny per minute. 700 cm} tidal volume 141; ~: 19.0 respirutiuna per minute, g011 cm} tidal volume 1331. TI BU 31589 tiltt~ fur part sis.lc For tribt 5CU r nu ; µm . the The mzt. Ho" the T the dt: v alreL thre reai the c;tu giut chi: mut beU see:

594 WERNER STOBER FIG. 1. Schematic diagratn of the pulmonary structure of the lung. tAl Terminal bronchiole ltra- cheobronchial compartment): (B. C) respiratory bronchioles; (U) alvrular duct; lEl atrium to the alveolar sacs, shown at moderate intlation. and the inhaled air takes place. In.many cases, the deposition in this region is of special importance, because an interference of the deposit with the gas exchange could impair the vital function of the lung. Furthermore, there is no ciliated epithelium in this compartment, so lung clearance is not supported by active directional transport. Instead, clearance depends either upon the solubility of the deposit or upon a very~slow phagocytic and lymphatic removal of the particles. Thus, insoluble deposits may persist and accumulate in the alveolar region for a long time. If, however, the same material is deposited in the other two compart- ments, it can .be cleared relatively quickly by ciliary transport of the mucus layer. Figure 2 gives the total relative deposition in the respiratory tract upon nose breathing. The curve represents the prediction by the Task Group model in com- parison with a variety of experimental data points. The agreement is of a quali- AERODYNAMIC DIAMETER Fu;. 2. Tutal depu,i(iun in the human respirutury iract upon nose breathing: Theoretical curve. Taak Group model 132): +: 13.3 roapiruttun per minutC. 750 cm} tidal volume I61; 0: 15.0 respirutiuny per minute. 700 cm} tidal volume 141; ~: 19.0 respirutiuna per minute, g011 cm} tidal volume 1331. TI BU 31589 tiltt~ fur part sis.lc For tribt 5CU r nu ; µm . the The mzt. Ho" the T the dt: v alreL thre reai the c;tu giut chi: mut beU see:

Itra- ~ the is of Inge ated :tive r the Aes . . . ~ora ~)art- tyer. 0 lose :om- uali- SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOK(NG 595 tative nature, but it is apparent that the human lung is an efficient filter, not only for particles exceeding 5 µm of aerodynamic diameter but also for very small particles below aerodynamic sizes of about 0. 1 µm. Neither mainstream nor sidestream smoke particles really extend into these high tilter efficiency ranges. For mainstream smoke particles, it is justitied to assume a log-normal size dis- tribution with a mass median aerodynamic diameter of, at most, 0.5 µm and a geometric standard deviation of 1.5 (31). This indicates that there are practically no particles exceeding a size of 2 µm, and very few particles below sizes of 0. 1 µm. Thus, all mainstream smoke particles are essentially within the range where the human respiratory tract deposits only a certain fraction of the inhaled aerosol. The same must be true for sidestream smoke, although a smaller value for the median size may slightly increase the relative deposition for the smallest sizes. However, this "tail" of the size distribution is rather insensitive in relation to the total mass deposit. The Task Group model permits a crude estimate of the regional deposition if the mass median aerodynamic diameter is known and the geometric standard deviation is between 1.2 and 4.5 (not a very restrictive requirement). The shaded areas in Fig. 3 indicate the corresponding variabilities of the deposition in the three compartments of the model. Since cigarette smoke will comply with the restraints for the size distribution, the deposition can be assessed according to the graph. The typical mass median aerodynamic diameter of 0.5 µ.m should cause, at most, about I t% nose-breathing deposition in the nasopharyngeal re- bion, while a deposit of between I and 6% may be expected for the tracheobron- chial compartment. However, the most signiticant deposition occurs in the pul- monary region and should range between 22 and 32%. In view of the actual geometric standard deviation of smoke particles, the lower deposition tigures seem to be more applicable. MASS MEOIAN OIAMETER-MICRONS F+a. 3. Depoeitwn ot polydiaparse aoruwts in the human respiratury tract accurding to Tusk Group ,;11(%;UtilpUnb (3 2). Lower llmltJn~,' curve: Geometric ltAnl(Lrlt dCY/BUOn uf IUg-nUrmat )Ize distribution 1.2 (qua.i-munudi,perne rerosul). Upper limiting curve: Geumetric ~tandard deviation uf log-normal etLc distribution 4.5 (very polydiopersa aerosol). TI BU 31590

0 .t .5 t OIAMETER Nm FIG. 4. Average total Jepusitiun ut' quasi-munudiaperac carnauba wax aerowl+ in 25 vulunteers breathing nuse•in/muuth•uut at 12 re5piratwnh per muiute und I.UUtI cmJ tidal volume according to Giacumelli-Maltuni et ul. (8): bruken line ~,howa theoretical curve uf the Tuak Group model (32). Experimental results on total deposition obtained by Giacomelli-Maltoni et al. (8) in 1972, with quasi-monodisperse hydrophobic carnauba wax particles inhaled by a human test panel at 1,000 cm3 tidal volume and a nose-in/mouth-out breathing frequency of 12 respirations per minute, confirmed the predictions of the Task Group model in this size range. Fipre 4 shows their comparison. The total de- position for 0.5-µm spheres amounted to about 23% and was close to the absolute minimum at 0.4 µ,m. However, a study by Heyder er ul. (9), using monudisperse hydrophobic sebacate droplets ranging in size from 0.2 to 1.0 µm at 500 cm3 tidal volume and a muuth-breathing frequency of 15 respirations per minute, gave considerably lower values for the total deposition in human lungs. Figure 5 shows three individual patterns of these results, which, in spite of the scatter, cunais- tently yielded values of about 10%a total deposition for all investigated sizes be- tween 0.2 and l:0 µm. Fur bigger particles, the total deposition increased, and amounted to 22% for I.9-µm particles. Not all of the differences between the two experimental studies can be ex- plained by the mouth-breathing pattern in the latter case. The authurs noted that variations in the deposition by a factor of 2 could be obtained for all particle sizes investigated, if the breathing level of the test person was not adjusted to the normal functional residual capacity of the lung. In a workshop study in 1977 (10), the two groups of investigators compared their methods and agreed finally on the validity of the lower data under carefully controlled instrumental and physiolog- ic:.if parameters. The previous differences indicate the enormous variability due to individual breathing patterns and the need of standardization for comparison between different studies. Much of this variability has been accounted for by recent improvements in deposition models. For mouth-breathing patterns, a semiempirical model by Heyder et ul. (11) reduces the number of dynamic and aerosol variables by in- troducing characteristic dimensiunless numbers like the aerodynamic deposition parameter. As a unique variable, this parameter determines the total deposition in mouth breathing for sizes above about I µm ot' aerodynamic diameter. The remaining i,ndividual variability observed in this regime is due to the intersubject TI BU 31591 I t rat en rer in: Th bhL 7. mu the wh Pa thL lat

590 W ERN ER STOBER oxide, can be disregarded here, because they do not influence the particulate matter composition of the smoke. While the multicomponent aerosol system is moving toward a chemical zqui- librium, the smoke kinetics involve an additional unidirectional physical procesb. which occurs simultaneously: contrary to hydrosols, the particulate phase uf an aerosol is never a truly stable dispersion. Instead, coagulation constantly reduct:a the degree of dispersity. Random particle collisions and subsequent coalescence of the colliding droplets diminish the number concentration of the droplets and increase the droplet sizes. This process is highly significant in the early stages ui' smoke formation when high number concentrations of very small droplets are generated. However, in aged smoke as it occurs in "smoke-tilled rooms," the process proceeds at an insignificantly slow rate. In practice, there are two types of tobacco smoke aerosols: one type is gen- erated by active smoking; i.e., it is created when the smoker takes a puff. Thc other type is more complex; it emerges through different routes from the burning zone of the tobacco and mixes directly with the surrounding air. The former type is known as mainstream smoke and differs in composition and physical charac- teristics from the latter type, appropriately called sidestream smoke. Presumably, the different ways of oxygen supply for combustion, i.e., active transport vs diffusion, and the pronounced occurrence of distillation and condensation within the tobacco strands during pufting cause qualitative and quantitative ditterenceh between mainstream and sidestream smoke. Although the inhalation of mainstream smoke is a feature of smoking, it must be noted that, in cigarette smoking, about 80% of the tobacco is consumed be- tween puffs by way of sidestream smoke formation. Thus, bn a per cigarette basis, more organic smoke components are released with the sidestream smuke than with the puffs inhaled by the smoker. Tables I and 2, based on the data uf Patriakanos and Hotfimann (25) and Pyriki and co-workers (28-30), respectively. TABLE I AROMATIC AMINES tN IVtAINSTREAM (MS) AND SIDESTREAM ($S) SMOKE OF A 70-mm NONFIt-71:R U.S. BLENU C1GAItETTE Compound (µging) (µgJcig) SSIMS Aniline 10.8 0.364 29.7 2-Toluidine 3.03 0.162 18.7 3-Toluidine . 2.0!! 0.0304 68.4 4-Tuluidine 1.73 0.0338 51.2 2-Ethylaniline + 2.6-dimethyianitine 1.24 0.0542 22.9 2.5-Dimethytaniline 2.37 0.0872 27.2 3-Ethylaniline + 2.4-dimethylaniline 1.20 0.0567 21.2 4-Ethylaniiine + 2,3-dimethytaniline 0.494 0.0273 18.1 1-Naphthylsunine U.067 0.0017 39.4 2-Aminubiphenyl 0.110 U.W3 36.7 3-Aminobiphenyl 0.132 0.005 26.4 4-Aminobiphenyl U. (43 0.0046 31.1 2-Methyl-l-naphthylaminc 0.117 0.0036 32.5 Source. Patrialwnos and Hoffmann (25). TI BU 31585 0 . ~ ~ ~ .. i tr.+ a,l nt.~ rc.~ th.~ nl. .ut FW; F1K '1t lt~~ uit t i ~ la~~ m.. t+~ t j~a t ~ {. ~ f.>t ~ /lt.u 1.>t• ~ ~ ul ~ ¢N1 ptt.

,mrticulatc nical equi- al process. hase of an ly reduces jalescencc iplets and r stages of oplets are oms." the pe is gen- putf. The ie burning ,rmer type al charac- esumably. .aspurt vs ion within . ifferenca:s, ig, it must ,umed be- - cigarette lm smoke he data uf pectively, 0 iUNPILTER f t t s1 sfPOStUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 591 TABLE 2 P1-ll4 Yl'L1C ABOMATIC HYDNOCARBONS IN MAINSTREAM (MS) AND SIDESTRfiAM (SS) SMOKE SS (ng/cig) MS (ng/cig) SSIMS lteniu(ulpyrene 25 12 :.1 lku(u Iatthracene 81 30 2.7 h rcnc 57 30 1.9 tficnrnrhrcne 60 29 2.1 S.wre. Pynki and co-workers (28-30). p,c t)ptcal examples. The mass ratio between sidestream and mainstream smoke t..umtstcntly well above I. In cuntraat to mainstream smoke, it is difficult to determine the actual concen- twuun% uf sitiestream smoke at the point of generation. Mainstream smoke has a Jctinctl air tluw past the combustion zone with a confined puff volume, so puff aw,% or druplet number measurements can be easily translated into original con- ccntrrtiuns. Sidestream smoke, however, is divided into a glow stream that mixes rc.Wtly with ambient air by convection, a dit'fusion stream of condensable vapor thrt txnetrates the cigarette paper, and a smoldering stream that exits from the muuthpict:c of the cigarette between puffs. Therefure, the emerging smoke con- wrtucnts can be determined accurately only by their mass or droplet number per pcttu,l ut' sampling time. They actually merge with the ambient air in differing 6...uuna and contribute in tiitterent ways to the total particulate matter in side- Wrc.un smoke. Thus, the smoke formation process of the three sidestream frac- twn, Jiffcra markedly from the smoke formation mechanism of mainstream .mukc and, 'therefore, produces smoke droplets of difterent composition and, pu+,tbly, nize. hnfurtunately, the current literature neither proves nor disproves the assump- trwt that mainstream and sidestream smoke droplets may be of different cum- la„tttun and site. Measurements on dry condensates indicate that the particulate aL,% ratio per cigarette between sidestream and mainstream smoke (SS/MS) is hctwccn 1.5 and 2.5 (34). Neurath-and Ehmke (23) quote respective total masses t+cr cigarette as 5?.U and 31.4 mg, thus yielding a ratio of 1.66, which is close to 1.7 ,u found by Brunnemann et u!. (5). Relating the Neurath and Ehmke figures tu r numinal tobacco consumption ratio of about 4.0 between sidestream and autnstrcam peritxls, the prorated condensate production per mass of consumed tuh"-cu indicates that sidestream is a less effective smoke generator by a factor ut ,,Ix,ut 2.5. This also indicates that gaseous combustion products are more prumtncnt in sidestream than in mainstream smoke. There is a sufficient body of c•ldcnt;c to support the existence of differences in the composition of the gaseous ui' sitiestream and mainstream smoke (see Table I). However, no cunsir tcnt ;ulnpu,itiun protiles of the dry condensates of sidestream and mainstream wwhc are a,•ailable to contirm the differences in composition. With regard to the potential differences in the size distribution and the mean ul'thc smoke particles, the situation in the literature is similar. Papers dealing 31586 ~

0 .t .5 t OIAMETER Nm FIG. 4. Average total Jepusitiun ut' quasi-munudiaperac carnauba wax aerowl+ in 25 vulunteers breathing nuse•in/muuth•uut at 12 re5piratwnh per muiute und I.UUtI cmJ tidal volume according to Giacumelli-Maltuni et ul. (8): bruken line ~,howa theoretical curve uf the Tuak Group model (32). Experimental results on total deposition obtained by Giacomelli-Maltoni et al. (8) in 1972, with quasi-monodisperse hydrophobic carnauba wax particles inhaled by a human test panel at 1,000 cm3 tidal volume and a nose-in/mouth-out breathing frequency of 12 respirations per minute, confirmed the predictions of the Task Group model in this size range. Fipre 4 shows their comparison. The total de- position for 0.5-µm spheres amounted to about 23% and was close to the absolute minimum at 0.4 µ,m. However, a study by Heyder er ul. (9), using monudisperse hydrophobic sebacate droplets ranging in size from 0.2 to 1.0 µm at 500 cm3 tidal volume and a muuth-breathing frequency of 15 respirations per minute, gave considerably lower values for the total deposition in human lungs. Figure 5 shows three individual patterns of these results, which, in spite of the scatter, cunais- tently yielded values of about 10%a total deposition for all investigated sizes be- tween 0.2 and l:0 µm. Fur bigger particles, the total deposition increased, and amounted to 22% for I.9-µm particles. Not all of the differences between the two experimental studies can be ex- plained by the mouth-breathing pattern in the latter case. The authurs noted that variations in the deposition by a factor of 2 could be obtained for all particle sizes investigated, if the breathing level of the test person was not adjusted to the normal functional residual capacity of the lung. In a workshop study in 1977 (10), the two groups of investigators compared their methods and agreed finally on the validity of the lower data under carefully controlled instrumental and physiolog- ic:.if parameters. The previous differences indicate the enormous variability due to individual breathing patterns and the need of standardization for comparison between different studies. Much of this variability has been accounted for by recent improvements in deposition models. For mouth-breathing patterns, a semiempirical model by Heyder et ul. (11) reduces the number of dynamic and aerosol variables by in- troducing characteristic dimensiunless numbers like the aerodynamic deposition parameter. As a unique variable, this parameter determines the total deposition in mouth breathing for sizes above about I µm ot' aerodynamic diameter. The remaining i,ndividual variability observed in this regime is due to the intersubject TI BU 31591 I t rat en rer in: Th bhL 7. mu the wh Pa thL lat

BRONCHIAI ALVEOLAR 0 I I -M1i 1 0 TOTAL J BRONCHIAI AlVEOIAR ~. QOS 0. 1 Q2 0.S 1 2 S 10 20 QOS 0.1 0.2 0.S I 2 5 t0 20 a PARTICIE OIAAETER (ym I b PARtICLE O/AMETER i Nm i Flc. 6. Deposition patterns of nose breathing according to Heyder (13). ta) Slow inspirution over a tung period of ume lti sec per respiration. 250 ctnt szc'I mean tlow rate): (b) Past insptrut/un over a short period of time (4 aec per respiratiun, 750 cm3 sec'' mean tluw rate). not exceed the shallow minimum of deposition where total deposition is deter- mined by alveolar deposition. Thus, all smoke particles should be deposited in the alveolar region and should not exceed a total deposition of some 30%. Both predictions, however, appear to be incorrect. Bronchial deposition is known to occur, and the literature data, though not abundant, refer to smoke depositions of 80% or more (3:, 26), with the exception of results by Porstendort'er and Schraub (27) who quote 15 to 20% without reporting experimental details. Some of the high deposition data involve retention of breath at'ter smoke inha- latiun, but the common explanation for an unaccounted increase in deposition is the hygroscopicity of the smoke particles, which causes them to grow instantly in the high humidity in the human airway. This may further speed up the coag- ulation process and provide larger smoke particles and, thus, higher total depo- sition. However, according to data in the literature. hygroscopic growth of smoke particles merely doubles their size (l5) and simultaneous coagulation increases this growth by another factor of ?('_0), which produces a maximal increase barely in excess of 2 µm. Heyder et ul. 02) reported that total deposition for particles of 3-µm diameter did not exceed 45%, which may indicatz that hygroscopicity cannot account for a total deposition in excess of 5U%. It appears most likely that. electrical charges usually prescnt. on fresh mainstream smoke may account for the efficient deposition. Melandri et ut. (21) reported that electrical forces can cause deposition rates larger than those produced by aerodynamic or diffusive effects. 31593 f a a h: ct st n. ac c:t

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 597 ~LeCrh rig lu 1. -t ul. taled .hing Tits k .1 tle- olute )erse tidal gave hows )n515- :s bz- .and . e zx- .1 that sizes u the ' ( lU). )n the iulug- v due arison :nts in 1el by by in- .)~itiun ~Sitiun :r. The ,ubject a4 0.3 a2 r 0.1 al s ~ ~ ; r; s 0 0 E i I i i I i i I 02 0.0 Ob 08 0.4 L MY *;aaA R t lD #}% I ;R 1 1 1 1 1 F- 0 < J W ~ 0 062 ab Qr1 as to 0.4 0.3 0.2 0.1 , I I I I L_1_J 0 a2 aA 0s ae OWdt7ER Nm L__A 1.0 Ftc. 5. individuat total depositiun patterns of mouth breathing monodisperse aerosuls at 15 respi- ratiuns per minute anJ 500 cm) tidal volume according to Heyder er a!. (9). variations in morphological lung structure as demonstrated by Yu and Diu (37) and lyeates c-t ul. (35). So tar. no approach of this kind has been made for nose breathing, although empirical data under various dynamic conditions have been collected (13). Heyder reports in his study the deposition for two extreme breathing patterns, i.e., slow inapiration over a long period and fast inspiration over•a short period of time. The corresponding total and regional deposition curves for nose breathing are shown in Fig. 6, while corresponding mouth-breathing results are given in Fig. 7. Fur sidestream smoke inhalation, the patterns of Fig. 6 would probably be more appropriate. By comparison, the nose-breathing deposition patterns show the efficient tiltering effect of the extrathoracic or nasopharyngeal compartment. which signiticantly reduces the deposition of the coarser sizes in all other com- partments. For 0.5-µm particles, the total deposition is surprisingly insensitive to the extreme breathing modes. A value of about 25% applies to both cases, but lu,t in,piratiuns cause part of the deposit to occur in the extruthuracic airways. Since active smoking pertains to mouth breathing, the semiempirical model of Heyder rt al. ( 11) should easily predict the total deposition of inhaled cigarette ,muke. As !,hown in Fig. 7. the entire size range of cigarette smoke particles does TI BU 31592

..... •.: te c xt.uta tNt: 13, 589-60I 1 1984) Effect of .tlerl. 79, posure un "ork. 1959. irauchens. 7IIC places. 1972). .co Nmuke Lung Dynamics and Uptake of Smoke Constituents by Nonsmokers-A Survey' WERNER STOBER IF .,,,,,..,rlrr lrtaatuie of Toxicology and Aerosol Research, Hannover, Federal Republic uj Germuny Mrxlcl, ui %muke kinetics and lung dynamics of inhaled particles are discussed and wm• t+wcJ wrth the availuble literature on mainstream and sidestreum smoke particles. The wc++turo sc:uch reveals a dearth of reliable informatiun on the deposition of inhaled par- t.ut.+te tuba.;cu !~moke components in the human lung. Scanty results on mainstream smuke r.nge Irunt unexpectedly high deposits to values in line'with predictions of conventionui Fwtncnwiw:,l Jcpusitiun models confirmed in tes(s with stable aerosols. Fur sidestreum .a.rac. unly one well-described experimental result is avuilable. It is in agreement with s.t.t)lnhcrl Jepunitiun probabilitics. Experimental and theoretical estimates of relative par- Lk Jepu+tuun in the human lung range from wmc lu'~7o for mdestream smukc particles to .wrc thrn Ku':F fur mainstream aerosol. This indicates a need for more, and better. cxptr- wtentrl J4W. c iv14t Ac:uternn; Preaa. Inc. (1k.untbu,tiun of tobacco in pipes, cigars, and cigarettes produces a relatively ,kn.c ,,nd. in most cases, visible smoke. From a physicuchemical point of view, th,. unukc is actually a multicomponent aerosol system which is in the process .4 ..ppru,,;hing a multiphase equilibrium between the vaporized smoke constit- r:,tt. tntacd into the air and a highly dispersed condensate phase made up of tnrny uny droplets of inhalable size. Each droplet contains at least one very small "tet curo uf carbonaceous and/or inorganic combustion residues. The cores serve ...uaJcnsatiun nuclei for the tinal droplet. They are enveloped by a multitude d uiyutic smoke constituents which adhere to the nuclei by adsorption and ..NtJ~n,atiun. These compounds may be products of either pyrolysis or incom- pktc cumbustiun of tobacco ingredients, or they may be original tobacco cum- t.,twnt, distiiled and recondensed near the combustion zone. Due to the various .,44ultuc, ut• the droplet constituents, the ensuing vapor phase is part of a system u)tnb Iu move toward an equilibrium state of stable partial pressures and droplet •+4c,. Lvtdcntly, all processes involving a decrease in the vapor concentration, 4" upcn air smoke dissipation or active dilution by clean air, will also diminish thc Iin..l ,mukc: particle size. Conversely, the addition of compounds occurring ,n tr,th the vapor phase and the condensed particulate state of the smoke will ..Au,c sume growth of the dispersed droplets. It is obvious that, for instance, .atcr r, a condensable combustion product will be a participant in the equili- tw.atng changes of the smoke aerosol system. Thus, tht; humidity of the gas phase *rll Intlucnct: the water content of the smoke droplets. On the other hand. cum- hv.ttun pru+lucts occurring exclusively in the gas phase, such as carbon mun- Nre+cntcJ at the Symposium -Medical Perspectives on Passive Smoking." April y-I:. 1984, Is.cnn.. Au.trta. 589 IW91-7345/H4 53.UQ C..pynght 6. IYril by icaJcmrc Prce.. In.. .VI nghn or repruJw;tuun in .my tnrm rexrvcJ. TI BU 31584

ro )n over .,n over deter- tted in tion is smoke AGrfer !etails. : inha- ttOn is ,tantly : coag- tlepo- ,smukz :reahts barely uticles opicity likely ccuunt .es can tfusive a 19 a6 [L 3 49 0.6 SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 016 0.3 0.9 3.6 a3 0 EXTRATRORACIC BRONCNIAL AlVEOIAR a0 It az 11.5 1 2 5 a PARTICtE OIAAtTER i No I to 20 a9 a6 R3 19 0.6 03 a9 016 0.3 0 EXTRATHORACIC BRONCHIAI AlVEOLAR aof ot nt ns 1 2 5 b PRRTICIF OIAAEIER 1 pA 1 599 to 20 Fw. 7. Depusition patternh. of mouth breathing according to Heyder (13). (a) Slow inspiration over a lung period uf umc ( as -sec per reaplratiun. 250 cros aec'' mean tlow nate): Ib) fust inspiration over a~hurt period of time (a nec per respirutiun. 750 cm-' sec I lleall tluw rate). A theoretical attempt to cope with the problem of unstable aerosol deposition has been made by Austin et ul. (1). but the approach has not yet been applied to cigarette smoke. Therefore. respiratory depositidn rates for cigarette smoke must still be acquired empirically. In view of the dearth of reliable data, a careful experimental attempt is very desirable. Sidestream smoke deposition may be unpredictable for the same reason. Aged and diluted sidestream smoke particles may no longer grow by coagulation, but aspontaneous growth in the high humidity uf the human respiratory tract may increase the deposition beyond theoretical expectation. Experimental evidence to date does not bear this out. Hiller et ul. ( 14) reported an oral inhalation study uf sidestream smoke and found an average total deposition of 11%, which is in perfect agreement with the empirical data shown in Fig. 5 for sebacate aerosol particles of the same size range. lf sidestream'smoke wuuld, indeed, behave like regular stable aerosols, a more appropriate nasal-breathing study may give higher total deposition values, but the intrathoracic deposit would probably be reduced due to the high tilter efficiency of the nasal passages. Figure 6 demonstrates this ettect. To further elucidate the behavior of aged and diluted sidestream smoke, a new study is being contemplated. "i'his would entail a human panel orally inhaling ,tde:htream Nmoke. with exposure concentrations and recovery rates at'ter inha- lation being determined by chemical tlosimetry, and would relate directly to the particulate mass concentration of the sidestream smoke. TI BU 31594

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 597 ~LeCrh rig lu 1. -t ul. taled .hing Tits k .1 tle- olute )erse tidal gave hows )n515- :s bz- .and . e zx- .1 that sizes u the ' ( lU). )n the iulug- v due arison :nts in 1el by by in- .)~itiun ~Sitiun :r. The ,ubject a4 0.3 a2 r 0.1 al s ~ ~ ; r; s 0 0 E i I i i I i i I 02 0.0 Ob 08 0.4 L MY *;aaA R t lD #}% I ;R 1 1 1 1 1 F- 0 < J W ~ 0 062 ab Qr1 as to 0.4 0.3 0.2 0.1 , I I I I L_1_J 0 a2 aA 0s ae OWdt7ER Nm L__A 1.0 Ftc. 5. individuat total depositiun patterns of mouth breathing monodisperse aerosuls at 15 respi- ratiuns per minute anJ 500 cm) tidal volume according to Heyder er a!. (9). variations in morphological lung structure as demonstrated by Yu and Diu (37) and lyeates c-t ul. (35). So tar. no approach of this kind has been made for nose breathing, although empirical data under various dynamic conditions have been collected (13). Heyder reports in his study the deposition for two extreme breathing patterns, i.e., slow inapiration over a long period and fast inspiration over•a short period of time. The corresponding total and regional deposition curves for nose breathing are shown in Fig. 6, while corresponding mouth-breathing results are given in Fig. 7. Fur sidestream smoke inhalation, the patterns of Fig. 6 would probably be more appropriate. By comparison, the nose-breathing deposition patterns show the efficient tiltering effect of the extrathoracic or nasopharyngeal compartment. which signiticantly reduces the deposition of the coarser sizes in all other com- partments. For 0.5-µm particles, the total deposition is surprisingly insensitive to the extreme breathing modes. A value of about 25% applies to both cases, but lu,t in,piratiuns cause part of the deposit to occur in the extruthuracic airways. Since active smoking pertains to mouth breathing, the semiempirical model of Heyder rt al. ( 11) should easily predict the total deposition of inhaled cigarette ,muke. As !,hown in Fig. 7. the entire size range of cigarette smoke particles does TI BU 31592

ro )n over .,n over deter- tted in tion is smoke AGrfer !etails. : inha- ttOn is ,tantly : coag- tlepo- ,smukz :reahts barely uticles opicity likely ccuunt .es can tfusive a 19 a6 [L 3 49 0.6 SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 016 0.3 0.9 3.6 a3 0 EXTRATRORACIC BRONCNIAL AlVEOIAR a0 It az 11.5 1 2 5 a PARTICtE OIAAtTER i No I to 20 a9 a6 R3 19 0.6 03 a9 016 0.3 0 EXTRATHORACIC BRONCHIAI AlVEOLAR aof ot nt ns 1 2 5 b PRRTICIF OIAAEIER 1 pA 1 599 to 20 Fw. 7. Depusition patternh. of mouth breathing according to Heyder (13). (a) Slow inspiration over a lung period uf umc ( as -sec per reaplratiun. 250 cros aec'' mean tlow nate): Ib) fust inspiration over a~hurt period of time (a nec per respirutiun. 750 cm-' sec I lleall tluw rate). A theoretical attempt to cope with the problem of unstable aerosol deposition has been made by Austin et ul. (1). but the approach has not yet been applied to cigarette smoke. Therefore. respiratory depositidn rates for cigarette smoke must still be acquired empirically. In view of the dearth of reliable data, a careful experimental attempt is very desirable. Sidestream smoke deposition may be unpredictable for the same reason. Aged and diluted sidestream smoke particles may no longer grow by coagulation, but aspontaneous growth in the high humidity uf the human respiratory tract may increase the deposition beyond theoretical expectation. Experimental evidence to date does not bear this out. Hiller et ul. ( 14) reported an oral inhalation study uf sidestream smoke and found an average total deposition of 11%, which is in perfect agreement with the empirical data shown in Fig. 5 for sebacate aerosol particles of the same size range. lf sidestream'smoke wuuld, indeed, behave like regular stable aerosols, a more appropriate nasal-breathing study may give higher total deposition values, but the intrathoracic deposit would probably be reduced due to the high tilter efficiency of the nasal passages. Figure 6 demonstrates this ettect. To further elucidate the behavior of aged and diluted sidestream smoke, a new study is being contemplated. "i'his would entail a human panel orally inhaling ,tde:htream Nmoke. with exposure concentrations and recovery rates at'ter inha- lation being determined by chemical tlosimetry, and would relate directly to the particulate mass concentration of the sidestream smoke. TI BU 31594

592 WERNER JTOHEtt with this problem are of varying quality, but there is a trend toward assigning somewhat smaller size values to siclestream smoke particles (16, 19. 24). The fact that sidestream smoke has a bluish tint, in contrast to the more whitish appear- ance of mainstream smoke, may point toward this probability. According to Baker (2), this color difference reflects the fact that the organic vapors in the two smoke streams experience ditferent rates of cooling and mixing with ambient air. Con- szquently, a change in particle composition is indicated. Actual size measurements have been reported by various authors. The results for mainstream particles range from 0.18 µm as count median diameter to 0.70 µm as mass median aerodynamic diameter. Sidestream particles, however, are in a somewhat lower range, from 0.10 to 0.54 µm. Porstendorfer and Schraub (27) found no significant difference in the two streams and reported values uf count medians of 0.10 to 0.12 µm in diameter. The generation rate for sidestream par- ticles was given as 6 x 109 sec' 1, which implies that sidestream smoke is subject to immediate coagulation, as long as ambient air dilution does not diminish its sidestream concentration too rapidly. In view of this, the nominal value of 3.6 x 1012 sidestream particles per cigarette as calculated for smoking one cigarette in 10 min is somewhat artificial, although it closely matches the data of Keith as quoted by Klus and Kuhn (17) and compares to a similarly artificial value uf 1.05 x 1012 particles per cigarette for mainstream smoke generation. With regard to the size distribution of the smpke droplets, investigators have measured or postulated log-normal size distributions and have given geometric standard dcs- viatiuns between 1.3 and 1.6. The current literature does not give any good reliable data confirming the as- sumptiun that differences exist in the chemical composition and physical size Ut primary mainstream and sidestream smoke particles. Aging cigarette smoke in room air will, on one hand, probably wipe out chemical differences by random coagulation, but may also change original composition by losing the more vola- tile components to the vapor phase. This is another field where the lack of ex- perimental data is prominent. Therefore, it is imperative to acquire more empirical data for all types of smoke and to investigate, in particular, the lung deposition of sidestream smoke inhaled from smoke-filled rooms. The suspicion is justified that these deposits are different from deposits of freshly generated mainstream smoke as inhaled by a smoker. In the context of this survey, primary impurtance is given to the actual depu- sition of sidestream smoke aerosul'particles in the human respiratory tract. be- cause this aerosol deposit and its bioavailability to the body by dissolution and resorption will determine the actual dose imparted to "passive smoktrs" as the body burden that may or may not cause a chronic health effect. To assess such burden, it is necessary to recall the physiological mechanics of lung deposition of inhaled particulate matter. The deposition of aerosol particles in the human respiratory tract has been a subject of theoretical and experimental research tur almost 50 years. The first theoretical model was published in 1935 by Findeisen (7) in Germany. !n subsequent years. in particular after World War 11, a variety of authors (3. 18, 36) added many theoretical and semiempirical refinements to TI BU 31587 k !; ~. tita h. it.6ht c .lt,.+a F. wn.~. Ira~t. e~ct) (.m:tui cIcr. M t11i Il .b.itk .kctct i t...! n 1 . upjW1 hrunc aa J.1t Jit:4 . a1J ). httKc N' i t and tt ,uui. i thL: k:.& JclxI. In I Utl 12.1 tnq~c,t the hc ut hui N,itc tn .kl- o( tlte modci t.4l • c%tcn.A tu'oo tI tht hrunc t cpithc. p.~r t lL: t~t rc~pir. l~lbi rc.3cnt

Itra- ~ the is of Inge ated :tive r the Aes . . . ~ora ~)art- tyer. 0 lose :om- uali- SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOK(NG 595 tative nature, but it is apparent that the human lung is an efficient filter, not only for particles exceeding 5 µm of aerodynamic diameter but also for very small particles below aerodynamic sizes of about 0. 1 µm. Neither mainstream nor sidestream smoke particles really extend into these high tilter efficiency ranges. For mainstream smoke particles, it is justitied to assume a log-normal size dis- tribution with a mass median aerodynamic diameter of, at most, 0.5 µm and a geometric standard deviation of 1.5 (31). This indicates that there are practically no particles exceeding a size of 2 µm, and very few particles below sizes of 0. 1 µm. Thus, all mainstream smoke particles are essentially within the range where the human respiratory tract deposits only a certain fraction of the inhaled aerosol. The same must be true for sidestream smoke, although a smaller value for the median size may slightly increase the relative deposition for the smallest sizes. However, this "tail" of the size distribution is rather insensitive in relation to the total mass deposit. The Task Group model permits a crude estimate of the regional deposition if the mass median aerodynamic diameter is known and the geometric standard deviation is between 1.2 and 4.5 (not a very restrictive requirement). The shaded areas in Fig. 3 indicate the corresponding variabilities of the deposition in the three compartments of the model. Since cigarette smoke will comply with the restraints for the size distribution, the deposition can be assessed according to the graph. The typical mass median aerodynamic diameter of 0.5 µ.m should cause, at most, about I t% nose-breathing deposition in the nasopharyngeal re- bion, while a deposit of between I and 6% may be expected for the tracheobron- chial compartment. However, the most signiticant deposition occurs in the pul- monary region and should range between 22 and 32%. In view of the actual geometric standard deviation of smoke particles, the lower deposition tigures seem to be more applicable. MASS MEOIAN OIAMETER-MICRONS F+a. 3. Depoeitwn ot polydiaparse aoruwts in the human respiratury tract accurding to Tusk Group ,;11(%;UtilpUnb (3 2). Lower llmltJn~,' curve: Geometric ltAnl(Lrlt dCY/BUOn uf IUg-nUrmat )Ize distribution 1.2 (qua.i-munudi,perne rerosul). Upper limiting curve: Geumetric ~tandard deviation uf log-normal etLc distribution 4.5 (very polydiopersa aerosol). TI BU 31590

suits 0.70 .re in ('_7) uunt par- bject .h its .6 x ae In thas .ie of •gard :d or d de- te as- ze of ke in tdum vola- )t' ex- 4rical -,ition titied seam .lc:pu- t. be- a and is the !wch ,itiun uman :h fur ieistn .iriety ats tu 0 SYMPUSIUM: MEDICAL PERSPECTIVES ON PASSiVE SMOKING 593 the basic model. However, for particles with simple physical characteristics, like stable droplet spheres, the general features of the model were not essentially changed. Lt was the process of checking the more sophisticated theoretical ap- proaches against experimental evidence during the last 15 years that provided a considerable body of reliable data on aerosol deposition in the human respiratory tract. Modern, experimentally confirmed theoretical models account for almost every cdnceivable factor that may have an impact on aerosol deposition. These factors may be divided into two basically different groups: (a) biological param- eters (e.g., morphological features of the lung and human breathing patterns, which need standardization) and (b) aerosol parameters (e.g., particle sizes and ~hapes. size distribution, and airborne concentration, which need experimental dctermination). Biological parameters, as commonly incorporated in mathemat- ical models of deposition, consist of morphological data on the structure of the upper airways and the lungs; the length, diameter, and number of generations of bronchi; and the number and features of bronchiolar and alveolar spaces, as well as data on body position and motion. Furthermore, they comprise physiological data on lung dynamics, i.e., breathing frequencies and profiles, tidal volumes, and vital and residual capacities of the lung. Finally, the models discriminate between mouth and nose breathing. With regard to the aerosol parameters. the models account for the aerodynamic and the thermodynamic size of the particies, their size distribution, their density, and, in the case of irregular particles and agglomerates, their shape factors or, in the case of elongated particfes (fibers or tlakes), their aspect ratio. For absolute' deposition values< the actual airborne concentration of the aerosol particles is needed. In 1966, the Task Group on Lung Dynamics of the International Commission on Radiological Protection (32) summarized the state of the art,at that time. The investigation was initiated because of the need for a conservative assessment of the health hazard associated with the deposition and retention of inhaled particles in human airways. Although the conservative protective purpose uf this com- posite model may overestimate deposition in critical regions, it is still a landmark in deposition modeling. The authors introduced a formal compartmentalization ut' the respiratory tract which, since then, has been retained in all subsequent models. The three compartments are: (a) the nasopharyngeal compartment. which begins at the anterior nares and extends beyond the anterior and posterior pharynx to the epiglottis (also referred to as the extrathoracic compartment); (b) the tracheobronchial compartment, which comprises the trachea and the bronchial tree ipcluding the terminal bronchioles; it is characterized by the ciliated epithelium and mucus production, which provide an effective lung clearance for particles; and (c:) the pulmonary or alveolar compartment, which consists of the nonciliated respiratory bronchioles, the atria. the alveoli, and the alveolar sacs. Figure t shows a schematic diagram of the pulmonary compartment. lt rep- resents the inflatable part of the lung, wherein the gas exchange between the body TI BU 31588

BRONCHIAI ALVEOLAR 0 I I -M1i 1 0 TOTAL J BRONCHIAI AlVEOIAR ~. QOS 0. 1 Q2 0.S 1 2 S 10 20 QOS 0.1 0.2 0.S I 2 5 t0 20 a PARTICIE OIAAETER (ym I b PARtICLE O/AMETER i Nm i Flc. 6. Deposition patterns of nose breathing according to Heyder (13). ta) Slow inspirution over a tung period of ume lti sec per respiration. 250 ctnt szc'I mean tlow rate): (b) Past insptrut/un over a short period of time (4 aec per respiratiun, 750 cm3 sec'' mean tluw rate). not exceed the shallow minimum of deposition where total deposition is deter- mined by alveolar deposition. Thus, all smoke particles should be deposited in the alveolar region and should not exceed a total deposition of some 30%. Both predictions, however, appear to be incorrect. Bronchial deposition is known to occur, and the literature data, though not abundant, refer to smoke depositions of 80% or more (3:, 26), with the exception of results by Porstendort'er and Schraub (27) who quote 15 to 20% without reporting experimental details. Some of the high deposition data involve retention of breath at'ter smoke inha- latiun, but the common explanation for an unaccounted increase in deposition is the hygroscopicity of the smoke particles, which causes them to grow instantly in the high humidity in the human airway. This may further speed up the coag- ulation process and provide larger smoke particles and, thus, higher total depo- sition. However, according to data in the literature. hygroscopic growth of smoke particles merely doubles their size (l5) and simultaneous coagulation increases this growth by another factor of ?('_0), which produces a maximal increase barely in excess of 2 µm. Heyder et ul. 02) reported that total deposition for particles of 3-µm diameter did not exceed 45%, which may indicatz that hygroscopicity cannot account for a total deposition in excess of 5U%. It appears most likely that. electrical charges usually prescnt. on fresh mainstream smoke may account for the efficient deposition. Melandri et ut. (21) reported that electrical forces can cause deposition rates larger than those produced by aerodynamic or diffusive effects. 31593 f a a h: ct st n. ac c:t

suits 0.70 .re in ('_7) uunt par- bject .h its .6 x ae In thas .ie of •gard :d or d de- te as- ze of ke in tdum vola- )t' ex- 4rical -,ition titied seam .lc:pu- t. be- a and is the !wch ,itiun uman :h fur ieistn .iriety ats tu 0 SYMPUSIUM: MEDICAL PERSPECTIVES ON PASSiVE SMOKING 593 the basic model. However, for particles with simple physical characteristics, like stable droplet spheres, the general features of the model were not essentially changed. Lt was the process of checking the more sophisticated theoretical ap- proaches against experimental evidence during the last 15 years that provided a considerable body of reliable data on aerosol deposition in the human respiratory tract. Modern, experimentally confirmed theoretical models account for almost every cdnceivable factor that may have an impact on aerosol deposition. These factors may be divided into two basically different groups: (a) biological param- eters (e.g., morphological features of the lung and human breathing patterns, which need standardization) and (b) aerosol parameters (e.g., particle sizes and ~hapes. size distribution, and airborne concentration, which need experimental dctermination). Biological parameters, as commonly incorporated in mathemat- ical models of deposition, consist of morphological data on the structure of the upper airways and the lungs; the length, diameter, and number of generations of bronchi; and the number and features of bronchiolar and alveolar spaces, as well as data on body position and motion. Furthermore, they comprise physiological data on lung dynamics, i.e., breathing frequencies and profiles, tidal volumes, and vital and residual capacities of the lung. Finally, the models discriminate between mouth and nose breathing. With regard to the aerosol parameters. the models account for the aerodynamic and the thermodynamic size of the particies, their size distribution, their density, and, in the case of irregular particles and agglomerates, their shape factors or, in the case of elongated particfes (fibers or tlakes), their aspect ratio. For absolute' deposition values< the actual airborne concentration of the aerosol particles is needed. In 1966, the Task Group on Lung Dynamics of the International Commission on Radiological Protection (32) summarized the state of the art,at that time. The investigation was initiated because of the need for a conservative assessment of the health hazard associated with the deposition and retention of inhaled particles in human airways. Although the conservative protective purpose uf this com- posite model may overestimate deposition in critical regions, it is still a landmark in deposition modeling. The authors introduced a formal compartmentalization ut' the respiratory tract which, since then, has been retained in all subsequent models. The three compartments are: (a) the nasopharyngeal compartment. which begins at the anterior nares and extends beyond the anterior and posterior pharynx to the epiglottis (also referred to as the extrathoracic compartment); (b) the tracheobronchial compartment, which comprises the trachea and the bronchial tree ipcluding the terminal bronchioles; it is characterized by the ciliated epithelium and mucus production, which provide an effective lung clearance for particles; and (c:) the pulmonary or alveolar compartment, which consists of the nonciliated respiratory bronchioles, the atria. the alveoli, and the alveolar sacs. Figure t shows a schematic diagram of the pulmonary compartment. lt rep- resents the inflatable part of the lung, wherein the gas exchange between the body TI BU 31588

i n i- r. 1- in th- ige na for .led 1.6, , of iere :spi- the : po- SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 0 605 0.10 0.zo 0.30 1 0 AEROOYNAMC DIAMETEF ( umJ Fic. I. The size distribution of sidestream smoke prior to and after inhalation is shown. The distri- butions are similar. suggesting little differential deposition by size. [By permission. American Review „/ Rrspvatury DiseLue 125, 406 (1982).] sition for particles in that size range is lower than for particles in any other size range. Particles in this size range have insufficient mass for inertial impaction, have a low settling velocity so that sedimentation is minimal, and diffuse slowly o so that deposition is also minimal. The size distribution of the exhaled aerosol does not appear to reflect any hygroScopic growth that might have occurred in the human respiratory tract. Since these studies were not designed to maintain respiratory tract humidity during measurements of exhaled sidestream smoke, water acquiredln the respiratory tract could be lost in the measurement process. The deposition fraction of 11% for sidestream cigarette smoke measured during mouth breathing is relatively low. However, the measured deposition fraction is cunsistent with previous predictions and measurements of deposition for particles in the 0.5-µm size range. The deposition of mainstream cigarette smoke measured by other investigators is seven to nine times higher than the deposition of side- stream smoke, even though particle size measurements for these two types of smoke are generally regarded to be similar. , It is likely that the deposition of mainstream smoke is altered by the smoking a process. The smoker often takes a puff into the-mouth, where it is held brietly before further inhalation to the lower respiratory tract. Particle growth in the mouth by coagulation and/or water accumulation could change the size of the particles considerably so that deposition would be higher. The very high concen- tratiun, IOs-10y particles/cm3 for mainstream smoke, is compatible with rapid cougulation. Other manipulations of respiratory patterns, such as breathholding by the smoker after inhalation, also account, in part, for the higher deposition of mainstream smoke. The present studies were done using oral inhalation. Total_deposition might be somewhat higher for nasal inhalation because of particle removal by the upper respiratory tract. However, predicted deposition in the nose of particles in the ,tie range of cigarette smoke is relatively low; thus, that deposition fraction is nut likely to change greatly (24). TI BU 31600

600 WERNER STOBER REFERENCES 24. 1. Austin, E.. Brock. J., and Wissler, E. A model for deposition of stable and unstable aerosols in the human respiratory tract. Amer. /nd. Hvg. Ae4uc. J. 40, 1U55-1066 (1979). '5. 2. Baker, R. R. Mechanism of smoke formation and delivery. Recent Adv. Tubucco Sci. 6, 184-224 (1980). 16. 3. Beeckmans. J. M. The deposition of asbestos particles in the human respiratory tract. lnt. J. I~nvirun. Stud. 1. 31-34 1197U/. 4. Brown. J. H., Cook. K. M.. Ney. F G., and Hatch. T. tnftuence of particle size upon ~he retention 17. of particulate matter in the human lung. Amer. J. Public Heulth 40. 450-458 (1950). 5. Brunnemann, K. U., Adams. J. D.. Hu, D. P. S., and Hot3'mann. D. The influence of tobacco smoke on indoor atmospheres. !l. Volatile and tobacco specific nitrusamines in main and side stream smoke and their contribution to indoor pollution. "Proceedings 4th Joipt Conference I a. 29. on Sensing of Environmental Pollutants, New Orleans, Louisiana (1977)," pp. 876-880. Amer- ican Chemical Society, 1978. 6. Dennis, W. L. Prepared discussion, in "lnhaled Particles and Vapuurs" (C. N. Davies. Ed.), pp. 30. 88-91. Pergamon Press, Oxford. 1961. 7. Findeisen, W. Uber das Absetzen kleiner. in der Luft suspendierter Teilchen in der menschlichen 31. Lunge bei der Attnung. PjTueger's Arch. 236, 367-379 ( I935). 8. Giacomelli-Maltuni. G.. Melandri. C.. Prodi. V., and Tarruni. G. Deposition efficiency of mono- 32. disperse particles in human respiratory tract. Ainer. Ind. Hy);. Assur. J. 33, 603-610 (1972). 9. Heyder, J.. Gebhurt, J., Heigwer. G.. Roth. C., and Stahlhufen. W. Experimental studies of the 33. total deposition of aerosol particles in the human respiratory tract. J. Aerua•ul Sci. 4, 191-208 11973). 34. 10. Heyder. J., Gebhart. J.. Roth. C.. Stuhlhofen, W., Stuck. B.. Tarruni. G., DeZuiacumu. T.. Fur- mignani, M.. Nlelandri. C., and Piudi. V. Intercumparisun of lung deposition data for aerosol 35. particles. J. Aerusul Sci. 9, 147-I55 (1978). tl. Heyder. J., Gebhart. J.. Rudolf. G., and Stahlhufen, W. Physical factors determining particle 36. deposition in the human respiratory tract. J. Arrusul Sci. 11, 505-515 (198U). 12. Heyder, J., Gebhart, J., Stahlhufen, W., and Stuck. B. Biological variability of particle deposition 37. in the human respiratory tract during controlled and spontaneous mouth breathing. .4nn. Occup. Hvg. 26, 137-147 (1y82). 13. Heyder. J. Studies of particle deposition and clearance in humans, in "Problems of Inhalatury Toxicity Studies," (P. Grosdanul'f et (1., eds.). BGA-Sehrtft.by MMV-Medizin-Verlag, Munchen. 1985. 14. Hiiler, F. C.. McCusker. K. T.. Mazumder. M. K.. Wilson. J. D.. and Bone. R. C. Deposition uf sidrstream cigarette smoke in the human respiratory tract. .tiner. Rev. Respir. Uis. 125. 4U6- 4UtS 1 1y2S2). 15. khizu. Y., Ohta. K., and Okada. T. The effect ut' moisture on the growth of cigarette tmuke particles. Beitr. Tubul:J'ursch. 10, 1hI-Ihl! (1980). 16. Keith. C. H.. and Derrick. J. C. Measurement ut' the particle size distribution and concentration of cigarette smoke by the "cunifuge." J. Culluid Sri. 15, 340-356 1196U1. 17. Klus, H., and Kuhn. H. Verteilung verschiedener Tabakrauchbestandteile auf Haupt- und Ne- benstromrauch (Eine Ubersicht). Britr. TubukJi,rsch. Inrerncu. 11,':9-265 (1982). 18. Landahl. H. D. On the removal of air-borne droplets by the human respiratory tract. 1. The lung. Bull. Nuth. Bruph,vs. 12. 43-56 11950). 19. McCusker, K., Killer, C.. Wilson. U.. :Mazumder. M.. and Bune, R. Characterizatiun of side- stream smoke from low tar cigarettes and cigars. Arrter. Rev. Respir. Uis. 123, 107 ( lytll). 20. McCusker, K.. Hiller, C.. Macumder. M., and Bone R. Dynamic growth of cigarette ~muke particles (Abstracu. Chest 80. 349 11981). 21. Melandri. C.. Tarruni- G.. Prodi. V., de Zaiacumu. T., Furmignani. M., and Lumbardi. C. C. Depusittun uf charged particles in the human airways, J. Aerusul Sci. 14, bS7-699 (1983). 22. :Kltchell, K. I. Controlled measurement of smUke particle retention in the resplratl)rytrdct. .3ntCr: Rev. Respir. Un. 85, 526-533 (1962). 23. Ntursth. G.. and Ehmke. H. Apparatur Lur Untersuchung des Ntbenstrumrauches. Beur. fu- hu,lJiiryc h. 2, 117- l21 t l y6d). TI BU 31595

i n i- r. 1- in th- ige na for .led 1.6, , of iere :spi- the : po- SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 0 605 0.10 0.zo 0.30 1 0 AEROOYNAMC DIAMETEF ( umJ Fic. I. The size distribution of sidestream smoke prior to and after inhalation is shown. The distri- butions are similar. suggesting little differential deposition by size. [By permission. American Review „/ Rrspvatury DiseLue 125, 406 (1982).] sition for particles in that size range is lower than for particles in any other size range. Particles in this size range have insufficient mass for inertial impaction, have a low settling velocity so that sedimentation is minimal, and diffuse slowly o so that deposition is also minimal. The size distribution of the exhaled aerosol does not appear to reflect any hygroScopic growth that might have occurred in the human respiratory tract. Since these studies were not designed to maintain respiratory tract humidity during measurements of exhaled sidestream smoke, water acquiredln the respiratory tract could be lost in the measurement process. The deposition fraction of 11% for sidestream cigarette smoke measured during mouth breathing is relatively low. However, the measured deposition fraction is cunsistent with previous predictions and measurements of deposition for particles in the 0.5-µm size range. The deposition of mainstream cigarette smoke measured by other investigators is seven to nine times higher than the deposition of side- stream smoke, even though particle size measurements for these two types of smoke are generally regarded to be similar. , It is likely that the deposition of mainstream smoke is altered by the smoking a process. The smoker often takes a puff into the-mouth, where it is held brietly before further inhalation to the lower respiratory tract. Particle growth in the mouth by coagulation and/or water accumulation could change the size of the particles considerably so that deposition would be higher. The very high concen- tratiun, IOs-10y particles/cm3 for mainstream smoke, is compatible with rapid cougulation. Other manipulations of respiratory patterns, such as breathholding by the smoker after inhalation, also account, in part, for the higher deposition of mainstream smoke. The present studies were done using oral inhalation. Total_deposition might be somewhat higher for nasal inhalation because of particle removal by the upper respiratory tract. However, predicted deposition in the nose of particles in the ,tie range of cigarette smoke is relatively low; thus, that deposition fraction is nut likely to change greatly (24). TI BU 31600

604 F. CHARLES tIILLER We have measured the deposition of a multimodal awrosol made up of 0.60-, 1.101-, and 2.01-µm polystyrene latex spheres. The deposition fractions for these sizes were, respectively, 0.08 *_- 0.009, 0.21 z 0.03, and 0.45 ; 0.05 (SE). These measurements were similar to those obtained by other investigators using mono- disperse aerosols of similar sizes (14, 19). For these studies of deposition of sidestream smoke, five healthy male volun- teers were studied. Each subject pertormed three separate studies, each initiated from functional residual capacity. All studies were done with mouth breathing using a 1-liter tidal volume and a respiratory rate of 12. The respiratory pattern was controlled by measuring the volume change in the chamber with a differential pressure pneumotachygraph. The signal from the pneumotachygraph was dis- played on a dual-channel oscilloscope along with a standard prerecorded respi- ratory pattern which the subject followed. With a little practice, all subjects were able to maintain the required 1-liter volume, 12 breaths/min, rate. Sidestream smoke was drawn into the chamber through the inhalation port from the burning tip of a cigarette held near the mouthpiece. Concentration was ad- justed to provide a particle counting rate appropriate for the SPART analyzer, i.e., 10 to 20 particlesl,ec. This rate could be achieved with approximately 50- 100 µg/m3 chamber concentration. After the chamber was loaded, fve consecu- tive 2-min samples were taken to detine the particle concentration decay rate in the chamber. The subject was then seated at the chamber and allowed to practice the respi- ratory pattern. After practicing the subject began to inhale aerosol from the chamber and exhale into a waste bag for 10 breaths. The subsequent 4 breaths were then exhaled into the sample collection bag. After 4 breaths were collected, analysis of the sample bag was initiated. The concentration in the sample collection bag was compared with the chamber smoke concentration at the time of inhulation. The deposition fraction was cal- culated using calibration factors to correct for the difference between the mea- sured concentration in the inhaled and exhaled aerosol (7, 8). RESULTS The mean deposition fraction of sidestream cigarette smoke for mouth- breathing subjects measured in this experiment was 0.11 = 0.04 (SE). The range of measurements was from 0.07 ± 0.06 to 0.20 z 0.05. The size distribution of the inhaled and exhaled aerosols was determined in a sample study (Fig. 1). The count and mass median aerodynamic diameters for the inhaled aerosols were, respectively, 0.32 and 0.41 µm and, for the exhaled aerosol, 0.31 and U.42 µm. Geometric standard deviations were 1.5 and 1.6, respectively. There was thus little, if any, difference in the size distributions of these two aerosols. DISCUSSION The differences in the size iiistributions of the inhaled and exhaled smoke were small, suggesting that there was minimal selective size deposition in the respi- ratory tract. This finding is not surprising, since there is little variation in the predicted deposition for aerosols in the size range 0.2-0.5 µm. The total depo- TI BU 31599 Fit butit. „j R, ranL hav, su t duc the resF watt T mou coM in th by ~ Strc: amu It proc bc:fui muuI partit tratit Lui1L' by U maic Tl sOnlt re"'p Si« nut i

604 F. CHARLES tIILLER We have measured the deposition of a multimodal awrosol made up of 0.60-, 1.101-, and 2.01-µm polystyrene latex spheres. The deposition fractions for these sizes were, respectively, 0.08 *_- 0.009, 0.21 z 0.03, and 0.45 ; 0.05 (SE). These measurements were similar to those obtained by other investigators using mono- disperse aerosols of similar sizes (14, 19). For these studies of deposition of sidestream smoke, five healthy male volun- teers were studied. Each subject pertormed three separate studies, each initiated from functional residual capacity. All studies were done with mouth breathing using a 1-liter tidal volume and a respiratory rate of 12. The respiratory pattern was controlled by measuring the volume change in the chamber with a differential pressure pneumotachygraph. The signal from the pneumotachygraph was dis- played on a dual-channel oscilloscope along with a standard prerecorded respi- ratory pattern which the subject followed. With a little practice, all subjects were able to maintain the required 1-liter volume, 12 breaths/min, rate. Sidestream smoke was drawn into the chamber through the inhalation port from the burning tip of a cigarette held near the mouthpiece. Concentration was ad- justed to provide a particle counting rate appropriate for the SPART analyzer, i.e., 10 to 20 particlesl,ec. This rate could be achieved with approximately 50- 100 µg/m3 chamber concentration. After the chamber was loaded, fve consecu- tive 2-min samples were taken to detine the particle concentration decay rate in the chamber. The subject was then seated at the chamber and allowed to practice the respi- ratory pattern. After practicing the subject began to inhale aerosol from the chamber and exhale into a waste bag for 10 breaths. The subsequent 4 breaths were then exhaled into the sample collection bag. After 4 breaths were collected, analysis of the sample bag was initiated. The concentration in the sample collection bag was compared with the chamber smoke concentration at the time of inhulation. The deposition fraction was cal- culated using calibration factors to correct for the difference between the mea- sured concentration in the inhaled and exhaled aerosol (7, 8). RESULTS The mean deposition fraction of sidestream cigarette smoke for mouth- breathing subjects measured in this experiment was 0.11 = 0.04 (SE). The range of measurements was from 0.07 ± 0.06 to 0.20 z 0.05. The size distribution of the inhaled and exhaled aerosols was determined in a sample study (Fig. 1). The count and mass median aerodynamic diameters for the inhaled aerosols were, respectively, 0.32 and 0.41 µm and, for the exhaled aerosol, 0.31 and U.42 µm. Geometric standard deviations were 1.5 and 1.6, respectively. There was thus little, if any, difference in the size distributions of these two aerosols. DISCUSSION The differences in the size iiistributions of the inhaled and exhaled smoke were small, suggesting that there was minimal selective size deposition in the respi- ratory tract. This finding is not surprising, since there is little variation in the predicted deposition for aerosols in the size range 0.2-0.5 µm. The total depo- TI BU 31599 Fit butit. „j R, ranL hav, su t duc the resF watt T mou coM in th by ~ Strc: amu It proc bc:fui muuI partit tratit Lui1L' by U maic Tl sOnlt re"'p Si« nut i

600 WERNER STOBER REFERENCES 24. 1. Austin, E.. Brock. J., and Wissler, E. A model for deposition of stable and unstable aerosols in the human respiratory tract. Amer. /nd. Hvg. Ae4uc. J. 40, 1U55-1066 (1979). '5. 2. Baker, R. R. Mechanism of smoke formation and delivery. Recent Adv. Tubucco Sci. 6, 184-224 (1980). 16. 3. Beeckmans. J. M. The deposition of asbestos particles in the human respiratory tract. lnt. J. I~nvirun. Stud. 1. 31-34 1197U/. 4. Brown. J. H., Cook. K. M.. Ney. F G., and Hatch. T. tnftuence of particle size upon ~he retention 17. of particulate matter in the human lung. Amer. J. Public Heulth 40. 450-458 (1950). 5. Brunnemann, K. U., Adams. J. D.. Hu, D. P. S., and Hot3'mann. D. The influence of tobacco smoke on indoor atmospheres. !l. Volatile and tobacco specific nitrusamines in main and side stream smoke and their contribution to indoor pollution. "Proceedings 4th Joipt Conference I a. 29. on Sensing of Environmental Pollutants, New Orleans, Louisiana (1977)," pp. 876-880. Amer- ican Chemical Society, 1978. 6. Dennis, W. L. Prepared discussion, in "lnhaled Particles and Vapuurs" (C. N. Davies. Ed.), pp. 30. 88-91. Pergamon Press, Oxford. 1961. 7. Findeisen, W. Uber das Absetzen kleiner. in der Luft suspendierter Teilchen in der menschlichen 31. Lunge bei der Attnung. PjTueger's Arch. 236, 367-379 ( I935). 8. Giacomelli-Maltuni. G.. Melandri. C.. Prodi. V., and Tarruni. G. Deposition efficiency of mono- 32. disperse particles in human respiratory tract. Ainer. Ind. Hy);. Assur. J. 33, 603-610 (1972). 9. Heyder, J.. Gebhurt, J., Heigwer. G.. Roth. C., and Stahlhufen. W. Experimental studies of the 33. total deposition of aerosol particles in the human respiratory tract. J. Aerua•ul Sci. 4, 191-208 11973). 34. 10. Heyder. J., Gebhart. J.. Roth. C.. Stuhlhofen, W., Stuck. B.. Tarruni. G., DeZuiacumu. T.. Fur- mignani, M.. Nlelandri. C., and Piudi. V. Intercumparisun of lung deposition data for aerosol 35. particles. J. Aerusul Sci. 9, 147-I55 (1978). tl. Heyder. J., Gebhart. J.. Rudolf. G., and Stahlhufen, W. Physical factors determining particle 36. deposition in the human respiratory tract. J. Arrusul Sci. 11, 505-515 (198U). 12. Heyder, J., Gebhart, J., Stahlhufen, W., and Stuck. B. Biological variability of particle deposition 37. in the human respiratory tract during controlled and spontaneous mouth breathing. .4nn. Occup. Hvg. 26, 137-147 (1y82). 13. Heyder. J. Studies of particle deposition and clearance in humans, in "Problems of Inhalatury Toxicity Studies," (P. Grosdanul'f et (1., eds.). BGA-Sehrtft.by MMV-Medizin-Verlag, Munchen. 1985. 14. Hiiler, F. C.. McCusker. K. T.. Mazumder. M. K.. Wilson. J. D.. and Bone. R. C. Deposition uf sidrstream cigarette smoke in the human respiratory tract. .tiner. Rev. Respir. Uis. 125. 4U6- 4UtS 1 1y2S2). 15. khizu. Y., Ohta. K., and Okada. T. The effect ut' moisture on the growth of cigarette tmuke particles. Beitr. Tubul:J'ursch. 10, 1hI-Ihl! (1980). 16. Keith. C. H.. and Derrick. J. C. Measurement ut' the particle size distribution and concentration of cigarette smoke by the "cunifuge." J. Culluid Sri. 15, 340-356 1196U1. 17. Klus, H., and Kuhn. H. Verteilung verschiedener Tabakrauchbestandteile auf Haupt- und Ne- benstromrauch (Eine Ubersicht). Britr. TubukJi,rsch. Inrerncu. 11,':9-265 (1982). 18. Landahl. H. D. On the removal of air-borne droplets by the human respiratory tract. 1. The lung. Bull. Nuth. Bruph,vs. 12. 43-56 11950). 19. McCusker, K., Killer, C.. Wilson. U.. :Mazumder. M.. and Bune, R. Characterizatiun of side- stream smoke from low tar cigarettes and cigars. Arrter. Rev. Respir. Uis. 123, 107 ( lytll). 20. McCusker, K.. Hiller, C.. Macumder. M., and Bone R. Dynamic growth of cigarette ~muke particles (Abstracu. Chest 80. 349 11981). 21. Melandri. C.. Tarruni- G.. Prodi. V., de Zaiacumu. T., Furmignani. M., and Lumbardi. C. C. Depusittun uf charged particles in the human airways, J. Aerusul Sci. 14, bS7-699 (1983). 22. :Kltchell, K. I. Controlled measurement of smUke particle retention in the resplratl)rytrdct. .3ntCr: Rev. Respir. Un. 85, 526-533 (1962). 23. Ntursth. G.. and Ehmke. H. Apparatur Lur Untersuchung des Ntbenstrumrauches. Beur. fu- hu,lJiiryc h. 2, 117- l21 t l y6d). TI BU 31595

TABLE : NICOTINE ANU COTLNINE LEVEL.S" IN SALIVA UF VOLUNTEERS E:(1'OSEU TO SLUESTNE.iM SMOKE Time (min) 2 Cigarettts" 3 Cigarettes Nicotine Cotinine Nicotine Cutinine 8 1.2 1 L7 372 7.0 505 U.3 347 0.3 712 2.5 427 0.8 837 5.0 386 2.3 993 2.5 76 2.3 157 1.5 26 1.0 46 3.3 13 1.5 26 2.3 6 1.5 17 2.8 3 2.5 9 2.3 13 3.3 14 2.0 5 1.5 t2 3.5 8 5.0 2 l.3 6 3.3 4 2.0 7 1.0 7 1.0 Note. E = Exposure to Nidestream smoke in 16-m•` chamber. ° Values given as ng/ml. ° Number of cigarettes being smoked throughout the exposure period. 4 Ciburettes T Nicotine Cutinine 1 3 1.0 Ba 458 1.3 E. 825 t.1 E- 878 ?.1 E 730 1.4 E~ 148 1.7 ~ 49 1.4 n 31 2.7 v 23 2.5 12 17 3.1 ti 24 2.8 ti+ 6 2.0 21 3 1.9 24 6 2.3 ,- 7 3.5 3U TABLE 3 NICOTLNE ANU CU'1'ININE LEVELS" IN PLASMA UF VOLUNTEERS EXNUSLiU TO SLDESTKLtAM SAtOKE Time 2 Cigarettes° 3 Cigarettes (min) Nicotine Cuunine Nicotinc Cuunine - BaaeiinC 1.1 1.7 ND LO E'_U 0.2 2.2 1 O.ii E 40 ND 1.5 2.1 U.19 E 60 N D - 1.3 4.2 0.7 E KU 1.1 1.6 1.3 1.1 +30 0.2 1.8 2.7 I.6 60 U.'_ 1.9 0.6 3.1 90 0.6 2.3 2.9 1.9 t_u 1.5 1.7 0.3 2.1 150 0.2 1.4 0.7 2.6 1250 N D 2.6 U.= 3.0 210 N D 1.8 0.7 :.0 :41) N D 2.1 0.2 1.9 ?7U 0.9 ?.1 U.2 2.4 300 ND - ND 2.5 Vute. E = Exposure to sLdestream omuke in 16-m'. chamber. ° Values given us ng/ml. ° Number of cigarettes being hmuked throughout the exposure period. N D = Not dctectelJ. 4 Cigarettes Niwune Cutinine 0.2 0.9 ND t.= 0.3 0.9 t1.3 1.2 0.5 1.3 0.4 1.6 r (1.8 2.1 0.6 i 2.6 s 1.4 2.9 1 ; 0.7 :.y 1.0 3.3 0.2 3.3 1.1 3.3 U.h 3.4 ll.b 3,'_ I TI BU 31605 L r' P' tt L.

TABLE : NICOTINE ANU COTLNINE LEVEL.S" IN SALIVA UF VOLUNTEERS E:(1'OSEU TO SLUESTNE.iM SMOKE Time (min) 2 Cigarettts" 3 Cigarettes Nicotine Cotinine Nicotine Cutinine 8 1.2 1 L7 372 7.0 505 U.3 347 0.3 712 2.5 427 0.8 837 5.0 386 2.3 993 2.5 76 2.3 157 1.5 26 1.0 46 3.3 13 1.5 26 2.3 6 1.5 17 2.8 3 2.5 9 2.3 13 3.3 14 2.0 5 1.5 t2 3.5 8 5.0 2 l.3 6 3.3 4 2.0 7 1.0 7 1.0 Note. E = Exposure to Nidestream smoke in 16-m•` chamber. ° Values given as ng/ml. ° Number of cigarettes being smoked throughout the exposure period. 4 Ciburettes T Nicotine Cutinine 1 3 1.0 Ba 458 1.3 E. 825 t.1 E- 878 ?.1 E 730 1.4 E~ 148 1.7 ~ 49 1.4 n 31 2.7 v 23 2.5 12 17 3.1 ti 24 2.8 ti+ 6 2.0 21 3 1.9 24 6 2.3 ,- 7 3.5 3U TABLE 3 NICOTLNE ANU CU'1'ININE LEVELS" IN PLASMA UF VOLUNTEERS EXNUSLiU TO SLDESTKLtAM SAtOKE Time 2 Cigarettes° 3 Cigarettes (min) Nicotine Cuunine Nicotinc Cuunine - BaaeiinC 1.1 1.7 ND LO E'_U 0.2 2.2 1 O.ii E 40 ND 1.5 2.1 U.19 E 60 N D - 1.3 4.2 0.7 E KU 1.1 1.6 1.3 1.1 +30 0.2 1.8 2.7 I.6 60 U.'_ 1.9 0.6 3.1 90 0.6 2.3 2.9 1.9 t_u 1.5 1.7 0.3 2.1 150 0.2 1.4 0.7 2.6 1250 N D 2.6 U.= 3.0 210 N D 1.8 0.7 :.0 :41) N D 2.1 0.2 1.9 ?7U 0.9 ?.1 U.2 2.4 300 ND - ND 2.5 Vute. E = Exposure to sLdestream omuke in 16-m'. chamber. ° Values given us ng/ml. ° Number of cigarettes being hmuked throughout the exposure period. N D = Not dctectelJ. 4 Cigarettes Niwune Cutinine 0.2 0.9 ND t.= 0.3 0.9 t1.3 1.2 0.5 1.3 0.4 1.6 r (1.8 2.1 0.6 i 2.6 s 1.4 2.9 1 ; 0.7 :.y 1.0 3.3 0.2 3.3 1.1 3.3 U.h 3.4 ll.b 3,'_ I TI BU 31605 L r' P' tt L.

TABLE I ESTIMATED SMOKE INTAKE ANO MASS DEPOSITION DUtt1NG PASSIVE SMOKING Deposition° lntake" (mgrday) Investigator 0.01-U.2 CE/hr 0.22-4.4 Hoegg, 1972 (12) 0.001-0.009 CFJhr 0.014-0.13 Hinds and First, 1975 (l0) 0.09 CE/hr 2.0 Hugod et al. 1978 (12a) 5-27 CE/day 0.3-1.6 Repace and Lowrey, 1980 (23) - 0.26 Hiller et al. 1982 (8) ° As calculated by author in cigarette equivalents. ° Approximate correction to mass per day. The human respiratory tract deposition of sidestream cigarette smake can be calculated using our measurment of deposition fraction and the measured con- centration of cigarette smoke in any environmc:nt. Mass deposition calculated on the basis of an 11% deposition fraction and the smoke intakes reported by several investigators are shown in Table 1. The calculations are not fully comparable, since the methodology employed was different in each study. Hinds (9. 10) esti- mated intake on the basis of nicotine r,ecovered in air samples, while Repace and Lowrey (23) determined total particulate mass, which included nonspecific dusts as well as tobacco smoke. Hugod (12a) estimated cigarette equivalents based on a 25-mg tar cigarette, Hoegg (12) used a 25-mg sidestream smoke measurement, Hinds used a 16. 1-mg tar cigarette, and Repace and Lowrey used a 0.55-mg low- tar cigarette. There is somewhat less disparity in the data when all tigures are reduced to mass inhaled or mass deposited over the same time interval. The estimation of mass deposition seems relatively low compared with the quantity consumed by cigarette smokers. The elevated content of some toxic constituents of sidestream smoke, however, may make it more potent than main- stream smoke. The seemingly low mass deposition estimated in this and other studies does not justify a conclusion that sidestream smoke is not a health hazard. REFERENCES I. Bridge. D. P., and Corn, M. Contribution to the assessment of exposure of nonsmokers to wr poilution from cigarette and cigar smoke in occupied spaces. Environ. Res. 5, 192-209 (1972). 2. Brunnemann, K. D., Yu. L., and Hoffmann. D. Assessment of carcinogenic volatile N-nitrosu- mines in tobacco and in mainstream and sidestream smoke from cigarettes. Cuncer Res. 37. 3218-3222) (1977). 3. Curn, M. Characteristics of tobacco sidestream smoke and factors intluencing its concentration and distribution in occupied spaces, i,t "Environmental Tobacco Smoke Effects on the Non- Smoker" (R. Rylander, Ed.). p. 21. Schmiclt, Vojens. 1974. 4. Dalhamn. T., Edfors. M.. and Rylander. R. Retention of cigarette smoke components in human lungs. Arch. Environ. Health 17, 746-71it (1968). 5. Davies. C. N., Heyder. J., and Subba Ramu. M. C. Breathing of half-micron aerosols. J. Appl. Physiul. 32. 591-60U (1972). 6. Hammond, E. C., Horn. D. Smoking and death r.ues-Repurt on forty-four months of fotlowup of 187,783 men. JAAlA 166, 1 15y-1 17_, 1294-13U8 (195h). 7. Hiller. F. C., Mucumder. M. K.. wilson, J. D.. McLeod. P. C.• and Bone. R. C. Human re~pi- ratory tract deposition using multimodal atrusuls. J. Aerusol Sci. 13. 337-343 ( tytS_). ii. Hiller. F. C., NicCusker• K. T., Macumder. M. K., Wilson. J. D., and Bone, R. C. Depubitiun of TI BU 31601 NiLlc 4U2f l0. Hinth. Nr ll. Hiray.. trut 12. Hoeg,__ 12a. Hu6L CUn 13. KeithI uf . 14. Lippr; Ph % W.I 15. ytatu. Rai si~,l 16. McC:: of (19 17. Mitch Rr ili. Muir. Sa 19. Mun. .i l. i 20. Ok:cu t ~m ~ -l. Poly4

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 603 irt uke t be r•.u:- thr tiun iube ,n u[ ut ~Iomz Lriety of „ant dif- nstream various content Experi- inbtream- areas varies from approximately 20 to 60 µg/m3 (23). Levels for respirable sus- pended particulates (RSP). ineasured indoors in the presence ut' smokers, gen- erally varies from about 100 to 600 µg/m3 (23). These measurements (23) include all particles, not only tobacco smoke. The accepted air quality standard averages fur total suspended particulates (TSP) are 75 µg/m3 annually and ?60 µglm3 per 24 hr in the United States, and 250 µglm3 per 24 hr in the United Kingdom. While the measured level of indoor cigarette smoke is often high and can exceed standards for suspended particulates, estimatiuit of respiratory tract dose is dif- hcult. A 70% deposition for sidestream smoke inhaled indoors has been assumed (3). By one estimate, the nonsmoker inhales in I hr the equivalent of smoking 0.01- 0.?0 cigarette (12). By another estimate, a person who works outdoors in a res- idential environment and lives in a tobacco smoke-free home environtnent is estimated to inhale 365 mg RSP annually. A person who works in a crowded, poorly ventilated nightclub 40 hr per week, 50 weeks per year, and who lives with heavy smokers, would inhale approximately 5,475 mg per year (23). The respiratory tract deposition of inhaled particles can be predicted using mathematical lung models and has been measured in a number of laboratories. In addition, the particle size of cigarette smoke has been found to be between 0.2 and 0.5 µm (9, 13, 16,'0, 21, 22). The deposition fraction for particles in that size range has been predicted to be 0.1-0.3 (26). Experimental studies generally aupport those predictions (5. 7, 8, 14, 18). ' There is a discrepancy between the predicted and measured deposition of l0- ?0% for particles in the size range 0.2-0.5 µm and the measured deposition of 70-90% for cigarette smoke, also thought to be in the same size range (9, 13, 16, 20). Several explanations for this discrepancy have been offered and include hy- groscopic growth of cigarette smoke particles during inhalation, coagulation of cigarette smoke during inhalation, and differences in respiratory pattern during cigarette smoking and passive smoking. The purpose of our work was to measure the deposition of sidestream cigarette i h i n the uman resp ratury tract. -latively t smoke snts. and ym} Was METtiODS rimental The inhalation and analysis system included a''40-liter chamber, a micropro- nass was cessur-controiled valve system, and a single particle aerodynamic relaxation time ns of 140 (SPART) analyzer (IS). Sidestream smoke was introduced into the chamber, and I of 50 to the subject inhaled by mouth from the chamber. of tiigniti- Smoke inhaled from the chamber was exhaled into collection bags inside the lz waiting iave been iuke cun- Ilecteti on > ,muking chamber. Deposttton was measured by comparing the size dtstrtbutton and con- ccntration of the inhaled smoke with that of the exhaled aerosol collected in a sample bag. • The exhaled smoke concentration was obtained by collecting the exhaled ,,crusul in the sample bag. Particle concentration in the sample bag was corrected for wall losses in the connecting tubing, diffusion, and sedimentation. These cor- rection factors were determined mathematically, and measured correction factors have been -,huwn to be very close to predicted correction factors. The details of j th"e Ntudiea have been described previously (7, 8). TI BU 31598

TABLE I ESTIMATED SMOKE INTAKE ANO MASS DEPOSITION DUtt1NG PASSIVE SMOKING Deposition° lntake" (mgrday) Investigator 0.01-U.2 CE/hr 0.22-4.4 Hoegg, 1972 (12) 0.001-0.009 CFJhr 0.014-0.13 Hinds and First, 1975 (l0) 0.09 CE/hr 2.0 Hugod et al. 1978 (12a) 5-27 CE/day 0.3-1.6 Repace and Lowrey, 1980 (23) - 0.26 Hiller et al. 1982 (8) ° As calculated by author in cigarette equivalents. ° Approximate correction to mass per day. The human respiratory tract deposition of sidestream cigarette smake can be calculated using our measurment of deposition fraction and the measured con- centration of cigarette smoke in any environmc:nt. Mass deposition calculated on the basis of an 11% deposition fraction and the smoke intakes reported by several investigators are shown in Table 1. The calculations are not fully comparable, since the methodology employed was different in each study. Hinds (9. 10) esti- mated intake on the basis of nicotine r,ecovered in air samples, while Repace and Lowrey (23) determined total particulate mass, which included nonspecific dusts as well as tobacco smoke. Hugod (12a) estimated cigarette equivalents based on a 25-mg tar cigarette, Hoegg (12) used a 25-mg sidestream smoke measurement, Hinds used a 16. 1-mg tar cigarette, and Repace and Lowrey used a 0.55-mg low- tar cigarette. There is somewhat less disparity in the data when all tigures are reduced to mass inhaled or mass deposited over the same time interval. The estimation of mass deposition seems relatively low compared with the quantity consumed by cigarette smokers. The elevated content of some toxic constituents of sidestream smoke, however, may make it more potent than main- stream smoke. The seemingly low mass deposition estimated in this and other studies does not justify a conclusion that sidestream smoke is not a health hazard. REFERENCES I. Bridge. D. P., and Corn, M. Contribution to the assessment of exposure of nonsmokers to wr poilution from cigarette and cigar smoke in occupied spaces. Environ. Res. 5, 192-209 (1972). 2. Brunnemann, K. D., Yu. L., and Hoffmann. D. Assessment of carcinogenic volatile N-nitrosu- mines in tobacco and in mainstream and sidestream smoke from cigarettes. Cuncer Res. 37. 3218-3222) (1977). 3. Curn, M. Characteristics of tobacco sidestream smoke and factors intluencing its concentration and distribution in occupied spaces, i,t "Environmental Tobacco Smoke Effects on the Non- Smoker" (R. Rylander, Ed.). p. 21. Schmiclt, Vojens. 1974. 4. Dalhamn. T., Edfors. M.. and Rylander. R. Retention of cigarette smoke components in human lungs. Arch. Environ. Health 17, 746-71it (1968). 5. Davies. C. N., Heyder. J., and Subba Ramu. M. C. Breathing of half-micron aerosols. J. Appl. Physiul. 32. 591-60U (1972). 6. Hammond, E. C., Horn. D. Smoking and death r.ues-Repurt on forty-four months of fotlowup of 187,783 men. JAAlA 166, 1 15y-1 17_, 1294-13U8 (195h). 7. Hiller. F. C., Mucumder. M. K.. wilson, J. D.. McLeod. P. C.• and Bone. R. C. Human re~pi- ratory tract deposition using multimodal atrusuls. J. Aerusol Sci. 13. 337-343 ( tytS_). ii. Hiller. F. C., NicCusker• K. T., Macumder. M. K., Wilson. J. D., and Bone, R. C. Depubitiun of TI BU 31601 NiLlc 4U2f l0. Hinth. Nr ll. Hiray.. trut 12. Hoeg,__ 12a. Hu6L CUn 13. KeithI uf . 14. Lippr; Ph % W.I 15. ytatu. Rai si~,l 16. McC:: of (19 17. Mitch Rr ili. Muir. Sa 19. Mun. .i l. i 20. Ok:cu t ~m ~ -l. Poly4

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 603 irt uke t be r•.u:- thr tiun iube ,n u[ ut ~Iomz Lriety of „ant dif- nstream various content Experi- inbtream- areas varies from approximately 20 to 60 µg/m3 (23). Levels for respirable sus- pended particulates (RSP). ineasured indoors in the presence ut' smokers, gen- erally varies from about 100 to 600 µg/m3 (23). These measurements (23) include all particles, not only tobacco smoke. The accepted air quality standard averages fur total suspended particulates (TSP) are 75 µg/m3 annually and ?60 µglm3 per 24 hr in the United States, and 250 µglm3 per 24 hr in the United Kingdom. While the measured level of indoor cigarette smoke is often high and can exceed standards for suspended particulates, estimatiuit of respiratory tract dose is dif- hcult. A 70% deposition for sidestream smoke inhaled indoors has been assumed (3). By one estimate, the nonsmoker inhales in I hr the equivalent of smoking 0.01- 0.?0 cigarette (12). By another estimate, a person who works outdoors in a res- idential environment and lives in a tobacco smoke-free home environtnent is estimated to inhale 365 mg RSP annually. A person who works in a crowded, poorly ventilated nightclub 40 hr per week, 50 weeks per year, and who lives with heavy smokers, would inhale approximately 5,475 mg per year (23). The respiratory tract deposition of inhaled particles can be predicted using mathematical lung models and has been measured in a number of laboratories. In addition, the particle size of cigarette smoke has been found to be between 0.2 and 0.5 µm (9, 13, 16,'0, 21, 22). The deposition fraction for particles in that size range has been predicted to be 0.1-0.3 (26). Experimental studies generally aupport those predictions (5. 7, 8, 14, 18). ' There is a discrepancy between the predicted and measured deposition of l0- ?0% for particles in the size range 0.2-0.5 µm and the measured deposition of 70-90% for cigarette smoke, also thought to be in the same size range (9, 13, 16, 20). Several explanations for this discrepancy have been offered and include hy- groscopic growth of cigarette smoke particles during inhalation, coagulation of cigarette smoke during inhalation, and differences in respiratory pattern during cigarette smoking and passive smoking. The purpose of our work was to measure the deposition of sidestream cigarette i h i n the uman resp ratury tract. -latively t smoke snts. and ym} Was METtiODS rimental The inhalation and analysis system included a''40-liter chamber, a micropro- nass was cessur-controiled valve system, and a single particle aerodynamic relaxation time ns of 140 (SPART) analyzer (IS). Sidestream smoke was introduced into the chamber, and I of 50 to the subject inhaled by mouth from the chamber. of tiigniti- Smoke inhaled from the chamber was exhaled into collection bags inside the lz waiting iave been iuke cun- Ilecteti on > ,muking chamber. Deposttton was measured by comparing the size dtstrtbutton and con- ccntration of the inhaled smoke with that of the exhaled aerosol collected in a sample bag. • The exhaled smoke concentration was obtained by collecting the exhaled ,,crusul in the sample bag. Particle concentration in the sample bag was corrected for wall losses in the connecting tubing, diffusion, and sedimentation. These cor- rection factors were determined mathematically, and measured correction factors have been -,huwn to be very close to predicted correction factors. The details of j th"e Ntudiea have been described previously (7, 8). TI BU 31598

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 609 an ive ars m- ,td :he are :he 1C5 las 'lIe ine .izs CtJ smoke ported umors 0 ~ przg- ontirrn irprise t weak ;roups >assive alogies .)bacco and to )a5ure. ) apply nd the ve not '_, 1984. significantly changed. While in many countries the sales-weighted average nicotine and carbon monoxide yields of the mainstream smoke of cigarettes have greatly decreased over the last two decades (7), cigarette sidestream smoke yields have remained stable (11). The uniqueness of these conditions has prompted us to initiate an in-depth research program on the uptake of sidestream smoke by nonsmokers and on the possible endogenous formation of carcinogens upon in- halation of sidestream smoke from tobacco products. A MODEL STUDY ON SIDESTREAM SMOKE EXPOSURE For the study on uptake of sidestream smoke by nonsmokers we utilized a ,mall, bare room of 16 m3 which was constantly polluted by the sidestream smoke of four reference cigarettes concurrently smoked by machine. The mainstream smoke of these cigarettes was directed outside the room. Six air exchanges per hour were made corresponding to the average ventilation conditions of offices in the United States. Within 10-15 min of smoking, the room reached the stable pollution levels presented in Table I. Before the cigarette smoking was begun, the nonsmoking volunteers were seated in the room, two at a time; an indwelling catheter inserted into the ante- cubital vein permitted blood sampling before, during, and after exposure. Dupli- cate samples were drawn at f0-min intervals up to the 80-min exposure and at larger intervals thereatter. up to 5 hr. Tables 2-4 show the analytical protiles of markers for exposures in saliva, plasma, and urine of the nonsmoking volunteers who spent 80 min in the smoke- polluted room. Thiocyanate in saliva, serum, and urine and carboxyhemoglobin in blood were not significantly elevated in the volunteers after exposure to side- stream smoke. lmportant observations lie in the facts that nicotine derived from sidestream smoke barely increased serum nicotine levels, whereas its metabolite, cutinine, was significantly elevated in the serum 2-3 hr after the first exposure to sidestream smoke. In saliva, nicotine rapidly rose to levels above gtx) ng/ml, but subsided just as rapidly when the volunteers left the exposure area. Nicotine TABLE I TEST LAtioKArokY Sice: 16.3 mi Temperature: 22 s I°C Air exchanges: Six times per hour Pollutants: Sidestream smoke of four concurrently smoked 1RI reference cigarettes Indoor pullutiun: Particulate mattCr 4,6W µglmJ Nicutine '_2SU µyml Hydrogen cyanide 56 µg/ml Carbon munuxide ,5 ppnt NO, 0.91 ppm Formaldehyde 160 µg/mj TI BU 31604

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 601 tobacco lnd "ide iference I. Amer- 'd.1. pp. ,ahlichen ,f mono- 11197_l. zs of the lyt -'_U8 T.. For- r aerusul ; {7.irticle cposuion ng. Atut. nhalatory n-Verlag. . w,itiun uf 125. 41J6- :ee ~muke ,:etttnuion unct Ne- The lung. ,n of eicle- I tNt1l). tte ~mukc :nli. C. C. (1983). Britr. Tu- '-a. Okada. T.. Ishizu. Y.. ;,nd Matsunuma. K. Determination of particle mie distribution and cun- centratiun of cigarette smoke by a light hcattcring method. Britr. Tubuk/urs(-h. 9, 153-16U (1y77). 25. Patriakanos. C.. and Hott'mann. D. Chemical stuJies on tobacco smoke LXIV: On the analysis uf arumatic amines in cigtuette smuke. J. Anu1. Tuxicul. 3, 15U-I54 11979). 26. Polydorova. M. An attempt to determine the retention uf tobucco smoke by means of membrane titters, in "lnhaleil Particles and Vapuurs" (C. N. Davtes. Ed.), pp. 142-144. Pergamun Press. Uxfurd. 1961. _7. Porstentlorfer. J.. and Schraub, A. Konzentration und mittlere Teilchengro8e des Haupt- und Nebenstromrauches der Zigarette. Srutth-Rrinh. Ltli 32. 4I)9-412 t 1972). 28. Pyrilii. C. Polycyclische Kuhlenwasscrstotfe im Zigarettenrauch. Mittriltucgsbl. GdCh-Fuclt- gruppe Lebrnswtittelchrnt. Gerichr. Chrtn. 1442). 27 (1960). 19. Pyrtki, C.. 4luller. R.. and Muldenhauer. W. Uber das Auftreten von polycyclischen Kuhlen- wussrrstuttcn tm Zigarettenrauch. tl. Mitteiiung: Untersuchung des lipuphtlen Anteils der ein- zelnen Rauchphasen suwie des Tabaks. Brr. lnw. 1'uhulr/iir.ic•!t. Dresden 7, 81-102 (1960). 30. Pyriki. C. Pulycyclische und aliphattsche Kuhtenw;userstolfe des labakrauchs. ,Vti1u•urtg 7, 439- 448 (19631. 31. Stuber, W. Generation. size distribution and composition of tubacco smoke aerosols. Recent Adt•. T6harcu Sci. 8, 3-4t (1982). 32. Task Group on Lung Dynamics. Deposition and retention models for internal dosimetry of the human respiratory tract. !leu/r!t P/trs. 12, 173-'-U8 (1966). 33. van Wijk. A. M.. and Patterson. H. S. The percentuge of particles of different sites removed from dust-laden air by breathing. J. lnd. Hy,q. Tu.ricul. 22. 31-35 (194U). 34. Wynder. E. L., and HulFinann. D. "Tobaccu and Tobacco Smoke. Studies in Experimental Car- ~;inugenesih," p. 230 ff. Academic Press. New Yo(kiLundun. 1967. 35. Yeutes. D. B.. Gerrity. T. R.. and Garrard, C. S. Charucteristics of tracheubronchial deposition and clearuncc in man. Anui. (h cup. HYg. 26. :59-272 (lylS2). 36. Yu. C. P.. Nicoluides, P., and Soong. T. T. Effect of random airway sites on aerosol deposition. ,aiaer. Ind. Hyg. Assuc•. J. 40, N99-I0U5 (1979). 37. Yu. C. P., anJ Diu. C. K. A probabilistic model fur intersubject deposition variability of inhaled particles. Aerosol Sci. Trclwul. 1, 353-362 (1982).

PREVENTIVE MEDtCINE 13, 608-617 (1984) Tobacco Sidestream Smoke: Uptake by Nonsrnokers'-2 DIETRICH HOFFMANN,3 NANCY J. HALEY, JOHN D. ADAMS, AND KLAUS D. BRUNNEMANN Nuylur Dana Institute for Diseuse Prevention, Atnerrcan Health Foundation. Valltalla. New York I0S95 Some epidemiological studies indicate an association between passive smoking and an increased risk for cancer, especially fur cancer of the lung. Other reports, however, have failed to contirm these findings. Biochemical analyses of the physiological tluids for markers of exposure to tobacco smoke are needed us measurements of the uptake of smoke com- ponents by nonsmokers and fur the estimation of relative cancer risk to passively exposed persons compared with that to active cigarette smokers. This communication reports the uptake of carbon monoxide, hydrogen cyanide. and nicotine after passive smoke exposure under controlled conditions. The results indicate that salivary nicotine values reflect the level of recent passive smoke exposure within an hour and that urinary cutinine values indicate the level of passive smoke exposure in the preceding hours. N-Nitrusopruline has been shown to serve as an indicator uf endogenous N-nitrosumine formation in cigarette smokers: yet. preliminary studies do not indicate that urinary excretion of N-nitrusuproline is increased following short-term passive smoke exposure. In infants, first field studies suggest a correlation between exposure to tobaccu-smoke-pulluted environment> and levels of wtinine in both serum and urine. o rvn. ,k:wemtc Press, Inc. INTRODUCTION Epidemiological studies have indicated an association between passive smoke exposure and lung cancer (1, 4, 5, 15). Weak associations have also been reported for passive smoke exposure and nasal sinus cancer and for childhood brain tumors in the offspring of women passively exposed to environmental smoke during preg- nancy (5, 13). However, a number of other investigations have failed to confirm such associations (2, 8, 9). This lack of full accord Jocs nut come as a surprise because of the many sources of bias and confounders inherent in cases of weak associations (14). Although additional studies with different population groups may lead to a bettel• understanding of the weak association between passive smoke exposure and cancer, the development of new biochemical methodologies enables us today to obtain more detinitive measurements of exposure to tobacco smoke by determining the uptake of specitic components in body tluids and to calculate the risk factors relative to those inherent in active smoke exposure. Furthermore, the results of the biochemical measurements taken today also apply to exposures in the past; the uptake of passive smoke by nonsmokers and the physicochemical nature and yield of sidestream smoke per cigarette have not ' Presented at the Symposium "Medical Perspectives on Passive Smoking." April 9-12, 1984, Vienna. Austria. = Supported by NIH Grants POI-CA-_y580 and ROl-CA35607. t To whom requests for reprints should be addressad. OWt-7435r84 S3.Utl CuPYnrAt C 19" hY AuJemm Press. Inc. Art nyhla ut reprwWaiun in any Wrm rexrvad. 608 TI BU 31603 t signit. nicouz greatl: havt r to init nonan halaut Fo t- small. of fou. smukL: hour v the Gr pullutt Bet', seateti cubita. CatZ s: larger Tab: plasm: puilut- in biu., streani Nidestr cutinin to sitic but .su.

PREVENTIVI: HEDICINE l3, 6(12-6U7 (1Yti41 Deposition of Sidestream Cigarette Smoke in the Human Respiratory Tract' F. CHARLES HILLER Department of Medicine. University uf Arkunaas f)r.Nedicu( Sciences, 4301 West ,tilart;hum, Little Ruck. Arkunsus 72207 The intake of sidestream cigarette smoke can be calculated if ventilation and smoke concentration are known. However, respiratory tract dose of smoke particul'ates cannot be determined unless the deposition fraction of inhaled smoke is known. The deposition fr.lc• tion of mainstream smoke is 7U-9llya, far higher than expected for most particles in the size range. 0.2-0.5 µm. which is also the range for cigarette smoke. The deposition fraction of sidestream smoke was measured in tive normal male subjects. The deposition. the ltose of smoke particulates to the respiratory tract can be estimated. The deposition fraction uf sidestreum smoke is much lower than for mainstream smoke. v Ivs+ r,;.,taroc Pre». lau. INTRODUCTION The health effects of sidestrean't cigarette smuke are controversial, but some epidemiolugical studies suggest sidestream•smuke intake can cause a variety of problems such as airway abnormalities (27) and lung cancer (1 i). Signiticant dif- Terences in the content of toxic constituents between sidestream and mainstream smoke have been described (2). Quantitative estimation of the intake of various cigarette smoke constituents requires knowledge not only of the chemical content of the particles, but also ot' the deposition fraction following inhalation. Experi- ments in the early 1960s gave deposition values of 70-90% during mainstream smoking (4, 17, 21). The concentration of sidestream cigarette smoke varies widely, from relatively low levels in churches to high levels in eating and drinking establishments, and very high levels in laboratory settings. A total particulate mass of 16.5 m1~,1m3 was measured 18 min after 24 cigarettes were smoked in a 25-m3 experimental chamber. After 4 cigarettes in the same chamber, the total particulate mass was slightly greater than 2 mg/m3 (121. The smoke concentrations for rooms of 140 m3 have been estimated at 2,0W-4,000 µg/ms based on an assumption of 50 to 73 persons present during a party; these exposures are assumed to be of signiti- cance (1). Levels of from 20 to 480 µg/m3 have been measuretd in bus and airline waiting rooms and on commuter trains and buses. Levels in cucktail iuungta have been measured as high as 640 µg/m3 (10). These tigures represent tobacco smoke cun- centrations, since measurements were made on [he basis uf nicotine cuilectzd un filters .(9). The level of respirable suspended particles measured in no smoking I Presented at the Symposium "MGdicul Perspectives on Passive Smuking," April 9-12, 1984, Vienna. Austria. 0091-7435/84 53.00 (:uqynµpl s'. IWW by At:;wleM; Pre». lno;. .all nlpqa ul repruJucuun in .ny lurm reierved. 602 TI BU 31597 area: penu erall all p: for t 24 h w stan, ticul (3). 1 0.20 icldn est[ti pu(N hea% Tf matt Ina, 0.2 . size 1upF TI _0114 7U-' _0). bros c:iga Ctga T? ~Imu T1 cea, (SP.- thC St chun cent 1i1n1; T aerl tur ret;.; havr the:a

PREVENTIVI: HEDICINE l3, 6(12-6U7 (1Yti41 Deposition of Sidestream Cigarette Smoke in the Human Respiratory Tract' F. CHARLES HILLER Department of Medicine. University uf Arkunaas f)r.Nedicu( Sciences, 4301 West ,tilart;hum, Little Ruck. Arkunsus 72207 The intake of sidestream cigarette smoke can be calculated if ventilation and smoke concentration are known. However, respiratory tract dose of smoke particul'ates cannot be determined unless the deposition fraction of inhaled smoke is known. The deposition fr.lc• tion of mainstream smoke is 7U-9llya, far higher than expected for most particles in the size range. 0.2-0.5 µm. which is also the range for cigarette smoke. The deposition fraction of sidestream smoke was measured in tive normal male subjects. The deposition. the ltose of smoke particulates to the respiratory tract can be estimated. The deposition fraction uf sidestreum smoke is much lower than for mainstream smoke. v Ivs+ r,;.,taroc Pre». lau. INTRODUCTION The health effects of sidestrean't cigarette smuke are controversial, but some epidemiolugical studies suggest sidestream•smuke intake can cause a variety of problems such as airway abnormalities (27) and lung cancer (1 i). Signiticant dif- Terences in the content of toxic constituents between sidestream and mainstream smoke have been described (2). Quantitative estimation of the intake of various cigarette smoke constituents requires knowledge not only of the chemical content of the particles, but also ot' the deposition fraction following inhalation. Experi- ments in the early 1960s gave deposition values of 70-90% during mainstream smoking (4, 17, 21). The concentration of sidestream cigarette smoke varies widely, from relatively low levels in churches to high levels in eating and drinking establishments, and very high levels in laboratory settings. A total particulate mass of 16.5 m1~,1m3 was measured 18 min after 24 cigarettes were smoked in a 25-m3 experimental chamber. After 4 cigarettes in the same chamber, the total particulate mass was slightly greater than 2 mg/m3 (121. The smoke concentrations for rooms of 140 m3 have been estimated at 2,0W-4,000 µg/ms based on an assumption of 50 to 73 persons present during a party; these exposures are assumed to be of signiti- cance (1). Levels of from 20 to 480 µg/m3 have been measuretd in bus and airline waiting rooms and on commuter trains and buses. Levels in cucktail iuungta have been measured as high as 640 µg/m3 (10). These tigures represent tobacco smoke cun- centrations, since measurements were made on [he basis uf nicotine cuilectzd un filters .(9). The level of respirable suspended particles measured in no smoking I Presented at the Symposium "MGdicul Perspectives on Passive Smuking," April 9-12, 1984, Vienna. Austria. 0091-7435/84 53.00 (:uqynµpl s'. IWW by At:;wleM; Pre». lno;. .all nlpqa ul repruJucuun in .ny lurm reierved. 602 TI BU 31597 area: penu erall all p: for t 24 h w stan, ticul (3). 1 0.20 icldn est[ti pu(N hea% Tf matt Ina, 0.2 . size 1upF TI _0114 7U-' _0). bros c:iga Ctga T? ~Imu T1 cea, (SP.- thC St chun cent 1i1n1; T aerl tur ret;.; havr the:a

612 HOFFMANN ET AL. These assays on the uptake of sidestream smoke by nonsmokers under con- trolled conditions are being repeated. We will also extend the determination of nicotine and cotinine in the urine to include measurements of their levels 10, 17, and 24 hr postexposure. Furthermore, it needs to be determined how the con- trolled conditions will affect the values for cotinine in serum and in urine and how these can be optimized for measurement. Upon establishing these baseline data under controlled environmental conditions, we intend to initiate field studies. These will consist of measurements of nicotine in saliva as an indicator of very recent passive smoke exposure and nicotine and cotinine in serum and urine as indicators of chronic sidestream smoke exposure. We are confident that these measurements will reflect the average and high uptake of sidestream smoke com- ponents by nonsmokers and, upon comparisuns with active cigarette smokers, that we will be able to assess relative risk factors. ENDOGENOUS FORMATION OF N-NITROSAMINES Calculations with respect to risk factors should also include an estimation of the potential of inhaled sidestream smoke to enhance the Cndugenous formation of carcinogenic N-nitrosamines. Recently, we reported on the endogenous formation of nitrosamines in cigarette smokers (6). That study was based on the observation ut' Ohshima and Bartsch that the urinary excretion of nitrosuproline (NPRO) can serve as an indicator for the endogenous formation of carcinogenic nitrosamines (12). NPRO is neither mutagenic nor carcinogenic and is excreted in urine in unmetabolized form; in contrast, most carcinogenic N-nitrosamines are,quickly metabolized in the tnam- malian system. This applies also to the carcinogenic tobacco-specific nitrosamines which yield some of the same urinary metabolites as are derived from nicotine and other Nicotiana alkaloids. Fit TABLE 5 + NICOTINE AND COTININE LEVELS IN SALIVA, StNUM. ANU UNINE Ot' VOLUNTEEH5" • (SUMMANY OF AVE1lAUE V.iLUtS) R t Saliva (ng/mU Serum tngrmi) 1 Urine ing/mg creatinine) t Nicotine Nicotine Countne Ti ° Cotinitie Cutinine mr tmin) 2 3 4 4 4 ~ 3 4 2 3 4 Baseline 8 1 3 1.0 0.9 24 20 17 14 14 14 1 40 350 719 830 1.1 0.9 t 60 430 K4U tltsu _.1 1.2 26 34 84 16 '_t '_S 0 30 76 157 148 t.7 1.8 0 120 6 17 '3 2.5 2.9 40 94 100 21 34 46 0 240 8 2 3 _.0 3.3 0 300 7 7 7 3.5 3.4 51 58 48 21 38 55 ° Numbers reprencnt room pollution by smuke uF 2. 3, or 4 cigarettes. b I= Inside exposure room during pulluuun: 0 = outside exposure ruum atter leaving the room. TI BU 31607 F un : on uf ~ tU~ turn bivc mc:n calt amt uth: l rc: i:. turi tn t

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 607 zan be d cort- tted on ieveral arable, 0) esti- :ce and .; dusts :5zd on emznt. i!; low- res are ith the z toxic , i main- J other "1:1zaCd. _rs to Lur 19 (197_). J-nitroea- Res•. 37, cntnuiun the Nun- sidestream cigarette smoke in the human respiratory tract. Anter. Rev. Respir. Dis. 125, 406- 408 (1982). 9. Hinds, W. C. Size characteristics of cigarette Smoke. .iiner. lnd. Hyg. Assuc. J. 39, a8-54 (1978). 10. Hinds, W. C.. and First. M. W. Concentrations of nicotine and tobacco smoke in public places. New Engl. J. Med. 292, 844-845 (1975). 11. Hirayama, T. Nonsmoking wives of heavy smokers have a higher risk of lung cancer: A study from Japan. Brit. a4ed. l. 282, 183-185 (1981). 12. Hoegg, U. R. Cigarette ~.moke in closed spaces. Environ. Health Perspect. 2, 117-128 (1972). 12a. Hugod. C., Hawkins. L. H.. and Astrup, P. Exposure of passive smokers to tobacco smoke constituents. Int. Arck. Occup. Envirun. Health 42, 21-29 (19781. 13. Keith. C. H., and Derrick. J. C. Measurement of the particle size distribution and concentration of cigarette smoke by the "Conifuge." J. Colloid Sci. 15, 340-356 ( l9Wl. 14. Lippmann, M. Regional deposition of particles in the human respiratory tract, in "Handbook of Physiology," Section 9: "Reuctions to Environmental Agents" (S. R. Geiger, Ed.), p. 213. Waveriy Press, kialtimore, 1977. 15. Mazumder. M. K.. Ware, R. E.. Wilson, 1. D.. Renninger, R. G., Hiller. F C.. McCleod. P. C., Raible, R. W., and Testerman. M. K. SPART Analyzer: Its application to aerodynamic size distribution measurement. J. Aerusol Sci. 10, 561-569 (1979). 16. MeCusker, K.. Hiller, F. C., Wilson. J. D., Mazumder. M. K., and Bone. R. Aerodynamic sizing of tobacco smoke particulate from commercial cigarettes. Arch. Ertvirun. Heulth 38, 215-218 (1983). 17. lvtitchell, R. 1. Controlled measurement of smoke-particle retention in the respiratory tract. ,iiner. Rev. Respir. Dis. 85, 526-533 (1962). 18. Muir. D. C. F. Tobacco smoke inhalation. in "Environmental Tobacco Smoke Effects on the Non- Smoker" (R. Ryiander,° Ed.), p. 44. Schmidt. Vujens, 1974. 19. Muir. D. C. F., and Davies, C. N. The deposition ot' 0.5 µm diameter aerosuls in the lungs of man. `[nn. uc•cup. Hyg. 10, 161-174 (1967). 20. Okada. T., and Matsunuma. K. Determination of particle size and concentration of cigarette Nmuke by a light scattering method. J. Culluid lnterJace Sci. 48, 461--169 (1974). :I. Polydurova, M. An attempt to determine the retention of tobacco smoke by means of membrane lilters, in "Inhaled Particles and Vapours" (C. N. Davies. Ed.), p. 1a2. Pergamon, London, 1961. '=. Porstendorter, J., and Schraub, A. Concentration and mean particle size of the main and side ' stream of cigarette smuke. Sruub-Reinhedt. LuJi. 32, 33-36 (1972). 23. Repace, J. L.. and Lowrey. A. H. Indoor air pollution. tobacco smoke, and public health. Science ( Wusltinxton. D.C.) 208, 464-a72 (1980). 24. Scott. W. R., Taulbee, D. B., and Yu, C. P. Theoretical study of nasal deposition. Bull, .bluth. Biul. 40, )81-604 (1978). 25. Spengler. J. D., and Sexton. K. Indoor air pollution: A public health perspective. Science ( Wush- in4tun, D.C.) 221, 9-16 (1983). 26. Task Group on Lung Dynamics Committee ll-ICRP. Deposition and retention models for internal dosimetry of the human respiratory tract. Health Phys. 12, 173-208 (1966). 17. White. J. R., and Fureb, H. F. Small-airways dysfunction in nonsmokers chronically exposed to tobacco smuke. New EngL J. Aled. 303, 720-723 (1980).

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 611 TABLE 4 NICOTINE ANU COTININE LEVELS" IN URINE OF VOLUNTEL•RS EXPOSED TO SIDESTREAM SMOKE Time 2 Cigareues° 3 Cigarettes 4 Cigarettes i ne (min) Nicotine Cotinine Nicotine Cotinine Nicotine Cotinine Baseline 24 14 20 14 17 14 E 20 E 40 E 60 E 80 26 16 34 21 84 28 T30 60 90 1210 150 40 21 94 34 100 46 i I80 ) 210 1 240 t 270 ; 300 51 21 58 38 48 55 Note. E= Exposure to ~,idestream smoke in t6-tn3 chamber. ° Values given us ng/mg creatinine. ° Number of cigarettes being ,muked throughout the exposure periW. .t ninc . t in saliva returned to the levels recorded before the smoke exposure within 90- 1?0 min. Cotinine increases in saliva were not observed until 2-3 hr after first exposure to sidestream smoke, reflecting the length of time needed for enzymatic oxidation of nicotine in the liver, which is also observed in serum measurements t'or cotinine. Similarly, the rapid decline of nicotine in blood and the concurrent increase of cotinine in the urine can be explained by the kinetics of nicotine metabolism. When the volunteers were exposed to the sidestream smoke of two, three, or fuur cigarettes smoked concurrently, we found an indication of a dose-response relationship only for nicotine in the• saliva and for cotinine in the urine (Tables ?-4). The most salient points of our study on the uptake of sidestream smoke by 10 nonsmokers can be summarized as t'ullows. Neither thiucyanate in saliva, serum, or urine nor carboxyhemoglobin in blood was significantly elevated. The average nicotine levels in saliva increased significantly, reaching maxima of 430. 840, and 680 nyml alier 60 min of exposure for the three degrees of pollution (Table 5). AI'tesr the volunteers Ict't the room, their salivary nicotine values decreased rap- idly, reaching preexposure levels after 2-3 hr. Salivary cotinine levels up to 5 ngl ml were recorded 2 hr puytexposure. Nicotine in serum was not significantly diCt'.:rent from the preexposure Icvcls. whereas cotinine increased to-t ngiml 3 hr pu,texpusure. The most ,igniticunt nicutine-wtinine results were obtained from the urine analysis. A dose-re,punse relationship t'or nicotine uptake was incli- ~.,tcd within 2 hr. and tor cotinine within 2-4 hr at'ter exposure. TI BU 31606 i t

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 601 tobacco lnd "ide iference I. Amer- 'd.1. pp. ,ahlichen ,f mono- 11197_l. zs of the lyt -'_U8 T.. For- r aerusul ; {7.irticle cposuion ng. Atut. nhalatory n-Verlag. . w,itiun uf 125. 41J6- :ee ~muke ,:etttnuion unct Ne- The lung. ,n of eicle- I tNt1l). tte ~mukc :nli. C. C. (1983). Britr. Tu- '-a. Okada. T.. Ishizu. Y.. ;,nd Matsunuma. K. Determination of particle mie distribution and cun- centratiun of cigarette smoke by a light hcattcring method. Britr. Tubuk/urs(-h. 9, 153-16U (1y77). 25. Patriakanos. C.. and Hott'mann. D. Chemical stuJies on tobacco smoke LXIV: On the analysis uf arumatic amines in cigtuette smuke. J. Anu1. Tuxicul. 3, 15U-I54 11979). 26. Polydorova. M. An attempt to determine the retention uf tobucco smoke by means of membrane titters, in "lnhaleil Particles and Vapuurs" (C. N. Davtes. Ed.), pp. 142-144. Pergamun Press. Uxfurd. 1961. _7. Porstentlorfer. J.. and Schraub, A. Konzentration und mittlere Teilchengro8e des Haupt- und Nebenstromrauches der Zigarette. Srutth-Rrinh. Ltli 32. 4I)9-412 t 1972). 28. Pyrilii. C. Polycyclische Kuhlenwasscrstotfe im Zigarettenrauch. Mittriltucgsbl. GdCh-Fuclt- gruppe Lebrnswtittelchrnt. Gerichr. Chrtn. 1442). 27 (1960). 19. Pyrtki, C.. 4luller. R.. and Muldenhauer. W. Uber das Auftreten von polycyclischen Kuhlen- wussrrstuttcn tm Zigarettenrauch. tl. Mitteiiung: Untersuchung des lipuphtlen Anteils der ein- zelnen Rauchphasen suwie des Tabaks. Brr. lnw. 1'uhulr/iir.ic•!t. Dresden 7, 81-102 (1960). 30. Pyriki. C. Pulycyclische und aliphattsche Kuhtenw;userstolfe des labakrauchs. ,Vti1u•urtg 7, 439- 448 (19631. 31. Stuber, W. Generation. size distribution and composition of tubacco smoke aerosols. Recent Adt•. T6harcu Sci. 8, 3-4t (1982). 32. Task Group on Lung Dynamics. Deposition and retention models for internal dosimetry of the human respiratory tract. !leu/r!t P/trs. 12, 173-'-U8 (1966). 33. van Wijk. A. M.. and Patterson. H. S. The percentuge of particles of different sites removed from dust-laden air by breathing. J. lnd. Hy,q. Tu.ricul. 22. 31-35 (194U). 34. Wynder. E. L., and HulFinann. D. "Tobaccu and Tobacco Smoke. Studies in Experimental Car- ~;inugenesih," p. 230 ff. Academic Press. New Yo(kiLundun. 1967. 35. Yeutes. D. B.. Gerrity. T. R.. and Garrard, C. S. Charucteristics of tracheubronchial deposition and clearuncc in man. Anui. (h cup. HYg. 26. :59-272 (lylS2). 36. Yu. C. P.. Nicoluides, P., and Soong. T. T. Effect of random airway sites on aerosol deposition. ,aiaer. Ind. Hyg. Assuc•. J. 40, N99-I0U5 (1979). 37. Yu. C. P., anJ Diu. C. K. A probabilistic model fur intersubject deposition variability of inhaled particles. Aerosol Sci. Trclwul. 1, 353-362 (1982).

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING con- Jn of ), 17, con- and and ,eline idies. very ne as these com- )kzrs. ion of iation :arette artsch .Ur for ieither rm; in mam- tmines icotine . ti \ ~ Z ~ i M 5 0 SO 40 ~ CONTROI 0/ET OIET •PROLINE OIET •~~I'~E •viTC 0/ET •v/TC .. h i ' , , ~ . . NS S NS S NS S NS S p•0.04) p•00JI p•0.46 g•0l® FIG. 1. Levels ot' NPRO in the urine of smokers and nonsmokers in tht four study groups. 613 For the comparison of smokers with nonsmokers, all volunteers were placed un a controlled diet for 12 days. Twenty-four-hour urine samples were collected 0 on Day 3. at'ter daily addition of profine to the diet (Day (3), after daily addition nine of ascorbic acid and proline (Day 9), and after'daily addition of ascorbic acid 4 ( Day 12). Figure I summarizes our findings. It is evident that the endogenous Iurmatiun of N PRO is increased in cigarette smokers when a controlled diet is ~ 14 givcn and that it is further increased when a controlled diet with proline supple- ~ ,g mcnt is given. The increased endogenous formation of NPRO, which hypotheti- - cally occurs concurrently with the endogenous formation of carcinogenic nitros- . 46 amines, is confirmed by the effectiveness of ascorbic acid supplements as an inhibitur of the NPRO formation. i 55- ` . 'rable 6bummarires the first NPRO data from the study on four nonsmokers rclative to the potential of passive smoke exposure as a factor in the endogenous ;he room. iurmatiun ot' NPRO. At present, it appears that the urinary excretion of NPRO ~ in nunsmukers does not measurably increase upon three daily exposure periods t F ~ TI BU 31608

612 HOFFMANN ET AL. These assays on the uptake of sidestream smoke by nonsmokers under con- trolled conditions are being repeated. We will also extend the determination of nicotine and cotinine in the urine to include measurements of their levels 10, 17, and 24 hr postexposure. Furthermore, it needs to be determined how the con- trolled conditions will affect the values for cotinine in serum and in urine and how these can be optimized for measurement. Upon establishing these baseline data under controlled environmental conditions, we intend to initiate field studies. These will consist of measurements of nicotine in saliva as an indicator of very recent passive smoke exposure and nicotine and cotinine in serum and urine as indicators of chronic sidestream smoke exposure. We are confident that these measurements will reflect the average and high uptake of sidestream smoke com- ponents by nonsmokers and, upon comparisuns with active cigarette smokers, that we will be able to assess relative risk factors. ENDOGENOUS FORMATION OF N-NITROSAMINES Calculations with respect to risk factors should also include an estimation of the potential of inhaled sidestream smoke to enhance the Cndugenous formation of carcinogenic N-nitrosamines. Recently, we reported on the endogenous formation of nitrosamines in cigarette smokers (6). That study was based on the observation ut' Ohshima and Bartsch that the urinary excretion of nitrosuproline (NPRO) can serve as an indicator for the endogenous formation of carcinogenic nitrosamines (12). NPRO is neither mutagenic nor carcinogenic and is excreted in urine in unmetabolized form; in contrast, most carcinogenic N-nitrosamines are,quickly metabolized in the tnam- malian system. This applies also to the carcinogenic tobacco-specific nitrosamines which yield some of the same urinary metabolites as are derived from nicotine and other Nicotiana alkaloids. Fit TABLE 5 + NICOTINE AND COTININE LEVELS IN SALIVA, StNUM. ANU UNINE Ot' VOLUNTEEH5" • (SUMMANY OF AVE1lAUE V.iLUtS) R t Saliva (ng/mU Serum tngrmi) 1 Urine ing/mg creatinine) t Nicotine Nicotine Countne Ti ° Cotinitie Cutinine mr tmin) 2 3 4 4 4 ~ 3 4 2 3 4 Baseline 8 1 3 1.0 0.9 24 20 17 14 14 14 1 40 350 719 830 1.1 0.9 t 60 430 K4U tltsu _.1 1.2 26 34 84 16 '_t '_S 0 30 76 157 148 t.7 1.8 0 120 6 17 '3 2.5 2.9 40 94 100 21 34 46 0 240 8 2 3 _.0 3.3 0 300 7 7 7 3.5 3.4 51 58 48 21 38 55 ° Numbers reprencnt room pollution by smuke uF 2. 3, or 4 cigarettes. b I= Inside exposure room during pulluuun: 0 = outside exposure ruum atter leaving the room. TI BU 31607 F un : on uf ~ tU~ turn bivc mc:n calt amt uth: l rc: i:. turi tn t

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 611 TABLE 4 NICOTINE ANU COTININE LEVELS" IN URINE OF VOLUNTEL•RS EXPOSED TO SIDESTREAM SMOKE Time 2 Cigareues° 3 Cigarettes 4 Cigarettes i ne (min) Nicotine Cotinine Nicotine Cotinine Nicotine Cotinine Baseline 24 14 20 14 17 14 E 20 E 40 E 60 E 80 26 16 34 21 84 28 T30 60 90 1210 150 40 21 94 34 100 46 i I80 ) 210 1 240 t 270 ; 300 51 21 58 38 48 55 Note. E= Exposure to ~,idestream smoke in t6-tn3 chamber. ° Values given us ng/mg creatinine. ° Number of cigarettes being ,muked throughout the exposure periW. .t ninc . t in saliva returned to the levels recorded before the smoke exposure within 90- 1?0 min. Cotinine increases in saliva were not observed until 2-3 hr after first exposure to sidestream smoke, reflecting the length of time needed for enzymatic oxidation of nicotine in the liver, which is also observed in serum measurements t'or cotinine. Similarly, the rapid decline of nicotine in blood and the concurrent increase of cotinine in the urine can be explained by the kinetics of nicotine metabolism. When the volunteers were exposed to the sidestream smoke of two, three, or fuur cigarettes smoked concurrently, we found an indication of a dose-response relationship only for nicotine in the• saliva and for cotinine in the urine (Tables ?-4). The most salient points of our study on the uptake of sidestream smoke by 10 nonsmokers can be summarized as t'ullows. Neither thiucyanate in saliva, serum, or urine nor carboxyhemoglobin in blood was significantly elevated. The average nicotine levels in saliva increased significantly, reaching maxima of 430. 840, and 680 nyml alier 60 min of exposure for the three degrees of pollution (Table 5). AI'tesr the volunteers Ict't the room, their salivary nicotine values decreased rap- idly, reaching preexposure levels after 2-3 hr. Salivary cotinine levels up to 5 ngl ml were recorded 2 hr puytexposure. Nicotine in serum was not significantly diCt'.:rent from the preexposure Icvcls. whereas cotinine increased to-t ngiml 3 hr pu,texpusure. The most ,igniticunt nicutine-wtinine results were obtained from the urine analysis. A dose-re,punse relationship t'or nicotine uptake was incli- ~.,tcd within 2 hr. and tor cotinine within 2-4 hr at'ter exposure. TI BU 31606 i t

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 609 an ive ars m- ,td :he are :he 1C5 las 'lIe ine .izs CtJ smoke ported umors 0 ~ przg- ontirrn irprise t weak ;roups >assive alogies .)bacco and to )a5ure. ) apply nd the ve not '_, 1984. significantly changed. While in many countries the sales-weighted average nicotine and carbon monoxide yields of the mainstream smoke of cigarettes have greatly decreased over the last two decades (7), cigarette sidestream smoke yields have remained stable (11). The uniqueness of these conditions has prompted us to initiate an in-depth research program on the uptake of sidestream smoke by nonsmokers and on the possible endogenous formation of carcinogens upon in- halation of sidestream smoke from tobacco products. A MODEL STUDY ON SIDESTREAM SMOKE EXPOSURE For the study on uptake of sidestream smoke by nonsmokers we utilized a ,mall, bare room of 16 m3 which was constantly polluted by the sidestream smoke of four reference cigarettes concurrently smoked by machine. The mainstream smoke of these cigarettes was directed outside the room. Six air exchanges per hour were made corresponding to the average ventilation conditions of offices in the United States. Within 10-15 min of smoking, the room reached the stable pollution levels presented in Table I. Before the cigarette smoking was begun, the nonsmoking volunteers were seated in the room, two at a time; an indwelling catheter inserted into the ante- cubital vein permitted blood sampling before, during, and after exposure. Dupli- cate samples were drawn at f0-min intervals up to the 80-min exposure and at larger intervals thereatter. up to 5 hr. Tables 2-4 show the analytical protiles of markers for exposures in saliva, plasma, and urine of the nonsmoking volunteers who spent 80 min in the smoke- polluted room. Thiocyanate in saliva, serum, and urine and carboxyhemoglobin in blood were not significantly elevated in the volunteers after exposure to side- stream smoke. lmportant observations lie in the facts that nicotine derived from sidestream smoke barely increased serum nicotine levels, whereas its metabolite, cutinine, was significantly elevated in the serum 2-3 hr after the first exposure to sidestream smoke. In saliva, nicotine rapidly rose to levels above gtx) ng/ml, but subsided just as rapidly when the volunteers left the exposure area. Nicotine TABLE I TEST LAtioKArokY Sice: 16.3 mi Temperature: 22 s I°C Air exchanges: Six times per hour Pollutants: Sidestream smoke of four concurrently smoked 1RI reference cigarettes Indoor pullutiun: Particulate mattCr 4,6W µglmJ Nicutine '_2SU µyml Hydrogen cyanide 56 µg/ml Carbon munuxide ,5 ppnt NO, 0.91 ppm Formaldehyde 160 µg/mj TI BU 31604

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 617 5. Hirayama. T. Passive smoking and lung cancer: Consistency of assuciation. Luncer 2. 1425-1426 (1983). 6. Huffmann, D., and Brunnemunn. K. D. Endogenous formation ut' N-nitrosoproline in cigarette smokers. Cancer Res. 43, 5570-5574 (1983). 7. Hoffmann, D., Tsu, T. C., and Gun, G. 8. The less harmful cigarette. Prev. Med. 9, 287-296 (1980). 8. Kabat. G. C.. and Wynder, E. L. Lung cancer in nonsmokers. Cancer 53, 1214-1221 (1984). 9. Koo. L. C.. I-lu, J. H. C., and Saw, D. Active and passive smoking among female lung cancer patients and controls in Hong Kong. J. E:rp. Clin. Cancer Res. 4, 365-375 (1983). 10. Pauishall, E.. and Haley, N. J. Unpublished data (1984). II. "Report 1979-Smoking and Health," Laboratory of the Government Chemist, pp. 170-183. London. 1980. 12. Ohshima, H., and Bartsch. H. Quantitative estimation of endogenous nitrosation in humans by monitoring N-nitrosuproline excreted in the urine. Cancer Res. 41, 3658-3662 (1981). 13. Preston-Martin, S.. Yu. M. C., Benton, B., and Henderson. B. E. N-Nitroso compounds and childhood brain tumors: A case control study. Cancer Res. 42, 5240-5245 (1982). 14. Schlessclman, J., and Wald, N,1 Etls. ). Weak associations in epidemiology and their interpretation. Prev. Wed. 11. 464-476 (1982). IS. Trichupuulos, D., Kalandidi. A.. and Sparros, L. Lung cancer and passive smoking: Conclusion of Greek study. Luncer 2, 677-678 (1983). . from ,)nly a >tream :rolled :men(s f non-. ;enous -cts of )or en- .id that .Lad lung t paa,ive ~f infants 0, Iu75- A ~tudy TI BU 31612

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING con- Jn of ), 17, con- and and ,eline idies. very ne as these com- )kzrs. ion of iation :arette artsch .Ur for ieither rm; in mam- tmines icotine . ti \ ~ Z ~ i M 5 0 SO 40 ~ CONTROI 0/ET OIET •PROLINE OIET •~~I'~E •viTC 0/ET •v/TC .. h i ' , , ~ . . NS S NS S NS S NS S p•0.04) p•00JI p•0.46 g•0l® FIG. 1. Levels ot' NPRO in the urine of smokers and nonsmokers in tht four study groups. 613 For the comparison of smokers with nonsmokers, all volunteers were placed un a controlled diet for 12 days. Twenty-four-hour urine samples were collected 0 on Day 3. at'ter daily addition of profine to the diet (Day (3), after daily addition nine of ascorbic acid and proline (Day 9), and after'daily addition of ascorbic acid 4 ( Day 12). Figure I summarizes our findings. It is evident that the endogenous Iurmatiun of N PRO is increased in cigarette smokers when a controlled diet is ~ 14 givcn and that it is further increased when a controlled diet with proline supple- ~ ,g mcnt is given. The increased endogenous formation of NPRO, which hypotheti- - cally occurs concurrently with the endogenous formation of carcinogenic nitros- . 46 amines, is confirmed by the effectiveness of ascorbic acid supplements as an inhibitur of the NPRO formation. i 55- ` . 'rable 6bummarires the first NPRO data from the study on four nonsmokers rclative to the potential of passive smoke exposure as a factor in the endogenous ;he room. iurmatiun ot' NPRO. At present, it appears that the urinary excretion of NPRO ~ in nunsmukers does not measurably increase upon three daily exposure periods t F ~ TI BU 31608

PREVENTIVE MEDtCINE 13, 608-617 (1984) Tobacco Sidestream Smoke: Uptake by Nonsrnokers'-2 DIETRICH HOFFMANN,3 NANCY J. HALEY, JOHN D. ADAMS, AND KLAUS D. BRUNNEMANN Nuylur Dana Institute for Diseuse Prevention, Atnerrcan Health Foundation. Valltalla. New York I0S95 Some epidemiological studies indicate an association between passive smoking and an increased risk for cancer, especially fur cancer of the lung. Other reports, however, have failed to contirm these findings. Biochemical analyses of the physiological tluids for markers of exposure to tobacco smoke are needed us measurements of the uptake of smoke com- ponents by nonsmokers and fur the estimation of relative cancer risk to passively exposed persons compared with that to active cigarette smokers. This communication reports the uptake of carbon monoxide, hydrogen cyanide. and nicotine after passive smoke exposure under controlled conditions. The results indicate that salivary nicotine values reflect the level of recent passive smoke exposure within an hour and that urinary cutinine values indicate the level of passive smoke exposure in the preceding hours. N-Nitrusopruline has been shown to serve as an indicator uf endogenous N-nitrosumine formation in cigarette smokers: yet. preliminary studies do not indicate that urinary excretion of N-nitrusuproline is increased following short-term passive smoke exposure. In infants, first field studies suggest a correlation between exposure to tobaccu-smoke-pulluted environment> and levels of wtinine in both serum and urine. o rvn. ,k:wemtc Press, Inc. INTRODUCTION Epidemiological studies have indicated an association between passive smoke exposure and lung cancer (1, 4, 5, 15). Weak associations have also been reported for passive smoke exposure and nasal sinus cancer and for childhood brain tumors in the offspring of women passively exposed to environmental smoke during preg- nancy (5, 13). However, a number of other investigations have failed to confirm such associations (2, 8, 9). This lack of full accord Jocs nut come as a surprise because of the many sources of bias and confounders inherent in cases of weak associations (14). Although additional studies with different population groups may lead to a bettel• understanding of the weak association between passive smoke exposure and cancer, the development of new biochemical methodologies enables us today to obtain more detinitive measurements of exposure to tobacco smoke by determining the uptake of specitic components in body tluids and to calculate the risk factors relative to those inherent in active smoke exposure. Furthermore, the results of the biochemical measurements taken today also apply to exposures in the past; the uptake of passive smoke by nonsmokers and the physicochemical nature and yield of sidestream smoke per cigarette have not ' Presented at the Symposium "Medical Perspectives on Passive Smoking." April 9-12, 1984, Vienna. Austria. = Supported by NIH Grants POI-CA-_y580 and ROl-CA35607. t To whom requests for reprints should be addressad. OWt-7435r84 S3.Utl CuPYnrAt C 19" hY AuJemm Press. Inc. Art nyhla ut reprwWaiun in any Wrm rexrvad. 608 TI BU 31603 t signit. nicouz greatl: havt r to init nonan halaut Fo t- small. of fou. smukL: hour v the Gr pullutt Bet', seateti cubita. CatZ s: larger Tab: plasm: puilut- in biu., streani Nidestr cutinin to sitic but .su.

616 HOFFMANN ET AL. 79 79 0 0 S. Hirw n n ~ fl~ 6. Hutt; T ~ nn 7. Hutt, 15 0 0 iS (!: 13 13 ~ 8. Kab:. ~ 9. Kuu. E ~ II 0 ~ PL+ C 2 z lU. Pat[r z ~ ll. "ReF ~ 0 . a LL u T . 7 3 ~ t=. Ohih 0 . m, S c. S { 13. Pres c 3 . 0 3 ch • • I 14. Schic Pr r o 0 1 (S. Trich Q i 2 3 4 Nae-yror.r SmoK.r ut Numbo Of xnoWn in tM hortr MO»rn01 smoMnp uadn FtG. 3. Relationship of serum cotinine levels to at-home exposure and maternal smoking status. (Pattishall .nd Haley, unpublished rtoults). ' DISCUSSION The data presented in this study indicate that the uptake of nicotine from sidestream smoke under controlled environmental conditions amounts to only a small percentage of the nicotine uptake from active inhalation of mainstream smoke. However, repeated daily exposures to sidestream smoke under controlled conditions and for at least I week are needed in order to obtain measurements that reflect the chronic level of nicutine:le;otinine in physiological fluids of non- smokers. These long-term exposures shut.tld also indicate whether the entlugenuus formation of N-nitrosamines is increased and is thus different from the effects of a limited exposure to sidestream smoke. The tirst assays of body tluids of infants living in smoke-polluted indoor en- vironments show that the uptake of sidestream smoke can be significant and that this aspect needs our attention and requires in-depth study. REFERENCES t. Correa, P.. Pickle, L. W., Funtham, L.. Lin. T., and Haenszet, W. Passive smoking and lung cancer. Lurerrr 2, 595-597 (19tl3). 2. Gariinkel. L. Time trends in lung cancer mortality among nonsmokers and a note on passive ~,moking. J. Nutl. Cautcer Insi. 66, lUbl-IUbb (1981). 3. Greenberg. R. A., Haley, N. J., Etzal. R. A.. und Luda. F. A. Measuring the exposure of intanta to tobacco imukc: Nicotine and cutinine in urnne and saliva. New EJ+gI. J. .tilyd. 310, (U75- IU7t3 I 192f4). 4. Hiraytuna. T. Non-smoking wives uf heavy smokers have a higher risk of lung cancer: At.tudy from Japan. Brir. Myd. J. 1. 183-185 (1981). TI BU 31611

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 615 The data indicate that both nicotine and cotinine are significantly elevated in the saliva of infants exposed to tobacco smoke in their home environments. Non- exposed intitnts had insignificant salivary levels of nicotine (2.6 ng/ml) and of cotinine (0.2 nglml) while corresponding levels for infants exposed to smoke were 30 and I i nglmi, respectively. However, a better indicator of chronic or previous exposure was demonstrated by quantitatiun of urinary cotinine excretion. Co- tinine/creatinine ratios provided average values of 19 ng/mg for nonexposed in- fants and 459 ng/mg for exposed infants. Figure 2 presents the correlation between urinary cotininelcreatinine ratios and maternal smoking behavior during the 24 hr prior to sample collection. A signif- icant direct correlation of 0.64 /P = 0.0001) was found in these analyses. These data suggest that an actual dose-response relationship can be established be- tween intensity of exposure and cotinine excretion in field studies with neonates and children (Fig. 3). Currently, we are enlarging these observations in a retrospective manner to evaluate cotinine levels in serum drawn from children in 1979 and their subse- quent rates of lower respiratory disease. Preliminary data collected with Dr. E. Pattishall and Dr. R. Etzel confirm our earlier tindings of significant relationships between tobacco smoke absorption and the exposure level of the children as determined by the numbers of smokers in the home and maternal smoking status. These observations are now being extended thrbugh analyses of the serum bank and epidemiologic data available at the Frank Porter Graham Child Care Center in North Carolina. These investigations suggest that biochemical measurements appear to provide guud indicaturs of passive smoking by neonates and young children. The sensi- tivity, the specificity, and the range of predictive values need still to be established in subjects receiving the spectrum of exposures found in the general population. 9 600F• - / r.0.6r ~ 0 1ro • . 0 5 i0 i52025303540 CIGARETTES / 24 riOURS I u.. 2. Rrlauunship between the number uf cigarette~, ,mokcd by mothers in the previous 24 hr .nJ chc mnnury concentratiuns ut cutinine from their inlants. From Greenberg et u/. (3). Reprinted -uh )xrmu.wn ut the uuthurs und Nrw Engf. J. ,tilrJ. O • a•0000W c7 a00 • n•4s ~ 200 TI BU 31610 .

616 HOFFMANN ET AL. 79 79 0 0 S. Hirw n n ~ fl~ 6. Hutt; T ~ nn 7. Hutt, 15 0 0 iS (!: 13 13 ~ 8. Kab:. ~ 9. Kuu. E ~ II 0 ~ PL+ C 2 z lU. Pat[r z ~ ll. "ReF ~ 0 . a LL u T . 7 3 ~ t=. Ohih 0 . m, S c. S { 13. Pres c 3 . 0 3 ch • • I 14. Schic Pr r o 0 1 (S. Trich Q i 2 3 4 Nae-yror.r SmoK.r ut Numbo Of xnoWn in tM hortr MO»rn01 smoMnp uadn FtG. 3. Relationship of serum cotinine levels to at-home exposure and maternal smoking status. (Pattishall .nd Haley, unpublished rtoults). ' DISCUSSION The data presented in this study indicate that the uptake of nicotine from sidestream smoke under controlled environmental conditions amounts to only a small percentage of the nicotine uptake from active inhalation of mainstream smoke. However, repeated daily exposures to sidestream smoke under controlled conditions and for at least I week are needed in order to obtain measurements that reflect the chronic level of nicutine:le;otinine in physiological fluids of non- smokers. These long-term exposures shut.tld also indicate whether the entlugenuus formation of N-nitrosamines is increased and is thus different from the effects of a limited exposure to sidestream smoke. The tirst assays of body tluids of infants living in smoke-polluted indoor en- vironments show that the uptake of sidestream smoke can be significant and that this aspect needs our attention and requires in-depth study. REFERENCES t. Correa, P.. Pickle, L. W., Funtham, L.. Lin. T., and Haenszet, W. Passive smoking and lung cancer. Lurerrr 2, 595-597 (19tl3). 2. Gariinkel. L. Time trends in lung cancer mortality among nonsmokers and a note on passive ~,moking. J. Nutl. Cautcer Insi. 66, lUbl-IUbb (1981). 3. Greenberg. R. A., Haley, N. J., Etzal. R. A.. und Luda. F. A. Measuring the exposure of intanta to tobacco imukc: Nicotine and cutinine in urnne and saliva. New EJ+gI. J. .tilyd. 310, (U75- IU7t3 I 192f4). 4. Hiraytuna. T. Non-smoking wives uf heavy smokers have a higher risk of lung cancer: At.tudy from Japan. Brir. Myd. J. 1. 183-185 (1981). TI BU 31611

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 607 zan be d cort- tted on ieveral arable, 0) esti- :ce and .; dusts :5zd on emznt. i!; low- res are ith the z toxic , i main- J other "1:1zaCd. _rs to Lur 19 (197_). J-nitroea- Res•. 37, cntnuiun the Nun- sidestream cigarette smoke in the human respiratory tract. Anter. Rev. Respir. Dis. 125, 406- 408 (1982). 9. Hinds, W. C. Size characteristics of cigarette Smoke. .iiner. lnd. Hyg. Assuc. J. 39, a8-54 (1978). 10. Hinds, W. C.. and First. M. W. Concentrations of nicotine and tobacco smoke in public places. New Engl. J. Med. 292, 844-845 (1975). 11. Hirayama, T. Nonsmoking wives of heavy smokers have a higher risk of lung cancer: A study from Japan. Brit. a4ed. l. 282, 183-185 (1981). 12. Hoegg, U. R. Cigarette ~.moke in closed spaces. Environ. Health Perspect. 2, 117-128 (1972). 12a. Hugod. C., Hawkins. L. H.. and Astrup, P. Exposure of passive smokers to tobacco smoke constituents. Int. Arck. Occup. Envirun. Health 42, 21-29 (19781. 13. Keith. C. H., and Derrick. J. C. Measurement of the particle size distribution and concentration of cigarette smoke by the "Conifuge." J. Colloid Sci. 15, 340-356 ( l9Wl. 14. Lippmann, M. Regional deposition of particles in the human respiratory tract, in "Handbook of Physiology," Section 9: "Reuctions to Environmental Agents" (S. R. Geiger, Ed.), p. 213. Waveriy Press, kialtimore, 1977. 15. Mazumder. M. K.. Ware, R. E.. Wilson, 1. D.. Renninger, R. G., Hiller. F C.. McCleod. P. C., Raible, R. W., and Testerman. M. K. SPART Analyzer: Its application to aerodynamic size distribution measurement. J. Aerusol Sci. 10, 561-569 (1979). 16. MeCusker, K.. Hiller, F. C., Wilson. J. D., Mazumder. M. K., and Bone. R. Aerodynamic sizing of tobacco smoke particulate from commercial cigarettes. Arch. Ertvirun. Heulth 38, 215-218 (1983). 17. lvtitchell, R. 1. Controlled measurement of smoke-particle retention in the respiratory tract. ,iiner. Rev. Respir. Dis. 85, 526-533 (1962). 18. Muir. D. C. F. Tobacco smoke inhalation. in "Environmental Tobacco Smoke Effects on the Non- Smoker" (R. Ryiander,° Ed.), p. 44. Schmidt. Vujens, 1974. 19. Muir. D. C. F., and Davies, C. N. The deposition ot' 0.5 µm diameter aerosuls in the lungs of man. `[nn. uc•cup. Hyg. 10, 161-174 (1967). 20. Okada. T., and Matsunuma. K. Determination of particle size and concentration of cigarette Nmuke by a light scattering method. J. Culluid lnterJace Sci. 48, 461--169 (1974). :I. Polydurova, M. An attempt to determine the retention of tobacco smoke by means of membrane lilters, in "Inhaled Particles and Vapours" (C. N. Davies. Ed.), p. 1a2. Pergamon, London, 1961. '=. Porstendorter, J., and Schraub, A. Concentration and mean particle size of the main and side ' stream of cigarette smuke. Sruub-Reinhedt. LuJi. 32, 33-36 (1972). 23. Repace, J. L.. and Lowrey. A. H. Indoor air pollution. tobacco smoke, and public health. Science ( Wusltinxton. D.C.) 208, 464-a72 (1980). 24. Scott. W. R., Taulbee, D. B., and Yu, C. P. Theoretical study of nasal deposition. Bull, .bluth. Biul. 40, )81-604 (1978). 25. Spengler. J. D., and Sexton. K. Indoor air pollution: A public health perspective. Science ( Wush- in4tun, D.C.) 221, 9-16 (1983). 26. Task Group on Lung Dynamics Committee ll-ICRP. Deposition and retention models for internal dosimetry of the human respiratory tract. Health Phys. 12, 173-208 (1966). 17. White. J. R., and Fureb, H. F. Small-airways dysfunction in nonsmokers chronically exposed to tobacco smuke. New EngL J. Aled. 303, 720-723 (1980).

614 HOFFMANN ET AL. TABLE 6 N-NITROSOPROUNE IN 24-hr URINE SAMPLES OF MALE NONSMOKERS EXPOSEU TO PASSIVE SMOKING° S t b Day I Day 2 Day 3° Day 4' Day 3° u jec No. N PRO CR COT NPRO CR COT NPRO CR COT NPRO CR COT NPRO CR COT I 1.6 1.4 6 0.5 1. 1 3 2.0 1.4 10 6.1 1.2 80 2.9 2.1 360 2 1.7 1.8 7 2.8 1.7 2 3.2 1.9 100 1.7 1.1 350 0.4 1.8 215 3 4.8 2.2 6 5.2 :.4 6 3.0 2.1 90 4.4 2.3 300 2.6 2.1 260 4 :.9 2.2 5 4.3 :.2 5 2.8 2.4 65 3.3 2.2 .00 2.0 2.1 330 Nean 2.75 3.20 2.75 3.88 1.98 Nore. NPRO =,V-mtrosoproline (ngJ24 hr), CR - creatinine (g24 hr). COT = counine Englmll. ° On cuncrolled diet, receiving 3 x 2W mg prutine per day. °Pa°aive imolfe e}lpoBLLre, 3 x 80 min per day. Vatues ad)uated to 24-hr urine. of 80 min each to a smoke-polluted environment, but these preliminary data require confirmation with data from a larger group of volunteers. EXPOSURE OF INFANTS TO TOBACCO SMOKE Currently, we are investigating the exposure to and absorption of sidestream smoke in neonates and infants. These studies are being carried out in a azries of collaborative projects with the Departments of Pediatrics and Community Pedi- atrics at the University of North Carolina School of Medicine. The pilot investigation, conducted with Dr. Robert Greenberg, included 32 infants from homes with at least one cigarette smoker and 19 infants who were reported as not exposed to tobacco smoke. Saliva and urine were collected from the infants who were less than 10 months of age, with 23 of them being under 2 months. The study's findings are summarized in Table 7 (3). i TABLE 7 } URINARY .4NU SALIV ARY CONCENTRATIO NS OF NICU'rINE AND CUrININE FROM [NFA NTS EX Y05F.D AN €u NoT ExPusEU To ToaACCO SMOKE [ r Urine Salivu f Parental Nicotine:creatinine Cotinine:creutinine Nicotine Coumnt ~ report (ng/mg) • (ng/mg) Ing/mt) (nglmil t Nonexposed Mean 7 19 2.6 t i 0.2 ~ R•rnge 0-59 0-125 0-17.6 0-3 (n = 18) In = 18) ln = 13) (n = 13) Exposed Mean 85 459 30 I I Range 0-370 41-1tlri5 U-166 (0-.5I (n = 28) In = 28) (1I = 29) ln = 27) Significance oP dit'fcrence P < 0.0t1U1 P < 0. 000 1 1 P < 0.0003 P< U.Utwl From Greenberg er uI. (3). Reprinted with permeawn of the authors and Ne+v Enyf. J. .bled. TI BU 31609 r b U y[I Nl

614 HOFFMANN ET AL. TABLE 6 N-NITROSOPROUNE IN 24-hr URINE SAMPLES OF MALE NONSMOKERS EXPOSEU TO PASSIVE SMOKING° S t b Day I Day 2 Day 3° Day 4' Day 3° u jec No. N PRO CR COT NPRO CR COT NPRO CR COT NPRO CR COT NPRO CR COT I 1.6 1.4 6 0.5 1. 1 3 2.0 1.4 10 6.1 1.2 80 2.9 2.1 360 2 1.7 1.8 7 2.8 1.7 2 3.2 1.9 100 1.7 1.1 350 0.4 1.8 215 3 4.8 2.2 6 5.2 :.4 6 3.0 2.1 90 4.4 2.3 300 2.6 2.1 260 4 :.9 2.2 5 4.3 :.2 5 2.8 2.4 65 3.3 2.2 .00 2.0 2.1 330 Nean 2.75 3.20 2.75 3.88 1.98 Nore. NPRO =,V-mtrosoproline (ngJ24 hr), CR - creatinine (g24 hr). COT = counine Englmll. ° On cuncrolled diet, receiving 3 x 2W mg prutine per day. °Pa°aive imolfe e}lpoBLLre, 3 x 80 min per day. Vatues ad)uated to 24-hr urine. of 80 min each to a smoke-polluted environment, but these preliminary data require confirmation with data from a larger group of volunteers. EXPOSURE OF INFANTS TO TOBACCO SMOKE Currently, we are investigating the exposure to and absorption of sidestream smoke in neonates and infants. These studies are being carried out in a azries of collaborative projects with the Departments of Pediatrics and Community Pedi- atrics at the University of North Carolina School of Medicine. The pilot investigation, conducted with Dr. Robert Greenberg, included 32 infants from homes with at least one cigarette smoker and 19 infants who were reported as not exposed to tobacco smoke. Saliva and urine were collected from the infants who were less than 10 months of age, with 23 of them being under 2 months. The study's findings are summarized in Table 7 (3). i TABLE 7 } URINARY .4NU SALIV ARY CONCENTRATIO NS OF NICU'rINE AND CUrININE FROM [NFA NTS EX Y05F.D AN €u NoT ExPusEU To ToaACCO SMOKE [ r Urine Salivu f Parental Nicotine:creatinine Cotinine:creutinine Nicotine Coumnt ~ report (ng/mg) • (ng/mg) Ing/mt) (nglmil t Nonexposed Mean 7 19 2.6 t i 0.2 ~ R•rnge 0-59 0-125 0-17.6 0-3 (n = 18) In = 18) ln = 13) (n = 13) Exposed Mean 85 459 30 I I Range 0-370 41-1tlri5 U-166 (0-.5I (n = 28) In = 28) (1I = 29) ln = 27) Significance oP dit'fcrence P < 0.0t1U1 P < 0. 000 1 1 P < 0.0003 P< U.Utwl From Greenberg er uI. (3). Reprinted with permeawn of the authors and Ne+v Enyf. J. .bled. TI BU 31609 r b U y[I Nl

616 HOFFMANN ET AL. 79 79 . . n 77 13 0 0 16 13 ~ E e I[ 0 it ~ C 2 2 ~ 9 . z 0 0 u T 0 7 3 0 . 5 S 3 0 0 3 I 0 I ~ 0 ~ ... 1 2 3 4 Npn Llqlt.r STaaw Nun10K of SR10Nfq m IfN nCrtU Mpi.rnCl srtq" st4Uf FIG. 3. Relationship of serum cotinine levels to at-home exposure and maternal smoking status. (Pattishall and Haley, unpublished results). DISCUSSION The data presented in this study indicate that the uptake of nicotine from sidestream smoke under controlled environmental conditions amounts to only a small percentage of the nicotine uptake from active inhalation of mainstream smoke. However, repeated daily exposures to sidestreatn smoke under controlled conditions and for at least I week are needed in order to obtain measurements that retlect the chronic level of nicotine/cotinine in physiological tluids of non- smokers. These long-term exposures should also indicate whether the endogenous formation of N-nitrosamines is increased and is thus different from the cffec:ts of a limited exposure to sidestream smoke. The'tirst a»ays of body tluids of infants living in smuke-pulluted indoor en- vironments show that the uptake ofCdestream smoke can be signiticant and that this aspect needs our attention and requires in-depth study. REFERENCES 1. Correa. P., Pickle. L. W., Fontham, L.. Lin, T., and Huenszel. W. Passive smoking and lung cancer. Luncrr 2. 595-597 (19tl3 ). 2. Gartinkel. L. Time trends in lung cancer mortality among nonsmokers and a note on passtvc • smoking. J. Nurl. Cancer lruv. 66, t1161-1066 (198 1). 0 3. Greenberg. R. A., Haley. N. J., Etzel. R. A.. and Luda. F A. Measuring the exposure of intanth to tobacco smoke: Nicutine and cotinine in unne and saliva. New Engl. J. Med. 310. 1075- 1078 (1yii4). 4. Hiniyuma, T. Non-smoking wives of heavy smokers have a higher risk of lung cancer: A StuJy Itytlt). from Japan. Brir. Med. J. t. 183-185 TI BU 31613 5. Hira~ (l'j 6. Hutfr 7. Hotfr tl', iS. Kaba 9. Kuu. Pa 10. Pattla l l. ~~(ZCt' 1=. Ohsh mi, 13. Prest ': h 14. Sch[c Pr 15. Trich of

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 617 5. Hirayama. T. Passive smoking and lung cancer: Consistency of assuciation. Luncer 2. 1425-1426 (1983). 6. Huffmann, D., and Brunnemunn. K. D. Endogenous formation ut' N-nitrosoproline in cigarette smokers. Cancer Res. 43, 5570-5574 (1983). 7. Hoffmann, D., Tsu, T. C., and Gun, G. 8. The less harmful cigarette. Prev. Med. 9, 287-296 (1980). 8. Kabat. G. C.. and Wynder, E. L. Lung cancer in nonsmokers. Cancer 53, 1214-1221 (1984). 9. Koo. L. C.. I-lu, J. H. C., and Saw, D. Active and passive smoking among female lung cancer patients and controls in Hong Kong. J. E:rp. Clin. Cancer Res. 4, 365-375 (1983). 10. Pauishall, E.. and Haley, N. J. Unpublished data (1984). II. "Report 1979-Smoking and Health," Laboratory of the Government Chemist, pp. 170-183. London. 1980. 12. Ohshima, H., and Bartsch. H. Quantitative estimation of endogenous nitrosation in humans by monitoring N-nitrosuproline excreted in the urine. Cancer Res. 41, 3658-3662 (1981). 13. Preston-Martin, S.. Yu. M. C., Benton, B., and Henderson. B. E. N-Nitroso compounds and childhood brain tumors: A case control study. Cancer Res. 42, 5240-5245 (1982). 14. Schlessclman, J., and Wald, N,1 Etls. ). Weak associations in epidemiology and their interpretation. Prev. Wed. 11. 464-476 (1982). IS. Trichupuulos, D., Kalandidi. A.. and Sparros, L. Lung cancer and passive smoking: Conclusion of Greek study. Luncer 2, 677-678 (1983). . from ,)nly a >tream :rolled :men(s f non-. ;enous -cts of )or en- .id that .Lad lung t paa,ive ~f infants 0, Iu75- A ~tudy TI BU 31612

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 615 The data indicate that both nicotine and cotinine are significantly elevated in the saliva of infants exposed to tobacco smoke in their home environments. Non- exposed intitnts had insignificant salivary levels of nicotine (2.6 ng/ml) and of cotinine (0.2 nglml) while corresponding levels for infants exposed to smoke were 30 and I i nglmi, respectively. However, a better indicator of chronic or previous exposure was demonstrated by quantitatiun of urinary cotinine excretion. Co- tinine/creatinine ratios provided average values of 19 ng/mg for nonexposed in- fants and 459 ng/mg for exposed infants. Figure 2 presents the correlation between urinary cotininelcreatinine ratios and maternal smoking behavior during the 24 hr prior to sample collection. A signif- icant direct correlation of 0.64 /P = 0.0001) was found in these analyses. These data suggest that an actual dose-response relationship can be established be- tween intensity of exposure and cotinine excretion in field studies with neonates and children (Fig. 3). Currently, we are enlarging these observations in a retrospective manner to evaluate cotinine levels in serum drawn from children in 1979 and their subse- quent rates of lower respiratory disease. Preliminary data collected with Dr. E. Pattishall and Dr. R. Etzel confirm our earlier tindings of significant relationships between tobacco smoke absorption and the exposure level of the children as determined by the numbers of smokers in the home and maternal smoking status. These observations are now being extended thrbugh analyses of the serum bank and epidemiologic data available at the Frank Porter Graham Child Care Center in North Carolina. These investigations suggest that biochemical measurements appear to provide guud indicaturs of passive smoking by neonates and young children. The sensi- tivity, the specificity, and the range of predictive values need still to be established in subjects receiving the spectrum of exposures found in the general population. 9 600F• - / r.0.6r ~ 0 1ro • . 0 5 i0 i52025303540 CIGARETTES / 24 riOURS I u.. 2. Rrlauunship between the number uf cigarette~, ,mokcd by mothers in the previous 24 hr .nJ chc mnnury concentratiuns ut cutinine from their inlants. From Greenberg et u/. (3). Reprinted -uh )xrmu.wn ut the uuthurs und Nrw Engf. J. ,tilrJ. O • a•0000W c7 a00 • n•4s ~ 200 TI BU 31610 .

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 619 and r. I n 4late uUur antL l in- ron- noY- nukr .hebe ccru iVC a paid :uch- lemic N tob-acco possible :ntrati.un for mea-~ room. ents in a ne. were uated by .le (NO), .ccurding 5 due to s before llues are 15 ~lmul- .buratury experiments-measurements of eye blink rate, considered to be an objective measure for eye irritations [for more details see (7) and (13)1. The laboratory studies were carried out in a climatic chamber of 30 m3, with an air temperature of ?U-'_4°C and a relative humidity between 40 and 60%. The ventilation rate could be varied between 0.1 and 16 changes/hr. The smoke was produced by a Borgwald smoking machine under standardized conditions. Air pollution was produced by the sidestream smoke of cigarettes only, the main- ,tream smoke having been conducted out of the chamber. Healthy students were exposed in groups of two to three to sidestream smoke in the climatic chamber. They all also participated in a control experiment with identical conditions, but without smoke. RESULTS AND DISCUSSION htcld SluclY Table I summarizes some results of the chemical measurements. The compar- isun of the :1 values ot' Table I with the measured absolute indoor concentrations reveals that 30-70% of the measured indoor concentrations of CO, NO, and PM are due to tobacco smoke. The correlation between the gas-phase components .1CO and ANO is relatively high (Pearson correlation coefficient r = 0.73). How- ever, the correlations with Anicotine and IPM are low. Therefore CO can be considered a useful indicator for nitrogen oxide, but not for particulate matter or nicotine. Figure I presents some results of the 472 employee interviews by smoking ,tatus (smoker vs nonsmoker). From this figure it can be deduced that: (a) Approximately 33% of the employees qualified the air at work as bad with regard to smoke. (b) Forty percent were disturbed by smoke. (c) Twenty-tive percent reported eye irritation at work. (d) Nonsmokers reacted significantly more to environmental tobacco smoke than did smokers. Empluyces suffering from hay .fever reported signiticantly more eye irritation at work than those without hay fi ver. Furthermure. 72% of the interviewed nun- ,ntukers and 67% of the smokers were in favor of sepsirate smoking and non- smoking workrounis; 49% supported a partial or total prohibition of smoking at work. TABLE I Atx Ptn.t.uru,N uun rct TuuM rct SMttKt: iN 44 WuKtatcwtits 0 Cumpunenl Number al" ,ampIc. MCUn values Standard deviation Maximum ~CU (ppm) 353 1.1 1.3 6.5 .1N() tpph) 348 3: 60 _tSU .1YM (µ(ym') 429 .133 130 962 1Nicwinr {dU 0.9 1.9 13.8 Y,nr. AValue ="indour cuncentr.itiun during work" minus "induur cunccntratiun beturc wurk.•" PM = parucuiate muttrr. TI BU 31615

PREVENTtYE MEUICINI: 13, 612i-62.5 11984) Annoyance and Irritation by Passive Smoking' ANNETTA WEBER Swiss Federu( Institute of Tec•hnolugy, Depurttnent uf Hygiene and Ergonuntics, CH-8092 Zurich, Switzerland The acute imtating and annoying effects of smoke have been investigated in tield and laboratory studies by exumining the concentration ot' some smoke components in air. In the workplace, 30 to 70% uf the indoor carbon monoxide, nitrogen oxide, and particulate concentrations are due to tobacco smoke; 25-41y7o of the employees are disturbed andJur annoyed by smoke and I5% suffer from eye irritatiun at work. Subjective eye, nuse, and throat irritations and eye blink rate increase with increasing smoke concentration and in- creasing exposure duration. Irritation is due nlatniy to the particulate phase of environ- mental tobacco smoke, whereas the gas phase is, to a large extent, responsible tor annoy- ance. It is concluded that healthy individuals can tolerate an environmental tobacco smoke level that corresponds to a carbun munuxida concentration of 1.5 to 2.0 ppm. Above these limits, countermeasures to protect passive smokers are necesaary. In order not to exceed the upper toler.tbie threshold Iimit of 2.0 ppm carbon monoxide, it is necesary to have a fresh air supply of 33 m} per hour per cigarette smoked. Special anentiun ~huuld be paid to groups of people with increased senaitivlty to environmental tobacco smuke. e.g., asth- matics, allergic individuals, chronic bronchitis sufferers, and children. c Ivna .a,;;,dem4G Pro», Inc. INTRODUCTION Several studies were carried laut to investigate air pollution due to tobacco smoke and its acute effects in order to draw some conclusions from a posaibie dose-regponse relationship on the ~ environmental tobacco smoke concentration tolerable for humans. Such a tulerance limit could give some indications for mea- sures to protect nonsmokers, such as the dimensioning of fresh air to a room. METHODS A field study in 44 workrooms and two series of laboratory experiments in a climatic chamber, in which cigarettes were smoked by a smoking machine, were carried out. The degree of air pollution due to tobacco smoke was evaluated by measuring the concentrationh of carbon monoxide (CO), nitrogen oxide (NO), formaldehyde, acrolein, particulate matter (PM), and nicotine in the air according to methods described elsewhere (7, 13. 15). The portion of pollutants due to tobacco smoke was estimated by subtracting the background levels before smoking from the concentrations during smoking. ThCSe difference values are hereafter called ACO. .1NO, and so on. The degree of acute irritation and annoyance to exposed persons was ~,imul- taneuusly determined by means of questiunnaire and-in the casz of laboratory ' Presented at the Symposium "Medica! Perspectives un Pasaive Smuking," April 9-12. 1984. VVienna, Austria. 618 l)tr9t •7d35/li4 $3.00 C,1pynghr •. FvBi by A,;aJemIc Preaa. Inc. NII nbOln ut reprodlNUUr1 in anY Iunn re%crvOd. TI BU 31614 cxp mc: ann vt;n pt'o pul stt'c exF Tht Wit

PREVENTtYE MEUICINI: 13, 612i-62.5 11984) Annoyance and Irritation by Passive Smoking' ANNETTA WEBER Swiss Federu( Institute of Tec•hnolugy, Depurttnent uf Hygiene and Ergonuntics, CH-8092 Zurich, Switzerland The acute imtating and annoying effects of smoke have been investigated in tield and laboratory studies by exumining the concentration ot' some smoke components in air. In the workplace, 30 to 70% uf the indoor carbon monoxide, nitrogen oxide, and particulate concentrations are due to tobacco smoke; 25-41y7o of the employees are disturbed andJur annoyed by smoke and I5% suffer from eye irritatiun at work. Subjective eye, nuse, and throat irritations and eye blink rate increase with increasing smoke concentration and in- creasing exposure duration. Irritation is due nlatniy to the particulate phase of environ- mental tobacco smoke, whereas the gas phase is, to a large extent, responsible tor annoy- ance. It is concluded that healthy individuals can tolerate an environmental tobacco smoke level that corresponds to a carbun munuxida concentration of 1.5 to 2.0 ppm. Above these limits, countermeasures to protect passive smokers are necesaary. In order not to exceed the upper toler.tbie threshold Iimit of 2.0 ppm carbon monoxide, it is necesary to have a fresh air supply of 33 m} per hour per cigarette smoked. Special anentiun ~huuld be paid to groups of people with increased senaitivlty to environmental tobacco smuke. e.g., asth- matics, allergic individuals, chronic bronchitis sufferers, and children. c Ivna .a,;;,dem4G Pro», Inc. INTRODUCTION Several studies were carried laut to investigate air pollution due to tobacco smoke and its acute effects in order to draw some conclusions from a posaibie dose-regponse relationship on the ~ environmental tobacco smoke concentration tolerable for humans. Such a tulerance limit could give some indications for mea- sures to protect nonsmokers, such as the dimensioning of fresh air to a room. METHODS A field study in 44 workrooms and two series of laboratory experiments in a climatic chamber, in which cigarettes were smoked by a smoking machine, were carried out. The degree of air pollution due to tobacco smoke was evaluated by measuring the concentrationh of carbon monoxide (CO), nitrogen oxide (NO), formaldehyde, acrolein, particulate matter (PM), and nicotine in the air according to methods described elsewhere (7, 13. 15). The portion of pollutants due to tobacco smoke was estimated by subtracting the background levels before smoking from the concentrations during smoking. ThCSe difference values are hereafter called ACO. .1NO, and so on. The degree of acute irritation and annoyance to exposed persons was ~,imul- taneuusly determined by means of questiunnaire and-in the casz of laboratory ' Presented at the Symposium "Medica! Perspectives un Pasaive Smuking," April 9-12. 1984. VVienna, Austria. 618 l)tr9t •7d35/li4 $3.00 C,1pynghr •. FvBi by A,;aJemIc Preaa. Inc. NII nbOln ut reprodlNUUr1 in anY Iunn re%crvOd. TI BU 31614 cxp mc: ann vt;n pt'o pul stt'c exF Tht Wit

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 619 and r. I n 4late uUur antL l in- ron- noY- nukr .hebe ccru iVC a paid :uch- lemic N tob-acco possible :ntrati.un for mea-~ room. ents in a ne. were uated by .le (NO), .ccurding 5 due to s before llues are 15 ~lmul- .buratury experiments-measurements of eye blink rate, considered to be an objective measure for eye irritations [for more details see (7) and (13)1. The laboratory studies were carried out in a climatic chamber of 30 m3, with an air temperature of ?U-'_4°C and a relative humidity between 40 and 60%. The ventilation rate could be varied between 0.1 and 16 changes/hr. The smoke was produced by a Borgwald smoking machine under standardized conditions. Air pollution was produced by the sidestream smoke of cigarettes only, the main- ,tream smoke having been conducted out of the chamber. Healthy students were exposed in groups of two to three to sidestream smoke in the climatic chamber. They all also participated in a control experiment with identical conditions, but without smoke. RESULTS AND DISCUSSION htcld SluclY Table I summarizes some results of the chemical measurements. The compar- isun of the :1 values ot' Table I with the measured absolute indoor concentrations reveals that 30-70% of the measured indoor concentrations of CO, NO, and PM are due to tobacco smoke. The correlation between the gas-phase components .1CO and ANO is relatively high (Pearson correlation coefficient r = 0.73). How- ever, the correlations with Anicotine and IPM are low. Therefore CO can be considered a useful indicator for nitrogen oxide, but not for particulate matter or nicotine. Figure I presents some results of the 472 employee interviews by smoking ,tatus (smoker vs nonsmoker). From this figure it can be deduced that: (a) Approximately 33% of the employees qualified the air at work as bad with regard to smoke. (b) Forty percent were disturbed by smoke. (c) Twenty-tive percent reported eye irritation at work. (d) Nonsmokers reacted significantly more to environmental tobacco smoke than did smokers. Empluyces suffering from hay .fever reported signiticantly more eye irritation at work than those without hay fi ver. Furthermure. 72% of the interviewed nun- ,ntukers and 67% of the smokers were in favor of sepsirate smoking and non- smoking workrounis; 49% supported a partial or total prohibition of smoking at work. TABLE I Atx Ptn.t.uru,N uun rct TuuM rct SMttKt: iN 44 WuKtatcwtits 0 Cumpunenl Number al" ,ampIc. MCUn values Standard deviation Maximum ~CU (ppm) 353 1.1 1.3 6.5 .1N() tpph) 348 3: 60 _tSU .1YM (µ(ym') 429 .133 130 962 1Nicwinr {dU 0.9 1.9 13.8 Y,nr. AValue ="indour cuncentr.itiun during work" minus "induur cunccntratiun beturc wurk.•" PM = parucuiate muttrr. TI BU 31615

n pptnl 'eriod 0 ~. 11 ~iide- SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING . © 0 623 10 20 30 40 50 60 mfn Eaposure duration FIG. 5. Effects of the total sidestream smoke of cigarettes and its gas phase on annoyance, both cxpoaures corresponding to a tobacco smoke induced ACO level of 10 ppm. Annoyance index cal- cutatcd from the answers to four questions. Scak: 1- minimum, 3= maximum. 32 subjects. 0 min: Me:uurement before smoke production. •P < 0.05: •'P < 0.0 1. -. Total sidestream smoke: ---, wta phase; -. control. There is a marked increase in the incidence oP strong eye irritations between ~moke levels corresponding to 1.3 and 2.5 ppm ACO. These levels seem to rep- resent a critical range of tobacco smoke concentration. Based on these results, a tentative threshold limit for environmental tobacco smoke expressed as ACO should lie between 1.5 and 2.0 ppm. It must be stressed, however. • that this prupusal is valid for healthy adults only and is aimed at protecting the well-being 0 E M Y C m 0 10 20 30 40 so ao min Eaposure duration I•K. e. Effects of total sidestream smoke of cigarettes and of its gaN phase on eye blink rate. both itrao+ures corresponding to a tobacco smoke induced .1CO level of 10 ppm. 32 subjects. 0 min: vc.t,urement before smoke production. •P < 0.05: •'P <' 0.01. -. Total sidestream smoke: ---. v% pttaae: •-, control. TI BU 31619

n pptnl 'eriod 0 ~. 11 ~iide- SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING . © 0 623 10 20 30 40 50 60 mfn Eaposure duration FIG. 5. Effects of the total sidestream smoke of cigarettes and its gas phase on annoyance, both cxpoaures corresponding to a tobacco smoke induced ACO level of 10 ppm. Annoyance index cal- cutatcd from the answers to four questions. Scak: 1- minimum, 3= maximum. 32 subjects. 0 min: Me:uurement before smoke production. •P < 0.05: •'P < 0.0 1. -. Total sidestream smoke: ---, wta phase; -. control. There is a marked increase in the incidence oP strong eye irritations between ~moke levels corresponding to 1.3 and 2.5 ppm ACO. These levels seem to rep- resent a critical range of tobacco smoke concentration. Based on these results, a tentative threshold limit for environmental tobacco smoke expressed as ACO should lie between 1.5 and 2.0 ppm. It must be stressed, however. • that this prupusal is valid for healthy adults only and is aimed at protecting the well-being 0 E M Y C m 0 10 20 30 40 so ao min Eaposure duration I•K. e. Effects of total sidestream smoke of cigarettes and of its gaN phase on eye blink rate. both itrao+ures corresponding to a tobacco smoke induced .1CO level of 10 ppm. 32 subjects. 0 min: vc.t,urement before smoke production. •P < 0.05: •'P <' 0.01. -. Total sidestream smoke: ---. v% pttaae: •-, control. TI BU 31619

616 HOFFMANN ET AL. 79 79 . . n 77 13 0 0 16 13 ~ E e I[ 0 it ~ C 2 2 ~ 9 . z 0 0 u T 0 7 3 0 . 5 S 3 0 0 3 I 0 I ~ 0 ~ ... 1 2 3 4 Npn Llqlt.r STaaw Nun10K of SR10Nfq m IfN nCrtU Mpi.rnCl srtq" st4Uf FIG. 3. Relationship of serum cotinine levels to at-home exposure and maternal smoking status. (Pattishall and Haley, unpublished results). DISCUSSION The data presented in this study indicate that the uptake of nicotine from sidestream smoke under controlled environmental conditions amounts to only a small percentage of the nicotine uptake from active inhalation of mainstream smoke. However, repeated daily exposures to sidestreatn smoke under controlled conditions and for at least I week are needed in order to obtain measurements that retlect the chronic level of nicotine/cotinine in physiological tluids of non- smokers. These long-term exposures should also indicate whether the endogenous formation of N-nitrosamines is increased and is thus different from the cffec:ts of a limited exposure to sidestream smoke. The'tirst a»ays of body tluids of infants living in smuke-pulluted indoor en- vironments show that the uptake ofCdestream smoke can be signiticant and that this aspect needs our attention and requires in-depth study. REFERENCES 1. Correa. P., Pickle. L. W., Fontham, L.. Lin, T., and Huenszel. W. Passive smoking and lung cancer. Luncrr 2. 595-597 (19tl3 ). 2. Gartinkel. L. Time trends in lung cancer mortality among nonsmokers and a note on passtvc • smoking. J. Nurl. Cancer lruv. 66, t1161-1066 (198 1). 0 3. Greenberg. R. A., Haley. N. J., Etzel. R. A.. and Luda. F A. Measuring the exposure of intanth to tobacco smoke: Nicutine and cotinine in unne and saliva. New Engl. J. Med. 310. 1075- 1078 (1yii4). 4. Hiniyuma, T. Non-smoking wives of heavy smokers have a higher risk of lung cancer: A StuJy Itytlt). from Japan. Brir. Med. J. t. 183-185 TI BU 31613 5. Hira~ (l'j 6. Hutfr 7. Hotfr tl', iS. Kaba 9. Kuu. Pa 10. Pattla l l. ~~(ZCt' 1=. Ohsh mi, 13. Prest ': h 14. Sch[c Pr 15. Trich of

SYMPOSIUM: MED[CAL PERSPECTIVES ON PASSIVE SMOKING 621 aye irritation •ye blink rataimin ,«y naeng S -1 r• 80 r 60 r 40 20 rkplace. eatl. 0. on the La were ri tation ialyzeJ -ealistic t to to- o these ing the naining ,r com- tfas•rn none I 'J ~ ~' 0 0 10 20 mm ACO 1 1/ 22 ]2 42 43 PPT ONO 0.0! 0.42 0.77 1,11 1.43 1.30 OYaM A NCNO 0.03 0.10 0.32 0.47 0.d2 0.64000 pwo/ou 0 0.06 0.11 0.1• 0.20 0.20ppin numtNr o/ c10. 0 10 20 F1c:. 2. Mean subjective eye irritations, mean eye blink rate, and concentrations of some pollutants Junng continuous smoke production in an unventilated climatic chamber, 33 subjects. Ventilation we. 0.01 H''. Eye irritation index calculated from the answers to four qubstions concerning eye irrntatwns. 0 min: Measurement before smoke production. -. Eye irritation index; •••, eye blink rate. indicator to evaluate the degree of air pollution due to environmental tobacco• tmuke. Figures 3 and 4 illustrate the results obtained for subjective eye irritation rnd eye blink rate. Two facts are obvious: (a) Mean eye irritation as well as eye blink rate increase with increasing smoke wncentration. (b) Both irritation parameters increase with the duration of exposure in spite of a constant smoke concentration. The same, but less pronounced result has Alw been observed for nose and throat irritations. The pattern of annoyance differs from that of irritation: it increases rapidly as wun as smoke production begins and, after 10 to 15 min of exposure, remains Approximately constant. Thus, the duration of exposure has nearly no intluence un the degree of annoyance. A third series of laboratory experiments (14, 16, 18) was carried out to deter- mtne which tobacco smoke components are mainly responsible for irritation and Annuyance. By comparing the degree of irritation and annoyance due to acrolein rJunC, formaldehyde alone, and gas phase alone with that of sidestream smoke o a whole, it was found that: (u) Acrolein and formaldehyde are only minimally responsible for the irritations Juc to the sidestream smoke of cigarettes, ib) The gas phase is, to a large extent, responsible for the annoyance due to thc ,idestream smoke of cigarettes (see Fig. 5). lc1 The particulate phase is, to a very large extent, responsible for the irritating atcct ut'sidestream smoke, since eye blink rate (see Fig. 6) as well as subjective ® ® ® TI BU 31617 ® a :: 4 0 © ® e

SYMPOSIUM: MED[CAL PERSPECTIVES ON PASSIVE SMOKING 621 aye irritation •ye blink rataimin ,«y naeng S -1 r• 80 r 60 r 40 20 rkplace. eatl. 0. on the La were ri tation ialyzeJ -ealistic t to to- o these ing the naining ,r com- tfas•rn none I 'J ~ ~' 0 0 10 20 mm ACO 1 1/ 22 ]2 42 43 PPT ONO 0.0! 0.42 0.77 1,11 1.43 1.30 OYaM A NCNO 0.03 0.10 0.32 0.47 0.d2 0.64000 pwo/ou 0 0.06 0.11 0.1• 0.20 0.20ppin numtNr o/ c10. 0 10 20 F1c:. 2. Mean subjective eye irritations, mean eye blink rate, and concentrations of some pollutants Junng continuous smoke production in an unventilated climatic chamber, 33 subjects. Ventilation we. 0.01 H''. Eye irritation index calculated from the answers to four qubstions concerning eye irrntatwns. 0 min: Measurement before smoke production. -. Eye irritation index; •••, eye blink rate. indicator to evaluate the degree of air pollution due to environmental tobacco• tmuke. Figures 3 and 4 illustrate the results obtained for subjective eye irritation rnd eye blink rate. Two facts are obvious: (a) Mean eye irritation as well as eye blink rate increase with increasing smoke wncentration. (b) Both irritation parameters increase with the duration of exposure in spite of a constant smoke concentration. The same, but less pronounced result has Alw been observed for nose and throat irritations. The pattern of annoyance differs from that of irritation: it increases rapidly as wun as smoke production begins and, after 10 to 15 min of exposure, remains Approximately constant. Thus, the duration of exposure has nearly no intluence un the degree of annoyance. A third series of laboratory experiments (14, 16, 18) was carried out to deter- mtne which tobacco smoke components are mainly responsible for irritation and Annuyance. By comparing the degree of irritation and annoyance due to acrolein rJunC, formaldehyde alone, and gas phase alone with that of sidestream smoke o a whole, it was found that: (u) Acrolein and formaldehyde are only minimally responsible for the irritations Juc to the sidestream smoke of cigarettes, ib) The gas phase is, to a large extent, responsible for the annoyance due to thc ,idestream smoke of cigarettes (see Fig. 5). lc1 The particulate phase is, to a very large extent, responsible for the irritating atcct ut'sidestream smoke, since eye blink rate (see Fig. 6) as well as subjective ® ® ® TI BU 31617 ® a :: 4 0 © ® e

15.0 ° 0 ~ ~ Y > W t p = C O a, Y ~ O 0.0 a I 0 1.3 2.5 5 10 ppm..CO Fte. 7. Percentage of persons with strong or very strong eye irritation reactions retued to the degree and duration of exposure. 32 to 43 subjects. •, 60 min expusure: Ir, 40 min exposure. of passive smokers in their everyday environment. Furthermore, it has to be pointed out that the results presented apply only to conditions of relative humidity between 40 and 50%. Since irritating effects are known to be enhanced in a dry environment (5), the threshold liuiit values should perhaps be slightly lower for naturally ventilated rooms in winter. Based on the above-mentioned threshold limit proposal, countermeasures to protect passive smokers are desirable when the ACO level reaches 1.5 ppm and are necessary when it hits 2.0 ppm. It is suggested that the lower limit be applied to workplaces where it is difficult for passive smokers to escape exposure, and the upper limit to restaurants and those places where people usually go voluntarily and for shorter periods of time. The upper concentration limit of 2.0 ppm ACO is reached when two cigarettes are smoked per hour in a room of 80 m3 with only one air exchange. Calculations (l2) show that a fresh air supply of 33 m3 per hour per cigarette smoked is necessary to keep the ACO concentration below the proposed upper limit of 2.0 ppm; for the lower limit, 50 m~ per hour per cigarette smoked is required. Depending on the number of persons present in a room, a fresh air supply of 25 to 45 m3 per hour per person is necessary to keep the ICO concen- tration below the upper limit. In other words, the ventilation in a room with smokers has to be two to four times greater than in a room where nobody smokes [in which only 12 to 15 m3 per hour per person is required (4)]. For that reason, increased ventilation as a measure to protect passive smokers is questionable from the point of view of energy conservation. Therefore, whenever possible, organizational measures, such as separation into smoking and nonsmoking roums or a prohibition againt smoking, rather than an increased fresh air supply, should be taken into consideration. It must be emphasized again that these considerations of threshold limit values TI BU 31620 and : At p aens !, uttrU hCna not ~ 3. C 5. J, b. L 7. N y. ,

622 3 ,,,,.-,,.- 10 ppm s C O 1 ANNETrA WEBER 0 10 20 30 40 50 B0tmon1 Exposure Duration FiG. 3. Mean subjective eye irritations related to smoke concentration (expressed as aCO in ppm) and duration of exposure. 32 to 43 subjects. 0 min: Measurement before smoke production. Period U to 5 min: Increasing smoke concentration. Period 6 to 60 min: Cunstant smoke concentration. eye irritations are much lower with the gas phase alone than with the total side- stream smoke. Tolerable Environmental Tobacco Smoke Concentrations and Protective Measures The mean incidences of strong and very strong subjective eye irritations after exposure to environmental tobacco smoke of 40 and.60 min duration, respectively, are reported in Fig. 7. [Results from the previously mentioned paper by Mura- matsu et al. (7)]. 10 pwn .co v i ~ / '"--- a oom .co / .. cO s pwM . . / ~ 2. I 1.3ppm ACO / FiG. 4. Mean effects on eye blink rate. See Fig. 3 for details. TI BU 31618 Flc expoa, culutC MeaIL bits pc Th smuh reser a ten ~,hou. propk

622 3 ,,,,.-,,.- 10 ppm s C O 1 ANNETrA WEBER 0 10 20 30 40 50 B0tmon1 Exposure Duration FiG. 3. Mean subjective eye irritations related to smoke concentration (expressed as aCO in ppm) and duration of exposure. 32 to 43 subjects. 0 min: Measurement before smoke production. Period U to 5 min: Increasing smoke concentration. Period 6 to 60 min: Cunstant smoke concentration. eye irritations are much lower with the gas phase alone than with the total side- stream smoke. Tolerable Environmental Tobacco Smoke Concentrations and Protective Measures The mean incidences of strong and very strong subjective eye irritations after exposure to environmental tobacco smoke of 40 and.60 min duration, respectively, are reported in Fig. 7. [Results from the previously mentioned paper by Mura- matsu et al. (7)]. 10 pwn .co v i ~ / '"--- a oom .co / .. cO s pwM . . / ~ 2. I 1.3ppm ACO / FiG. 4. Mean effects on eye blink rate. See Fig. 3 for details. TI BU 31618 Flc expoa, culutC MeaIL bits pc Th smuh reser a ten ~,hou. propk

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 627 :» ,P- rs. -Icrt hy o- ms letl to ur ule or- here is Aester- of the )f non- sponse -eases, s to in- iophys- may be effects ironary ions. In >genesis id. long- the dis- less zx- paasivc :. VICIIn:1. smoking were observed (27). But, in 1983, Hirayama (36) reported a signiticant risk elevation for ischemic heart disease in nonsmoking women married to smokers. The author offered a combined action of CO and nicotine as a possible explanation. In this report, we will try to summarize the present state of research into the influence of passive smoking on the development of CHD. The concentrations of CO and nicotine in indoor air and their uptake by passive smoking as described in the relevant literature will be reviewed. Possible cardiovascular effects of these substances in concentrations found in passive smokers will be discussed. We have omitted thrombangiitis obliterans, as it does not seem to be related to CHD. CARBON MONOXIDE AND NICOTINE CONCENTRATIONS IN INDOOR AIR Numerous studies have focused on CO and nicotine, both of which, depending on their concentrations, may affect the cardiovascular system (51, 52). Since other tobacco smoke constituents are not yet known to have such an effect, we contine ourselves to discussing these two substances. The fact that CO originates from a number of sources apart from tobacco smoke must be taken into consideration when evaluating CO concentrations in indoor air. The mean concentration of CO arising from tobacco smoke is assumed to amount to less than 5 ppm (24, 29, 66), although maximum values of approximately 40 ppm may occur for short periods under extreme conditions (32, 33). Nicotine is an alkaloid specific to tobacco; its indoor air levels can thus be regarded as, in fact, originating from tobacco smoke. The relevant literature cites levels of between 0.9 (71) and 3100 µg/m3 (33). Ac- tually, nicotine concentrations in rooms with smokers are estimated to range from 5 to 50 µgJm3 (8, 35, 71). CARBON MONOXIDE AND NICOTINE UPTAKE BY PASSIVE SMOKING Measuring CO, nicotine, and nicotine's main metabolite, cotinine, in blood and other bodily fluids is a better method of estimating their biological activity than drawing conclusions from indoor air concentrations. We are, however, faced with uncertainty as to the origin of CO in blood. In addition to various exogenous suurces, there is a permanent endogenous CO formation in the human organism. Therefore, in order to measure accurately the CO uptake by passive smoking, carbuxyhemoglobin (COHb) concentrations have to be quantitied under well- Jctined conditions before and after exposure. The results obtained on the basis of these requirements are shown in Table 1. Taking into account a CO content b,,luw S ppm in indoor air originating from tobacco smoke, the COHb amount is cstimated to increase over the day by less than 1%. Nonsmoking subjects, under rc:,l-life conditions representing the average German population, reveal COHb levels ranging from t to 1.5% (Table 2). In comparisun, cigarette smokers are tuund to have COHb levels between 3 and 10% during the day (59). Nicotine uptake from passive smoking has been assessed by measuring nicotine And cutinine concentrations in serum, saliva, and urine. Table 3 summarizes the Jat:e obtained under experimental conditions. Due to varying test procedures, they are not absolutely comparable. Nevertheless, they demonstrate that urinary c.cr6tion of nicotine and cotinine is several times higher in passive smokers TI BU 31623

15.0 ° 0 ~ ~ Y > W t p = C O a, Y ~ O 0.0 a I 0 1.3 2.5 5 10 ppm..CO Fte. 7. Percentage of persons with strong or very strong eye irritation reactions retued to the degree and duration of exposure. 32 to 43 subjects. •, 60 min expusure: Ir, 40 min exposure. of passive smokers in their everyday environment. Furthermore, it has to be pointed out that the results presented apply only to conditions of relative humidity between 40 and 50%. Since irritating effects are known to be enhanced in a dry environment (5), the threshold liuiit values should perhaps be slightly lower for naturally ventilated rooms in winter. Based on the above-mentioned threshold limit proposal, countermeasures to protect passive smokers are desirable when the ACO level reaches 1.5 ppm and are necessary when it hits 2.0 ppm. It is suggested that the lower limit be applied to workplaces where it is difficult for passive smokers to escape exposure, and the upper limit to restaurants and those places where people usually go voluntarily and for shorter periods of time. The upper concentration limit of 2.0 ppm ACO is reached when two cigarettes are smoked per hour in a room of 80 m3 with only one air exchange. Calculations (l2) show that a fresh air supply of 33 m3 per hour per cigarette smoked is necessary to keep the ACO concentration below the proposed upper limit of 2.0 ppm; for the lower limit, 50 m~ per hour per cigarette smoked is required. Depending on the number of persons present in a room, a fresh air supply of 25 to 45 m3 per hour per person is necessary to keep the ICO concen- tration below the upper limit. In other words, the ventilation in a room with smokers has to be two to four times greater than in a room where nobody smokes [in which only 12 to 15 m3 per hour per person is required (4)]. For that reason, increased ventilation as a measure to protect passive smokers is questionable from the point of view of energy conservation. Therefore, whenever possible, organizational measures, such as separation into smoking and nonsmoking roums or a prohibition againt smoking, rather than an increased fresh air supply, should be taken into consideration. It must be emphasized again that these considerations of threshold limit values TI BU 31620 and : At p aens !, uttrU hCna not ~ 3. C 5. J, b. L 7. N y. ,

620 Question Answer Air aualltv r+ttn re9ard to snlbke Dtsturbed by smoke never ANNETrA WEBER 9ood bad lntervtewed aersons 0 20 40 so so x somettmes/ often 21 ''///////////~///. %////// ,. , ~ ~/" ® s. .. .a nciw a Eye trrlta- ttons at work sorrst 1 mes/ often 0 I Fle. 1. Evaluation ot air quality and effects Jue to environmental tobacco smoke at the workplace. Reaults uf 472 employees in 4-1 workrooms. "-Oifference significant with P < 0.01 (Y'- tcat). u. Nonsmokers + exsmokers (N = 237); M, smokers (N = 235). Luburutvey Studies , !n a first experiment (15), 33 subjects were exposed to continuously increasing smoke concentrations. The main results are summarized in Fig. 2. The con- centrations of CO, NO, formaldehyde (HCHO), and acrolein increase linearly with the number of cigarettes smoked. Buth tncan subjective eye irritation and mean eye blink rate increase in a nearly linear way with increasing smoke cun- centratiun. j Subjective nose and throat irritation were also determined. The effects on the nose were less pronounced than those on the eyes, and those on the throat were the ~,mallest of all three. Thus, eyes are the organs most sensitive to irritation due to passive smoking. In a second series of laboratory experiments (7, 17), acute effects were analyzt:d in relation to smoke concentration and duration of exposure. The chosen realistic smoke concentrations corresponded to 1.3, 2.5, 5.0, and 10 ppm CO due to tu- bacco smoke (JCO) . During a l hr period, 32-43 subjects were exposed to theao smoke concentrations. Each smoke concentration increased linearly during the tirst 5 to 10 min, and was then held constant at the desired level for the remaining part of the hour. Since very high correlations (r > 0.9) between CO and each of the other cum- pounds were obtained in the tirst experimental series, only JCO was used as an TI BU 31616 Fic. 2_ during cu rute. 0.01 trruattuns indicstu hmuke. and eye (a) M concent, (b) 8, Ut il Cur alsu bee The p, wun aa rpproxir un tht: tlL A thirL tnine wht 4n1110yanc. .tlunc:, fu, .a. i whu l,tl Ac due to tF (b) Th ttlc iidea (c) Th cliet;t ut

Authors Harke. 1970 (28) Russell and Feycrabcnd, 1975 (561 Cana rr al., TAHLE 3 - NICOTINk UPTAKt OI NONSMOKEKS EXPOSED 70 TOtlACCO SMOKE UNDtn DLFINEU CONDITIONS No. of nonsmokers Exposure conditions investigated Results ta) Experiment 1: Room /170 7 Nicotine excretion in urine until 6 hr after exposure m31; 7 nonsmokers; 11 10 a 6.8Lg/6hr smokers smoked a total of Cotinine excretion about 100 cig:utueh: duration 35 ± 34.5 µg/6 hr of exposure, 2 hr; no ventilation; indoor air concentration, 30 ppm CO. (b) Experiment 3: Same as 7 Nicotine excretion in urine until 6 hr after exposure Experiment I but with fair 18 t 7 µgl6 hr. ventilation (according to Cotinine excretion German stand:u.l for industry, . 19 s 9.4 µg/6 hr 33.4 m3/person hrr, indoor air concentration. 5 ppm CC). Room (43 nl)); 12 nonsmokers; 9 12 Nicotine in plasma (ng/ml) smokers; altogether 80 Before: 0.73 --t 1.6 cigarettes and 2 cigan were After: 0.90 -t 11.29 smuked; duration of exposure, Nicotine in urine (15 min aftCr exposure) 78 min; no ventilation; indoor 80 = 58.7 air concentration, 38 ppm CO. (a) Experiment 1: Room (66 mj); ~ Nicotine excretion in urine tµg/?4 hr) 1970 (13) 2 nonsmokers and 4 smokers lived together for 5 days; ventilation: 90 mtlhr. D.q I (without Smoking): 0 Day 2 (98 cig): 35-44 Day 3 (121 cig): 50-61 Day 4 1'I-l c ig l: 62 S- 711 Da~ 5 (NK cig): 47-SO Nici,unr r+,rt.•tium m 11iinC

SYMPOSIUM: MEDICAL PERSPECTiVES ON PASSIVE SMOKING 625 Lsures to ppm and ~ applied ure, and jluntarily• pm _~CO .vith only cigarette zd upper nuked ih trehh air ,cunct:n- )utn with y smokes .t reasun, stiunablC possible, ng rooma y, should 1it values and the resulting requirements of fresh air supply are valid only for healthy adults. At present, only little is known about the acute effects of passive smoking on sensitive or at-risk groups such as asthmatics. allergic persons, chronic bronchitis sufferers, and children. Several studies (1-3, 6, 8, 10, ll) indicate a greater sensitivity of these groups to environmental tobacco smoke, but other studies do not (9). Further systematic research in this field is needed. REFERENCES I. Culley, J. Passive smoking in children. Nursing Times 71, 1858-185911975). 2. Dahms, T. E.. Bolin. J. F., and Slavin, R. G. Passive smoking: Effects on bronchial asthma. Chest 80, 530-534 (1981). 3. Gortmakar, S. L.. Walker. D. W., Jacobs. F. H., and Ruch-Ross, H. Parental smoking and the risk of childhood asthma. Arner. J. Public Heulth 72, 574-579 (1982). 4. Huber, G.. and Wanner, H. U. Indoor air quality and minimum ventilation rate. Environ. /nt. 9, 153- l5b (1983). 5. Johansson. C., and Ronge, H. Climatic influence on smell and irritation effects from tobacco smoke. Nurd. Hyg. T. 47, 33-39 (1966). 6. Leederer, S. R.. Corkhill. R.. Irwig, L. M., and Holland. W. W. Influence of family factors on asthma, wheezing incidence of lower repir.uury illness, and ventilation function in children. Brit. J. Prev. Suc•. Med. 30, 203-224 (1976). 7. Muramatsu, T.. Weber. A.. Murumatau. S., and Akermann. F. Experimental study on irritations and annoyance due to passive smoking. hu. Arc•h. O(uup. Environ. Health 51, 305-317 (1y83). 8. Shephard. R. J., Ponsfurd. E.. LaBarre. R., and Ba;u. P. K. Effects of cigarette smoke on the eyes and airways. Int. Arch. Occup. Environ. Hrult/r 43, 135-I44 11979). 9. Shephard, R. J., Collins. R.. and Silverman. F. Passive exposure of asthmatic subjects to cigarette smoke. Environ. Res. 20. 392--W2 (1.979). 10. Speer, F. Tobacco and the nonsmoker. Arch. Environ. Healtlr 16, 443-446 (1968). 11. Tager. 1. B.. Weiss. S. T.. Rosnzr. B.. and Speizer. F. G. Effect ot' parental cigarette smoking on the pulmonary function of children. Amer. J. Epirlenriul. 110, 15-26 (079). 12. Weber. A. Luftungsmassnahmen zunt Schutze der Passivraucher. Huusrechnik-Buuphysik-Um- we/ttechnik-Gesundheits-ingenieur 104, 37-42 (1983). 13. Weber, A.. and Fischer. T. Passive smoking at work. Int. Arch. Occup. Em•i.un. Health 47, _09- 11_t (1980). 14. Weber, A.. Fischer. T.. Gierer. R., and Grandjean. E. Experimentelle Reizwirkungen von Ak- rolein auf den Menschen. Int. Arch. Occup. E'rn•rrun. Heulth 40, l 17-130 (1977). 15. Weber. A.. Fischer. T., and GrJndjean. E. Objektive und subjektive physiulogische Wirkungen Jeo P.rsaivrauchertb. Int. Arch. Ovcup: Environ. Health 37, 277-288 (1976). 16. Weber. A., Fischer, T., and Grandjean. E. Reizwirkungen des Furmaldchyds auf den Menschen. /ru. Arch. Occnp. Envirun. Health 39, 207-218 11977). 17. Weber, A., Fischer. T., and Grandjean, E. Passive smoking in experimental and tield conditions. tnvirun. Res. 20. 205-216 (1979). I8. Weber. A.. Fischer. T.. and Grandjean, E. Passive smoking: Irritating effects of total smoke and gaa phase. Int. Arch. Ocrup. Environ. Health 43, 183-193 (1979). TI BU 31621

PREVEtvTIVE MEUICINE 13, 626-644 (1984) The Influence of Passive Smoking on the Cardiovascular System' H. SCHIEVELBEIN AND F. RiCHTER Institute o/'Clinical CJtemislry, Gerinun Heart Center, Luthstrusse !!, B(111U Munich 2. West Cierrnuny Cardiovascular effects of tobacco smoke have been studied in passive smokers far tehs extensively than in active smokers. Under real-life conditions, passive smokers inhale ap- proximately 0.02 to 0.01 ut' the amount of particulate matter taken up by active smukere. Their nicotine concentration in serum is within a range that is barely distinguishable from the background level. The increase in carboxyhemoglobin rarely exceeds 1%. In healthy subjects heavily exposed to tobacco smoke, no or only slightly acute effects on the cardio- vascular system are found. Whether or not passive smoking is likely to aggravate symptoms in patients with advanced coronary heart disease has not yet been unequivocally established and requires further investigation. From a few studies on occupatiunal groups exposed to carbon monoxide (CO) and from experiments with animals chronically treated with CO or nicotine, the conclusion can be drawn that neither CO nor nicotine is likely to play a rolo in the development and progression of coronary heart tlisease in those concentratiuns nur- mally found in passive smokers. a 1964 nc;,uemtc Press. Inc. INTRODUCTION The pathogenesis of coronary heart d'isease (CHD) remains unclear. There is some evidence, however, that a number of risk t'actors (such as hyperchofester- olemia, hypertonia, cigarette smoking) are involved in the development of the disease. The risk of cigarette smokers dying from CHD is twice that of non- smokers. In heavy smokers it is even higher, thus demonstrating a dose-respomie relationship. After smoking has been given up, the risk rapidly decreases, reaching the rate ut' nonsmokers after several years (51). Though cigarette smoking is known to interact with other risk factors to in- crease the rate of cardiovascular morbidity and mortality, the precise pathophys- iological basis of these clinical manifestations is nut understood (44). It may be related to several cardiovascular•effects of cigarette smoking. Short-term effects mainly attributed to carbon monoxide (CO) and nicotine may cause coronary heart infarction and sudden death in the presence of atherosclerotic lesions. In addition, cumulative short-term effects may accelerate the pace of atherogenesis by other mechanisms not yet known. Consequently, both short-term and long- term effects of cigarette smoking may contnbute to the pathogenesis of the dis- ease. The cardiovascular effects of passive smoking have been investigated less ex- tensively. In earlier studies, no chronic cardiovascular consequences of passive ' Presented at the Symposium `',Medical Perspectives on Passive Smoking," April y-l=, Vienna. Austria. 626 11091-7435/84 33.00 CoP)n6ht s: 14114 by AC.ulemlc Prena. lnc. AIi nghta ui rcpruJucuun m any lurm reaerved. TI BU 31622 hmul ri4k ~lmu cxpl tn intlt. CO in tt aub: umi CA ti utl t tubz uur: nun wht artS aith unu inci, Tht tua 5 tL , uth dru unL ~.uL Th car det, of t bei,_ dsit. real lCvr tut da tht ex

PREVEtvTIVE MEUICINE 13, 626-644 (1984) The Influence of Passive Smoking on the Cardiovascular System' H. SCHIEVELBEIN AND F. RiCHTER Institute o/'Clinical CJtemislry, Gerinun Heart Center, Luthstrusse !!, B(111U Munich 2. West Cierrnuny Cardiovascular effects of tobacco smoke have been studied in passive smokers far tehs extensively than in active smokers. Under real-life conditions, passive smokers inhale ap- proximately 0.02 to 0.01 ut' the amount of particulate matter taken up by active smukere. Their nicotine concentration in serum is within a range that is barely distinguishable from the background level. The increase in carboxyhemoglobin rarely exceeds 1%. In healthy subjects heavily exposed to tobacco smoke, no or only slightly acute effects on the cardio- vascular system are found. Whether or not passive smoking is likely to aggravate symptoms in patients with advanced coronary heart disease has not yet been unequivocally established and requires further investigation. From a few studies on occupatiunal groups exposed to carbon monoxide (CO) and from experiments with animals chronically treated with CO or nicotine, the conclusion can be drawn that neither CO nor nicotine is likely to play a rolo in the development and progression of coronary heart tlisease in those concentratiuns nur- mally found in passive smokers. a 1964 nc;,uemtc Press. Inc. INTRODUCTION The pathogenesis of coronary heart d'isease (CHD) remains unclear. There is some evidence, however, that a number of risk t'actors (such as hyperchofester- olemia, hypertonia, cigarette smoking) are involved in the development of the disease. The risk of cigarette smokers dying from CHD is twice that of non- smokers. In heavy smokers it is even higher, thus demonstrating a dose-respomie relationship. After smoking has been given up, the risk rapidly decreases, reaching the rate ut' nonsmokers after several years (51). Though cigarette smoking is known to interact with other risk factors to in- crease the rate of cardiovascular morbidity and mortality, the precise pathophys- iological basis of these clinical manifestations is nut understood (44). It may be related to several cardiovascular•effects of cigarette smoking. Short-term effects mainly attributed to carbon monoxide (CO) and nicotine may cause coronary heart infarction and sudden death in the presence of atherosclerotic lesions. In addition, cumulative short-term effects may accelerate the pace of atherogenesis by other mechanisms not yet known. Consequently, both short-term and long- term effects of cigarette smoking may contnbute to the pathogenesis of the dis- ease. The cardiovascular effects of passive smoking have been investigated less ex- tensively. In earlier studies, no chronic cardiovascular consequences of passive ' Presented at the Symposium `',Medical Perspectives on Passive Smoking," April y-l=, Vienna. Austria. 626 11091-7435/84 33.00 CoP)n6ht s: 14114 by AC.ulemlc Prena. lnc. AIi nghta ui rcpruJucuun m any lurm reaerved. TI BU 31622 hmul ri4k ~lmu cxpl tn intlt. CO in tt aub: umi CA ti utl t tubz uur: nun wht artS aith unu inci, Tht tua 5 tL , uth dru unL ~.uL Th car det, of t bei,_ dsit. real lCvr tut da tht ex

, 3 6 lo UY ) 5 1 3 ndi- .~r~. ~~zd tt uf ) unt! SYMPOSIUM: IMEDICAL PERSPECTIVES ON PASSIVE SMOKING 629 TABLE 2 COHb IN NONSMOKEiiS UNDEN REAL-LIFE CONDITIONS No. of nonsmokers Authors Subjects (exposure) investlgated Mean 1: SID) Range Waid rr ul.. Participants in a hualth 6.641 97% 1981 169) screenmg program 0.0-1.5 Carsky rr u1.. Blood donors in Bratislava 130 U.9 U.3-3.7 1'lttl/ 1141 lanzun r( ul.. il 1981 Participants in a health 509 Median: 0.5 0-2 ) ( Stewurt et u/.. screentng program Blood donors in United States 16.Ik{2 1.42 (= 0.23) 90% 1975 1641 0.4-3.7 Stewurt rr ul.. (a) Blood donors in Chicago 406 2.04 90% 1976 (65) 1970 1.0-3.2 (b) Blood donors in Chicago 426 1.53 vll% 11)74-1975 0.7-2.4 Einbnxlt et ul.. Total adult inhubitants uf a 103 0,71 1= 0.36) I97n (.0) small village Cule. 1975 lal Subjects working in a•'CO- 78 0.68 0.2-2.6 1151 I'ree" atmosphere I~- I ppm CU/ (bt Subjects working in a huapital where smuking was not allowed 1:-•t ppm CO) 120 0.97 0.2-2.5 tct Subjects working in an uflice where ,muking was 1(!0 1.12 0.1-2.7 alluwcu 13-iS ppm CO) Stadkuw,kt rr al.. lal Subjects working in an 130 Morning: 0.82 I= 0,37) 1976 l6bl utlice with emukers present Ibl Subjects working in an 118 il Afternoon: 0.63 (= 0.3U1 Morning: 0.92 t= 0.40) ullice wuhuut smuken 37 Afternuun: 0.72 I s 0.25) in our laboratory (Table 3). On the basis of their own tindings, Jarvis et a!. (42) calculated urinary nicotine concentrations of about 7.5% for passive smokers and urinary cotinine concentrations ranging from 0.9 to 2.4% of the levels found in hmukers (Table 4). EFFECTS OF CARBON MONOXIDE, NICOTINE, AND CIGARETTE SMOKE ON THE CARDIOVASCULAR SYSTEM OF PASSIVE SMOKERS As already mentioned, CO and nicotine have been proposed as the components of cigarette smoke that may be responsible for the development and progression of CHD in smokers (5l). Carbon monoxide exposure causes the same symptoms or disturbances, during rest or exercise, as those encountered in low oxygen or hypobaric hypoxia (67). In healthy individuals at rest, alterations in physiological lunctions such as change in pulse rate, cardiac output, or blood pressure are not u,ually found at levels uf COHb below 20% (60). During exercise, while breathing (;(). there is a decrease in the maximum uxyguu uptake and an increase in heart ratt: and ventilation. Adaptive metabolic changes may occur in response to chron- tt:ally increased COHb concentrations resembling the processes in subjects who Ilve at high altitudes (57). It is not certain whether the advantages of such an ,uljustment outweigh possible disadvantages. TI BU 31625

Aulhurs . W A TABLE 5 CHNC)NIC EttkCll Uf CO IN ANIAtALS Species Exptrimtntal conditions Results EcAurdl rt ul., 197209) Monkey (n = 27) Exposure level: 20 ppm (and 67.5 ppnl) Exposure tirne: 22 hrldal. 7 dayslweck Heart wt (g) Exposed: 15.9 = 1.4(14.6 •_ ?3) • fur 2 years Control: 16.6 = 4.6 Mean COHb: I - 1. 5e7 (6-B~lc ) SGOT, SGPT, LHll, alkaline phusphatase No dose-related differences Histology of the heart No differences Armitage e•l ul., 1970(3) Pigeon (n = 180) Expuswe level: 150 ppm CO Exposure time: 6 hNday, 7 days/wrrk Serum cholesterol (mg/100 nil) Exposed: 321 zt 13 for 84 weeks Mean ('UI Ib: 10-14% Control: 326 t 12 Aurtic sudanuphilia (70 Exposed: 10.0 s 1.5 Control: 6.3 = 0.9 Affected coronary arteries (9r) Exposed: 5.3 i 1.3 Control: 3.4 * 1.1 Hugod el uf., Rabbil Exposuee level: 200 ppm 1978 (391 (n = 60) Exposure lime: 24 hr/day, 7 daysMeek '% of animals with intimal changes k•xposed Cuntrul fur 12 weeks Coronary artery 32 55 Mean CC)Hb: i0-ISSi Aurtic arch 74 75 Hutulugiical criteria Upper thoracic aorta 62 63 Subcndulhelial vacuules Lower thoracic aorta 62 1--1 Subcn.luthclial celcma furmmg arc:rdes Abduminal aorla 65 53 Srtrrr' tiuhrnJoll c h.d eJcvn.r r>r..+.c .... . . ~ -

Aulhurs . W A TABLE 5 CHNC)NIC EttkCll Uf CO IN ANIAtALS Species Exptrimtntal conditions Results EcAurdl rt ul., 197209) Monkey (n = 27) Exposure level: 20 ppm (and 67.5 ppnl) Exposure tirne: 22 hrldal. 7 dayslweck Heart wt (g) Exposed: 15.9 = 1.4(14.6 •_ ?3) • fur 2 years Control: 16.6 = 4.6 Mean COHb: I - 1. 5e7 (6-B~lc ) SGOT, SGPT, LHll, alkaline phusphatase No dose-related differences Histology of the heart No differences Armitage e•l ul., 1970(3) Pigeon (n = 180) Expuswe level: 150 ppm CO Exposure time: 6 hNday, 7 days/wrrk Serum cholesterol (mg/100 nil) Exposed: 321 zt 13 for 84 weeks Mean ('UI Ib: 10-14% Control: 326 t 12 Aurtic sudanuphilia (70 Exposed: 10.0 s 1.5 Control: 6.3 = 0.9 Affected coronary arteries (9r) Exposed: 5.3 i 1.3 Control: 3.4 * 1.1 Hugod el uf., Rabbil Exposuee level: 200 ppm 1978 (391 (n = 60) Exposure lime: 24 hr/day, 7 daysMeek '% of animals with intimal changes k•xposed Cuntrul fur 12 weeks Coronary artery 32 55 Mean CC)Hb: i0-ISSi Aurtic arch 74 75 Hutulugiical criteria Upper thoracic aorta 62 63 Subcndulhelial vacuules Lower thoracic aorta 62 1--1 Subcn.luthclial celcma furmmg arc:rdes Abduminal aorla 65 53 Srtrrr' tiuhrnJoll c h.d eJcvn.r r>r..+.c .... . . ~ -

Suhcuciuthelwl vactlulcs Suhrnduthchal edetpa (orminE arcaJes • Sc.cic .uhcnJ.~thcl~.~l cJcm.. scp:1raung the mumw from thc undcrlymE tissue t'resence of smooth muscle cells in Ibe subcnJothrlial spuce Fiagmcntation of the internal elastic membrane Munkey Expusurr level: up to 400 ppm CO (,t = 11) Eapusme time: 10 x 30 min/day for 12 months Monkey Exposure: 43 cig/day (,t = 36) Exposure time: 14- 19 months CO level in blood: 0.24 m1/dl Animals were conditioned to cigarette smoking. t.amcl Ihot alta Ahduminal aona b5 53 Serum levels tµmul/mU Ex useJ Cuntrul Free cholesterul 0.60 Y 0.08 0.55 *_ (1.(18 Cholesterol ehlers 1.74 = 0.24 1.63 T 0,18 Aor1iL cholesterul content (Nmullg) Cholesterol Thoracic aorta 4.34 Y 0.61 3.95 s 0,58 Abdominal aorta 4.98 ± 1.'_0 3.06 s 0.29 Chulesrerinesler Thoracic aorta 0.77 T 0.12 0.71 ~ 0.(16 Abdominal aorta 0.69 z 0.06 0.76 ~ 0.09 Histology of the aorta No structural damages or fat depusirion Serum levels (ng/afl/ Smokers Coouul 7otal cholesterol 171.6 162.9 7otal tnglycerlde 41.3 41.2 Hht. cholesterol 83.(t H8.1 VLDL + U)L 78.3 66.4

11,1111.1 ~ ~r.rlr..rrr..n~r.~ r.1 Y ]IlLrnl.1 ~ lived togcther fot 5 days; ventilation: 90 m'/hr. 0 Hoflmann rt ul.. 1983 (31) (bl kxpaiment'_: Same a+ kxperimcnt I: subjects lived together for 4 llays : indoor air nicotine concentration was measured. Chamber (16 m3): 2 nonsmokers: 4 cigarettes Nere smokeJ continuously by machine; duration of exposure, 80 min; ventilation, 6 air changes/hr. indoor air concentration, 4.6 mg TPM/m3, 280 I,g nicutine/ 1113, 56 µg HCN/m'. 25 ppm CO. (The experiment was repeated three times with different subjects.) I 6 . Ii..~ 1 1. r........ ~.... .....br ~ Day 2 (98 cig): 35-44 Day 3 (121 cig): 5(i-61 1).,} 4 tva iigI t.' 1. -7U D.o S tt<r cigl. a7-tu Nicotine excretion in urine Day 1(97 cig. 15 µg mc/nH'1: 23-34 Ng%'_4 hr Day 2 196 cig. '_2 µg nic/m'): 22.5-58 µgI24 hr Day 3(9a cig. 35 pg nic/m'): 47.5-69 µg/24 hr Day 4 (103 cig. 33 µg nic/m'): 32-65 µg/24 hr Nicotine in plasma (ng'mll Before: - After: 0.5 Cotinine in plasma (ngAnU Uefore: 1.0 2.5 hr after expusure: 2.9 Nicotine in saliva (ng/m) Before: 3 After: 730 Cotinine in saliva (ngJml) Before: 1.3 2.5 hr after exposure: 3.1 Nicotine in urine (ng/ml) Qefore: 18 2.5 hr after exposure: 86 Cotinine in urine (ng/ml) Before: 17 2.5 hr afier exposure: 42

632 SCHIEVELBEIN AND RICHTER In cigarette smokers, the average COHb levels are in the range of 4 to 6%. Concentrations of more than 12% are rarely found (59). Nonsmoking bridge and tunnel workers, blast furnace workers, and traftic officers chronically exposed to CO were found to have COHb levels between 2 and 3%. The rate of CHD in these groups was not higher than that of nonsmoking workers who were not so exposed (7, 43, 63). [n contrast to healthy individuals, in patients with coronary artery disease, an acute 2% increaNc in COHb may reduce the exercise tolerance until the onset of anginal pain by aggravating the ischemia of the heart muscle (2, 4, 5). Based on the theory that CO is an important factor in the pathogenesis of CHD, numerous animal experiments have been conducted (Table 5). The results of these studies thus far have neither proved nor disproved an atherogenic effect of CO. By reevaluating their own findings (6), Hugod and co-workers came to the fol- lowing conclusion: "Even when rabbits are exposed to a lethal level of CO, the criteria hitherto believed to be indicators of CO-mediated endotheliai damage in normucholesterolemic rabbits appear not to be valid in the present study, leaving us with the problem that we-so far-do not know which criteria to apply, whcn trying to estimate a toxic action of CO on the intima of normocholesterolemic animals, and attempting to relate this to early atherosclerosis" (39). The data obtained in rabbits were confirmed in studies with primates whose metabolic status is more comparable to that of humans. Murucu fascicularis chronically treated with CO were found to show neither structural lesions nor lipid deposi- tions in the arteries (10). Ln a similar experiment with baboons, the cholesterol and triglyceride concentrations, which are thought to be related to the develop- ment of arteriosclerosis, were not increased (53). There have been no experimental investigations into the acute effects of nic- otine on the cardiovascular system in concentrations such as those found in pas- sive smokers. [f one extrapolates from the dose-effect relationship in smokers, no measurable acute effects of nicotine inhaled by passive smokers should be expected. Several animal experiments have been pertormed to study the chronic effects of nicotine (Table 6). Their results, again, neither prove nor disprove an atherogenic effect of nicotine. ln a few experimental studies, the influence of cigarette smoke on the cardiu- vascular system was measured directly (Table 7). Although exposure in these studies was much higher than'under real-life conditions, an increase in heart rate and blood pressure was either undetectable or only marginal. As shown in astuJy by Harke and Bleichert (30), the skin temperature, which decreases considerably while smoking one cigarette, remained constant. DISCUSSION For more than 25 years cigarette smoking has been associated with an elevated risk of death from CHD. This has been confirmed by numerous epidemiological studies. Yet, the mechanism by which smoking might increase the risk of deve- loping cardiovascular diseases is not understood. A large number of chemicals have been identitied in tobacco smoke (18), some of which may be implicated in the development of cardiovascular diseases. Interest primarily has concentrated ~ 4 { K. h- «

Authors Harke. 1970 (28) Russell and Feycrabcnd, 1975 (561 Cana rr al., TAHLE 3 - NICOTINk UPTAKt OI NONSMOKEKS EXPOSED 70 TOtlACCO SMOKE UNDtn DLFINEU CONDITIONS No. of nonsmokers Exposure conditions investigated Results ta) Experiment 1: Room /170 7 Nicotine excretion in urine until 6 hr after exposure m31; 7 nonsmokers; 11 10 a 6.8Lg/6hr smokers smoked a total of Cotinine excretion about 100 cig:utueh: duration 35 ± 34.5 µg/6 hr of exposure, 2 hr; no ventilation; indoor air concentration, 30 ppm CO. (b) Experiment 3: Same as 7 Nicotine excretion in urine until 6 hr after exposure Experiment I but with fair 18 t 7 µgl6 hr. ventilation (according to Cotinine excretion German stand:u.l for industry, . 19 s 9.4 µg/6 hr 33.4 m3/person hrr, indoor air concentration. 5 ppm CC). Room (43 nl)); 12 nonsmokers; 9 12 Nicotine in plasma (ng/ml) smokers; altogether 80 Before: 0.73 --t 1.6 cigarettes and 2 cigan were After: 0.90 -t 11.29 smuked; duration of exposure, Nicotine in urine (15 min aftCr exposure) 78 min; no ventilation; indoor 80 = 58.7 air concentration, 38 ppm CO. (a) Experiment 1: Room (66 mj); ~ Nicotine excretion in urine tµg/?4 hr) 1970 (13) 2 nonsmokers and 4 smokers lived together for 5 days; ventilation: 90 mtlhr. D.q I (without Smoking): 0 Day 2 (98 cig): 35-44 Day 3 (121 cig): 50-61 Day 4 1'I-l c ig l: 62 S- 711 Da~ 5 (NK cig): 47-SO Nici,unr r+,rt.•tium m 11iinC

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 627 :» ,P- rs. -Icrt hy o- ms letl to ur ule or- here is Aester- of the )f non- sponse -eases, s to in- iophys- may be effects ironary ions. In >genesis id. long- the dis- less zx- paasivc :. VICIIn:1. smoking were observed (27). But, in 1983, Hirayama (36) reported a signiticant risk elevation for ischemic heart disease in nonsmoking women married to smokers. The author offered a combined action of CO and nicotine as a possible explanation. In this report, we will try to summarize the present state of research into the influence of passive smoking on the development of CHD. The concentrations of CO and nicotine in indoor air and their uptake by passive smoking as described in the relevant literature will be reviewed. Possible cardiovascular effects of these substances in concentrations found in passive smokers will be discussed. We have omitted thrombangiitis obliterans, as it does not seem to be related to CHD. CARBON MONOXIDE AND NICOTINE CONCENTRATIONS IN INDOOR AIR Numerous studies have focused on CO and nicotine, both of which, depending on their concentrations, may affect the cardiovascular system (51, 52). Since other tobacco smoke constituents are not yet known to have such an effect, we contine ourselves to discussing these two substances. The fact that CO originates from a number of sources apart from tobacco smoke must be taken into consideration when evaluating CO concentrations in indoor air. The mean concentration of CO arising from tobacco smoke is assumed to amount to less than 5 ppm (24, 29, 66), although maximum values of approximately 40 ppm may occur for short periods under extreme conditions (32, 33). Nicotine is an alkaloid specific to tobacco; its indoor air levels can thus be regarded as, in fact, originating from tobacco smoke. The relevant literature cites levels of between 0.9 (71) and 3100 µg/m3 (33). Ac- tually, nicotine concentrations in rooms with smokers are estimated to range from 5 to 50 µgJm3 (8, 35, 71). CARBON MONOXIDE AND NICOTINE UPTAKE BY PASSIVE SMOKING Measuring CO, nicotine, and nicotine's main metabolite, cotinine, in blood and other bodily fluids is a better method of estimating their biological activity than drawing conclusions from indoor air concentrations. We are, however, faced with uncertainty as to the origin of CO in blood. In addition to various exogenous suurces, there is a permanent endogenous CO formation in the human organism. Therefore, in order to measure accurately the CO uptake by passive smoking, carbuxyhemoglobin (COHb) concentrations have to be quantitied under well- Jctined conditions before and after exposure. The results obtained on the basis of these requirements are shown in Table 1. Taking into account a CO content b,,luw S ppm in indoor air originating from tobacco smoke, the COHb amount is cstimated to increase over the day by less than 1%. Nonsmoking subjects, under rc:,l-life conditions representing the average German population, reveal COHb levels ranging from t to 1.5% (Table 2). In comparisun, cigarette smokers are tuund to have COHb levels between 3 and 10% during the day (59). Nicotine uptake from passive smoking has been assessed by measuring nicotine And cutinine concentrations in serum, saliva, and urine. Table 3 summarizes the Jat:e obtained under experimental conditions. Due to varying test procedures, they are not absolutely comparable. Nevertheless, they demonstrate that urinary c.cr6tion of nicotine and cotinine is several times higher in passive smokers TI BU 31623

11,1111.1 ~ ~r.rlr..rrr..n~r.~ r.1 Y ]IlLrnl.1 ~ lived togcther fot 5 days; ventilation: 90 m'/hr. 0 Hoflmann rt ul.. 1983 (31) (bl kxpaiment'_: Same a+ kxperimcnt I: subjects lived together for 4 llays : indoor air nicotine concentration was measured. Chamber (16 m3): 2 nonsmokers: 4 cigarettes Nere smokeJ continuously by machine; duration of exposure, 80 min; ventilation, 6 air changes/hr. indoor air concentration, 4.6 mg TPM/m3, 280 I,g nicutine/ 1113, 56 µg HCN/m'. 25 ppm CO. (The experiment was repeated three times with different subjects.) I 6 . Ii..~ 1 1. r........ ~.... .....br ~ Day 2 (98 cig): 35-44 Day 3 (121 cig): 5(i-61 1).,} 4 tva iigI t.' 1. -7U D.o S tt<r cigl. a7-tu Nicotine excretion in urine Day 1(97 cig. 15 µg mc/nH'1: 23-34 Ng%'_4 hr Day 2 196 cig. '_2 µg nic/m'): 22.5-58 µgI24 hr Day 3(9a cig. 35 pg nic/m'): 47.5-69 µg/24 hr Day 4 (103 cig. 33 µg nic/m'): 32-65 µg/24 hr Nicotine in plasma (ng'mll Before: - After: 0.5 Cotinine in plasma (ngAnU Uefore: 1.0 2.5 hr after expusure: 2.9 Nicotine in saliva (ng/m) Before: 3 After: 730 Cotinine in saliva (ngJml) Before: 1.3 2.5 hr after exposure: 3.1 Nicotine in urine (ng/ml) Qefore: 18 2.5 hr after exposure: 86 Cotinine in urine (ng/ml) Before: 17 2.5 hr afier exposure: 42

Suhcuciuthelwl vactlulcs Suhrnduthchal edetpa (orminE arcaJes • Sc.cic .uhcnJ.~thcl~.~l cJcm.. scp:1raung the mumw from thc undcrlymE tissue t'resence of smooth muscle cells in Ibe subcnJothrlial spuce Fiagmcntation of the internal elastic membrane Munkey Expusurr level: up to 400 ppm CO (,t = 11) Eapusme time: 10 x 30 min/day for 12 months Monkey Exposure: 43 cig/day (,t = 36) Exposure time: 14- 19 months CO level in blood: 0.24 m1/dl Animals were conditioned to cigarette smoking. t.amcl Ihot alta Ahduminal aona b5 53 Serum levels tµmul/mU Ex useJ Cuntrul Free cholesterul 0.60 Y 0.08 0.55 *_ (1.(18 Cholesterol ehlers 1.74 = 0.24 1.63 T 0,18 Aor1iL cholesterul content (Nmullg) Cholesterol Thoracic aorta 4.34 Y 0.61 3.95 s 0,58 Abdominal aorta 4.98 ± 1.'_0 3.06 s 0.29 Chulesrerinesler Thoracic aorta 0.77 T 0.12 0.71 ~ 0.(16 Abdominal aorta 0.69 z 0.06 0.76 ~ 0.09 Histology of the aorta No structural damages or fat depusirion Serum levels (ng/afl/ Smokers Coouul 7otal cholesterol 171.6 162.9 7otal tnglycerlde 41.3 41.2 Hht. cholesterol 83.(t H8.1 VLDL + U)L 78.3 66.4

I TABLE 7 CAKPIUVASCULAR EFFLCIS IN PASSI%E SM/i6LKS Authur Conditions n Results" Eaprrimcnwl roont: 9 m3 40 Heart rate Cigarettes sntuked: 6 Bcfure 89 Exposure time: 15 min Afttr 97 Blood pressure btlure 116/67 Alter 120/72 Esperimental roum: 170 nr; 10 Hzart rate Cigarettrs smuked: 150 Before 72 ~ 8 Exposure time: 20 min Atter 74 * 12 CO lavel: max 60 ppm Blood pressure Bcfure 123 ~ 15/23a *_ 10 Alter 121 ± 11/84 m 6 ilal skin tem th er re l-°CY t u / g p u a n n (icfurc 0 After U.0?73 T 0_0314 ?-6 !f He ait ralc 71

SYMPOSIUM: MEDICAL PERSPECTiVES ON PASSIVE SMOKING 625 Lsures to ppm and ~ applied ure, and jluntarily• pm _~CO .vith only cigarette zd upper nuked ih trehh air ,cunct:n- )utn with y smokes .t reasun, stiunablC possible, ng rooma y, should 1it values and the resulting requirements of fresh air supply are valid only for healthy adults. At present, only little is known about the acute effects of passive smoking on sensitive or at-risk groups such as asthmatics. allergic persons, chronic bronchitis sufferers, and children. Several studies (1-3, 6, 8, 10, ll) indicate a greater sensitivity of these groups to environmental tobacco smoke, but other studies do not (9). Further systematic research in this field is needed. REFERENCES I. Culley, J. Passive smoking in children. Nursing Times 71, 1858-185911975). 2. Dahms, T. E.. Bolin. J. F., and Slavin, R. G. Passive smoking: Effects on bronchial asthma. Chest 80, 530-534 (1981). 3. Gortmakar, S. L.. Walker. D. W., Jacobs. F. H., and Ruch-Ross, H. Parental smoking and the risk of childhood asthma. Arner. J. Public Heulth 72, 574-579 (1982). 4. Huber, G.. and Wanner, H. U. Indoor air quality and minimum ventilation rate. Environ. /nt. 9, 153- l5b (1983). 5. Johansson. C., and Ronge, H. Climatic influence on smell and irritation effects from tobacco smoke. Nurd. Hyg. T. 47, 33-39 (1966). 6. Leederer, S. R.. Corkhill. R.. Irwig, L. M., and Holland. W. W. Influence of family factors on asthma, wheezing incidence of lower repir.uury illness, and ventilation function in children. Brit. J. Prev. Suc•. Med. 30, 203-224 (1976). 7. Muramatsu, T.. Weber. A.. Murumatau. S., and Akermann. F. Experimental study on irritations and annoyance due to passive smoking. hu. Arc•h. O(uup. Environ. Health 51, 305-317 (1y83). 8. Shephard. R. J., Ponsfurd. E.. LaBarre. R., and Ba;u. P. K. Effects of cigarette smoke on the eyes and airways. Int. Arch. Occup. Environ. Hrult/r 43, 135-I44 11979). 9. Shephard, R. J., Collins. R.. and Silverman. F. Passive exposure of asthmatic subjects to cigarette smoke. Environ. Res. 20. 392--W2 (1.979). 10. Speer, F. Tobacco and the nonsmoker. Arch. Environ. Healtlr 16, 443-446 (1968). 11. Tager. 1. B.. Weiss. S. T.. Rosnzr. B.. and Speizer. F. G. Effect ot' parental cigarette smoking on the pulmonary function of children. Amer. J. Epirlenriul. 110, 15-26 (079). 12. Weber. A. Luftungsmassnahmen zunt Schutze der Passivraucher. Huusrechnik-Buuphysik-Um- we/ttechnik-Gesundheits-ingenieur 104, 37-42 (1983). 13. Weber, A.. and Fischer. T. Passive smoking at work. Int. Arch. Occup. Em•i.un. Health 47, _09- 11_t (1980). 14. Weber, A.. Fischer. T.. Gierer. R., and Grandjean. E. Experimentelle Reizwirkungen von Ak- rolein auf den Menschen. Int. Arch. Occup. E'rn•rrun. Heulth 40, l 17-130 (1977). 15. Weber. A.. Fischer. T., and GrJndjean. E. Objektive und subjektive physiulogische Wirkungen Jeo P.rsaivrauchertb. Int. Arch. Ovcup: Environ. Health 37, 277-288 (1976). 16. Weber. A., Fischer, T., and Grandjean. E. Reizwirkungen des Furmaldchyds auf den Menschen. /ru. Arch. Occnp. Envirun. Health 39, 207-218 11977). 17. Weber, A., Fischer. T., and Grandjean, E. Passive smoking in experimental and tield conditions. tnvirun. Res. 20. 205-216 (1979). I8. Weber. A.. Fischer. T.. and Grandjean, E. Passive smoking: Irritating effects of total smoke and gaa phase. Int. Arch. Ocrup. Environ. Health 43, 183-193 (1979). TI BU 31621

Authun Wesilalt, 1970 472) TABLE 6 CHRONIC EFFECTS C/1 NIC(/1INE IN ANIMALS Species Eaperimental conditions Results Rat Dusage: I mg'kg tK'ice daily Duration of treatment: 47. days M)ucardtal norepmephrine (NE) Nicotine Contrul - - Administratton: sc Stcady-state NE level lµg+gl 0.66 :t 0.04 O.03 0.6'V ± Rate cunstanl of NE loss (hr-'/ 0.103 a 0.01 0 073 ± 0.05 P < 01101 Turnuver rate (µg/glhr) (1070 0.047 Systahc btuod pressure (mm iig) 134 122 P< 0.5 -~ Schievdbein rr uf.. 1970 158 1 Rabbit /n = 391 Dosage: 1.14 ml;•'kg twice daily Duration of treatment: 20 munth> Concentratiuns in serum Nicotine m X Cunuul Administrauun: Solured with Tutal lipids tmg'1011 ml/ 540 480 drinking w•aler Free fatq acids (µval/1) 170 155 Mean nicoune bhw'd level: Concentrations in heart tissue 0.2 µg'ml /15 min after administrarrun) Fuher rr ul.. Rabbn DUSaEL- 0 5 nlg 1NlCe daily 1973 1_'5! (a = I6) Uurauun of ue:,nnent 90 days Admimslruuun si Total hpids (mg'10u g) '_,779 2,517 Esterified fatty acids /mg'1/10 g) 1.286 1,48(/ Aonic calcium content (mg'100 g) 19 8 (DdTerence is significant) Histulugy Nu dillcrences between nicuune- treated and .:untrul gruups. Syswh: bluud pressure of nicuune- treatcd animals (mm Hgl Initial 105 = B Final ION = 11

632 SCHIEVELBEIN AND RICHTER In cigarette smokers, the average COHb levels are in the range of 4 to 6%. Concentrations of more than 12% are rarely found (59). Nonsmoking bridge and tunnel workers, blast furnace workers, and traftic officers chronically exposed to CO were found to have COHb levels between 2 and 3%. The rate of CHD in these groups was not higher than that of nonsmoking workers who were not so exposed (7, 43, 63). [n contrast to healthy individuals, in patients with coronary artery disease, an acute 2% increaNc in COHb may reduce the exercise tolerance until the onset of anginal pain by aggravating the ischemia of the heart muscle (2, 4, 5). Based on the theory that CO is an important factor in the pathogenesis of CHD, numerous animal experiments have been conducted (Table 5). The results of these studies thus far have neither proved nor disproved an atherogenic effect of CO. By reevaluating their own findings (6), Hugod and co-workers came to the fol- lowing conclusion: "Even when rabbits are exposed to a lethal level of CO, the criteria hitherto believed to be indicators of CO-mediated endotheliai damage in normucholesterolemic rabbits appear not to be valid in the present study, leaving us with the problem that we-so far-do not know which criteria to apply, whcn trying to estimate a toxic action of CO on the intima of normocholesterolemic animals, and attempting to relate this to early atherosclerosis" (39). The data obtained in rabbits were confirmed in studies with primates whose metabolic status is more comparable to that of humans. Murucu fascicularis chronically treated with CO were found to show neither structural lesions nor lipid deposi- tions in the arteries (10). Ln a similar experiment with baboons, the cholesterol and triglyceride concentrations, which are thought to be related to the develop- ment of arteriosclerosis, were not increased (53). There have been no experimental investigations into the acute effects of nic- otine on the cardiovascular system in concentrations such as those found in pas- sive smokers. [f one extrapolates from the dose-effect relationship in smokers, no measurable acute effects of nicotine inhaled by passive smokers should be expected. Several animal experiments have been pertormed to study the chronic effects of nicotine (Table 6). Their results, again, neither prove nor disprove an atherogenic effect of nicotine. ln a few experimental studies, the influence of cigarette smoke on the cardiu- vascular system was measured directly (Table 7). Although exposure in these studies was much higher than'under real-life conditions, an increase in heart rate and blood pressure was either undetectable or only marginal. As shown in astuJy by Harke and Bleichert (30), the skin temperature, which decreases considerably while smoking one cigarette, remained constant. DISCUSSION For more than 25 years cigarette smoking has been associated with an elevated risk of death from CHD. This has been confirmed by numerous epidemiological studies. Yet, the mechanism by which smoking might increase the risk of deve- loping cardiovascular diseases is not understood. A large number of chemicals have been identitied in tobacco smoke (18), some of which may be implicated in the development of cardiovascular diseases. Interest primarily has concentrated ~ 4 { K. h- «

620 Question Answer Air aualltv r+ttn re9ard to snlbke Dtsturbed by smoke never ANNETrA WEBER 9ood bad lntervtewed aersons 0 20 40 so so x somettmes/ often 21 ''///////////~///. %////// ,. , ~ ~/" ® s. .. .a nciw a Eye trrlta- ttons at work sorrst 1 mes/ often 0 I Fle. 1. Evaluation ot air quality and effects Jue to environmental tobacco smoke at the workplace. Reaults uf 472 employees in 4-1 workrooms. "-Oifference significant with P < 0.01 (Y'- tcat). u. Nonsmokers + exsmokers (N = 237); M, smokers (N = 235). Luburutvey Studies , !n a first experiment (15), 33 subjects were exposed to continuously increasing smoke concentrations. The main results are summarized in Fig. 2. The con- centrations of CO, NO, formaldehyde (HCHO), and acrolein increase linearly with the number of cigarettes smoked. Buth tncan subjective eye irritation and mean eye blink rate increase in a nearly linear way with increasing smoke cun- centratiun. j Subjective nose and throat irritation were also determined. The effects on the nose were less pronounced than those on the eyes, and those on the throat were the ~,mallest of all three. Thus, eyes are the organs most sensitive to irritation due to passive smoking. In a second series of laboratory experiments (7, 17), acute effects were analyzt:d in relation to smoke concentration and duration of exposure. The chosen realistic smoke concentrations corresponded to 1.3, 2.5, 5.0, and 10 ppm CO due to tu- bacco smoke (JCO) . During a l hr period, 32-43 subjects were exposed to theao smoke concentrations. Each smoke concentration increased linearly during the tirst 5 to 10 min, and was then held constant at the desired level for the remaining part of the hour. Since very high correlations (r > 0.9) between CO and each of the other cum- pounds were obtained in the tirst experimental series, only JCO was used as an TI BU 31616 Fic. 2_ during cu rute. 0.01 trruattuns indicstu hmuke. and eye (a) M concent, (b) 8, Ut il Cur alsu bee The p, wun aa rpproxir un tht: tlL A thirL tnine wht 4n1110yanc. .tlunc:, fu, .a. i whu l,tl Ac due to tF (b) Th ttlc iidea (c) Th cliet;t ut

13igrual sl in lcmpciattnc I-`0tnnnl fielore After 0 0.0273 = 0.0314 Hurshman rt ul.. (lgarcnes smoF,eJ: _'-A 10 min t ftcan ralc BClore 73 197$ (anl Exposure lime: 2/1 ppm After 79 CO level. Hlood pressure Hcfore 107/67 ' After 114/68 V1 -C Pirnm et u[.. Experimental room: 15 tn 20 Hcart rate S 1y7>; (50) Cigarettes smoF.ed: 12 Before 6y(d)74 t~)" ~ Exposure time: 2 hr Afttr 7U(d)7K(;) ~ CO level: ' 24 ppm / ' C ~ Fischer 1979 TI M: erimental room: Ex 4 mg m 30 m' 33 Heart rate . (23) p Cigarettes smoked: 20 Before 76.3 m (7 Exposure time: 28 min After 75.2 n Rummel rt ul., CO level: Experimental room: 43 ppm 30 m' 56 He:,rt rate y r v 1975 (54) Cigarettes smoked: 6-8 Before 7'' = 1(1 ~' ~ Exposure time: 20 min After 71 *_ 11 ~ CO level . Blood pressure fiefore 117 s 21 /71 ~ m n 12 After 117 ~ 15/71 ~ 12 < rr, v ° Citations are to figures of inean values. None of ttie Jifferences are significant. 0 " In the female subgroup the increase in heart rate was partially significant during the exposure period. z ~ ' v v < ~. m v C z C) o1 w ~

Authun Wesilalt, 1970 472) TABLE 6 CHRONIC EFFECTS C/1 NIC(/1INE IN ANIMALS Species Eaperimental conditions Results Rat Dusage: I mg'kg tK'ice daily Duration of treatment: 47. days M)ucardtal norepmephrine (NE) Nicotine Contrul - - Administratton: sc Stcady-state NE level lµg+gl 0.66 :t 0.04 O.03 0.6'V ± Rate cunstanl of NE loss (hr-'/ 0.103 a 0.01 0 073 ± 0.05 P < 01101 Turnuver rate (µg/glhr) (1070 0.047 Systahc btuod pressure (mm iig) 134 122 P< 0.5 -~ Schievdbein rr uf.. 1970 158 1 Rabbit /n = 391 Dosage: 1.14 ml;•'kg twice daily Duration of treatment: 20 munth> Concentratiuns in serum Nicotine m X Cunuul Administrauun: Solured with Tutal lipids tmg'1011 ml/ 540 480 drinking w•aler Free fatq acids (µval/1) 170 155 Mean nicoune bhw'd level: Concentrations in heart tissue 0.2 µg'ml /15 min after administrarrun) Fuher rr ul.. Rabbn DUSaEL- 0 5 nlg 1NlCe daily 1973 1_'5! (a = I6) Uurauun of ue:,nnent 90 days Admimslruuun si Total hpids (mg'10u g) '_,779 2,517 Esterified fatty acids /mg'1/10 g) 1.286 1,48(/ Aonic calcium content (mg'100 g) 19 8 (DdTerence is significant) Histulugy Nu dillcrences between nicuune- treated and .:untrul gruups. Syswh: bluud pressure of nicuune- treatcd animals (mm Hgl Initial 105 = B Final ION = 11

, 3 6 lo UY ) 5 1 3 ndi- .~r~. ~~zd tt uf ) unt! SYMPOSIUM: IMEDICAL PERSPECTIVES ON PASSIVE SMOKING 629 TABLE 2 COHb IN NONSMOKEiiS UNDEN REAL-LIFE CONDITIONS No. of nonsmokers Authors Subjects (exposure) investlgated Mean 1: SID) Range Waid rr ul.. Participants in a hualth 6.641 97% 1981 169) screenmg program 0.0-1.5 Carsky rr u1.. Blood donors in Bratislava 130 U.9 U.3-3.7 1'lttl/ 1141 lanzun r( ul.. il 1981 Participants in a health 509 Median: 0.5 0-2 ) ( Stewurt et u/.. screentng program Blood donors in United States 16.Ik{2 1.42 (= 0.23) 90% 1975 1641 0.4-3.7 Stewurt rr ul.. (a) Blood donors in Chicago 406 2.04 90% 1976 (65) 1970 1.0-3.2 (b) Blood donors in Chicago 426 1.53 vll% 11)74-1975 0.7-2.4 Einbnxlt et ul.. Total adult inhubitants uf a 103 0,71 1= 0.36) I97n (.0) small village Cule. 1975 lal Subjects working in a•'CO- 78 0.68 0.2-2.6 1151 I'ree" atmosphere I~- I ppm CU/ (bt Subjects working in a huapital where smuking was not allowed 1:-•t ppm CO) 120 0.97 0.2-2.5 tct Subjects working in an uflice where ,muking was 1(!0 1.12 0.1-2.7 alluwcu 13-iS ppm CO) Stadkuw,kt rr al.. lal Subjects working in an 130 Morning: 0.82 I= 0,37) 1976 l6bl utlice with emukers present Ibl Subjects working in an 118 il Afternoon: 0.63 (= 0.3U1 Morning: 0.92 t= 0.40) ullice wuhuut smuken 37 Afternuun: 0.72 I s 0.25) in our laboratory (Table 3). On the basis of their own tindings, Jarvis et a!. (42) calculated urinary nicotine concentrations of about 7.5% for passive smokers and urinary cotinine concentrations ranging from 0.9 to 2.4% of the levels found in hmukers (Table 4). EFFECTS OF CARBON MONOXIDE, NICOTINE, AND CIGARETTE SMOKE ON THE CARDIOVASCULAR SYSTEM OF PASSIVE SMOKERS As already mentioned, CO and nicotine have been proposed as the components of cigarette smoke that may be responsible for the development and progression of CHD in smokers (5l). Carbon monoxide exposure causes the same symptoms or disturbances, during rest or exercise, as those encountered in low oxygen or hypobaric hypoxia (67). In healthy individuals at rest, alterations in physiological lunctions such as change in pulse rate, cardiac output, or blood pressure are not u,ually found at levels uf COHb below 20% (60). During exercise, while breathing (;(). there is a decrease in the maximum uxyguu uptake and an increase in heart ratt: and ventilation. Adaptive metabolic changes may occur in response to chron- tt:ally increased COHb concentrations resembling the processes in subjects who Ilve at high altitudes (57). It is not certain whether the advantages of such an ,uljustment outweigh possible disadvantages. TI BU 31625

6°To. and al to D itt A su nary ance .1Scfe 'f•f D, :hese Co. ~ foi- t. tl`tt: ge in .Avin6 w hen lt:m1C data LbollC ically :posi- sterol relop- >f nic- n pas- Aers,' ild be ironic ,vean ardiu- these rt rate study rably .vatt:d .ugical tte v e- tnicaiN tttal. in tt rateti SY,MPO51UM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING TABLE -t NICOTINE U['TAKJ: OF NONSMItKtittJ UNUL:H R<:AL•Llt•E CONUI'rIONs 633 Authors Sub)ects (exposure) No. ul nonsmokers investigated Results tng/mll kusNell and Feyerabend. 1975 (56) Employees in a hospital (urtne was delivered I hr after lunch) (al Group I la Nicutine in urtne: 12.4 s 16.9 [b) Group 2 13 Nicotine in urine: 8.9 x 9.1 Feyerabend rt Employees 31) Nicotine in urine; 7.5 x 8.3 ul., 1962 (a) Nunexposed to passive Nicotine in saliva: 5.9 s 4.4 121) smoking Jurmg the morning tself-repurted) (bl Exposed to passive :h Nicobne in urine: _1 •, : 289 nmuking during the morning lr / Nicotine in ealiva: 10.1 : 9.7 Fuhart rr ul.. (.e -repurtet ) Flight auendants (women) l8-hr 6 Nicotine in serum 1963 (I61 tlightl Before: 1.6 x 0.8 Atter: 3.2 : I.0 Nicutine excretion in urine: 12.9 s 6.5 µ(yM hr WrW rr d.. Outdoor patients and clinic atatT 19ri4 (615) (u) Nunexpustd to paaaive smoktng tselr-repurted) (bl Exposed to passive smok- ing (selt•repurted) W:dd und Ritchie. (a) Husbands (wife is a nun- 19tlb 1701 emuker) tbl HusbanJs twire is a smoker) Jurvin r( ul.. Empluyees in an uttice 1983 142) IBIuuJ, suliva, and urine sum- pies were deliveret Jt I t:3U .1M and 7:45 aM: period uf ubser- vauon included a.•hr stay in a ~muky pub.) ,7 Cutinine in urine: Median 2.0 ( lOth percentde: U.U: 911th percentile: 6.3) 199 Cutinine in urine: Median 6.0 ( IUth percentile: 1.4: 90th percentile: 2_.W IUI Cutinine in unne: S.5 z 1.3 20 Cunnine in urtne: 25.'_ = 14.8 7 (3efure rltter Nicotine in plasma = 7-44 Nicotine in ~altva I.vt1 43.63 Nicotine in urine 11.1.5 1 92.63 Cuunine ut plasma 1.0 7.33 Cuunme in saliva 1.50 S.U4 Counine in urine 4.80 12.94 on CO and nicotine. There is little evidence to suggest that other substances fuund in cigarette smoke may adversely affect the cardiovascular systtem (3g). fhuubh substances with such properties may exist, none have yet been identified. Carbon monoxide causes increased COHb levels, thereby reducing the blood's uxygen-carrying capacity and tissue oxygenation. Nicotine releases adrenalin t'rum the adrenal gland and other chromaffin tissues, and noradrenalin from the hy- puthallunus and nerve endings. The cardiovascular responses to nicotine, in gen- t:ral, parallel those that follow stimulation of the sympathico-adrenal system. Increase in pulse rate and blood pressure correlates with nicotine levels in blood (45). Carbon monoxide and nicotine act together insofar as CO diminishes the lrampurt capacity fur oxygen to the heart muscle, while nicotine increases the need of the activated heart muscle for oxygen. Since the heart is known to con- sume most of the oxygen available to it, any disorder that enhances the need for uxygen or decreases oxygen in the circulating blood could directly impair cardiac mtucle metabolism, thus leading to irreversible ischemic necrosis. Nicotine could %;:r,u,e a fatal arrhythmia by increasing myocardial irritability (9). It could also TI BU 31629

I TABLE 7 CAKPIUVASCULAR EFFLCIS IN PASSI%E SM/i6LKS Authur Conditions n Results" Eaprrimcnwl roont: 9 m3 40 Heart rate Cigarettes sntuked: 6 Bcfure 89 Exposure time: 15 min Afttr 97 Blood pressure btlure 116/67 Alter 120/72 Esperimental roum: 170 nr; 10 Hzart rate Cigarettrs smuked: 150 Before 72 ~ 8 Exposure time: 20 min Atter 74 * 12 CO lavel: max 60 ppm Blood pressure Bcfure 123 ~ 15/23a *_ 10 Alter 121 ± 11/84 m 6 ilal skin tem th er re l-°CY t u / g p u a n n (icfurc 0 After U.0?73 T 0_0314 ?-6 !f He ait ralc 71

IY/S 1-'11 D+Ig tn = 241 1iu1.+4.n~ •,1 ucnuuail. yu d.+) 4 .4dm10istr.+nun ec ........... ~ ... I+C~ In/t1al 105- 1B I-ma1 Itkt - 11 Scrum:un.cMr.+u~+m nuE I/MI olll N+:..unc Cunnol 7nglyccr+Jr, 9T-~V(5Y 10;-'i41 Cholc.rrr+d 154 = hO 197 2 66 No alhcrusclcrwr, or vaacular dalnagic uhscrvr:d by corunary angl.+graphy Dusage o/ hislolugy I Gruup ] Cunlrol Gr /UU Group I. 7 clgarcue+'duy Hcarl ratt (hcars min +I .- " 148 = 13 144 = B Group 2. 2141 pg niruUnetkg End diasrohc prtseurc (mm Hgl K.4 - 1/1 8.3 = 1.7 N,4 = 1.3 INI:C dally Duratiun of Ircatmcnl. ]2 nwnths E)ccuun fra.rrun 1`il 35= 3.U 27= 3.4 44='.6 Arlnriniutralron /Gruup 2 vs cuntrol P~ 0011 Group 1: Trachcolumy Group 2. inl T+la1 sys.cmlc reslslance Idyn sc:-'1 7691 _ : 11145 4'166 = SII6 3tS?ll = 534 IGroup 2 ve cunlrul P< 0.01) (Gruup 2 vs group I P- QOSI Serum ciwltslcral /mmol/l3 4.11 - 0 2 3.K5 z 0.94 3.95 = 0.65 Elcctrun microscopy Nu abnurmahuc~ of myohbrils or olher cell organtlles in group I or 2. Rabbu Dosagc. 2 4 mg/kglJay tf/ ° 26) Duratlorl- 25 "eef.s Administrrlion: Solured in dnnking N'ater Serum lipid concenlrations (mg/dI! Nicounc Tnglycerrdes 41.6 - 8.3 HDL chuleaterul 22 9= 3.1 LDL chuleslcrol 23.6 s 3.4 Evans hlue dyc uplal.e of aorta No drfferencC. Enduthchal cyroplasmic ailvcr dcpusulun Incrca,ed rn nrcoune-treated ammals. Control 33.1-5.2 lP< 0,001) 234x3.7 10.4 = 2.4 (P - (LUlll/ ~ w J

13igrual sl in lcmpciattnc I-`0tnnnl fielore After 0 0.0273 = 0.0314 Hurshman rt ul.. (lgarcnes smoF,eJ: _'-A 10 min t ftcan ralc BClore 73 197$ (anl Exposure lime: 2/1 ppm After 79 CO level. Hlood pressure Hcfore 107/67 ' After 114/68 V1 -C Pirnm et u[.. Experimental room: 15 tn 20 Hcart rate S 1y7>; (50) Cigarettes smoF.ed: 12 Before 6y(d)74 t~)" ~ Exposure time: 2 hr Afttr 7U(d)7K(;) ~ CO level: ' 24 ppm / ' C ~ Fischer 1979 TI M: erimental room: Ex 4 mg m 30 m' 33 Heart rate . (23) p Cigarettes smoked: 20 Before 76.3 m (7 Exposure time: 28 min After 75.2 n Rummel rt ul., CO level: Experimental room: 43 ppm 30 m' 56 He:,rt rate y r v 1975 (54) Cigarettes smoked: 6-8 Before 7'' = 1(1 ~' ~ Exposure time: 20 min After 71 *_ 11 ~ CO level . Blood pressure fiefore 117 s 21 /71 ~ m n 12 After 117 ~ 15/71 ~ 12 < rr, v ° Citations are to figures of inean values. None of ttie Jifferences are significant. 0 " In the female subgroup the increase in heart rate was partially significant during the exposure period. z ~ ' v v < ~. m v C z C) o1 w ~

TABLE I COHb IN NONSMOKERS EXYOSEu r0 TUBACCU SMOKE UNUEK DEFINED CONUI'rIQN5 Authors Exposure conditions Harke. 1970I:8) lat Experiment I: Room 117U m'); 7 nonsnwkers: I I emukers. whu smoked a total of about 100 ctg•arettes; duration of exposure, 2 hr; no venttlatiun: indoor uir cuncentrattun. 30 ppm CO. lbl Experiment 2: Satne as Experiment I but with muderate ventilation (16.7 m'lpersun/hr); indoor air concentration, 5 ppm CO. Ict Experiment 3: Same as Experiment I but with fair ventilation (according to German standard fur industry, 33.4 mJ1 persunihr); indoor air concentration. <5 ppm CO. Russell ee ul.. Room 143 ml): IZ nonsmukerst 9 1973 (55) smokers; altogether 8U cigarettes and 2 cigars were 3muked; duration ut' exposure. 7g min: no ventilatiun: indoor air cuncentr.uiun. 38 ppm CO. Seppanen. 1977 Room (37.5 m1): 4 nunsmuken; 3 (6l) emukers, who ~muked a tutal of IS cigarettes; duration of expusuru, 1.5 hr; no venttlutiun: indoor ar Zuncentratiun (at the end of exposure period). 30 ppm CO. (The expenment wao repeated seven times with different iub3ects.) Nu. ut nunsmukera invesugatad Reoufts I% COHbI 7 Before 0.9 _ 0.1 After: 2.1 z 0.1 Cdlr 1` )4n. 7 Before: 0.9 _ 0.1 Stc, 1` 7 .i)'ter: Before: 1.3 s 0.1 1.1 = 0.1 Jte' t, 2 After: Before: 1.6 = 0.t 1.6 = 0.6 Eln i Cul After: 2.6 :0J 28 Befure: 1.6 x 03 St. Atter: :.; = 0,6 Dahms et ul.. Chamber (30 m'): cigarettes were 20 (a) Il/ Nonsmokers 198 1 /16) omuked by machtne: venuiatwn, 5 BCCUre: 0.62 x U.UIS cur changes/hr; duration uf Alter: I.U6 _ U.Utl expusure. I hr: indoor air (b) 10 Asthmaucs cuncentrotiun. 15-30 ppm CO Befure: U.tl' = 0.10 Pimm er al.. Wrlculated). Chamber (14.6 tnlI; tl cigarettes were 20 After: Before: I.]u z U.tP1 0.5 1'+7tS l5U) eppanen and ~muked by machine; no ventilation; duration of exposure. 2 hr; indoor air concentration. 23.1 = 2.9 ppm CO. -4 mg TPM"/my. (a) Restaurants: durauun uf ,~tay. 5 7 After: Befure: 0.tf 2.1 = u.5 Uuaualu, 1977 hr; indoor air cuncentration. 2.5- Al'ter: 2.1 x 0.5 (b2) 15 ppm CO. (b/ Office; duration of stay, g hr: 15 Before: 2.3 _ 0.4 indoor air cuncentration. 2.5 ppm After. 2.3 x 0.3 CO. ' TPM - Total particutate matter. compared with nonexposed subjects. In passive smokers under real-life condi- tions, the nicotine uptake may vary From 0.02 to O.Uk that ot' cigarette-smokzrs... Cano rr ul. ({3) reported a nicotine excretion in two nonsmokers heavily exposed to tobacco smoke over several days amounting to approximatt:ly 0.04 that of cigarette smokers, the latter ranging from about 1,000 to 2,000 µg/24 h as found TI BU 31624 in t Cal, uri l hm EF A of ~ ut C uf L hyp tUn• U]L Cc rat it:;. Itv, atl.

TABLE I COHb IN NONSMOKERS EXYOSEu r0 TUBACCU SMOKE UNUEK DEFINED CONUI'rIQN5 Authors Exposure conditions Harke. 1970I:8) lat Experiment I: Room 117U m'); 7 nonsnwkers: I I emukers. whu smoked a total of about 100 ctg•arettes; duration of exposure, 2 hr; no venttlatiun: indoor uir cuncentrattun. 30 ppm CO. lbl Experiment 2: Satne as Experiment I but with muderate ventilation (16.7 m'lpersun/hr); indoor air concentration, 5 ppm CO. Ict Experiment 3: Same as Experiment I but with fair ventilation (according to German standard fur industry, 33.4 mJ1 persunihr); indoor air concentration. <5 ppm CO. Russell ee ul.. Room 143 ml): IZ nonsmukerst 9 1973 (55) smokers; altogether 8U cigarettes and 2 cigars were 3muked; duration ut' exposure. 7g min: no ventilatiun: indoor air cuncentr.uiun. 38 ppm CO. Seppanen. 1977 Room (37.5 m1): 4 nunsmuken; 3 (6l) emukers, who ~muked a tutal of IS cigarettes; duration of expusuru, 1.5 hr; no venttlutiun: indoor ar Zuncentratiun (at the end of exposure period). 30 ppm CO. (The expenment wao repeated seven times with different iub3ects.) Nu. ut nunsmukera invesugatad Reoufts I% COHbI 7 Before 0.9 _ 0.1 After: 2.1 z 0.1 Cdlr 1` )4n. 7 Before: 0.9 _ 0.1 Stc, 1` 7 .i)'ter: Before: 1.3 s 0.1 1.1 = 0.1 Jte' t, 2 After: Before: 1.6 = 0.t 1.6 = 0.6 Eln i Cul After: 2.6 :0J 28 Befure: 1.6 x 03 St. Atter: :.; = 0,6 Dahms et ul.. Chamber (30 m'): cigarettes were 20 (a) Il/ Nonsmokers 198 1 /16) omuked by machtne: venuiatwn, 5 BCCUre: 0.62 x U.UIS cur changes/hr; duration uf Alter: I.U6 _ U.Utl expusure. I hr: indoor air (b) 10 Asthmaucs cuncentrotiun. 15-30 ppm CO Befure: U.tl' = 0.10 Pimm er al.. Wrlculated). Chamber (14.6 tnlI; tl cigarettes were 20 After: Before: I.]u z U.tP1 0.5 1'+7tS l5U) eppanen and ~muked by machine; no ventilation; duration of exposure. 2 hr; indoor air concentration. 23.1 = 2.9 ppm CO. -4 mg TPM"/my. (a) Restaurants: durauun uf ,~tay. 5 7 After: Befure: 0.tf 2.1 = u.5 Uuaualu, 1977 hr; indoor air cuncentration. 2.5- Al'ter: 2.1 x 0.5 (b2) 15 ppm CO. (b/ Office; duration of stay, g hr: 15 Before: 2.3 _ 0.4 indoor air cuncentration. 2.5 ppm After. 2.3 x 0.3 CO. ' TPM - Total particutate matter. compared with nonexposed subjects. In passive smokers under real-life condi- tions, the nicotine uptake may vary From 0.02 to O.Uk that ot' cigarette-smokzrs... Cano rr ul. ({3) reported a nicotine excretion in two nonsmokers heavily exposed to tobacco smoke over several days amounting to approximatt:ly 0.04 that of cigarette smokers, the latter ranging from about 1,000 to 2,000 µg/24 h as found TI BU 31624 in t Cal, uri l hm EF A of ~ ut C uf L hyp tUn• U]L Cc rat it:;. Itv, atl.

644 SCHIEVELBEIN AND RICHTER 65, Stewart. R. U.. Huke. C. L., Wu. A., Stewart. T. A.. and Kulbtleisch, J. H. Carbuxyhemugtubin trend in Chicago blood donors. 1970-1974. Arch. Envirnn. He•utrh 31. _2iQ-=86 11'J76). 66. iL;kdkuwski. D.. Harke. H.•P., and Angerer. J. KuhlenmunuxiJbetaslung durch Pasaivrauchen in liururaumen. Imt. Med. 3. 310-313 11y76). 67. Vogel. J. A.. Gleser. .Lt. A., Wheeler. lt. C., unJ Whitten. B. K. Carbon monoxide and physical work capacity, Arclt. Errvirun. Nruldt 24, 198-31)3 11971). 68. Wald, N. J., t3urchum. J., Bailey, A., Ritchie. C.. Hudduw, J. E., and Knight, G. Urinary cutimne as marker ut breathing other peupte's tobacco smuke. Lacu er 1, 230-231 (1984). 69. WatJ. N. J.. Idle, M., Borehum. J.. anJ k3uiley, A. Carbon monoxide in breath in relation to smoking nnJ carbuxyhuemuglobin levels. Thorux 36, 366-369 (1981). 70. Wald. N. J., and Ritchie. C. Validation uF studies on lung cancer in non-smokers married tu ~mukers. Lurcer 1, U)67 (1984). 71. Wehar, A., and Fischer, T. Passive smoking at work. Iru. Arch. Uc•c•up. Emvirun. Health 47, '_09- 2=l (1980). 72. Westtall. T. C. Influence of chronic nicotine administration on blood pressure and heart nurept- nephrtne turnover. Eur. J. Phurmuc•u1. 10, 19-24 (19'7U). I'KtV t`4I l IR Dit•ttturr A Chtlc. ette~ tha.t un ~ tU =l the , ettt. RCrc EruF AItht hntukin t:artliw, butiutt- huftici` Sc:vc ~itlc:~tr~ 1b, ~17 ) has bcL 'rhd in tliuxitle cartrorl (36. 4= 1ULIiltc: tl'tac:uac (sc:n (I. ?3) ti = f'rc- V ienn:.. a . TI BU 31640

644 SCHIEVELBEIN AND RICHTER 65, Stewart. R. U.. Huke. C. L., Wu. A., Stewart. T. A.. and Kulbtleisch, J. H. Carbuxyhemugtubin trend in Chicago blood donors. 1970-1974. Arch. Envirnn. He•utrh 31. _2iQ-=86 11'J76). 66. iL;kdkuwski. D.. Harke. H.•P., and Angerer. J. KuhlenmunuxiJbetaslung durch Pasaivrauchen in liururaumen. Imt. Med. 3. 310-313 11y76). 67. Vogel. J. A.. Gleser. .Lt. A., Wheeler. lt. C., unJ Whitten. B. K. Carbon monoxide and physical work capacity, Arclt. Errvirun. Nruldt 24, 198-31)3 11971). 68. Wald, N. J., t3urchum. J., Bailey, A., Ritchie. C.. Hudduw, J. E., and Knight, G. Urinary cutimne as marker ut breathing other peupte's tobacco smuke. Lacu er 1, 230-231 (1984). 69. WatJ. N. J.. Idle, M., Borehum. J.. anJ k3uiley, A. Carbon monoxide in breath in relation to smoking nnJ carbuxyhuemuglobin levels. Thorux 36, 366-369 (1981). 70. Wald. N. J., and Ritchie. C. Validation uF studies on lung cancer in non-smokers married tu ~mukers. Lurcer 1, U)67 (1984). 71. Wehar, A., and Fischer, T. Passive smoking at work. Iru. Arch. Uc•c•up. Emvirun. Health 47, '_09- 2=l (1980). 72. Westtall. T. C. Influence of chronic nicotine administration on blood pressure and heart nurept- nephrtne turnover. Eur. J. Phurmuc•u1. 10, 19-24 (19'7U). I'KtV t`4I l IR Dit•ttturr A Chtlc. ette~ tha.t un ~ tU =l the , ettt. RCrc EruF AItht hntukin t:artliw, butiutt- huftici` Sc:vc ~itlc:~tr~ 1b, ~17 ) has bcL 'rhd in tliuxitle cartrorl (36. 4= 1ULIiltc: tl'tac:uac (sc:n (I. ?3) ti = f'rc- V ienn:.. a . TI BU 31640

IY/S 1-'11 D+Ig tn = 241 1iu1.+4.n~ •,1 ucnuuail. yu d.+) 4 .4dm10istr.+nun ec ........... ~ ... I+C~ In/t1al 105- 1B I-ma1 Itkt - 11 Scrum:un.cMr.+u~+m nuE I/MI olll N+:..unc Cunnol 7nglyccr+Jr, 9T-~V(5Y 10;-'i41 Cholc.rrr+d 154 = hO 197 2 66 No alhcrusclcrwr, or vaacular dalnagic uhscrvr:d by corunary angl.+graphy Dusage o/ hislolugy I Gruup ] Cunlrol Gr /UU Group I. 7 clgarcue+'duy Hcarl ratt (hcars min +I .- " 148 = 13 144 = B Group 2. 2141 pg niruUnetkg End diasrohc prtseurc (mm Hgl K.4 - 1/1 8.3 = 1.7 N,4 = 1.3 INI:C dally Duratiun of Ircatmcnl. ]2 nwnths E)ccuun fra.rrun 1`il 35= 3.U 27= 3.4 44='.6 Arlnriniutralron /Gruup 2 vs cuntrol P~ 0011 Group 1: Trachcolumy Group 2. inl T+la1 sys.cmlc reslslance Idyn sc:-'1 7691 _ : 11145 4'166 = SII6 3tS?ll = 534 IGroup 2 ve cunlrul P< 0.01) (Gruup 2 vs group I P- QOSI Serum ciwltslcral /mmol/l3 4.11 - 0 2 3.K5 z 0.94 3.95 = 0.65 Elcctrun microscopy Nu abnurmahuc~ of myohbrils or olher cell organtlles in group I or 2. Rabbu Dosagc. 2 4 mg/kglJay tf/ ° 26) Duratlorl- 25 "eef.s Administrrlion: Solured in dnnking N'ater Serum lipid concenlrations (mg/dI! Nicounc Tnglycerrdes 41.6 - 8.3 HDL chuleaterul 22 9= 3.1 LDL chuleslcrol 23.6 s 3.4 Evans hlue dyc uplal.e of aorta No drfferencC. Enduthchal cyroplasmic ailvcr dcpusulun Incrca,ed rn nrcoune-treated ammals. Control 33.1-5.2 lP< 0,001) 234x3.7 10.4 = 2.4 (P - (LUlll/ ~ w J

648 MICHAEL LEBOWLT2 chosen in a stratified sample of families in uniform geographic clusters from a representative community population [already under study (:0)], as previou,ly described (23, 26). Studies of the total population sample indicated the presence of relationships between weekly symptoms and both air pollutants and allergens, controlling for weather (44). Stratification ensured that the families were repre- sentative of families with and without asthma in their geographic clusters. They were studied over a 2-year period, using daily symptom diaries and peak flow meters; daily response rates and peak flows were acceptable for a majority of days in all seasons (23, 28). All families provided environmental information as to characteristics of the house, yard, and streets, heating, cooling, type of stove, water heater, washer and dryer, and smoking in the home (26). Daily ambient monitoring data for air pollutants, aero-allergens, and meteoro- logical variables with regional and geographic clusters were used (23). Indoorand yard (micro-outdoor) monitorings were conducted in a random cluster sample representative of all study households (23, 26) for particulate matter [RSP and total suspended particulate matter (TSP)], pollen, bacilli, fungi, and algae, for 72 hr each, and grab samples were made for CO and 03. Microenvironmental sam- pling of pollen, algae, and fungi have been described previously (23, 37, 38). Temperature and humidity varied by season, but the former did not vary very much indoors over the year and the latter did not differ much between indoors and outdoors over the year. TSP indoors ranged from 5.7 to 68.5 µg/m3, and simultaneous micro-outdoor TSP ranged from 2.1 to 169.6 µg/m3. Cyclone measurements of RSP had a max- imum of 49.8 µglm3 indoors and 124.5 µg/m' outdoors. Electron microscopy yielded little identifying information on the type of dust found on the tluors of homes. Intiltration of outdoor TSP indoors was similar to pollen intiltration: total micro-indoor pollen counts were approximately 5% of the micro-outdoor pollen counts, which were about 5% of macro-pollen rates. Both TSP and RSP were significantly associated with environmental tobacco smoke (Table 1), but not with gas stove usage (26). Averages of the two spot CO readings using infrared klecurminations were lower than 2.4 ppm; micro-outdoor readings were less than 3.8 ppm. Indoor CO was significantly associated with gas stove usage (26), but not with environmental tobacco smoke (Table 1), nor with gas-burning furnaces, water heaters, or washer-dryers (that is, vented appliances). Although low, it was used as the indicator pollutant measure of combustion emission from the gas stoves. Neither environmental tobacco smoke nor gas stove usage was associated with social status in this sample. RESULTS We found many interactions between indoor and indoor-outdoor pullutants, and weather, with respect to health effects. Children's standardized peak tlow (PEF) was related to outdoor environmental factors only (03, TSP, temperature, and relative humidity); outdoor factors at'fected nonsmokers more than passive amokers (Table 2). Adult PEF was related to active smoking (as were chronic symptoms), gas TI BU 31644 Tor.{L. Sc • Dii "" t3ui StoVe U and no relativc PEF micro-i attack~ RS 1', b Indu and nu ductiv, smukir creaazl lzrSich Micro- Crtlu

SYMPOSIUhI: MEDICAL PERSPEC'CIVES ON PASSIVE SMOKING 647 pairment function .o effects. mptoms ; passive n France :s). They lealthiest .iy might tounding akia (52) ,mokers, - relation r adults. ,:ity) and oking on -)g3 N1H : on pul- ?toms in orkplace heir vul- -d popu- nu clear- ~ kers and conclu- nuke ex- tive with cre used mmental zgs. One reas the ive asth- ig habits those uf Ale (17). )n; huw- I lut;uus. In other studies in which measurements were not made, contradictory results were found. Hasselblad et al. (15) demonstrated that maternal smoking habits explain 0. I~r of the variance in children's forced expiratory volume (FEV) in the first 0.75 sec of the expiratory maneuver (FEVO75), but they utilized data t'rum previous studies that had shown no such trend when examining other risk factors at the time (4, 14). Gas stove usage was not found to be significant in either study. Ekwo et al. (10) found respiratory effects on children from gas stove usage but inconsistent effects of passive smoking. Tager et al. (57) and Schenker et a/. (48) found effects of parental smoking but not gas stoves; social status factors were again shown to be, important. Recently, several studies of passive smoking effects that included environ- mental monitoring have been performed. The results have been quite contradic- tory. In the studies by Florey et al. and Melia (11, 30, 31), looking at other indoor pollutants, they could tind no relation of passive smoking to children's lung func- tions. Speizer et al. (55, 56) demonstrated the effects of gas stoves but not of passive smoking on children. In the same study, the reverse was shown with different analyses of further data (59), but there were inconsistencies between different cohorts, even in the same city. Dodge (8, 9) examined children's lung function cruss-sectiunally and logitudinally; he did not find any effect of gas stove usage or passive smoking, although self-reported cough was higher in children expu,ed to parental smoking. Studies by Binder et al. and Schilling et al. (2, 49) fuund indoor levels of RSP higher in Connecticut homes with smokers, but no respiratory effects were seen in adults or children. Often, results have been in- cunsistent within the same study, or a single positive tinding has occurred without consideration for the number of comparisons examined. The 1983 Geneva Workshop on Environmental Smoke (47) concluded that no definitive studies of acute effects in children have been performed. Furthermore, there was difticulty in interpreting several of the studies due to the large num- bers of children involved and the qualitative responses found. Additionally the !,pecitic mechanisms of some irritant gases [such as sulfur dioxide (SO,_), ozone (U,), and formaldehyde] involved are also unknown. Retrospective studies of the etiology of respiratory diseases. such as asthma (13), have been discounted because of the major problem of reporting bias (4, 21). Patients with allergies appear to complain more of negative responses to environmental tobacco smoke (35. 40, 45, 54); as do many nonsmokers, but sub- jective responses generally have not been validated by epidemiological studies. The question also has been raised as to whether levels of environmental smoke per se found in the workplace could be considered sufficient to produce lung function effects in workers (58). There has been little evidence that passive smoking has an effect on chronic symptoms in adults (33, 47). In Tucson, we have attempted to investigate the relationship between peak flow ;,nd reported respiratory symptoms and passive smoking, air pollutants, and aCro- allergens (pollen-, bacilli, fungi, algae) in the microenvironment of the home as well as the macro- (ambient) environment. METHODS A total of 117 families (229 adult and child subjects) in Tucson were randomly TI BU 31643

6°To. and al to D itt A su nary ance .1Scfe 'f•f D, :hese Co. ~ foi- t. tl`tt: ge in .Avin6 w hen lt:m1C data LbollC ically :posi- sterol relop- >f nic- n pas- Aers,' ild be ironic ,vean ardiu- these rt rate study rably .vatt:d .ugical tte v e- tnicaiN tttal. in tt rateti SY,MPO51UM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING TABLE -t NICOTINE U['TAKJ: OF NONSMItKtittJ UNUL:H R<:AL•Llt•E CONUI'rIONs 633 Authors Sub)ects (exposure) No. ul nonsmokers investigated Results tng/mll kusNell and Feyerabend. 1975 (56) Employees in a hospital (urtne was delivered I hr after lunch) (al Group I la Nicutine in urtne: 12.4 s 16.9 [b) Group 2 13 Nicotine in urine: 8.9 x 9.1 Feyerabend rt Employees 31) Nicotine in urine; 7.5 x 8.3 ul., 1962 (a) Nunexposed to passive Nicotine in saliva: 5.9 s 4.4 121) smoking Jurmg the morning tself-repurted) (bl Exposed to passive :h Nicobne in urine: _1 •, : 289 nmuking during the morning lr / Nicotine in ealiva: 10.1 : 9.7 Fuhart rr ul.. (.e -repurtet ) Flight auendants (women) l8-hr 6 Nicotine in serum 1963 (I61 tlightl Before: 1.6 x 0.8 Atter: 3.2 : I.0 Nicutine excretion in urine: 12.9 s 6.5 µ(yM hr WrW rr d.. Outdoor patients and clinic atatT 19ri4 (615) (u) Nunexpustd to paaaive smoktng tselr-repurted) (bl Exposed to passive smok- ing (selt•repurted) W:dd und Ritchie. (a) Husbands (wife is a nun- 19tlb 1701 emuker) tbl HusbanJs twire is a smoker) Jurvin r( ul.. Empluyees in an uttice 1983 142) IBIuuJ, suliva, and urine sum- pies were deliveret Jt I t:3U .1M and 7:45 aM: period uf ubser- vauon included a.•hr stay in a ~muky pub.) ,7 Cutinine in urine: Median 2.0 ( lOth percentde: U.U: 911th percentile: 6.3) 199 Cutinine in urine: Median 6.0 ( IUth percentile: 1.4: 90th percentile: 2_.W IUI Cutinine in unne: S.5 z 1.3 20 Cunnine in urtne: 25.'_ = 14.8 7 (3efure rltter Nicotine in plasma = 7-44 Nicotine in ~altva I.vt1 43.63 Nicotine in urine 11.1.5 1 92.63 Cuunine ut plasma 1.0 7.33 Cuunme in saliva 1.50 S.U4 Counine in urine 4.80 12.94 on CO and nicotine. There is little evidence to suggest that other substances fuund in cigarette smoke may adversely affect the cardiovascular systtem (3g). fhuubh substances with such properties may exist, none have yet been identified. Carbon monoxide causes increased COHb levels, thereby reducing the blood's uxygen-carrying capacity and tissue oxygenation. Nicotine releases adrenalin t'rum the adrenal gland and other chromaffin tissues, and noradrenalin from the hy- puthallunus and nerve endings. The cardiovascular responses to nicotine, in gen- t:ral, parallel those that follow stimulation of the sympathico-adrenal system. Increase in pulse rate and blood pressure correlates with nicotine levels in blood (45). Carbon monoxide and nicotine act together insofar as CO diminishes the lrampurt capacity fur oxygen to the heart muscle, while nicotine increases the need of the activated heart muscle for oxygen. Since the heart is known to con- sume most of the oxygen available to it, any disorder that enhances the need for uxygen or decreases oxygen in the circulating blood could directly impair cardiac mtucle metabolism, thus leading to irreversible ischemic necrosis. Nicotine could %;:r,u,e a fatal arrhythmia by increasing myocardial irritability (9). It could also TI BU 31629

648 MICHAEL LEBOWLT2 chosen in a stratified sample of families in uniform geographic clusters from a representative community population [already under study (:0)], as previou,ly described (23, 26). Studies of the total population sample indicated the presence of relationships between weekly symptoms and both air pollutants and allergens, controlling for weather (44). Stratification ensured that the families were repre- sentative of families with and without asthma in their geographic clusters. They were studied over a 2-year period, using daily symptom diaries and peak flow meters; daily response rates and peak flows were acceptable for a majority of days in all seasons (23, 28). All families provided environmental information as to characteristics of the house, yard, and streets, heating, cooling, type of stove, water heater, washer and dryer, and smoking in the home (26). Daily ambient monitoring data for air pollutants, aero-allergens, and meteoro- logical variables with regional and geographic clusters were used (23). Indoorand yard (micro-outdoor) monitorings were conducted in a random cluster sample representative of all study households (23, 26) for particulate matter [RSP and total suspended particulate matter (TSP)], pollen, bacilli, fungi, and algae, for 72 hr each, and grab samples were made for CO and 03. Microenvironmental sam- pling of pollen, algae, and fungi have been described previously (23, 37, 38). Temperature and humidity varied by season, but the former did not vary very much indoors over the year and the latter did not differ much between indoors and outdoors over the year. TSP indoors ranged from 5.7 to 68.5 µg/m3, and simultaneous micro-outdoor TSP ranged from 2.1 to 169.6 µg/m3. Cyclone measurements of RSP had a max- imum of 49.8 µglm3 indoors and 124.5 µg/m' outdoors. Electron microscopy yielded little identifying information on the type of dust found on the tluors of homes. Intiltration of outdoor TSP indoors was similar to pollen intiltration: total micro-indoor pollen counts were approximately 5% of the micro-outdoor pollen counts, which were about 5% of macro-pollen rates. Both TSP and RSP were significantly associated with environmental tobacco smoke (Table 1), but not with gas stove usage (26). Averages of the two spot CO readings using infrared klecurminations were lower than 2.4 ppm; micro-outdoor readings were less than 3.8 ppm. Indoor CO was significantly associated with gas stove usage (26), but not with environmental tobacco smoke (Table 1), nor with gas-burning furnaces, water heaters, or washer-dryers (that is, vented appliances). Although low, it was used as the indicator pollutant measure of combustion emission from the gas stoves. Neither environmental tobacco smoke nor gas stove usage was associated with social status in this sample. RESULTS We found many interactions between indoor and indoor-outdoor pullutants, and weather, with respect to health effects. Children's standardized peak tlow (PEF) was related to outdoor environmental factors only (03, TSP, temperature, and relative humidity); outdoor factors at'fected nonsmokers more than passive amokers (Table 2). Adult PEF was related to active smoking (as were chronic symptoms), gas TI BU 31644 Tor.{L. Sc • Dii "" t3ui StoVe U and no relativc PEF micro-i attack~ RS 1', b Indu and nu ductiv, smukir creaazl lzrSich Micro- Crtlu

640 SCHIEVELBEIN AND RICHTER impair blood flow in the microcirculation of the heart by provoking arteriolar (5y)• spasm. In addition, it may play a role in thrombus formation (46). less t: The cardiovascular effects of cigarette smoke on passive smokers have been broul- investigated much less extensively than in active smokers. A number of studies ~'mul.- have shown that under real-life conditions, passive smokers inhale only approx- Frt imately 0.02 to 0.01 the amount of particulate matter taken up by active smokers an ct* (Tables 3 and 4). Carboxyhemoglobin increase in passive smokers under real-life that ~ conditions rarely exceeds 1% (Tables I and 2), while nicotine uptake is in the rcter- range where it can just be detected in blood stream (37, 56). Nicotine levels up wntL to 5 nglml obtained by gas chromatography should be interpreted with caution, because no zero level is found using this method even in subjects not exposed to tobacco smoke (22). In order to find measurable body tluid concentrations of nicotine and cotinine and a significant increase in COHb, experimental studies with high exposure to tobacco smoke were performed (Tables I and 3). Even in such experiments, it was not possible-to measure any cardiovascular effects such be conciuded that CO ~ =A. ai those found in active smokers From this tindin it ma g y . and nicotine uptake by passive smokers, under real-life conditions, do not exert an influence on the cardiovascular system or on metabolic processes. 1 3A Based on his findings Aronow (4, 5) claimed that patients with severely com- promised cardiovascular systems were adversely affected by CO and environ- 1 4. A mental tobacco smoke. He also reported a decrease in exercise time duration in i F '' i c tobacco smoke compared with the exercise t me patients exposed to atmospher of the same patients not exposed. Aronow's work, however, has been criticized y hIL for several reasons. Cardiovascular responses to passive smoking are difficult to measure, because influences other than those of tobacco smoke must first be excluded. Also, these tindings must he viewed from the standpoint that the cruual lG . 7.A study variable, the onset of angina pectoris, is highly subjective and that the stress factor was not controlled for. Most importantly, Aronow's work has been ~ a u called into question in light of his own admission to having reported false evidence in drug testing in humans (12). 1 9. z As mentioned above, Hirayama (36), in his large-scale cohort study following i~ up 265,1 lS adults in Japan over the period from 1966 to 1981. found an elevated risK ror tung cancer anu iscnemic neart utsease in nonsmotctng women marrteu to smokers. Risk increased steadily with the increase in the number of cigarettes smoked by husbands. However, in light of the findings already discusted, and in I 11. t, view of the followin a causal relationshi between facts as~ive smokin and g , p p g CHD must be doubted: t=- t. (a) In female smokers; the association between cigarette smoking and CHD is t3. ~ much weaker than in male smokers. In the Framingham Study, no significant effect of smoking was observed among actively smoking women (17, 44). lf in l4~ female active smokers an effect of smoking on the development of CHD cannot convincingly be demonstrated, it is difficult to assume that such an effect is ls ; possible in female passive smokers. (b) The incidence rate of CHD in pipe or cigar smokers is not, or is only slightly, . elevated compared with nonsmokers (34, 49) despite the fact that in many pipe 11 17. e~ and cigar smokers a high uptake of carbon monoxide and nicotine can be tound k I 11 BU 31636

640 SCHIEVELBEIN AND RICHTER impair blood flow in the microcirculation of the heart by provoking arteriolar (5y)• spasm. In addition, it may play a role in thrombus formation (46). less t: The cardiovascular effects of cigarette smoke on passive smokers have been broul- investigated much less extensively than in active smokers. A number of studies ~'mul.- have shown that under real-life conditions, passive smokers inhale only approx- Frt imately 0.02 to 0.01 the amount of particulate matter taken up by active smokers an ct* (Tables 3 and 4). Carboxyhemoglobin increase in passive smokers under real-life that ~ conditions rarely exceeds 1% (Tables I and 2), while nicotine uptake is in the rcter- range where it can just be detected in blood stream (37, 56). Nicotine levels up wntL to 5 nglml obtained by gas chromatography should be interpreted with caution, because no zero level is found using this method even in subjects not exposed to tobacco smoke (22). In order to find measurable body tluid concentrations of nicotine and cotinine and a significant increase in COHb, experimental studies with high exposure to tobacco smoke were performed (Tables I and 3). Even in such experiments, it was not possible-to measure any cardiovascular effects such be conciuded that CO ~ =A. ai those found in active smokers From this tindin it ma g y . and nicotine uptake by passive smokers, under real-life conditions, do not exert an influence on the cardiovascular system or on metabolic processes. 1 3A Based on his findings Aronow (4, 5) claimed that patients with severely com- promised cardiovascular systems were adversely affected by CO and environ- 1 4. A mental tobacco smoke. He also reported a decrease in exercise time duration in i F '' i c tobacco smoke compared with the exercise t me patients exposed to atmospher of the same patients not exposed. Aronow's work, however, has been criticized y hIL for several reasons. Cardiovascular responses to passive smoking are difficult to measure, because influences other than those of tobacco smoke must first be excluded. Also, these tindings must he viewed from the standpoint that the cruual lG . 7.A study variable, the onset of angina pectoris, is highly subjective and that the stress factor was not controlled for. Most importantly, Aronow's work has been ~ a u called into question in light of his own admission to having reported false evidence in drug testing in humans (12). 1 9. z As mentioned above, Hirayama (36), in his large-scale cohort study following i~ up 265,1 lS adults in Japan over the period from 1966 to 1981. found an elevated risK ror tung cancer anu iscnemic neart utsease in nonsmotctng women marrteu to smokers. Risk increased steadily with the increase in the number of cigarettes smoked by husbands. However, in light of the findings already discusted, and in I 11. t, view of the followin a causal relationshi between facts as~ive smokin and g , p p g CHD must be doubted: t=- t. (a) In female smokers; the association between cigarette smoking and CHD is t3. ~ much weaker than in male smokers. In the Framingham Study, no significant effect of smoking was observed among actively smoking women (17, 44). lf in l4~ female active smokers an effect of smoking on the development of CHD cannot convincingly be demonstrated, it is difficult to assume that such an effect is ls ; possible in female passive smokers. (b) The incidence rate of CHD in pipe or cigar smokers is not, or is only slightly, . elevated compared with nonsmokers (34, 49) despite the fact that in many pipe 11 17. e~ and cigar smokers a high uptake of carbon monoxide and nicotine can be tound k I 11 BU 31636

Parental smoking N Log IgE Both smoke 58 1.76 Father smokes tsb 1.85 Mother smokes =tl :.U`J Neither smoke 184 I.9l 650 MICHAEL LEBOWITZ TABLE 3 ASSOCIATIONS OF INDOOR TSP WITH AVERAGE PEAK FLOWS (STANDARllIZEll DEVIATION SCORES) IN ADULT ASTHMATICS PEF tnduur "CSP Mean - 1.17 - 1.92 - 1.56 SD U.~J9 1.83 1.54 Person days I.?U5 1.334 '_.539 Note. Analysis of variance. P< 0.0001. Deviations of individual values from seasonal means. standardized for seasonal variation. ° 25 + µyml. Thus, sensitive groups (children, asthmatics, allergics. AOD subjects) hacd more reactions to intiuur and outdoor pollutants. Children were found to spend more time outdoors, usually with exercise, and were affected by outdoor pollutants more; these results contirm previous findings in Tucson (19). Indoor pollutants affected sensitive adults more than did outdoor pollutants. Indoor TSP was as- sociated with health effects, but not passive smoking per se, indicating that other characteristics of indoor TSP were important (28). We have shown that active smukers have higher circulating immunoglubulin E (IgE), which with smuking is related to airway obstructive diseases (3). However, we were not able to see any increase in IgE in passive-smoking children (Table 4). Generally, our recent findings are compatible with our previous negative find- ings (21, ?2. 33). However, the blunting of irritant effects in children (seen in Table 2) needs further explanation. DISCUSSION Other effects of passive smoking have been reviewed recently (27, 47). The studies performed so far continue to provide confusing results vis-a-vis most public health issues. Many of the problems of pubitive studies are due to poor study design related to reporting bias, lack of control for family concordance. interaction among various pollutant emitters, lack of appropriate measurements (61), and crude statistical studies of geographical differences. Confounding factors need to be incorporated more explicitly and completely. Nute. Analysis ul' variance. P = 0.258. TABLE 4 CIRCULATING LOG IMMUNOGLOtiULIN E(IyE) IN CHILDREN BY CURRENT PARENTAL SMOKING TI BU 31646 hN m, al; PL to ru nL ri: Ct, (i) br CL w bL Ci Pz rlc ,E rc ar bL Pt. Nt

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING -Fur- encc. from .vuhn- c atttl rtMc- ,1. 31. unter attcn. xpeti- tlight Cur- -2334 e CO- •d. 29 lungen rctten_ as de:1 Archc umrn. au :utcl '_3 -426 pluces. :~cnted gWCItC )ke»." ., 1983. ti pwed, irbunyl onclun, 643 41. Janzon. L., Lindell, S.-E., Trell. E., and Larme. P. Smoking habits and earbuxyhemuglubin. A cross-sectional study of an urban population of middle-aged men. J. Epidrrniul. Cunanunity HrulfJt 35, 271-273 1l'Itfl). 42. Jarvis, M. J., Russell. M. A. H., and Feyer,tbcnd, C. Absorption of nicotine and carbon monoxide from passive smoking under natural conditions of exposure. Thurux 38, 829-833 (1y1S3). 43. Jones, J. G.. and Sinclair. A. Arterial disease amongst blast furnace workers. Ann. Uccup. Hyg. 18, I5-'_0 (1975). 44. Kannel, W. li. Update on the role of cigarette smoking in coronary artery disease. .4nrer. Heart J. 101, 31N-322f 11981). 45. Koch. A.. Hufl'mann, K.. Steck. W., Horsch, A., Hengen. N., Morl, H., Harenberg, J., Spohr, U.. and Weber, E. Acute cardiovascular reactions ufter cigarette smoking. Allrerwrlerusis 35, 67-75 (1980). 46. Levine. P. H. An acute effect of cigarette smoking on platelet function. A possible link between Nmuking and arterial thrombosis. Cirridutiun 48. 619-623 (1973). 47. Luyuettc. A. J., Landiss, C. W., and Merki, D. J. Some immediate effects of a smoking environ- ment on children of elementarv school atse. l. S'ch..Hrulth 40. 533-536 (1970). 48. Matsukura. S.. Sucoka, S., Yushimi. H., Yokota. M.. Hiratu, Y., and Fujita. T. Effect of passive smoking due to indoor tobacco smoke on plasm and urinary cotinine levels in nonsmokers. Jupun. J. NucL Mrd. 18, 700 /19111). 49. Mulcahy, R. Influence uf cigarette smoking on morbidity and mortality after myocardial infarction. in "Secondary Prevention of Coronary Heart Disease" (K. Pyurala, E. Rapaport, K. Konig, G. Schettler. and C. Diehm, Eds.). p. 23. Thieme, Stuttgart. 1983. 50. Pimm, P. E:.. Silverman, F., and Shephard, R. J. Physiological effects o1 acute passive'exposure to cigarette smoke. Arch. Environ. Heulth 33, 201-213 11978). 51. Report of the Surgeon General. "The Health Cunbcclucnces of Smoking. Cardiovascular Dis- eabe." Ruckville, Md.. 1983. 52. Report of the Surgeon General. "The Health Cunaeyuences of Smoking. Chronic Obstructive Lung Dise:ue." Rockville, Md.. 1984. 53. Rogers. W. R.. Bass. R. L., Juhnson, D. E., Kruski, A. W., McMahan, C. A., Montiel, M. M., Mutt, G. E.. Wilbur. R. L., and McGill. H. C. Atherosclerusis-relatetl responses tu cigarette smoking in the baboon. Circulation 61, I lii8-1193 119801. 54. Rummel. R. M.. Crawford. M., and )3ruce, P. The physiological effects of inhaling exhaled cig- arette smoke in relation to attitude ul' the nonsmoker. J. ScJt. Health 45. 524-529 /1975). 15. Russell. M. A. H., Cole. P. V., and Srown, E. Absorption by non-smokers of carbon monoxide from room air polluted by tobacco smoke. Lancet 1, 576-579 11973). ch. kus+ell, M. A. H., and Feyerabend. C. Blood and urinary nicuttne in non-smukers. Laacet 1, 179 - 1li 1 1 19751. s7. Schievelbein, H. Evaluation of the role of carbon monoxide and nicotine in the pathogenesis ut arteriuecleroais and cardiovascular disease. Prev. Med. 8. 379-389 119791. fn. Schieveibein. H.. Lundong. V., Lundong, W., Grumbach, H., and Remplik, V. Nicotine and arteriosclerosis. Z. Klin. C/trm. Klin. Biuchent. 8, 190-196 1 1970). !1). Schtevelbetn. H.. Richter. F., and Heller, W.-D. Nikotinaufnahme bei Zigarettcn-, Pfeifen- und Zigarrenrauchern (Unpublished data. 1984). hu. Jchulte. J. H. E1Teets of mild carbon monoxide intoxication. Arch. Environ. Health 7, 524-530 (1963). hl. Seppanen. A. Smoking in closed space and its effects on carboxyhaemoglobin saturation uf smoking and nonsmoking subjects. Ann. Clin. Rrs. 9, 281-283 (1977). n:. Seppanen, A., and Uusitalo, A. J. Carboxyhaemuglubin saturation in relation to smoking and v;,riuun occupational conditions. Ann. Clin. Res. 9, 261-268 (1977). Al Sievers, R. F.. Eth..ird.. 'l'. 1., Murr,ty, A. L., and Schrenk, H. H. Effect of exposure to known ,;arbun k;uncentratiuns uf carbon monoxide. Study uf tratlic ufticers ,tatiuned at the Holland tunnel u.twn. tur thirteen years. JAMrI 118, 585-588 11942). t*t. Stewart. R. D.. Eiarett'u, E. D.. Platte. L. R., Stewart, E. B., Kalblleix:h, J. H.. Van Yserlou, :y:creete I B., and Rimm. A. A. Carbuxyhetnuglubin levels in American blood donors. J.i.hlA 229, 1 IM7- 1195 (1'174). TI BU 31639

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING -Fur- encc. from .vuhn- c atttl rtMc- ,1. 31. unter attcn. xpeti- tlight Cur- -2334 e CO- •d. 29 lungen rctten_ as de:1 Archc umrn. au :utcl '_3 -426 pluces. :~cnted gWCItC )ke»." ., 1983. ti pwed, irbunyl onclun, 643 41. Janzon. L., Lindell, S.-E., Trell. E., and Larme. P. Smoking habits and earbuxyhemuglubin. A cross-sectional study of an urban population of middle-aged men. J. Epidrrniul. Cunanunity HrulfJt 35, 271-273 1l'Itfl). 42. Jarvis, M. J., Russell. M. A. H., and Feyer,tbcnd, C. Absorption of nicotine and carbon monoxide from passive smoking under natural conditions of exposure. Thurux 38, 829-833 (1y1S3). 43. Jones, J. G.. and Sinclair. A. Arterial disease amongst blast furnace workers. Ann. Uccup. Hyg. 18, I5-'_0 (1975). 44. Kannel, W. li. Update on the role of cigarette smoking in coronary artery disease. .4nrer. Heart J. 101, 31N-322f 11981). 45. Koch. A.. Hufl'mann, K.. Steck. W., Horsch, A., Hengen. N., Morl, H., Harenberg, J., Spohr, U.. and Weber, E. Acute cardiovascular reactions ufter cigarette smoking. Allrerwrlerusis 35, 67-75 (1980). 46. Levine. P. H. An acute effect of cigarette smoking on platelet function. A possible link between Nmuking and arterial thrombosis. Cirridutiun 48. 619-623 (1973). 47. Luyuettc. A. J., Landiss, C. W., and Merki, D. J. Some immediate effects of a smoking environ- ment on children of elementarv school atse. l. S'ch..Hrulth 40. 533-536 (1970). 48. Matsukura. S.. Sucoka, S., Yushimi. H., Yokota. M.. Hiratu, Y., and Fujita. T. Effect of passive smoking due to indoor tobacco smoke on plasm and urinary cotinine levels in nonsmokers. Jupun. J. NucL Mrd. 18, 700 /19111). 49. Mulcahy, R. Influence uf cigarette smoking on morbidity and mortality after myocardial infarction. in "Secondary Prevention of Coronary Heart Disease" (K. Pyurala, E. Rapaport, K. Konig, G. Schettler. and C. Diehm, Eds.). p. 23. Thieme, Stuttgart. 1983. 50. Pimm, P. E:.. Silverman, F., and Shephard, R. J. Physiological effects o1 acute passive'exposure to cigarette smoke. Arch. Environ. Heulth 33, 201-213 11978). 51. Report of the Surgeon General. "The Health Cunbcclucnces of Smoking. Cardiovascular Dis- eabe." Ruckville, Md.. 1983. 52. Report of the Surgeon General. "The Health Cunaeyuences of Smoking. Chronic Obstructive Lung Dise:ue." Rockville, Md.. 1984. 53. Rogers. W. R.. Bass. R. L., Juhnson, D. E., Kruski, A. W., McMahan, C. A., Montiel, M. M., Mutt, G. E.. Wilbur. R. L., and McGill. H. C. Atherosclerusis-relatetl responses tu cigarette smoking in the baboon. Circulation 61, I lii8-1193 119801. 54. Rummel. R. M.. Crawford. M., and )3ruce, P. The physiological effects of inhaling exhaled cig- arette smoke in relation to attitude ul' the nonsmoker. J. ScJt. Health 45. 524-529 /1975). 15. Russell. M. A. H., Cole. P. V., and Srown, E. Absorption by non-smokers of carbon monoxide from room air polluted by tobacco smoke. Lancet 1, 576-579 11973). ch. kus+ell, M. A. H., and Feyerabend. C. Blood and urinary nicuttne in non-smukers. Laacet 1, 179 - 1li 1 1 19751. s7. Schievelbein, H. Evaluation of the role of carbon monoxide and nicotine in the pathogenesis ut arteriuecleroais and cardiovascular disease. Prev. Med. 8. 379-389 119791. fn. Schieveibein. H.. Lundong. V., Lundong, W., Grumbach, H., and Remplik, V. Nicotine and arteriosclerosis. Z. Klin. C/trm. Klin. Biuchent. 8, 190-196 1 1970). !1). Schtevelbetn. H.. Richter. F., and Heller, W.-D. Nikotinaufnahme bei Zigarettcn-, Pfeifen- und Zigarrenrauchern (Unpublished data. 1984). hu. Jchulte. J. H. E1Teets of mild carbon monoxide intoxication. Arch. Environ. Health 7, 524-530 (1963). hl. Seppanen. A. Smoking in closed space and its effects on carboxyhaemoglobin saturation uf smoking and nonsmoking subjects. Ann. Clin. Rrs. 9, 281-283 (1977). n:. Seppanen, A., and Uusitalo, A. J. Carboxyhaemuglubin saturation in relation to smoking and v;,riuun occupational conditions. Ann. Clin. Res. 9, 261-268 (1977). Al Sievers, R. F.. Eth..ird.. 'l'. 1., Murr,ty, A. L., and Schrenk, H. H. Effect of exposure to known ,;arbun k;uncentratiuns uf carbon monoxide. Study uf tratlic ufticers ,tatiuned at the Holland tunnel u.twn. tur thirteen years. JAMrI 118, 585-588 11942). t*t. Stewart. R. D.. Eiarett'u, E. D.. Platte. L. R., Stewart, E. B., Kalblleix:h, J. H.. Van Yserlou, :y:creete I B., and Rimm. A. A. Carbuxyhetnuglubin levels in American blood donors. J.i.hlA 229, 1 IM7- 1195 (1'174). TI BU 31639

Parental smoking N Log IgE Both smoke 58 1.76 Father smokes tsb 1.85 Mother smokes =tl :.U`J Neither smoke 184 I.9l 650 MICHAEL LEBOWITZ TABLE 3 ASSOCIATIONS OF INDOOR TSP WITH AVERAGE PEAK FLOWS (STANDARllIZEll DEVIATION SCORES) IN ADULT ASTHMATICS PEF tnduur "CSP Mean - 1.17 - 1.92 - 1.56 SD U.~J9 1.83 1.54 Person days I.?U5 1.334 '_.539 Note. Analysis of variance. P< 0.0001. Deviations of individual values from seasonal means. standardized for seasonal variation. ° 25 + µyml. Thus, sensitive groups (children, asthmatics, allergics. AOD subjects) hacd more reactions to intiuur and outdoor pollutants. Children were found to spend more time outdoors, usually with exercise, and were affected by outdoor pollutants more; these results contirm previous findings in Tucson (19). Indoor pollutants affected sensitive adults more than did outdoor pollutants. Indoor TSP was as- sociated with health effects, but not passive smoking per se, indicating that other characteristics of indoor TSP were important (28). We have shown that active smukers have higher circulating immunoglubulin E (IgE), which with smuking is related to airway obstructive diseases (3). However, we were not able to see any increase in IgE in passive-smoking children (Table 4). Generally, our recent findings are compatible with our previous negative find- ings (21, ?2. 33). However, the blunting of irritant effects in children (seen in Table 2) needs further explanation. DISCUSSION Other effects of passive smoking have been reviewed recently (27, 47). The studies performed so far continue to provide confusing results vis-a-vis most public health issues. Many of the problems of pubitive studies are due to poor study design related to reporting bias, lack of control for family concordance. interaction among various pollutant emitters, lack of appropriate measurements (61), and crude statistical studies of geographical differences. Confounding factors need to be incorporated more explicitly and completely. Nute. Analysis ul' variance. P = 0.258. TABLE 4 CIRCULATING LOG IMMUNOGLOtiULIN E(IyE) IN CHILDREN BY CURRENT PARENTAL SMOKING TI BU 31646 hN m, al; PL to ru nL ri: Ct, (i) br CL w bL Ci Pz rlc ,E rc ar bL Pt. Nt

1'RL•V(:rrIVF MtiU1C1Nt l3, tii5-655 (l'/Ndl nlul.~'tUhin I), iulhetl in Influence of Passive Smoking on Pulmonary Function: A Survey''2 MICHAEL D. LEIIOWIT'L Uivh/,.n uj Respiruruiy Scienres, Wesrend Research Luburafurres. Univer.srt,y oj'Arrtuna Health Sciences Crnrc•r, Culh•tr o/',Y(edreine, Tru•sun, ariLuna 85724 A review uf the ctfects uf passive emuking on pulmunary function has been made. In ,;hildren. there is still a great deal uf confusiun and controversy. The range of the many eftects studied is limited: most prospective studies show small ctfects. in general. it appears that respiratory infecttuns in children may be increasrd with pa.mivc ,moking. The ettects un children's pulmonary function ranges from 0 to 3% over the aga range from conception to 2U yCar]: thLniC 111ItCrences in absolute magnitude are con5ttieredSmall. in heallhy uduits. Ihe effects on pulmonary function and symptoms are not considered to be ut cuncern: the Ctfech in asthmatics rCt{ulre turther study. A study to evaluate the above effect5 is presented here. It derives from a prospective study of atrway obstructive diseases in Ttlcson. Arizona. From that etudy. it was concluded that the effect of passive smoking on pulmonary function ur respiratory symptoms (recurt/ed on a daily babis) in'children is not positive. Eitects may be ,een with regurd to renpun.e+ to 8ther irritants in pabsivt-bmoking children. In adults, no etfect was seen, even in those with asthma or airway obstructive distaaes: this may be due to low dosage. It is concluded that turther, more Jppropl'lale studies are needed to understand the rolC uf paasive smoking on pulmonary function. 9 1vw+Ac„demic Nre», trw. INTRODUCTION Although there is substantial agreement that passive, or involuntary, tobacco smoking produces annoyance and sensory irritant effects, its role in producing cardiovascular or respiratory diseases is less clear (34, 47). The specitie contri- hutiuns of passive smoke to personal exposures and their effects have not been ,uliicicntly documented. Several studies have shown that the levels of carbon monoxide and other ,idcatrt:am smoke constituents in indoor areas are high in many instances (34, 43, 46. 47). Deposition of sidestream cigarette smoke in the human respiratory tract has been shown to be only about 11%, much less than mainstream smoke (16). The impact of environmental tobacco on indoor concentrations of nitrogen diuxide (NO,) is usually small compared with the impact of gas stove, (6-t). Blood carbon monoxide (CO) levels may be higher in those exposed to passive smoke 136, 4Z), but are probably lower than the carboxyhemoglobin (COHb) levels as- ,uciated with exposures to unvented gas stoves (34). Some of the same respiratory diseases may occur in indoor situations in which hit;h levels of respirablc sus- pended paniculates (RSP) and gases are due to wood-burning or kerosene heaters t 1, 53) and-not to environmentaL tobact:u smoke. . Iht+ wurk was ~uppurtetd by NHLBI SCOR Grant HL14136. Nrr,cntrd at the Symposium ",Ivtedical Perspectives on Passive Smuking," April 9-12. 19K4. ~ icnnu. Austria. 645 1Nri11-7435r1S1 53.t111 lupynght ~ IYtl4 hy A-Jcmic Preao. lnc. All nghb ut repruJucuun m rny lurm roaenod. TI BU 31641

X nI; ed ~tl :es in ) is -tnt in Zot tN SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKINU 641 i59). Cigar and pipe smokers inhale tobacco smoke both actively, although much less than cigarette smokers, and passively. They may, in fact, be regarded as the group of passive smokers exposed to the highest concentrations of tobacco smoke. From the data available we conclude that passive smoking is not likely to have an effect on the development and progression of CHD. We conclude, theretore, that the 1983 KReport of the Surgeon General on CHD is correct in making no reference to passive smoking ah a possible cause .(5 t). Our view also seems to contorm with the prevailing scientit7c opinion as expressed thus far (27, 44). REFERENCES 1. Ahmed. S. S., ,Moschos. C. B., Lyons, M. M., Oldewurtel, H. A.. Cuumbis, R. J., Regan. T. J., and lenkins, B. Cardiuva4cular effects of long-term cigarette smoking and nicotine adminis- tratiun. Aneer. J. Curdiul. 37, 33-4t1(1976). :. Anderson. E. W.. Andelman, R. J., Sirauch, 1. M., Fortuin, N. J.. and Knelson, J. H. Effect of low-level carbon monoxide exposure on onset and duration of angina pectoris. Ann. Intern. ,t?ed. 79, 46-50 (1973). 3. Armitage, A. K., D•avies, R. F., and Turner, D. M. The effects of carbon monoxide on the de- velupment of athcrosclerosis in the white carneuu pigeon. Atherosc•lerusis 23, 333-344 (1976). 4. Arunuw, W. S. Eff'ect of passtve smoking on angina prctor'ts. New Enl;l. J. Sh•d. 299, 21-24 I ly7tf). 5. Aronow, W. S. Aggravation of angina pectoris by two percent carboxyhemuglobin. Antrr. Heart I. 101. 154-157 ( t9tl 1). h. Astrup. P., Kjcldhon, K., and Wanstrup, 1. Enhancing influence of c:icbun monoxide on the development of atheromatosis in chulesteroi-fed rabbits. J. Athervsc•(er. Res. 7, 343-354 1190). 7. Ayres, S. M.. Evans, R. G., and Buehler, M. E. The prevalence of carboxyhemoglubinemia in New Yorkers and its effects on the coronary and systemic circulation. Prev. .b1ed. 8, 323-332 (1979). tl. Badre, lt., Guillerm. R., Abran. N., Bourdin, M., and Dumas, C. Pollution atmospheriyue par la tumre de tabac. Ann. Ph,,,,n. Fr. 36, 443-452 (1978). 9. Bellct. S., DeGurman. N. Z., Kustis, J. B., Roman. L., and Fleischmann, D. The effect of in- halation of cigarette smoke on ventricular fibrillation threshold in normal dogs and dogs with acute myucardial infarction. Ainer. Heart J. 83, 67-76 ('1972). tU. liing, R. J.. Sarma, J. S. M., Weishaar, R., Rackl, A.. and P•awlik. G. Biochemical and histological effects of intermittent carbon monoxide exposure in cynumulgus monkeys (Muruc•u Jascic•u- luris) in relation to atherosclerosis. J. Clin. Phurntuc•ot. 20, 487-499 (19801. 11. Buoyse, F. M.. Osikuwicz. G.. and Quarfoot. A. J. Effects of chronic oral consumption of nicotine on the rabbit aurtic endothelium. Amer. J. Purhul. 102, 229-238 (1981). 12. Budiansky, S. t}ata falsification. Food and drug data fudged. Nature ILondoa) 302. 5611 (1983). 13. Cano. J. P.. Catulin, J.. Budre. R.. Dumas. C.. Vialla. A., and Guillerm. R. Datermtnation de la nicotine par chromatographie en phase gazeuse. 11. Applications. Ann. Pharm. Fr. 28, b33- 640 (1y70). la. Carsky, J.. Hrnciarova, M., Rusnakova, S., Busuva. 8., and P•ajdlhauserova, An tarboxyhemo- glubm level in blood donors in Bratislava. 1lrutisl. Lek. Listy 74, 273-278 (1980). ls Cole. P V. Comparative effects of atmospheric pollution and cigarette smoking on carboxyhe- muglubin levels in man. Nuture ILundurrl 255, 699-701 (1975). 1n. Dahms, T. E.. Bolin. J. F., and Slavin, R. G. P•assive smoking. Eff"t, on bronchial asthma. Chr.rr 80. 53U-534(1ytit). 17 Uawber. T. R. "The Framingham Study. The Epidemiology of Atherosclerotic Disease." Harvard Univ. Press, Cumbrtdge. Muss., 1980. TI BU 31637

X nI; ed ~tl :es in ) is -tnt in Zot tN SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKINU 641 i59). Cigar and pipe smokers inhale tobacco smoke both actively, although much less than cigarette smokers, and passively. They may, in fact, be regarded as the group of passive smokers exposed to the highest concentrations of tobacco smoke. From the data available we conclude that passive smoking is not likely to have an effect on the development and progression of CHD. We conclude, theretore, that the 1983 KReport of the Surgeon General on CHD is correct in making no reference to passive smoking ah a possible cause .(5 t). Our view also seems to contorm with the prevailing scientit7c opinion as expressed thus far (27, 44). REFERENCES 1. Ahmed. S. S., ,Moschos. C. B., Lyons, M. M., Oldewurtel, H. A.. Cuumbis, R. J., Regan. T. J., and lenkins, B. Cardiuva4cular effects of long-term cigarette smoking and nicotine adminis- tratiun. Aneer. J. Curdiul. 37, 33-4t1(1976). :. Anderson. E. W.. Andelman, R. J., Sirauch, 1. M., Fortuin, N. J.. and Knelson, J. H. Effect of low-level carbon monoxide exposure on onset and duration of angina pectoris. Ann. Intern. ,t?ed. 79, 46-50 (1973). 3. Armitage, A. K., D•avies, R. F., and Turner, D. M. The effects of carbon monoxide on the de- velupment of athcrosclerosis in the white carneuu pigeon. Atherosc•lerusis 23, 333-344 (1976). 4. Arunuw, W. S. Eff'ect of passtve smoking on angina prctor'ts. New Enl;l. J. Sh•d. 299, 21-24 I ly7tf). 5. Aronow, W. S. Aggravation of angina pectoris by two percent carboxyhemuglobin. Antrr. Heart I. 101. 154-157 ( t9tl 1). h. Astrup. P., Kjcldhon, K., and Wanstrup, 1. Enhancing influence of c:icbun monoxide on the development of atheromatosis in chulesteroi-fed rabbits. J. Athervsc•(er. Res. 7, 343-354 1190). 7. Ayres, S. M.. Evans, R. G., and Buehler, M. E. The prevalence of carboxyhemoglubinemia in New Yorkers and its effects on the coronary and systemic circulation. Prev. .b1ed. 8, 323-332 (1979). tl. Badre, lt., Guillerm. R., Abran. N., Bourdin, M., and Dumas, C. Pollution atmospheriyue par la tumre de tabac. Ann. Ph,,,,n. Fr. 36, 443-452 (1978). 9. Bellct. S., DeGurman. N. Z., Kustis, J. B., Roman. L., and Fleischmann, D. The effect of in- halation of cigarette smoke on ventricular fibrillation threshold in normal dogs and dogs with acute myucardial infarction. Ainer. Heart J. 83, 67-76 ('1972). tU. liing, R. J.. Sarma, J. S. M., Weishaar, R., Rackl, A.. and P•awlik. G. Biochemical and histological effects of intermittent carbon monoxide exposure in cynumulgus monkeys (Muruc•u Jascic•u- luris) in relation to atherosclerosis. J. Clin. Phurntuc•ot. 20, 487-499 (19801. 11. Buoyse, F. M.. Osikuwicz. G.. and Quarfoot. A. J. Effects of chronic oral consumption of nicotine on the rabbit aurtic endothelium. Amer. J. Purhul. 102, 229-238 (1981). 12. Budiansky, S. t}ata falsification. Food and drug data fudged. Nature ILondoa) 302. 5611 (1983). 13. Cano. J. P.. Catulin, J.. Budre. R.. Dumas. C.. Vialla. A., and Guillerm. R. Datermtnation de la nicotine par chromatographie en phase gazeuse. 11. Applications. Ann. Pharm. Fr. 28, b33- 640 (1y70). la. Carsky, J.. Hrnciarova, M., Rusnakova, S., Busuva. 8., and P•ajdlhauserova, An tarboxyhemo- glubm level in blood donors in Bratislava. 1lrutisl. Lek. Listy 74, 273-278 (1980). ls Cole. P V. Comparative effects of atmospheric pollution and cigarette smoking on carboxyhe- muglubin levels in man. Nuture ILundurrl 255, 699-701 (1975). 1n. Dahms, T. E.. Bolin. J. F., and Slavin, R. G. P•assive smoking. Eff"t, on bronchial asthma. Chr.rr 80. 53U-534(1ytit). 17 Uawber. T. R. "The Framingham Study. The Epidemiology of Atherosclerotic Disease." Harvard Univ. Press, Cumbrtdge. Muss., 1980. TI BU 31637

646 v11Ct1AEL LEkit)WITZ PULMONARY FUNCTION IN ADULTS There have been several studies of passive 5mokin5 an,l re,piratory impairment in adults. Comstock et u/. (6) evaluated respiratory symptoms and lung function in a community population and found some effects from gas stoves but no effects from passive smoking; no exposure measurements were made. Lehuwitz and his co-workers (21, 22) studied chronic respiratory symptoms and lung function in a community population and fuund no effects ot passive smoking. Kautfimann et u!. (18) studied pulmonary function in adult populations in France (without indoor measurements or information about use of gas appliances). They tound some effects of spouse's smoking on nonsmoking women, but the healthiest population lived in the most polluted areas. This indicated that the study might have been flawed by biased population selection, biased testing, or confuundirig factors. One epidemiological study of chronic bronchitis in Czechoslovakia l5'_) showed a signiticant increase of symptoms of this disease in passive smokers, but this occurred almost entirely among men and showed a much greater relation to other adverse factors and/or exposures. The Harvard Six City Study (59) pertormed covariance analyses for adults, examining passive smoking, gas stove usage, and ambient pollution (by city) and controlling for at;e. and Smokint;.-!t found n, o effect of passive smoking on the adult respiratory system. Other population studies reported at a 1983 NIH workshop (33) led to the conclusion that there was basi,:ally no effect on pul- monary function and maybe only slight, inconsistent ctfects on symptoms in adults. White and Froeb (60) studied self-selected volunteers with regard to workplace exposure and reported some effects ut' ra,sive smoking in a subset of their vul- unteers; their study h:ts been discussed in detail (23). Even-«ith a biased pupu- latiun, poor study design, and incorrect statistical evaluation, there were nu clear- cut, consistent, medically meaningful differences between passivc ,mukers and groups of nonsmokers; a corrected statistical analysis strengthened this cunclu- siun. Three physiological studies of the effects of environmental tobacco smoke ex- posure on healthy subjects have been conducted, and all have been negative with regard to ,uch effects (31). 51), even when massive amounts ot' smokz were used to expose nonsmokers (41). Two similar studies of the effects of environmental smoking on asthmatics and controls have demonstrated cuntlicting findings. One study (50) showed no important effect on asthmatics uj :untrols, whereas the other study (7) demonstrated a broncho-constrictor effect in highly reactivc a,th- matics at higher concentrations. PULMONARY FUNCTION IN CHILOREN Studies of the relationship of children's lung function to parental smoking habits again show somewhat contradictory evidence. In some studies, such as those of Martinez et u!. (29), Yarnell and St. Leger (63). ;and Kauffmann and Brille (17), effects of parental smoking were seen in terms uf childrc:n's lung tLnctian: how- ever, these did nut measure or control for other indoor (or uutduur) pollutants. TI BU 31642 In uthc were t( exptain tirst U. previor at the t Ekwu incunsi founti L again n. Rzc,~ mental tory. tn pottuta: tions. : passive differer differet func:tiu usage L expus.: found i re~spira consist conhitlc The detinitii there '+ bers ut ~peciti. I Uti), ac Retrk ( 13), h~ '_t). P:: envirott '. jective The qu,- per ye ' func:tiut !,mukin. In Tu and rcp allerzer well ah A wt

646 v11Ct1AEL LEkit)WITZ PULMONARY FUNCTION IN ADULTS There have been several studies of passive 5mokin5 an,l re,piratory impairment in adults. Comstock et u/. (6) evaluated respiratory symptoms and lung function in a community population and found some effects from gas stoves but no effects from passive smoking; no exposure measurements were made. Lehuwitz and his co-workers (21, 22) studied chronic respiratory symptoms and lung function in a community population and fuund no effects ot passive smoking. Kautfimann et u!. (18) studied pulmonary function in adult populations in France (without indoor measurements or information about use of gas appliances). They tound some effects of spouse's smoking on nonsmoking women, but the healthiest population lived in the most polluted areas. This indicated that the study might have been flawed by biased population selection, biased testing, or confuundirig factors. One epidemiological study of chronic bronchitis in Czechoslovakia l5'_) showed a signiticant increase of symptoms of this disease in passive smokers, but this occurred almost entirely among men and showed a much greater relation to other adverse factors and/or exposures. The Harvard Six City Study (59) pertormed covariance analyses for adults, examining passive smoking, gas stove usage, and ambient pollution (by city) and controlling for at;e. and Smokint;.-!t found n, o effect of passive smoking on the adult respiratory system. Other population studies reported at a 1983 NIH workshop (33) led to the conclusion that there was basi,:ally no effect on pul- monary function and maybe only slight, inconsistent ctfects on symptoms in adults. White and Froeb (60) studied self-selected volunteers with regard to workplace exposure and reported some effects ut' ra,sive smoking in a subset of their vul- unteers; their study h:ts been discussed in detail (23). Even-«ith a biased pupu- latiun, poor study design, and incorrect statistical evaluation, there were nu clear- cut, consistent, medically meaningful differences between passivc ,mukers and groups of nonsmokers; a corrected statistical analysis strengthened this cunclu- siun. Three physiological studies of the effects of environmental tobacco smoke ex- posure on healthy subjects have been conducted, and all have been negative with regard to ,uch effects (31). 51), even when massive amounts ot' smokz were used to expose nonsmokers (41). Two similar studies of the effects of environmental smoking on asthmatics and controls have demonstrated cuntlicting findings. One study (50) showed no important effect on asthmatics uj :untrols, whereas the other study (7) demonstrated a broncho-constrictor effect in highly reactivc a,th- matics at higher concentrations. PULMONARY FUNCTION IN CHILOREN Studies of the relationship of children's lung function to parental smoking habits again show somewhat contradictory evidence. In some studies, such as those of Martinez et u!. (29), Yarnell and St. Leger (63). ;and Kauffmann and Brille (17), effects of parental smoking were seen in terms uf childrc:n's lung tLnctian: how- ever, these did nut measure or control for other indoor (or uutduur) pollutants. TI BU 31642 In uthc were t( exptain tirst U. previor at the t Ekwu incunsi founti L again n. Rzc,~ mental tory. tn pottuta: tions. : passive differer differet func:tiu usage L expus.: found i re~spira consist conhitlc The detinitii there '+ bers ut ~peciti. I Uti), ac Retrk ( 13), h~ '_t). P:: envirott '. jective The qu,- per ye ' func:tiut !,mukin. In Tu and rcp allerzer well ah A wt

SYMPOSIUhI: MEDICAL PERSPEC'CIVES ON PASSIVE SMOKING 647 pairment function .o effects. mptoms ; passive n France :s). They lealthiest .iy might tounding akia (52) ,mokers, - relation r adults. ,:ity) and oking on -)g3 N1H : on pul- ?toms in orkplace heir vul- -d popu- nu clear- ~ kers and conclu- nuke ex- tive with cre used mmental zgs. One reas the ive asth- ig habits those uf Ale (17). )n; huw- I lut;uus. In other studies in which measurements were not made, contradictory results were found. Hasselblad et al. (15) demonstrated that maternal smoking habits explain 0. I~r of the variance in children's forced expiratory volume (FEV) in the first 0.75 sec of the expiratory maneuver (FEVO75), but they utilized data t'rum previous studies that had shown no such trend when examining other risk factors at the time (4, 14). Gas stove usage was not found to be significant in either study. Ekwo et al. (10) found respiratory effects on children from gas stove usage but inconsistent effects of passive smoking. Tager et al. (57) and Schenker et a/. (48) found effects of parental smoking but not gas stoves; social status factors were again shown to be, important. Recently, several studies of passive smoking effects that included environ- mental monitoring have been performed. The results have been quite contradic- tory. In the studies by Florey et al. and Melia (11, 30, 31), looking at other indoor pollutants, they could tind no relation of passive smoking to children's lung func- tions. Speizer et al. (55, 56) demonstrated the effects of gas stoves but not of passive smoking on children. In the same study, the reverse was shown with different analyses of further data (59), but there were inconsistencies between different cohorts, even in the same city. Dodge (8, 9) examined children's lung function cruss-sectiunally and logitudinally; he did not find any effect of gas stove usage or passive smoking, although self-reported cough was higher in children expu,ed to parental smoking. Studies by Binder et al. and Schilling et al. (2, 49) fuund indoor levels of RSP higher in Connecticut homes with smokers, but no respiratory effects were seen in adults or children. Often, results have been in- cunsistent within the same study, or a single positive tinding has occurred without consideration for the number of comparisons examined. The 1983 Geneva Workshop on Environmental Smoke (47) concluded that no definitive studies of acute effects in children have been performed. Furthermore, there was difticulty in interpreting several of the studies due to the large num- bers of children involved and the qualitative responses found. Additionally the !,pecitic mechanisms of some irritant gases [such as sulfur dioxide (SO,_), ozone (U,), and formaldehyde] involved are also unknown. Retrospective studies of the etiology of respiratory diseases. such as asthma (13), have been discounted because of the major problem of reporting bias (4, 21). Patients with allergies appear to complain more of negative responses to environmental tobacco smoke (35. 40, 45, 54); as do many nonsmokers, but sub- jective responses generally have not been validated by epidemiological studies. The question also has been raised as to whether levels of environmental smoke per se found in the workplace could be considered sufficient to produce lung function effects in workers (58). There has been little evidence that passive smoking has an effect on chronic symptoms in adults (33, 47). In Tucson, we have attempted to investigate the relationship between peak flow ;,nd reported respiratory symptoms and passive smoking, air pollutants, and aCro- allergens (pollen-, bacilli, fungi, algae) in the microenvironment of the home as well as the macro- (ambient) environment. METHODS A total of 117 families (229 adult and child subjects) in Tucson were randomly TI BU 31643

642 SCHIEVELBEIN AND RICHTER 18. Dube. 1q. F. • aud Green. C. R. Methods of collection of ,mukc fur analytical purpuses, irt "Fur- mation. Analysts, and Composition of Tobacco Smuke." Recent Advances in Tobacco Science. Vol. 8. p. 42. 36th Tobacco Chemists Research Cunterence. Raleigh, 1982. 19. Eckardt. R. E.. [vlacFarland, H. N., Alarte, Y. C. E., and Busty, W. M. The biologic effect from lung-term expusure ot' primates to carbon monoxide. A.rcG. Ern•irurt. Hrulcfi 25, 381-387I 1972). 20. Einbrodt. M. J.. kohmanith. J., and Prajsnar. D. Der Catlmiumgehult im dlut und Rauchgewohn- heiten. Nuucrwissensc•huJrrrt 63, 148 (19761. 21. Fryerabend, C.. Higenbottam, T., and Russell. M. A. H. Nicotine concentrations in urine and saliva of smokers and nun-amokers. Urit. Med. J. 284, 1W2-11x14 (1982). 22. Feyerabend, C.. and Russell, M. A. H. Improved gas-chrumacogruphic method and microextruc- tion technique !ur the measurement of nicotine in biological fluids. J. Pkurnt. Phurrnacu(. 31, 73-76 (1979). 23. Fischer. A. Passivrauchen. Ausmass und Wirkung der Luftverunreinigung durch Tabakrauch untcr cxperunrntnllcn Bedingungen und in FcldversuchCn. Doctoral disacrtatiun. Zurich. 1979. 24. Fischer, C• Weber, A., and GrandJean, E. Luftverunreinigung durch Tabakruuch in Guststatten. tnt. Arc•h• Occ•up. Environ. HrultJt 41. 267-280 (19781. 25. Fisher, E. R., Rothstein. R.. Wholey, M. H., and Nelson, R. (ntluence of nicotine on eupcri- mental atherusclerusis and its determinants. Arch. PcultuJ. 96, 298-304 (1973). 26. Fuliart. D.. dtnuwitz, N. L.. and Becker. C. E. Passive absorption of nicotine in airline flight attendants. New Engl. J. Mrd. 308, 1105 (1983). 27. Frishman, W. H. Involuntary ,ntuking: Cardiovascular effects ot' smoke on nonsmokers. Cur- Jiuvusc•. ,Ned. 4, aSy-296 (1979). 28. Harkr, H.-P. Zum Problem des "Passiv-Rauchens.'• Muenc•h. Med. Wuc•/tr,nE 1tr. 112, 2328-2334 (1970). 29. Harke. H.-P. Zum Problem des Passivrauchensv 1. Uber den Eintluss des Rauchens auf die CO- Kunzentration in Hururaumen. lut. Arc•la Arbrits+nrd. 33, 199-?tw (1y74). 30. Harke. H.•P., and Bleichert. A. Zum Problem des Passivrauchens. lctr. Arc•!t. Arbei(snted. 29, 312-322 11972). 31. Harke. H.-P.. Lictll, W., and Denker, D. Zum Problem des Pttssivrauchcns. ll. Untersuchungen uber den Kohlenmonoxidgehult der Lut't im KraftCahrteug durch daa Rauchen von Zigarettcn. tnt. Arch. Arbritsmrd. 33, 207-220 (1974). ~ 32. Harke. H.-P., and Peters, H. Zum Problem des Passivrauchens. 111. Uber den Eintluss des Rauchens auf die CO-Konzentration im Kraftfahrteug bei Fahrten im Statltgebict. Du. Arclt. Arbrirsmrd. 33, 2_1-229 (1974). 33. Harmsen. H.. and Effenberger. E. Tubakrauch in Verkehrsmitteln, Wuhn- und Arbcitsraumen. Arc•It. Hyg. 8akteriol. 141, 383--i1f0 (1957). 34. Hickey, N.. Mulcahy, R.. Daly. L.. Graham. I., O'Donughue, S., and Kennedy, C. Cigar and pipe smoking related tu fuur-year survival uP coronary patients. Brit. Heurr J. 49, 4'_3-4'_b (1983). 35. Hinds. W. C., and First. M. W. Cuncenu•ations of nicotine and tobacco smoke in public places. New Engf. J. Mcd. 292, 844-845 (1975). 36. Hirayama, T. "Lung Cancer in Japan. Effects ut' Nutrition and Passive Smoking." Paper presented at the U.S-Japan meeting on New Etiology of Lung Cancer. Honolulu, 1983. 37. Hoffmann, D., Haley, N. J., Brunnemann, K. D.. Adams. 1. D., and Wynder, E. L. "Cigarette sidestream smoke: Formation, analysis and model studies on the uptake by nonsmokers." Paper presented at the U.S.-Japan meeting on New Etiology of Lung Cancer. Hunululu, 1983. 38. Hugod. C.. and Asttup, P. Studies of wronary and aurtic intimal morphology in rabbits exposed to gas phase constituents of tobacco smoke (hydruyen cyanicie. nitne oxide, and carbunyl sulphide), irt "Smuking and Arterial DisCase" (R. M. GrCenha)6ii, tid.)• p. 89. Pitman, London. 1981. 39. Hugtxi. C.. Hawkins. L. H.. Kjeldsen. K., Thomsen, H. K.. and Astrup• P. Effect of wrbun monoxide exposure on aurtic and coronary intitnal morphology in the rabbit. A revaluation. Atlrerusc•lerusis 30, 333-342 (1978). 4/)• Hurshman. L. G.. Brown, B. S.. and Guyton. R. G. The implicatiuns of sidestream cigarette smoke fur cardiovascular health. J. Ern•irurt. Heult{t 41, 145-149 (1978). TI BU 31638 47. Lu, 48. 1 ta; 44. Ntu ~ 50. Pia

642 SCHIEVELBEIN AND RICHTER 18. Dube. 1q. F. • aud Green. C. R. Methods of collection of ,mukc fur analytical purpuses, irt "Fur- mation. Analysts, and Composition of Tobacco Smuke." Recent Advances in Tobacco Science. Vol. 8. p. 42. 36th Tobacco Chemists Research Cunterence. Raleigh, 1982. 19. Eckardt. R. E.. [vlacFarland, H. N., Alarte, Y. C. E., and Busty, W. M. The biologic effect from lung-term expusure ot' primates to carbon monoxide. A.rcG. Ern•irurt. Hrulcfi 25, 381-387I 1972). 20. Einbrodt. M. J.. kohmanith. J., and Prajsnar. D. Der Catlmiumgehult im dlut und Rauchgewohn- heiten. Nuucrwissensc•huJrrrt 63, 148 (19761. 21. Fryerabend, C.. Higenbottam, T., and Russell. M. A. H. Nicotine concentrations in urine and saliva of smokers and nun-amokers. Urit. Med. J. 284, 1W2-11x14 (1982). 22. Feyerabend, C.. and Russell, M. A. H. Improved gas-chrumacogruphic method and microextruc- tion technique !ur the measurement of nicotine in biological fluids. J. Pkurnt. Phurrnacu(. 31, 73-76 (1979). 23. Fischer. A. Passivrauchen. Ausmass und Wirkung der Luftverunreinigung durch Tabakrauch untcr cxperunrntnllcn Bedingungen und in FcldversuchCn. Doctoral disacrtatiun. Zurich. 1979. 24. Fischer, C• Weber, A., and GrandJean, E. Luftverunreinigung durch Tabakruuch in Guststatten. tnt. Arc•h• Occ•up. Environ. HrultJt 41. 267-280 (19781. 25. Fisher, E. R., Rothstein. R.. Wholey, M. H., and Nelson, R. (ntluence of nicotine on eupcri- mental atherusclerusis and its determinants. Arch. PcultuJ. 96, 298-304 (1973). 26. Fuliart. D.. dtnuwitz, N. L.. and Becker. C. E. Passive absorption of nicotine in airline flight attendants. New Engl. J. Mrd. 308, 1105 (1983). 27. Frishman, W. H. Involuntary ,ntuking: Cardiovascular effects ot' smoke on nonsmokers. Cur- Jiuvusc•. ,Ned. 4, aSy-296 (1979). 28. Harkr, H.-P. Zum Problem des "Passiv-Rauchens.'• Muenc•h. Med. Wuc•/tr,nE 1tr. 112, 2328-2334 (1970). 29. Harke. H.-P. Zum Problem des Passivrauchensv 1. Uber den Eintluss des Rauchens auf die CO- Kunzentration in Hururaumen. lut. Arc•la Arbrits+nrd. 33, 199-?tw (1y74). 30. Harke. H.•P., and Bleichert. A. Zum Problem des Passivrauchens. lctr. Arc•!t. Arbei(snted. 29, 312-322 11972). 31. Harke. H.-P.. Lictll, W., and Denker, D. Zum Problem des Pttssivrauchcns. ll. Untersuchungen uber den Kohlenmonoxidgehult der Lut't im KraftCahrteug durch daa Rauchen von Zigarettcn. tnt. Arch. Arbritsmrd. 33, 207-220 (1974). ~ 32. Harke. H.-P., and Peters, H. Zum Problem des Passivrauchens. 111. Uber den Eintluss des Rauchens auf die CO-Konzentration im Kraftfahrteug bei Fahrten im Statltgebict. Du. Arclt. Arbrirsmrd. 33, 2_1-229 (1974). 33. Harmsen. H.. and Effenberger. E. Tubakrauch in Verkehrsmitteln, Wuhn- und Arbcitsraumen. Arc•It. Hyg. 8akteriol. 141, 383--i1f0 (1957). 34. Hickey, N.. Mulcahy, R.. Daly. L.. Graham. I., O'Donughue, S., and Kennedy, C. Cigar and pipe smoking related tu fuur-year survival uP coronary patients. Brit. Heurr J. 49, 4'_3-4'_b (1983). 35. Hinds. W. C., and First. M. W. Cuncenu•ations of nicotine and tobacco smoke in public places. New Engf. J. Mcd. 292, 844-845 (1975). 36. Hirayama, T. "Lung Cancer in Japan. Effects ut' Nutrition and Passive Smoking." Paper presented at the U.S-Japan meeting on New Etiology of Lung Cancer. Honolulu, 1983. 37. Hoffmann, D., Haley, N. J., Brunnemann, K. D.. Adams. 1. D., and Wynder, E. L. "Cigarette sidestream smoke: Formation, analysis and model studies on the uptake by nonsmokers." Paper presented at the U.S.-Japan meeting on New Etiology of Lung Cancer. Hunululu, 1983. 38. Hugod. C.. and Asttup, P. Studies of wronary and aurtic intimal morphology in rabbits exposed to gas phase constituents of tobacco smoke (hydruyen cyanicie. nitne oxide, and carbunyl sulphide), irt "Smuking and Arterial DisCase" (R. M. GrCenha)6ii, tid.)• p. 89. Pitman, London. 1981. 39. Hugtxi. C.. Hawkins. L. H.. Kjeldsen. K., Thomsen, H. K.. and Astrup• P. Effect of wrbun monoxide exposure on aurtic and coronary intitnal morphology in the rabbit. A revaluation. Atlrerusc•lerusis 30, 333-342 (1978). 4/)• Hurshman. L. G.. Brown, B. S.. and Guyton. R. G. The implicatiuns of sidestream cigarette smoke fur cardiovascular health. J. Ern•irurt. Heult{t 41, 145-149 (1978). TI BU 31638 47. Lu, 48. 1 ta; 44. Ntu ~ 50. Pia

The WHO/EURO Working Group (62) examined passive smoking within the context of all indoor exposures. It was felt that the current levels of knowledge did not yet allow quantitative assessment ot' the impact on the public health, as the knowledge of exposure effect relationships was inadequate, especially fur delayed effects of chronic exposures. ln terms of exposure levels, there were no data on how many cigarettes were being smoked in each of those indoor places where passive smoking can occur, and the ability to estimate how this is distrib- uted within the building stock is marginal. It was felt that the amount of current information about the fraction of the population exposed to levels high enough to be a public health concern is smaller. The amount of information about the specitic pollutants that might be responsible for the effects of concern is also very small. Of the variety of potential adverse affects, it was felt that only for irritation, odor, and airway response is their even a marginal amount of current knowledge. This group also felt that there were many possible pollutant interactions and that there were many other confounding exposures involved in examining the effects of passive smoking. Thus, in terms of relative priurity, effects ut' passive smoking were low in comparison with other concerns, based on knowledge levels. Since, however, passive smoking may indeed be an important public health concern, it should be studied further. Attempts are currently being made by various groups tu ;unsider potential efticacious methods for further studies of the health effects ot' passive smoking (33, 47). REFERENCES I. Anderson. H. R. Respiratory abnurmalitits, smoking habits, and ventilatory capacity in a high- land community in Papua, New Guinea: Prevalence and effect on murtality. lnt. J. Epidrmiul. 8. 127-135 (1979). 2. Binder, R. E.. Mitchell, C. A.. Hubein.. H. R., and Bouhuys. A. Importance of the indoor cnvi- ronmenl in air pollution exposure. Arch. Eirvirurr. Heult/r 31, 277-279 (1976). 3. Burrows. B., Lebowitz, M. D., Barbee. R. A.. Knudson. R. J., and Haiunen. M. tnteractiuns ut smoking and immunologic lucturs in relation to airways obstruction. Chrsr 84, 657-661 (1y83). 4. Cederluf, R.. and Cullry, J. Epidcmiulugical investigations on environmental tobacco omukr. S'cunrl. J. Rrspir. Dis. Srrppl. 91, 47-49 (1974). 5. Chapman, R. S., Habstlblad. V.. Hayes, C. G.. Williams. J. V. R.. and Hummer. D. I. Air pul- lution and childhood ventitatury function. l. E.tipu.ure tu particulate mutter in two southeastern cities. 1971-1972, in "Clinical lmplicatiunb of Air Pollution Resettrch" (A. J. Finkel and W. C. Duel. Eds.), pp. 285-303. Publ. Sci. Group. inc.. Actun. Ytubs., 1976. 6. Cumstuc:k. G., Meyer, M. B.. Hrlsing. K. J., and Tuckman. M. S. Respiratory effects of huuae• huld exposures to tobacco smukc and gub cooking. .irr+er. Rev. Rrspir. Uis. 124, 143-148 (19t31). 7. Dahms. T. E.. Bulin. M. D., and Slavin. R. G, Passive smoking: Effects of bronchial abthma. Chest 80, 53U-534 (1981). 8. Dodge. R. R. The effects of indoor air pollution un Arizona children. Arc•Jt. Envirun. Health 37, 151-155 (1982). 9. Dodge. R. R. The respiratory hcalth and lung function of Anglo-American children in a,melter , town. Anrrr. Ri-s•. Rrapir. Uix. 127; 1"-161 11yt4}). - - 10. Ekwo, E. E., Weinberger, M. M., Lachenbruch. P. A.. and Hunttay. W. H. Relationship u[ pa- rontal smoktng and gas cooking to respiratory disease in children. C/rr3t 84, 6h2-6tN4 (14zsit 11. Florey. C. du Ve, Melia. R. 1. W., Chinn. S.. Guldatein, 13. lJ . liruuks, A. G. F., John. H. II . Br.tigherul, 1. B.. and Webhter, X. The relation between respiratury illness in primary xhuul- TI BU 31648 15. Ha 16. Hil 17. K. i s. K:: ly. Lc :0. Lc =l. L~ 22. Lc _3. Lt 24. L~ 35. L '_6. L 27. L _8. L~ 3:. ti

SYMPOSIUM: MEDICAL 1'EkSPECCIVES ON PASSIVE SMOKING 655 .4sh- t2- .ung_ -I len 'uM ,sit': un- lrg. ;up. b of csC: ; _k!, ,i. 7-1- cn: -n14 on SB. Valent[n. H.. 13USt. H.-P, and WBwru, E. Das PAlbtvrauChen am ArbeltsplACz-elne GeSunllhetts- ,chadigung" Zentralblatt fur k3akteriulugie. P,trasitenUuncle, 1nCekttunsrankheiten und Hygiene; Erste Abteilung: Onginale. Reihe B: Hygiene. Prdventivr .tilecli,in 167, 405-434 (197;t1. 59. Wure. J. H.. Dockery. D. W.. Spiro. A., III, Speizer. F E...cnj Ferris. B. G.. Jr. Passrve smuking, gas cooking. and resptrutury health uF chddren living in ~rx cities. amrr. Rev. Rrspir. Uis. 129. 366-374 (1980. M). White. 3.. and Froeb. H. Small-airways Jystunctiun in non-smokers chrunieally exposed to tu- baccu smoke. Yrn f:rr41. J. .Wrd. 302, 720-723 (19KU). bl. Wurld Health Orgunirrtiun (WHO). 1982. "rstimating Human Exposure to Air Pollutants." WHO, CupenhageNGeneva, 19ri2. n.. WHO1EU1tU. "Indoor Air Pollutants: Exposure and Health Effects Assessments." EURO Publ. N,r 71l. World Health Organization. Copenhagen, 1983. h3. Yarneil, J. W. G.. and St. Leger. A. S. Respiratory illness, maternui smoking habit and lung tunctiun in children. Brit. J. Uii. Chesr 73. 230-236 (1979). td,. Yucum. J. E. Indour-uutduur air quality relatiunships: A critical review. J. Air Pulfru. Control rl.seuc. 32, 5UU-5_U (19M2). • TI BU 31651

SYMPOSIUM: MEDICAL 1'EkSPECCIVES ON PASSIVE SMOKING 655 .4sh- t2- .ung_ -I len 'uM ,sit': un- lrg. ;up. b of csC: ; _k!, ,i. 7-1- cn: -n14 on SB. Valent[n. H.. 13USt. H.-P, and WBwru, E. Das PAlbtvrauChen am ArbeltsplACz-elne GeSunllhetts- ,chadigung" Zentralblatt fur k3akteriulugie. P,trasitenUuncle, 1nCekttunsrankheiten und Hygiene; Erste Abteilung: Onginale. Reihe B: Hygiene. Prdventivr .tilecli,in 167, 405-434 (197;t1. 59. Wure. J. H.. Dockery. D. W.. Spiro. A., III, Speizer. F E...cnj Ferris. B. G.. Jr. Passrve smuking, gas cooking. and resptrutury health uF chddren living in ~rx cities. amrr. Rev. Rrspir. Uis. 129. 366-374 (1980. M). White. 3.. and Froeb. H. Small-airways Jystunctiun in non-smokers chrunieally exposed to tu- baccu smoke. Yrn f:rr41. J. .Wrd. 302, 720-723 (19KU). bl. Wurld Health Orgunirrtiun (WHO). 1982. "rstimating Human Exposure to Air Pollutants." WHO, CupenhageNGeneva, 19ri2. n.. WHO1EU1tU. "Indoor Air Pollutants: Exposure and Health Effects Assessments." EURO Publ. N,r 71l. World Health Organization. Copenhagen, 1983. h3. Yarneil, J. W. G.. and St. Leger. A. S. Respiratory illness, maternui smoking habit and lung tunctiun in children. Brit. J. Uii. Chesr 73. 230-236 (1979). td,. Yucum. J. E. Indour-uutduur air quality relatiunships: A critical review. J. Air Pulfru. Control rl.seuc. 32, 5UU-5_U (19M2). • TI BU 31651

1'RL•V(:rrIVF MtiU1C1Nt l3, tii5-655 (l'/Ndl nlul.~'tUhin I), iulhetl in Influence of Passive Smoking on Pulmonary Function: A Survey''2 MICHAEL D. LEIIOWIT'L Uivh/,.n uj Respiruruiy Scienres, Wesrend Research Luburafurres. Univer.srt,y oj'Arrtuna Health Sciences Crnrc•r, Culh•tr o/',Y(edreine, Tru•sun, ariLuna 85724 A review uf the ctfects uf passive emuking on pulmunary function has been made. In ,;hildren. there is still a great deal uf confusiun and controversy. The range of the many eftects studied is limited: most prospective studies show small ctfects. in general. it appears that respiratory infecttuns in children may be increasrd with pa.mivc ,moking. The ettects un children's pulmonary function ranges from 0 to 3% over the aga range from conception to 2U yCar]: thLniC 111ItCrences in absolute magnitude are con5ttieredSmall. in heallhy uduits. Ihe effects on pulmonary function and symptoms are not considered to be ut cuncern: the Ctfech in asthmatics rCt{ulre turther study. A study to evaluate the above effect5 is presented here. It derives from a prospective study of atrway obstructive diseases in Ttlcson. Arizona. From that etudy. it was concluded that the effect of passive smoking on pulmonary function ur respiratory symptoms (recurt/ed on a daily babis) in'children is not positive. Eitects may be ,een with regurd to renpun.e+ to 8ther irritants in pabsivt-bmoking children. In adults, no etfect was seen, even in those with asthma or airway obstructive distaaes: this may be due to low dosage. It is concluded that turther, more Jppropl'lale studies are needed to understand the rolC uf paasive smoking on pulmonary function. 9 1vw+Ac„demic Nre», trw. INTRODUCTION Although there is substantial agreement that passive, or involuntary, tobacco smoking produces annoyance and sensory irritant effects, its role in producing cardiovascular or respiratory diseases is less clear (34, 47). The specitie contri- hutiuns of passive smoke to personal exposures and their effects have not been ,uliicicntly documented. Several studies have shown that the levels of carbon monoxide and other ,idcatrt:am smoke constituents in indoor areas are high in many instances (34, 43, 46. 47). Deposition of sidestream cigarette smoke in the human respiratory tract has been shown to be only about 11%, much less than mainstream smoke (16). The impact of environmental tobacco on indoor concentrations of nitrogen diuxide (NO,) is usually small compared with the impact of gas stove, (6-t). Blood carbon monoxide (CO) levels may be higher in those exposed to passive smoke 136, 4Z), but are probably lower than the carboxyhemoglobin (COHb) levels as- ,uciated with exposures to unvented gas stoves (34). Some of the same respiratory diseases may occur in indoor situations in which hit;h levels of respirablc sus- pended paniculates (RSP) and gases are due to wood-burning or kerosene heaters t 1, 53) and-not to environmentaL tobact:u smoke. . Iht+ wurk was ~uppurtetd by NHLBI SCOR Grant HL14136. Nrr,cntrd at the Symposium ",Ivtedical Perspectives on Passive Smuking," April 9-12. 19K4. ~ icnnu. Austria. 645 1Nri11-7435r1S1 53.t111 lupynght ~ IYtl4 hy A-Jcmic Preao. lnc. All nghb ut repruJucuun m rny lurm roaenod. TI BU 31641

w lns. ore ore nts .nu, as- hi:r iE er. ble nd- in 'he Ust )or :its SYMPOStUM: MEDICAL PEKSPECTIVES ON PASSIVE SMOKING 651 Better ascertainment and control of the covariables and confounders are needed (12). These include the following: (a) Unvented combustion products from combustiun appliances: gas (especially ,tuves). wood burning, and kerosene burning. (b) Effects of other indoor pollutants (e.g., formaldehyde, consumer products), mostly irritant in nature: including interaction effects. (c) Other indoor conditions of biological importance, such as (i) temperature and humidity conditions, size of house, ventilation (e.g., air exchange rates); (ii) pollen, molds, mites, and other ailergens; and (iii) infectious organisms. lcU Outdoor and occupational exposures and personal mobility, representing tutallintegrated exposure (includes consideration of geographic region, urban/ rural setting). (e) Socioeconomic status, cultural, ethnic, and religious factors (re: crowding, number of siblings, household conditions, occupation of parents, host conditions, risk of infectious spread, medical knowledge, medical care, reporting biases, ctc. ). ltl Demographic and medical characteristics of the study population, such as li) age, sex, marital status; (ii) presence of underlying respiratory condition (e.g., bronchial reactivity, clinical airways obstruction, atopy); (ii) disability and/or cumorbidity. (g) Parental syndromes, like productive cough, t,vhich would mask other pro- cesses/mechanisms (e.g., genetic, infectious) or affect reporting (4, 21) (e.g., winter morning phlegm). Ihl Maternal smoking during pregnancy and resulting effects on fetus and new- born, including low birth weight (33, 46). (i) Annoyance responses, including odor problems, other psychological or so- cia! responses to tobacco smoking in a nonsmoker. (j) Unknown smoking by subject. A major problem to be resolved is the determination of the actual dose to the passive smoker (34, 61, 62). Monitoring of airborne particulates and organics is needed (47). Respir•able- particulates indoors are usually dritirn by smoking (2, ?6, 64). However, the smoking-generated fraction of these cunc:Lntratiuns is nut related to health effects ('_li. 49). This implies that sizing and speciation of particles are necessary (61, 62). The differentiation of sources and concentrations would be very useful in determining mechanism of effects as well as the sources of primary concern. Friedman and colleagues (12) showed, by questionnaire responses of about 3n.Ul)0 adults, that 63% of people were exposed to environmental tobacco smoke, but their reported exposure was poorly correlated with serum thiocyanate levels. Exposures were correlated significantly with other health-related characteristics, including age, use of alcohol and marijuana, and exposure to occupational haz- arda. They weru ,il,u :orrelated with race, lower educational !,tatus, anti ,ingle marital statua. The reported concordance of marital smoking habits by categorization of spouse's smoking resulted in considerable misclassification, in- dicating spouse's smoking habits are a very inaccurate index of exposure. Future ,tudies of passive-smoking effects should thus take these tindings into account. TI BU 31647

w lns. ore ore nts .nu, as- hi:r iE er. ble nd- in 'he Ust )or :its SYMPOStUM: MEDICAL PEKSPECTIVES ON PASSIVE SMOKING 651 Better ascertainment and control of the covariables and confounders are needed (12). These include the following: (a) Unvented combustion products from combustiun appliances: gas (especially ,tuves). wood burning, and kerosene burning. (b) Effects of other indoor pollutants (e.g., formaldehyde, consumer products), mostly irritant in nature: including interaction effects. (c) Other indoor conditions of biological importance, such as (i) temperature and humidity conditions, size of house, ventilation (e.g., air exchange rates); (ii) pollen, molds, mites, and other ailergens; and (iii) infectious organisms. lcU Outdoor and occupational exposures and personal mobility, representing tutallintegrated exposure (includes consideration of geographic region, urban/ rural setting). (e) Socioeconomic status, cultural, ethnic, and religious factors (re: crowding, number of siblings, household conditions, occupation of parents, host conditions, risk of infectious spread, medical knowledge, medical care, reporting biases, ctc. ). ltl Demographic and medical characteristics of the study population, such as li) age, sex, marital status; (ii) presence of underlying respiratory condition (e.g., bronchial reactivity, clinical airways obstruction, atopy); (ii) disability and/or cumorbidity. (g) Parental syndromes, like productive cough, t,vhich would mask other pro- cesses/mechanisms (e.g., genetic, infectious) or affect reporting (4, 21) (e.g., winter morning phlegm). Ihl Maternal smoking during pregnancy and resulting effects on fetus and new- born, including low birth weight (33, 46). (i) Annoyance responses, including odor problems, other psychological or so- cia! responses to tobacco smoking in a nonsmoker. (j) Unknown smoking by subject. A major problem to be resolved is the determination of the actual dose to the passive smoker (34, 61, 62). Monitoring of airborne particulates and organics is needed (47). Respir•able- particulates indoors are usually dritirn by smoking (2, ?6, 64). However, the smoking-generated fraction of these cunc:Lntratiuns is nut related to health effects ('_li. 49). This implies that sizing and speciation of particles are necessary (61, 62). The differentiation of sources and concentrations would be very useful in determining mechanism of effects as well as the sources of primary concern. Friedman and colleagues (12) showed, by questionnaire responses of about 3n.Ul)0 adults, that 63% of people were exposed to environmental tobacco smoke, but their reported exposure was poorly correlated with serum thiocyanate levels. Exposures were correlated significantly with other health-related characteristics, including age, use of alcohol and marijuana, and exposure to occupational haz- arda. They weru ,il,u :orrelated with race, lower educational !,tatus, anti ,ingle marital statua. The reported concordance of marital smoking habits by categorization of spouse's smoking resulted in considerable misclassification, in- dicating spouse's smoking habits are a very inaccurate index of exposure. Future ,tudies of passive-smoking effects should thus take these tindings into account. TI BU 31647

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 649 GiT a. .1sly nce -ns, pre- hey TU w y uf n a5b JVe, jro- and zple and 1r T! ,am- 3S). 'rery our5 luur nax- :Upy 'sot _utaE >I(en vcre .v i th )wer w;,ts :ntal or the ther >Caal .nts, :low urc:, sivt~ TABLE I TOTAL SUSPENDED P.VtTICULATES (TSP), RESPIKAbLE SUSPENDED P.,,K'rtCULi'rES (RSP), AND CAKtlON MONUXIllE (CO) BY ENVIKONMENTAL TOIfACCU SMOKE ETS TSP (µgtms) Mean 17.7 33.0• SD 10.7 18.2 N 14 19 RSP (µglml) Mean 7.2 13.t•• S D 6.0 13.1 ,V 15 19 CO (µg/ms) Mean 1.2 1.., S D 0.6 0.6 N 14 11. • Difference by two-tailed t test, P < 0.005. •• Borderline significant (two-tailetl t test) 0.05 < P < 0.10. stove use, and outdoor NO, (independent and interactive) in asthmatics, allergics, and normals. In some cases, PEF was related to outdoor 03, temperature, and relative humidity in nunnurmal groups combined. • PEF in adult asthmatics was related also to temperature, high relative humidity, micro-pollen, and indoor TSP (Table 3), but not to passive smoking. "Asthma attacks" in adult asthmatics were related to temperature and to indoor TSP and RSP. but not to passive smoking or outdoor TSP (28). Indoor TSP (but not passive smoking) was associated with rhinitis in asthmatics and nonasthmatics. as was RSP. Eye irritation in all adults and increased pro- ductive cough in nonasthmatics also were related to indoor TSP, but not to passive Nmuking. In additiun, gas stove usage (and indoor CO) was associated with in- crcaazd rhinitis in all adults, increased productive cough in asthmatics and al- lergics, increased sore throat in nonasthmatics, and eye irritation in all adults. Micro-pollen was associated also with adult eye irritation and rhinitis (in summer). TABLE 2 - CHILDKEN'S MEAN PEAK FLOWS (EXPKESSED AS STANDARDIZED DEVIATION SCOKhS)°: EFFECT OF OUTSIDE OZONE IN PASSIVE AND NONSMOKERS (PEKSON DAYS = 1.366) OUtSide ozone Nunsmokers` 0.075 0.059 -0.175 -U.4ii44 tla»ive smukers 0.071 -U.U41 -0.046 -U.U79 ,vurr. Overall ANOVA P < 0.QUUt. ° Ueviutiuna ut' individual values from seasonal means, uand;trdized for sC;uunai variatiun. ° High: Max hourly O, (ppm) > 0,08. Analyms of variance, P < O.W1. 'W/ )_- IS+1y,, ?igtttlicantly greater day-to-day change than neen in normal. 1.2). TI BU 31645

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 649 GiT a. .1sly nce -ns, pre- hey TU w y uf n a5b JVe, jro- and zple and 1r T! ,am- 3S). 'rery our5 luur nax- :Upy 'sot _utaE >I(en vcre .v i th )wer w;,ts :ntal or the ther >Caal .nts, :low urc:, sivt~ TABLE I TOTAL SUSPENDED P.VtTICULATES (TSP), RESPIKAbLE SUSPENDED P.,,K'rtCULi'rES (RSP), AND CAKtlON MONUXIllE (CO) BY ENVIKONMENTAL TOIfACCU SMOKE ETS TSP (µgtms) Mean 17.7 33.0• SD 10.7 18.2 N 14 19 RSP (µglml) Mean 7.2 13.t•• S D 6.0 13.1 ,V 15 19 CO (µg/ms) Mean 1.2 1.., S D 0.6 0.6 N 14 11. • Difference by two-tailed t test, P < 0.005. •• Borderline significant (two-tailetl t test) 0.05 < P < 0.10. stove use, and outdoor NO, (independent and interactive) in asthmatics, allergics, and normals. In some cases, PEF was related to outdoor 03, temperature, and relative humidity in nunnurmal groups combined. • PEF in adult asthmatics was related also to temperature, high relative humidity, micro-pollen, and indoor TSP (Table 3), but not to passive smoking. "Asthma attacks" in adult asthmatics were related to temperature and to indoor TSP and RSP. but not to passive smoking or outdoor TSP (28). Indoor TSP (but not passive smoking) was associated with rhinitis in asthmatics and nonasthmatics. as was RSP. Eye irritation in all adults and increased pro- ductive cough in nonasthmatics also were related to indoor TSP, but not to passive Nmuking. In additiun, gas stove usage (and indoor CO) was associated with in- crcaazd rhinitis in all adults, increased productive cough in asthmatics and al- lergics, increased sore throat in nonasthmatics, and eye irritation in all adults. Micro-pollen was associated also with adult eye irritation and rhinitis (in summer). TABLE 2 - CHILDKEN'S MEAN PEAK FLOWS (EXPKESSED AS STANDARDIZED DEVIATION SCOKhS)°: EFFECT OF OUTSIDE OZONE IN PASSIVE AND NONSMOKERS (PEKSON DAYS = 1.366) OUtSide ozone Nunsmokers` 0.075 0.059 -0.175 -U.4ii44 tla»ive smukers 0.071 -U.U41 -0.046 -U.U79 ,vurr. Overall ANOVA P < 0.QUUt. ° Ueviutiuna ut' individual values from seasonal means, uand;trdized for sC;uunai variatiun. ° High: Max hourly O, (ppm) > 0,08. Analyms of variance, P < O.W1. 'W/ )_- IS+1y,, ?igtttlicantly greater day-to-day change than neen in normal. 1.2). TI BU 31645

SY+APUSIU.M: MEDICAL PERSPECTIVES UN PASSIVL SMOKING 657 1Ctlatl- t'LE <•usrs uf tke :u !~muke nvuiving wbacw ,!)iblc tu .Cr). CX- vC) ;,nd knalybis. piraory ned and by this un. It u let weert ne hand It dules answered lnction in a certainr '13). Other :zcent re- ~) ;utd by t trassive ,tigatiuns pri maril y rnevitably )nsidered ,;ontrast, ying any :m using - I :. 1984. PATIENTS AND METHODS A total of 1,351 people (941 men and 410 women) took part in the study, which wa, carried out in lytS2-fyg3. Participation in the study was voluntary. All par- ticipants were white-collar workers, the majority of whom were employed in ,etlentary office jobs. Test subjects were recruited from the staff of an adminis- tratiun authority and two large industrial companies located in Northern Bavaria, a region with a relatively low level of air pollution. All those participating anon- ymously tilled out a standardized questionnaire specially developed for this study, and upon completion handed it in a closed envelope to the person carrying out the examinations. The pulmonary function tests were carried out with an electronic spirometer.3 The device was calibrated either daily or prior to each new series of measure- ments. The spirometric examination was conducted by two trained examiners with a knowledge of medicine. Special attention was paid to good cooperation un the part of the test subject with respect to the respiration exercises, which were generally repeated two to three times. Of the various pulmonary functiun parameters, those listed in Table I were included in the t'urther analysis. These parameters were identical to those used in comparable studies. These data were transferred, anonymously, to documents designed for t:lec- tronic data processing. Evaluation of the data was effected on a data processing hystem of the type Cyber t345.' After various plausibility checks and appropriate adjustment ot' the data record and group, the overall group was subdivided into live subgruups (Fig. I ): (a) Nrrrr sircvkers (NS), defined as persons who have never been regularly exposed to tobacco smoke, either actively or passively. (b) Passive smukers (PS). defined as subjects who have never actively smoked. but who are currently exposed to passive smoking. Here, three subgroups were differentiated: (i) PS exclusively with household exposure to passive smoking (PS/H). (ii) PS exposed passively to tobacco smoke only at their place of work (PS/W), and (iii) persons with a combination of passive smoke exposure (PS/HW). (c) E.r-smukers (ES). defined as persons who had given up active tobacco smoking at least 6 months previously. (d) Current smukers (CS), representing the group of persons who, at the time ut' the investigation were actively smoking and inhaling the smoke. (e) Utlwr sinukers (US), representing a remaining category in which noninhaling cigarette smokers and cigar and pipe smokers were grouped together. Cun,iderable care was expended on the standardization of the analytical pul- munary function data. For all the pulmonary function parameters, on the basis of the NS group, formulae for intertial predicted values were computed, using ' Siregnost FDIO. XY Recorder E::Itf. Siemens. Ertangen, FRG. ' Cuntrul Data Corporation. Minneapolis. Minn. TI BU 31653

SY+APUSIU.M: MEDICAL PERSPECTIVES UN PASSIVL SMOKING 657 1Ctlatl- t'LE <•usrs uf tke :u !~muke nvuiving wbacw ,!)iblc tu .Cr). CX- vC) ;,nd knalybis. piraory ned and by this un. It u let weert ne hand It dules answered lnction in a certainr '13). Other :zcent re- ~) ;utd by t trassive ,tigatiuns pri maril y rnevitably )nsidered ,;ontrast, ying any :m using - I :. 1984. PATIENTS AND METHODS A total of 1,351 people (941 men and 410 women) took part in the study, which wa, carried out in lytS2-fyg3. Participation in the study was voluntary. All par- ticipants were white-collar workers, the majority of whom were employed in ,etlentary office jobs. Test subjects were recruited from the staff of an adminis- tratiun authority and two large industrial companies located in Northern Bavaria, a region with a relatively low level of air pollution. All those participating anon- ymously tilled out a standardized questionnaire specially developed for this study, and upon completion handed it in a closed envelope to the person carrying out the examinations. The pulmonary function tests were carried out with an electronic spirometer.3 The device was calibrated either daily or prior to each new series of measure- ments. The spirometric examination was conducted by two trained examiners with a knowledge of medicine. Special attention was paid to good cooperation un the part of the test subject with respect to the respiration exercises, which were generally repeated two to three times. Of the various pulmonary functiun parameters, those listed in Table I were included in the t'urther analysis. These parameters were identical to those used in comparable studies. These data were transferred, anonymously, to documents designed for t:lec- tronic data processing. Evaluation of the data was effected on a data processing hystem of the type Cyber t345.' After various plausibility checks and appropriate adjustment ot' the data record and group, the overall group was subdivided into live subgruups (Fig. I ): (a) Nrrrr sircvkers (NS), defined as persons who have never been regularly exposed to tobacco smoke, either actively or passively. (b) Passive smukers (PS). defined as subjects who have never actively smoked. but who are currently exposed to passive smoking. Here, three subgroups were differentiated: (i) PS exclusively with household exposure to passive smoking (PS/H). (ii) PS exposed passively to tobacco smoke only at their place of work (PS/W), and (iii) persons with a combination of passive smoke exposure (PS/HW). (c) E.r-smukers (ES). defined as persons who had given up active tobacco smoking at least 6 months previously. (d) Current smukers (CS), representing the group of persons who, at the time ut' the investigation were actively smoking and inhaling the smoke. (e) Utlwr sinukers (US), representing a remaining category in which noninhaling cigarette smokers and cigar and pipe smokers were grouped together. Cun,iderable care was expended on the standardization of the analytical pul- munary function data. For all the pulmonary function parameters, on the basis of the NS group, formulae for intertial predicted values were computed, using ' Siregnost FDIO. XY Recorder E::Itf. Siemens. Ertangen, FRG. ' Cuntrul Data Corporation. Minneapolis. Minn. TI BU 31653

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 653 lha Jge . ~~ tur nu ces :ib- k:nt lgh the :ry -)n, ge. h~tt ~ts tn~ it tps .:ts nvi- ~ uf ~3). ke. wl- ern C. IJe- I 4a n4- 37, ter ,:htldren and the use of g,u fur cooking. lll. Nitrogen diuxide, respiratory illnebs and lung lunctiun. Int. J. E'pidc•tniul. 8, 347-353 (1979). t_. Frtcdman, G. D., Petitti, D. B.. and Bawol, R. D. Prevalence and correlates uf passive -smuking. rtmrr. J. Piiblic Health 73, 401-dt15 (1983). 0. Gunmaker, S. L.. Walker. D. K.. Jacobs, F. H.. and Ruch-Ruas, H. Parental smoking and the risk of childhood asthma. Arnrr. J. Public Health 72, 574-579 (1982). Id. H•rmmer. t). 1.. Miller. F J., Stead. A. G., and Hayes, C. G. Air pollution and childhuod venti- Iatury function. 1. Exposure to particulates matter in two southeastern cities, 1971-1972, in "Clinical Implications of Air Pollution Research" (A. 1. Finkel and W. C. Duel. Eds). pp. 321- 337. Publishing Sciences Group, Acton, Mass., 1976. 15. H:uxlblad, V.. Humble. C. G.., Gr.tham, M. G.. and Anderson, H. S. Indoor environmental determinants uf lung function in children. Antrr. Rev. Respir. Dir. 123, 479-485 (1981). lb. Hiller, F. C., McCuskcr. K. T., 4latumber, M. K., Wilson. 1. D., and Bone, R. C. Deposition of ~iJ"uN:un cigarette smoke in the human respiratory tract. Ainer. Rrv. Rerpir. Dis. 125, 4Ub- 4UIS (1982). 17. K.,uffmtlnn, F., and Brille, D. Respiratory function in passive smokers' children, in PROC lEA, . Edinburgh. August 1981. !b. Kauffmann. F., Tessier. J. F.. and Oriul. P. Adult passive smoking in the home environment: A risk factor for chronic aiRlow limitation. atner. J. EpidrnLiuL 11l, _b9-2ML) t 1983). 19. Lebowitz, M. D., Bundheim, P., Cristea, G., Markuwitc. U., %aiaazek, J., Staniec, M., and Van Wyck, U. The effect of air pollution and weather on lung function in exercising children and adolescents. atnrr. Rev. Rrspir. Dfs. 109, 262-273 (19741. :U. Lebuwitz, M. D.. Knudson. R. J., and Burrows. B. The Tucson epidemiology study of chronic obstructive lung disease. 1. Methodology and prevalence uf disease. .4mer. J. Epidetruul. 102, 137-152 (19751. 21. Lebuwitt. M. U., and Burrows. B. Respiratory symptoms related tu smoking habi[s of family adults. Chest 69, 4S-5(1 (1976). ::. Lebowitz, M. I).. Armet, D. B., and Knudson. R. The effect of passive smoking on pulmonary tunctiun in children. Envirorr. Int. 8, 371-374 (I9tl2). _3. Lebowitz. M. D.. O'Rourke. M. K., Dlxlge, R.. Holberg. C. 1., Corman. G., Hushaw, R. W., Pinnas, J. L.. Barbee, R. A., and Sneller. M. R. The adverse health effects of biological aeru- suls, other aerosols, and indoor micruclimute on asthmatics and nun-asthmutics. Envirun. Int. 8, 375-3li0119ii2). 24. Lebowitz. M. D. Cong. Rec.. 97th Cungress. 2nd session. 16 December 1982. E5253. '5, Lebuwitz, M. D.. Knudsun, R. J.. Robertson. G.. and Burrows. B. Signiticance uf intruindividuul - ~:hangeb in maximum expiratory tlow volume and peak expiraWry lluw mca,urements. Chest 81, 566-57U I INtl_). :b. Lebowitz. M. D.. Hulberg, C. 1.. O'Rourke. M. K., Curman. G.. Dodge. R. R. Gas stove usage. CO and TSP, and respiratory et3'ects, in "Proceedings: Indoor Air Pollutanu-Health ElTeca." Air Polln. Control Assuciatiun, Pittsburgh. 1983. :27. LebuwitY. M. U., Health effects ut' induur pollution. Annnl. Rev. Pnhlic• Hrcdth 4, 2U3-2.1 119tf31. :8. Lebowitz. M. D. The etfects of environmental tobacco smuke exposure and gas stoves on daily peak tlow rates in asthmatic and nun-asthmatic families. Eur. J. Respir. Dis. 65 (Suppl. 133), tyU-ly5 (1984). 29. Martinez. F., and Runchetti. R., in "Proceedings of the National Congress on Pediatric Bron- 6;hupneumunolugy" (U. Castello. Ed.). Italian Ped. Brunchupneumonol. Association. Turin. I'Ili.. 111. ;Nelw, R. J. W.. F(orey, C. du Ve. Altman. D. G.. and Swan. A. V. AssuCtaLlun between gas wuking and rc+pirUtory disease in children. Brit. .t/rd. J. 2. (4y-15_ (1y77) 31. Stclia. R. J. W.. I lurcy. t.du Ve, Morris, (c. W.. Goldstein. B. U.. Clark. D.. and Juhn, H. H. Childhood respiratory IUnesS and the home environment. 1. RtlaUonb between n1trUgen diuxLLle. tcmpcnuure. and relative humidity. Int. J. EpiJemiul. 1/, 155-163 (IyM2). 32. MClla. R. J. W.. Flurey, C. du V.. Ahman. U. G.. and Swan. A. V. The association between respiratory illness dnd tatO,, tCWpttrature anlt relative humidity. /!!t. J. EVidtlAt1Ui. 11, ItY4-l6y I INii21. TI BU 31649

The WHO/EURO Working Group (62) examined passive smoking within the context of all indoor exposures. It was felt that the current levels of knowledge did not yet allow quantitative assessment ot' the impact on the public health, as the knowledge of exposure effect relationships was inadequate, especially fur delayed effects of chronic exposures. ln terms of exposure levels, there were no data on how many cigarettes were being smoked in each of those indoor places where passive smoking can occur, and the ability to estimate how this is distrib- uted within the building stock is marginal. It was felt that the amount of current information about the fraction of the population exposed to levels high enough to be a public health concern is smaller. The amount of information about the specitic pollutants that might be responsible for the effects of concern is also very small. Of the variety of potential adverse affects, it was felt that only for irritation, odor, and airway response is their even a marginal amount of current knowledge. This group also felt that there were many possible pollutant interactions and that there were many other confounding exposures involved in examining the effects of passive smoking. Thus, in terms of relative priurity, effects ut' passive smoking were low in comparison with other concerns, based on knowledge levels. Since, however, passive smoking may indeed be an important public health concern, it should be studied further. Attempts are currently being made by various groups tu ;unsider potential efticacious methods for further studies of the health effects ot' passive smoking (33, 47). REFERENCES I. Anderson. H. R. Respiratory abnurmalitits, smoking habits, and ventilatory capacity in a high- land community in Papua, New Guinea: Prevalence and effect on murtality. lnt. J. Epidrmiul. 8. 127-135 (1979). 2. Binder, R. E.. Mitchell, C. A.. Hubein.. H. R., and Bouhuys. A. Importance of the indoor cnvi- ronmenl in air pollution exposure. Arch. Eirvirurr. Heult/r 31, 277-279 (1976). 3. Burrows. B., Lebowitz, M. D., Barbee. R. A.. Knudson. R. J., and Haiunen. M. tnteractiuns ut smoking and immunologic lucturs in relation to airways obstruction. Chrsr 84, 657-661 (1y83). 4. Cederluf, R.. and Cullry, J. Epidcmiulugical investigations on environmental tobacco omukr. S'cunrl. J. Rrspir. Dis. Srrppl. 91, 47-49 (1974). 5. Chapman, R. S., Habstlblad. V.. Hayes, C. G.. Williams. J. V. R.. and Hummer. D. I. Air pul- lution and childhood ventitatury function. l. E.tipu.ure tu particulate mutter in two southeastern cities. 1971-1972, in "Clinical lmplicatiunb of Air Pollution Resettrch" (A. J. Finkel and W. C. Duel. Eds.), pp. 285-303. Publ. Sci. Group. inc.. Actun. Ytubs., 1976. 6. Cumstuc:k. G., Meyer, M. B.. Hrlsing. K. J., and Tuckman. M. S. Respiratory effects of huuae• huld exposures to tobacco smukc and gub cooking. .irr+er. Rev. Rrspir. Uis. 124, 143-148 (19t31). 7. Dahms. T. E.. Bulin. M. D., and Slavin. R. G, Passive smoking: Effects of bronchial abthma. Chest 80, 53U-534 (1981). 8. Dodge. R. R. The effects of indoor air pollution un Arizona children. Arc•Jt. Envirun. Health 37, 151-155 (1982). 9. Dodge. R. R. The respiratory hcalth and lung function of Anglo-American children in a,melter , town. Anrrr. Ri-s•. Rrapir. Uix. 127; 1"-161 11yt4}). - - 10. Ekwo, E. E., Weinberger, M. M., Lachenbruch. P. A.. and Hunttay. W. H. Relationship u[ pa- rontal smoktng and gas cooking to respiratory disease in children. C/rr3t 84, 6h2-6tN4 (14zsit 11. Florey. C. du Ve, Melia. R. 1. W., Chinn. S.. Guldatein, 13. lJ . liruuks, A. G. F., John. H. II . Br.tigherul, 1. B.. and Webhter, X. The relation between respiratury illness in primary xhuul- TI BU 31648 15. Ha 16. Hil 17. K. i s. K:: ly. Lc :0. Lc =l. L~ 22. Lc _3. Lt 24. L~ 35. L '_6. L 27. L _8. L~ 3:. ti

)dy ~ les nt. Lnd lue er- VaJ : in SYMPOSIUM: MEDICAL PERSPECTIVES ON PaSSLVE SMOKING 659 TAHLE'_ CuKKt•:CTION FACI'OKJ FUR THE SrANUAKU1LA17UN OF l'HE PULMONARY FUNCTtON PAKAMETF.KS Age Height Weight (years) (m) Ikg) CJnst. FVC M -0.028 5.71 -0.011 -3.37 iL) F -0.U38 4.87 -U.1x)3 -2.94 FEF'_5/75 M -0.020 4.30 -u.U14 - 1.7'_ ILS) F -0.043 1.37 -U.U?U 1.27 FEF 75/85 M -0.027 13.85 -0.U" - 12.01 (L''S) F -U.lM/1 6.48 -0.018 -0.36 MEF 25/75 M -0.006 3.51 -0.029 -U.U3 (L'S) F - 0.U?3 1.84 0.031 0.03 each subgroup. The average age varies between 39 and 43 years for the men, and between 29 and 41 years for the women. Since almost all the pulmonary function parameters correlated negatively with age, the need for a standardization was thus underscored. The CS group was most frequently affected by bronchopulmonary prior dis- cabes, but the tlit'ference, were not significant when compared with the other groups. Since on the onc hand, passive inhalatiun ot'tubacco smoke may induce, or at least aggravate, chrunic bronchitis, while active smoking may be the cause ut this disease, the analysis of pulmonary function parameters would be affected by both including or excluding bronchitic subjects. It was, however, found that this had essentially no intlttence on the median values or the regression analysis. For this rea~un. a basic broup, excluding subjects with bronchopulmonary dis- eases such as pneumonia, asthma, and tuberculosis in their history, but including the bronchitics, was used tur further evaluation. TABLE 3 SI'KUCTUKE UF THE GKUUP INTO SUNGKUUPs, NUHHIiK OF SUHJECTS. AVEKAGE AGE. .•1NU I3KONCHOPULMONAKY PKIUK t)15E.isli! Pneumonia. bronchial No. Age aathma. of yubjccts (median tuberc. of Chronic examined valuea) the lungs bronchitis NS 142 66 43 41 21 7 3 2 PS '-51 132 41) 36 30 17 8 3 L•J 301 i9 43 39 42 8 9 3 0S 64 12 41 29 5 2 1 2 CS 1253 141 39 31 28 II 9 6 Su. 941 410 41 35 136 45 30 18

660 KFVTNEIt, rltIE13IG, AND WELTLE followed by PS/H and PS/HW (Fig. 6). For the hta' 25/75 values, all the mulc: PS groups are at about the same level (in the IUU''c range). By contrast, in the female PS group, a gradation is seen, with PS/W showing the smallest negative deviatiun from the predicted value, way of example. parameters among the passive smoker subgroups, a heterogeneous picture is ob- tained. Here, FEF 25/75 (Fig. 6) and FEF 75/85 (Fig. 7) values are considered by If we compare the deviations in the predicted values of the pulmonary function from CS. ES group differs fruut both the PS and CS, while in the women, NS and ES differ parameters (Fig. 5). Nevertheless, here, statistically signiticant differentiation between the individual subgroups is possible. Thus, in the case of the men, the The MEF 25/75 values show an even greater range of scatter than the prior both from NS and from PS. This, however, applies only to men. In Fig. 4, the FEF 75/85 values are shown. Here, ES and CS dif'i'c:r signiticantly eter. deviation of the medians from the predicted FEF 25/75 values in both sexes was found in the CS group. However, no statistically signiticant differences were Neen between the groups, which revealed a considerable scatter range for this param- The comparison of FEF 25/75 (Fig. 3) values shows results similar to thuw seen for FVC. as do the results of FEVi, not shown here. The greatest negative the three groups of nonsmokers. In the case of FVC (Fig. ?), the median values for men are similar in all tivc subgroups. whilt, in women, only that of the CS group differs significantly t'rom Pulmutzury Frrrtc•tiun Purunteters in a Cunrpcu•ison uftlte Snb,qrorrps 3.30 120 ua eo 70 i 1 PAEDICTED 90 II F FVC rS PS ES OS CS n F ~ Fia. 2. irledians and 66 n cuntidtncc levch u[ the percentage Javiauons from prrcdictrd values of FVC in Sex-xparated wbgroup cumparisun. TI BU 31656 Ftc;. FEF == For :5175 PS/ H wume Neith ~pzcit Duse- Put In with cigar: A(t ~'muh 8). .-k a tun Fu

660 KFVTNEIt, rltIE13IG, AND WELTLE followed by PS/H and PS/HW (Fig. 6). For the hta' 25/75 values, all the mulc: PS groups are at about the same level (in the IUU''c range). By contrast, in the female PS group, a gradation is seen, with PS/W showing the smallest negative deviatiun from the predicted value, way of example. parameters among the passive smoker subgroups, a heterogeneous picture is ob- tained. Here, FEF 25/75 (Fig. 6) and FEF 75/85 (Fig. 7) values are considered by If we compare the deviations in the predicted values of the pulmonary function from CS. ES group differs fruut both the PS and CS, while in the women, NS and ES differ parameters (Fig. 5). Nevertheless, here, statistically signiticant differentiation between the individual subgroups is possible. Thus, in the case of the men, the The MEF 25/75 values show an even greater range of scatter than the prior both from NS and from PS. This, however, applies only to men. In Fig. 4, the FEF 75/85 values are shown. Here, ES and CS dif'i'c:r signiticantly eter. deviation of the medians from the predicted FEF 25/75 values in both sexes was found in the CS group. However, no statistically signiticant differences were Neen between the groups, which revealed a considerable scatter range for this param- The comparison of FEF 25/75 (Fig. 3) values shows results similar to thuw seen for FVC. as do the results of FEVi, not shown here. The greatest negative the three groups of nonsmokers. In the case of FVC (Fig. ?), the median values for men are similar in all tivc subgroups. whilt, in women, only that of the CS group differs significantly t'rom Pulmutzury Frrrtc•tiun Purunteters in a Cunrpcu•ison uftlte Snb,qrorrps 3.30 120 ua eo 70 i 1 PAEDICTED 90 II F FVC rS PS ES OS CS n F ~ Fia. 2. irledians and 66 n cuntidtncc levch u[ the percentage Javiauons from prrcdictrd values of FVC in Sex-xparated wbgroup cumparisun. TI BU 31656 Ftc;. FEF == For :5175 PS/ H wume Neith ~pzcit Duse- Put In with cigar: A(t ~'muh 8). .-k a tun Fu

~. ~ PKEVLNfIVE MEDICINE 13, 656-66911yM4) The Influence of Passive Smoking on Pulmonary Function- A Study of 1,351 Office Workers' MICHAEL KENTNEti,' GERHARD TRIEBIG, AND DIETER WELTLE lnsriaue uf Occuputionul and Suc•iul Medic•ine and Polyclinic uj'Oc•c•uputiranrd Diseases uf the Univers•ity of Erlanxen-Nurentberx, Sc•lullerstrusse 2S/29, U-1l52U ErfumtKrn. FeJerul Republle• uj Germary Until now it has been difticult to ascertain how much passive inhalation ut tobucco smoke atTects bronchupulmunary function. To answer thts question. an Investigation involving 1,351 white cullar workers wus carried out. lnformatiun abuut active and passive tobacco smoke exposure was ubtained by a standartlized questionnaire. This made it pus.ible tu subdivide the overall group into tive subgroups: Never smukCrs, passive smukers. ex- smukers. current ,mukers, and other smokers. Forced expiratory vital capacity (FVC) and maximal expiratory tlow-volume (MEFV) curves were used fur lung function analysis. From these curves FVC, forced mid-expiratory tlow (FEF I5175), forced entl-expiratury flow (FEF 75/85). and maxtmai mid-expiratory tlow (MEF 35/75) were determined and standardized for .cx. age, hcight, and body weight. Passive smokers evaluated by this method showed essentially no decfease in parameters describing ventilatory function. It is concluded from the dose-antl time-ttfuct rclationships ubtatned in active smokers between the lung function parameters and the duration of tobacco smoke expuaure on the one hand and the daily consumption of cigarettes on the other thai t•.,,,rvo smoking in small doses may have no essential effect on pulmonary function. s• tvn4 A-aenuc Pre~s. Inc. INTRODUCTION One of the numerous questions that have not yet been unequivocally answercd with respect to passive smoking is its effect on bronchdpulmunary function in people with healthy lunbs. A number of studies in children indicate a certain negative influence, particularly in children ut preschool age (11. 30, 33). Other studies, huwever, fail to confirm such an association (5, 18, 27, 29). ltcccnt re- views on this subject can be found in publications by Weiss et u!. (34) and by Lebowitz (19). To date, only a few studies have been published on the effects of passive smoking and pulmonary function in adults (7, 14, 27, 35). These investigatiuns were usually initiated within the framework of research projects not primarily concerned with passive smoking. As a result, certain re,trictiuns must inevitably be made regarding the relevance of these findings. This point will be considered in more detail in the discussion. The investigation reported herein, in contrast, was designed and carried out for the sole purpose ut' confirming or denying any significant effects of passive smoking on the bronchopulmonary system uaing pulmonary function measurement. A t k. with L. ti.:iput ,~~cnL tratwt a rCLI ymuu- and u. the c: t. Th,_ The .. mt:nt: with un th were Ot' i nca u. in co Th trUnt, sy"tc atlju- tivc: (u, expu (b but ciift'c: ' Presented at the Symposium "Medical Perspectives on Passive Smuking," April 9-12. 1984. Vienna. ~\ustria. ' To whom reprint requests should be addressed. u1)91-7435 fS4 h3.00 • Cupynglu [t"41 bv AcaJemw Pre.,. Inc. UI nµnt+ ur repruduUUon in -. wnn rc,er.cJ. 656 TI BU 31652

SYNIIP(1SIUy1: MEDIC.IL PI:RSPEC't'tVES ON PASSIVE SMUK1N(; 661 I five Cru m :husc :atlve ~ Wa1 bt:en tram- antly priur tUUn the litfer aiun , ub- J by evc{ cen, .l ue. 150 -1 140 ~ 130 -i 120 , 110 1 100 90 ~ aoa 70 a 60~ S aREDEC7ED 0 N F FEF 25175 50 J nS PS ES 0S CS n F l:us. 3. Medians and 66 i~ confidcnce levels uf the percentage deviutions t'rum preUicted vatuea Uf FEF _5/75 in Ne+t-sepuracd subgroup cumpurisun. For the FEF 75/85 measurement, the situation is almost the reverse of the FEF ?5175 medians (Fig. 7). Among the men a "ranking" can be observed in that f'~,, H have the best function figures, followed by PS/W and PS/HW. Among the women, the medians are scattered around the 100% range in all three groups. Neither fur FEF'_5/75, nor for FEF 75/85, were any statistically sibniticant group- ,pecitic differences evident. Dose- and Time-E,/'f~ct Relationships between Tobacco Smoke Exposure and Pulmonary Function In the CS group. the individual pulmonary function parameters were correlated with both the duration of tobacco smoke exposure and the daily consumption of cigarettes. The time-effect regressions are briefly aiscussc;td for FEF 75/85. Among men, this parameter deteriorates with increasing duration of tobacco ,muke cxposure. This relationship is statistically significant at the 5~~. level (Fig. 8). Among women, however, no statistically significant change in FEF 75it35 as a function of the duration of exposure to tobacco smoke can be sten (Fig. 9). For parameters FVC. FEF 25/75. and MEF 25/75, tlose- and time-et'fect re- TI BU 31657

658 KENTNER, TRIEH[G, AND WELTLE TABLE I THE PAftAMETERS EMPLUY6D t'UK PULStUNAIdY FUNLTIUN ANALYSIS Method Forced vital capactty curve Maximul expiratory tlow-volume curve Pulmunury function purameter :\bbrcviatiun Forced expiratury vital capacity Forced mitl-expirutory flow FVC FEF 25/75 Forced end-expirutury flow FEF 75/85 Maximal mid-expiratory flow MEF '_5/75 multiple regression analyses. and taking into account sex, agc;, height, and body weight. A comparison of the reference values thus obtained, with standard values established by other authors (4, 17, 23, 26) shows a good level of agreement. According to these tigures, the standard values decrease with increasing age and weight, while an increase in height results in an elevation in the predicted value (Table 2). All the individual values actually measured were converted into a per- centage deviation from the predicted standard value. The differences in the pulmonary function parameters standardized in this way were statistically checked in a subgroup comparison, using a nunparametric tcst procedure. This procedure invulveti•a comparison ut'several independent samples as described by Kruskal and Wallis (36). A signiticance level of P s 0.05 was chosen. RESULTS Group Structure Table 3 shows the size, average age, and prevalence of pulmonary disease in RhUt.G .wNPLE Fto. I. Subgruups in accordance with tubaccu Smuke expunure criteria. CURHtt FF M: each sut between paramet thus unt The L ca»s, r gruupb. or at lc: uf thia ~ by both th+s haL. Fur thi- cabes „ the bro

658 KENTNER, TRIEH[G, AND WELTLE TABLE I THE PAftAMETERS EMPLUY6D t'UK PULStUNAIdY FUNLTIUN ANALYSIS Method Forced vital capactty curve Maximul expiratory tlow-volume curve Pulmunury function purameter :\bbrcviatiun Forced expiratury vital capacity Forced mitl-expirutory flow FVC FEF 25/75 Forced end-expirutury flow FEF 75/85 Maximal mid-expiratory flow MEF '_5/75 multiple regression analyses. and taking into account sex, agc;, height, and body weight. A comparison of the reference values thus obtained, with standard values established by other authors (4, 17, 23, 26) shows a good level of agreement. According to these tigures, the standard values decrease with increasing age and weight, while an increase in height results in an elevation in the predicted value (Table 2). All the individual values actually measured were converted into a per- centage deviation from the predicted standard value. The differences in the pulmonary function parameters standardized in this way were statistically checked in a subgroup comparison, using a nunparametric tcst procedure. This procedure invulveti•a comparison ut'several independent samples as described by Kruskal and Wallis (36). A signiticance level of P s 0.05 was chosen. RESULTS Group Structure Table 3 shows the size, average age, and prevalence of pulmonary disease in RhUt.G .wNPLE Fto. I. Subgruups in accordance with tubaccu Smuke expunure criteria. CURHtt FF M: each sut between paramet thus unt The L ca»s, r gruupb. or at lc: uf thia ~ by both th+s haL. Fur thi- cabes „ the bro

662 KENTNER, rRIEBIG, AND WELTLE FEF 73l83 100 90 : e0 ; 704 60 J FEF 75Ri5 in scx-sepuntted ,ubgruup cumparuun. 0 M F r I NS PS ES 0S CS t FiG. 4. Medians and hn '~ ;omtiJence levet+ ol the percentuge Jeviatiuns frum prcdicteU valueb of r! F lationships with only low correlation coefficients. not statistically ,iSniticant. were observed. DISCUSSION Inflammation af the airways is the starting point of all chronic noncarcinogenic bronchupulmonary effects of tobacco smoke inhalation (3l). Although the pru- cehses which bring about functionally significant changes following the inhalation of tobacco smoke are not yet fully known, it may be assumeti that they represent multifacturial processes, for which the following pathumechani,m, are under tiib- cussion: (a) Enlargement of the mucous gland mass with resulting hypersecretion (9). tb) Impairment of the function ot'the respiratory ciliated epithelium, with build- up of secretion (3). (c) Increase in bronchial muscles ('_5), due to retlectory bronchoconstrictiun or direct bronchospasmic effects (2). (d) Release of proteolytic enzymes from alveolar macrophages. with consecu- tive formation of a centro-acinar pulmonary emphysema aftc:r years uF smuk- ing (9). (e) "Displacement" of the surfactant factor (25). with pus!,ible promotion ut ateiecta.sis. TI BU 31658 Flu. i Ltt!-" __ The bron;h invebti_ Nive ex in the the am, the pn di5c:1~,_ The thint_,. ut vit:: or tlu% larbet'

SYNIIP(1SIUy1: MEDIC.IL PI:RSPEC't'tVES ON PASSIVE SMUK1N(; 661 I five Cru m :husc :atlve ~ Wa1 bt:en tram- antly priur tUUn the litfer aiun , ub- J by evc{ cen, .l ue. 150 -1 140 ~ 130 -i 120 , 110 1 100 90 ~ aoa 70 a 60~ S aREDEC7ED 0 N F FEF 25175 50 J nS PS ES 0S CS n F l:us. 3. Medians and 66 i~ confidcnce levels uf the percentage deviutions t'rum preUicted vatuea Uf FEF _5/75 in Ne+t-sepuracd subgroup cumpurisun. For the FEF 75/85 measurement, the situation is almost the reverse of the FEF ?5175 medians (Fig. 7). Among the men a "ranking" can be observed in that f'~,, H have the best function figures, followed by PS/W and PS/HW. Among the women, the medians are scattered around the 100% range in all three groups. Neither fur FEF'_5/75, nor for FEF 75/85, were any statistically sibniticant group- ,pecitic differences evident. Dose- and Time-E,/'f~ct Relationships between Tobacco Smoke Exposure and Pulmonary Function In the CS group. the individual pulmonary function parameters were correlated with both the duration of tobacco smoke exposure and the daily consumption of cigarettes. The time-effect regressions are briefly aiscussc;td for FEF 75/85. Among men, this parameter deteriorates with increasing duration of tobacco ,muke cxposure. This relationship is statistically significant at the 5~~. level (Fig. 8). Among women, however, no statistically significant change in FEF 75it35 as a function of the duration of exposure to tobacco smoke can be sten (Fig. 9). For parameters FVC. FEF 25/75. and MEF 25/75, tlose- and time-et'fect re- TI BU 31657

662 KENTNER, rRIEBIG, AND WELTLE FEF 73l83 100 90 : e0 ; 704 60 J FEF 75Ri5 in scx-sepuntted ,ubgruup cumparuun. 0 M F r I NS PS ES 0S CS t FiG. 4. Medians and hn '~ ;omtiJence levet+ ol the percentuge Jeviatiuns frum prcdicteU valueb of r! F lationships with only low correlation coefficients. not statistically ,iSniticant. were observed. DISCUSSION Inflammation af the airways is the starting point of all chronic noncarcinogenic bronchupulmonary effects of tobacco smoke inhalation (3l). Although the pru- cehses which bring about functionally significant changes following the inhalation of tobacco smoke are not yet fully known, it may be assumeti that they represent multifacturial processes, for which the following pathumechani,m, are under tiib- cussion: (a) Enlargement of the mucous gland mass with resulting hypersecretion (9). tb) Impairment of the function ot'the respiratory ciliated epithelium, with build- up of secretion (3). (c) Increase in bronchial muscles ('_5), due to retlectory bronchoconstrictiun or direct bronchospasmic effects (2). (d) Release of proteolytic enzymes from alveolar macrophages. with consecu- tive formation of a centro-acinar pulmonary emphysema aftc:r years uF smuk- ing (9). (e) "Displacement" of the surfactant factor (25). with pus!,ible promotion ut ateiecta.sis. TI BU 31658 Flu. i Ltt!-" __ The bron;h invebti_ Nive ex in the the am, the pn di5c:1~,_ The thint_,. ut vit:: or tlu% larbet'

666 KENTNEIt, TRIEBIG, AND WGLCLE 10 20 30 40 50 TOBACCO SMOKE IXP. ( YP.ARS I Fla. Zi. Regres.iun between FE 1- 75t85 and tubuccu smuke expuhure in yeur5 in male current smukers (r = U._U3. P = U.U13). imum of tive cigarette equivalents (depending upon the smoke constituents) may be assumed. In addition. thu ,ariuus types uC inhalation on the part of active and passive smokers must be utken into account. While active smokers inhale pri- marll} mainstream smoke constituents via mouth breathing, passive smokers in- hale largely sidestream smoke constituents which enter the bronchial system via the nasopharynx. lf additional con,idCration is given to the tiltration function ut nasal breathing, it may be stated that in the case of passivC smoking une is dealing with the effects of a relatively low duse. Since, despite optimized methodulogy, clear and unequivocal dose- and time-effect relationships between daily cigarette consumption or smoking duration in year5 and pulmonary tunction have proved M . • 1ao = 200 : 160 : a k 140 m 120 ~ h 100 Eb 90 • t; • . • 60 40 20 I 0 i0 20 30 40 50 T08ACC0 SMOKE IXP. tYEARSI Fio. 9. Regre,Niun between FEF 7S!rii and tubaccu smoke expusure in ye:u, in lemate currew smukern Ir = 0.042, P = ll.(r111. TI BU 31662 KL W. C K.

666 KENTNEIt, TRIEBIG, AND WGLCLE 10 20 30 40 50 TOBACCO SMOKE IXP. ( YP.ARS I Fla. Zi. Regres.iun between FE 1- 75t85 and tubuccu smuke expuhure in yeur5 in male current smukers (r = U._U3. P = U.U13). imum of tive cigarette equivalents (depending upon the smoke constituents) may be assumed. In addition. thu ,ariuus types uC inhalation on the part of active and passive smokers must be utken into account. While active smokers inhale pri- marll} mainstream smoke constituents via mouth breathing, passive smokers in- hale largely sidestream smoke constituents which enter the bronchial system via the nasopharynx. lf additional con,idCration is given to the tiltration function ut nasal breathing, it may be stated that in the case of passivC smoking une is dealing with the effects of a relatively low duse. Since, despite optimized methodulogy, clear and unequivocal dose- and time-effect relationships between daily cigarette consumption or smoking duration in year5 and pulmonary tunction have proved M . • 1ao = 200 : 160 : a k 140 m 120 ~ h 100 Eb 90 • t; • . • 60 40 20 I 0 i0 20 30 40 50 T08ACC0 SMOKE IXP. tYEARSI Fio. 9. Regre,Niun between FEF 7S!rii and tubaccu smoke expusure in ye:u, in lemate currew smukern Ir = 0.042, P = ll.(r111. TI BU 31662 KL W. C K.

)dy ~ les nt. Lnd lue er- VaJ : in SYMPOSIUM: MEDICAL PERSPECTIVES ON PaSSLVE SMOKING 659 TAHLE'_ CuKKt•:CTION FACI'OKJ FUR THE SrANUAKU1LA17UN OF l'HE PULMONARY FUNCTtON PAKAMETF.KS Age Height Weight (years) (m) Ikg) CJnst. FVC M -0.028 5.71 -0.011 -3.37 iL) F -0.U38 4.87 -U.1x)3 -2.94 FEF'_5/75 M -0.020 4.30 -u.U14 - 1.7'_ ILS) F -0.043 1.37 -U.U?U 1.27 FEF 75/85 M -0.027 13.85 -0.U" - 12.01 (L''S) F -U.lM/1 6.48 -0.018 -0.36 MEF 25/75 M -0.006 3.51 -0.029 -U.U3 (L'S) F - 0.U?3 1.84 0.031 0.03 each subgroup. The average age varies between 39 and 43 years for the men, and between 29 and 41 years for the women. Since almost all the pulmonary function parameters correlated negatively with age, the need for a standardization was thus underscored. The CS group was most frequently affected by bronchopulmonary prior dis- cabes, but the tlit'ference, were not significant when compared with the other groups. Since on the onc hand, passive inhalatiun ot'tubacco smoke may induce, or at least aggravate, chrunic bronchitis, while active smoking may be the cause ut this disease, the analysis of pulmonary function parameters would be affected by both including or excluding bronchitic subjects. It was, however, found that this had essentially no intlttence on the median values or the regression analysis. For this rea~un. a basic broup, excluding subjects with bronchopulmonary dis- eases such as pneumonia, asthma, and tuberculosis in their history, but including the bronchitics, was used tur further evaluation. TABLE 3 SI'KUCTUKE UF THE GKUUP INTO SUNGKUUPs, NUHHIiK OF SUHJECTS. AVEKAGE AGE. .•1NU I3KONCHOPULMONAKY PKIUK t)15E.isli! Pneumonia. bronchial No. Age aathma. of yubjccts (median tuberc. of Chronic examined valuea) the lungs bronchitis NS 142 66 43 41 21 7 3 2 PS '-51 132 41) 36 30 17 8 3 L•J 301 i9 43 39 42 8 9 3 0S 64 12 41 29 5 2 1 2 CS 1253 141 39 31 28 II 9 6 Su. 941 410 41 35 136 45 30 18

~. ~ PKEVLNfIVE MEDICINE 13, 656-66911yM4) The Influence of Passive Smoking on Pulmonary Function- A Study of 1,351 Office Workers' MICHAEL KENTNEti,' GERHARD TRIEBIG, AND DIETER WELTLE lnsriaue uf Occuputionul and Suc•iul Medic•ine and Polyclinic uj'Oc•c•uputiranrd Diseases uf the Univers•ity of Erlanxen-Nurentberx, Sc•lullerstrusse 2S/29, U-1l52U ErfumtKrn. FeJerul Republle• uj Germary Until now it has been difticult to ascertain how much passive inhalation ut tobucco smoke atTects bronchupulmunary function. To answer thts question. an Investigation involving 1,351 white cullar workers wus carried out. lnformatiun abuut active and passive tobacco smoke exposure was ubtained by a standartlized questionnaire. This made it pus.ible tu subdivide the overall group into tive subgroups: Never smukCrs, passive smukers. ex- smukers. current ,mukers, and other smokers. Forced expiratory vital capacity (FVC) and maximal expiratory tlow-volume (MEFV) curves were used fur lung function analysis. From these curves FVC, forced mid-expiratory tlow (FEF I5175), forced entl-expiratury flow (FEF 75/85). and maxtmai mid-expiratory tlow (MEF 35/75) were determined and standardized for .cx. age, hcight, and body weight. Passive smokers evaluated by this method showed essentially no decfease in parameters describing ventilatory function. It is concluded from the dose-antl time-ttfuct rclationships ubtatned in active smokers between the lung function parameters and the duration of tobacco smoke expuaure on the one hand and the daily consumption of cigarettes on the other thai t•.,,,rvo smoking in small doses may have no essential effect on pulmonary function. s• tvn4 A-aenuc Pre~s. Inc. INTRODUCTION One of the numerous questions that have not yet been unequivocally answercd with respect to passive smoking is its effect on bronchdpulmunary function in people with healthy lunbs. A number of studies in children indicate a certain negative influence, particularly in children ut preschool age (11. 30, 33). Other studies, huwever, fail to confirm such an association (5, 18, 27, 29). ltcccnt re- views on this subject can be found in publications by Weiss et u!. (34) and by Lebowitz (19). To date, only a few studies have been published on the effects of passive smoking and pulmonary function in adults (7, 14, 27, 35). These investigatiuns were usually initiated within the framework of research projects not primarily concerned with passive smoking. As a result, certain re,trictiuns must inevitably be made regarding the relevance of these findings. This point will be considered in more detail in the discussion. The investigation reported herein, in contrast, was designed and carried out for the sole purpose ut' confirming or denying any significant effects of passive smoking on the bronchopulmonary system uaing pulmonary function measurement. A t k. with L. ti.:iput ,~~cnL tratwt a rCLI ymuu- and u. the c: t. Th,_ The .. mt:nt: with un th were Ot' i nca u. in co Th trUnt, sy"tc atlju- tivc: (u, expu (b but ciift'c: ' Presented at the Symposium "Medical Perspectives on Passive Smuking," April 9-12. 1984. Vienna. ~\ustria. ' To whom reprint requests should be addressed. u1)91-7435 fS4 h3.00 • Cupynglu [t"41 bv AcaJemw Pre.,. Inc. UI nµnt+ ur repruduUUon in -. wnn rc,er.cJ. 656 TI BU 31652

SYMt'USIUbt: `IEDICAL f'ERSI'EC'ffVES ON PASSIVE SMOKING 665 150 140 1 130 1 120 ~ 110 -~ 100 90 ~ 80 1 70 ~ 69 ~ 6a rways a FEF uaetu! tteEtty. tU air- rnetric iticant rwayi I = ?Si _SCntti L large: ,asun, inter- than bia.S. .1N al- " uIts. tzd in onary ~tu ull L PREDICfED FEf 7S1Ei M F PS/H P$/W P5/HM P$/N P$/N PS/MM n• 39 146 36 18 61 36 Fia. 7. %ttdians and 66 fi confidence laveln of the percentuge deviutiuns from prctticteJ valuCb uf hEF 75iSj in the hex-~,aparated cumpari,un, Cur varwua types ut pussive yrnuke expuwre. categories of nonsmokers and from nonactive smokers. In this conneetion, FEF 75/235 and MEF 25/75 revealed the greatest Jiscriminatiun. Even time-and duse- e/'t'ect relationships in the CS group between tobacco smoke exposure and number ut ~ibarettes smoked per day on the one hand. and pulmonary function on the uther, revealed only a fvw statistically significant correlations which, huwever, revealed inconsistencies in the sex-separated comparison. This result is possibly due to the relatively low average age of the members of the groups investigatc;d, particularly among women. The precision of the measuring methodology and curve evaluation may be considered good. This was demonstrated by a parallel determination of the tlow volume curve tur all subjects with the aid of a;ewncl electronic 5pirumeter which determined the measured values automatically. It may, therefure. Ue ussumed that the heterogeneity and inconsistency of the pulmonary tuncttuu Pattern results ntainly from the large interindividual variability of the measured data. Thus, ap- parent correlations may also he simulated. On the basis of the calculatiuns ut' variuus authors ( 1'_. 13, 15, 3'_), tor passive ~mukers in general. a daily cxpubwc of the order of magnitude ut' ll ' to a max- TI BU 31661

SYS1NOSIUM: .14EUIC:\L PERSPECTIVES ON PASSIVE SMOKING NEf 25175 120 ~ 110 ~ taa 9oa ao ~ ,.a lues ot 60 50 ~ tcant. u0 9S PS ES OS CS Y F i. 1 663 ,genllti } Ftl/. ?. Medians and 66 '1 contdence Ie«hi UI the ptlrcentllge <levllltlun5 from predicted v:llUe~~ ut : pru- latiun. 'escnt :r dis- t y). build- un or ,zcu- .muk- M1=F '-5/75 in the tex•.epurated ,ubgruup cumpuribun. The intlammation of the smaA airways represents the initial reaction of the brunchopulmonary system to the active inhalation of tobacco smoke (9. 24). This investigation is primarily concerned with clarifying the question of whether pas- ,ive expuNure to tobacco smoke results in tunction-analytical signilicant changes trt the small airways. There is a close correlation between structural changes of the small airways and the associated effects on pulmonary function (8), although the prognostic value of these lesions with respect to severe obstructive airway di,ea,es has not yet been established 16, '_b). 'I"he function-analytic correlate of a,mall airways dysfunction is, among other things, the (low limitation in the largely elturt-indepentient. enal-c:xpiratury purt ul vital capacity ('_U. 21, 28). This results in characteristic deformations of FVC ur flow volume curves, which are relatively easy to obtain in an investigation of larger groups of subjects. Although doubts have been expressed ( 1. 10) on the TI BU 31659

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 653 lha Jge . ~~ tur nu ces :ib- k:nt lgh the :ry -)n, ge. h~tt ~ts tn~ it tps .:ts nvi- ~ uf ~3). ke. wl- ern C. IJe- I 4a n4- 37, ter ,:htldren and the use of g,u fur cooking. lll. Nitrogen diuxide, respiratory illnebs and lung lunctiun. Int. J. E'pidc•tniul. 8, 347-353 (1979). t_. Frtcdman, G. D., Petitti, D. B.. and Bawol, R. D. Prevalence and correlates uf passive -smuking. rtmrr. J. Piiblic Health 73, 401-dt15 (1983). 0. Gunmaker, S. L.. Walker. D. K.. Jacobs, F. H.. and Ruch-Ruas, H. Parental smoking and the risk of childhood asthma. Arnrr. J. Public Health 72, 574-579 (1982). Id. H•rmmer. t). 1.. Miller. F J., Stead. A. G., and Hayes, C. G. Air pollution and childhuod venti- Iatury function. 1. Exposure to particulates matter in two southeastern cities, 1971-1972, in "Clinical Implications of Air Pollution Research" (A. 1. Finkel and W. C. Duel. Eds). pp. 321- 337. Publishing Sciences Group, Acton, Mass., 1976. 15. H:uxlblad, V.. Humble. C. G.., Gr.tham, M. G.. and Anderson, H. S. Indoor environmental determinants uf lung function in children. Antrr. Rev. Respir. Dir. 123, 479-485 (1981). lb. Hiller, F. C., McCuskcr. K. T., 4latumber, M. K., Wilson. 1. D., and Bone, R. C. Deposition of ~iJ"uN:un cigarette smoke in the human respiratory tract. Ainer. Rrv. Rerpir. Dis. 125, 4Ub- 4UIS (1982). 17. K.,uffmtlnn, F., and Brille, D. Respiratory function in passive smokers' children, in PROC lEA, . Edinburgh. August 1981. !b. Kauffmann. F., Tessier. J. F.. and Oriul. P. Adult passive smoking in the home environment: A risk factor for chronic aiRlow limitation. atner. J. EpidrnLiuL 11l, _b9-2ML) t 1983). 19. Lebowitz, M. D., Bundheim, P., Cristea, G., Markuwitc. U., %aiaazek, J., Staniec, M., and Van Wyck, U. The effect of air pollution and weather on lung function in exercising children and adolescents. atnrr. Rev. Rrspir. Dfs. 109, 262-273 (19741. :U. Lebuwitz, M. D.. Knudson. R. J., and Burrows. B. The Tucson epidemiology study of chronic obstructive lung disease. 1. Methodology and prevalence uf disease. .4mer. J. Epidetruul. 102, 137-152 (19751. 21. Lebuwitt. M. U., and Burrows. B. Respiratory symptoms related tu smoking habi[s of family adults. Chest 69, 4S-5(1 (1976). ::. Lebowitz, M. I).. Armet, D. B., and Knudson. R. The effect of passive smoking on pulmonary tunctiun in children. Envirorr. Int. 8, 371-374 (I9tl2). _3. Lebowitz. M. D.. O'Rourke. M. K., Dlxlge, R.. Holberg. C. 1., Corman. G., Hushaw, R. W., Pinnas, J. L.. Barbee, R. A., and Sneller. M. R. The adverse health effects of biological aeru- suls, other aerosols, and indoor micruclimute on asthmatics and nun-asthmutics. Envirun. Int. 8, 375-3li0119ii2). 24. Lebowitz. M. D. Cong. Rec.. 97th Cungress. 2nd session. 16 December 1982. E5253. '5, Lebuwitz, M. D.. Knudsun, R. J.. Robertson. G.. and Burrows. B. Signiticance uf intruindividuul - ~:hangeb in maximum expiratory tlow volume and peak expiraWry lluw mca,urements. Chest 81, 566-57U I INtl_). :b. Lebowitz. M. D.. Hulberg, C. 1.. O'Rourke. M. K., Curman. G.. Dodge. R. R. Gas stove usage. CO and TSP, and respiratory et3'ects, in "Proceedings: Indoor Air Pollutanu-Health ElTeca." Air Polln. Control Assuciatiun, Pittsburgh. 1983. :27. LebuwitY. M. U., Health effects ut' induur pollution. Annnl. Rev. Pnhlic• Hrcdth 4, 2U3-2.1 119tf31. :8. Lebowitz. M. D. The etfects of environmental tobacco smuke exposure and gas stoves on daily peak tlow rates in asthmatic and nun-asthmatic families. Eur. J. Respir. Dis. 65 (Suppl. 133), tyU-ly5 (1984). 29. Martinez. F., and Runchetti. R., in "Proceedings of the National Congress on Pediatric Bron- 6;hupneumunolugy" (U. Castello. Ed.). Italian Ped. Brunchupneumonol. Association. Turin. I'Ili.. 111. ;Nelw, R. J. W.. F(orey, C. du Ve. Altman. D. G.. and Swan. A. V. AssuCtaLlun between gas wuking and rc+pirUtory disease in children. Brit. .t/rd. J. 2. (4y-15_ (1y77) 31. Stclia. R. J. W.. I lurcy. t.du Ve, Morris, (c. W.. Goldstein. B. U.. Clark. D.. and Juhn, H. H. Childhood respiratory IUnesS and the home environment. 1. RtlaUonb between n1trUgen diuxLLle. tcmpcnuure. and relative humidity. Int. J. EpiJemiul. 1/, 155-163 (IyM2). 32. MClla. R. J. W.. Flurey, C. du V.. Ahman. U. G.. and Swan. A. V. The association between respiratory illness dnd tatO,, tCWpttrature anlt relative humidity. /!!t. J. EVidtlAt1Ui. 11, ItY4-l6y I INii21. TI BU 31649

SYS1NOSIUM: .14EUIC:\L PERSPECTIVES ON PASSIVE SMOKING NEf 25175 120 ~ 110 ~ taa 9oa ao ~ ,.a lues ot 60 50 ~ tcant. u0 9S PS ES OS CS Y F i. 1 663 ,genllti } Ftl/. ?. Medians and 66 '1 contdence Ie«hi UI the ptlrcentllge <levllltlun5 from predicted v:llUe~~ ut : pru- latiun. 'escnt :r dis- t y). build- un or ,zcu- .muk- M1=F '-5/75 in the tex•.epurated ,ubgruup cumpuribun. The intlammation of the smaA airways represents the initial reaction of the brunchopulmonary system to the active inhalation of tobacco smoke (9. 24). This investigation is primarily concerned with clarifying the question of whether pas- ,ive expuNure to tobacco smoke results in tunction-analytical signilicant changes trt the small airways. There is a close correlation between structural changes of the small airways and the associated effects on pulmonary function (8), although the prognostic value of these lesions with respect to severe obstructive airway di,ea,es has not yet been established 16, '_b). 'I"he function-analytic correlate of a,mall airways dysfunction is, among other things, the (low limitation in the largely elturt-indepentient. enal-c:xpiratury purt ul vital capacity ('_U. 21, 28). This results in characteristic deformations of FVC ur flow volume curves, which are relatively easy to obtain in an investigation of larger groups of subjects. Although doubts have been expressed ( 1. 10) on the TI BU 31659

664 l<ENTNt'tt. TRIEBIG, AND WELTLl: 130 120 u0 100 90 80 70 ~ 60 j T. PREDICTED FEF 25/75 M F PS/H PS/W PS/HW PS/H PS/M PS/HW .- 39 • 146 36 i8 61 36 Fu;. h. MeJians and 66 =it contidence levcls uf the percentage deviatiune frum prcJicted val=, ut FEF =5i75 in the ,ea-~.eparated cumparihun, tur variuua types ut pasolVe -,muke expu,ure. specificity of such pulmonary function analyses with respect to a small airways ciysfunctiun diagnosis, the view that the determination of parameters such as FEF ?5/75. FEF 75185. and MEF ?5/75, in this respect, represc:nt, a,ensible and useful tunctiun-analytie method (6, ??, 28) has largely become accepted. Admittedly. the determination of tlow rates or tluw aifferenc:es in the diagnosis of small air- ways dysfunction makes sense only when normal values fur such spirumeurlc parameters as FVC and FEV, are present ( lU. _?, '_ts), that is. when signitu;ant ventilatory and respiratiun-mechanical disturbances affecting the large airway, can be excluded. This may be assumeJ, t'or the present study. since in all ,ub- groups, independent of sex, the FVC values scatter unly in the range ut a_Yi deviation from the predicted norm. Only the temalc CS group. at -M(., represtnts an exception. Similar remarks alsu apply to the FEV t parameter. It is well known that end-txpirutury flow determinations are subject to a large intraindividual (6) and interindivieluat (16) variation spectrum. For this reason, particular care was taken in this study to eliminate as far as possible, such inter- fering factors as prior bronchupulmonary diseases or inhaled noxae other than tobacco smoke. Furthermore, in order to limit the danger of an evaluation hia,, the study was performed under blind conditions and great importance was at- tachccl to the precise stanJardiZatiun of the pulmonary function measuring results. Nevertheless, it wa, nut possible to dit'ferentiate all t7vz subgroups UrtinCJ in accordance with tubct.:cu smoke exposure criteria, on the ba,is of pulmunury function analysis. It proved possible only to discriminate the CS group trum all TI BU 31660 cut vui du th. ttt. pa sit

33. National Hean Lung and Blood (nst. "Repon ut' a Workshop un ihe Respiratory Effects uf invuluntary Smoking: Epidemiological Stutlies." U.S. Govt. 1'rinting Ottice. Washington. D.C., December 19153. 34. National Research Council (NRC). "Indour Pollutants." National Academy of Sciences. Waah- ingtun. D.C.. 1981. 35. O'Connell. E. J.. and Logan. G. B. Parental smoking in childhood asthma. Ann. Allergy 32, l42- 145 (1y74). 36. Olansky, S. 1. Is smoker/nonsmoker segregation effective in reducing passive inhalation among nonsmokers'! Autrr. J. Public Krulth 72, 737-739 (1982). 37. O'Rourke, M. K. Pollen dispersal and its relationship to respiratory illness. in "Proceedings of the First International Cunference on Aerobiology" (German), pp. 81-88. Symposium Pub- lishers, Munich. 198U. 38. O'Rourke. M. K.. and Lebowitz. M. D. A comparison of regional atmosphere pollen with pollen collected at and near homes. Grrunu 22, 1-lU (1984). 39. Pimm. P. E.. Silverman. F.. and Shephard. R. J. Physiological effects of acute passive exposure to cigarette ,muke. Arc•1t. Environ. Ileu/th 33, 201-213 (19781. 40. Pipes, D. .4. Allergy to tobacco smoke. Ann. Allergy 28, 277-82 (1945). the pulmonary function of children. :tntrr. J. EpiJrntiul. 110. 15-25 (1979). 57. Tager. I. B., Weiss. S. T.. Rosner. B.. and Speizer. F. E. Effect u(purental cigarette smoking un 3- IU (1980). pulmonary function in children associated with NO: exposure. Anter: Rev. Rrspir. Uis. 121. 56. Speizer. F. E.. Ferris, B., Jr., Bishop. Y. M. M., and Spenylcr. J. Respiratory disease rates and EJ.), pp. 343-359. Ann Arbor Press. Ann Arbor. Nlich.. 1980. 55. Speizer. F E.. Ferris. B.. Jr.. Bishop. Y. M. M.. and Spengler. 1. Health effects of indoor NU; exposures: Preliminary results. i/t "Nitrogen Oxides and fhelr EftCcth on Health" IS. F. Lee. 16, 443-446 (19681. 54. Speer. F. Tubaccu and the non-smoker: A study of subjective symptoms. Arc•h. Environ. Nrulth Heulrlt 16, 670-672 (1968). 53. Sufoiuwe. G. O. Smoke pollution in dwellings of infants with bruncho-pneumunia. Arch. Envirun. bronchitis nun-smukers. Czech. Med. 3. 308-310 (1980). 52. Simcs:ek. C. Reflection of passive exposure to smoking in the home on the prevalence of chronic ,ive" cigarette smoke exposure. Envirun. Rrs. 19, 279-291 (1y791. 5 1. Shephard. R. J., Cullings. R., and Silvermun. F. Responses of exercising subjects to acute "pus- smoke. Environ. Rrs. 20. 392-402 (1972). 50. Shephard. R. J.. Collins. R.. and Silverman. F. Passive exposure of asthmatic subjects to cigarette :83 11977). Lung functiun, respiratory disease, and smoking in families. A»trr. J. Epidrutiul. IU6, 274- 49. Schilling, R. S. F.. Letai. A. D., Hui. S. L.. Beck. G. J., Schuenbert. J. B.. and Buuhuys, A. 1038-1043 (1983). The effects uf host facturs and home environmental exposures. Amer. Rrv. Respir. Uis. 128, 48. Schenker. M. B., Samet. J. M., and Speiter, F. E. Risk factors for childhood respiratory disease: a workshop. Eur. J. Respir. Dis. 65 (Suppi. 133), 1-52 (1984). 47. Rylunder, R. (Ed.). Environmental tobacco smuke: effects and exposure levels: Proceedings of Uitiversity of Geneva. 1974. Scund. J. Respir. Dis. (Suppl. yl). (t974). 46. Rylunder, R. Environmental tobacco smoke etlr.t, un the nonsmoker: Report from a workshop. 145, IU25-IU27 (lyri3). 45. Romer. J.. and Hermann. H. Tobaksrygnmgs Betydning for asthma ug Rhinitis. USeskr. Lueg. , mental facturs over time. Envirun. Rrs. 33, 130-143 (1984). 44. Robertson. G.. and Lebuwitz. M. D. Analysis of relatiunshhios between symptoms and envirun- l Wushingt(tn, D. C. ) 208, 464-472 11980). 43. Repace. J. L., and Luwrcy, A. H. Indoor air pollution tobacco smoke, and public health. Science United States. 1976-80. NCHS Advance Dutu 76, 1-24 (1982). 42. Radl'urd. E. P., and Drizd. 1'. A. Blood carbon monoxide levels in persons 3-74 years of age: Notion Nouvelte Preciste. Kruxcltes-Med. 57. 335-340 (1977). 41. Polak, E. Le Papier a cigarette. Son Rule Duns la Pollution de Lieux Habites. Tabagisme Pasbif: TI BU 31650 58. Vale

SYMt'USIUbt: `IEDICAL f'ERSI'EC'ffVES ON PASSIVE SMOKING 665 150 140 1 130 1 120 ~ 110 -~ 100 90 ~ 80 1 70 ~ 69 ~ 6a rways a FEF uaetu! tteEtty. tU air- rnetric iticant rwayi I = ?Si _SCntti L large: ,asun, inter- than bia.S. .1N al- " uIts. tzd in onary ~tu ull L PREDICfED FEf 7S1Ei M F PS/H P$/W P5/HM P$/N P$/N PS/MM n• 39 146 36 18 61 36 Fia. 7. %ttdians and 66 fi confidence laveln of the percentuge deviutiuns from prctticteJ valuCb uf hEF 75iSj in the hex-~,aparated cumpari,un, Cur varwua types ut pussive yrnuke expuwre. categories of nonsmokers and from nonactive smokers. In this conneetion, FEF 75/235 and MEF 25/75 revealed the greatest Jiscriminatiun. Even time-and duse- e/'t'ect relationships in the CS group between tobacco smoke exposure and number ut ~ibarettes smoked per day on the one hand. and pulmonary function on the uther, revealed only a fvw statistically significant correlations which, huwever, revealed inconsistencies in the sex-separated comparison. This result is possibly due to the relatively low average age of the members of the groups investigatc;d, particularly among women. The precision of the measuring methodology and curve evaluation may be considered good. This was demonstrated by a parallel determination of the tlow volume curve tur all subjects with the aid of a;ewncl electronic 5pirumeter which determined the measured values automatically. It may, therefure. Ue ussumed that the heterogeneity and inconsistency of the pulmonary tuncttuu Pattern results ntainly from the large interindividual variability of the measured data. Thus, ap- parent correlations may also he simulated. On the basis of the calculatiuns ut' variuus authors ( 1'_. 13, 15, 3'_), tor passive ~mukers in general. a daily cxpubwc of the order of magnitude ut' ll ' to a max- TI BU 31661

33. National Hean Lung and Blood (nst. "Repon ut' a Workshop un ihe Respiratory Effects uf invuluntary Smoking: Epidemiological Stutlies." U.S. Govt. 1'rinting Ottice. Washington. D.C., December 19153. 34. National Research Council (NRC). "Indour Pollutants." National Academy of Sciences. Waah- ingtun. D.C.. 1981. 35. O'Connell. E. J.. and Logan. G. B. Parental smoking in childhood asthma. Ann. Allergy 32, l42- 145 (1y74). 36. Olansky, S. 1. Is smoker/nonsmoker segregation effective in reducing passive inhalation among nonsmokers'! Autrr. J. Public Krulth 72, 737-739 (1982). 37. O'Rourke, M. K. Pollen dispersal and its relationship to respiratory illness. in "Proceedings of the First International Cunference on Aerobiology" (German), pp. 81-88. Symposium Pub- lishers, Munich. 198U. 38. O'Rourke. M. K.. and Lebowitz. M. D. A comparison of regional atmosphere pollen with pollen collected at and near homes. Grrunu 22, 1-lU (1984). 39. Pimm. P. E.. Silverman. F.. and Shephard. R. J. Physiological effects of acute passive exposure to cigarette ,muke. Arc•1t. Environ. Ileu/th 33, 201-213 (19781. 40. Pipes, D. .4. Allergy to tobacco smoke. Ann. Allergy 28, 277-82 (1945). the pulmonary function of children. :tntrr. J. EpiJrntiul. 110. 15-25 (1979). 57. Tager. I. B., Weiss. S. T.. Rosner. B.. and Speizer. F. E. Effect u(purental cigarette smoking un 3- IU (1980). pulmonary function in children associated with NO: exposure. Anter: Rev. Rrspir. Uis. 121. 56. Speizer. F. E.. Ferris, B., Jr., Bishop. Y. M. M., and Spenylcr. J. Respiratory disease rates and EJ.), pp. 343-359. Ann Arbor Press. Ann Arbor. Nlich.. 1980. 55. Speizer. F E.. Ferris. B.. Jr.. Bishop. Y. M. M.. and Spengler. 1. Health effects of indoor NU; exposures: Preliminary results. i/t "Nitrogen Oxides and fhelr EftCcth on Health" IS. F. Lee. 16, 443-446 (19681. 54. Speer. F. Tubaccu and the non-smoker: A study of subjective symptoms. Arc•h. Environ. Nrulth Heulrlt 16, 670-672 (1968). 53. Sufoiuwe. G. O. Smoke pollution in dwellings of infants with bruncho-pneumunia. Arch. Envirun. bronchitis nun-smukers. Czech. Med. 3. 308-310 (1980). 52. Simcs:ek. C. Reflection of passive exposure to smoking in the home on the prevalence of chronic ,ive" cigarette smoke exposure. Envirun. Rrs. 19, 279-291 (1y791. 5 1. Shephard. R. J., Cullings. R., and Silvermun. F. Responses of exercising subjects to acute "pus- smoke. Environ. Rrs. 20. 392-402 (1972). 50. Shephard. R. J.. Collins. R.. and Silverman. F. Passive exposure of asthmatic subjects to cigarette :83 11977). Lung functiun, respiratory disease, and smoking in families. A»trr. J. Epidrutiul. IU6, 274- 49. Schilling, R. S. F.. Letai. A. D., Hui. S. L.. Beck. G. J., Schuenbert. J. B.. and Buuhuys, A. 1038-1043 (1983). The effects uf host facturs and home environmental exposures. Amer. Rrv. Respir. Uis. 128, 48. Schenker. M. B., Samet. J. M., and Speiter, F. E. Risk factors for childhood respiratory disease: a workshop. Eur. J. Respir. Dis. 65 (Suppi. 133), 1-52 (1984). 47. Rylunder, R. (Ed.). Environmental tobacco smuke: effects and exposure levels: Proceedings of Uitiversity of Geneva. 1974. Scund. J. Respir. Dis. (Suppl. yl). (t974). 46. Rylunder, R. Environmental tobacco smoke etlr.t, un the nonsmoker: Report from a workshop. 145, IU25-IU27 (lyri3). 45. Romer. J.. and Hermann. H. Tobaksrygnmgs Betydning for asthma ug Rhinitis. USeskr. Lueg. , mental facturs over time. Envirun. Rrs. 33, 130-143 (1984). 44. Robertson. G.. and Lebuwitz. M. D. Analysis of relatiunshhios between symptoms and envirun- l Wushingt(tn, D. C. ) 208, 464-472 11980). 43. Repace. J. L., and Luwrcy, A. H. Indoor air pollution tobacco smoke, and public health. Science United States. 1976-80. NCHS Advance Dutu 76, 1-24 (1982). 42. Radl'urd. E. P., and Drizd. 1'. A. Blood carbon monoxide levels in persons 3-74 years of age: Notion Nouvelte Preciste. Kruxcltes-Med. 57. 335-340 (1977). 41. Polak, E. Le Papier a cigarette. Son Rule Duns la Pollution de Lieux Habites. Tabagisme Pasbif: TI BU 31650 58. Vale

664 l<ENTNt'tt. TRIEBIG, AND WELTLl: 130 120 u0 100 90 80 70 ~ 60 j T. PREDICTED FEF 25/75 M F PS/H PS/W PS/HW PS/H PS/M PS/HW .- 39 • 146 36 i8 61 36 Fu;. h. MeJians and 66 =it contidence levcls uf the percentage deviatiune frum prcJicted val=, ut FEF =5i75 in the ,ea-~.eparated cumparihun, tur variuua types ut pasolVe -,muke expu,ure. specificity of such pulmonary function analyses with respect to a small airways ciysfunctiun diagnosis, the view that the determination of parameters such as FEF ?5/75. FEF 75185. and MEF ?5/75, in this respect, represc:nt, a,ensible and useful tunctiun-analytie method (6, ??, 28) has largely become accepted. Admittedly. the determination of tlow rates or tluw aifferenc:es in the diagnosis of small air- ways dysfunction makes sense only when normal values fur such spirumeurlc parameters as FVC and FEV, are present ( lU. _?, '_ts), that is. when signitu;ant ventilatory and respiratiun-mechanical disturbances affecting the large airway, can be excluded. This may be assumeJ, t'or the present study. since in all ,ub- groups, independent of sex, the FVC values scatter unly in the range ut a_Yi deviation from the predicted norm. Only the temalc CS group. at -M(., represtnts an exception. Similar remarks alsu apply to the FEV t parameter. It is well known that end-txpirutury flow determinations are subject to a large intraindividual (6) and interindivieluat (16) variation spectrum. For this reason, particular care was taken in this study to eliminate as far as possible, such inter- fering factors as prior bronchupulmonary diseases or inhaled noxae other than tobacco smoke. Furthermore, in order to limit the danger of an evaluation hia,, the study was performed under blind conditions and great importance was at- tachccl to the precise stanJardiZatiun of the pulmonary function measuring results. Nevertheless, it wa, nut possible to dit'ferentiate all t7vz subgroups UrtinCJ in accordance with tubct.:cu smoke exposure criteria, on the ba,is of pulmunury function analysis. It proved possible only to discriminate the CS group trum all TI BU 31660 cut vui du th. ttt. pa sit

SYytPUSIUM: %,tEDICAL PERSPECTIVES ON PASSIVE SMOK1NG 671 damaging effect ot' nitrogen dioxide derived from tobacco smoke and air pollution t_1). Such an effect of nitrogen dioxide is known from the literature ( I4), ln ,c:verai studies involving more than 400 subjects we have investigated the urinary hydroxyproline excretion in cigarette smokers, pipe and/or cigar smokers, and nonsmokers. These studies were originally designed to measure the nicotine uptake in relation to smoking habit. However, they also allowed the examination uf urinary hydruxypruline excretion in relation to the subject's smoking habit or exposure to anibient tobacco smoke. SUBJECTS AND METHODS .5uhjet•ts Male cigarette ,mokers, pipe and/or cigar smokers. and nonsmokers were re- cruiteal by newspaper advertisements, by posters in shups, facturies, public build- ings, and universities as well as by word uF muuth. All subjects tilled in a detailed questionnaire related to lifestyle, occupational status, and psychological traits. Srrrdy 1. In the tirst tield study. which lasted from November 1980 to March 1981, _200 cigarette smokers were recruitcd. Three blood samples were taken from each subject on 3 different days over a period of 3 weeks. The blood samples were taken in our laboratory between 4 and 7 PM. Urine was collected from murning until the time of blood sampling, then continued until the following nturning. The samples were categorized as "day" or "night" urine, respectively. Data for day and currespunding consecutive night urines were evaluated as 24- hr urines. To an aliquot of each urine sample U.11/c (v/v) toluene was added. tiamplCS were stored at -?U°C until December 1983, when the hydruxypruline ;wd creatinine analyses were performed. Results t'ur day and night urines were cvaluatecl separately. Since not all six urine samples for each subject were avail- uhle at the time uf hydroxyproline determination,. it was decided to include the tir,t day or night urine available from each subject for evaluation. Due to phy,- iulu5ically raised values in adolescents ( lU) and, perhaps in sume.casCs, pathu- lugtcally elevated values in elderly pipe and cigar smukers ( 13), it was also decided tu base our evaluation only on subjects betweCn ages 21 and 60. Thus 116 day urtnes. 120 night w•ines, and 88 24-hr urines from cigarette smokers could be c~:iluatecl. St«cl~, 2. In asecund study, from May to July 1981, 199 pipe and/or cigar Snwkers were recruited. From each subject a blood sample was takzn between 4 and 7 P%t. Before blood sampling, 24-hr urine was collected. Due t.) :Lvailability ut' urine samples at the time of hydroxyproline and creatinine analyses and after cxclu,iuna for age, 154 urine samples from pipe/cigar smokers could be evaluated. .1rud r 3. In a third -study, from March to July 1983, '24 nonsmokers were re- ~ruwtcJ. From each subject a blood ,ample was taken between 4 anJ 7 Ptit. Urine Na, Lullectcd over a period of 24 hr before blood sampling. This group was irucnded, to be the c:untrul group tur thc smokcrs: One subject had to be excluded due to age younger than 21. lending 23 urine samples from nonsmokers for tval- uwuun. TI BU 31667

It SYMPO!sIUti1: 'Ot.l)IL'AL PERSPECTIVES ON PASSIVE SMUKING 667 impossible to establish or have been established only in an inconsistent form (as in the ca,c of CS), this is all the more likely to be the case for the low-dose ,ituatiun int:t with in passive smoking. As mentioned at the beginning of this article, only a few stutlies have so far cunsideretl the question of possible impairment of pulmonary function in adult passive smokers. These studies all have the disadvantage, owing to the nature of the data collectiun. of not being able to form groups fur all possible factors of active and passive tobacco smoke exposure. This means that there is a danger of ,clectiun bias involving the nonconsideration of possible positive or negative results. Since, however, a similar pulmonary function methodology was em- pluyed, comparisons between our results and those reported in other studies are pussible. In common with us. Schilling t•t al. ('_7) and Comstock et irl. (7) were unable tu establish any statistically significant reduction uf pulmunary function in passive ,mukers (Table 4). In cunlrast. Kauffmann et ul. (14) t'ountl tlittcrences between TAl3LE 4 IZI.vI1:W UF 'IHF. LIIERArUKIi ON AKIlCL1iS SU FAR PUBLISHGU ON THN: Topic PULMUNAKY FUNCTIUN IN PASJIVts SMUOKIditS Pulmonary Authors Subgroups" function parameter, Results" Schilling NSiH N = 138 FVC. FEVi No statistically significant rt ul.. 1977 PS/H 114 PEF. %tEF 50 reduction of lung function CS 78 14EN 25 for PS/H. White and NS N = 400 FVC. FEVi Concerning FEF 25/75 and Frutb. Iyi30 PS/W 400 FEF 25/75 FEF 75/ts5: PS/W show ~ CS 1300 statistically significantly lower values than NS. Values in PS1W do nut differ essentially from values in CS smoking I- l0 cigarettes per day and in CS without inhalation. Cumstuck PS/ H N= 539 FEV, No stutistically significant ES/H 232 FEVI/FVC reduction of lung function CS 953 fur PS/H. K.wtfmann NS/H N= 1675 FVC. FEVI ' Concerning FEF '5/75 in PS/H 1'?3 F EF 25/75 femalCs aged atl years or more: statistically signiticant differences between NS anJ PS/H. Duse-tffect relationship between reductiun uf lung function in PS/H and daily tobacco consumption ul the hunbanJ. - See text for abbreviations. TI BU 31663 ,

672 ADLKC)Ft K. SCttEkER. AND tiELLEl2 rtnul vticul .til etitocls Hvclru.c.vprulhw anul creatinine in urine. Hydroxyproline and creatinine were analyzed3 with an autuanalyzer (Autoanalyzer Type 11. Technicon Co., Ltd.) ac- curding to the methods used by Husley et al. (9) and Technicon Clinical Methotls No. SEZ-O01 I FC4 ( l7), respectively. Nicutine in arine and cotinine in seruin. Nicotine in urine and cotinine in serum were determined by gas-liquid chromatography based on the method tiescribCd by Hengen and Hengen (8). Scrum and urine samples were stored at -?U°C. All analyses were pertbrmed within 6 months after sampling. Curbuxyhelnuylubin (COHb). COHb levels were measured with a CO-Oxitncter (lnstrumentatiun Laboratories Ltd., Model 182) immetliately after taking the blood samples. Statistical e rulnutiun. Differences between subgroups were tested for signiti- cance using, as appropriate, analysis of variance, the t test, the Wilcuxtm test. or the test for linearity. RESULTS Hydru.eypruline/Creutinine Ratio For cigarette smokers, a small, but statistically significant correlation between the hydroxyproline/creatinine ratio and smoke uptake variables such as daily cigarette consumption, COHb. serum cutinine. and nicotine in urine was found, irrespective of whether the analysis was based on urine collected by day or by night (Fig. I, Tables I and 2). In pipe andlor cigar smokers the association between hydruxyproline/creatininc Yatiu and smoke uptake variable, was either nut bis_s- nificant or merely of borderline significance (Table I). The passive smoke expo- sure is estimated as shown in Table 3. No association between hydroxyprolinel creatinine ratio and passive smoke exposure in nonsmokers was seen, irrespective of whether the analysis was based -on the summarizing score (Table I) or on individual exposures at home or at work (data not shown). Since changes in the hydroxypruline/creatinine ratio could be caused by changes in cither the urinary hydruxyproline or the creatinine concentrations or in buth, the association between smoking and urinary creatinine concentrations was checked. Urinary creatinine concentration was found to be negatively cur- rclatCd with smoke uptake in cigarette und pipe/cicar smokers (Tables 2 and 4). HYc/ro.rypruline L.ecretiun Stcurektrdi;.ec! fur Body Surface Hydroxyproline excretion in urine sampled during day, during night, or over 24 hr in relation to different smoking categories for cigarette and pipe/cigar smukers is summarized in Table 5. A significant linear increase in hydroxyproline excretion was found only for night and 24-hr urine of cigarette smokers with `AII urine ,ampteh were analyzed under blind wndittons. All :wailuhle intuPmutiun on the ,ubi"t. smoking habits was fir.rt deposited with Ur. Dietrich Huft'mann. Nuylor Dana Institute Fur Uue;,,e Prevention. American Health Fuundutiun. Valhutla.,This waa passed on to PrutCssur Kasuga in lap,tn I"flokat Univcr.it}. Schuut ut Medicinr, Bohscidai. Isehara :5N-1 t), immrdtatrty alier the .uiatyuul data had been scnt tu uur laboratory. TI BU 31668 I Fit,. arCtte incre to be Hy passi passi work !nc body are h 39. a tivel~

It SYMPO!sIUti1: 'Ot.l)IL'AL PERSPECTIVES ON PASSIVE SMUKING 667 impossible to establish or have been established only in an inconsistent form (as in the ca,c of CS), this is all the more likely to be the case for the low-dose ,ituatiun int:t with in passive smoking. As mentioned at the beginning of this article, only a few stutlies have so far cunsideretl the question of possible impairment of pulmonary function in adult passive smokers. These studies all have the disadvantage, owing to the nature of the data collectiun. of not being able to form groups fur all possible factors of active and passive tobacco smoke exposure. This means that there is a danger of ,clectiun bias involving the nonconsideration of possible positive or negative results. Since, however, a similar pulmonary function methodology was em- pluyed, comparisons between our results and those reported in other studies are pussible. In common with us. Schilling t•t al. ('_7) and Comstock et irl. (7) were unable tu establish any statistically significant reduction uf pulmunary function in passive ,mukers (Table 4). In cunlrast. Kauffmann et ul. (14) t'ountl tlittcrences between TAl3LE 4 IZI.vI1:W UF 'IHF. LIIERArUKIi ON AKIlCL1iS SU FAR PUBLISHGU ON THN: Topic PULMUNAKY FUNCTIUN IN PASJIVts SMUOKIditS Pulmonary Authors Subgroups" function parameter, Results" Schilling NSiH N = 138 FVC. FEVi No statistically significant rt ul.. 1977 PS/H 114 PEF. %tEF 50 reduction of lung function CS 78 14EN 25 for PS/H. White and NS N = 400 FVC. FEVi Concerning FEF 25/75 and Frutb. Iyi30 PS/W 400 FEF 25/75 FEF 75/ts5: PS/W show ~ CS 1300 statistically significantly lower values than NS. Values in PS1W do nut differ essentially from values in CS smoking I- l0 cigarettes per day and in CS without inhalation. Cumstuck PS/ H N= 539 FEV, No stutistically significant ES/H 232 FEVI/FVC reduction of lung function CS 953 fur PS/H. K.wtfmann NS/H N= 1675 FVC. FEVI ' Concerning FEF '5/75 in PS/H 1'?3 F EF 25/75 femalCs aged atl years or more: statistically signiticant differences between NS anJ PS/H. Duse-tffect relationship between reductiun uf lung function in PS/H and daily tobacco consumption ul the hunbanJ. - See text for abbreviations. TI BU 31663 ,

SY,MPOSIUM: .14EDICAL PFKSPECCIVES ON PASSIVE Sr1d[)KING 673 ,Sr u~i were ~J I DAY I N+IIC ~J ~ '+IJIT N~I:J ) aC- 55 ~ 55 . I t tius,ls r SO 501 45 45 drum ri beti ~ 40 40i r :. All. < ~ JS 351 7 10 30 neter ~ thC 'a =5 • t . . . ~ i 25 t , i ~ . • . - . t 20 201 gntti- LdSt. r0{ `.._- _ 10 . 0 510 15 20 25 30 35 4 45 50 0 5 10 15 20 25 30 35 40 45 50 NUlBER uF CSGAAETTFS/WY NU!®EA UF C1-AAE•I'CFS/WY hlt,. I. Correlation between hytlruxypruline (Hf)t')/crcalinine ratio in urine (y' axi,) and daity cig- ,srenr cunhumptiun (.r uxis) in Smukers ages 31-4U years. w eetL daily )und. _)r by wc:en ( sig- :rpu- .)lind :talvt: )r ua cl hy ns or tiun, cur- 1 44. over .:igar utint: with incrt:asing serum cotinine levels. All other associations investigated were found to hc nunsigniticant. Hyclroxyproline excretion in 24-hr urine of nonsmokers was unrelated to their pas,ive smoke expusure, irrespective ut' whether the analysis was based on their pa„ive smoking scores (Fig. 2) or on individual exposures at home or at the worksite tdata not shuw•n). lndividual hydroxypruline excretions in 24-hr urine samples standardized for body surface for cigarette smokers, pipe and/or cigar smokers, and nonsmokers are shown in Fig. 3. The coCfticient of intcrintlividual variation amounted to 48, 39. and 37% for cigarette smokers, pipe/cigar smokers, and nonsmokers, respec- tively. Following up a limited number of subjects (tive cigarette smokers and tive TABLE I CUL•1-FICILNTS OF CUKKF:LAnUN NETWF.F.'N HYUKUXYYK<lI.1NE E.CCKfillt7N ANU SMOKI: UNG\KH VAKLitlI.IkS IN JMUKEKS ANU PAS5IVl; SAfI/KEKS" Smoke uptake variables Cigarettes per day COHb Serum cutininc Score for Nicutine pa»ive in urine ymuking Cigarette ~mukers Day o.3U (S) 0.39 IS) 0.161s) U.3- (S) - :Vlght U.'__ (S) U.171S1 0.11 (S) 0. 10 INS) - Pipeiclgar,mukers - 1).N (NS) U.I4IJI U.117(NS) - {lasl1YC ,mokcrs - - - - -U.uS (NSI S. ,Igndicantty diiferent trum cero IP < U.US): NS• nut Signiticuntty dItf'ereni trulit ceru tP -;s U.U5t. TI BU 31669

SY,MPOSIUM: .14EDICAL PFKSPECCIVES ON PASSIVE Sr1d[)KING 673 ,Sr u~i were ~J I DAY I N+IIC ~J ~ '+IJIT N~I:J ) aC- 55 ~ 55 . I t tius,ls r SO 501 45 45 drum ri beti ~ 40 40i r :. All. < ~ JS 351 7 10 30 neter ~ thC 'a =5 • t . . . ~ i 25 t , i ~ . • . - . t 20 201 gntti- LdSt. r0{ `.._- _ 10 . 0 510 15 20 25 30 35 4 45 50 0 5 10 15 20 25 30 35 40 45 50 NUlBER uF CSGAAETTFS/WY NU!®EA UF C1-AAE•I'CFS/WY hlt,. I. Correlation between hytlruxypruline (Hf)t')/crcalinine ratio in urine (y' axi,) and daity cig- ,srenr cunhumptiun (.r uxis) in Smukers ages 31-4U years. w eetL daily )und. _)r by wc:en ( sig- :rpu- .)lind :talvt: )r ua cl hy ns or tiun, cur- 1 44. over .:igar utint: with incrt:asing serum cotinine levels. All other associations investigated were found to hc nunsigniticant. Hyclroxyproline excretion in 24-hr urine of nonsmokers was unrelated to their pas,ive smoke expusure, irrespective ut' whether the analysis was based on their pa„ive smoking scores (Fig. 2) or on individual exposures at home or at the worksite tdata not shuw•n). lndividual hydroxypruline excretions in 24-hr urine samples standardized for body surface for cigarette smokers, pipe and/or cigar smokers, and nonsmokers are shown in Fig. 3. The coCfticient of intcrintlividual variation amounted to 48, 39. and 37% for cigarette smokers, pipe/cigar smokers, and nonsmokers, respec- tively. Following up a limited number of subjects (tive cigarette smokers and tive TABLE I CUL•1-FICILNTS OF CUKKF:LAnUN NETWF.F.'N HYUKUXYYK<lI.1NE E.CCKfillt7N ANU SMOKI: UNG\KH VAKLitlI.IkS IN JMUKEKS ANU PAS5IVl; SAfI/KEKS" Smoke uptake variables Cigarettes per day COHb Serum cutininc Score for Nicutine pa»ive in urine ymuking Cigarette ~mukers Day o.3U (S) 0.39 IS) 0.161s) U.3- (S) - :Vlght U.'__ (S) U.171S1 0.11 (S) 0. 10 INS) - Pipeiclgar,mukers - 1).N (NS) U.I4IJI U.117(NS) - {lasl1YC ,mokcrs - - - - -U.uS (NSI S. ,Igndicantty diiferent trum cero IP < U.US): NS• nut Signiticuntty dItf'ereni trulit ceru tP -;s U.U5t. TI BU 31669

iYMl't1SIL'M: MGDI(':\1. I't{K51'EC'TIVES ON 1'ASSIVG SMOKINU 675 TABLE 3 St'UKL FUK PASJIVtd SMUKINU Partner -smukes A = U Nu I Yes Smoking at workplace B= 1) vu I Ucca,iunally 2 Heavy Frc~,h ;ur upply C = U Unmati,lactory I Sauslactory 2 Fair 3 Good Score Cor nonsmokers = A T B -~ , nunsmukt:rs) who had been seen three to four times. intruintlividual variations of 36 and iU% for nonsmokers antd cigarette smokers, respectively, were tuund. DISCUSSION A statistically significant dose-response relationship between cigarette ,,moking and urinary hydroxyproline/creatinine ratio was found. This i~ in line with findings by Kasuga et ul. (1 1). In contrast to Ka,uga's results, no aig- niticant association was fuund between passive smuking ancd hydroxypruiine/cre- aunine ratio. Our tindings are not detinitive because they are based on an ad- mttteclly low number of nonsmokers who reported on their passive smoke ex- pusure (Study 3). On the other hand. they should be viewetl.in conjunction with the results fur pipe antllur cigar smokers (Study 2), who may be regarded as highly exposed passive smokers. Huwever, expressing hydroxyproline excretion by means of the urinary hy- Jruxypruiincicreutininc ratio is only justified if the urinary creatinine cuncentra- tiun is unrelated to smoking. Negative correlations were found between smoke uptake and creatinine concentrations in urine for cigarctte: and pipe/cigar.mukers, TABLE 4 CUF.FHC1P.Nt5 uf CUKKtsLA('1UN Iit:rwt:EN CKfiAr1NINE CUNCEN1'K:\rIUN IN UKINfi ANU S\I4)KE UPr:\Kt: V.\KIABLL•S IN SMUKY.Kb 1 -1 CWAKt•:ITGS -V4 U hlPh." iNUKIK CIGAKS Smoke uptake variablwi Cigarettes per day COHb Serum cutinine Nicotine in urine Cigarette smuken ('-4-hr urine) -U'8"• -U..9'• -tl.tb -U.ly I'iperca6ar ,muktr-, 1'-4-hr urlncl - -u.lU -U._t)"• -U.ly" ' 1' 0.05. " Y •• U.UI. TI BU 31671

YRflVLNILVE \1L•:ULCINt 13, 67D-679(19t94) Hydroxyproline Excretion in Urine of Smokers and Passive Smokers' FRANZ ADLKOFEK,''2 GERtiA1tU SCHERER,* AND W.-D. HELLERt 'Forsrluul,4sgrsellsc•Itufr Ruut•hrit und Gestutdhrir ItthH, Mitreluveg !7, D-Il1Ut9 Hutrthur,t; !1. wld tYnstitut /iir Stutistik mtd .Ylutrtemutisrhr 4Yirtsc•ltattst/ret~rie, Uaiverxirat KurAruilc. Pustfurh 6380. U-75r10 Kurl.yruke l. Fedl•rul RepuhFic• uf Grrrllurv Urinary hydroxyproline,excretiun was investigated in 125 malr cigarette ,mukers. 194 male pipe andlur cigar smokers, and 24 male nunamukers. Hydroxypruline excretion was calcUtaLCd either as hyLlrUxyproline/crCatlnine ratio or as body surface-standardiZed amounls ot hydruxyproiine excreted in urine Santpled during day. during night, or over 24 hr. The assuciauun of hyllroxypruline excretion with smoke uptake variables such as daily cigarette consumption, c;uboxyhemogiobtn. serum cutininc, and nicotine in unnr ;uLd with Neit'-reported paasive smoking exposure in nonsmokers was anaiy2ed. The hyJru.vyruolinel creatimne ratio wus found lo be unsuitable as a measure of hydroxyproline excretion since creutinine urine concentrations currelate invereely with ,moke uptake in cigarette and pipe/ cigar +mukers. The amount ul hydruxyproline excreted in 24-hr urine and ,tandardized fur body Surtace was not signLlicantly aabuciutell with Smoke uptake in pipetcigar smukera or exposure tu ru„ivc Smuking in nonsmokers. In cigarette smukers the situation appeared SImllar, although the results were Ie]s clear-cut. Che daLa do not favor the preml,C that measuring urinary hydroxyproline excretion is au accurate method uf investigating a lung- damagtng rltbct of smuking, passive ,muking, or air pollution. e 1984 Acadrmic r«». Inc. INTRODUCTION Knowledge of the effects of smoking and passive smoking has been obtained mainly through epidemiological studies. ln order to gain a greater insight into thL: pathogenesis of smoking-related diseases it is necessary to inve,tigiutc carly Ct'- tect5 of smoking on the human organism. Whether or not passive smoking or even active smoking stimulates the degradation of lung collagen and elastin and causes an increaseld excretion of hydroxyproline Ln urine is a question that has been discussed over the last few years. Kasuga and his co-workers (11, 12. 31) have extensively studied urinary hy- droxyproline excretion in rc;lutiun to exposure to tobacco smoke and polluted .ur. For practical re;L,uns they used, as is widely accepted, hydroxyproline/creatinine ratios as indicators of hydroxyproline ercretion. These authors report a dose- related increase in urinary hydroxyproline excretion in male and female smokers, wives exposed to their husbands' smoking, children exposed to their parents' smoking, and children exposed to automobile exhaust-polluted air (11). They ~,uggest that elevated hydroxyproline/creatinine ratios may be caused by a lung- ' Presented at the Symposium "'.btellicul Perspectives on PusJLVe Smuking." April 9-12. 19Aa. vienna. Austria. ='fu whom requests for reprints >huuld be alldrz,.ed at: Funchungegebellnchaft Rauchen und Ge• ~undhcu mbH. MILLCIWeg 17, U-2LN/U Hamburg 13, Federal Republic ot Germany. 670 IxM/t-73a5/tS-t $3.(N) l,.m righ, ~ j 'hs4 b> -%crJcmac Prce.. Inc .UI non u/ ropruuueuun in eny turm i euncd. TI BU 31666 ljiln Ir urin and upt. ut u exp Srtb :~I Qrui Inbh quc Sr 1981 c:a'+h wcrL mur nwr 1)ilt~ hr u Sam ~ and wuii ablc: tirst Luluz lugi.: to b: urin% w,tlt S'rt ,rnul., :Lnd 7 ul urt cx'Ju .5'rt/ ~;rutt, a'.Lti IntCrl due t u.Ulu

672 ADLKC)Ft K. SCttEkER. AND tiELLEl2 rtnul vticul .til etitocls Hvclru.c.vprulhw anul creatinine in urine. Hydroxyproline and creatinine were analyzed3 with an autuanalyzer (Autoanalyzer Type 11. Technicon Co., Ltd.) ac- curding to the methods used by Husley et al. (9) and Technicon Clinical Methotls No. SEZ-O01 I FC4 ( l7), respectively. Nicutine in arine and cotinine in seruin. Nicotine in urine and cotinine in serum were determined by gas-liquid chromatography based on the method tiescribCd by Hengen and Hengen (8). Scrum and urine samples were stored at -?U°C. All analyses were pertbrmed within 6 months after sampling. Curbuxyhelnuylubin (COHb). COHb levels were measured with a CO-Oxitncter (lnstrumentatiun Laboratories Ltd., Model 182) immetliately after taking the blood samples. Statistical e rulnutiun. Differences between subgroups were tested for signiti- cance using, as appropriate, analysis of variance, the t test, the Wilcuxtm test. or the test for linearity. RESULTS Hydru.eypruline/Creutinine Ratio For cigarette smokers, a small, but statistically significant correlation between the hydroxyproline/creatinine ratio and smoke uptake variables such as daily cigarette consumption, COHb. serum cutinine. and nicotine in urine was found, irrespective of whether the analysis was based on urine collected by day or by night (Fig. I, Tables I and 2). In pipe andlor cigar smokers the association between hydruxyproline/creatininc Yatiu and smoke uptake variable, was either nut bis_s- nificant or merely of borderline significance (Table I). The passive smoke expo- sure is estimated as shown in Table 3. No association between hydroxyprolinel creatinine ratio and passive smoke exposure in nonsmokers was seen, irrespective of whether the analysis was based -on the summarizing score (Table I) or on individual exposures at home or at work (data not shown). Since changes in the hydroxypruline/creatinine ratio could be caused by changes in cither the urinary hydruxyproline or the creatinine concentrations or in buth, the association between smoking and urinary creatinine concentrations was checked. Urinary creatinine concentration was found to be negatively cur- rclatCd with smoke uptake in cigarette und pipe/cicar smokers (Tables 2 and 4). HYc/ro.rypruline L.ecretiun Stcurektrdi;.ec! fur Body Surface Hydroxyproline excretion in urine sampled during day, during night, or over 24 hr in relation to different smoking categories for cigarette and pipe/cigar smukers is summarized in Table 5. A significant linear increase in hydroxyproline excretion was found only for night and 24-hr urine of cigarette smokers with `AII urine ,ampteh were analyzed under blind wndittons. All :wailuhle intuPmutiun on the ,ubi"t. smoking habits was fir.rt deposited with Ur. Dietrich Huft'mann. Nuylor Dana Institute Fur Uue;,,e Prevention. American Health Fuundutiun. Valhutla.,This waa passed on to PrutCssur Kasuga in lap,tn I"flokat Univcr.it}. Schuut ut Medicinr, Bohscidai. Isehara :5N-1 t), immrdtatrty alier the .uiatyuul data had been scnt tu uur laboratory. TI BU 31668 I Fit,. arCtte incre to be Hy passi passi work !nc body are h 39. a tivel~

iYMl't1SIL'M: MGDI(':\1. I't{K51'EC'TIVES ON 1'ASSIVG SMOKINU 675 TABLE 3 St'UKL FUK PASJIVtd SMUKINU Partner -smukes A = U Nu I Yes Smoking at workplace B= 1) vu I Ucca,iunally 2 Heavy Frc~,h ;ur upply C = U Unmati,lactory I Sauslactory 2 Fair 3 Good Score Cor nonsmokers = A T B -~ , nunsmukt:rs) who had been seen three to four times. intruintlividual variations of 36 and iU% for nonsmokers antd cigarette smokers, respectively, were tuund. DISCUSSION A statistically significant dose-response relationship between cigarette ,,moking and urinary hydroxyproline/creatinine ratio was found. This i~ in line with findings by Kasuga et ul. (1 1). In contrast to Ka,uga's results, no aig- niticant association was fuund between passive smuking ancd hydroxypruiine/cre- aunine ratio. Our tindings are not detinitive because they are based on an ad- mttteclly low number of nonsmokers who reported on their passive smoke ex- pusure (Study 3). On the other hand. they should be viewetl.in conjunction with the results fur pipe antllur cigar smokers (Study 2), who may be regarded as highly exposed passive smokers. Huwever, expressing hydroxyproline excretion by means of the urinary hy- Jruxypruiincicreutininc ratio is only justified if the urinary creatinine cuncentra- tiun is unrelated to smoking. Negative correlations were found between smoke uptake and creatinine concentrations in urine for cigarctte: and pipe/cigar.mukers, TABLE 4 CUF.FHC1P.Nt5 uf CUKKtsLA('1UN Iit:rwt:EN CKfiAr1NINE CUNCEN1'K:\rIUN IN UKINfi ANU S\I4)KE UPr:\Kt: V.\KIABLL•S IN SMUKY.Kb 1 -1 CWAKt•:ITGS -V4 U hlPh." iNUKIK CIGAKS Smoke uptake variablwi Cigarettes per day COHb Serum cutinine Nicotine in urine Cigarette smuken ('-4-hr urine) -U'8"• -U..9'• -tl.tb -U.ly I'iperca6ar ,muktr-, 1'-4-hr urlncl - -u.lU -U._t)"• -U.ly" ' 1' 0.05. " Y •• U.UI. TI BU 31671

'L:ABL_k 2 F1)'UKUXI'1'KULINE CUN('ENIKAIIUN. CKLMININL CUNCEN1KAlICIN, ANU HIUKUXYPK(H.INUCKEA77NINL RAIIU (= S1A NUAKU DL\'IA IIiIV) IN UKINL ul y CIGAKLITL SMQKLKS CL.ASSIFICU ACCOKUIN(. 1U THLIK Dqn.Y CIGAKETIE CONSUMI'l1uN d "Day" urine "Night" urinz - Hydruxyprulinel HyJruxqpro6ne/ creatinine creatimne Hy'druxypruline Creatinine r:~liu Nydiuxyprulinr Crentinine ratiu CigareFrrs%Jay N lµg'n)1) (mgtdll im_ gl N (µg/mI) (mg dll lmg/gl x M z = I(1 20 31.2 '~ 21.N 162.0 = 95 19.9 ~ 5.5 20 33.2 s 12.9 141.9 ~ 63 25.3 ~ 7.8 y 11-20 38 28.5 ? 20.1 133.8 Y 85 22.6 ? 611 40 37.5 ~ 21.6 145.0 ~ 86 26.8 ~ 8. 1 21- 30 41 26.7 ~ 13.2 116.9 :t i8 24.1 ~ 8. t 41 30.6 ~ 16.1 128.9 ~ 77 26.2 ? 11.4 z C 31 rt 17 27.1 ~ 12.7 103.9 :t 55 27.4 ~ 7.5 19 26.9 s 1(1.8 83.8 ~ 42 34.5 ? 12.4 ~ rr ~ /"' 0.37 0.02 (t.tll Q(IB 0.02 0.412 ~ -- -- -- -- -- t- " 7rst !ur lincaritp. x ' ( I I

SYMPOSIUM: NtEDIC.-kL PEkSNECTiVES UN P.4SStVt- SMOKING 677 I PIFE .na/aa ilUA•EitE $naFEMS ~-aM S'MlFEFS Man•SNr<ER5 • _ `b • • • • NI N• • ~ • . SruDY 1 SruDr [ StuDr S ~Wv 90 to n,u[M 81) ItUv ra mtr 8U Uutr 831 6 CoEFFI(IEVt OF vYIWE hk.. 3. Hydrurypruline tHOP) excrettun in 24-hr urine ,tunE/urJizeEl fur body ,urth5x with ipecial rclcrence to ,mukinb habus. ishecl urinary creatinine cuncentratiun in smokers. The cause of this is as yet unclear. Our data suggest, huwever, that heavy smokers have higher urine vol- umes as compared with light smokers, leading to a lower urinary concentration of creatinine. This, together with the fact that hydroxyrruline, in contrast to crcatinine, exhibits a high renal reabsurptiun rate I I, Ib). crtight explain at least part of the assuLiatiun IiwnJ. On the other hantl. ~,ome investigat5u', report ,Itghtly reduced crc:atinine ,crum levels tur smokers as cumpareJ with nun- ,mukE:rti, which could ;,l,u cause luwcr urinary creatinine concentrations l5. 7, 15/. This finding has nut been confirmed by other authors 1'_. 4). Whatever the cauaE; tur the decreased urinary creatinine concentrations with increased smoke

676 ADLKUFER. SCHEKEtt. AND HELLEK TABLE 5 tJIKINANY HYUKOXYPKULINL EXCKI:rIUN I=SIANUAKU DEYIATIUNI SI'.iNUANDIZfiII FOK I3UUY SUKhAI't: IN CI(iAKl:'rrF: ANU PiPLIC1UAK SMUKERS CLAJJtFILU ACCUKUINU IU UwI'I:KIiNI' SMUKL UI'fAKL CAIL(,t1KIL•5 Cigarette smukers (N) Pipelclgar muhcn I,\'1 "Day" urine "Night" urine 24-hr urine 24-hr urine G.tugury (m1:n1~,Jay) (mytn, inight) Imynr/a hrl ImNm- ~'_1 ho Ci6;,rcttewJay --10 6.4 _ 4.4 II(1) 10.5 = 4.5 1191 16.5 r 7,1 117) 11-_'0 5.h x 3.d U2S) 12.= = 7.3 140) 17.1 _ 8.9 1.7) a-30 5.6 = 3.61a11 11.8 s h.y1411 18.4 _ 8.9011 >30 5.9 r d.0(17) 13.4 x 5.9119) 18.9 = tl.l 1131 p° 1).70 0._6 0.35 I"ct COHb s 1.5 5 6:'_.6 [37) 10.7 = 5.8 (35) 16.4 _ 7.3 (29/ J.6 n 6.8 11111) 1.6-3.U 5.9 s 3.9121) 11.8 x'4 (_6) Iti.1 : I(1.6 (17) 24.8 = 11. 1 (.1) 3.1-4.5 4.7 : 3.0 (20) 13.8 z 7.1(21) 19.0 _ 9.8 113) 7h.5 _ IU.a[111 4.6-6.0 6.1 = i.6(16) 11.4 x 6.N (18) 16.7 z 7.7 1I:) 23.9 = 10.8001 %6.tl 6.9 z 5,31211 11.3 _ 5.8(18) IPS.Y = 7.61171 20.9 = 8.6(9) r~ tl._7 0.:0 0.47 0.x5 Serum Cunnme Intymtl ,75 5.1 -?.6I4?) 10.5 s 5.9 139! 15.7 = 7.4 1301 23.0 = Y.I 1771 76_15/) 6.8 = 5,1 1_31 10.8 =>.8 12'_) 17.6 = 7.11151 _5 4 : 11.1 I:U ISI-,25 4.9 _ 2.4 1'_21 1..1 ^_ 7.4 1251 18.3 _ 10.3 (171 _'3.h x 7.4 i Ihl =2b-3Ut1 7 y r 4.8 I1_'1 • 1_.I _ 7.U ( Itll 18.7 : 10.0 17) ?25.7 = 13 = 1111 >11N/ 6.U ~ 3.9 117) 15.1 = 6.± 1_'3) 21.0 n 8.1 t lil :_.7 z 9.5 12h1 p• II t' , 1).0 1 t1.0 11.5h ' reat Iou Imcanty. the association being stronger t•ur cigarette smokers than for pipe/cigar smokers. Thus, the (:ft'(:ct of cigarette Smuking on hyJruxypruline excretion (expressed aa hytlruxyproline/creatinine ratio) is partly or cumpletely siniulated by the climin- e 30.0 27.$ .> J > ° 22.5 v > 0 1 I ~ 20.0 rcl, .. 17.51 0 • G L 5.0 , Ish- ~ 12.5 ~ 0 ? un, x ' lo.u > um a i 7.5 5.0 Pa, 5l J.J J.4 1.U 3.J l~l Bh55LVE 5MUK2NG.S1JhG._ NKi. 2. HyJruxypruline (H relauun to pu,bive ,muking sc . UP) exc ore l.r .t retio rl,) n in :1•hr urine stundarJt[ in nonsmokers. eJ tur body ,urta" ty .L\u1 lu Ca. TI BU 31672

YRflVLNILVE \1L•:ULCINt 13, 67D-679(19t94) Hydroxyproline Excretion in Urine of Smokers and Passive Smokers' FRANZ ADLKOFEK,''2 GERtiA1tU SCHERER,* AND W.-D. HELLERt 'Forsrluul,4sgrsellsc•Itufr Ruut•hrit und Gestutdhrir ItthH, Mitreluveg !7, D-Il1Ut9 Hutrthur,t; !1. wld tYnstitut /iir Stutistik mtd .Ylutrtemutisrhr 4Yirtsc•ltattst/ret~rie, Uaiverxirat KurAruilc. Pustfurh 6380. U-75r10 Kurl.yruke l. Fedl•rul RepuhFic• uf Grrrllurv Urinary hydroxyproline,excretiun was investigated in 125 malr cigarette ,mukers. 194 male pipe andlur cigar smokers, and 24 male nunamukers. Hydroxypruline excretion was calcUtaLCd either as hyLlrUxyproline/crCatlnine ratio or as body surface-standardiZed amounls ot hydruxyproiine excreted in urine Santpled during day. during night, or over 24 hr. The assuciauun of hyllroxypruline excretion with smoke uptake variables such as daily cigarette consumption, c;uboxyhemogiobtn. serum cutininc, and nicotine in unnr ;uLd with Neit'-reported paasive smoking exposure in nonsmokers was anaiy2ed. The hyJru.vyruolinel creatimne ratio wus found lo be unsuitable as a measure of hydroxyproline excretion since creutinine urine concentrations currelate invereely with ,moke uptake in cigarette and pipe/ cigar +mukers. The amount ul hydruxyproline excreted in 24-hr urine and ,tandardized fur body Surtace was not signLlicantly aabuciutell with Smoke uptake in pipetcigar smukera or exposure tu ru„ivc Smuking in nonsmokers. In cigarette smukers the situation appeared SImllar, although the results were Ie]s clear-cut. Che daLa do not favor the preml,C that measuring urinary hydroxyproline excretion is au accurate method uf investigating a lung- damagtng rltbct of smuking, passive ,muking, or air pollution. e 1984 Acadrmic r«». Inc. INTRODUCTION Knowledge of the effects of smoking and passive smoking has been obtained mainly through epidemiological studies. ln order to gain a greater insight into thL: pathogenesis of smoking-related diseases it is necessary to inve,tigiutc carly Ct'- tect5 of smoking on the human organism. Whether or not passive smoking or even active smoking stimulates the degradation of lung collagen and elastin and causes an increaseld excretion of hydroxyproline Ln urine is a question that has been discussed over the last few years. Kasuga and his co-workers (11, 12. 31) have extensively studied urinary hy- droxyproline excretion in rc;lutiun to exposure to tobacco smoke and polluted .ur. For practical re;L,uns they used, as is widely accepted, hydroxyproline/creatinine ratios as indicators of hydroxyproline ercretion. These authors report a dose- related increase in urinary hydroxyproline excretion in male and female smokers, wives exposed to their husbands' smoking, children exposed to their parents' smoking, and children exposed to automobile exhaust-polluted air (11). They ~,uggest that elevated hydroxyproline/creatinine ratios may be caused by a lung- ' Presented at the Symposium "'.btellicul Perspectives on PusJLVe Smuking." April 9-12. 19Aa. vienna. Austria. ='fu whom requests for reprints >huuld be alldrz,.ed at: Funchungegebellnchaft Rauchen und Ge• ~undhcu mbH. MILLCIWeg 17, U-2LN/U Hamburg 13, Federal Republic ot Germany. 670 IxM/t-73a5/tS-t $3.(N) l,.m righ, ~ j 'hs4 b> -%crJcmac Prce.. Inc .UI non u/ ropruuueuun in eny turm i euncd. TI BU 31666 ljiln Ir urin and upt. ut u exp Srtb :~I Qrui Inbh quc Sr 1981 c:a'+h wcrL mur nwr 1)ilt~ hr u Sam ~ and wuii ablc: tirst Luluz lugi.: to b: urin% w,tlt S'rt ,rnul., :Lnd 7 ul urt cx'Ju .5'rt/ ~;rutt, a'.Lti IntCrl due t u.Ulu

676 ADLKUFER. SCHEKEtt. AND HELLEK TABLE 5 tJIKINANY HYUKOXYPKULINL EXCKI:rIUN I=SIANUAKU DEYIATIUNI SI'.iNUANDIZfiII FOK I3UUY SUKhAI't: IN CI(iAKl:'rrF: ANU PiPLIC1UAK SMUKERS CLAJJtFILU ACCUKUINU IU UwI'I:KIiNI' SMUKL UI'fAKL CAIL(,t1KIL•5 Cigarette smukers (N) Pipelclgar muhcn I,\'1 "Day" urine "Night" urine 24-hr urine 24-hr urine G.tugury (m1:n1~,Jay) (mytn, inight) Imynr/a hrl ImNm- ~'_1 ho Ci6;,rcttewJay --10 6.4 _ 4.4 II(1) 10.5 = 4.5 1191 16.5 r 7,1 117) 11-_'0 5.h x 3.d U2S) 12.= = 7.3 140) 17.1 _ 8.9 1.7) a-30 5.6 = 3.61a11 11.8 s h.y1411 18.4 _ 8.9011 >30 5.9 r d.0(17) 13.4 x 5.9119) 18.9 = tl.l 1131 p° 1).70 0._6 0.35 I"ct COHb s 1.5 5 6:'_.6 [37) 10.7 = 5.8 (35) 16.4 _ 7.3 (29/ J.6 n 6.8 11111) 1.6-3.U 5.9 s 3.9121) 11.8 x'4 (_6) Iti.1 : I(1.6 (17) 24.8 = 11. 1 (.1) 3.1-4.5 4.7 : 3.0 (20) 13.8 z 7.1(21) 19.0 _ 9.8 113) 7h.5 _ IU.a[111 4.6-6.0 6.1 = i.6(16) 11.4 x 6.N (18) 16.7 z 7.7 1I:) 23.9 = 10.8001 %6.tl 6.9 z 5,31211 11.3 _ 5.8(18) IPS.Y = 7.61171 20.9 = 8.6(9) r~ tl._7 0.:0 0.47 0.x5 Serum Cunnme Intymtl ,75 5.1 -?.6I4?) 10.5 s 5.9 139! 15.7 = 7.4 1301 23.0 = Y.I 1771 76_15/) 6.8 = 5,1 1_31 10.8 =>.8 12'_) 17.6 = 7.11151 _5 4 : 11.1 I:U ISI-,25 4.9 _ 2.4 1'_21 1..1 ^_ 7.4 1251 18.3 _ 10.3 (171 _'3.h x 7.4 i Ihl =2b-3Ut1 7 y r 4.8 I1_'1 • 1_.I _ 7.U ( Itll 18.7 : 10.0 17) ?25.7 = 13 = 1111 >11N/ 6.U ~ 3.9 117) 15.1 = 6.± 1_'3) 21.0 n 8.1 t lil :_.7 z 9.5 12h1 p• II t' , 1).0 1 t1.0 11.5h ' reat Iou Imcanty. the association being stronger t•ur cigarette smokers than for pipe/cigar smokers. Thus, the (:ft'(:ct of cigarette Smuking on hyJruxypruline excretion (expressed aa hytlruxyproline/creatinine ratio) is partly or cumpletely siniulated by the climin- e 30.0 27.$ .> J > ° 22.5 v > 0 1 I ~ 20.0 rcl, .. 17.51 0 • G L 5.0 , Ish- ~ 12.5 ~ 0 ? un, x ' lo.u > um a i 7.5 5.0 Pa, 5l J.J J.4 1.U 3.J l~l Bh55LVE 5MUK2NG.S1JhG._ NKi. 2. HyJruxypruline (H relauun to pu,bive ,muking sc . UP) exc ore l.r .t retio rl,) n in :1•hr urine stundarJt[ in nonsmokers. eJ tur body ,urta" ty .L\u1 lu Ca. TI BU 31672

'L:ABL_k 2 F1)'UKUXI'1'KULINE CUN('ENIKAIIUN. CKLMININL CUNCEN1KAlICIN, ANU HIUKUXYPK(H.INUCKEA77NINL RAIIU (= S1A NUAKU DL\'IA IIiIV) IN UKINL ul y CIGAKLITL SMQKLKS CL.ASSIFICU ACCOKUIN(. 1U THLIK Dqn.Y CIGAKETIE CONSUMI'l1uN d "Day" urine "Night" urinz - Hydruxyprulinel HyJruxqpro6ne/ creatinine creatimne Hy'druxypruline Creatinine r:~liu Nydiuxyprulinr Crentinine ratiu CigareFrrs%Jay N lµg'n)1) (mgtdll im_ gl N (µg/mI) (mg dll lmg/gl x M z = I(1 20 31.2 '~ 21.N 162.0 = 95 19.9 ~ 5.5 20 33.2 s 12.9 141.9 ~ 63 25.3 ~ 7.8 y 11-20 38 28.5 ? 20.1 133.8 Y 85 22.6 ? 611 40 37.5 ~ 21.6 145.0 ~ 86 26.8 ~ 8. 1 21- 30 41 26.7 ~ 13.2 116.9 :t i8 24.1 ~ 8. t 41 30.6 ~ 16.1 128.9 ~ 77 26.2 ? 11.4 z C 31 rt 17 27.1 ~ 12.7 103.9 :t 55 27.4 ~ 7.5 19 26.9 s 1(1.8 83.8 ~ 42 34.5 ? 12.4 ~ rr ~ /"' 0.37 0.02 (t.tll Q(IB 0.02 0.412 ~ -- -- -- -- -- t- " 7rst !ur lincaritp. x ' ( I I

SYytPUSIUM: %,tEDICAL PERSPECTIVES ON PASSIVE SMOK1NG 671 damaging effect ot' nitrogen dioxide derived from tobacco smoke and air pollution t_1). Such an effect of nitrogen dioxide is known from the literature ( I4), ln ,c:verai studies involving more than 400 subjects we have investigated the urinary hydroxyproline excretion in cigarette smokers, pipe and/or cigar smokers, and nonsmokers. These studies were originally designed to measure the nicotine uptake in relation to smoking habit. However, they also allowed the examination uf urinary hydruxypruline excretion in relation to the subject's smoking habit or exposure to anibient tobacco smoke. SUBJECTS AND METHODS .5uhjet•ts Male cigarette ,mokers, pipe and/or cigar smokers. and nonsmokers were re- cruiteal by newspaper advertisements, by posters in shups, facturies, public build- ings, and universities as well as by word uF muuth. All subjects tilled in a detailed questionnaire related to lifestyle, occupational status, and psychological traits. Srrrdy 1. In the tirst tield study. which lasted from November 1980 to March 1981, _200 cigarette smokers were recruitcd. Three blood samples were taken from each subject on 3 different days over a period of 3 weeks. The blood samples were taken in our laboratory between 4 and 7 PM. Urine was collected from murning until the time of blood sampling, then continued until the following nturning. The samples were categorized as "day" or "night" urine, respectively. Data for day and currespunding consecutive night urines were evaluated as 24- hr urines. To an aliquot of each urine sample U.11/c (v/v) toluene was added. tiamplCS were stored at -?U°C until December 1983, when the hydruxypruline ;wd creatinine analyses were performed. Results t'ur day and night urines were cvaluatecl separately. Since not all six urine samples for each subject were avail- uhle at the time uf hydroxyproline determination,. it was decided to include the tir,t day or night urine available from each subject for evaluation. Due to phy,- iulu5ically raised values in adolescents ( lU) and, perhaps in sume.casCs, pathu- lugtcally elevated values in elderly pipe and cigar smukers ( 13), it was also decided tu base our evaluation only on subjects betweCn ages 21 and 60. Thus 116 day urtnes. 120 night w•ines, and 88 24-hr urines from cigarette smokers could be c~:iluatecl. St«cl~, 2. In asecund study, from May to July 1981, 199 pipe and/or cigar Snwkers were recruited. From each subject a blood sample was takzn between 4 and 7 P%t. Before blood sampling, 24-hr urine was collected. Due t.) :Lvailability ut' urine samples at the time of hydroxyproline and creatinine analyses and after cxclu,iuna for age, 154 urine samples from pipe/cigar smokers could be evaluated. .1rud r 3. In a third -study, from March to July 1983, '24 nonsmokers were re- ~ruwtcJ. From each subject a blood ,ample was taken between 4 anJ 7 Ptit. Urine Na, Lullectcd over a period of 24 hr before blood sampling. This group was irucnded, to be the c:untrul group tur thc smokcrs: One subject had to be excluded due to age younger than 21. lending 23 urine samples from nonsmokers for tval- uwuun. TI BU 31667

C 678 ADLKOFER, SCFIEEZEK, AND HELLEft uptake may be, the conclusion must he drawn that hydroxyprolinelcreatinine ratiu is nut a suitable parauneter to express urinary h~droyvproline excretion in ,mukers. We, therefore, reevaluated our data using body surface-standardized hydroxyproline excretion in day, night, and 24-hr urine samples. None of the investigated associations were significant except those between s4rum cotinine and hydroxyproline in night and 24-hr urine samples of cigarette smokers. We do not believe that this result gives any evidence of a cullagen- and elastin-degrading ettect of tobacco smoke in the smoker's lung, because there are at Ic;ast three other possibilities that must be considered: (a) this single result may have been obtained by chance; (b) nicotine may enhance the release of growth hormone (20). which seems to be related to an increased urinary hydroxyproline excretion (18); and (c) urinary hydroxyproline excretion may depend on urine volume, which is found to be higher in smukers. In our view, it is rather unlikely that any assumed damaging effect of smoking on lung collagen or elastin coultd be demonstrated by measuring hytlruxypruline excretion in urine. A large number of factors exert their influence on hydroxy- proline excretion in urine (1). Our finding of high inter- and intraindividual vari- atiuns support this view. The large amounts ot' urinary hydroxyproline originating from bone and skin collagen metabolism may prevent the detection of small changes in hydroxyproline derived from lung tissue. Lung collagen and elastin constitute only 1~Io ot'total body collagen and elastin (3). Fur this reason Wentlel and Rehpenning (19) du not recommend hydroxyproline measurements for the diagnosis of chronic bronchitis and emphysema. These authors did not tintf any difference in hydroxyproline excretion hetwet:n normal subjects and patients with chronic bronchitis and emphysema as long as the heart function of these patients was not det:umpen,ated. The same applies for urine desmu5ine, the crosslinking amino acid unique to elastin. Davies e1 cri. (6) found that urinr desmosine is unrelatcnd tu cigarette smuking, and attributed their result to high ba.al urine levels of this amino acid. This suggests that dc.musine also cannot be regarded as a useful indicator of lung tissue degradation. Altobether, our studies do nut titvur the idea that mt:astu-inis h}.Iruryproline excretion in urine is a suitable method for investigating c:arly effects ut" smokins, passive smoking, and air pollution in man. ACKNOWLEDGMENTS rhe authors are indebted to Protes,ur Hitunhi Kanugu and his co-workers at Tokai University tur determining ~;rCatinine and hydruxypruhnc in urine: to Dr. Gi,eia Hrnce and Mrs. Claire Hruby tur expert a1listance in preparing lhe tableS anlt ttSurel and the manU.%crtPl: and to Dr. Dietrich Huttmafln. Naylor Dana Institute. American Health Foundation, Vulhallu, New York, who enabted the study tu be pertormCd under blind conditions. REFERENCES I. Adams. E.. and Frank. L. Vletabulian ul ptuttne and thu hydroxyprutines.Airnu. Rcc Hr(,t Irevn. 49. IIN)5_ IUhI I IkJHU)_. 2. Atidrtu. L. H., Miller. D. C.. Slullunr.. R. A.. Ehrlit;h, S. P.. and Jones. 1. P. tpidcuuulugicst in~liviJual ~1uJy .tutly of coronary dieuaae risk facturti. I. Study tlesign and chatuctrriaUcs of ,uhjecu. .-Iinrr. J. L•pula,nial. 87, 73-16b r 1%2i). TI BU 31674 3 4 K IU It 13 14 V. ,

668 KENTNER, TRIEBIG. AND WELTLE never smokers and passive smokers at home with regard to the FEF'-5/75 pa- ranuter, and also a dose-effect relationship between the reduction in pulmonary function in passive smokers and the daily tobacco consumption of the 5ubject's spouse. Htiwuver, thc:ae relationships were statistically "ignit'icant only for women, and then only for certain age groups. To date, White and Froeb (35) are the only authors who unreservedly tind that passive smoking leads to an impair- ment of pulmonary function. The predicted value reductions in the case of persons passively exposed to tobacco smoke at the workplace do not, however, differ essentially from those seen in active smokers consuming between I and 10 cig- arettes a day. This result is difficult to reconcile with a rational dose-elft:Qt relationship under the assumptions made earlier (sze also Lebowitz (1y)J. Taking our own results into consider•ation, it seems that the passive inhalation of tobacco smoke at home or in the workplace by healthy individuals probably does not lead to any essential impairment of pulmonary function. To what extent this also applics to extreme passive exposure to smoke must remain the subject of further investigation. Thus, for example, persons working in restaurants or in poorly ventilated rooms may be exposed to high levels of smoke pollution. ACKNOWLEDGMENTS We thank Dr. Mentzel. Bundesanstatt t7ir ArbCit. Nuremberg; Dr. Juhn, from the tirm of Diehl. Nuremberg: Dr. Bressel and l)r. Heynen frum the tirm uf Siemens. trlangcn. Metlit:al Engtneenng Group. and the management uf the various tirms, the personnel directurs. and the employee repre- sentatives who uppurte6 our investigatiun and made the study possible. Further• we express uur Npectul thunks to Mrs. Schmidt and Mr~. Weydanz fur carrying out the pulmonary function analyses and estubiishing the data records. REFERENCES I. Bake. B. ls maximum mid-e.piratory tluw rute sensitive to small airways obstruction! Eur. J. Rrspir. Dis. 62, 150-151 (1981). 2. Berend. N., Wright. J. L., Thurlbeck, W. M.. Marlin. G. E., and Woolcock, A. J. Small airways dise:uc: Reproducibility of measurement and corretatiun with lung function. Chrsf 79, 263- _'bK 11981). 3. Camner. P.. Philipson. K., and Arvidsson, T. Withdrawal ut'cigarette smoking: A study on tra- cheubronchial clearance. Are•h. Environ. Heu(th 26, 90-92 (1973). 4. Cherniack. R. M.. and Raber, M. B. Normal standards fur ventilatory function using an auto- mated wedge spirometer. Airier. Rev. Rrapir. pis. 106, 38-461 l97'_). 5. Dodge. R. The effects of indoor pollution on Arituna children. Arc•h. Environ. Heulth 37. 151- 155 (1982). 6. Cochrane. G. M.. Prieto, F.. and Clark. T. J. H. lntrasubject variability of maximal expiratury tlow volume curve. Thorax 32. 171-176 (1977). 7. Cumstock, G. W., Meyer, M. B., Helsing, K. J., and Tockman. M. S. Respiratory effects of huuaehoW exposures to tuhaccu smoke and gas cooking. Airier. Rev. Respir. Ui.t. 124, 143- I4>+ I I`ttl l l. 8. Cusiu, M. H.. Gheuo, H., Hogg. 1. C.. Corbin. R., Loveland. M., Dosman. J.. and Macklem. P. T. The relation between ,truourul changes in small atrwa" mnd pulmonary function tesu. New Enxl. J. Mrrt. 298, I.'.77-1281 (1477). 9. Cusiu, M. G., Hale, K. A., Niewoehner. D. E. Morphologic and murphumetric effects ut' pro- longed cigarette smoking on the .mail airways. Airier. Rrv. Resptr. Uia•. 122, =65-=71 (1y2S1/1. 10. Gelb, A. F.. Williams, A. J., antl Lamel. N. Spirometry-FEV, vs FEF 25-75 Percent. Chr.1t 84. 47i-74711ytl3). 11. Ha,.,clMa.tt• U., Humble, C. G., Graltam, M. G., and Anderson, H. S. Indoor environmental Jeterminants of lung function in children. Airier. Rev. Respa. Dis. 123, 479-85 (1981). TI BU 31664 12. Hug'ULL. CUna 13. Hugucl. .ti1 e c1 14. K.rutiitt risk 15. Ktu,tet .{ht. 1h. Knurh. Its u Dr,. 17. KnuJst Curc. I1y7, IkS. Lebuw, tu m" 19. LebuwE puhlt 20. J. s1, 21. 1dcFad.. Mrd. 22. Mc F,td~ 'Sv- 23. Murrts. adult 2a. Niewuc of y~=5. Petty. l patht hum.I .6. QuanJer and ~ :1. Schtltinh Lunu (1y7' '_1S. S,;hulz. 29. Jpeicer pulm. 3-Itt )11. Tager. I the p 31. Thurlbr. UrsD t 3:. Vutuc. t nuL , .`. 33. Weiss, respu ot at. la. Weu~. ~muti 15, Whtte.. tuba, 4n. Krusk.u Stuu.'

668 KENTNER, TRIEBIG. AND WELTLE never smokers and passive smokers at home with regard to the FEF'-5/75 pa- ranuter, and also a dose-effect relationship between the reduction in pulmonary function in passive smokers and the daily tobacco consumption of the 5ubject's spouse. Htiwuver, thc:ae relationships were statistically "ignit'icant only for women, and then only for certain age groups. To date, White and Froeb (35) are the only authors who unreservedly tind that passive smoking leads to an impair- ment of pulmonary function. The predicted value reductions in the case of persons passively exposed to tobacco smoke at the workplace do not, however, differ essentially from those seen in active smokers consuming between I and 10 cig- arettes a day. This result is difficult to reconcile with a rational dose-elft:Qt relationship under the assumptions made earlier (sze also Lebowitz (1y)J. Taking our own results into consider•ation, it seems that the passive inhalation of tobacco smoke at home or in the workplace by healthy individuals probably does not lead to any essential impairment of pulmonary function. To what extent this also applics to extreme passive exposure to smoke must remain the subject of further investigation. Thus, for example, persons working in restaurants or in poorly ventilated rooms may be exposed to high levels of smoke pollution. ACKNOWLEDGMENTS We thank Dr. Mentzel. Bundesanstatt t7ir ArbCit. Nuremberg; Dr. Juhn, from the tirm of Diehl. Nuremberg: Dr. Bressel and l)r. Heynen frum the tirm uf Siemens. trlangcn. Metlit:al Engtneenng Group. and the management uf the various tirms, the personnel directurs. and the employee repre- sentatives who uppurte6 our investigatiun and made the study possible. Further• we express uur Npectul thunks to Mrs. Schmidt and Mr~. Weydanz fur carrying out the pulmonary function analyses and estubiishing the data records. REFERENCES I. Bake. B. ls maximum mid-e.piratory tluw rute sensitive to small airways obstruction! Eur. J. Rrspir. Dis. 62, 150-151 (1981). 2. Berend. N., Wright. J. L., Thurlbeck, W. M.. Marlin. G. E., and Woolcock, A. J. Small airways dise:uc: Reproducibility of measurement and corretatiun with lung function. Chrsf 79, 263- _'bK 11981). 3. Camner. P.. Philipson. K., and Arvidsson, T. Withdrawal ut'cigarette smoking: A study on tra- cheubronchial clearance. Are•h. Environ. Heu(th 26, 90-92 (1973). 4. Cherniack. R. M.. and Raber, M. B. Normal standards fur ventilatory function using an auto- mated wedge spirometer. Airier. Rev. Rrapir. pis. 106, 38-461 l97'_). 5. Dodge. R. The effects of indoor pollution on Arituna children. Arc•h. Environ. Heulth 37. 151- 155 (1982). 6. Cochrane. G. M.. Prieto, F.. and Clark. T. J. H. lntrasubject variability of maximal expiratury tlow volume curve. Thorax 32. 171-176 (1977). 7. Cumstock, G. W., Meyer, M. B., Helsing, K. J., and Tockman. M. S. Respiratory effects of huuaehoW exposures to tuhaccu smoke and gas cooking. Airier. Rev. Respir. Ui.t. 124, 143- I4>+ I I`ttl l l. 8. Cusiu, M. H.. Gheuo, H., Hogg. 1. C.. Corbin. R., Loveland. M., Dosman. J.. and Macklem. P. T. The relation between ,truourul changes in small atrwa" mnd pulmonary function tesu. New Enxl. J. Mrrt. 298, I.'.77-1281 (1477). 9. Cusiu, M. G., Hale, K. A., Niewoehner. D. E. Morphologic and murphumetric effects ut' pro- longed cigarette smoking on the .mail airways. Airier. Rrv. Resptr. Uia•. 122, =65-=71 (1y2S1/1. 10. Gelb, A. F.. Williams, A. J., antl Lamel. N. Spirometry-FEV, vs FEF 25-75 Percent. Chr.1t 84. 47i-74711ytl3). 11. Ha,.,clMa.tt• U., Humble, C. G., Graltam, M. G., and Anderson, H. S. Indoor environmental Jeterminants of lung function in children. Airier. Rev. Respa. Dis. 123, 479-85 (1981). TI BU 31664 12. Hug'ULL. CUna 13. Hugucl. .ti1 e c1 14. K.rutiitt risk 15. Ktu,tet .{ht. 1h. Knurh. Its u Dr,. 17. KnuJst Curc. I1y7, IkS. Lebuw, tu m" 19. LebuwE puhlt 20. J. s1, 21. 1dcFad.. Mrd. 22. Mc F,td~ 'Sv- 23. Murrts. adult 2a. Niewuc of y~=5. Petty. l patht hum.I .6. QuanJer and ~ :1. Schtltinh Lunu (1y7' '_1S. S,;hulz. 29. Jpeicer pulm. 3-Itt )11. Tager. I the p 31. Thurlbr. UrsD t 3:. Vutuc. t nuL , .`. 33. Weiss, respu ot at. la. Weu~. ~muti 15, Whtte.. tuba, 4n. Krusk.u Stuu.'

SYMPOSIUM: NtEDIC.-kL PEkSNECTiVES UN P.4SStVt- SMOKING 677 I PIFE .na/aa ilUA•EitE $naFEMS ~-aM S'MlFEFS Man•SNr<ER5 • _ `b • • • • NI N• • ~ • . SruDY 1 SruDr [ StuDr S ~Wv 90 to n,u[M 81) ItUv ra mtr 8U Uutr 831 6 CoEFFI(IEVt OF vYIWE hk.. 3. Hydrurypruline tHOP) excrettun in 24-hr urine ,tunE/urJizeEl fur body ,urth5x with ipecial rclcrence to ,mukinb habus. ishecl urinary creatinine cuncentratiun in smokers. The cause of this is as yet unclear. Our data suggest, huwever, that heavy smokers have higher urine vol- umes as compared with light smokers, leading to a lower urinary concentration of creatinine. This, together with the fact that hydroxyrruline, in contrast to crcatinine, exhibits a high renal reabsurptiun rate I I, Ib). crtight explain at least part of the assuLiatiun IiwnJ. On the other hantl. ~,ome investigat5u', report ,Itghtly reduced crc:atinine ,crum levels tur smokers as cumpareJ with nun- ,mukE:rti, which could ;,l,u cause luwcr urinary creatinine concentrations l5. 7, 15/. This finding has nut been confirmed by other authors 1'_. 4). Whatever the cauaE; tur the decreased urinary creatinine concentrations with increased smoke

SYMPOSIUM: MEDICAL. PERSPECTIVES ON P;tSSIVE SMOKING 669 75 pa- aunary bject'!, lly for ±51 are mpair- erwns differ 10 ci.g- -Zttect autiun obub{y exLt:nt ,ubject 12. Hugotl. C.. Hawkins. L. H.. and Astrup. P. Exposure of passive smokers to tobacco smoke cunstitwents. IrH. Arclt. Occup. Etrvirun. Heulrh 42. -'1 -'-9 i 1978). 13. Hugud, C. Indoor air pollution with smoke constituents-An experimental investigation. Prev. .Yled. 13, 5!f'_-jtitl (1984). 14. Kaultmunn, F.. Tessier. J. F., and (7rtol, P. Adult passive smoking in the home environment: AA risk tactur tur chrunic airtluw limtlauun. .9+trer. J. EprJiul. 117, 26N-2t;li (1983). 15. Klusterkutter, W., and Gunu. E. Zum Problem des Passivrauchens. Lrwru(bl. Bukteriul. Hr,4. I. .-Ihr. (Iri,p. B 162, 51-6y (1976). 16. Knudson. R. J.. Burrows, B., and Lebuwitz. M. D. The maximal expiratory tlow-vl,lume curve: Its use in the detection of ventilatuiy abnormalities in a population study..line•r. l;rr. Rr.+pir. 1)a. 114, tS71-M7y 11976). 11 Knudson. R. J.. Slutin. R. C.. Lebuwitz. M. D.. Burrow. B. The maximal expirutory tluw-vulume curve. Normal standurds, variability and ettectS uf age. Amrr. Rev. Respir. Dis. 113, 5tf7-60(1 119761. 18. Lebuwitz. M. L).. Armet. D. B., and Knudson. R. The effect of pa»ive smoking on pulmonary tunctiun in children. i:)n•irwl. lnternut. 8, 371-373 (19tS2). 19. Lcbuwitz. M. D. "Intluence uP Passive Smoking on the Pulmonary Function-A Survey." Un- published d•ata. ~s or iR I :0. McFndden, E. R., and Linden. D. A. A reduction in maximum mid-expiratory flow rate. Amer. 7 7 7 2). (17 t. I J. Lled. 52, 7:5- 3 21. McFadden. E. R.. Kiker. R.. Holmes. B., and Degroot. W. J. Small airway disease. Amer. l. ,Dtrd. 57, 171-1 ii _' 11974). ?1. McFadden. E. 12._ and lneram. R. H. Periohentl airwav ob.truction_ J. Amer. Wed. A.ssuc. 235. : y-_nu t i7ini. """°`ff. i :3 N d Johnwn L C S irumetric standards f A r h lth i J h K l k p un u y nur,mu .......,... orr s l~. u ea ing ``j" ` a A17 (1y71) Di 103 57 l i l A R R - inrr rv rsp r s s.....,. .u u t a. I`/ICwUCnnCC. U. 6.. 111cInCTman. J.. anu 1cIGC, U. o. rllnuHVllc lnilnRCN In mC PCRPnCfUI airways •u• E l 1 bled lU 75j_7~rs I t)7~1 V """r"' t d' n i r tte k r l .,...,. , y,ru g c ga smu e s : ni e :3. Peuy, f. L.. Silvers. G. W.. Stanfurd, R. E.. Baird. M. A., and Mitchell. R. S. Small airway pathology is related to increased closing capacity anll abnormal slope of Phase 111 in excised Eur. 1_ i human lungs. Antrr. Rev. Respir. Uis. 121, tJy-~?o (19140). :A. Quanjer, P. H., (Ed.). "Standurdized Lung Function Tc,ung." European Community for Coal .t,rwav, I and Stecl. Luxembourg. July 1983. 9, 263- )! =t ~cmutng, K. J. r.. LClal. n. u.. rrnuw, 3. L.. 6CCA, u. j., acnuenoerg, j. n., anu ouunuys. n. Lung tuncttun. resptratury uw]ease anu SmuKmg in tamttles. Amer. J. tpolrnuut. 11.16, Sti-L83 t 19771 . - ,-_ a 2M. Schulz, V. Small airways disease. Prar. Pneumuf. 34, 187-218 11'Ii{0). V d 1 (i 13i M 1t l R t li 2y S r i h J i d F E 1 J . t,cng . up,. . ., ;tn ra ory c seaae rates an rt;t,tt,r ~ p r~er. rrr +. er. . esp .. .. pulnwnary function in children associated with NU, expuaure. Antrr. Rrv. Respir. Uis. 121, -r rtr_ n 3-Il)t192S0/. )11. Tager. I. B., WCis]. S. T., Rosner. B.. and Speizer. F. E. Effect of pJrental cigarette ~mUking on 1y791 hdd J id i l 15 6 h f i f t E 110 ' ren. ~ mer. . rnr u . unet on o c p , - _ , e pulmonary 1 nri,torv  t 31. fhurlbeck. W. M. Smoking, airtlua limitatton, and the pulmonary circulation. Amer. Rrv. Rrspir. :,,,.,. „F g Urs. 122. 12t3-ItS611Htf01. 3_ 143- >f t_, vutua. L. t_ungenr;rcosnstt:u unu ntasrvraucnen: VuanrntauvC voortegungen. `enrnuor. aur;ae- riul. Hs).'. l. Abi. Oriy. 8 177, 90-95 (19tl3). ,ckfeta_ g Ji. Weiss. S. T.. Tager, 1. B.. Speizer. F. E., and Rosner. B. Persistent wheeze: lts rclatiunship to d I l l i k t t i l i ill eve pu munury „uun samp ing en u unct un in a pupu e ratury ness. cigarette ,mu , tc,t, r re,p ut children. A»rer. Ra v. Respir. U6. 122, 697-707 (1980). ,)f nro- i t;. Weis.. S. 1.. Tager. 1. B.. Schcnker. M.. and Speizer, t. The health effects ul involuntary i t')}nll. li mrtuning. ninrr. rtrr. ice.,plr. tn.l. l4o, »J-r.: t 17wr. Chr.lt ~ iS White. J. R.. and Froeb. H. F. Small-airways, dysfunction in nonsmokers chruntcaly espu~ed to tobacco smoke. ,Veu, Ent;l. J. Nied. 302. 7'_lt-7?3 1197SU). l i J a f W W ntrr. ys h. . . . H.. and Wullis. . A. Use o ranka in one-criterion variance ana mrnts,l a w+ Kruskal. .5 tw t., t. A., t. 47. 5 tS 3-627 11952). TI BU 31665

SYYIPUSIIIM: MEDICAL PEttSPECftVES ON PASSIVE SMOKING 681 Ob- -tc:d ,ted Ay _ of due ,ing ults .Nu~b ed. Cen ~ rds ldy vated risk of lung cancer (n ='_00), na,al sinus cancer (n =?8), brain tumors (/l = 34). and cancer of all sites (n ='_7O5). Lun,4• Cunrer A total of 429 deaths from lung cancer in women was recorded during the 16 y ears of lilllow-up ( 1966- lyti l). Of these deaths, 303 occurred among nonsmokers and 200 among 91,540 nonsmoking married women whose husbands' smoking habits were known. The standardized mortality ratios (SMRs) of lung cancer in nonsmoking women were 1.00, 1.36, 1.42. 1.58, and 1.91 when husbands were nonsmokers, ex- ,mukc;rs, or' daily smokers of 1- 14, 15- l9, or 20 or more cigarettes per day, respectively (one-tail P value = 0.00178) (Table I). A similar dose-response relationship was observecd by age and occupation of the husband (Table 2). This tendency is in Nharp contrast with that of stomach cancer, where no re- Wionship at all exists between the risk in nonsmoking wives and the amount of smuking by the husband (Tables 3 and 4, Figs. I and 2). Similur trends of lung cancer risk elevation in nonsmoking women with the increase in the extent ut'the husband's smoking were observed in each time period uf observation, in each age broup, both by age of husbands and by age of wives. in each occupational group, and in most areas under ubsc;rvatiun (internal con- ,istency) ( hig. 3) (7). No other characteristics of husbands or wives themselves were t'uund to elevate the risk of lung cancer in their nonsmoking partners (7) (Table 5). Nonsmoking husbands with smoking wives also showed an elevated risk of lung cancer, the SMRs being 1.00, 3.14, and 2.31 in nonsmoking wives, wives ,muking I-ly cigarettes, and wives smoking 20 or more cigarettes daily, respec- tively (P = 0.0177). This observation also strengthens the evidence listed above (Table 6). TABLE I LLNG CAN['h.R WliI'AIJTY IN WOMt:N BY AGE GkI)UP AND BY HU5NANUS' 5Ml1KING H.4Urr lPAC1ENr HI{k1I-.1.1 .l tNUNSMIIKb.kl" Husband'> >muking habit Husbnnd'. ,gc group Nunamukar Ex+muker I-14/day IS-19lday '-U. day Tuuil 40-49 9U_Sy M1-6v 70-74 Tutat 4 Itl is 5 37 6.229 7.791 7.120 755 :1.My5 1 I._55 3 1,922 II =,6257 2 348 17 6.212 B 8,621 20 9.668 28 7.243 2 612 58 .6.1i4 6 5,158 8 4.1.152 9 :,513 1 105 24 11.828 16 10.764 24 9,620 23 4.651 1 236 64 25.461 35 65 219 11 '-1111 32,t127 33.233 :4,:14 2.1146 91.54/1 I he wcightcd putnl I iN) 36 '. 113 1 ~ 1.42~?.ll l I. SIS ~ 38 ~ 1.91 ' 71 ~ Con mute ut ratr ratio . . U.tl5 tUl I v11 1.3a Manmi cxtenswn .nJ test-buxd 9014 :ht :.Y15 U n- . wnlitlenco linuts 45<1.t/2 1 One-tail .a C- , I t>,i P vulue II.I)U171S Nanlel-Haen.zcl chi 1.0855 1.81911 3.U?95 y1f-~. _ t)nc•tsd P value U.13Ny U.t1337 U.UUI2

,Kwa3 SYMPOSIUM: !AE:DICAL PERSPECTIVES ON PASSIVE SMOKING 683 TABLE 4 SR1AfACH CANCEN MORTALI'rY IN WUMI:N HY AGL: GHUUP. HY OCCUPAr1UN. ANU HY Itt:SHANUS' SM(/KING HAHrt (P.111L•N'r HL:H1tr1.1.i, NUNSMUKI:H1" Husband's hmuking hubit Huabantl's Husband's Exsmuker uccupuuun age group Nonsmoker 1-19/tiay 2t)tiJay Tutal Agrtcultural 40-49 13 2.502 41 5,941 25 3,636 79 I.,U79 worker SU-59 37 3.497 56 6,812 37 3,514 130 13,823 6U-6y 77 4,U2i4 116 6,845 43 ?,153 236 13,081 70- 3 323 13 446 3 ts9 19 858 l'atal 13U IU:-tU6 226 _ll•U41 IUM 9,341 464 39.841 Other 40-49 Itf 3,7?7 38 9,093 23 7,1_8 79 19.948 50-59 23 4..94 76 9,830 44) 6,3U6 139 19,430 60-69 44 3.036 83 5,598 35 2.499 162 11,133 70- 4 432 3 619 3 137 10 1,188 TutUl 89 11,489 ?0U 24.140 10 1 16.U7U 390 51,699 , ~ y Mantel extension The weighted point estimate I.W 1.03 < ~:Hy 1.05 < :~ chi 0.234 of rate ratio and teat-bascJ vth/, cuntiUent:e limlts OnC-tail P value .40749 1 Muntel-Haanszel chi - 0.298 0.486 Une-tall P value 0.38285 U.3134>f Pru7pective 3tuJy, 1966-1981. Japan. .IGN...Y ..NCtR ~(: a j541 ir 1 ]3 r G 5 Gelatlve NISK 3y3jtt 0. ,..O.nJ. I-u,. I. RcLttive n,k, ui lung cuncer and ,tunr.lch cancer In'11.54U nonsmoking wives by hu.bunJs' ,luuking habit. iNru,rective 1tuJy. I'Nlh-lyrit. Japun.t TI BU 31679

684 T. HIKAYAMA ,MartatiCY ~ natl0 i yusouna'> U<a ltrouu • 0-y9 5C - 59 60 ~ Can<f. .I n< I 1 ~ <.nce. 0 I 1. C t i 0 1 0 1 0 1 s I I I I I` a~nON i~t•t9 :0• 0 vON t.l-t9 ;0• ~so.nas +n..•~nq n.aa n.ao.nos +ro.~ny naa~t j n 1 nON E~.t-19 20• ..O.nY4 v u..n9 nltl+t Fic. 2. Mortality ratios Yor lung cancer and stomach cancer in smoking habits. (Yrospcctivc 5tudy, 1y66-1981, Japan.) nonsmoking wives by huabund+' Nusu! Sinus Canc•er A significant risk elevation of cancer of para nasal sinuses in nonsmoking wives was observed according to the amount that husbands smoked, the SMRs being 1.00, 1.67, 2.02, and 2.55 when husbands were nonsmokers or smokers of 10- 14, 15-1y, or 20 or more cigarettes daily, respectively (P = 0.02482) (Tabla 7). No other risk factors studied were identitied as signilicantty altering the risk of nasal sinus cancer in women. CL. Bv age of nusnwnis Bv oCCuontton of nusoan;!•:• ,A ayf`f.lt~ff. / 1 n•105 J~ ' ,,,nawv~a ;~..~ 3 N0N X.I-t9 20• ~ naN x, i-~9 20• Bv nertoa of ooservatton• ]r 0 Ot 0 naN x.l-13 20• • i9e•Stanaarol:aa Fto. 3 Mortality ratios tur lung cancer in nonsmoking wivt!. by husbands' smoking II.,tlit5. (Pru- spctuva I)tuJy. 1y66-ly2ll. Japun.) ] S 2 5 '• gr 2 i ^ Sa-59 2 t 5 ~ TI BU 31680 th~~ no I

SYYIPUSIIIM: MEDICAL PEttSPECftVES ON PASSIVE SMOKING 681 Ob- -tc:d ,ted Ay _ of due ,ing ults .Nu~b ed. Cen ~ rds ldy vated risk of lung cancer (n ='_00), na,al sinus cancer (n =?8), brain tumors (/l = 34). and cancer of all sites (n ='_7O5). Lun,4• Cunrer A total of 429 deaths from lung cancer in women was recorded during the 16 y ears of lilllow-up ( 1966- lyti l). Of these deaths, 303 occurred among nonsmokers and 200 among 91,540 nonsmoking married women whose husbands' smoking habits were known. The standardized mortality ratios (SMRs) of lung cancer in nonsmoking women were 1.00, 1.36, 1.42. 1.58, and 1.91 when husbands were nonsmokers, ex- ,mukc;rs, or' daily smokers of 1- 14, 15- l9, or 20 or more cigarettes per day, respectively (one-tail P value = 0.00178) (Table I). A similar dose-response relationship was observecd by age and occupation of the husband (Table 2). This tendency is in Nharp contrast with that of stomach cancer, where no re- Wionship at all exists between the risk in nonsmoking wives and the amount of smuking by the husband (Tables 3 and 4, Figs. I and 2). Similur trends of lung cancer risk elevation in nonsmoking women with the increase in the extent ut'the husband's smoking were observed in each time period uf observation, in each age broup, both by age of husbands and by age of wives. in each occupational group, and in most areas under ubsc;rvatiun (internal con- ,istency) ( hig. 3) (7). No other characteristics of husbands or wives themselves were t'uund to elevate the risk of lung cancer in their nonsmoking partners (7) (Table 5). Nonsmoking husbands with smoking wives also showed an elevated risk of lung cancer, the SMRs being 1.00, 3.14, and 2.31 in nonsmoking wives, wives ,muking I-ly cigarettes, and wives smoking 20 or more cigarettes daily, respec- tively (P = 0.0177). This observation also strengthens the evidence listed above (Table 6). TABLE I LLNG CAN['h.R WliI'AIJTY IN WOMt:N BY AGE GkI)UP AND BY HU5NANUS' 5Ml1KING H.4Urr lPAC1ENr HI{k1I-.1.1 .l tNUNSMIIKb.kl" Husband'> >muking habit Husbnnd'. ,gc group Nunamukar Ex+muker I-14/day IS-19lday '-U. day Tuuil 40-49 9U_Sy M1-6v 70-74 Tutat 4 Itl is 5 37 6.229 7.791 7.120 755 :1.My5 1 I._55 3 1,922 II =,6257 2 348 17 6.212 B 8,621 20 9.668 28 7.243 2 612 58 .6.1i4 6 5,158 8 4.1.152 9 :,513 1 105 24 11.828 16 10.764 24 9,620 23 4.651 1 236 64 25.461 35 65 219 11 '-1111 32,t127 33.233 :4,:14 2.1146 91.54/1 I he wcightcd putnl I iN) 36 '. 113 1 ~ 1.42~?.ll l I. SIS ~ 38 ~ 1.91 ' 71 ~ Con mute ut ratr ratio . . U.tl5 tUl I v11 1.3a Manmi cxtenswn .nJ test-buxd 9014 :ht :.Y15 U n- . wnlitlenco linuts 45<1.t/2 1 One-tail .a C- , I t>,i P vulue II.I)U171S Nanlel-Haen.zcl chi 1.0855 1.81911 3.U?95 y1f-~. _ t)nc•tsd P value U.13Ny U.t1337 U.UUI2

682 1'. HIttAY:%MA TABLE = LUNG CANCI.k MUKTAL.ITY IN WUMEN ISY AGh CiRUUP. BY OCCUPA'I'IUN, ANU f3Y HUStiANU5' 5MUKIN(; HABI1 IPArIL•NT HEKShLF A NUNSMUKhk)° Husband's occupation Husbund's agc group Agricultural 40-49 worker 50-59 60-6y Total 70- Other 40-49 50-59 bU-69 70- Tot;tl The weighted point estimate of rate ratio and test-basetl 901%5 cUntldence limits MantCl-Haenazet chi OOe-t'Jlt P value ~ Prospective stutly, 1y66-1981, Japan. 1.786 3.053 0.03705 0.0UIIl TABLE 3 STIIMAIH CANCEK MUKrALIrY IN WOMEN UY AUt: GKUUp AmU HY HUSBANDS' SMUKING HAUIT (PATIFNI' HHNShLF A NUN5MUKhR)° HuabunJ'i age gruup Nonsmoker Examuker I= li/day 15-19/day :0+lduy Tutal = HuabanJ'+ uwking hubit 111-19 31 6.229 13 1.255 44 8.1121 :3 5.158 49 10.764 15a! 3:.u27 511-iy MI 7,791 14 1,922 M: 9,hb8 36 4,052 77 9,y_'u 269 33,?53 Ml-hy l' l 7,1_0 50 2.687 Illy 7.243 10 :.513 7ZS 1,b51 3`lt1 :1,:14 70-79 7 755 4 348 II 612 1 105 6 226 29 '_.W6 Tiu'rI 219 :1.ii95 1!0 6,_1: :4h :6,t31 100 11.lt'_tf 209 25.461 854 yt.54U 1'he welghted pulnt ~ U.93 l1.Mb ~ 11.8I Il.tl6 \duntel extemwn t 15.., 1';3 1 IN)Z1.17 1 IH),~ 1'3- 1 tll z 1. 19' eiunmte ut rutc ,..° chi -tl._7U rauu unJ tctt-buaad tlnr•iad - AM wninlcnee hmHs- - - - P ~a1uc Mantet-Hueneznl chi 1.09 -tl.tlt6 -0.U33 U.UYI t)nn•wil P value ' U.li4iftl 0.49162 U.l+f6hi U.16375 Hubbantl'> >muking habit Nonsmoker Exbmokrr I-I91duy :U+Iduy Total 1 2.502 6 5.941 9 3.636 16 1:,079 4 3,497 16 6,812 9 3.514 29 13.623 13 4,U84 33 0.845 10 _,t52 56 13,uzil 3 323 I 446 U 89 4 858 21 10.406 56 '_0,1Y14 28 9,391 1U5 39.841 3 3,7_7 9 9,093 7 7,1?8 l9 19,948 6 4.294 15 8,830 IS 6.306 36 19,430 5 3,036 15 5,59i3 13 _,499 33 11,133 2 432 4 619 1 137 7 1,I6tf 16 11.489 43 ,4,140 36 t6,U70 95 51,049 OU I 1 41 1'94 93 < 2.74 1 Mantel extenaiun . . < 1.0 . 1.35 chi 3.145 One-tail P value U.UUt183 0.39358 Hubbam.t v, upuu, Agrlcultu: wurker Tuta1 Other Total The we- ut r.Lt~ ylll. i .. Mantel- One-wt, ° Pru-

C 678 ADLKOFER, SCFIEEZEK, AND HELLEft uptake may be, the conclusion must he drawn that hydroxyprolinelcreatinine ratiu is nut a suitable parauneter to express urinary h~droyvproline excretion in ,mukers. We, therefore, reevaluated our data using body surface-standardized hydroxyproline excretion in day, night, and 24-hr urine samples. None of the investigated associations were significant except those between s4rum cotinine and hydroxyproline in night and 24-hr urine samples of cigarette smokers. We do not believe that this result gives any evidence of a cullagen- and elastin-degrading ettect of tobacco smoke in the smoker's lung, because there are at Ic;ast three other possibilities that must be considered: (a) this single result may have been obtained by chance; (b) nicotine may enhance the release of growth hormone (20). which seems to be related to an increased urinary hydroxyproline excretion (18); and (c) urinary hydroxyproline excretion may depend on urine volume, which is found to be higher in smukers. In our view, it is rather unlikely that any assumed damaging effect of smoking on lung collagen or elastin coultd be demonstrated by measuring hytlruxypruline excretion in urine. A large number of factors exert their influence on hydroxy- proline excretion in urine (1). Our finding of high inter- and intraindividual vari- atiuns support this view. The large amounts ot' urinary hydroxyproline originating from bone and skin collagen metabolism may prevent the detection of small changes in hydroxyproline derived from lung tissue. Lung collagen and elastin constitute only 1~Io ot'total body collagen and elastin (3). Fur this reason Wentlel and Rehpenning (19) du not recommend hydroxyproline measurements for the diagnosis of chronic bronchitis and emphysema. These authors did not tintf any difference in hydroxyproline excretion hetwet:n normal subjects and patients with chronic bronchitis and emphysema as long as the heart function of these patients was not det:umpen,ated. The same applies for urine desmu5ine, the crosslinking amino acid unique to elastin. Davies e1 cri. (6) found that urinr desmosine is unrelatcnd tu cigarette smuking, and attributed their result to high ba.al urine levels of this amino acid. This suggests that dc.musine also cannot be regarded as a useful indicator of lung tissue degradation. Altobether, our studies do nut titvur the idea that mt:astu-inis h}.Iruryproline excretion in urine is a suitable method for investigating c:arly effects ut" smokins, passive smoking, and air pollution in man. ACKNOWLEDGMENTS rhe authors are indebted to Protes,ur Hitunhi Kanugu and his co-workers at Tokai University tur determining ~;rCatinine and hydruxypruhnc in urine: to Dr. Gi,eia Hrnce and Mrs. Claire Hruby tur expert a1listance in preparing lhe tableS anlt ttSurel and the manU.%crtPl: and to Dr. Dietrich Huttmafln. Naylor Dana Institute. American Health Foundation, Vulhallu, New York, who enabted the study tu be pertormCd under blind conditions. REFERENCES I. Adams. E.. and Frank. L. Vletabulian ul ptuttne and thu hydroxyprutines.Airnu. Rcc Hr(,t Irevn. 49. IIN)5_ IUhI I IkJHU)_. 2. Atidrtu. L. H., Miller. D. C.. Slullunr.. R. A.. Ehrlit;h, S. P.. and Jones. 1. P. tpidcuuulugicst in~liviJual ~1uJy .tutly of coronary dieuaae risk facturti. I. Study tlesign and chatuctrriaUcs of ,uhjecu. .-Iinrr. J. L•pula,nial. 87, 73-16b r 1%2i). TI BU 31674 3 4 K IU It 13 14 V. ,

3 SYMI'USIUM: XtF-UICAL PF.RSI'I.l. 1'IVES ON PASSIVE SMOKING TAHLE 5 L<,NG CAN('iik iV1UHI U I I Y IY ;VUN5MUKING WUML'N: f2,\rlU t1Y SltLtiC"rEU KISK h':U"rl)Kti" 685 '.Hurtality raliu (relative risk) Chi-.yu:u e values Husban(1's churacteris tica Smoking (a) clg. +iJay) I.y1 y 11i Drinking I .U6 0.04 Population density: 6tN)-/-61)U 1.10 0.30 Wumen's charactcriaticb Uccupatiun: Agricu(ture/olhers 0.95 (1.17 Number uf children: U-3/4-y 1.09 U.-th Drinking: +/- 1.02 0.01 Meau I)ady/uthers 1.12 U.U`/ Green-yellow vegetable: Daily/others 0.88 0.93 Soybean p;a,te soup: Uaily/uthers I.t12S 0.29 ° Prospective study. 1964-191i1. Japun. Br(rin Tumurs The risk of brain tumor was also observed to increase with an increase in the c:xtent ut husbands' smoking habits, the risk for nunbmuking women being 1.00, 3.03, 6.25, and 4.32 when husbands were nonsmokers or smokers of 1-14, I5- 1(1, or 20 or more :l .tt-ettes daily, respectively (P = 0.00376) (Table 8). Cancer of All Sitcs In the case of cancers of all sites, a significant elevation inYisk was ubservetl, the SMRs fur nonsmoking women being 1.00, I.a'_', and 1.23 when hu.hancis were nunsmukers, ex-smokers, or ~,mukcrs of !-19 ur,2U or ntore cigarettes daily, TABLE 6 LOV(i C:1NCIiK M(1K1'ALITY IN N(/NSMUKIN(, HIJ:iHANl35 HY WIVF:S' SMULIN(i H,Xulr' Hu,ban(1'r .Ige group Nonsmoker itt-59 24 10,741 6U- 33 8.538 Wife's smoking habit Tulul 57 19.279 I he weighted point estimate 1.00 uf ratc ratio and test-ha,eJ `itP;r, wntidence limits 1.00 I -19/(lay j 20 + I(lay Total 1 321 t 1754 26 11._'46 3 276 2 229 38 y.U43 4 597 3 413 64 '_U._S'! ~ 1,~4 3.65 3 1 5.94 2 ~ - U.9ii . U.yt) Mantel extensiun - Jhl 1.989 ::5 < 't''2 Une-tai( 1.19 P value U.U?335 Munte!-Hatnstel chi _.1U46 Unc I.ul P v;tlue 0.0177 " Yruspecuve ,tu(ly. I`X+h-19tl1. Japan. TI BU 31681

3 SYMI'USIUM: XtF-UICAL PF.RSI'I.l. 1'IVES ON PASSIVE SMOKING TAHLE 5 L<,NG CAN('iik iV1UHI U I I Y IY ;VUN5MUKING WUML'N: f2,\rlU t1Y SltLtiC"rEU KISK h':U"rl)Kti" 685 '.Hurtality raliu (relative risk) Chi-.yu:u e values Husban(1's churacteris tica Smoking (a) clg. +iJay) I.y1 y 11i Drinking I .U6 0.04 Population density: 6tN)-/-61)U 1.10 0.30 Wumen's charactcriaticb Uccupatiun: Agricu(ture/olhers 0.95 (1.17 Number uf children: U-3/4-y 1.09 U.-th Drinking: +/- 1.02 0.01 Meau I)ady/uthers 1.12 U.U`/ Green-yellow vegetable: Daily/others 0.88 0.93 Soybean p;a,te soup: Uaily/uthers I.t12S 0.29 ° Prospective study. 1964-191i1. Japun. Br(rin Tumurs The risk of brain tumor was also observed to increase with an increase in the c:xtent ut husbands' smoking habits, the risk for nunbmuking women being 1.00, 3.03, 6.25, and 4.32 when husbands were nonsmokers or smokers of 1-14, I5- 1(1, or 20 or more :l .tt-ettes daily, respectively (P = 0.00376) (Table 8). Cancer of All Sitcs In the case of cancers of all sites, a significant elevation inYisk was ubservetl, the SMRs fur nonsmoking women being 1.00, I.a'_', and 1.23 when hu.hancis were nunsmukers, ex-smokers, or ~,mukcrs of !-19 ur,2U or ntore cigarettes daily, TABLE 6 LOV(i C:1NCIiK M(1K1'ALITY IN N(/NSMUKIN(, HIJ:iHANl35 HY WIVF:S' SMULIN(i H,Xulr' Hu,ban(1'r .Ige group Nonsmoker itt-59 24 10,741 6U- 33 8.538 Wife's smoking habit Tulul 57 19.279 I he weighted point estimate 1.00 uf ratc ratio and test-ha,eJ `itP;r, wntidence limits 1.00 I -19/(lay j 20 + I(lay Total 1 321 t 1754 26 11._'46 3 276 2 229 38 y.U43 4 597 3 413 64 '_U._S'! ~ 1,~4 3.65 3 1 5.94 2 ~ - U.9ii . U.yt) Mantel extensiun - Jhl 1.989 ::5 < 't''2 Une-tai( 1.19 P value U.U?335 Munte!-Hatnstel chi _.1U46 Unc I.ul P v;tlue 0.0177 " Yruspecuve ,tu(ly. I`X+h-19tl1. Japan. TI BU 31681

682 1'. HIttAY:%MA TABLE = LUNG CANCI.k MUKTAL.ITY IN WUMEN ISY AGh CiRUUP. BY OCCUPA'I'IUN, ANU f3Y HUStiANU5' 5MUKIN(; HABI1 IPArIL•NT HEKShLF A NUNSMUKhk)° Husband's occupation Husbund's agc group Agricultural 40-49 worker 50-59 60-6y Total 70- Other 40-49 50-59 bU-69 70- Tot;tl The weighted point estimate of rate ratio and test-basetl 901%5 cUntldence limits MantCl-Haenazet chi OOe-t'Jlt P value ~ Prospective stutly, 1y66-1981, Japan. 1.786 3.053 0.03705 0.0UIIl TABLE 3 STIIMAIH CANCEK MUKrALIrY IN WOMEN UY AUt: GKUUp AmU HY HUSBANDS' SMUKING HAUIT (PATIFNI' HHNShLF A NUN5MUKhR)° HuabunJ'i age gruup Nonsmoker Examuker I= li/day 15-19/day :0+lduy Tutal = HuabanJ'+ uwking hubit 111-19 31 6.229 13 1.255 44 8.1121 :3 5.158 49 10.764 15a! 3:.u27 511-iy MI 7,791 14 1,922 M: 9,hb8 36 4,052 77 9,y_'u 269 33,?53 Ml-hy l' l 7,1_0 50 2.687 Illy 7.243 10 :.513 7ZS 1,b51 3`lt1 :1,:14 70-79 7 755 4 348 II 612 1 105 6 226 29 '_.W6 Tiu'rI 219 :1.ii95 1!0 6,_1: :4h :6,t31 100 11.lt'_tf 209 25.461 854 yt.54U 1'he welghted pulnt ~ U.93 l1.Mb ~ 11.8I Il.tl6 \duntel extemwn t 15.., 1';3 1 IN)Z1.17 1 IH),~ 1'3- 1 tll z 1. 19' eiunmte ut rutc ,..° chi -tl._7U rauu unJ tctt-buaad tlnr•iad - AM wninlcnee hmHs- - - - P ~a1uc Mantet-Hueneznl chi 1.09 -tl.tlt6 -0.U33 U.UYI t)nn•wil P value ' U.li4iftl 0.49162 U.l+f6hi U.16375 Hubbantl'> >muking habit Nonsmoker Exbmokrr I-I91duy :U+Iduy Total 1 2.502 6 5.941 9 3.636 16 1:,079 4 3,497 16 6,812 9 3.514 29 13.623 13 4,U84 33 0.845 10 _,t52 56 13,uzil 3 323 I 446 U 89 4 858 21 10.406 56 '_0,1Y14 28 9,391 1U5 39.841 3 3,7_7 9 9,093 7 7,1?8 l9 19,948 6 4.294 15 8,830 IS 6.306 36 19,430 5 3,036 15 5,59i3 13 _,499 33 11,133 2 432 4 619 1 137 7 1,I6tf 16 11.489 43 ,4,140 36 t6,U70 95 51,049 OU I 1 41 1'94 93 < 2.74 1 Mantel extenaiun . . < 1.0 . 1.35 chi 3.145 One-tail P value U.UUt183 0.39358 Hubbam.t v, upuu, Agrlcultu: wurker Tuta1 Other Total The we- ut r.Lt~ ylll. i .. Mantel- One-wt, ° Pru-

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SM(it.lNG 689 hus- arci- ~~4 ). i the _Is uf ~.ults m to suits ,aive :'CU- clc:- IunS lUst; t tur ared t the 'v1Rs (fect .t ot risk r = Illc:r L: to ~. t! ) Uf and 44.3 in the subticqucut 6 years, respE:ctively (P = 0.00373). This phenomenon can be interpreted as the intluenct: of widespread exposure to passive smoking in Japan. As also emphasized in this earlier report, these observations strongly question the validity ot'thu conventional method of assessing the relative risk ut .leveloping lung cancer in smokers by comparing it with nonsmokers. This study shows that nonsmokers are detinitely not a homogenous group and should be subdivided according to the extent of previous exposure to indirect or passive smoking. The ubservatiun of the effect of passive and active smoking on lung cancer risk in men and women revealed a similar effect of both active anti passive smoking on lung cancer when nonsmokers without exposure to intrahousehuld passive smoking were used a, the unit risk group (Fig. 4). The observation ut an elevated risk of brain tumors in nonsmoking women with smoking husbands is of importance in considering the etiology of brain tumors lan area in which our current knowledge is quite limited), especially in relation to a similar report on the influence of pasaivt; smoking on childhood brain tu- mors (9). 9.0 8.0 7.0 6.0 5.0 4.0 3.0 STAXOANOIZEO 1MpTA{.{ rr RATIO 2.0 1.0 0 0XX SIIOKlNO XASIr N 0 N E 1•1l10• FEM.~E MA~E i0.10 5POU5E'5 S/OK1NG M~/IT 11..~/w (4f~ra r~W.n9J 40. OF OEAfN Il 5! 161 1 I10 611 II $SJ PePULAf10X JltlS 69NS ~60t0 lit{ lY.i9 1010 59077 1$6S6 O 1.01 YO t COXFIOEXCE ~P..• t.~~ ~ JE ~.if 1 !0 1 11 IXfEXVAI ~ c ISS i.ES 1 U l.» 5.16 4.H c~..r 1.10 I.I! LI! I.IS 1.11 S.~f ~..clw.n9 su.~wa w .. va.r.. Flti. 4. Active and pu.,rve ,muking .Ind lung cancer murtality~ Relative risks I IiR) with ytM7 :un- (iJcnce intervals. (Pruspectlvp JtuJy. I9M-lytSt. Japun.) TI BU 31685

688 T. H[IZAYAMA Histology ot''_l ca.,c, of lung cancer in nonsmoking wives with smoking hus- bands was not essentially different from that of smoking women (adc:nucarci- noma. 57.M squamous cell carcinoma. 19.0%; and small-cell carcinoma, The current results of elevated risk of nasal sinus cancer in addition to the risk of lung cancer must strengthen the plausibility of carcinogenic hazards ut ,idestream smoke inhalation through the nose, as they are in line with the results of measurements of various carcinogens in sidestream smoke showing them to be present in higher concentrations than in mainstream smoke (2. 3). These results are also compatible with known evidence showing a possible influence of passive smoking on health including elevation of carbuxyhemoglubin and nicutinc/co- tinine levels in saliva, blood, and urine after exposure tu passive smoking: ele- vation of hydroxyproline levels in urine (a marker of cullagen destruction in lung tissue); the presence of mutagens in urine (1), small airway dysfunction in those expused daily to passive smoking in the workplace ( I I): and risk elevation t'ur pneumonia, bronchitis, and asthma in chiCdr`en with smoking parent(s). When the effects of passive smoking due to husbands' smoking were compared with the effects of direct smoking in women, the results clearly indicated that the effect of passive smoking is less than une-titth that ut' direct smoking, the SMR, being 1.55 and 3.81, respectively. In terms of attributable risk, huwever, the ettect of passive smoking on lung cancer in women is nearly as important as that ul' direct smoking because the population of intrahousehold passive smokers at risk is fuur times greater (n = 69,645) than the population of active smokers ln = 17. jf,b). Therefore, althougli the relative risk of indirect smoking is much hmallcr than that of direct smoking, the absolute excess deaths from lung cancer due to passive smoking may be quite important because of the large size of the exposed group--especially in countries such as Japan where the majority (nCarly17U1~(,) of adult men smoke, but only a minority ( I5"; or less) of adult women smoke. Passive smoking.can be divided into (a) direct passive smoking (direct inhala- tion of sidestream smoke before being diluted by room air) and (b) indirect pas,ive smoking (inhalation of room air polluted by sidestream smoke) according to the extent of proximity effect, just as droplet infection is separated from droplet nuclei infection in acute respiratory communicable Jiseases; the effect of venti- latiun is of limited importance in the former case, although quite significant in the latter. Small room size and congested li.ving conditions in Japan (and possibly also in countries like Greece) are naturally more conducive to direct passive smoking. As described in a previous report, the age-adjusted mortality rates for lung cancer are increasing rapidly for bu(h mcn and women in Japan. As only a fraction of Japanese women with lung cancer smoke cigarettes, the reasons why their mortality from lung cancer nearly parallels that of men have been unclear. The current study attempts to explain at least a part of this long-,tanding riddle. Although the average rate of female smokers in Japan has remained fairly stable over the past 20 years, a statistically signiticant increase in the mortality rate fur lung cancer in nonsmoking women was observed in our long-term follow-up study of a large-size population. Mortality rates per lut),UUtI t'ur ages 50-59, 60-6y, and 70 and above were 7.1, 17.7, and 31.0 in tirst 10 years of follow-up and 9.9, 27. l, and 44. can be in Japa As a the vali lung C.1 nunsnt, accura ubserv. mc:n ar lung C' smukin The imukir (an arL tuabi murs t' F'u,. lidrn~c

PKIA I.NIIVL s11:Ul1'INV: 13. hX0-6yl) l I'1ti11 Cancer Mortality in Nonsmoking Women with Smoking Husbands Based on a Large-Scale Cohort Study in Japan' TAtcLsElt HIISa,YAMA Vuriunul Cunc•er Center Research !na'firlue, 1-! f.+uAiji 5-c'{lunte, C1rur,-ku, TuAyu IOd, Jupu,t Murtality ut 91,5411 nonsmoking wives was studied in relutiun to the ~,muklug habits of their husbands by means ul' a whurt study in Japan. During 16 years ut fulluw-up, :W deaths from lung cancer took place. The relative risks of lung cancer in these nonnmuking wives were I.M. 1.36, 1•a2, 1.58. and 1.91 when husbands were nunamukcrs, nx-,mukcrs. or daily smokers of 1-14, 15-19, or 20 or more cigarettes daily, respectively. Currespundiny relative risks I'ur stomach k:ancer were LW, 1.16. I.l1U. L011. and 1.01, respectively. Spec- dicity of aasuciation ;ind internal cunsiatencies were ubatrveJ. Among canccrs ut each ,ite, a vimdur tendency tuward risk elevation in nonsmoking wives with smoking hueb:,nds was observed li)r• nu.ul !.inus canccr, brain wnwrs, and cancer of all ~.itcs btlsR1Cs lung cancer. In interpreting these resultb, the ,lgntlicunce of proximity in exposure to Nidebtream smoke in J„Q.rn"e homes was Dtre»eJ. I rvNd Acadenu: Yra... In.. INl`RODUCTION The possible health hazard due to passive smoking was evaluated by the ub- bervatiun ut• murtality in nonsmoking wives with smoking husbands. As reported previously (6), nonsmoking wives ot• heavy smokers had a significantly elevated risk ot' lung cancer. Results of our large-scale cohort study repurtett here not unly confirm the rcsults of the rrcviuus report, but also reveal additional evidence of the health consequences of passive smoking by pointing out excess deaths due to cancer of other selected sites. MATERIAL$ AND METHODS A prospective cohort study on the health consequences of cigarette smoking has been in progress in Japan since the tail of 1965. In total, '_65,1IS adults ,( 1??,'-6I men and 14'-,1i57 women) ages 40 years anti above, 94.8% of the census population in the study area in 29 Health Center Districts in Japan, participated. They were interviewed from October I to December 31, 1965. and have been trackctl by establishing a record linkage System between the risk tactor records and death certificates. The 16-year follow-up results of this census-population-based cohort study were used as the materiah for the study. RESULTS In a large-scale cohort study carried out in Japan from 1966 tto IINI, nun- amuking wives with smoking husbands were found tu carry a significantly clc- ' Presented at the 5yntpu,/um •'.Lteclical Per.prctives on P.c,+rvt Smukinb," April '1-1_, 1984. Vl,c:nnJ, Austria. tNN1I-73ti/i4•i )3.1N1 t.,pyntihr i`Ma M:1-Jcmrc Pre... Inc. :\II nghh.d reprwl"uim in am wrrn rc.crvcJ. TI BU 31676 LUII L, A - yCar- antl : habit 1'h, werc: smul. rCSPL relal Th latiut smul. Sit incru of ul• ln e:. si.titc:, werQ (Tab. v, lunu smur tivel (Tab rul rhe ,~.

PKIA I.NIIVL s11:Ul1'INV: 13. hX0-6yl) l I'1ti11 Cancer Mortality in Nonsmoking Women with Smoking Husbands Based on a Large-Scale Cohort Study in Japan' TAtcLsElt HIISa,YAMA Vuriunul Cunc•er Center Research !na'firlue, 1-! f.+uAiji 5-c'{lunte, C1rur,-ku, TuAyu IOd, Jupu,t Murtality ut 91,5411 nonsmoking wives was studied in relutiun to the ~,muklug habits of their husbands by means ul' a whurt study in Japan. During 16 years ut fulluw-up, :W deaths from lung cancer took place. The relative risks of lung cancer in these nonnmuking wives were I.M. 1.36, 1•a2, 1.58. and 1.91 when husbands were nunamukcrs, nx-,mukcrs. or daily smokers of 1-14, 15-19, or 20 or more cigarettes daily, respectively. Currespundiny relative risks I'ur stomach k:ancer were LW, 1.16. I.l1U. L011. and 1.01, respectively. Spec- dicity of aasuciation ;ind internal cunsiatencies were ubatrveJ. Among canccrs ut each ,ite, a vimdur tendency tuward risk elevation in nonsmoking wives with smoking hueb:,nds was observed li)r• nu.ul !.inus canccr, brain wnwrs, and cancer of all ~.itcs btlsR1Cs lung cancer. In interpreting these resultb, the ,lgntlicunce of proximity in exposure to Nidebtream smoke in J„Q.rn"e homes was Dtre»eJ. I rvNd Acadenu: Yra... In.. INl`RODUCTION The possible health hazard due to passive smoking was evaluated by the ub- bervatiun ut• murtality in nonsmoking wives with smoking husbands. As reported previously (6), nonsmoking wives ot• heavy smokers had a significantly elevated risk ot' lung cancer. Results of our large-scale cohort study repurtett here not unly confirm the rcsults of the rrcviuus report, but also reveal additional evidence of the health consequences of passive smoking by pointing out excess deaths due to cancer of other selected sites. MATERIAL$ AND METHODS A prospective cohort study on the health consequences of cigarette smoking has been in progress in Japan since the tail of 1965. In total, '_65,1IS adults ,( 1??,'-6I men and 14'-,1i57 women) ages 40 years anti above, 94.8% of the census population in the study area in 29 Health Center Districts in Japan, participated. They were interviewed from October I to December 31, 1965. and have been trackctl by establishing a record linkage System between the risk tactor records and death certificates. The 16-year follow-up results of this census-population-based cohort study were used as the materiah for the study. RESULTS In a large-scale cohort study carried out in Japan from 1966 tto IINI, nun- amuking wives with smoking husbands were found tu carry a significantly clc- ' Presented at the 5yntpu,/um •'.Lteclical Per.prctives on P.c,+rvt Smukinb," April '1-1_, 1984. Vl,c:nnJ, Austria. tNN1I-73ti/i4•i )3.1N1 t.,pyntihr i`Ma M:1-Jcmrc Pre... Inc. :\II nghh.d reprwl"uim in am wrrn rc.crvcJ. TI BU 31676 LUII L, A - yCar- antl : habit 1'h, werc: smul. rCSPL relal Th latiut smul. Sit incru of ul• ln e:. si.titc:, werQ (Tab. v, lunu smur tivel (Tab rul rhe ,~.

ratio rn in litcti `' the ninc: C tsu d ing. hree Det:n ione :tiun ln1c, lirr tur nn, tu .n. JYM I't)SIUM: ,MGt)IC•;\I. f'IittSPGC')'IVES ON PASSIVLi Si<tOKING 679 3. Buddecke. E. tEd.l. "Grundriss der kiiuchcmie." pp. 471-474. De Gruyter. l3eriintNew Yurk, 1973. s4. Chun-Yrung, M.. Ferrcir.r. 1'.. 1-ruhlich, J., Schulzrr. M., and T:rn. F. The etl"t, ui age, smuking, and alwhul on routine lahurutury Ie.ta. Anrrr. J. (7rn. Pculrr,l. 75, 302-326 ( lyril l. S. Dales. L. G., Friedman. G. l).. Sieyelaub, A. B.. and Seltzer, C. C. Cig;uctte smoking and serum chemntrv tent,. J. Ch,•r,na- 1Jrs. 27, 293-307 t 19741. h. Davies. S. F., UtTun), K. R. tiruwn, M. G., Campe, H.. and Niewoehner, D. Urine Je~.musine is unrtlweJ tu ~rgarcua ,muking or to sprrumetric function. ,intrr. Kev. Respir. Dr.r. 122S, 473- 475 I 19M3). 7. Gulin. J.. Kark, J. U., Haltim, S. T.. Friecllander. Y., and Stein. Y. Smoking and its relation to anthrupumctric chur.rcteristiCh and biochemical variables in Jerusalctn 17-year-ulals and adults. 1s. J. .tlcrl. .1•, r. 18, 1'_33-I241 (1982). M. Hengen. N., uml I Icngen. M. Gnn-IiyuiJ chromatugraphic determination of nicotine and cotinine in plasma. Clrtr. 1 luvrr. 24, 511-53 (19781. v. Hu,lcy, H. F,. Olson. K. f3., Hurtun, J., .'vtichelsCn. l?, antl Atkins. R. Automated analysia of urinary hydroxyproline for cancer research. Tr•e•lurrcun Inr. Cu,ryr. 1, Il15-11U (1969). Il). Jones. C. R.. l3ergman, Ad. W.. Kittner. P. J.. and Pigman. W. W. Urinary hyJroxypruline cxcrc- tiun in nurm.rl children and aclulCscents. Pruc•. Suc•. Lrp. Sru(. Myd. 115, 85-M7 (1964). II. Kasuga. H.. Mutsuki. H., and Osaka. F. "Hydroxypruline and Pas!,ivt Smoking." Paper pre- sented at the 9th International Scientilic Mecting of 1.1...\., Edinburgh. 19ti1. 13. Kasuga, H., M:,t,uki. H.. Usuka, F., and Inone, M. The study on the relationship between urinury Ityclruxypruhne ;rnd creutininc ratio Irum the viewpoint of public health. Tukcu J. rr:rp• Sdc•ri. 4, 343-351 11Y7y). 13. Langnc». U. HyJruxyprulinau„chcidung und Kullagcnstult•wcchael. Utsch. ,LA,I Wuc'lrrrt.cchr. 50, :53u-'_53i ( (H71)1. 14. Lewu, 'f. R. Crtterra relevant tu an uccuputiunal health ,tandard for nitrogen ,liuxicle. in "Ni- trogen Oxides and Their k•:Itc~r, on I lealth" IS. D. Lee, Et1.1, pp. 361-375. Ann Arbor Science PuMtxhcn. Ann Arbor, Mich., 1980. 15. Mellstrurm. 1).. Kun.lgrrn. A.. Jagenburg, R.. Steen. B.. and Svanborg. A. 11,b.,"u smoking. ;,ging, and health among the Clclerly:,a longitudinal pupulation .tudy of 7Q-year-uIJ men and an age arhurt comparison. A,ce .i,4errr,e 11, 45-5ts t 1Nii.2). (6. Sibernugl. S. Renal transport of amino acids. l;lur. W,u lrtvrsrlrr. 57, 1U0H-11119 (1979). 17. Tcchnicun :\utuunulycer 11: Clinical Method No. Sk:'2-W1I hC1. IN. Wharton. B. A., Brown. G., Rayner. P. H. W.. Il,rwrlls, G., and Pennuck, C. A. Urintlry hy- Jroxyprulinr in children with growth hurmone tJeliciency. Are~lr. Urs. CIu1J. 49, 159- 162 1 IN74/. 19. Wendcl, H.. and lirhprnning. W. Untersuchungen zur HydruxyprulinausachctJung im Harn bci chrunis,:hrr ob.truktiver lirunch,u,. L. ,r;esu,tue uur. Slecl. CJr,vr;,4eb. 36. 247-249 (t'1rit1. '_u. Wilkins. J. N.. C:ulurn. H. E.. van Vunukis. H., Hill. M. A.. Gritz. E., and Jarvik. M. E. Nic- otine trum ;rg.,ratte -smuking incrtaaes circulating levels of curti,ul. growth hormone, anJ prulucun in male chronic srnukrrs. i'%rckuplrurnm, odr,gy 78. 3U5-3UB 11'1tS_). =1. Yanagi+awa, Y., Ni,hrnwr.r, H., Matauki. H.. Osaka. F., Kaauga, H. "txposure-tflCct relation tur NU, with pcr•oma rxposure and urinary hydroxypruline to creauninC ratio ut schoolchil- dren as tndrc.,tur,." I'.iper presented at the 76th :\nnual Meeting of the Air Pollution Control Aiauciatiun. Atlanta. CieUrbta, June 19-23, 19143. TI BU 31675

684 T. HIKAYAMA ,MartatiCY ~ natl0 i yusouna'> U<a ltrouu • 0-y9 5C - 59 60 ~ Can<f. .I n< I 1 ~ <.nce. 0 I 1. C t i 0 1 0 1 0 1 s I I I I I` a~nON i~t•t9 :0• 0 vON t.l-t9 ;0• ~so.nas +n..•~nq n.aa n.ao.nos +ro.~ny naa~t j n 1 nON E~.t-19 20• ..O.nY4 v u..n9 nltl+t Fic. 2. Mortality ratios Yor lung cancer and stomach cancer in smoking habits. (Yrospcctivc 5tudy, 1y66-1981, Japan.) nonsmoking wives by huabund+' Nusu! Sinus Canc•er A significant risk elevation of cancer of para nasal sinuses in nonsmoking wives was observed according to the amount that husbands smoked, the SMRs being 1.00, 1.67, 2.02, and 2.55 when husbands were nonsmokers or smokers of 10- 14, 15-1y, or 20 or more cigarettes daily, respectively (P = 0.02482) (Tabla 7). No other risk factors studied were identitied as signilicantty altering the risk of nasal sinus cancer in women. CL. Bv age of nusnwnis Bv oCCuontton of nusoan;!•:• ,A ayf`f.lt~ff. / 1 n•105 J~ ' ,,,nawv~a ;~..~ 3 N0N X.I-t9 20• ~ naN x, i-~9 20• Bv nertoa of ooservatton• ]r 0 Ot 0 naN x.l-13 20• • i9e•Stanaarol:aa Fto. 3 Mortality ratios tur lung cancer in nonsmoking wivt!. by husbands' smoking II.,tlit5. (Pru- spctuva I)tuJy. 1y66-ly2ll. Japun.) ] S 2 5 '• gr 2 i ^ Sa-59 2 t 5 ~ TI BU 31680 th~~ no I

SYytPt1SIUM: MEDICAL PE)tSPECfIVES ON PAS1IVE SMOKING 687 CAHLE 9 ur :U.l. SICI:J CANIt.K Ml1KIAt.1lY IN WUMI:N IiY AGt~ GKUUP. I!Y OCCCLP:\IIUN, ANU IflHU."HANUS' SMl1KlN4 HAnrl (PAIIkNtHIiKSlLI- A V/)NJAIUKF.K1" HubbunJ', ,muk)ng habit .3.; FIu,band', Hush uccupatiun )ge g anJ', roup Nunemuker l:.cantuker I-l9/day :0-tday Tutul ~ .~ Abrncultur;rl 4u -4) 40 '_.5U: 119 5,941 76 3.636 235 1?,U79 .u wurker 50 -59 96 3,497 201 6.612 115 3,514 412 l3,iS?3 60 -69 '_U5 4.084 373 6,845 127 '_,l5? 7(15 13.uKI R 7U - 17 323 22 446 5 89 44 rs 3 Tutal 358 IU,4U6 715 :0,U44 323 9.391 1.396 31/ X41 Q12. t)ther 4U -4y 414 3,727 118 9,093 103 7,1'_k9 :6y ly,y4NS 5U -59 79 4.294 :4is 8,810 I6y 6.306 4ye 19.410 hU -6y 132 3,036 _39 5.598 129 2.499 RU II,133 7U - 17 432 :1 619 6 137 44 1,188 7i)lal 276 11,489 h:h ?4.14U 4117 16.070 1.3(ly 51,699 The weighteJ point I.UU 1.12 < I'~I I U3 1.23 < I.3~ t l Mantel txten.iun ehumute u( rate rati u . . _ hc ,al and test-bu,cd 1NI°;v cuntidencc lunn, chi 3.540 One-tail P vutue U.INN)?U or Mantel-Haenacel chi 2.232 3.628 Ju t)ne•tail P value U.UI'_kSl 0.(NN)14 hc: U- ll- " Pru,pective Audy, I`166-lyiil. Japan. -h kemia, the direction of this trend being evenly distributed to both the plus-side (risk increases with the extent of husband's smoking habit) and the minus-side (risk decreases with the extent of husband's smoking habit), DISCUSSION This study cuntirms the correlation between lung cancer and spousal smqking reported previously. The correlation is quite specific in terms of diseases. For instance, no risk elevation at all was observed for,stomach cancer. A striking internal consistency of association was also observed. The results were essen- tially similar when observed in terms of age of husbands, age of wives, occupation of hushantl,. and dit'ferint: periods of observation. The results are in line with a Greek study by Trichopoulos :utd others (Il)) and a U.S. study by Correa and others (4) (external consistency), although they are slightly at variance with an American Cancer Society study in the United States (5) and a case-cuntrol study cuntlu,~tcal by Kabat and Wynder tsl. lliFfcrencta tn proximity between husband and wife in daily liFe. room size, room ventilation. and frequency of wives who work in offices in these countries are potentially influential factors in enhancing the extent of risk posed by hus- bands' smoking. TI BU 31683

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SM(it.lNG 689 hus- arci- ~~4 ). i the _Is uf ~.ults m to suits ,aive :'CU- clc:- IunS lUst; t tur ared t the 'v1Rs (fect .t ot risk r = Illc:r L: to ~. t! ) Uf and 44.3 in the subticqucut 6 years, respE:ctively (P = 0.00373). This phenomenon can be interpreted as the intluenct: of widespread exposure to passive smoking in Japan. As also emphasized in this earlier report, these observations strongly question the validity ot'thu conventional method of assessing the relative risk ut .leveloping lung cancer in smokers by comparing it with nonsmokers. This study shows that nonsmokers are detinitely not a homogenous group and should be subdivided according to the extent of previous exposure to indirect or passive smoking. The ubservatiun of the effect of passive and active smoking on lung cancer risk in men and women revealed a similar effect of both active anti passive smoking on lung cancer when nonsmokers without exposure to intrahousehuld passive smoking were used a, the unit risk group (Fig. 4). The observation ut an elevated risk of brain tumors in nonsmoking women with smoking husbands is of importance in considering the etiology of brain tumors lan area in which our current knowledge is quite limited), especially in relation to a similar report on the influence of pasaivt; smoking on childhood brain tu- mors (9). 9.0 8.0 7.0 6.0 5.0 4.0 3.0 STAXOANOIZEO 1MpTA{.{ rr RATIO 2.0 1.0 0 0XX SIIOKlNO XASIr N 0 N E 1•1l10• FEM.~E MA~E i0.10 5POU5E'5 S/OK1NG M~/IT 11..~/w (4f~ra r~W.n9J 40. OF OEAfN Il 5! 161 1 I10 611 II $SJ PePULAf10X JltlS 69NS ~60t0 lit{ lY.i9 1010 59077 1$6S6 O 1.01 YO t COXFIOEXCE ~P..• t.~~ ~ JE ~.if 1 !0 1 11 IXfEXVAI ~ c ISS i.ES 1 U l.» 5.16 4.H c~..r 1.10 I.I! LI! I.IS 1.11 S.~f ~..clw.n9 su.~wa w .. va.r.. Flti. 4. Active and pu.,rve ,muking .Ind lung cancer murtality~ Relative risks I IiR) with ytM7 :un- (iJcnce intervals. (Pruspectlvp JtuJy. I9M-lytSt. Japun.) TI BU 31685

,Kwa3 SYMPOSIUM: !AE:DICAL PERSPECTIVES ON PASSIVE SMOKING 683 TABLE 4 SR1AfACH CANCEN MORTALI'rY IN WUMI:N HY AGL: GHUUP. HY OCCUPAr1UN. ANU HY Itt:SHANUS' SM(/KING HAHrt (P.111L•N'r HL:H1tr1.1.i, NUNSMUKI:H1" Husband's hmuking hubit Huabantl's Husband's Exsmuker uccupuuun age group Nonsmoker 1-19/tiay 2t)tiJay Tutal Agrtcultural 40-49 13 2.502 41 5,941 25 3,636 79 I.,U79 worker SU-59 37 3.497 56 6,812 37 3,514 130 13,823 6U-6y 77 4,U2i4 116 6,845 43 ?,153 236 13,081 70- 3 323 13 446 3 ts9 19 858 l'atal 13U IU:-tU6 226 _ll•U41 IUM 9,341 464 39.841 Other 40-49 Itf 3,7?7 38 9,093 23 7,1_8 79 19.948 50-59 23 4..94 76 9,830 44) 6,3U6 139 19,430 60-69 44 3.036 83 5,598 35 2.499 162 11,133 70- 4 432 3 619 3 137 10 1,188 TutUl 89 11,489 ?0U 24.140 10 1 16.U7U 390 51,699 , ~ y Mantel extension The weighted point estimate I.W 1.03 < ~:Hy 1.05 < :~ chi 0.234 of rate ratio and teat-bascJ vth/, cuntiUent:e limlts OnC-tail P value .40749 1 Muntel-Haanszel chi - 0.298 0.486 Une-tall P value 0.38285 U.3134>f Pru7pective 3tuJy, 1966-1981. Japan. .IGN...Y ..NCtR ~(: a j541 ir 1 ]3 r G 5 Gelatlve NISK 3y3jtt 0. ,..O.nJ. I-u,. I. RcLttive n,k, ui lung cuncer and ,tunr.lch cancer In'11.54U nonsmoking wives by hu.bunJs' ,luuking habit. iNru,rective 1tuJy. I'Nlh-lyrit. Japun.t TI BU 31679

686 Hu.banJ', a6e group TABLE 7 SINUS CANCL'K MUKT.aLITY IN W11Mt::V U1At,t: GKUUP:Wt) Hl' HUSHANUS' SMOKING HaHir (PAllL-N1' HI:KSL:LP .i :V/)NSMt)KL•K1° 10-49 51/-59 (10-119 70-79 Tutal Iho wctghted point eshm•ate ul iute rattu and 1C]t-basCd 4iho cunlldence limits M:,ntcl-Hatmszt:l chi Clne-lcul P valuc 1'. HIRAYAMA ALL Huabund', ,muktna_ habit Nonsmoker Ex-'muker I-I-uday 15-IVrday ?U+,Jay Tucll 0 6.229 l/ 1.255 1 8,h21 1 5.138 10,764 4 1..U27 Hunb.. uc';up. 1 7,791 0 1,92. 3 9.668 1 4.02 2 9.92ll 7 31,251 4 7 120 0 2,60 5 7 243 2 2,513 6 4 651 17 24.214 u , 755 0 34s , 0 612 0 105 . 0 0 1.u46 Ab'1'I~.:. 5 :1,595 1) 6.212 9 :6,144 4 11.818 IU _'5.-i61 _'tl 91,51U wut: LINI - 1.67< 4-2U , - U.<6.33 ,.55< 6.]7 - U.67 U.6•t Ltl4 StunlCl C[Ienmun 'hl I.`1h3 Tul.. .916 .UI? .713 OnC-tLLll P value U.llall? UthCr - - U.1783 1L1557T U.114336 Nwr. In wmputation, ages MI-69 and 70-79 wcre combined. ' PrUSpecttve stutly, 1966-1981. Japan. respectively (P = 0.00020) (Table 9). This risk elevation is influenced by the elevated risk of lung cancer and cancers of other sclected sites such as nasal sinus cancer, brain tumor, and possibly also breast cancer. Risk elevation fur cancer of all sites becomes nonsienit7cant when these cancers are excluded. No signiticant association was ubstrved with other cancers such as those of the mouth, pharynx, esuphagus, stuhtat:h. colon, rqctum, liver, pancreas, perito- neum, cervix, ovary, urinary bladder, skin, bone, malignant lymphoma, or leu- TABLE 8 BRAIN TUMUK \'tUKI'ALI"rY IN WUfd1iN HY AGF. GKl)UP ANU HY HuSHANL)S' SMUKI.-4U H.aHlr lP.xrttsul' HEKSLLr'.i NuNSntuKLxl•• H usbanJ's uge group Husbund'S smoking habit Nonsmoker Ex-smukcr I-1-I/duy 15-191J•ay _U+•Jay Tutal The welghtcJ point estimate ul rate rauu and test-buaed YIY-i ;unliJCnCu 11t111ts Mantcl-Hacnsua chi Onc-uul P valuc 1.00 3.1/3<x:5~ I9..lll i.3'-<I,~.i3 .In Une-t~d P value U.Ix137h 1.7;6 :.h5h ?.317 0.03934 11.1K1345 I I.II I U:5 .Vu(r. In wmputauun, agc. 6U-h9 anJ 7(1-79 were cumbined. ' Pruspecttvc tuJy, INM,1`/151. Japan. U 6.:=N (1 I.-'S5 1 7.791 0 1.922 1 7,12U 0 2.687 I 755 1) 348 1 8.621 6 5.1511 4 10.7fv3 II 32,U7 19.6hd 3 1.052 4 9.820 12 33.253 5 7.243 11 :,513 4 4.h51 11) '_4._'I4 0 612 U ILIS 0 226 1 2.U46 Tot.. stant (3nc-e kern. (risk (riSk Z*1 repuo inst~: intet tl'all`• Ot ht GreL Utht: Am. ~UI1~ L) ruut al'C ban•

686 Hu.banJ', a6e group TABLE 7 SINUS CANCL'K MUKT.aLITY IN W11Mt::V U1At,t: GKUUP:Wt) Hl' HUSHANUS' SMOKING HaHir (PAllL-N1' HI:KSL:LP .i :V/)NSMt)KL•K1° 10-49 51/-59 (10-119 70-79 Tutal Iho wctghted point eshm•ate ul iute rattu and 1C]t-basCd 4iho cunlldence limits M:,ntcl-Hatmszt:l chi Clne-lcul P valuc 1'. HIRAYAMA ALL Huabund', ,muktna_ habit Nonsmoker Ex-'muker I-I-uday 15-IVrday ?U+,Jay Tucll 0 6.229 l/ 1.255 1 8,h21 1 5.138 10,764 4 1..U27 Hunb.. uc';up. 1 7,791 0 1,92. 3 9.668 1 4.02 2 9.92ll 7 31,251 4 7 120 0 2,60 5 7 243 2 2,513 6 4 651 17 24.214 u , 755 0 34s , 0 612 0 105 . 0 0 1.u46 Ab'1'I~.:. 5 :1,595 1) 6.212 9 :6,144 4 11.818 IU _'5.-i61 _'tl 91,51U wut: LINI - 1.67< 4-2U , - U.<6.33 ,.55< 6.]7 - U.67 U.6•t Ltl4 StunlCl C[Ienmun 'hl I.`1h3 Tul.. .916 .UI? .713 OnC-tLLll P value U.llall? UthCr - - U.1783 1L1557T U.114336 Nwr. In wmputation, ages MI-69 and 70-79 wcre combined. ' PrUSpecttve stutly, 1966-1981. Japan. respectively (P = 0.00020) (Table 9). This risk elevation is influenced by the elevated risk of lung cancer and cancers of other sclected sites such as nasal sinus cancer, brain tumor, and possibly also breast cancer. Risk elevation fur cancer of all sites becomes nonsienit7cant when these cancers are excluded. No signiticant association was ubstrved with other cancers such as those of the mouth, pharynx, esuphagus, stuhtat:h. colon, rqctum, liver, pancreas, perito- neum, cervix, ovary, urinary bladder, skin, bone, malignant lymphoma, or leu- TABLE 8 BRAIN TUMUK \'tUKI'ALI"rY IN WUfd1iN HY AGF. GKl)UP ANU HY HuSHANL)S' SMUKI.-4U H.aHlr lP.xrttsul' HEKSLLr'.i NuNSntuKLxl•• H usbanJ's uge group Husbund'S smoking habit Nonsmoker Ex-smukcr I-1-I/duy 15-191J•ay _U+•Jay Tutal The welghtcJ point estimate ul rate rauu and test-buaed YIY-i ;unliJCnCu 11t111ts Mantcl-Hacnsua chi Onc-uul P valuc 1.00 3.1/3<x:5~ I9..lll i.3'-<I,~.i3 .In Une-t~d P value U.Ix137h 1.7;6 :.h5h ?.317 0.03934 11.1K1345 I I.II I U:5 .Vu(r. In wmputauun, agc. 6U-h9 anJ 7(1-79 were cumbined. ' Pruspecttvc tuJy, INM,1`/151. Japan. U 6.:=N (1 I.-'S5 1 7.791 0 1.922 1 7,12U 0 2.687 I 755 1) 348 1 8.621 6 5.1511 4 10.7fv3 II 32,U7 19.6hd 3 1.052 4 9.820 12 33.253 5 7.243 11 :,513 4 4.h51 11) '_4._'I4 0 612 U ILIS 0 226 1 2.U46 Tot.. stant (3nc-e kern. (risk (riSk Z*1 repuo inst~: intet tl'all`• Ot ht GreL Utht: Am. ~UI1~ L) ruut al'C ban•

SYMPOSIUM: MEDICAL. PERSPECTIVES ON P;tSSIVE SMOKING 669 75 pa- aunary bject'!, lly for ±51 are mpair- erwns differ 10 ci.g- -Zttect autiun obub{y exLt:nt ,ubject 12. Hugotl. C.. Hawkins. L. H.. and Astrup. P. Exposure of passive smokers to tobacco smoke cunstitwents. IrH. Arclt. Occup. Etrvirun. Heulrh 42. -'1 -'-9 i 1978). 13. Hugud, C. Indoor air pollution with smoke constituents-An experimental investigation. Prev. .Yled. 13, 5!f'_-jtitl (1984). 14. Kaultmunn, F.. Tessier. J. F., and (7rtol, P. Adult passive smoking in the home environment: AA risk tactur tur chrunic airtluw limtlauun. .9+trer. J. EprJiul. 117, 26N-2t;li (1983). 15. Klusterkutter, W., and Gunu. E. Zum Problem des Passivrauchens. Lrwru(bl. Bukteriul. Hr,4. I. .-Ihr. (Iri,p. B 162, 51-6y (1976). 16. Knudson. R. J.. Burrows, B., and Lebuwitz. M. D. The maximal expiratory tlow-vl,lume curve: Its use in the detection of ventilatuiy abnormalities in a population study..line•r. l;rr. Rr.+pir. 1)a. 114, tS71-M7y 11976). 11 Knudson. R. J.. Slutin. R. C.. Lebuwitz. M. D.. Burrow. B. The maximal expirutory tluw-vulume curve. Normal standurds, variability and ettectS uf age. Amrr. Rev. Respir. Dis. 113, 5tf7-60(1 119761. 18. Lebuwitz. M. L).. Armet. D. B., and Knudson. R. The effect of pa»ive smoking on pulmonary tunctiun in children. i:)n•irwl. lnternut. 8, 371-373 (19tS2). 19. Lcbuwitz. M. D. "Intluence uP Passive Smoking on the Pulmonary Function-A Survey." Un- published d•ata. ~s or iR I :0. McFndden, E. R., and Linden. D. A. A reduction in maximum mid-expiratory flow rate. Amer. 7 7 7 2). (17 t. I J. Lled. 52, 7:5- 3 21. McFadden. E. R.. Kiker. R.. Holmes. B., and Degroot. W. J. Small airway disease. Amer. l. ,Dtrd. 57, 171-1 ii _' 11974). ?1. McFadden. E. 12._ and lneram. R. H. Periohentl airwav ob.truction_ J. Amer. Wed. A.ssuc. 235. : y-_nu t i7ini. """°`ff. i :3 N d Johnwn L C S irumetric standards f A r h lth i J h K l k p un u y nur,mu .......,... orr s l~. u ea ing ``j" ` a A17 (1y71) Di 103 57 l i l A R R - inrr rv rsp r s s.....,. .u u t a. I`/ICwUCnnCC. U. 6.. 111cInCTman. J.. anu 1cIGC, U. o. rllnuHVllc lnilnRCN In mC PCRPnCfUI airways •u• E l 1 bled lU 75j_7~rs I t)7~1 V """r"' t d' n i r tte k r l .,...,. , y,ru g c ga smu e s : ni e :3. Peuy, f. L.. Silvers. G. W.. Stanfurd, R. E.. Baird. M. A., and Mitchell. R. S. Small airway pathology is related to increased closing capacity anll abnormal slope of Phase 111 in excised Eur. 1_ i human lungs. Antrr. Rev. Respir. Uis. 121, tJy-~?o (19140). :A. Quanjer, P. H., (Ed.). "Standurdized Lung Function Tc,ung." European Community for Coal .t,rwav, I and Stecl. Luxembourg. July 1983. 9, 263- )! =t ~cmutng, K. J. r.. LClal. n. u.. rrnuw, 3. L.. 6CCA, u. j., acnuenoerg, j. n., anu ouunuys. n. Lung tuncttun. resptratury uw]ease anu SmuKmg in tamttles. Amer. J. tpolrnuut. 11.16, Sti-L83 t 19771 . - ,-_ a 2M. Schulz, V. Small airways disease. Prar. Pneumuf. 34, 187-218 11'Ii{0). V d 1 (i 13i M 1t l R t li 2y S r i h J i d F E 1 J . t,cng . up,. . ., ;tn ra ory c seaae rates an rt;t,tt,r ~ p r~er. rrr +. er. . esp .. .. pulnwnary function in children associated with NU, expuaure. Antrr. Rrv. Respir. Uis. 121, -r rtr_ n 3-Il)t192S0/. )11. Tager. I. B., WCis]. S. T., Rosner. B.. and Speizer. F. E. Effect of pJrental cigarette ~mUking on 1y791 hdd J id i l 15 6 h f i f t E 110 ' ren. ~ mer. . rnr u . unet on o c p , - _ , e pulmonary 1 nri,torv  t 31. fhurlbeck. W. M. Smoking, airtlua limitatton, and the pulmonary circulation. Amer. Rrv. Rrspir. :,,,.,. „F g Urs. 122. 12t3-ItS611Htf01. 3_ 143- >f t_, vutua. L. t_ungenr;rcosnstt:u unu ntasrvraucnen: VuanrntauvC voortegungen. `enrnuor. aur;ae- riul. Hs).'. l. Abi. Oriy. 8 177, 90-95 (19tl3). ,ckfeta_ g Ji. Weiss. S. T.. Tager, 1. B.. Speizer. F. E., and Rosner. B. Persistent wheeze: lts rclatiunship to d I l l i k t t i l i ill eve pu munury „uun samp ing en u unct un in a pupu e ratury ness. cigarette ,mu , tc,t, r re,p ut children. A»rer. Ra v. Respir. U6. 122, 697-707 (1980). ,)f nro- i t;. Weis.. S. 1.. Tager. 1. B.. Schcnker. M.. and Speizer, t. The health effects ul involuntary i t')}nll. li mrtuning. ninrr. rtrr. ice.,plr. tn.l. l4o, »J-r.: t 17wr. Chr.lt ~ iS White. J. R.. and Froeb. H. F. Small-airways, dysfunction in nonsmokers chruntcaly espu~ed to tobacco smoke. ,Veu, Ent;l. J. Nied. 302. 7'_lt-7?3 1197SU). l i J a f W W ntrr. ys h. . . . H.. and Wullis. . A. Use o ranka in one-criterion variance ana mrnts,l a w+ Kruskal. .5 tw t., t. A., t. 47. 5 tS 3-627 11952). TI BU 31665

SYytPt1SIUM: MEDICAL PE)tSPECfIVES ON PAS1IVE SMOKING 687 CAHLE 9 ur :U.l. SICI:J CANIt.K Ml1KIAt.1lY IN WUMI:N IiY AGt~ GKUUP. I!Y OCCCLP:\IIUN, ANU IflHU."HANUS' SMl1KlN4 HAnrl (PAIIkNtHIiKSlLI- A V/)NJAIUKF.K1" HubbunJ', ,muk)ng habit .3.; FIu,band', Hush uccupatiun )ge g anJ', roup Nunemuker l:.cantuker I-l9/day :0-tday Tutul ~ .~ Abrncultur;rl 4u -4) 40 '_.5U: 119 5,941 76 3.636 235 1?,U79 .u wurker 50 -59 96 3,497 201 6.612 115 3,514 412 l3,iS?3 60 -69 '_U5 4.084 373 6,845 127 '_,l5? 7(15 13.uKI R 7U - 17 323 22 446 5 89 44 rs 3 Tutal 358 IU,4U6 715 :0,U44 323 9.391 1.396 31/ X41 Q12. t)ther 4U -4y 414 3,727 118 9,093 103 7,1'_k9 :6y ly,y4NS 5U -59 79 4.294 :4is 8,810 I6y 6.306 4ye 19.410 hU -6y 132 3,036 _39 5.598 129 2.499 RU II,133 7U - 17 432 :1 619 6 137 44 1,188 7i)lal 276 11,489 h:h ?4.14U 4117 16.070 1.3(ly 51,699 The weighteJ point I.UU 1.12 < I'~I I U3 1.23 < I.3~ t l Mantel txten.iun ehumute u( rate rati u . . _ hc ,al and test-bu,cd 1NI°;v cuntidencc lunn, chi 3.540 One-tail P vutue U.INN)?U or Mantel-Haenacel chi 2.232 3.628 Ju t)ne•tail P value U.UI'_kSl 0.(NN)14 hc: U- ll- " Pru,pective Audy, I`166-lyiil. Japan. -h kemia, the direction of this trend being evenly distributed to both the plus-side (risk increases with the extent of husband's smoking habit) and the minus-side (risk decreases with the extent of husband's smoking habit), DISCUSSION This study cuntirms the correlation between lung cancer and spousal smqking reported previously. The correlation is quite specific in terms of diseases. For instance, no risk elevation at all was observed for,stomach cancer. A striking internal consistency of association was also observed. The results were essen- tially similar when observed in terms of age of husbands, age of wives, occupation of hushantl,. and dit'ferint: periods of observation. The results are in line with a Greek study by Trichopoulos :utd others (Il)) and a U.S. study by Correa and others (4) (external consistency), although they are slightly at variance with an American Cancer Society study in the United States (5) and a case-cuntrol study cuntlu,~tcal by Kabat and Wynder tsl. lliFfcrencta tn proximity between husband and wife in daily liFe. room size, room ventilation. and frequency of wives who work in offices in these countries are potentially influential factors in enhancing the extent of risk posed by hus- bands' smoking. TI BU 31683

690 r. HIRAYAMA The results of the present study must be etfectively utilized in planning pro- grams for thc control uf lung cancer and other selected cliseasc,, . The results clearly indicate that lung cancer, especially in women, can only be controlled satisfactorily when proper measures are taken against passive smoking as well as against active smoking, especially in countries like Japan. A similar statement may also be valid for cancers ut' other selected sites. REFERENCES 1. Bos, R. P., Theuws, J. K. G., and Henderson, P. Excretion of mtuagens in human urine after passive bmuking. Cancer Lett. 19, 85-90 (1983). 2. Brunnemann, K. D., Adams,1. D., Hu, D. ('. S., er u(. The intluence of tobucco smoke on indoor atmospheres. 11. Volatile and tobacco specitic nitrosamines in main• and si~lestreucn Smokc and their cuntribution to indoor pollutiun, in "Proceedings of the ~th 7uint Conference on the Jensing ut Environmental Yollutants. New Orleans 1977", pp. 876-880. American Chemical Sucirty, Wushingtun, D.C.. 1978. 3. Brunnemann, K. D., and Hot'fmann, D. Chemical ,tudies on tobacco smoke L1X. Analysis of volatile nitrusumines in tubuccu smoke and polluted indoor envirunments, in "EnvironmCntal Aspects of N-nitroso Cumpounds" (E. A. Walter, L. Griciutc .,nd M. Gustegnaro, Eds.). pp. 343- 156. [ARC Scientific Publications No. 19, WHU, I},,n, 1978. 4. Correa, N.. Pickle. L. W., Funthum. E.. Lin, Y., and H.+cnseel, W. Passive smoking and lung cancer. Lunret 1, 595-597 (1983). 5. Gartinkcl. L. Time trends in lung cancer mortality among non-smukers and a note on passive ,muking. J. Nut(. Cancer /nsY. 66, 1116 1-1 066 (lytl I). 6. Hirayama, T. Non-smoking wives of heavy emukcrs h;,ve a higher risk of lung cancer: A~tualy in Japan. Bret. .bled J. 282, Iri3-lH5 (19tf1). 7. Hirayuma, T. Passive smoking and lung cancer: Consistency of aswciatiun. Lancet 1, 1425-1a:6 (1983). 8. Kabut, G. C., and Wynder. E. L. Lung cancer in nonsmokers. Ccuncer 53, 1214-1221 11984). 9. Preston-Martin, S.. Yu, M. C.. lienton, B., and Henderson, B. E. N-Nitroso cumpuunda and childhood brain tumors: A caae-cuntrol study, Cancer Rcs. 42. 524t1-5245 11982). I(1. T'richupOulos, D.. Knlandidi. A., and Sparrus. L. Lung cancer and passive smoking; conclusion ul Greek study. Lcuu•rt 1. 677-678 119tl3). I I. White, R, 1.. and Froeb. F. H. Smull-utrways dysfunction in non-smokers chronically cxpused tu tobacco smoke. New Engi. J. Med. 302, 7.0-723 /{yMU). :'RE~l. ThL majur This : at th prus t+ antl I t. uf thr, Tht: he:,lth relat4k Hcal t t marit anci ~, (a) cluac:L inwm tbl ntay h tiun, ;: (c) IlUtl5t1 pc:Up14 (d) Pnc:un i I,i.c, Vienn; TI BU 31686 t

Nltt.~,P.NfIVE .1It:U1CINE 13, 691-697 (19841 . pru- :~,ll)th -olled ell as : ment ,C after indoor )kC and Un thC icmtcal (y,~is of lmrntal ..I. pp. n1 lung Passive Smoking and Cancer-American Experience' LAWRENCE GAKF[NKEL Auteric•utr Cwrcer Sucirrv, Inc.. 4 West 35th 5'trerr. .Vrw Yurk, New York Ill(Xll Although one U.S. casc-control study has ,huwn that nonsmoking women married to ,mukers have a greater risk of lung cancer than do nonsmoking women married to nun- hmukors. LIata trum the Arnerieun Cancer Society's pru.rccuve study have failed to show .uch riak: and, a caae-cuntrul ,tudy baaed on reports uC whether or not nonsmoking women and men were cxpused to smoke of uthers alsu tadrsd to ,huw an nlcVatCd risk of lung cancer. Uata prc,ented here indicate that 4l)-5U%• of nonsmuking American women marnud to nUnsmUkCh Illily be exposed lt) passive smoke. More ltu/jie5 are neCtlett to etCnlutlstrilte the rule Ut palslve smoke in the development of lung cancer in nonsmokers. The American Cancer Society is currently collecting such data in a case-cuntrul und a prospective ,tudy. %. lyiat A.uJemt. Pre.a, Inc. pa."tvC The role of passive smoking in the development of lung cancer has become a study major issue in many segments of the scientific community in the United States. This topic is under investigation at a number of leading U.S. scientific centers. .5-1.r._6 At the American Cancer Society, we are undertaking twu separate ,tudic:, ta prospective study and a case-control study) on the hrobiem of passive smoking anti lung cancer. Fullowing is a description of both studie,, as well as a review "'t' and ul three published U.S. studies on the subject. The ill effects of passive smoking pertain not only to lung can,:er, hut tu other ~ health conditions us well. Several studies have indicated that pussive smoking is )o,etl to rclated to nunneuplastic lung disease (12. 14). !n 1975, the Office of Smoking and Health in its annual issue of "The Health Consequences of Smoking" ( 0) sum- maritt:d the available evidence on the effects of passive Nmoking to other diseases and conditions at that time. It concluded that: la) Tobacco smoke can be a significant Suurce of atmospheric pollution in en- cluwed areas. Occasionally, with heavy smoking and poor ventilation the max- imum limit for an 8-hr work exposure to carbon monoxide (CO) may be exceeded. (b) CO may produce some deterioration in psychomotor performance. This may become important when added to factors such u, tatigue, alcohol consump- tiun, and motor vehicle operation. 0 (c) Unrestricted smoking on buses and planes is annoying to a majority of nonsmoking passengers even under conditions of adequate ventilation. To some peupie, even slight exposure to smoke causes eye and throat irritatiun. (tl) Children of parents who smoke are more likely to have bronchitis and pncumunia during the lirst year of life. ' Nre,rntett at the Symposium "Medicat Penpective% on P.t,sive Smuking," April 9-13, 1984. Vtcnna. Auntria. 691 I091-7345181 )3.00 , I',.pynFhl - Iv1si by AcrJemic Preas, Inc. UI righta ut rcpruduUwn in an1 turne ieaerveJ. TI BU 31687

688 T. H[IZAYAMA Histology ot''_l ca.,c, of lung cancer in nonsmoking wives with smoking hus- bands was not essentially different from that of smoking women (adc:nucarci- noma. 57.M squamous cell carcinoma. 19.0%; and small-cell carcinoma, The current results of elevated risk of nasal sinus cancer in addition to the risk of lung cancer must strengthen the plausibility of carcinogenic hazards ut ,idestream smoke inhalation through the nose, as they are in line with the results of measurements of various carcinogens in sidestream smoke showing them to be present in higher concentrations than in mainstream smoke (2. 3). These results are also compatible with known evidence showing a possible influence of passive smoking on health including elevation of carbuxyhemoglubin and nicutinc/co- tinine levels in saliva, blood, and urine after exposure tu passive smoking: ele- vation of hydroxyproline levels in urine (a marker of cullagen destruction in lung tissue); the presence of mutagens in urine (1), small airway dysfunction in those expused daily to passive smoking in the workplace ( I I): and risk elevation t'ur pneumonia, bronchitis, and asthma in chiCdr`en with smoking parent(s). When the effects of passive smoking due to husbands' smoking were compared with the effects of direct smoking in women, the results clearly indicated that the effect of passive smoking is less than une-titth that ut' direct smoking, the SMR, being 1.55 and 3.81, respectively. In terms of attributable risk, huwever, the ettect of passive smoking on lung cancer in women is nearly as important as that ul' direct smoking because the population of intrahousehold passive smokers at risk is fuur times greater (n = 69,645) than the population of active smokers ln = 17. jf,b). Therefore, althougli the relative risk of indirect smoking is much hmallcr than that of direct smoking, the absolute excess deaths from lung cancer due to passive smoking may be quite important because of the large size of the exposed group--especially in countries such as Japan where the majority (nCarly17U1~(,) of adult men smoke, but only a minority ( I5"; or less) of adult women smoke. Passive smoking.can be divided into (a) direct passive smoking (direct inhala- tion of sidestream smoke before being diluted by room air) and (b) indirect pas,ive smoking (inhalation of room air polluted by sidestream smoke) according to the extent of proximity effect, just as droplet infection is separated from droplet nuclei infection in acute respiratory communicable Jiseases; the effect of venti- latiun is of limited importance in the former case, although quite significant in the latter. Small room size and congested li.ving conditions in Japan (and possibly also in countries like Greece) are naturally more conducive to direct passive smoking. As described in a previous report, the age-adjusted mortality rates for lung cancer are increasing rapidly for bu(h mcn and women in Japan. As only a fraction of Japanese women with lung cancer smoke cigarettes, the reasons why their mortality from lung cancer nearly parallels that of men have been unclear. The current study attempts to explain at least a part of this long-,tanding riddle. Although the average rate of female smokers in Japan has remained fairly stable over the past 20 years, a statistically signiticant increase in the mortality rate fur lung cancer in nonsmoking women was observed in our long-term follow-up study of a large-size population. Mortality rates per lut),UUtI t'ur ages 50-59, 60-6y, and 70 and above were 7.1, 17.7, and 31.0 in tirst 10 years of follow-up and 9.9, 27. l, and 44. can be in Japa As a the vali lung C.1 nunsnt, accura ubserv. mc:n ar lung C' smukin The imukir (an arL tuabi murs t' F'u,. lidrn~c

ratio rn in litcti `' the ninc: C tsu d ing. hree Det:n ione :tiun ln1c, lirr tur nn, tu .n. JYM I't)SIUM: ,MGt)IC•;\I. f'IittSPGC')'IVES ON PASSIVLi Si<tOKING 679 3. Buddecke. E. tEd.l. "Grundriss der kiiuchcmie." pp. 471-474. De Gruyter. l3eriintNew Yurk, 1973. s4. Chun-Yrung, M.. Ferrcir.r. 1'.. 1-ruhlich, J., Schulzrr. M., and T:rn. F. The etl"t, ui age, smuking, and alwhul on routine lahurutury Ie.ta. Anrrr. J. (7rn. Pculrr,l. 75, 302-326 ( lyril l. S. Dales. L. G., Friedman. G. l).. Sieyelaub, A. B.. and Seltzer, C. C. Cig;uctte smoking and serum chemntrv tent,. J. Ch,•r,na- 1Jrs. 27, 293-307 t 19741. h. Davies. S. F., UtTun), K. R. tiruwn, M. G., Campe, H.. and Niewoehner, D. Urine Je~.musine is unrtlweJ tu ~rgarcua ,muking or to sprrumetric function. ,intrr. Kev. Respir. Dr.r. 122S, 473- 475 I 19M3). 7. Gulin. J.. Kark, J. U., Haltim, S. T.. Friecllander. Y., and Stein. Y. Smoking and its relation to anthrupumctric chur.rcteristiCh and biochemical variables in Jerusalctn 17-year-ulals and adults. 1s. J. .tlcrl. .1•, r. 18, 1'_33-I241 (1982). M. Hengen. N., uml I Icngen. M. Gnn-IiyuiJ chromatugraphic determination of nicotine and cotinine in plasma. Clrtr. 1 luvrr. 24, 511-53 (19781. v. Hu,lcy, H. F,. Olson. K. f3., Hurtun, J., .'vtichelsCn. l?, antl Atkins. R. Automated analysia of urinary hydroxyproline for cancer research. Tr•e•lurrcun Inr. Cu,ryr. 1, Il15-11U (1969). Il). Jones. C. R.. l3ergman, Ad. W.. Kittner. P. J.. and Pigman. W. W. Urinary hyJroxypruline cxcrc- tiun in nurm.rl children and aclulCscents. Pruc•. Suc•. Lrp. Sru(. Myd. 115, 85-M7 (1964). II. Kasuga. H.. Mutsuki. H., and Osaka. F. "Hydroxypruline and Pas!,ivt Smoking." Paper pre- sented at the 9th International Scientilic Mecting of 1.1...\., Edinburgh. 19ti1. 13. Kasuga, H., M:,t,uki. H.. Usuka, F., and Inone, M. The study on the relationship between urinury Ityclruxypruhne ;rnd creutininc ratio Irum the viewpoint of public health. Tukcu J. rr:rp• Sdc•ri. 4, 343-351 11Y7y). 13. Langnc». U. HyJruxyprulinau„chcidung und Kullagcnstult•wcchael. Utsch. ,LA,I Wuc'lrrrt.cchr. 50, :53u-'_53i ( (H71)1. 14. Lewu, 'f. R. Crtterra relevant tu an uccuputiunal health ,tandard for nitrogen ,liuxicle. in "Ni- trogen Oxides and Their k•:Itc~r, on I lealth" IS. D. Lee, Et1.1, pp. 361-375. Ann Arbor Science PuMtxhcn. Ann Arbor, Mich., 1980. 15. Mellstrurm. 1).. Kun.lgrrn. A.. Jagenburg, R.. Steen. B.. and Svanborg. A. 11,b.,"u smoking. ;,ging, and health among the Clclerly:,a longitudinal pupulation .tudy of 7Q-year-uIJ men and an age arhurt comparison. A,ce .i,4errr,e 11, 45-5ts t 1Nii.2). (6. Sibernugl. S. Renal transport of amino acids. l;lur. W,u lrtvrsrlrr. 57, 1U0H-11119 (1979). 17. Tcchnicun :\utuunulycer 11: Clinical Method No. Sk:'2-W1I hC1. IN. Wharton. B. A., Brown. G., Rayner. P. H. W.. Il,rwrlls, G., and Pennuck, C. A. Urintlry hy- Jroxyprulinr in children with growth hurmone tJeliciency. Are~lr. Urs. CIu1J. 49, 159- 162 1 IN74/. 19. Wendcl, H.. and lirhprnning. W. Untersuchungen zur HydruxyprulinausachctJung im Harn bci chrunis,:hrr ob.truktiver lirunch,u,. L. ,r;esu,tue uur. Slecl. CJr,vr;,4eb. 36. 247-249 (t'1rit1. '_u. Wilkins. J. N.. C:ulurn. H. E.. van Vunukis. H., Hill. M. A.. Gritz. E., and Jarvik. M. E. Nic- otine trum ;rg.,ratte -smuking incrtaaes circulating levels of curti,ul. growth hormone, anJ prulucun in male chronic srnukrrs. i'%rckuplrurnm, odr,gy 78. 3U5-3UB 11'1tS_). =1. Yanagi+awa, Y., Ni,hrnwr.r, H., Matauki. H.. Osaka. F., Kaauga, H. "txposure-tflCct relation tur NU, with pcr•oma rxposure and urinary hydroxypruline to creauninC ratio ut schoolchil- dren as tndrc.,tur,." I'.iper presented at the 76th :\nnual Meeting of the Air Pollution Control Aiauciatiun. Atlanta. CieUrbta, June 19-23, 19143. TI BU 31675

own lblic vhen an I t de- .bits. a of huse. Part- huse iS of t the is of ;n«1 ,Nive nd's uber ~ are hese keLt.L ume hese tver _ttes tudy ,kut ~iun t the ancis, ,7r4. men =r of age'. ~: Cv .tte5 :CrS. ality :r in; . SYW'utitl:',l: tiIEDICAL tIEttSPECTivES ON P.ASSIVE SMOKING 693 TAIiLE I P!!KCL•NI'AUL UF NUNSMttKI~K> \LIKNIEU TO NONSMOKERS CLAStiIFIEU BY NUMBER OF HUUKS PHK ' WL•!:K THL•Y WL•Kt t,XNUSL•U fU SMUKL• UP UfHEKS Hours per week exposed Female 1%1 .41ule ('70) 0 59.5 50.8 1-9 27.8 32.3 tU-39 6.7 13.3 40 + 6.0 3.1 Tutat 100.0 100.0 nonsmoking men and women classified by the number of cigarettes smoked per day by their spuuses (8). Relative risks in women ranged up to 3.5 in those married to men who smoked 40 or more cigarettes a day. The relative risk of lung cancer in nonsmoking men mar.ried to smoking women was 2.0. This study was based un ?2 lung cancers in nonsmoking women and 8 in nonsmoking men. The authors ul,u examined whether or not there was an assueiatiun with par- ents' smoking habits. No assuciatiun was found with paternal ,muking, but there was an elevated risk in smokers whose mothers smoked. No elevated risk was tuund for nonsmokers. In 1984. Kabat und Wynder (IU) examined various demographic factors in 97 temale and 37 male lung cancer cases who reported being lifetime nonsmokers and equal numbers of matc:hcd nonsmoking contruls. The smoking history and histologic diagnoses on the hospital charts were rechecked unJ validatecl. They also presented preliminary data on exposure to ,idastream smoke fur a,ub5et of cases and controls. They reported on 25 nonsmoking men and 53 nonsmoking women classitiecd not only by whether or not their spouses smoked, but also by whether or not they were exposed to the cigarette smoke of others at work or at hume. In wumen, there was no increase in lung cancer risk fur any of the three categories. In men, there was an increased risk t'ur lung cancer in those exposed at work (marginally statistically significant). but no increase in risk for those cxpu.ed at home or from spouse's smoking. The authors concluded that more Jata must be collected before any conclusion can be drawn reiating to the effects uf passive smoking. With respect to classification ut'passive smoking. Friedman et ul., at the Kai.cr- 1'ermanente Hospital in Calil'urnia, questioned 38.000 nonsmokers (c:x-smukers plus those who had never smoked) about their exposure to smoke of others (5). rable I shows the percentage of nonsmokers rnarried to nonsmokers classified by the numbcr ut hours per week they were exposed tu smoke of others. About 1u' , of women and itl' i ut' men were exposed to the smoke of others for some pcnud of time. and of women and 3% of men married to nonsmokers were C.\puwd tur_4U or mure huur5.per week. Among women nonsmokers married to ,mukers, 47% repurted that they were nur exposed to tobacco smoke at home. There is another type of evidence thut is related to the cla,siticatiun of exposure to -3muke of others. Over a period of nearly 3U years, Dr. Oscar Auerbach, a

LAWRENCE GARFINKEL TABLE 2 CLASSII•k:\rIUN oF LUNG CANCEk CASES BY SMOKING HABIT AND CATEGORY OF BRONCHIAL EPITHELIAL ATYPISM IN WHICH THE MAJUR1TY UF SLIDES IN rHE CASE WERE Kt:C'URllL•tl Smoking habit Never smoked regufuly < 10 Cigarettes a day 10-19 Cigarettes a day 20-39 Cigarettes a day a0+ Cigarettes a day Numher of cases in which the percentage with atypism in majority ul' s6des was: Total 0 10-29 30-49 50-69 70 r 57 57 1 ii 5 8 5 - 32 I 4 27 - 68 1 7 53 7 35 - ~ 7 28 pathologist at Veterans Hospital in New Jersey, in cooperation with Dr. E. C. Hammond and the author, has been studying the histologic changes in the tra- cheubronchial tree ot'thousands ot'autopsiud men and women, including hundreds of persons who had never smoked (1-3). Usually, about 40-50 cross-sectioned slides are selected for each case, taken from different parts of the tracheubrun- chial tree and read in random order. The histologic lesions recorded include presence of cClls with atypiu.kl nuclei, basal cell hyperplasia, carcinoma in silu. and early preclinical invasive cancer. These changes have been observetd in direct proportion to the number of cigarettes the person smoked per day whilt: they liveil; they decreased after cessation of smoking. Table 2 shows results from a study of bronchial epithelia. Slides were classified according to the percentage ot'epithelial cells in which atypical nuclei were found. Each case was then classified by the category of atypical cell most often seen in the case. With an increase in Smoking, an increase in the proportion of cases talling into groups with higher proportions of atypical cells wx, seen. In the 57 nonsmoking cases, virtually all slides had "U%" atypical cells. If passive smoking does indeed lead to lung cancer, it would hu expected that some of the non- smokers who hatd been zxpu,«I t,„moke tit others would have shown histological changes in their bronchial epithelium sirnilar to those Iuund in the light (less than 10 cibarettes a day) smokers, but no such cases were observed. This does nut exclude the possibility that the mechanism by which passive smoking could lead to cancer might be totally dit'ferenr from that in smokers. At the Americun Cancer Society. we are pursuing two avenues of investigation related to passive smoking and lung cancer. One is a case-control study in which lung cancer cases in nonsmoking women diagnosed during the years 1971 -1yzS 1 are being abstracted in tour hospitals. We are matching these cases with non- smoking women with colon/rectum cancers (a 1:3 match by age). Culon/rectum cancers wcre chosen as controls because this site has been shown to be unrelated to smoking habits (7, 11). The slides of every eligible case and control are being reviewed2 to establish whether the lung or culun/rectum cancer diagnosia was appropriate. In addition. intcrvicwers with no knowledge of the tliagnuais are ' The slide review is being implemented by l)r. Uscar Auerbach. TI BU 31690 NL. ubtait of utt who • Tat cabe~ dia6n wctre the h tnicri mcUt ~'cu{ll at di;. In , ttun"i. in nu, pruut W~: have D. \~ tis ~'tu& w nt Ieaat tiun alSu

LAWRENCE GARFINKEL TABLE 2 CLASSII•k:\rIUN oF LUNG CANCEk CASES BY SMOKING HABIT AND CATEGORY OF BRONCHIAL EPITHELIAL ATYPISM IN WHICH THE MAJUR1TY UF SLIDES IN rHE CASE WERE Kt:C'URllL•tl Smoking habit Never smoked regufuly < 10 Cigarettes a day 10-19 Cigarettes a day 20-39 Cigarettes a day a0+ Cigarettes a day Numher of cases in which the percentage with atypism in majority ul' s6des was: Total 0 10-29 30-49 50-69 70 r 57 57 1 ii 5 8 5 - 32 I 4 27 - 68 1 7 53 7 35 - ~ 7 28 pathologist at Veterans Hospital in New Jersey, in cooperation with Dr. E. C. Hammond and the author, has been studying the histologic changes in the tra- cheubronchial tree ot'thousands ot'autopsiud men and women, including hundreds of persons who had never smoked (1-3). Usually, about 40-50 cross-sectioned slides are selected for each case, taken from different parts of the tracheubrun- chial tree and read in random order. The histologic lesions recorded include presence of cClls with atypiu.kl nuclei, basal cell hyperplasia, carcinoma in silu. and early preclinical invasive cancer. These changes have been observetd in direct proportion to the number of cigarettes the person smoked per day whilt: they liveil; they decreased after cessation of smoking. Table 2 shows results from a study of bronchial epithelia. Slides were classified according to the percentage ot'epithelial cells in which atypical nuclei were found. Each case was then classified by the category of atypical cell most often seen in the case. With an increase in Smoking, an increase in the proportion of cases talling into groups with higher proportions of atypical cells wx, seen. In the 57 nonsmoking cases, virtually all slides had "U%" atypical cells. If passive smoking does indeed lead to lung cancer, it would hu expected that some of the non- smokers who hatd been zxpu,«I t,„moke tit others would have shown histological changes in their bronchial epithelium sirnilar to those Iuund in the light (less than 10 cibarettes a day) smokers, but no such cases were observed. This does nut exclude the possibility that the mechanism by which passive smoking could lead to cancer might be totally dit'ferenr from that in smokers. At the Americun Cancer Society. we are pursuing two avenues of investigation related to passive smoking and lung cancer. One is a case-control study in which lung cancer cases in nonsmoking women diagnosed during the years 1971 -1yzS 1 are being abstracted in tour hospitals. We are matching these cases with non- smoking women with colon/rectum cancers (a 1:3 match by age). Culon/rectum cancers wcre chosen as controls because this site has been shown to be unrelated to smoking habits (7, 11). The slides of every eligible case and control are being reviewed2 to establish whether the lung or culun/rectum cancer diagnosia was appropriate. In addition. intcrvicwers with no knowledge of the tliagnuais are ' The slide review is being implemented by l)r. Uscar Auerbach. TI BU 31690 NL. ubtait of utt who • Tat cabe~ dia6n wctre the h tnicri mcUt ~'cu{ll at di;. In , ttun"i. in nu, pruut W~: have D. \~ tis ~'tu& w nt Ieaat tiun alSu

690 r. HIRAYAMA The results of the present study must be etfectively utilized in planning pro- grams for thc control uf lung cancer and other selected cliseasc,, . The results clearly indicate that lung cancer, especially in women, can only be controlled satisfactorily when proper measures are taken against passive smoking as well as against active smoking, especially in countries like Japan. A similar statement may also be valid for cancers ut' other selected sites. REFERENCES 1. Bos, R. P., Theuws, J. K. G., and Henderson, P. Excretion of mtuagens in human urine after passive bmuking. Cancer Lett. 19, 85-90 (1983). 2. Brunnemann, K. D., Adams,1. D., Hu, D. ('. S., er u(. The intluence of tobucco smoke on indoor atmospheres. 11. Volatile and tobacco specitic nitrosamines in main• and si~lestreucn Smokc and their cuntribution to indoor pollutiun, in "Proceedings of the ~th 7uint Conference on the Jensing ut Environmental Yollutants. New Orleans 1977", pp. 876-880. American Chemical Sucirty, Wushingtun, D.C.. 1978. 3. Brunnemann, K. D., and Hot'fmann, D. Chemical ,tudies on tobacco smoke L1X. Analysis of volatile nitrusumines in tubuccu smoke and polluted indoor envirunments, in "EnvironmCntal Aspects of N-nitroso Cumpounds" (E. A. Walter, L. Griciutc .,nd M. Gustegnaro, Eds.). pp. 343- 156. [ARC Scientific Publications No. 19, WHU, I},,n, 1978. 4. Correa, N.. Pickle. L. W., Funthum. E.. Lin, Y., and H.+cnseel, W. Passive smoking and lung cancer. Lunret 1, 595-597 (1983). 5. Gartinkcl. L. Time trends in lung cancer mortality among non-smukers and a note on passive ,muking. J. Nut(. Cancer /nsY. 66, 1116 1-1 066 (lytl I). 6. Hirayama, T. Non-smoking wives of heavy emukcrs h;,ve a higher risk of lung cancer: A~tualy in Japan. Bret. .bled J. 282, Iri3-lH5 (19tf1). 7. Hirayuma, T. Passive smoking and lung cancer: Consistency of aswciatiun. Lancet 1, 1425-1a:6 (1983). 8. Kabut, G. C., and Wynder. E. L. Lung cancer in nonsmokers. Ccuncer 53, 1214-1221 11984). 9. Preston-Martin, S.. Yu, M. C.. lienton, B., and Henderson, B. E. N-Nitroso cumpuunda and childhood brain tumors: A caae-cuntrol study, Cancer Rcs. 42. 524t1-5245 11982). I(1. T'richupOulos, D.. Knlandidi. A., and Sparrus. L. Lung cancer and passive smoking; conclusion ul Greek study. Lcuu•rt 1. 677-678 119tl3). I I. White, R, 1.. and Froeb. F. H. Smull-utrways dysfunction in non-smokers chronically cxpused tu tobacco smoke. New Engi. J. Med. 302, 7.0-723 /{yMU). :'RE~l. ThL majur This : at th prus t+ antl I t. uf thr, Tht: he:,lth relat4k Hcal t t marit anci ~, (a) cluac:L inwm tbl ntay h tiun, ;: (c) IlUtl5t1 pc:Up14 (d) Pnc:un i I,i.c, Vienn; TI BU 31686 t

HIAL tU E. C. 4 tra- drcch iUnet1 brun- Caudc sf!!t. lit'ect they iticcl uncl. :n in ases e 57 k ing wn- ical han nUt =ad a SYV1POSIUNt: MEDICAL PEkSPECrIVES ON PASSIVE SMOKING TABLE 3 Nl•>i1/L•k UF WOMEN WITH LUNG C.{NCEk ANU NuMNfik Vf;RIFI(;D TO BE NONSMOKERS WITH CONFIRMED MICkuS/'UPIC PkUUF Hospital A 284 Women with lung cancer on hospital record 7i Were nurtsmukers or smoking habit was not ,tuted 37 No histologic pruol' or not lung cancer mlcroscupically ~0 Microscopic proof ut lung cana;er-smuker Ih Microscopic proof of lung c;utcer-nunsmuker Hospital B 114 Women with lung cancer on hospital record 31 Were nunsmukers or smoking habit was nut statet/ t I Nu histological proof or nut lung cancer microscopically 7 Microscopic pruul' uf lung cancer-smuker 13 Microscopic proof of lung eancer-nonsmuker Hospital C 277 Women with lung cancer on hospital record 104 Were nonsmokers or smoking not stated 69 Not yet processed 7 No histologic proof or not lung k;ancer microscopically 15 Microscopic proof uflung cancer-smoker 13 Microscopic pruuf of lung cuncer-=nunsmoker ubtaining detailed answers to questions on current and past exposure to the smuke of others. childhood smoke exposure, husbantl's smoking habit, and, tur those who smoke, the number of cigarettes smoked at home. Table 3shuws the number of cases that had to be reviewed to obtain verified case5 of nunsmui.ing women with lung cancer. in Hospital A. ':c4 women were diagnosed with lung cancer in the 11-year period. St;venty-tive of them either were rcpurtzcl to be nonsmokers or their smoking habits were not mentioned on the hospital record. After histologic revit:w of the slides, 37 either were lacking microscopic pru0l' uF lung cancer or were misdiagnosed. Most ut' the latter were metastutic cancers to the lung frutn another site. Twenty of those with micru- ,cupic prout'were confirmed tu bc either current smokers or cx-cigarettt: smokers at Jiagnusis; 18 were nonsmokers. ln Hospital B, ut' 114 lung cancers in I1 years, 31 either were reported to be nonsmokers or their smoking habits were not known. Eleven ut'them were lacking in microscopic proof or were misdiagnosed as lung cancer; 7 with microscopic pruuf wcrz smokers, and 13 were nonsmokers. We have not yet confirmed all lung cancer cuses in Hospital C, but we currently have 13 nonsmokers with confirmed microscopic pruut'uFlung cancer. In Hospital D, we have not yet accumulated enough cases to show similar distributions. in 1910. the American Cancer Society beban a new. l.u-.uc-scalce prospective ,tucfy: 77,000 volunteers enrolled l'0U,UUU persons, asking them to complete a cunttdcntial questiunnaire. They will trace these individuals tur a periud of at least n years. Among thc 400 questions on the ti)ur-page que5tiunnairc is it que5- tiun related to passive smoking. Sincu enrollment is by household, the ,uhjects ai,u will be cla,sitietl according to the smoking habits of their spouses. The

Nltt.~,P.NfIVE .1It:U1CINE 13, 691-697 (19841 . pru- :~,ll)th -olled ell as : ment ,C after indoor )kC and Un thC icmtcal (y,~is of lmrntal ..I. pp. n1 lung Passive Smoking and Cancer-American Experience' LAWRENCE GAKF[NKEL Auteric•utr Cwrcer Sucirrv, Inc.. 4 West 35th 5'trerr. .Vrw Yurk, New York Ill(Xll Although one U.S. casc-control study has ,huwn that nonsmoking women married to ,mukers have a greater risk of lung cancer than do nonsmoking women married to nun- hmukors. LIata trum the Arnerieun Cancer Society's pru.rccuve study have failed to show .uch riak: and, a caae-cuntrul ,tudy baaed on reports uC whether or not nonsmoking women and men were cxpused to smoke of uthers alsu tadrsd to ,huw an nlcVatCd risk of lung cancer. Uata prc,ented here indicate that 4l)-5U%• of nonsmuking American women marnud to nUnsmUkCh Illily be exposed lt) passive smoke. More ltu/jie5 are neCtlett to etCnlutlstrilte the rule Ut palslve smoke in the development of lung cancer in nonsmokers. The American Cancer Society is currently collecting such data in a case-cuntrul und a prospective ,tudy. %. lyiat A.uJemt. Pre.a, Inc. pa."tvC The role of passive smoking in the development of lung cancer has become a study major issue in many segments of the scientific community in the United States. This topic is under investigation at a number of leading U.S. scientific centers. .5-1.r._6 At the American Cancer Society, we are undertaking twu separate ,tudic:, ta prospective study and a case-control study) on the hrobiem of passive smoking anti lung cancer. Fullowing is a description of both studie,, as well as a review "'t' and ul three published U.S. studies on the subject. The ill effects of passive smoking pertain not only to lung can,:er, hut tu other ~ health conditions us well. Several studies have indicated that pussive smoking is )o,etl to rclated to nunneuplastic lung disease (12. 14). !n 1975, the Office of Smoking and Health in its annual issue of "The Health Consequences of Smoking" ( 0) sum- maritt:d the available evidence on the effects of passive Nmoking to other diseases and conditions at that time. It concluded that: la) Tobacco smoke can be a significant Suurce of atmospheric pollution in en- cluwed areas. Occasionally, with heavy smoking and poor ventilation the max- imum limit for an 8-hr work exposure to carbon monoxide (CO) may be exceeded. (b) CO may produce some deterioration in psychomotor performance. This may become important when added to factors such u, tatigue, alcohol consump- tiun, and motor vehicle operation. 0 (c) Unrestricted smoking on buses and planes is annoying to a majority of nonsmoking passengers even under conditions of adequate ventilation. To some peupie, even slight exposure to smoke causes eye and throat irritatiun. (tl) Children of parents who smoke are more likely to have bronchitis and pncumunia during the lirst year of life. ' Nre,rntett at the Symposium "Medicat Penpective% on P.t,sive Smuking," April 9-13, 1984. Vtcnna. Auntria. 691 I091-7345181 )3.00 , I',.pynFhl - Iv1si by AcrJemic Preas, Inc. UI righta ut rcpruduUwn in an1 turne ieaerveJ. TI BU 31687

own lblic vhen an I t de- .bits. a of huse. Part- huse iS of t the is of ;n«1 ,Nive nd's uber ~ are hese keLt.L ume hese tver _ttes tudy ,kut ~iun t the ancis, ,7r4. men =r of age'. ~: Cv .tte5 :CrS. ality :r in; . SYW'utitl:',l: tiIEDICAL tIEttSPECTivES ON P.ASSIVE SMOKING 693 TAIiLE I P!!KCL•NI'AUL UF NUNSMttKI~K> \LIKNIEU TO NONSMOKERS CLAStiIFIEU BY NUMBER OF HUUKS PHK ' WL•!:K THL•Y WL•Kt t,XNUSL•U fU SMUKL• UP UfHEKS Hours per week exposed Female 1%1 .41ule ('70) 0 59.5 50.8 1-9 27.8 32.3 tU-39 6.7 13.3 40 + 6.0 3.1 Tutat 100.0 100.0 nonsmoking men and women classified by the number of cigarettes smoked per day by their spuuses (8). Relative risks in women ranged up to 3.5 in those married to men who smoked 40 or more cigarettes a day. The relative risk of lung cancer in nonsmoking men mar.ried to smoking women was 2.0. This study was based un ?2 lung cancers in nonsmoking women and 8 in nonsmoking men. The authors ul,u examined whether or not there was an assueiatiun with par- ents' smoking habits. No assuciatiun was found with paternal ,muking, but there was an elevated risk in smokers whose mothers smoked. No elevated risk was tuund for nonsmokers. In 1984. Kabat und Wynder (IU) examined various demographic factors in 97 temale and 37 male lung cancer cases who reported being lifetime nonsmokers and equal numbers of matc:hcd nonsmoking contruls. The smoking history and histologic diagnoses on the hospital charts were rechecked unJ validatecl. They also presented preliminary data on exposure to ,idastream smoke fur a,ub5et of cases and controls. They reported on 25 nonsmoking men and 53 nonsmoking women classitiecd not only by whether or not their spouses smoked, but also by whether or not they were exposed to the cigarette smoke of others at work or at hume. In wumen, there was no increase in lung cancer risk fur any of the three categories. In men, there was an increased risk t'ur lung cancer in those exposed at work (marginally statistically significant). but no increase in risk for those cxpu.ed at home or from spouse's smoking. The authors concluded that more Jata must be collected before any conclusion can be drawn reiating to the effects uf passive smoking. With respect to classification ut'passive smoking. Friedman et ul., at the Kai.cr- 1'ermanente Hospital in Calil'urnia, questioned 38.000 nonsmokers (c:x-smukers plus those who had never smoked) about their exposure to smoke of others (5). rable I shows the percentage of nonsmokers rnarried to nonsmokers classified by the numbcr ut hours per week they were exposed tu smoke of others. About 1u' , of women and itl' i ut' men were exposed to the smoke of others for some pcnud of time. and of women and 3% of men married to nonsmokers were C.\puwd tur_4U or mure huur5.per week. Among women nonsmokers married to ,mukers, 47% repurted that they were nur exposed to tobacco smoke at home. There is another type of evidence thut is related to the cla,siticatiun of exposure to -3muke of others. Over a period of nearly 3U years, Dr. Oscar Auerbach, a

692 LAWRENCE GARFINKEL (e) CO levels commonly found in smoke-fifled environments have been shown to decrease the exercise tolerance of patients with angina pectoris. For all of the above rea,uns, there has long been reason to prutcct the public trum involuntarily inhaling the smoke of others. The American Cancer Society's Cancer Prevention Study I began in 1959 when 68.000 American Cancer Society volunteers in 25 states enrolled more than I million men and women for long-term follow-up. These subjects tilled out a de- tailed questionnaire about themselves, including questions on smoking habits. The volunteers reported annually on deaths, changes of address, and changes of names by marriage. Over a l'-year period, the volunteers traced yii.4r/o ot thuse originally enrolled. Death certificates were obtained from state health depart- mcM, and related to the data on the original questionnaires. For those whose death certificates specified cancer, verification of the rHmary site and basis of diagnosis was obtained from physicians and hospitals. It was this data bank that served as the source for many publicatiun, on the zitccts of smoking (7-9) as well as the American Cancer Society's analysis of p:u,ive smoking and lung cancer. The questions in the study were ttut designed to quantify the number of hours a day each person was exposed to passive smoking. Instead, in this study, as in several other published studies, a husband's smoking habit was identified as the source of passive smoking, and the number of cigarettes that the husband smoked per day was wwd to quantify the degree of passive exposure to smoke. In the United States partI,:ularly, nonsmokers are exposed to cigarette smoke from suurcc~ other than in the home, but these sources were not included in the c:lassiticatiun of passive smoke :xpusure. The study included 375,UUU women who reported that they had never smokzJ. Many were widowed or single, divorced or separated, and the husbands of some married women did not till out questiunnuires. "['he husbands of 176.000 of these women, however, did complete questionnaires indicating whether they had never smoked, smoked less than 20 cigarettes a day, or smoked 'U or more cigarettes a day. The majority of the men had smoked for 20 or more years beturc: the study began. Women whose husbands wvre ex-smokers or whose husbands smoked only cigars or pipes were excluded trum the analysis. Expected numbers of death, in women with smoking husbands were based on lung cancer rates t'ur the 12-year period ( ly6U-l97?) by 5-year age groups of the women with nonsmoking husbands: The nonsmoking women whose husbands smoked less than 20 cigarettes a• day had a riak of dying from lung cancer ?71/r higher than those married to nonsmokers. Nonsmoking women married to men who smoked 20 or more cigarettes a day had a mortality ratio of 1. lU. Neither of the ratios was statistically significant (6). A separate analysis was made for women in the three groups matched for age. race, educational status, urban or rural residence, and husband's expo,ure to du,ts or chemicals. Wunien married tu men who smoked less than _U cigareue, a day had a mortality ratio of 1.37 compared with those married to nonsmokers. Thu,t married to men who smoked 20 or more cigarettes a day had a mortality ratio of 1.04. Neither of these was statistically significant. In lyti3, Correa et ul. (4) published a Case-cuntrol study of lung cancer in TI BU 31688 PLxCtN nunsmu day by i to men in nonsi on 22 lu The a ents' sw W as an found tL In lW tenialc: . and equ histulua also pre caaes at women whether home. 1: categu6 at work exposed Jata mu of pas,i% With rL 1'ermane: plus thus 'rable I , by the nt 4U~'r ut ' periud u cxpu,ed smuker, There to .~niuk

696 t.AWKENCE GAftFINKEL. question is designed to determine thi: amount of smuke to which a person i~ expused. It reads: Whether or noi you smoke. on the avcrage, how muny hours a Jay are you exposed lu hmoke from uthers:' At Home At Work In Other Arnas . studies have been based on relatively small numbers of nonsmoking lung cancer cases. It is important to emphasize that smoking habits need to be verified and pathological diagnoses of lung cancQr have to be confirmed. In our own study we hope to accumulate up to I(X) ca,cs before making an analysis. As yet, no con- clusions have been drawn by us as to whether or nut passive smuke exposure i, related to IunL cancer. Continuing research will provide the inturmation needed to do so. Preliminary analysis of the answers to this question shows that. among non- smuking women. 3 1% reported they were not exposed to snwke at all, 25% were exposecd only occasionally, 15% were exposed 1-7 hr a day, und 29% were ex- posed 8 hr or more a day. In the case-cuntrul study in Hospitals A and B, preliminary data indicate that 64% of the control women reported they were not exposed to passive smoke at all, 25% were exposed 1-9 hr a week. 6% were exposed 10-39 hr a weck, and 5% were exposed 40 hr or more a week. These figures are very similar to those reported by Friedman et ul. (5). Ot' the total number of hours during which the control subjects were subject to smoke of others, 75% was in the home, 20% was at work, and 5% occurred in other areas. We also classified nonsmoking women according to the average percentage ut the total amount of their husbands' cigarette smoking that occurred in the home. P%liminary data from the lung cancer group showed that an average of -t1% of the cigarettes smoked by husbands of nonsmokers with lung cancer were smoked at home, compared with an average of about 465''o of cigarettes smoked by hus- bands of the colun/rectum cancer cases. Thus, it is apparent that in our U.S. tudies, classification of nonsmokers by the number of cigarettes smoked by a spouse is only a crude estimate uf the degree of passive exposure. The method for measuring passive stnuking is an important consideration. It should also be recognized that the effects of passive exposure are highly related to the proximity of the smuker, the amount of ven- tilation in the room, the number of smokers in the ruom, and the number of hours exposed. In atltlitiun, smokers themselves inhale a great deal more passive smoke by virtue of their proximity to their smouldering cigarettes than do others who may be in the same area. Another point to be considered in relutiun to the biologic effects of passive smoking is the lack of evidence at autopsy of atypical cells in the lungs of nonsmokers. At this point there are both positive and negative studies in the United States on the issue of passive smoking and lung cancer. More data are needed antd will be published by our group and by others. Most of the published case-contrul IU ClbarCtte lntu(\Inb. Iy55-t10 V-,. I97U-77. NPw L:IIgl. f. ,Ncd. 3011. 1FiI-3HA /I'1791. I. Auerhach. 0.. Gartinkal. L.. and Hcunmund. E. C. Change, in brunchial epithrlium in rtlauun REFERENCES TI BU 31692 2. Auerbal !,ntuk 3. AuCrbaL. in ret. (1962 4. Correu. and IL 5. Frieilm:, A,lrr, n. Gartink. ~muk 7. Hammu c'al,< t!. Hammu in thc 9. Hammu ttr,l. 10. Kabat. ~ I t. Kahn. I and .: 12. Kauttm. ii~k t 13. U.S. Pu uHE 14. White. tuhu.

692 LAWRENCE GARFINKEL (e) CO levels commonly found in smoke-fifled environments have been shown to decrease the exercise tolerance of patients with angina pectoris. For all of the above rea,uns, there has long been reason to prutcct the public trum involuntarily inhaling the smoke of others. The American Cancer Society's Cancer Prevention Study I began in 1959 when 68.000 American Cancer Society volunteers in 25 states enrolled more than I million men and women for long-term follow-up. These subjects tilled out a de- tailed questionnaire about themselves, including questions on smoking habits. The volunteers reported annually on deaths, changes of address, and changes of names by marriage. Over a l'-year period, the volunteers traced yii.4r/o ot thuse originally enrolled. Death certificates were obtained from state health depart- mcM, and related to the data on the original questionnaires. For those whose death certificates specified cancer, verification of the rHmary site and basis of diagnosis was obtained from physicians and hospitals. It was this data bank that served as the source for many publicatiun, on the zitccts of smoking (7-9) as well as the American Cancer Society's analysis of p:u,ive smoking and lung cancer. The questions in the study were ttut designed to quantify the number of hours a day each person was exposed to passive smoking. Instead, in this study, as in several other published studies, a husband's smoking habit was identified as the source of passive smoking, and the number of cigarettes that the husband smoked per day was wwd to quantify the degree of passive exposure to smoke. In the United States partI,:ularly, nonsmokers are exposed to cigarette smoke from suurcc~ other than in the home, but these sources were not included in the c:lassiticatiun of passive smoke :xpusure. The study included 375,UUU women who reported that they had never smokzJ. Many were widowed or single, divorced or separated, and the husbands of some married women did not till out questiunnuires. "['he husbands of 176.000 of these women, however, did complete questionnaires indicating whether they had never smoked, smoked less than 20 cigarettes a day, or smoked 'U or more cigarettes a day. The majority of the men had smoked for 20 or more years beturc: the study began. Women whose husbands wvre ex-smokers or whose husbands smoked only cigars or pipes were excluded trum the analysis. Expected numbers of death, in women with smoking husbands were based on lung cancer rates t'ur the 12-year period ( ly6U-l97?) by 5-year age groups of the women with nonsmoking husbands: The nonsmoking women whose husbands smoked less than 20 cigarettes a• day had a riak of dying from lung cancer ?71/r higher than those married to nonsmokers. Nonsmoking women married to men who smoked 20 or more cigarettes a day had a mortality ratio of 1. lU. Neither of the ratios was statistically significant (6). A separate analysis was made for women in the three groups matched for age. race, educational status, urban or rural residence, and husband's expo,ure to du,ts or chemicals. Wunien married tu men who smoked less than _U cigareue, a day had a mortality ratio of 1.37 compared with those married to nonsmokers. Thu,t married to men who smoked 20 or more cigarettes a day had a mortality ratio of 1.04. Neither of these was statistically significant. In lyti3, Correa et ul. (4) published a Case-cuntrol study of lung cancer in TI BU 31688 PLxCtN nunsmu day by i to men in nonsi on 22 lu The a ents' sw W as an found tL In lW tenialc: . and equ histulua also pre caaes at women whether home. 1: categu6 at work exposed Jata mu of pas,i% With rL 1'ermane: plus thus 'rable I , by the nt 4U~'r ut ' periud u cxpu,ed smuker, There to .~niuk

0 TAt3Lf: I c 0 LUNG C-ANCEK INC/DLNCL IN NEVEK-SAfOKEKS AND SMOKERS Of VAR1•ING A?tOL'N7S Of LIGAKL1IES PLR DAY FOK DSrFEKENi AGLS AND Ru.AiIvL RISKs (RR) FOK Ll'!aa C AN/ l K03) lncidence I4-.IKNl, n SmoF,crs Iag day) z z v c Never smoked A 1110 1/4 1:2 . 34 1 l 5 10 -4 g i 40 39 0 0.40 0.42 0A5 0.4'1 0.52 0.69 1.30 z 2.74 50 . 2.95 3.05 3?(1 3.46 3.73 4(ll 5.24 9.91 < 20.96 c 60 11 90 30 12 91 12 13 97 15.06 16.20 21.16 40.00 ria.63 -i 70 . 34.48 . 35.64 . 37.41 . 40.47 43.64 46.93 61.30 115.90 C '4S'0 n 79 75.21s 77.81 81.68 88.35 95.27 102.47 133.83 25'_.03 535.33 RR 00 1 03 1 1.08 1.17 1.27 1.36 1.78 3.36 7.11 -3 . . ~ ~ ° Caicutalrd accorJing Doll and Pe1o (5). ci L= a o~ C3 '° C~ '~ < t~ 11 Z G? G~C ^ n t''t Er I" -, r_-, > Gr_; C~ V ry G(:. ~. .- .. r. C.. .- . . I

on is U non- were e Zx- : that keat and hu,e I the Was ge u€ urne. `•~ uf okc:cd hus- -s by t the : ~- an ,'ivc v4n- uurs noke who loSic Is in SYV(P()titlJ41: vIEDICAL PtiRSPECTIVES ON PASSIVE SMOKING 697 2. Auerbaeh, () . ~icwt. A. P.. Hammund, E. C., and Gartinkel. L. l3runchiul epithelium in (urmer -3mukers. Vccr Lntl. J. Myd. 267, 1 IN-125 11962). 3. Auerbach. 0.. Stuut. A. P.. Hammond. E. C.. and Gurtinkel. L. Changes in bronchial epithelium in relatwn to ,ex, age, residence. smoking anJ pncumunra. Vrsr Eir;l. J. Mrd. 267, 11I-I IiS (I'ki'-). 4. Currea. P., Pickle. L. W., Fontham. E.. PiLklc, W., Lin. Y.. and Haenstel, W. l".c„ive smoking and lung cancrr. Lunc•at 2, 595-597 11'!tl3). 5. Friedman. G. U.. l'euut. I). 13., and Bawul. R. D. Prevalence and currelutes ut Iw„ive smuking. Airrrr. J. Public lfrulNr 73, i01-4U5 /1y2f)). h. Gartinkd, L. l'ime trends in lung cancer mortality among nonsmokers and a note on passive smoking. J. Nurl. Cunrer /n.st. 66, 1061- I llhb (1981). 7. Hammond. E. C. Smoking in relation tu the denth rates ufune million men and women. Nud. Cuncrr ln.sf. Nlucrcpqr. 19, 127-204 11966). S. Hammond, E. C., and (;artinkei. L. Curonury heart clisen.e. ,truke. and aortic aneurism: Factors in the etiology. Are'h. Etmirua. Heulth 19, 167-IK'- I 1%'ll. y. Hammond. E. C., and Gurtinkci. L. General air pollution anad cancer in the United States. Prcn•. L/c•c/. 9, 2Ub-21 1 (1980). 10. Kahat, G. C., and WyntJer, E. L. Lung cancer in nonsmokers. Cunc•rr 53, 1214-1221 (1984). I I. Kahn. H. A. The Dorn Study ut' smoking and mortality among U.S. veterans: Report on eight and une•half ol' oh.ervatiun. Ncetl. Cancer btev. ,tlucru,qr. 19, 1-126 119661. 12. Kaullmunn, F.. Ic,.icr. J. F., and (hiuL W. :Ndutt passive smoking in the home environment: A risk factur tur chronic airtlow limitation. Airrer. J. lipidecnia(. 117, 269-?M0(1'J233). 13. U.S. Public Health Service. "The Health Cunxyuences ul Smoking: A Reference tclitiun." U.S. UHEW, U.S. Govt. Printing Oftice. Wu,hingtun. D.C.. 1976. 14. White, 1. R., and 1•rueb, Il. I• Surall airways Jysfunctiun in nonsmokers chronicy lly expuseJ to tobacco smoke. Vr+r @'n;f. J. MeJ. 302, 720-723 c lytf0): TI BU 31693

i SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 699 tations are based on a cigarette with a tar content of 16 mg, which roughly corresponds to the average tar content of cigarettes in the United States (20), in Great Britain (24), and in Austria ( I), tur the passive ,muking situations indicated by fZepace and Luwrey, we obtain cigarette equivalents of ihe order of 0.2 to I cigarette per day. The tigures t'ur cigarette equivalents reported by other authors are also found to lie within this range; Klusterkotter and Gono (12): 0.32 cigarette/ ti hr; Hugud et al. 1 t0): 0.5 to I cigarette/day. For the present study, this range was taken as the actual level of exposure to noxae of nonsmokers under passive ~muking conditions. LUNG CANCER RISK In Table I. the incidence of lung cancer per 100,000 is ihown for never-smokers and for smokers who have smoked varying numbers of cigarettes per day. The incidence is reported for dit'ferent age groups, and the relative risk oF smokers compared with nevcr-smokers has been calculated (23). DISCUSSION As indicated above, the exposure to noxae under passive smoking conditions is comparable to the exposure to nu.rae represented by one actively smoked cigarette per day. ln the present investigation, therefore, the results for smokers in this exposure range have been correspondingly tpplied to passive smokers. The inference made by Doll and Peto (>) was developed on the basis of data obtained from studies uf ma:n. Since there is no evidence to suggest that the risk uf contracting bronchial carcinoma is affected by ,ex-specitic facturs, this pru- cetiure may legitimately also be applied to women. A., the present analysis shows, in the range of.small doses (exposures), carcin- ogenic processes manifest an approximately ltnear dose-response relationship (3) for the course of the risk in the range 0.1 to I cigarette/day. The question of whether a threshold value exists (below which there is no risk) cannot be an- ,wered. However, the risk is expressed by such low tigures that the range for very small exposures can be considered "safe." For a Pcr,on exposed to passive ,muking (U.l to I cigarette/day), compared with a nonexposed prr,un, the risk lies within the range 1.03 to 1.36. Such a risk may be ignored. since in comparison with the greater risks, it has no appreciable effect on lung canccr incidence t'l:Lble I). The intluence becogtes more marked only in the higher age gruups, where the "ircinogenic effect of tobacco smoke is very strongly•intluenced by the eJtu'ation ut cxposure (5). Since risk is also dependent upon age at initial exposure (7, 11, '-a. a more marked effect of passive .ntuking is conceivable when exposure is cuntinuous from childhood to old age. In practice, however, such a situation is probably unlikely. The dosimetrically determinetd risk range of 1.03 to 1.36 for passive smokers u calculateci on the basis of the results of the studies cited above (2. 6. K, I1)t (t=iy. I). In addition. lunp cancer risk for female smokers in Austria has been included in the figture (22). Female smukers, cumh.u.LHc in age: to bronchial cancer patients, indicate an average daily cigarette cunsumptiun of 10 to 14 ciga- rett" 121). TI BU 31695

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 'rUl Rk ~ -7.3 - ,j 6 5 r 4 3 2 ; 2 3 4 5 1 4 1 10 I cig./day 23 Fio. I. Lunc cancer rink undrr pa»ive ~,muking cunditiun, i:. 6. 23. 191 linduding fumatt active smuktra in Auhtria (2_)1 und Jaily Ogurette cunsumpuun culcutatcd in accordunce wuh labtc 1. The adjusted risk for nonsmoking women whose husbands smoked fewer than 20 cigarettes a day as determined by Gartinkel (6) is R = 1.37 and t'or women whose husbands smoked 20 cigarettes anti more per tlay. R = l.(kL. These figures are in agreement with the results of the present study; that is. the t:\ttu,ure of the;sc rassivcly smoking women to noxac is less than or equal to one actively ,mokcd cigarette a day. The absence of a dose-response relationship appears plausible, because the passive ,nwking exposure of never-smokers need not be in agrccmt:nt with the smoking habits of their spouses, a, has also been empha- 'flLed by Gartinkel. - According to Trichopoulos er a/. (19), nonsmoking wives whose husbands have ,tupped smoking have a risk ut' R = 1.8. Fur women whose husbands smoke fewer than 20 Liyarettes per tlay. the risk is 2.4. Those whose hu.h;rnds smoke more than 20 cigarettes a day hnve a risk ut' R = 3.4. These figures, which correspond to'_-i actively smoked cigarettes per day, are higher than the c:um- putecl range of the present study and are also higher than the figures reported by Gartinkel (6). The dose-response relationship presupposes that the increasing daily cigarette cunsumptirtn by .puuses must bear a relationship to the increasing exposure to passive smuking; that is. the greater consumption does not take place in the workplace or in public settings. There is no indication in this report of any evidence tur this assumption. I:t, therefore, would appear to be improbable. Hirayama Iti) has reported the risk borne by pas,ivzly ,muking women to be R = 1.6 when the husband was an ex-smoker or smoked fewer than 19 cigarettes a day and as R = 2.08 when the husband smoked more than 20 cigarettes per day. 'flte highest figures were ubtained tor passively smoking wives of smoking TI BU 31697 I

0 TAt3Lf: I c 0 LUNG C-ANCEK INC/DLNCL IN NEVEK-SAfOKEKS AND SMOKERS Of VAR1•ING A?tOL'N7S Of LIGAKL1IES PLR DAY FOK DSrFEKENi AGLS AND Ru.AiIvL RISKs (RR) FOK Ll'!aa C AN/ l K03) lncidence I4-.IKNl, n SmoF,crs Iag day) z z v c Never smoked A 1110 1/4 1:2 . 34 1 l 5 10 -4 g i 40 39 0 0.40 0.42 0A5 0.4'1 0.52 0.69 1.30 z 2.74 50 . 2.95 3.05 3?(1 3.46 3.73 4(ll 5.24 9.91 < 20.96 c 60 11 90 30 12 91 12 13 97 15.06 16.20 21.16 40.00 ria.63 -i 70 . 34.48 . 35.64 . 37.41 . 40.47 43.64 46.93 61.30 115.90 C '4S'0 n 79 75.21s 77.81 81.68 88.35 95.27 102.47 133.83 25'_.03 535.33 RR 00 1 03 1 1.08 1.17 1.27 1.36 1.78 3.36 7.11 -3 . . ~ ~ ° Caicutalrd accorJing Doll and Pe1o (5). ci L= a o~ C3 '° C~ '~ < t~ 11 Z G? G~C ^ n t''t Er I" -, r_-, > Gr_; C~ V ry G(:. ~. .- .. r. C.. .- . . I

I'RtVt•.NTIVE MI:UtCINIi 13, 6ytS-7t)4 111)84) Quantitative Aspects of Passive Smoking and Lung Cancer' CHRISTI:\tV VUTUC Instirure uJ Suciul .Nrdicine. Uttit•ersirv of Vienntt. KinderspituJ,>;usse 15, ,i-1U95 Viertn(j, Austriu The txposure ut' passive ,mukers to cigarette smoke is estimated to be equivalent to U.1- t.U ctguretteltlay actively smoked. According to the reported relationships ut du.e and Ume, lung cancer incidence and other relative risk tigures relating to nunamuker, have been calculated fUr ages 40, 5l), 60, 70. and 79. Risks for smokrrs with a ILulV Cria.umptiun ut U.I-1.U cigurcttr were t'uund to be in the range uf R = 1.03 to 1.36. As tt .Ippltc?~ to puaaive !~mukers, this range of exposure nwy be neglected bncuuat it has no major ctfect un twtg cancer incidence. The results ut luur previous studies dealing with passive ~,tnukinb and lung cancer are compared with the current calculated riaks, and the dltlerences are lir CLLSbett. 0 1964 Ac:ulemrc Prena. Inc. INTRODUCTION In 1981, the first studies dealing with the question of passive smoking and the risk of contracting cancer of thc lung were published. They comprised two cohort studies, one from Japan tg) aad the other from the U nitetJ States (6), and a case- cuntrol study from Greece ( ly). Ln all cases, the study subjects were women exposed to passive smoking. A t'urthercasc-control study from the United Statt:~ (2) was puhli,hed in 1983. The results reported in these four studies shuw t:on- sitJerable disagreement and are controversial; a number of authurs (13, 14, 15, (7) have adopted a critical attitude toward them-particulariy to the Japanese study. Based on data from a prospective study of the smoking habits and incitlent;c: of lung.canczr among British physicians, Doll and Peto (4) analyzed the duve- response relationship of carcinoma of thc: lung in regular smokers and in persons who had never smoked. They describe a method (5) for computing the incidence of bronchial carcinoma. On the basis of the incidence figures. risk fur bronchial carcinoma can also be computed for low exposure to nuxur (a, muy be assumed to be the case under conditions of passive smoking) (23). In the present study. the calculated risk figures are compared with the results reported in the studies just mentioned. DOSIMETRY On the basis of experimental and theoretical investigations. Repace and Luwrey (18) have quantitatively gstablished the passive smoker's exposure to noxae. De- pending upon the passive smoking situation, they computed an exposure to nuxaC for nonsmokers that correspuncJed to between 5 and 27 actively smukztl cigarettes per day. These figures ret'er, however, to cigarettes with a smoke condensate Itarl content of 0.55 mglctbarette, which should be: considered unrealistic. !f cumpu- ' Prenented at the Symposium "Mcdicat Perspectives on P:uaivc Smuking,•' April 9-12. 19K4. Vienna. Austria. 698 IH19I-7435/84 S3.1N1 topyribnr i. IvNy h. .{-Jennc Pro,.,. Int. VI nghu al repr,id"uun m~ny Wrm rcneneJ. TI BU 31694 s i I k t , t [

702 CHRISTIAN vUTUc agricultural workers (R = 3.17 t'or husbands who were ex-smokers or who con- .)untn;c! fewer than 19 cigarettes per day, and R = 4.57 for husbands smoking 20 or more cigarettes per tiay). The latter figure is higher than that determined for actively smoking wumen, who have a lung cancer risk of R = 3.8. In thia report, too, the risk is higher than the computed values of our study and those reported by Gartinkel (6). With respect to the dose-response relationship, the same re- marks apply here as to the Trichopoulous study. A risk of R=-t.57 corresponds to the polluting effect of about 7 actively smoked cigarettes a day. This is to he compared with the risk R = 3.8 for actively smoking women, which corresponds to a consumption of about 5.5 cigarettes per day. In an interview (9), Hirayama has said that women are exposed to a level of passive smoking corresponding to the polluting effect of from 5 to more than 20 actively smoked cigarettes. Com- puted in accordance with Doll and Peto's method (5). this represents a relative risk of more than 19.U--a rigure that cannot be explained ciusimetric:ally, because it corresponds to the risk of active anwi.crs with a very high daily cigarette consumption. By way of comparison, in Austria. the risk borne by female smokers is R = 7.3 (22). Hirayama did not indicate hum he arrived at these high cigarette equivalents. Although in agreement with the data of Repace and Lowrey (18), these latter figures are based on standardized cigarettes with an unrealistic tar content of 0.55 mg/cigarette. If these data are re-evaluated for cigarettes with a realistic tar content, cigarette equivalents of the order of 0.2 to I cigarctte per day are obtaineti. This corresponds to a real risk range of R = 1.03 to 1.36. Furthermore, a striking point in the Japanese study is that stan.t:u'diration in accordance with the alcohol consumption hahits ut' the husbands results in the loss of the marked effect on lung cancer risk in women. Women whose husbands drink alcohol only occasionally have a risk of R = l. 13; those whose husbands consume alcohol daily have a risk of R = 1. 18. As the initial results ot' a study carried out in the United Statzs, France, Scutland, the Federal Republic oC Ger- many, and Austria (16) show, smoking habit is frequently asbuciatcd with the alcohol consumption. If this also applies in Japan, us we assume to be the case, the risk for women ought to increase with an increase in their husbands' alcohol consumption. Since this does not appear to be the case, the husbands involved must have been either drinking and not smoking or smoking but not consuming alcohol. Currea rt u!. (2) have calculated that passiveby smoking men have a lung cancer risk of R=?.0; passively smoking women have a risk equal to 2.07. In women. a dose-response relationship was demonstrated: R = I.t;U for exposure to l-4l) pack-years and R= 3.52 iur exposure to 41 pack-years and more. The value K = 1.18 corresponds to the risk borne by an active smoker who consumes half a cigarette per day-a tigure that appears realistic. The value R = 3.52 correspunJ, roughly to the risk associated with 5 actively smoked cig arettes per day. This large discrepancy can be explained by the great ail'Cci4nce in the average tigurc, for pack-years in the two exposure groups. This dues not, however, appear very plausiblc; but ~uggests, rather. thut in the group "41 pack-years and mure," a confounding t'actor is falsifying the results. lndeed, the authors of this study emphasize that the results do not necessarily reflect the biological effect of paa- TI BU 31698 sive bezc rzm, in ti Wh:, art i f. tocU unaf smuis inrlit the ~. trihu ctzm: Furtt uccu: pas~.i band titati• d itt~t !n the d is, tt•, . u'acti 7. H... ~ S. H u 9. H1i

HIAL tU E. C. 4 tra- drcch iUnet1 brun- Caudc sf!!t. lit'ect they iticcl uncl. :n in ases e 57 k ing wn- ical han nUt =ad a SYV1POSIUNt: MEDICAL PEkSPECrIVES ON PASSIVE SMOKING TABLE 3 Nl•>i1/L•k UF WOMEN WITH LUNG C.{NCEk ANU NuMNfik Vf;RIFI(;D TO BE NONSMOKERS WITH CONFIRMED MICkuS/'UPIC PkUUF Hospital A 284 Women with lung cancer on hospital record 7i Were nurtsmukers or smoking habit was not ,tuted 37 No histologic pruol' or not lung cancer mlcroscupically ~0 Microscopic proof ut lung cana;er-smuker Ih Microscopic proof of lung c;utcer-nunsmuker Hospital B 114 Women with lung cancer on hospital record 31 Were nunsmukers or smoking habit was nut statet/ t I Nu histological proof or nut lung cancer microscopically 7 Microscopic pruul' uf lung cancer-smuker 13 Microscopic proof of lung eancer-nonsmuker Hospital C 277 Women with lung cancer on hospital record 104 Were nonsmokers or smoking not stated 69 Not yet processed 7 No histologic proof or not lung k;ancer microscopically 15 Microscopic proof uflung cancer-smoker 13 Microscopic pruuf of lung cuncer-=nunsmoker ubtaining detailed answers to questions on current and past exposure to the smuke of others. childhood smoke exposure, husbantl's smoking habit, and, tur those who smoke, the number of cigarettes smoked at home. Table 3shuws the number of cases that had to be reviewed to obtain verified case5 of nunsmui.ing women with lung cancer. in Hospital A. ':c4 women were diagnosed with lung cancer in the 11-year period. St;venty-tive of them either were rcpurtzcl to be nonsmokers or their smoking habits were not mentioned on the hospital record. After histologic revit:w of the slides, 37 either were lacking microscopic pru0l' uF lung cancer or were misdiagnosed. Most ut' the latter were metastutic cancers to the lung frutn another site. Twenty of those with micru- ,cupic prout'were confirmed tu bc either current smokers or cx-cigarettt: smokers at Jiagnusis; 18 were nonsmokers. ln Hospital B, ut' 114 lung cancers in I1 years, 31 either were reported to be nonsmokers or their smoking habits were not known. Eleven ut'them were lacking in microscopic proof or were misdiagnosed as lung cancer; 7 with microscopic pruuf wcrz smokers, and 13 were nonsmokers. We have not yet confirmed all lung cancer cuses in Hospital C, but we currently have 13 nonsmokers with confirmed microscopic pruut'uFlung cancer. In Hospital D, we have not yet accumulated enough cases to show similar distributions. in 1910. the American Cancer Society beban a new. l.u-.uc-scalce prospective ,tucfy: 77,000 volunteers enrolled l'0U,UUU persons, asking them to complete a cunttdcntial questiunnaire. They will trace these individuals tur a periud of at least n years. Among thc 400 questions on the ti)ur-page que5tiunnairc is it que5- tiun related to passive smoking. Sincu enrollment is by household, the ,uhjects ai,u will be cla,sitietl according to the smoking habits of their spouses. The

702 CHRISTIAN vUTUc agricultural workers (R = 3.17 t'or husbands who were ex-smokers or who con- .)untn;c! fewer than 19 cigarettes per day, and R = 4.57 for husbands smoking 20 or more cigarettes per tiay). The latter figure is higher than that determined for actively smoking wumen, who have a lung cancer risk of R = 3.8. In thia report, too, the risk is higher than the computed values of our study and those reported by Gartinkel (6). With respect to the dose-response relationship, the same re- marks apply here as to the Trichopoulous study. A risk of R=-t.57 corresponds to the polluting effect of about 7 actively smoked cigarettes a day. This is to he compared with the risk R = 3.8 for actively smoking women, which corresponds to a consumption of about 5.5 cigarettes per day. In an interview (9), Hirayama has said that women are exposed to a level of passive smoking corresponding to the polluting effect of from 5 to more than 20 actively smoked cigarettes. Com- puted in accordance with Doll and Peto's method (5). this represents a relative risk of more than 19.U--a rigure that cannot be explained ciusimetric:ally, because it corresponds to the risk of active anwi.crs with a very high daily cigarette consumption. By way of comparison, in Austria. the risk borne by female smokers is R = 7.3 (22). Hirayama did not indicate hum he arrived at these high cigarette equivalents. Although in agreement with the data of Repace and Lowrey (18), these latter figures are based on standardized cigarettes with an unrealistic tar content of 0.55 mg/cigarette. If these data are re-evaluated for cigarettes with a realistic tar content, cigarette equivalents of the order of 0.2 to I cigarctte per day are obtaineti. This corresponds to a real risk range of R = 1.03 to 1.36. Furthermore, a striking point in the Japanese study is that stan.t:u'diration in accordance with the alcohol consumption hahits ut' the husbands results in the loss of the marked effect on lung cancer risk in women. Women whose husbands drink alcohol only occasionally have a risk of R = l. 13; those whose husbands consume alcohol daily have a risk of R = 1. 18. As the initial results ot' a study carried out in the United Statzs, France, Scutland, the Federal Republic oC Ger- many, and Austria (16) show, smoking habit is frequently asbuciatcd with the alcohol consumption. If this also applies in Japan, us we assume to be the case, the risk for women ought to increase with an increase in their husbands' alcohol consumption. Since this does not appear to be the case, the husbands involved must have been either drinking and not smoking or smoking but not consuming alcohol. Currea rt u!. (2) have calculated that passiveby smoking men have a lung cancer risk of R=?.0; passively smoking women have a risk equal to 2.07. In women. a dose-response relationship was demonstrated: R = I.t;U for exposure to l-4l) pack-years and R= 3.52 iur exposure to 41 pack-years and more. The value K = 1.18 corresponds to the risk borne by an active smoker who consumes half a cigarette per day-a tigure that appears realistic. The value R = 3.52 correspunJ, roughly to the risk associated with 5 actively smoked cig arettes per day. This large discrepancy can be explained by the great ail'Cci4nce in the average tigurc, for pack-years in the two exposure groups. This dues not, however, appear very plausiblc; but ~uggests, rather. thut in the group "41 pack-years and mure," a confounding t'actor is falsifying the results. lndeed, the authors of this study emphasize that the results do not necessarily reflect the biological effect of paa- TI BU 31698 sive bezc rzm, in ti Wh:, art i f. tocU unaf smuis inrlit the ~. trihu ctzm: Furtt uccu: pas~.i band titati• d itt~t !n the d is, tt•, . u'acti 7. H... ~ S. H u 9. H1i

708 JOHNSON AND LETZEL rule summarizes the entire life ut' an indivillual on the uC ,~tu•rent exposure con. ~tatus: what one says about one's life today may not be an'alleyuate summary of Lt ut' what has taken place in the past. What is needed, therefore, is a method of so i assessing the exposure that has occurred over the entire life of the individual l being interviewed, the Such a method has been proposed by Wynder and his coworkers at the Amer- pro ican Health Foundation. As can be seen in Fig. 3, the questionnaire divides the the life of the individual being interviewed into three time frames: prenatal, first 16 por years, and adulthood. Within these time frames the questionnaire focuses upon L;yu relevant exposure situations. Over the first 16 years of life, this exposure takes place primarily in the home. During the adult phases of life, relevant exposure C"` can occur at home, at work, on public transportation, and/ur during leisure time t activities. the We have drawn this figure such that the size ut' each rectangle indicates the tha emphasis andlor detail of the infurmatiun collected concerning this life segment eq1; and exposure situation. This questionnaire places substantial emphasis upon the exr first 16 ycars ut' life and also allows for relevant exposure in situations other than in i. in the home. Using this scheme one can incurporatl; a broad spectrum of lit'e in t Thi int, FETUS _ FIRST t6 YEARS OF LIFE NOME I AnuLT PHASE OF LIFE HOME WORK iRANSPORT LEISURE TIME Fk:. 3. I)istributiun ul'yur,tiuns in the WynI,1er yueitiunnuire. i TI BU 31704 hut Vc, re r tei Ptb

696 t.AWKENCE GAftFINKEL. question is designed to determine thi: amount of smuke to which a person i~ expused. It reads: Whether or noi you smoke. on the avcrage, how muny hours a Jay are you exposed lu hmoke from uthers:' At Home At Work In Other Arnas . studies have been based on relatively small numbers of nonsmoking lung cancer cases. It is important to emphasize that smoking habits need to be verified and pathological diagnoses of lung cancQr have to be confirmed. In our own study we hope to accumulate up to I(X) ca,cs before making an analysis. As yet, no con- clusions have been drawn by us as to whether or nut passive smuke exposure i, related to IunL cancer. Continuing research will provide the inturmation needed to do so. Preliminary analysis of the answers to this question shows that. among non- smuking women. 3 1% reported they were not exposed to snwke at all, 25% were exposecd only occasionally, 15% were exposed 1-7 hr a day, und 29% were ex- posed 8 hr or more a day. In the case-cuntrul study in Hospitals A and B, preliminary data indicate that 64% of the control women reported they were not exposed to passive smoke at all, 25% were exposed 1-9 hr a week. 6% were exposed 10-39 hr a weck, and 5% were exposed 40 hr or more a week. These figures are very similar to those reported by Friedman et ul. (5). Ot' the total number of hours during which the control subjects were subject to smoke of others, 75% was in the home, 20% was at work, and 5% occurred in other areas. We also classified nonsmoking women according to the average percentage ut the total amount of their husbands' cigarette smoking that occurred in the home. P%liminary data from the lung cancer group showed that an average of -t1% of the cigarettes smoked by husbands of nonsmokers with lung cancer were smoked at home, compared with an average of about 465''o of cigarettes smoked by hus- bands of the colun/rectum cancer cases. Thus, it is apparent that in our U.S. tudies, classification of nonsmokers by the number of cigarettes smoked by a spouse is only a crude estimate uf the degree of passive exposure. The method for measuring passive stnuking is an important consideration. It should also be recognized that the effects of passive exposure are highly related to the proximity of the smuker, the amount of ven- tilation in the room, the number of smokers in the ruom, and the number of hours exposed. In atltlitiun, smokers themselves inhale a great deal more passive smoke by virtue of their proximity to their smouldering cigarettes than do others who may be in the same area. Another point to be considered in relutiun to the biologic effects of passive smoking is the lack of evidence at autopsy of atypical cells in the lungs of nonsmokers. At this point there are both positive and negative studies in the United States on the issue of passive smoking and lung cancer. More data are needed antd will be published by our group and by others. Most of the published case-contrul IU ClbarCtte lntu(\Inb. Iy55-t10 V-,. I97U-77. NPw L:IIgl. f. ,Ncd. 3011. 1FiI-3HA /I'1791. I. Auerhach. 0.. Gartinkal. L.. and Hcunmund. E. C. Change, in brunchial epithrlium in rtlauun REFERENCES TI BU 31692 2. Auerbal !,ntuk 3. AuCrbaL. in ret. (1962 4. Correu. and IL 5. Frieilm:, A,lrr, n. Gartink. ~muk 7. Hammu c'al,< t!. Hammu in thc 9. Hammu ttr,l. 10. Kabat. ~ I t. Kahn. I and .: 12. Kauttm. ii~k t 13. U.S. Pu uHE 14. White. tuhu.

704 CtlltlsrlAN b t; rUC 14. Lee. P. N. Passivr Smoking. Fur,J Chrr+r. Tu.tie oL 20, '_.3-._9 1 Ic1N_'1. 15. Lehnert. G. Krank durch Pa..,rvrauchen° Slurnclr. .Ned. Wl,clrrrteclu•. 123, 147i5-);gl{ I1y241). 16. Lynch. C. J. "lntnrnatiunul Epidemtuluglc Study ut ttta Relationship between Smoking and Lung Cancer: First "rwu Years of Survr~ in We,tcrn Europe. Interim Report Smoking and Health Prugnwt." Vols. l. ll. National lanccr tn,utute. Bethesda. Mel., Dec. 31. 1979. 17. i4fucDunalJ. E. Nun-smoking wives uf heavy ,mukcrs have a higher risk ul lung cancer (letter(. Brit. ,1-leJ. J. 283, 914 (1y81). 18. Rcl'.M;e. J. L.. and Luwrey, A. H. Indoor air pulluttun. tobacco smuke, and public health. Scie,rrr ( Wuslrirepruaj v.C.) 208. 464-472 /1980/. 19. Trichupaulus. D.. Kalantlitli. A.. Sparrus, L.. and MacMahon. B. Lung cancer and pa,eivr smuking. Irtr. J. Cune•er 27, 1-4 (1981). 20. U.S. Public Health Servtce. "The Changing Cigarette. A Report (lf the Surgeun-GCnCr;el" p. 230. U.S. Department %,t liealth Education and Welfara. DHHS Publ. No (PHS) KI-5U156, 1981. :21. Vutuc. C. anJ Kunte, M. Brunchuskurzinum: F:pieltmiulubioche Gruntlluyen. in "Suualmeth/tn des Sronchuakurzinums". ( M. Kunze and L. Vutuc. EJs. ), p. 53. Fucultus Verlag, V irnna. 1980. :22. Vutuc, C. Bronchuakarztnum und Zitwrettenkunsum bei Fruuen (Ergebnisst eincr Fall-Kuntrull• SluJiel. Zerttrulbi. BuRteriul.. PurusitevekJ.. In1, 6reuwtskr. Hy,q., Abt. 1: Uri,4.. Rriler B 176, 329-338 (1982). 23. Vutuc, C. Lungcnkrebsritilku unJ Passivrauchcn: Quantitative Uberlegungen. Zerrtrulbl. Bukrr• riul.. PunwitertkJ.. lrrjekriurr.,6r. l1v,4.. Aht. 1: Or1,q.. R.•tlrr 11 177, yU-y5 (1983). 24. Wald. N.. Doll. lt., and Cupcland. G. Trends in tar. nicotine, anel carbon monoxide yields ut U.K. cigurettc manufactured since 1935. Brit. .Noel. J. 282. 763-765 11981). NKL•VL.r1I PC M Prc an>J t pu~,cf. huvc: furth in th. Hirs, with with atiun been ~,c:nt~ Cllck' lur I TI BU 31700

706 Jt)HNSUN AND L.F.TZEL SOME EMPIRICAL MEASURING TECHNIQUES status. While smoking history is obviously relevant to a classification schcmc, information occurs because a series ut' population selection facturs has heen in- cluded in the definition. These include smoking histury, age. szx. and marital Figure 1 ha, been cunsu•uctecl to draw attention to the t:Lct that this loss ut have been located would represent an unaccuntttble loss of information. cancer patients will be relatively rare, inability to classify such cases once they in which this hierarchical classification rule can be represented. Using this rule on the German population as a whole we can identify 3.5"' ,t, exposed individuals and 9.1% as not exposed. Unfortunately, 50.7% of the population cannot be classitieEi. For this reason the rule seems inappropriate fur a representative ex- aminatiun of exposure. In addition, this loss of int'or,matiun would szriuusly hand- icap a retrospective epidemiological stuEly designed to examine the relationship between lung cancer and passive smoking exposure. Since nonsmoking lung of h:tssive smoking with the Hirayama definition. Figure I depicts another way It is appropriate that we begin uur analysis ot' the Cpiclemiulue:ical definitions Fta. 2. i InAqpi I~ I ~ + ~ SMOninO 3.5 q ( II ~tl I Il aeELaSSIFI[Dj ~ iiI'IIIf PExmns I ~ \`~- NA81r Of ~ rrr j9.L .! I _ - 50.1 : !wJ •an;aFIEO 9Y nlN.lrA.-.l Ftu. I. An apprutimute rccurtaructiun at Hirayama', Jctintuun ul pasave ,mukin4 applicJ tu Ihr Gcrman pupulatiun. TI BU 31702 the oth htatus. yue,ti, lllvtlle. ~,mu{:c:' l:labsltl the hu• roum,. "not Q and th tndlvtt.: w hu rt. atlEl. : U~,L eieht _'-5.3`. t11tWn numht.

SY,ti1N(ISIU:tit: MEll1CAL PEkSPI:C'I'IVF.S ON P:ISSIVE S,blOK1NG 707 " I PESENrIY 16T/. I Acr,vE ~ NEYEA~tvr \~~ FOAMEAtr t~ 3•.: \\\ i1fl \ 1 i 11 III I ill ~L 'SMOREA,MfME rf.5 ~~a -~ rfS E L SAME = SAM E 1 nauS EMOlO I H, OuS E Ol0 I ~ i ~ I NO II I il ` I .~~- , ~ li OGiEN iN I YES ~., OGrEN ,M { YES_ I '1, 'SMOKY NOOMS I SMOaY a00M9 { , ~ INO 1 ' ~~ NI ~ll I ~II ~'25.4 :i 6.0 12.b Z j8.J 1j 35 J S A 8 C 0 E , ` . 25,5 L EXPOSED tB • 0) a.0 : .vBt EXPOSEB tA • 0 Fu,. 2. Preliminary detinitiun: three•yue,uun cla»iticatiun rule oppiicd i., the Gernt:ut pupulatwn. the other selection tacturs are less clearly related to the clasiit'iLatiun ot'expusure .tatu.. As an alternative to this definition, we have been working with the tturc:e- yuestiun cla,siticatiun scheme shown in Fig. 2. Here we see the smoking history dividetd into tVce categories: "nevrr smoked." "prcviuus ,muker," and "currznt ,.muker." For those individuals who are nut currently smuking, two additional cla,:,iticatiun questions are used. The tirst determines whether a,ntukzr lives in the home. and the second determines whether the indiviJ'ual is utien in smoky ruums. inEiividuals who answer no to both of these yutstiurt, are clabsitiad us "nut exposed." Since we differentiate between those who have never smoked and those who have previously smokeEi, we have two classes of not exposed individuals. These have been given the group designations A and C. An individual whu responds altirmatively to one ur both uY the questions is t;Iaa,i.Fizd aS exposed anJ. depending upon previous smoking hi,tury, is assigned to group B or D. Using this definition we are ahle to classify all nunsmuking, ituliviLluals. Thirty- riLht percent ui"the population is classitieJ as nut cauo.:J using this rule, while '_i,11'; receives an -expused" cla.,it'icatiun. Although our three-question deti- ttttiun has the advantage of ctassttying all of the nonsmoking population. it has a number of important ,hurtcomings, the most important being the fact that the TI BU 31703

SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 'rUl Rk ~ -7.3 - ,j 6 5 r 4 3 2 ; 2 3 4 5 1 4 1 10 I cig./day 23 Fio. I. Lunc cancer rink undrr pa»ive ~,muking cunditiun, i:. 6. 23. 191 linduding fumatt active smuktra in Auhtria (2_)1 und Jaily Ogurette cunsumpuun culcutatcd in accordunce wuh labtc 1. The adjusted risk for nonsmoking women whose husbands smoked fewer than 20 cigarettes a day as determined by Gartinkel (6) is R = 1.37 and t'or women whose husbands smoked 20 cigarettes anti more per tlay. R = l.(kL. These figures are in agreement with the results of the present study; that is. the t:\ttu,ure of the;sc rassivcly smoking women to noxac is less than or equal to one actively ,mokcd cigarette a day. The absence of a dose-response relationship appears plausible, because the passive ,nwking exposure of never-smokers need not be in agrccmt:nt with the smoking habits of their spouses, a, has also been empha- 'flLed by Gartinkel. - According to Trichopoulos er a/. (19), nonsmoking wives whose husbands have ,tupped smoking have a risk ut' R = 1.8. Fur women whose husbands smoke fewer than 20 Liyarettes per tlay. the risk is 2.4. Those whose hu.h;rnds smoke more than 20 cigarettes a day hnve a risk ut' R = 3.4. These figures, which correspond to'_-i actively smoked cigarettes per day, are higher than the c:um- putecl range of the present study and are also higher than the figures reported by Gartinkel (6). The dose-response relationship presupposes that the increasing daily cigarette cunsumptirtn by .puuses must bear a relationship to the increasing exposure to passive smuking; that is. the greater consumption does not take place in the workplace or in public settings. There is no indication in this report of any evidence tur this assumption. I:t, therefore, would appear to be improbable. Hirayama Iti) has reported the risk borne by pas,ivzly ,muking women to be R = 1.6 when the husband was an ex-smoker or smoked fewer than 19 cigarettes a day and as R = 2.08 when the husband smoked more than 20 cigarettes per day. 'flte highest figures were ubtained tor passively smoking wives of smoking TI BU 31697 I

on is U non- were e Zx- : that keat and hu,e I the Was ge u€ urne. `•~ uf okc:cd hus- -s by t the : ~- an ,'ivc v4n- uurs noke who loSic Is in SYV(P()titlJ41: vIEDICAL PtiRSPECTIVES ON PASSIVE SMOKING 697 2. Auerbaeh, () . ~icwt. A. P.. Hammund, E. C., and Gartinkel. L. l3runchiul epithelium in (urmer -3mukers. Vccr Lntl. J. Myd. 267, 1 IN-125 11962). 3. Auerbach. 0.. Stuut. A. P.. Hammond. E. C.. and Gurtinkel. L. Changes in bronchial epithelium in relatwn to ,ex, age, residence. smoking anJ pncumunra. Vrsr Eir;l. J. Mrd. 267, 11I-I IiS (I'ki'-). 4. Currea. P., Pickle. L. W., Fontham. E.. PiLklc, W., Lin. Y.. and Haenstel, W. l".c„ive smoking and lung cancrr. Lunc•at 2, 595-597 11'!tl3). 5. Friedman. G. U.. l'euut. I). 13., and Bawul. R. D. Prevalence and currelutes ut Iw„ive smuking. Airrrr. J. Public lfrulNr 73, i01-4U5 /1y2f)). h. Gartinkd, L. l'ime trends in lung cancer mortality among nonsmokers and a note on passive smoking. J. Nurl. Cunrer /n.st. 66, 1061- I llhb (1981). 7. Hammond. E. C. Smoking in relation tu the denth rates ufune million men and women. Nud. Cuncrr ln.sf. Nlucrcpqr. 19, 127-204 11966). S. Hammond, E. C., and (;artinkei. L. Curonury heart clisen.e. ,truke. and aortic aneurism: Factors in the etiology. Are'h. Etmirua. Heulth 19, 167-IK'- I 1%'ll. y. Hammond. E. C., and Gurtinkci. L. General air pollution anad cancer in the United States. Prcn•. L/c•c/. 9, 2Ub-21 1 (1980). 10. Kahat, G. C., and WyntJer, E. L. Lung cancer in nonsmokers. Cunc•rr 53, 1214-1221 (1984). I I. Kahn. H. A. The Dorn Study ut' smoking and mortality among U.S. veterans: Report on eight and une•half ol' oh.ervatiun. Ncetl. Cancer btev. ,tlucru,qr. 19, 1-126 119661. 12. Kaullmunn, F.. Ic,.icr. J. F., and (hiuL W. :Ndutt passive smoking in the home environment: A risk factur tur chronic airtlow limitation. Airrer. J. lipidecnia(. 117, 269-?M0(1'J233). 13. U.S. Public Health Service. "The Health Cunxyuences ul Smoking: A Reference tclitiun." U.S. UHEW, U.S. Govt. Printing Oftice. Wu,hingtun. D.C.. 1976. 14. White, 1. R., and 1•rueb, Il. I• Surall airways Jysfunctiun in nonsmokers chronicy lly expuseJ to tobacco smoke. Vr+r @'n;f. J. MeJ. 302, 720-723 c lytf0): TI BU 31693

708 JOHNSON AND LETZEL rule summarizes the entire life ut' an indivillual on the uC ,~tu•rent exposure con. ~tatus: what one says about one's life today may not be an'alleyuate summary of Lt ut' what has taken place in the past. What is needed, therefore, is a method of so i assessing the exposure that has occurred over the entire life of the individual l being interviewed, the Such a method has been proposed by Wynder and his coworkers at the Amer- pro ican Health Foundation. As can be seen in Fig. 3, the questionnaire divides the the life of the individual being interviewed into three time frames: prenatal, first 16 por years, and adulthood. Within these time frames the questionnaire focuses upon L;yu relevant exposure situations. Over the first 16 years of life, this exposure takes place primarily in the home. During the adult phases of life, relevant exposure C"` can occur at home, at work, on public transportation, and/ur during leisure time t activities. the We have drawn this figure such that the size ut' each rectangle indicates the tha emphasis andlor detail of the infurmatiun collected concerning this life segment eq1; and exposure situation. This questionnaire places substantial emphasis upon the exr first 16 ycars ut' life and also allows for relevant exposure in situations other than in i. in the home. Using this scheme one can incurporatl; a broad spectrum of lit'e in t Thi int, FETUS _ FIRST t6 YEARS OF LIFE NOME I AnuLT PHASE OF LIFE HOME WORK iRANSPORT LEISURE TIME Fk:. 3. I)istributiun ul'yur,tiuns in the WynI,1er yueitiunnuire. i TI BU 31704 hut Vc, re r tei Ptb

SYtitPOSllJA: 4cEDll'AL PEkSPLI IIVLS ON PASSiV[ SMUKING 709 0 contiguratiuns. A large body uFdetailE;J information can be obtained for a number ut differing but relevant expusurc ,ituatiuns. It can be. applied to every individual so that, in theory, everyone can be ~Jas,itizd. Two issues, however, remain unresolved: specifically, it is unclear how to score the exposure. and there seems to be no way to validate the results. The scoring problem occurs because, as outside ubservers. we cannot know how to weight the various c:,~pusure components in this questionnaire. Do work, home, trans- rurtatiun. and leisure time exposures that occur in the same time frame all count cqually'' If not how should they be wzighted'•' Cunre pt n/' ,L!-"I'iirre Ei pusure One way to solve the scoring problem is to allow the interviewee to estimate the exposure intensity independent of the specific lite events that contribute to that exposure. 'l'u do this we divide the time interval of interest into a series of equal time intervals and ask the intlivi(luals to provide a global asussment of the exposure received in that time interval. We have used this technique, for example, in interviews designed to examine the exposure which an individual has received in the last 24 hr. The left panel of Fig. 4,hows the results of such an interview. This data matrix can be read as an exposure profile during the 24 hr prior to the interview. The ruw, are divided into 24 one-hour segments. For each of these hour-long scgment,. the individual has provided an estimate ut'the exposure that occurred. ranging trutn nuuC at all to strongly exposed. A, yct we have not been able to account for the probability that the inEiivitiual repurttnb ;ome exposure in an interval was nut ,u exposed over the entire in- te rval. Counting this interval as rc presenting 60 min of exposure will tend to T~rf rE~t 0 1, 2 3 S+~ Y 6 e ~9 :0r x k EAPOSED M-iIIE OF A PERSON (t" 1 PERCENt EXPOSED K-iItE OF A PERSON E+PDSED I V , I/1 1 2 3 ~/. I EtPOSEO 01 1 1 2 l v r \ I ~ ) 1 1'/ PPA'SON t1QWS ~ 9 PEp~1 t10uA5 ~7 '7 ~---~ ~ EaPUSED 012 I rr 1 / 1 r 1 0 PERSON HOtit5 t_--7 A EY. 3/5'/. OD% Flt.. 4. Maximum individu;il cxpu,ure ume a, a funtaiun of cuunting rulc. TI BU 31705

PKI~.YI{N I-IVP. Nh.llll INP. 13. 7115-7Iti ( I9tS«t1 I 1yH 11. intt Lung d Helth Measuring Passive Smoking: Methods, Problems, and Perspectives' L. C. 1AMNSUN' AND H. W. LET"LEL p. 230. 1981. lmccltcin Vienna, :untrult- ti 176. . Brl/irr- 0t U. K. -Grsr!!scltnjl /ier Ifr/urrnutiuns-vrrahriftrrtr nnd.5'ttui.wik in der Y/etliziu r.V., Pertertku)i•rslrusse 15, U-NIIW blurticlt 2. Fedrrul Rrpuhlic• uj Grrawrrv The definition of passive smoking used by T. Hirayuma [Brir. ~tiled. J. 282, t83-t85 119iS11] and other authors hus ;rt least three mujur ,hurtcumingh: it is only applicable to a highly sclective sub-pupulatiun: varying lifCstylC patterns cauu.rt be taken into account: and exposure tlutsttttl the home is neglected. To overcome thele NhUCICUmIngS, a preliminary qualitative and Cla]5ltlcatory definition of pilNStve smoking that Inclulle5 the total population as well as any smoke exposure. regardless ul' its lucauun, wu, developed. 'I'hu Jetinitiun dues nut include the yuantity of exposure or changes over unte. liasr:d on a representutive survey. 17.84'r of the population over 35 years are potential pusaive .ntukCrs according tu thia definitiun. l'he questiunnairr developed by E. l.. Wyndar and aswciate:+ rAmCrican Health Foundation. New Yurkl pre.Cnu Jifticultir, in .curing the reaults fur different life- style!,. Theretiure a yuantitative concept was cleveiupeJ fur estimating erru.utc over any period ut time aa a t;rncraltcetl assessment instrumrnt. The cun.r.l-t ul maXtntwn c\puaetl 4t•time I'ur an individual 1/ i'1 a+ well ab !ur a gioup 1T;Sl is i/u, rrduceJ. The results are prCbenttd graphically by a cumuL•juvu ,tanaardizrd exposed M-ttmC diagt:un. ttCaults ub- tatncd -,u far Iean toward plausibility and stability uf the data and the concept. The interview form hus still to bn validated. [ Ivur Ac;,Jem/c t+re... In.. INTRODUCTION Previous epidemiological tletinitiuns of passive smoking have been pragmatic, antl tt1 ,t1me extent simplistic. Nonsmoking women married to smokers are ex- p,r~ctl tu ,tnukt: through passive inhalutiun. while those married to nonsmokers have been considered not exposed. Our research group is interested not only in turther epidemiological resrarch, but also in estimating passive smoking exposure in (he German population. The tletinition of passive smoking exposure used by Hirayamu 1 l f is inatlcaluate for our objectives tor three reasons. First, it deals with a highly selective -,ubset of the general populatiun. Seconct, it is intlexible with regard to varying lifestyle patterns. And finally, it fails to account for situ- atiuns outside of the home in which exposure can take place. We have therefore been working on the development of a generalized a„ussment instrument. Pre- sented here is an overview ut' our method, a summary of the prutilcros we have encountered, and some insights into the perspectives and/or opportunities wC ,Ce for turther retinement. . ' Presented at Ihe Sympusiutn "MCdical Perspcctivas on P.tssive Smoking." April y-12, 19144, Vienna, Austria. : lu whom reyuratu Fur reprints ,huuW be adtlreabcJ. 705 lN91-71 i i/2td S3.1N1 l,q,ynghr - 1984 hv I`:.,. Inc. \I1 ii.hl, ul r~pr.Nrwuun in ao~ L.un rcaorvrJ. TI BU 31701

i SYMPOSIUM: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 699 tations are based on a cigarette with a tar content of 16 mg, which roughly corresponds to the average tar content of cigarettes in the United States (20), in Great Britain (24), and in Austria ( I), tur the passive ,muking situations indicated by fZepace and Luwrey, we obtain cigarette equivalents of ihe order of 0.2 to I cigarette per day. The tigures t'ur cigarette equivalents reported by other authors are also found to lie within this range; Klusterkotter and Gono (12): 0.32 cigarette/ ti hr; Hugud et al. 1 t0): 0.5 to I cigarette/day. For the present study, this range was taken as the actual level of exposure to noxae of nonsmokers under passive ~muking conditions. LUNG CANCER RISK In Table I. the incidence of lung cancer per 100,000 is ihown for never-smokers and for smokers who have smoked varying numbers of cigarettes per day. The incidence is reported for dit'ferent age groups, and the relative risk oF smokers compared with nevcr-smokers has been calculated (23). DISCUSSION As indicated above, the exposure to noxae under passive smoking conditions is comparable to the exposure to nu.rae represented by one actively smoked cigarette per day. ln the present investigation, therefore, the results for smokers in this exposure range have been correspondingly tpplied to passive smokers. The inference made by Doll and Peto (>) was developed on the basis of data obtained from studies uf ma:n. Since there is no evidence to suggest that the risk uf contracting bronchial carcinoma is affected by ,ex-specitic facturs, this pru- cetiure may legitimately also be applied to women. A., the present analysis shows, in the range of.small doses (exposures), carcin- ogenic processes manifest an approximately ltnear dose-response relationship (3) for the course of the risk in the range 0.1 to I cigarette/day. The question of whether a threshold value exists (below which there is no risk) cannot be an- ,wered. However, the risk is expressed by such low tigures that the range for very small exposures can be considered "safe." For a Pcr,on exposed to passive ,muking (U.l to I cigarette/day), compared with a nonexposed prr,un, the risk lies within the range 1.03 to 1.36. Such a risk may be ignored. since in comparison with the greater risks, it has no appreciable effect on lung canccr incidence t'l:Lble I). The intluence becogtes more marked only in the higher age gruups, where the "ircinogenic effect of tobacco smoke is very strongly•intluenced by the eJtu'ation ut cxposure (5). Since risk is also dependent upon age at initial exposure (7, 11, '-a. a more marked effect of passive .ntuking is conceivable when exposure is cuntinuous from childhood to old age. In practice, however, such a situation is probably unlikely. The dosimetrically determinetd risk range of 1.03 to 1.36 for passive smokers u calculateci on the basis of the results of the studies cited above (2. 6. K, I1)t (t=iy. I). In addition. lunp cancer risk for female smokers in Austria has been included in the figture (22). Female smukers, cumh.u.LHc in age: to bronchial cancer patients, indicate an average daily cigarette cunsumptiun of 10 to 14 ciga- rett" 121). TI BU 31695

704 CtlltlsrlAN b t; rUC 14. Lee. P. N. Passivr Smoking. Fur,J Chrr+r. Tu.tie oL 20, '_.3-._9 1 Ic1N_'1. 15. Lehnert. G. Krank durch Pa..,rvrauchen° Slurnclr. .Ned. Wl,clrrrteclu•. 123, 147i5-);gl{ I1y241). 16. Lynch. C. J. "lntnrnatiunul Epidemtuluglc Study ut ttta Relationship between Smoking and Lung Cancer: First "rwu Years of Survr~ in We,tcrn Europe. Interim Report Smoking and Health Prugnwt." Vols. l. ll. National lanccr tn,utute. Bethesda. Mel., Dec. 31. 1979. 17. i4fucDunalJ. E. Nun-smoking wives uf heavy ,mukcrs have a higher risk ul lung cancer (letter(. Brit. ,1-leJ. J. 283, 914 (1y81). 18. Rcl'.M;e. J. L.. and Luwrey, A. H. Indoor air pulluttun. tobacco smuke, and public health. Scie,rrr ( Wuslrirepruaj v.C.) 208. 464-472 /1980/. 19. Trichupaulus. D.. Kalantlitli. A.. Sparrus, L.. and MacMahon. B. Lung cancer and pa,eivr smuking. Irtr. J. Cune•er 27, 1-4 (1981). 20. U.S. Public Health Servtce. "The Changing Cigarette. A Report (lf the Surgeun-GCnCr;el" p. 230. U.S. Department %,t liealth Education and Welfara. DHHS Publ. No (PHS) KI-5U156, 1981. :21. Vutuc. C. anJ Kunte, M. Brunchuskurzinum: F:pieltmiulubioche Gruntlluyen. in "Suualmeth/tn des Sronchuakurzinums". ( M. Kunze and L. Vutuc. EJs. ), p. 53. Fucultus Verlag, V irnna. 1980. :22. Vutuc, C. Bronchuakarztnum und Zitwrettenkunsum bei Fruuen (Ergebnisst eincr Fall-Kuntrull• SluJiel. Zerttrulbi. BuRteriul.. PurusitevekJ.. In1, 6reuwtskr. Hy,q., Abt. 1: Uri,4.. Rriler B 176, 329-338 (1982). 23. Vutuc, C. Lungcnkrebsritilku unJ Passivrauchcn: Quantitative Uberlegungen. Zerrtrulbl. Bukrr• riul.. PunwitertkJ.. lrrjekriurr.,6r. l1v,4.. Aht. 1: Or1,q.. R.•tlrr 11 177, yU-y5 (1983). 24. Wald. N.. Doll. lt., and Cupcland. G. Trends in tar. nicotine, anel carbon monoxide yields ut U.K. cigurettc manufactured since 1935. Brit. .Noel. J. 282. 763-765 11981). NKL•VL.r1I PC M Prc an>J t pu~,cf. huvc: furth in th. Hirs, with with atiun been ~,c:nt~ Cllck' lur I TI BU 31700

712 JOHNSON AND LETZEL 7ds% 910% 0% 125%. 250N. 379% 500%. 62Y/. 750% 875%. 100Y. : OF 08SEevED INiEBVAL PERCEHt ExPOSED 11-TIME (PERSON - HOUAS) OF A GROUP (T" ) G Ta" : AYERAGE ExPOSUflE 3 12.2 : 1 1.68 Z 2 4.21 Z FIG. 6. e,, viewed ,,. an aver.lge. We also have some data that clcurly emphasize the potential tor instability. In one ut'-the three-question classiticatiun rules we maEle what we thought was a minor change: we changed the question "Are you frequently in very smoky ruums7" to, "Are you regularly in very smoky rooms?" As a result ut'this change of a single word, the proportion of individuals responding "yzs" dropped from 37 to a highly signiticant tran,itiun (see Fig. Il)1. This suggcbts that thc: ansessment of passive amuking can be strungly affected by apparently trivial changes in the questions or t'urmat of the intcrview. PERSPECTIVES FOR FUTURE WORK The scoring procedure outlined in the preceding section follows naturally from the ,tyle of questionnaire used. Actually, this method of passive smoking expo- TI BU 31708

i ~ MO ExPOSURE i i ZI, NuwER aF nauRt 0% 25% SO% 75% ~~ Z OF 08SERVED INtERVµ Fia. 7. Cumuiative. ,IunJardiad. erp6.ed M-time dias_tram. TM 713 • SYMNClSIU41: 1A[?DIC':'LL I'ERSPECI'[vL•'S UN PASSIVE StitUKING 8 6 v 2 a a 2 4 G a lu 12 14 (EvEM n0uA5) n = dBo r - 0.91 m u- Ftc,. tt. Stabilitv u[' T'it+core icuunting rule I1. 4 TI BU 31709 COUnfIMG RaLE 47.7 18.5 41.0

PKI~.YI{N I-IVP. Nh.llll INP. 13. 7115-7Iti ( I9tS«t1 I 1yH 11. intt Lung d Helth Measuring Passive Smoking: Methods, Problems, and Perspectives' L. C. 1AMNSUN' AND H. W. LET"LEL p. 230. 1981. lmccltcin Vienna, :untrult- ti 176. . Brl/irr- 0t U. K. -Grsr!!scltnjl /ier Ifr/urrnutiuns-vrrahriftrrtr nnd.5'ttui.wik in der Y/etliziu r.V., Pertertku)i•rslrusse 15, U-NIIW blurticlt 2. Fedrrul Rrpuhlic• uj Grrawrrv The definition of passive smoking used by T. Hirayuma [Brir. ~tiled. J. 282, t83-t85 119iS11] and other authors hus ;rt least three mujur ,hurtcumingh: it is only applicable to a highly sclective sub-pupulatiun: varying lifCstylC patterns cauu.rt be taken into account: and exposure tlutsttttl the home is neglected. To overcome thele NhUCICUmIngS, a preliminary qualitative and Cla]5ltlcatory definition of pilNStve smoking that Inclulle5 the total population as well as any smoke exposure. regardless ul' its lucauun, wu, developed. 'I'hu Jetinitiun dues nut include the yuantity of exposure or changes over unte. liasr:d on a representutive survey. 17.84'r of the population over 35 years are potential pusaive .ntukCrs according tu thia definitiun. l'he questiunnairr developed by E. l.. Wyndar and aswciate:+ rAmCrican Health Foundation. New Yurkl pre.Cnu Jifticultir, in .curing the reaults fur different life- style!,. Theretiure a yuantitative concept was cleveiupeJ fur estimating erru.utc over any period ut time aa a t;rncraltcetl assessment instrumrnt. The cun.r.l-t ul maXtntwn c\puaetl 4t•time I'ur an individual 1/ i'1 a+ well ab !ur a gioup 1T;Sl is i/u, rrduceJ. The results are prCbenttd graphically by a cumuL•juvu ,tanaardizrd exposed M-ttmC diagt:un. ttCaults ub- tatncd -,u far Iean toward plausibility and stability uf the data and the concept. The interview form hus still to bn validated. [ Ivur Ac;,Jem/c t+re... In.. INTRODUCTION Previous epidemiological tletinitiuns of passive smoking have been pragmatic, antl tt1 ,t1me extent simplistic. Nonsmoking women married to smokers are ex- p,r~ctl tu ,tnukt: through passive inhalutiun. while those married to nonsmokers have been considered not exposed. Our research group is interested not only in turther epidemiological resrarch, but also in estimating passive smoking exposure in (he German population. The tletinition of passive smoking exposure used by Hirayamu 1 l f is inatlcaluate for our objectives tor three reasons. First, it deals with a highly selective -,ubset of the general populatiun. Seconct, it is intlexible with regard to varying lifestyle patterns. And finally, it fails to account for situ- atiuns outside of the home in which exposure can take place. We have therefore been working on the development of a generalized a„ussment instrument. Pre- sented here is an overview ut' our method, a summary of the prutilcros we have encountered, and some insights into the perspectives and/or opportunities wC ,Ce for turther retinement. . ' Presented at Ihe Sympusiutn "MCdical Perspcctivas on P.tssive Smoking." April y-12, 19144, Vienna, Austria. : lu whom reyuratu Fur reprints ,huuW be adtlreabcJ. 705 lN91-71 i i/2td S3.1N1 l,q,ynghr - 1984 hv I`:.,. Inc. \I1 ii.hl, ul r~pr.Nrwuun in ao~ L.un rcaorvrJ. TI BU 31701

712 JOHNSON AND LETZEL 7ds% 910% 0% 125%. 250N. 379% 500%. 62Y/. 750% 875%. 100Y. : OF 08SEevED INiEBVAL PERCEHt ExPOSED 11-TIME (PERSON - HOUAS) OF A GROUP (T" ) G Ta" : AYERAGE ExPOSUflE 3 12.2 : 1 1.68 Z 2 4.21 Z FIG. 6. e,, viewed ,,. an aver.lge. We also have some data that clcurly emphasize the potential tor instability. In one ut'-the three-question classiticatiun rules we maEle what we thought was a minor change: we changed the question "Are you frequently in very smoky ruums7" to, "Are you regularly in very smoky rooms?" As a result ut'this change of a single word, the proportion of individuals responding "yzs" dropped from 37 to a highly signiticant tran,itiun (see Fig. Il)1. This suggcbts that thc: ansessment of passive amuking can be strungly affected by apparently trivial changes in the questions or t'urmat of the intcrview. PERSPECTIVES FOR FUTURE WORK The scoring procedure outlined in the preceding section follows naturally from the ,tyle of questionnaire used. Actually, this method of passive smoking expo- TI BU 31708

SY:tiIP+,,~tUht: MEDICAL PERSI't•.l I tVI:S ON PASSIVE SMUKING 703 ;h y .c h .r >. n ~ sive smuking, hut may putisibiy be the result of confuuntling tactors that had not been eliminated. The considerations presented in this study, together with critical remarks made by various uthcr authors (13-15, 17), permit the cunclusion that, in the puhlicatiuns citcd abuve, some of the results may be statistical artifacts. What particular factors tbias, conti>unding tacturs) may be responsible for such artifacts cannot he determined without an accurate knowledge of the study pro- tocol. especially as it relates to the collection and evaluation uf the data. A further unanswered question is whether a given passive smoker had been an active smoker at some time in his or her life. Such a situation might explain how the individual high risk figures would arise. As represented in the present study, all the evidence seems to indicate that the risk involved in passive smoking is at- tributable to a low dose aa't'ea:t, which has no appreciable iMlucnca: on the inci- dence of lung cancer and is not in conformity with the valucs for active smoking. Furthermore, it must be said that in this dose-response range, air pollution and occupational exposures must be considered minimal risk factors. This means that passive smoking exposutr Qannot be established merely on the basis of the hus- band's cigarette consumption (6), since: such consumption does not permit a quan- titative determination of the exposure and its temporal correlation with, or its tlttterentiatiun from, other noxac. In common with discussions on the question of air pollution and lung cancer, the discussion of passive smoking and lung %;ancer is charged with emotion and is, therta'ure, g(:ncrally overstated. Unfortunately, this situation leads to a dis- traction from the still unresolved, and greater, problem of active smoking. ' I REFERENCES 1. Austria IanuKWerke Ali. F'ucnllCnC MlttttlungCtl. vlennU. lyli. ~ 1 2. lurrea. R. Pickle. L. W., Funtham. E., Lin, Y., and Haenstel, W. f'assive smoking and lung cancer. Laupcrt 2. iy5-597 ( 19tt3). ~~ 3. Crusnp. K. S.. Huet. D. G.. Lanyley. C. H.. and Peto, R. FundamCnttsl carcinogenesis pri>Lesbes -und their imp(icatita,, lur low Juse risk assessment. Cuuru•er Ria•. 36, 2973-2979 I t97(+/. 4. Doll, k., anJ Neto, K. Mortality iu relation tu smoking: 20 yeur, observations on male tiritish dix:turs. Brir. ,Ni•c/. J. 2, 15_5-t53t+ 111)76). 5 . Doll. R., und Petu. K. Cigareue ,moking and bronchial carcinoma: Dose and time relutiunehips uswng regular ,muker. :+nd tiictunL nun-smukers. J. EphlensinL Cunuutuuty Nrulrh 32. 303- 313 119781. h. Gattinkel. L. Time trends in lung cancer mortality among nonsmokers and a note on passive ,muking. J. .Vutl. t:'un<'er hot. h6, IOti)-I(Ab 11981). 7. H:unmuuJ. t. C. Smoking in relation to the death races of one million men and women. Nurl. Cirnrer lnst. Slnntukr. 19, l27-2t)411yhb). 8. Hir:+}u,n:r. "r. N++a-,muktng wives of heuvy smokers have a higher risk of lung cancer: A study tram Japan. Brit. ,Wed. 1. 282, 183- IM5 (19tS 1). 'l, tiirayamu. T. Was spricht tur, was gegen die Schudlichkeit Lies Passivr,tuchens:' (interview). 41ucm ft. .bfed. Wuchrnst•hr. 123, t4b(ll-Idti3 (19M11. lu. Hupu.t C'.. Hawkins. L. H., .ind ANtrup, R Expo,ure uf pas,ive ,mukers to tobuccu ,muke ~un~tituem.. lrtt. .ar, (+ !)rrup. btrnrun. ffe•ultlt 42. et t 14714), II K.rhn. If. A. The llurn ,tuJy „( ,ntukin3 and murtality among U S. veterans. Report on eight anJ une-ha1C ye•art of observation. Nwl. Cuiner li,.rt, .lfurrr;.,;r. 19, t-125 I19hAl. 1:. Klu,tctkuter, W.. and Gunu, E. Zuni Pruhiem Jes t'ansivrauchen.. ZcnrrtrlhL &,ktvrk.l.. Yur,r .itendJ., /nlrr<rrunxkr. fh•t•. lthr. I' oriq.. Rriltr 8 162, 5t-6y (1N7b). I t Kurneg:ry. H. R.. and Kastenbuum. yt. A. Non-smoking wives uf heavy smokers have it higher rt,k ut lung cuncer Uetterl. 13rrt. Med. J. 283, 9t6-yt6 ( tytil l. TI BU 31699

SYMPOSiUNt: VIEDICAt. PEll>PE:CTIVI?S (1N P:\15IVF. SNtOK1N(i 711 I 1 uvvEA LIN1T Illr' ~ ~ (cOUNrcN6 NuLE U 2 EST111AfE 1 (COuNfING NuLE 2) 3 LONEN LINIT (COUNfING RuLE 3) AREA uHOER THE CURVE = TG" - PERCENTAGE EXPOSED M-riME (PERSON HOURS) FIG. 5. Group muxirtwm oxpuseJ ume 17,,) Vlewed en al1 area. be used t'or other forms ut exposure that have been measured by questionnaires (e:.g., other types of air pollution). to the next. The fact that the estimates rGmain constant acruss days ut i hc week be rxpectc;Ld that the exposure estimates will not dit'fc;r,ubstantially from one day ut the week in which the interview was conducted (1=ig. y). T-hc survey was conducted so that these subgroups were also represcntative. As a result, it can a,c:cund stability check we calculatctd the I~t valua:s tur t;ruups defined by day The foregoing analysis has been checked t'ur within-individual consistency. As The tact that this correlation is above 0.9 5uggc,ts that this is nut the case. calculation. A low correlation would indicate a highly random response pattern. Figure ti shows a,catter diagram tur the score pairs that result t'rum such a ., uhinJices yields score pairs which should be hi~:hly rclutc;d. The stabiiity of the T~index can he measured in two ways. First. one can divide the 24 exposure estimates into the odd and even hours ut' the day. Calculating Sruhilitv ,uggcsts that our ntcthud is ,upplying fairly stahle estimates. 12.2 X 4.21 Z ,

706 Jt)HNSUN AND L.F.TZEL SOME EMPIRICAL MEASURING TECHNIQUES status. While smoking history is obviously relevant to a classification schcmc, information occurs because a series ut' population selection facturs has heen in- cluded in the definition. These include smoking histury, age. szx. and marital Figure 1 ha, been cunsu•uctecl to draw attention to the t:Lct that this loss ut have been located would represent an unaccuntttble loss of information. cancer patients will be relatively rare, inability to classify such cases once they in which this hierarchical classification rule can be represented. Using this rule on the German population as a whole we can identify 3.5"' ,t, exposed individuals and 9.1% as not exposed. Unfortunately, 50.7% of the population cannot be classitieEi. For this reason the rule seems inappropriate fur a representative ex- aminatiun of exposure. In addition, this loss of int'or,matiun would szriuusly hand- icap a retrospective epidemiological stuEly designed to examine the relationship between lung cancer and passive smoking exposure. Since nonsmoking lung of h:tssive smoking with the Hirayama definition. Figure I depicts another way It is appropriate that we begin uur analysis ot' the Cpiclemiulue:ical definitions Fta. 2. i InAqpi I~ I ~ + ~ SMOninO 3.5 q ( II ~tl I Il aeELaSSIFI[Dj ~ iiI'IIIf PExmns I ~ \`~- NA81r Of ~ rrr j9.L .! I _ - 50.1 : !wJ •an;aFIEO 9Y nlN.lrA.-.l Ftu. I. An apprutimute rccurtaructiun at Hirayama', Jctintuun ul pasave ,mukin4 applicJ tu Ihr Gcrman pupulatiun. TI BU 31702 the oth htatus. yue,ti, lllvtlle. ~,mu{:c:' l:labsltl the hu• roum,. "not Q and th tndlvtt.: w hu rt. atlEl. : U~,L eieht _'-5.3`. t11tWn numht.

SY,ti1N(ISIU:tit: MEll1CAL PEkSPI:C'I'IVF.S ON P:ISSIVE S,blOK1NG 707 " I PESENrIY 16T/. I Acr,vE ~ NEYEA~tvr \~~ FOAMEAtr t~ 3•.: \\\ i1fl \ 1 i 11 III I ill ~L 'SMOREA,MfME rf.5 ~~a -~ rfS E L SAME = SAM E 1 nauS EMOlO I H, OuS E Ol0 I ~ i ~ I NO II I il ` I .~~- , ~ li OGiEN iN I YES ~., OGrEN ,M { YES_ I '1, 'SMOKY NOOMS I SMOaY a00M9 { , ~ INO 1 ' ~~ NI ~ll I ~II ~'25.4 :i 6.0 12.b Z j8.J 1j 35 J S A 8 C 0 E , ` . 25,5 L EXPOSED tB • 0) a.0 : .vBt EXPOSEB tA • 0 Fu,. 2. Preliminary detinitiun: three•yue,uun cla»iticatiun rule oppiicd i., the Gernt:ut pupulatwn. the other selection tacturs are less clearly related to the clasiit'iLatiun ot'expusure .tatu.. As an alternative to this definition, we have been working with the tturc:e- yuestiun cla,siticatiun scheme shown in Fig. 2. Here we see the smoking history dividetd into tVce categories: "nevrr smoked." "prcviuus ,muker," and "currznt ,.muker." For those individuals who are nut currently smuking, two additional cla,:,iticatiun questions are used. The tirst determines whether a,ntukzr lives in the home. and the second determines whether the indiviJ'ual is utien in smoky ruums. inEiividuals who answer no to both of these yutstiurt, are clabsitiad us "nut exposed." Since we differentiate between those who have never smoked and those who have previously smokeEi, we have two classes of not exposed individuals. These have been given the group designations A and C. An individual whu responds altirmatively to one ur both uY the questions is t;Iaa,i.Fizd aS exposed anJ. depending upon previous smoking hi,tury, is assigned to group B or D. Using this definition we are ahle to classify all nunsmuking, ituliviLluals. Thirty- riLht percent ui"the population is classitieJ as nut cauo.:J using this rule, while '_i,11'; receives an -expused" cla.,it'icatiun. Although our three-question deti- ttttiun has the advantage of ctassttying all of the nonsmoking population. it has a number of important ,hurtcomings, the most important being the fact that the TI BU 31703

i ~ Ra ExPOSURE i i ZI, NuwER aF nauRt 0% 25% SO% 75% ~~ Z OF 08SERVED INtERVµ Fia. 7. Cumuiative. ,IunJardiad. erp6.ed M-time dias_tram. TM 713 • SYMNC>SIU41: 1A[?DI(':'LL I'ERSPECI'[vL•'S UN PASSIV[_ Stit(>KING 8 6 v 2 a a 2 4 G a lu 12 14 (EvEM n0uA5) n = dBo r - 0.91 m u- Ftc,. tt. Stabilitv u[' T'it+core icuunting rule I1. 4 TI BU 31709 COUnfIMG RaLE 47.7 18.S 41.0

I'RtVt•.NTIVE MI:UtCINIi 13, 6ytS-7t)4 111)84) Quantitative Aspects of Passive Smoking and Lung Cancer' CHRISTI:\tV VUTUC Instirure uJ Suciul .Nrdicine. Uttit•ersirv of Vienntt. KinderspituJ,>;usse 15, ,i-1U95 Viertn(j, Austriu The txposure ut' passive ,mukers to cigarette smoke is estimated to be equivalent to U.1- t.U ctguretteltlay actively smoked. According to the reported relationships ut du.e and Ume, lung cancer incidence and other relative risk tigures relating to nunamuker, have been calculated fUr ages 40, 5l), 60, 70. and 79. Risks for smokrrs with a ILulV Cria.umptiun ut U.I-1.U cigurcttr were t'uund to be in the range uf R = 1.03 to 1.36. As tt .Ippltc?~ to puaaive !~mukers, this range of exposure nwy be neglected bncuuat it has no major ctfect un twtg cancer incidence. The results ut luur previous studies dealing with passive ~,tnukinb and lung cancer are compared with the current calculated riaks, and the dltlerences are lir CLLSbett. 0 1964 Ac:ulemrc Prena. Inc. INTRODUCTION In 1981, the first studies dealing with the question of passive smoking and the risk of contracting cancer of thc lung were published. They comprised two cohort studies, one from Japan tg) aad the other from the U nitetJ States (6), and a case- cuntrol study from Greece ( ly). Ln all cases, the study subjects were women exposed to passive smoking. A t'urthercasc-control study from the United Statt:~ (2) was puhli,hed in 1983. The results reported in these four studies shuw t:on- sitJerable disagreement and are controversial; a number of authurs (13, 14, 15, (7) have adopted a critical attitude toward them-particulariy to the Japanese study. Based on data from a prospective study of the smoking habits and incitlent;c: of lung.canczr among British physicians, Doll and Peto (4) analyzed the duve- response relationship of carcinoma of thc: lung in regular smokers and in persons who had never smoked. They describe a method (5) for computing the incidence of bronchial carcinoma. On the basis of the incidence figures. risk fur bronchial carcinoma can also be computed for low exposure to nuxur (a, muy be assumed to be the case under conditions of passive smoking) (23). In the present study. the calculated risk figures are compared with the results reported in the studies just mentioned. DOSIMETRY On the basis of experimental and theoretical investigations. Repace and Luwrey (18) have quantitatively gstablished the passive smoker's exposure to noxae. De- pending upon the passive smoking situation, they computed an exposure to nuxaC for nonsmokers that correspuncJed to between 5 and 27 actively smukztl cigarettes per day. These figures ret'er, however, to cigarettes with a smoke condensate Itarl content of 0.55 mglctbarette, which should be: considered unrealistic. !f cumpu- ' Prenented at the Symposium "Mcdicat Perspectives on P:uaivc Smuking,•' April 9-12. 19K4. Vienna. Austria. 698 IH19I-7435/84 S3.1N1 topyribnr i. IvNy h. .{-Jennc Pro,.,. Int. VI nghu al repr,id"uun m~ny Wrm rcneneJ. TI BU 31694 s i I k t , t [

714 JOHNSON AND LETZEL I UPPER UMI T (COUMTIRG RULE 11 0 MONDAY _ 2 ESTIMATE L (cOUnTIMG RuLE 2) 3 LONER L)MIT (COUNTIMG R1/LE 3~ SUNDAY DAY OF TME ME£K wEDMESCAY TUESOAP THURSDAY FRIOAY Fu;. 9 Stabitity ut 741,1 acruss day uf the interview. sure estimation was selected because it offered an opportunity to validate 'the information obtained. Aaking a question will almost always produce an answer. In our survey, 92.4% of those who answered at all supplied estimates for all 24 hr. It is, however, somewhat naive to asaumt, that the answers received are necessarily an accurate measure of the concept one wishes to measure. This is all the more true if the answers given rrtauire some subjective estimate on the part of the interviewee. tF YOU ASK: ARE YOU fREOUEHTIY 'Itt VERY SMOKY ROORS' REtiuLARLY _ _ ARSNER: YES SATURDAY n0 130 215 36.9 f e1.1 2 113 286 2a.1 z 71.1 s N - 744 7fj• 7.37 u • (1.007) E:ia. tU. .l'en,iuvity of resutts to wording. TI BU 31710 J Tf prr% timc ut tl quet Thu tunt ,.tr.l - Catl V sal l tht- tt~L r:Ln pledu~L phi t t ,. li. t,

SYMPOSiUNt: VIEDICAt. PEll>PE:CTIVI?S (1N P:\15IVF. SNtOK1N(i 711 I 1 uvvEA LIN1T Illr' ~ ~ (cOUNrcN6 NuLE U 2 EST111AfE 1 (COuNfING NuLE 2) 3 LONEN LINIT (COUNfING RuLE 3) AREA uHOER THE CURVE = TG" - PERCENTAGE EXPOSED M-riME (PERSON HOURS) FIG. 5. Group muxirtwm oxpuseJ ume 17,,) Vlewed en al1 area. be used t'or other forms ut exposure that have been measured by questionnaires (e:.g., other types of air pollution). to the next. The fact that the estimates rGmain constant acruss days ut i hc week be rxpectc;Ld that the exposure estimates will not dit'fc;r,ubstantially from one day ut the week in which the interview was conducted (1=ig. y). T-hc survey was conducted so that these subgroups were also represcntative. As a result, it can a,c:cund stability check we calculatctd the I~t valua:s tur t;ruups defined by day The foregoing analysis has been checked t'ur within-individual consistency. As The tact that this correlation is above 0.9 5uggc,ts that this is nut the case. calculation. A low correlation would indicate a highly random response pattern. Figure ti shows a,catter diagram tur the score pairs that result t'rum such a ., uhinJices yields score pairs which should be hi~:hly rclutc;d. The stabiiity of the T~index can he measured in two ways. First. one can divide the 24 exposure estimates into the odd and even hours ut' the day. Calculating Sruhilitv ,uggcsts that our ntcthud is ,upplying fairly stahle estimates. 12.2 X 4.21 Z ,

716 1OIiNSON A:.It LETZEL = U --• not at all exposed = 3 --y strongly where i= person index = 1.2 ........, 1V and j= time of day index = 1.2. ......'_4. Counting rule: 5,j =0 54, = I Exposed ;tif-time of a person: IFFT~t<<K Rule1=~ K= 1 IhF7~j -K kule24> K=2 Rule3_#~ K=3 I = I Percentage exposed al-time of a person: Tm, x IUU. 24 • Percentage exposed S1-time (person-hours) of a group: VG TGM 100.4 Ncj x 24 ACKNOWLEDGMENT We wish to thank Prut. K. K. Uberla. tnstitut fur Metlizint,cht Infurmauunsvzrurbettung, 5tatuuk und Hinmathemauk. Klinikum Grosehurlern. Murchiunlnutrabbc 1?, U-ti111k/ Munrch 'l), I-NG, Iiir his valu:,hie methudulugical advicC. REFERENCE 1. Hiruyama. T. Nun-smuking wives uf heavy sm0kers have a higher risk ut' lung "ncer: Astudy frum Japan. 8rir. ,Dlyd. J. 282, 183-1 ? II`/HIr. TI BU 31712 YNI:~ t:r I. B;,ht. .. other is the with ~ and ~:. pupul:. wu{. .. this in' cunce; The . work stitutc tainetl

718 LETZEL AND JOHNSUN viewees wcre approached nationwide, via a random-route procedure. Con,e- quently, we ubtaineJ a sample that can be rrg:)rded as being representative tur West Germany. The survey questions can be classified into t'uur major categories: (a) demo- graphic clata: (b) general questions about active and passive smoking; (c) detailed history ot' active and passive smoking during the last 24 hr; and (d) a liteiung history of active and passive smoking. The present paper focuses on current data on active smoking, as well as the analysis of passive smoking during the 24 hr preceding the interview. 1u what extent different definitions of passive smoking are consistently retlected in the exposure during the last 24 hr is of additiunal interest. THE CURRENT ACTIVE SMOKING SITUATION Any discussion about passive smoking has to be compared with its back- ground-the active smoking situation. In our survey of 1,670 persons, !.07 (63.31Yo) were nonsmokers and 613 (36.7%) were current smokers (Fig. I). Nun- smokers can be divided into furmer smokers ('_ 1.3`%'0) and those persuns who claim NEVER 42 0 % 701 ~ PRESENTLY 367% 613 n: 1670 AGE 1G-65 YEARS FORMERLY /- OCCASIONALLY 213% 3S6 ~ ME_N ~ REGULARLY <3 YEARS WQMEN 1376%0) ISC]A9 22 i- REGULARLY >3 YEARS c,niarl 13 L 401 so 114 (52.31 399% 61.6'~0 Fu;. t. Active smoking 13unJcsrrnuhlik t)eutschland I Nuvember .yZS31. TI BU 31714 I 1

718 LETZEL AND JOHNSUN viewees wcre approached nationwide, via a random-route procedure. Con,e- quently, we ubtaineJ a sample that can be rrg:)rded as being representative tur West Germany. The survey questions can be classified into t'uur major categories: (a) demo- graphic clata: (b) general questions about active and passive smoking; (c) detailed history ot' active and passive smoking during the last 24 hr; and (d) a liteiung history of active and passive smoking. The present paper focuses on current data on active smoking, as well as the analysis of passive smoking during the 24 hr preceding the interview. 1u what extent different definitions of passive smoking are consistently retlected in the exposure during the last 24 hr is of additiunal interest. THE CURRENT ACTIVE SMOKING SITUATION Any discussion about passive smoking has to be compared with its back- ground-the active smoking situation. In our survey of 1,670 persons, !.07 (63.31Yo) were nonsmokers and 613 (36.7%) were current smokers (Fig. I). Nun- smokers can be divided into furmer smokers ('_ 1.3`%'0) and those persuns who claim NEVER 42 0 % 701 ~ PRESENTLY 367% 613 n: 1670 AGE 1G-65 YEARS FORMERLY /- OCCASIONALLY 213% 3S6 ~ ME_N ~ REGULARLY <3 YEARS WQMEN 1376%0) ISC]A9 22 i- REGULARLY >3 YEARS c,niarl 13 L 401 so 114 (52.31 399% 61.6'~0 Fu;. t. Active smoking 13unJcsrrnuhlik t)eutschland I Nuvember .yZS31. TI BU 31714 I 1

714 JOHNSON AND LETZEL I UPPER UMI T (COUMTIRG RULE 11 0 MONDAY _ 2 ESTIMATE L (cOUnTIMG RuLE 2) 3 LONER L)MIT (COUNTIMG R1/LE 3~ SUNDAY DAY OF TME ME£K wEDMESCAY TUESOAP THURSDAY FRIOAY Fu;. 9 Stabitity ut 741,1 acruss day uf the interview. sure estimation was selected because it offered an opportunity to validate 'the information obtained. Aaking a question will almost always produce an answer. In our survey, 92.4% of those who answered at all supplied estimates for all 24 hr. It is, however, somewhat naive to asaumt, that the answers received are necessarily an accurate measure of the concept one wishes to measure. This is all the more true if the answers given rrtauire some subjective estimate on the part of the interviewee. tF YOU ASK: ARE YOU fREOUEHTIY 'Itt VERY SMOKY ROORS' REtiuLARLY _ _ ARSNER: YES SATURDAY n0 130 215 36.9 f e1.1 2 113 286 2a.1 z 71.1 s N - 744 7fj• 7.37 u • (1.007) E:ia. tU. .l'en,iuvity of resutts to wording. TI BU 31710 J Tf prr% timc ut tl quet Thu tunt ,.tr.l - Catl V sal l tht- tt~L r:Ln pledu~L phi t t ,. li. t,

SYtitPOSllJA: 4cEDll'AL PEkSPLI IIVLS ON PASSiV[ SMUKING 709 0 contiguratiuns. A large body uFdetailE;J information can be obtained for a number ut differing but relevant expusurc ,ituatiuns. It can be. applied to every individual so that, in theory, everyone can be ~Jas,itizd. Two issues, however, remain unresolved: specifically, it is unclear how to score the exposure. and there seems to be no way to validate the results. The scoring problem occurs because, as outside ubservers. we cannot know how to weight the various c:,~pusure components in this questionnaire. Do work, home, trans- rurtatiun. and leisure time exposures that occur in the same time frame all count cqually'' If not how should they be wzighted'•' Cunre pt n/' ,L!-"I'iirre Ei pusure One way to solve the scoring problem is to allow the interviewee to estimate the exposure intensity independent of the specific lite events that contribute to that exposure. 'l'u do this we divide the time interval of interest into a series of equal time intervals and ask the intlivi(luals to provide a global asussment of the exposure received in that time interval. We have used this technique, for example, in interviews designed to examine the exposure which an individual has received in the last 24 hr. The left panel of Fig. 4,hows the results of such an interview. This data matrix can be read as an exposure profile during the 24 hr prior to the interview. The ruw, are divided into 24 one-hour segments. For each of these hour-long scgment,. the individual has provided an estimate ut'the exposure that occurred. ranging trutn (IuuC at all to strongly exposed. A, yct we have not been able to account for the probability that the inEiivitiual repurttnb ;ome exposure in an interval was nut ,u exposed over the entire in- te rval. Counting this interval as rc presenting 60 min of exposure will tend to T~rf s. 1, 2 3 0 S+~ Y 6 x C e ~9 :0r+ x rr) EAPOSED M-iIIE OF A PERSON (t" 1 PERCENt EXPOSED K-iItE OF A PERSON E+PDSED I Vt I/1 1 2 3 ~/. I EtPOSEO 01 1 1 2 l v \ I ~ ) 1 1'/ PfA'SON t1QWS ~ 9 PEp~1 t10uAS ~7 '7 ~---~ ~ EaPUSED 012 I rr 1 / ` i r 1 0 PERSON HOtit5 t_--7 A EY. 3/5'/. OD% Flt.. 4. Maximum individu;il cxpu,ure ume a, a fuwtiun of cuunting rulc. TI BU 31705

720 HOUSEHULt7 MARITAL STATUS SELF-EMPLOYED 5d5%. CIVIL-SERVANTS 72.2•/. WHITE- COLLAR WORKERS 62.G% BLUE-COLLAR WORKERS 575%. RETIRED I NO LONGER WORKING FtG. 5. Frequcncy of nonsmokers and uccuputiun. TI BU 31716 PERSONS ss.ev~ 6a~ . 2 t 142 131 Lt i"LEI. AND JOHNSON 612% 3 c 26 667•/. 4. 70.6% Fu;. 4. Living suuatiun at home uf nonsmokers. (up to 10 cigarettes, 27.9%) are somewhat more numerous than heavy smokers (more than'_0 cigarettes per day, 23.2%). The proportions ut smokers and nonsmokers vary t'ur a wide range of ,ub- groups. Starting with age and sex. Fig. 3 shows that, up to 35 years, smokers and nonsmokers are comparably distributed over both sex groups. Above the age ut' 3~, however, the frequency of nonsmokers among women increases markedly with age. The home environment al,u shows some cunnectiun with ,muking habits (Fig. 4). Interviewees living in large households (tive or nture persons) were nun- ,mukers tar more often (7U.64''c) than were people living alone (55.8`%). Marital ~,tatu. shows little variation except t'ur the fact that two-thirds of divorced people are smokers in comparison to one-third in the other groups. . With regard to uc:cuputiun. the frequency of nonsmokers can be classified into three levels (Fig. i)~ :ivil servants and persons no longer working (mu~t ut'the latter being t'emalc,l avCrage around 72%, white-collar workers and rztir.ttl persons around 6•'_r/'(1; and szlf-empluyetd persons and blue-collar workers around 581/o. . There t 7U1/~ of Bavaria. the uthCt in larger The I r,

1Y'~1PUSIUVI: N11a)IC':YL I't•at.l'PIi( llVtiS llN PASSIVt: SMOKING 6YJ/i EG) '~ I I I i LLb- :c:rS :Lge :dly Fig. wn- rital jpi<: ~ 7 } ~~ ~ V / 5 65 / Fu;. 6. Nun.moking in West German states. 721 There are Nurpri,ingly marked clitference, among German states I Fib. 6). About 7l1`.~ of the population are uku,lnukers lL1 West I Berlin, Baden-Wurttemberg. an<1 l3avaria. but only 44"~ ot th<: Hamburg population are nonsmokers. The rates tur the uther ,tate, lie in between. Cities show <litterences in that more people smoke In I:u'g,:r Cities than in small ones (Fig. 7). ACTIVE AND PASSIVE SMOKING DURING 24 HOURS The frequency of active and passive smoking is ,huwn in Fig. 8 as summarized 6e5 72a/ '/. < 5000 5000-20000 2D000-100000 100 000-500 000 60e r- 577 )500000 mMaWraNts 1=i<.. 7. SiLC ut' cumnwnuy iperceiuage ul' nun,mukcr.l. TI BU 31717

710 JOHNSON AND LETZEL I overestimate the actual exposure. To call attention to the fact that this is a max- imum exposure estimate we label these intcrvals as M-times lT"t). • In cunstructing a scoring procedure tur this data matrix we must also recognize that the global assessment of exposure intensity has a strong subjective cumpu- nent. Since virtually no objective basis exists tur this assessment, each individual must rely on some form of heuristics by which he classities the exposure intensity. In r"ognitiun ut'this tact we have defined three counting rules'shuwn in Fig. 4. "1'hcse rules define whether or not a particular interval will he counted as exposed or not exposed. From there, the individual's score is calculated as the sum of the expuSetl hours. This sum can then be viewed as the exposed M-time of a person (1'~), or it can also be expressed ;0, a pcrcentage of the total time under observation (`.'4 exposed M-timc: ut a pcrsun). As can be seen in Fig. 4 these values can vary markedly, depending upon the counting rule u,eLl. The counting rule that counts all entries other than "nut at all" as exposed yields the highest scures. The rule that counts only those hours in which the individual indicated a very high intensity of exposure yields the lowest. Due to the subjective nature of these global assessments it is not possible tur this estiin,ae the percentage of exposed M-time t'ur a group in persun-huur5. be u~4 tt' we calculate the "1~ values for each of the three counting rules, we obtain (e.9.. scores that range from 1.68 to 1'_.?S4: of the observed person-hours. In Fig. 6«u have plotted the same data as a ti u4uency distribution. Displaying Sruhit. the data in this way highlights the rurul c.,%husure fur each individual in the group. The Fur ;uunting rule l, 47.7% had no exposure, wnile 7.2% reported a single hour the '_ • in which exposure took place. As might be expected, very few individuals report subin, more than IU hr of exposure. When displayed in this way the 7~, or percentage Fiu: ut' exposeJ M-time. represents the mean value of the frequency distribution. calcul Figure 7 displays the data in a cumulative form useful for visual comparibuns The t, of group diFferenc'es. We call it the cumulative standardiLed exposed M-timc Th, diagram. From such a diagram one can easily see what percentage uf ,i group had a~,6~ an exposure ufa magnitude between cero and a specified percentage u[' th.; oh- of th• . served time. Each of these presentations and ,coring techniques can, in principle, wndt us to know which of these counting rules is the most appropriate. It i,,. however, reasonable to expect that the "true" exposure lies somewhere between the two extreme values. In Fig. 5 we have plotted the percentage of exposed persons as a function of the time ut day tur a representative sample of nunsmukcr, in the Federal Republic ot Germany. As expectc:d, the three counting rules produce estimates of exposure that ditCer markedly. The overall pattern. howqver, is consistent in that they all rise and tiall at roughly the same time of day. • "1'hc area und.:r the curve for 1.1uunting rule 3 has- been highlighted. The mag- nituJe of this area represents the total number ut'expused person-hours observecd in the sample using this counting rule. When standarJizeJ using the total-per5un- huurs ubsc:rvec1, it can be used as an estimate oYc;xpusure for the group. We call 5ee appe:ndix. TI BU 31706 he ex to th, ~uL~~

SY:tiIP+,,~tUht: MEDICAL PERSI't•.l I tVI:S ON PASSIVE SMUKING 703 ;h y .c h .r >. n ~ sive smuking, hut may putisibiy be the result of confuuntling tactors that had not been eliminated. The considerations presented in this study, together with critical remarks made by various uthcr authors (13-15, 17), permit the cunclusion that, in the puhlicatiuns citcd abuve, some of the results may be statistical artifacts. What particular factors tbias, conti>unding tacturs) may be responsible for such artifacts cannot he determined without an accurate knowledge of the study pro- tocol. especially as it relates to the collection and evaluation uf the data. A further unanswered question is whether a given passive smoker had been an active smoker at some time in his or her life. Such a situation might explain how the individual high risk figures would arise. As represented in the present study, all the evidence seems to indicate that the risk involved in passive smoking is at- tributable to a low dose aa't'ea:t, which has no appreciable iMlucnca: on the inci- dence of lung cancer and is not in conformity with the valucs for active smoking. Furthermore, it must be said that in this dose-response range, air pollution and occupational exposures must be considered minimal risk factors. This means that passive smoking exposutr Qannot be established merely on the basis of the hus- band's cigarette consumption (6), since: such consumption does not permit a quan- titative determination of the exposure and its temporal correlation with, or its tlttterentiatiun from, other noxac. In common with discussions on the question of air pollution and lung cancer, the discussion of passive smoking and lung %;ancer is charged with emotion and is, therta'ure, g(:ncrally overstated. Unfortunately, this situation leads to a dis- traction from the still unresolved, and greater, problem of active smoking. ' I REFERENCES 1. Austria IanuKWerke Ali. F'ucnllCnC MlttttlungCtl. vlennU. lyli. ~ 1 2. lurrea. R. Pickle. L. W., Funtham. E., Lin, Y., and Haenstel, W. f'assive smoking and lung cancer. Laupcrt 2. iy5-597 ( 19tt3). ~~ 3. Crusnp. K. S.. Huet. D. G.. Lanyley. C. H.. and Peto, R. FundamCnttsl carcinogenesis pri>Lesbes -und their imp(icatita,, lur low Juse risk assessment. Cuuru•er Ria•. 36, 2973-2979 I t97(+/. 4. Doll, k., anJ Neto, K. Mortality iu relation tu smoking: 20 yeur, observations on male tiritish dix:turs. Brir. ,Ni•c/. J. 2, 15_5-t53t+ 111)76). 5 . Doll. R., und Petu. K. Cigareue ,moking and bronchial carcinoma: Dose and time relutiunehips uswng regular ,muker. :+nd tiictunL nun-smukers. J. EphlensinL Cunuutuuty Nrulrh 32. 303- 313 119781. h. Gattinkel. L. Time trends in lung cancer mortality among nonsmokers and a note on passive ,muking. J. .Vutl. t:'un<'er hot. h6, IOti)-I(Ab 11981). 7. H:unmuuJ. t. C. Smoking in relation to the death races of one million men and women. Nurl. Cirnrer lnst. Slnntukr. 19, l27-2t)411yhb). 8. Hir:+}u,n:r. "r. N++a-,muktng wives of heuvy smokers have a higher risk of lung cancer: A study tram Japan. Brit. ,Wed. 1. 282, 183- IM5 (19tS 1). 'l, tiirayamu. T. Was spricht tur, was gegen die Schudlichkeit Lies Passivr,tuchens:' (interview). 41ucm ft. .bfed. Wuchrnst•hr. 123, t4b(ll-Idti3 (19M11. lu. Hupu.t C'.. Hawkins. L. H., .ind ANtrup, R Expo,ure uf pas,ive ,mukers to tobuccu ,muke ~un~tituem.. lrtt. .ar, (+ !)rrup. btrnrun. ffe•ultlt 42. et t 14714), II K.rhn. If. A. The llurn ,tuJy „( ,ntukin3 and murtality among U S. veterans. Report on eight anJ une-ha1C ye•art of observation. Nwl. Cuiner li,.rt, .lfurrr;.,;r. 19, t-125 I19hAl. 1:. Klu,tctkuter, W.. and Gunu, E. Zuni Pruhiem Jes t'ansivrauchen.. ZcnrrtrlhL &,ktvrk.l.. Yur,r .itendJ., /nlrr<rrunxkr. fh•t•. lthr. I' oriq.. Rriltr 8 162, 5t-6y (1N7b). I t Kurneg:ry. H. R.. and Kastenbuum. yt. A. Non-smoking wives uf heavy smokers have it higher rt,k ut lung cuncer Uetterl. 13rrt. Med. J. 283, 9t6-yt6 ( tytil l. TI BU 31699

710 JOHNSON AND LETZEL I overestimate the actual exposure. To call attention to the fact that this is a max- imum exposure estimate we label these intcrvals as M-times lT"t). • In cunstructing a scoring procedure tur this data matrix we must also recognize that the global assessment of exposure intensity has a strong subjective cumpu- nent. Since virtually no objective basis exists tur this assessment, each individual must rely on some form of heuristics by which he classities the exposure intensity. In r"ognitiun ut'this tact we have defined three counting rules'shuwn in Fig. 4. "1'hcse rules define whether or not a particular interval will he counted as exposed or not exposed. From there, the individual's score is calculated as the sum of the expuSetl hours. This sum can then be viewed as the exposed M-time of a person (1'~), or it can also be expressed ;0, a pcrcentage of the total time under observation (`.'4 exposed M-timc: ut a pcrsun). As can be seen in Fig. 4 these values can vary markedly, depending upon the counting rule u,eLl. The counting rule that counts all entries other than "nut at all" as exposed yields the highest scures. The rule that counts only those hours in which the individual indicated a very high intensity of exposure yields the lowest. Due to the subjective nature of these global assessments it is not possible tur this estiin,ae the percentage of exposed M-time t'ur a group in persun-huur5. be u~4 tt' we calculate the "1~ values for each of the three counting rules, we obtain (e.9.. scores that range from 1.68 to 1'_.?S4: of the observed person-hours. In Fig. 6«u have plotted the same data as a ti u4uency distribution. Displaying Sruhit. the data in this way highlights the rurul c.,%husure fur each individual in the group. The Fur ;uunting rule l, 47.7% had no exposure, wnile 7.2% reported a single hour the '_ • in which exposure took place. As might be expected, very few individuals report subin, more than IU hr of exposure. When displayed in this way the 7~, or percentage Fiu: ut' exposeJ M-time. represents the mean value of the frequency distribution. calcul Figure 7 displays the data in a cumulative form useful for visual comparibuns The t, of group diFferenc'es. We call it the cumulative standardiLed exposed M-timc Th, diagram. From such a diagram one can easily see what percentage uf ,i group had a~,6~ an exposure ufa magnitude between cero and a specified percentage u[' th.; oh- of th• . served time. Each of these presentations and ,coring techniques can, in principle, wndt us to know which of these counting rules is the most appropriate. It i,,. however, reasonable to expect that the "true" exposure lies somewhere between the two extreme values. In Fig. 5 we have plotted the percentage of exposed persons as a function of the time ut day tur a representative sample of nunsmukcr, in the Federal Republic ot Germany. As expectc:d, the three counting rules produce estimates of exposure that ditCer markedly. The overall pattern. howqver, is consistent in that they all rise and tiall at roughly the same time of day. • "1'hc area und.:r the curve for 1.1uunting rule 3 has- been highlighted. The mag- nituJe of this area represents the total number ut'expused person-hours observecd in the sample using this counting rule. When standarJizeJ using the total-per5un- huurs ubsc:rvec1, it can be used as an estimate oYc;xpusure for the group. We call 5ee appe:ndix. TI BU 31706 he ex to th, ~uL~~

716 1OIiNSON A:.It LETZEL = U --• not at all exposed = 3 --y strongly where i= person index = 1.2 ........, 1V and j= time of day index = 1.2. ......'_4. Counting rule: 5,j =0 54, = I Exposed ;tif-time of a person: IFFT~t<<K Rule1=~ K= 1 IhF7~j -K kule24> K=2 Rule3_#~ K=3 I = I Percentage exposed al-time of a person: Tm, x IUU. 24 • Percentage exposed S1-time (person-hours) of a group: VG TGM 100.4 Ncj x 24 ACKNOWLEDGMENT We wish to thank Prut. K. K. Uberla. tnstitut fur Metlizint,cht Infurmauunsvzrurbettung, 5tatuuk und Hinmathemauk. Klinikum Grosehurlern. Murchiunlnutrabbc 1?, U-ti111k/ Munrch 'l), I-NG, Iiir his valu:,hie methudulugical advicC. REFERENCE 1. Hiruyama. T. Nun-smuking wives uf heavy sm0kers have a higher risk ut' lung "ncer: Astudy frum Japan. 8rir. ,Dlyd. J. 282, 183-1 ? II`/HIr. TI BU 31712 YNI:~ t:r I. B;,ht. .. other is the with ~ and ~:. pupul:. wu{. .. this in' cunce; The . work stitutc tainetl

SY,b1POStU,'VI: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 715 The fundamental advantage ut'the passive smoking instrument presented in the prcviuus section lies in its ability to be applied for any relevant time period. The time intervals selected can be adjusted to span a day, a munth, a year, or the life of the interviewee. No objective criteria exist for lifetime exposure. Conse- quently, a direct valieiatiun ut' a lifclunb exposure instrument cannot be obtained. The fact that the foregoing methodology is completely flexible offers an uppur- tunity to validate it on a time trame that can be observed (e.g., 24 hr) and to argue by analogy that the methudulugy, when expanded to a longer time frame, can also be rs:garcied as valicl. We regarded this as a trivial exercise at first, and conducted a survey in which saliva samples were collectetd from each individual interviewed. The results of this prcliminary validation effort were not c;ncuuraging. Individuals who hubjec- tively indicated virtually no exposure had cotinine values over the entire observed range. Similarly, individuals with high subjective exposure ratings showed a com= plete spectrum of objective values in the saliva. This may, howzver, have been due to imperfect laboratory techniques. We are presently I.:unducting more so- phisticated validation experiments. Whether or not one is willing to believe that an expansion of the methodology to a lifelong data collection instrument is apprupriate, one thing is clear: it' we are unable to show a relationship between exposure and an objective measure taken in the last 24 hr, then it is very difficult to argue that there will he any relatiunship between the subjectively mc;~,~irv~l values for an entire lifetime and the ubjectivt. exposure that an incliviclual ucruall ' ~ received. This is true for any definition of passive smukittg, including those that have been published to date. None of the published drfinitiun., have as yet been validateJ. SUMMARY In this paper we have focused on the design ot'a definition for passive smoking cxpusure. In pointing out the shortcomings of existing definitions we have isolated a,eries of charncteristies that the final definition must possess. Ot' particular importance are the following: (a) the definition must make efficient use of all cases, and (b) the assessment instrument must possess the flexibility to determine accurately the expu,ure ut an individual across his life irrespective of the specific hI'e pattcrn that the individual has had. Still, despite this increasecl Jetail, we wish to be able to construct a scoring procedure that will allow us to arrive at a single value uP exposure (ur that individual. In this regard we have proposed the cun- cepts ut' zxpu,cd M-time uf a person (71), or group ( T~) us well as the cumulati ~ e standardi/-ed expusecl tit-timc diagram. Finally, and perhaps most importantly, it is c„ential that the inarumc:nt be validated. Only then will it be possible to ubtain intcrprctable results in any epidemiological study designed to invzstisate the re- latiun,hip between passive smoking exposure and lung cancer. APPENDIX: DEFINITIONS person hour = an hour rn which a person reports exposure. Degree of exposure: TI BU 31711

SY,b1POStU,'VI: MEDICAL PERSPECTIVES ON PASSIVE SMOKING 715 The fundamental advantage ut'the passive smoking instrument presented in the prcviuus section lies in its ability to be applied for any relevant time period. The time intervals selected can be adjusted to span a day, a munth, a year, or the life of the interviewee. No objective criteria exist for lifetime exposure. Conse- quently, a direct valieiatiun ut' a lifclunb exposure instrument cannot be obtained. The fact that the foregoing methodology is completely flexible offers an uppur- tunity to validate it on a time trame that can be observed (e.g., 24 hr) and to argue by analogy that the methudulugy, when expanded to a longer time frame, can also be rs:garcied as valicl. We regarded this as a trivial exercise at first, and conducted a survey in which saliva samples were collectetd from each individual interviewed. The results of this prcliminary validation effort were not c;ncuuraging. Individuals who hubjec- tively indicated virtually no exposure had cotinine values over the entire observed range. Similarly, individuals with high subjective exposure ratings showed a com= plete spectrum of objective values in the saliva. This may, howzver, have been due to imperfect laboratory techniques. We are presently I.:unducting more so- phisticated validation experiments. Whether or not one is willing to believe that an expansion of the methodology to a lifelong data collection instrument is apprupriate, one thing is clear: it' we are unable to show a relationship between exposure and an objective measure taken in the last 24 hr, then it is very difficult to argue that there will he any relatiunship between the subjectively mc;~,~irv~l values for an entire lifetime and the ubjectivt. exposure that an incliviclual ucruall ' ~ received. This is true for any definition of passive smukittg, including those that have been published to date. None of the published drfinitiun., have as yet been validateJ. SUMMARY In this paper we have focused on the design ot'a definition for passive smoking cxpusure. In pointing out the shortcomings of existing definitions we have isolated a,eries of charncteristies that the final definition must possess. Ot' particular importance are the following: (a) the definition must make efficient use of all cases, and (b) the assessment instrument must possess the flexibility to determine accurately the expu,ure ut an individual across his life irrespective of the specific hI'e pattcrn that the individual has had. Still, despite this increasecl Jetail, we wish to be able to construct a scoring procedure that will allow us to arrive at a single value uP exposure (ur that individual. In this regard we have proposed the cun- cepts ut' zxpu,cd M-time uf a person (71), or group ( T~) us well as the cumulati ~ e standardi/-ed expusecl tit-timc diagram. Finally, and perhaps most importantly, it is c„ential that the inarumc:nt be validated. Only then will it be possible to ubtain intcrprctable results in any epidemiological study designed to invzstisate the re- latiun,hip between passive smoking exposure and lung cancer. APPENDIX: DEFINITIONS person hour = an hour rn which a person reports exposure. Degree of exposure: TI BU 31711

720 HOUSEHULt7 MARITAL STATUS SELF-EMPLOYED 5d5%. CIVIL-SERVANTS 72.2•/. WHITE- COLLAR WORKERS 62.G% BLUE-COLLAR WORKERS 575%. RETIRED I NO LONGER WORKING FtG. 5. Frequcncy of nonsmokers and uccuputiun. TI BU 31716 PERSONS ss.ev~ 6a~ . 2 t 142 131 Lt i"LEI. AND JOHNSON 612% 3 c 26 667•/. 4. 70.6% Fu;. 4. Living suuatiun at home uf nonsmokers. (up to 10 cigarettes, 27.9%) are somewhat more numerous than heavy smokers (more than'_0 cigarettes per day, 23.2%). The proportions ut smokers and nonsmokers vary t'ur a wide range of ,ub- groups. Starting with age and sex. Fig. 3 shows that, up to 35 years, smokers and nonsmokers are comparably distributed over both sex groups. Above the age ut' 3~, however, the frequency of nonsmokers among women increases markedly with age. The home environment al,u shows some cunnectiun with ,muking habits (Fig. 4). Interviewees living in large households (tive or nture persons) were nun- ,mukers tar more often (7U.64''c) than were people living alone (55.8`%). Marital ~,tatu. shows little variation except t'ur the fact that two-thirds of divorced people are smokers in comparison to one-third in the other groups. . With regard to uc:cuputiun. the frequency of nonsmokers can be classified into three levels (Fig. i)~ :ivil servants and persons no longer working (mu~t ut'the latter being t'emalc,l avCrage around 72%, white-collar workers and rztir.ttl persons around 6•'_r/'(1; and szlf-empluyetd persons and blue-collar workers around 581/o. . There t 7U1/~ of Bavaria. the uthCt in larger The I r,

~ti~ti~tt~it yt: Mt:DtC:\i. PFKSPt:C°rivrs c>h PASsive S\WKtyc, 719 ue- ~- C]9•/. tur mu- i ~ ile(1 j z,an ung nev 01% the .hnt t 6 Y. / 2.9'/.• the <5 ; 6-10 11-20 21-40 >40 LIGARETTES A DAY I iRREGW.ARIY NOT EYERY DAY MISSING DATA ~.60 Fl(1. 2. CltdAfCtle Cun5u117p11Un U[ aclive ~1llUkeTh. never to have smoked (4?.OC/ /. Former smoking FuUuws various patterns ( regu- laritv. (lurutiun) and -,huw, some sex (liFterences. The cigarette consumption of active smokers is shown in Fig. 2. The majority ut smUICCrS (43.y1~'( ) -~„nhumc hetwc;en I I an(1 20 cigarettes per (luy. Light smokers 5 n .so) /63.6 n 1 56-65 • 160, / I .'• / 46 - 55 36-v 26 - 35 14 - 25 WU^IE.N i / / / l-1(,. i. :\ge and ,ex Jl.irlhuu0m ul nummuker.. /

PtNCEeI CxPOSCU FERSO:+S ,~ ~ 0 1 -4 ACTIVE SMOKERS HORSMOKERS.ExPOSEDTO OTNER PEOPIE'S SMOKE t~ r ••~ T-T• ~ I n I f~'1 ~' r i. s e r, ro r n W s r r e r:o n:r ar r~ Tt!£ OF 00.Y AREA UNDER TRE CuRVE: t"• PERCENT DF EXPOSED N-TtrvE (PERSON NuuRS) Fu;. M. Twenty-tuur hour time patarrn uf activr anJ pa..ivc ,nm6ing. P[RECNI E%POSEU PERSaYS I 'GOIl --{ A • BOT EXPOSED NEVER SnOKER • mv- B • EXPOSED NEVER SnORER • "J GROUP B GROUP A r e ~ o a u n u ro y n . e =o a, tr n x TIPE OF DAT Fuo. 9. fwenty-tirur hour timC pattern ul Pu„iVr ,nwking in two .uhgruups of ;r pralimtnarv Jeli- nuwun. 722 TI BU 31718 ! T 20 1/_ ,5t. 10 % Si a

~ti~ti~tt~it yt: Mt:DtC:\i. PFKSPt:C°rivrs c>h PASsive S\WKtyc, 719 ue- ~- C]9•/. tur mu- i ~ ile(1 j z,an ung nev 01% the .hnt t 6 Y. / 2.9'/.• the <5 ; 6-10 11-20 21-40 >40 LIGARETTES A DAY I iRREGW.ARIY NOT EYERY DAY MISSING DATA ~.60 Fl(1. 2. CltdAfCtle Cun5u117p11Un U[ aclive ~1llUkeTh. never to have smoked (4?.OC/ /. Former smoking FuUuws various patterns ( regu- laritv. (lurutiun) and -,huw, some sex (liFterences. The cigarette consumption of active smokers is shown in Fig. 2. The majority ut smUICCrS (43.y1~'( ) -~„nhumc hetwc;en I I an(1 20 cigarettes per (luy. Light smokers 5 n .so) /63.6 n 1 56-65 • 160, / I .'• / 46 - 55 36-v 26 - 35 14 - 25 WU^IE.N i / / / l-1(,. i. :\ge and ,ex Jl.irlhuu0m ul nummuker.. /

.SYMl'(11tUVl: `IEC)IC.aL 1'1{KSI'L•'C°CIVLS ON p:\JSIVIi SM(>K1NG I ; /. 3 ESTI/YTE Z:~/T M-11NE (COURTIRG RULE 1) ESTIIMTE I(COUIITIUG RULE 2) i /i- 4 LOME0. t.1111T (COUbTING NULE j) Fw. f(1. 1'ercentage exposed 41-timr (perwn hours) tur men IT'~). from uw-'_-t-hr history. Both groups have a corresponding time pattern and. in principle, ,huw a plausible picture. There is practically no smoking during the night. Smuking ,tart, between 6 and 7Atit and reaches a pluteau at around y,\Nt which last~ utllil 4 PM. The Crcyuency then increases again until a peak ut about 35% is reached between 7 and 9 vM. This is followed by a rapid drop toward midnight. Overall. the percentage ui' active smokers smoking during a certain hour is two to three times higher than the hourly incidence of nonsmokers re- porting some degree of exposure to other peuple's smoke. In Fie. 1). the 24-hr tintc pattcrn is depicted tor two subgroups uf uur preliminary detinitiun which was base(.I on three questions. According to this Jelinitiun, uruup Ashuultl bc; without expusure, whereas group B includes persons who live with a,mokzr (not necessarily the interviewee's Spouse) and/or are frequently in ruums where a lot ut'smoking takes place. The 24-hr lttnu patterns (1it't'er markCtlly between these two groups. On average, the percentage of exposed persons is two. to three times higher in group B. This suggests that our preliminary (letinitiun and the results of our 24-hr history representetd by 7' t seem to measure the same qualities. ` Percentage cxpustd M-umc IperSUn-huursl of a group: ~' .fel , =1 30 YV 1 ( I) 723 u ER LIMIT I ~ - COUMrIN6 RULE TI BU 31719

1Y'~1PUSIUVI: N11a)IC':YL I't•at.l'PIi( llVtiS llN PASSIVt: SMOKING 6YJ/i EG) '~ I I I i LLb- :c:rS :Lge :dly Fig. wn- rital jpi<: ~ 7 } ~~ ~ V / 5 65 / Fu;. 6. Nun.moking in West German states. 721 There are Nurpri,ingly marked clitference, among German states I Fib. 6). About 7l1`.~ of the population are uku,lnukers lL1 West I Berlin, Baden-Wurttemberg. an<1 l3avaria. but only 44"~ ot th<: Hamburg population are nonsmokers. The rates tur the uther ,tate, lie in between. Cities show <litterences in that more people smoke In I:u'g,:r Cities than in small ones (Fig. 7). ACTIVE AND PASSIVE SMOKING DURING 24 HOURS The frequency of active and passive smoking is ,huwn in Fig. 8 as summarized 6e5 72a/ '/. < 5000 5000-20000 2D000-100000 100 000-500 000 60e r- 577 )500000 mMaWraNts 1=i<.. 7. SiLC ut' cumnwnuy iperceiuage ul' nun,mukcr.l. TI BU 31717

726 LETZEL AND JOHNSON e GROUP: A - 1wP E11Pf1SEO MEVER SMOKQRS B • ExPOSEO nEYER Sl+OUERS C• ttlr ExPOSEO FORMER S'qKERS D • EAPUSEO FURMER ShIIKERS 50 Z 75 Z 1UU !: OF oesEAVEo etFRVAL. Flc. 13. Cumulative ,tnnJardiLed r(puscd M-timc Jiagram Pur prcliminury aefinniun. PaEUnltaih! OEFIUI1t04 t3-tl4ESIlON CIASSIFiCAIION) aPPRO%iMAtE AECONSiAUC11(fN OF MIPAYN7A'S AEFINIIIUIf ` 3 uNEf LIw1T 4COU.ftw .ut,E tl ! I i I t6 S ESTI/YrE 1:f(lh-TIPt / (COUfrlwr. aulE 1) 2 ESTtlu7E t(COYFTtM6 EULE 2s I W QMEN > LO MARRIEO TO A N0N•SMGKER WOMEN ) LO MARRIEO Tp A SMOKER Flc. 14. Prrcrntagc m.xpu,cE! M•tlmr (person huunl cunlp;trfng Iwu detinitiuns. TI BU 31722 J F report,- i Cla~sit. to utlt' R7?.1,Y , latiun. at {Ca- the ~kI a:+ nut This i: t'a re nL is CUrt rate ut as ha-., ' Johi ,tuElie, 1IU Sufll. (. Utl ir

.SYMl'(11tUVl: `IEC)IC.aL 1'1{KSI'L•'C°CIVLS ON p:\JSIVIi SM(>K1NG I ; /. 3 ESTI/YTE Z:~/T M-11NE (COURTIRG RULE 1) ESTIIMTE I(COUIITIUG RULE 2) i /i- 4 LOME0. t.1111T (COUbTING NULE j) Fw. f(1. 1'ercentage exposed 41-timr (perwn hours) tur men IT'~). from uw-'_-t-hr history. Both groups have a corresponding time pattern and. in principle, ,huw a plausible picture. There is practically no smoking during the night. Smuking ,tart, between 6 and 7Atit and reaches a pluteau at around y,\Nt which last~ utllil 4 PM. The Crcyuency then increases again until a peak ut about 35% is reached between 7 and 9 vM. This is followed by a rapid drop toward midnight. Overall. the percentage ui' active smokers smoking during a certain hour is two to three times higher than the hourly incidence of nonsmokers re- porting some degree of exposure to other peuple's smoke. In Fie. 1). the 24-hr tintc pattcrn is depicted tor two subgroups uf uur preliminary detinitiun which was base(.I on three questions. According to this Jelinitiun, uruup Ashuultl bc; without expusure, whereas group B includes persons who live with a,mokzr (not necessarily the interviewee's Spouse) and/or are frequently in ruums where a lot ut'smoking takes place. The 24-hr lttnu patterns (1it't'er markCtlly between these two groups. On average, the percentage of exposed persons is two. to three times higher in group B. This suggests that our preliminary (letinitiun and the results of our 24-hr history representetd by 7' t seem to measure the same qualities. ` Percentage cxpustd M-umc IperSUn-huursl of a group: ~' .fel , =1 30 YV 1 ( I) 723 u ER LIMIT I ~ - COUMrIN6 RULE TI BU 31719

t'Kr.vt:4nvt: str•:uictNt. 13, 717-739IIYtSiI The Extent of Passive Smoking in the Federal Republic of Germany' H. W. L.E'rZEL= AND L. CLARK .IOHNSUN Gr.rrllsc•huJ't Jiir lnJilrnuuiurts veruhrit/ul~f and 5latistil: in der :Nedizin r l' PettenkuJ'rrstru.%.ee 15, U-dUUO,Niutici/ 2. FeJerul Rrpublic uf (iermuny A respreaentutive survey of 1.670 persons between 14 anll6i years of age was contlucted in order to obtain current data on actlve and passive smoking in the Federal Rcpubliu of Germany. Overult. 36.74 were smokers. =1.3f%r were ti/rmer Nmuken. and 42.llf:i were noionwkers. These rates vary Por Suciudrmugraphic ,ubgruups and t'ur ,latea. %atles. and rurat areab. The time pattern during the 24 hr prcccJiug the interview is identical in shape tur active as well ab passive smoking. The exposed maxlmum time tior passive smoking varies with ige t<Illl wx. It lies Nu/llewhere between 2 and l5"r UI the tlhservelt'-4 hr: the hCS( CSllnlalCl -,eem tt) he S<Glur nlln.lmUklng men and ?-4`;f t'ur nonsmoking women. It wn,i,lrntty cumpuren with our preliminary dalinitiun uf pas,ive smoking. A rewn,tructiun ut Hirayama', delinitiun reveah. parallel results in terms of maximum expu,ure ume when cumtueJ with our preliminary detinuiun. A direct comparison between both JetinitiunN ,huweLI incunNi,tencies to an extent lhat could jwlparJicc the results uf a ca~C-cunuul ,,tudy. t)utu ~etmrouraln a ma.,ive ctTeet of measuring techmyues on study results with regard Iu the Ireyuencv anJ extent ut passive Nmukmg. l'hey A,u show the ,Idnerabiltty ut Ihe cal.l,l.,uuu .1 cyuivalcncc ut activety smoked cl[!,'illYttes. IIvtl3 A...JC,u- rI,,,. lnc. INTRODUCTION Basic research in the field of passive smoking is difficult (as is apparent from uther articles herein), but practical applications are by no means easier. Our goal is the investigatiun of an association between passive smoking anld lung cancer with sound Ncicntitic methods. For the planning of such studies we need reliable and valid data with regard to the t'reyucncy and extent of pa,sive smoking in the population. Up to nuw. such data have nut been available. Theretorz, we under- took a natilutwiJe representative survey in the German pupulation. In additiun, this investigatiun provided us with the opportunity to gathcr experience with our concept f'ur measuring passive smoking. DATA BASE The survey included 1.670 persons ranging In age from 14 to 65 years. Field work and basic data processing were carried out h} a commercial reseurch in- stitute that ypecializes in conducting representative surveys. Interviews were ub- tainerd by professional interviewers. E:,ch interview lasted about 4_5 min. inter- k'resented at the Symposium "MeJicut Pcr,pective, on Ra»ive Jmuking," April 4-I_. 1984. VVienna. Austria. = Ii) whom reyueNta tur reprints ,huutJ he adJreaacd. 717 IN141 •711iin.y N;.00 ~,qIN I.au ~ 14144 by :\..,denuc Hre+.. In.. \II i.FLh UI rCf1flMIuoUUn In Jnv Alf/n rClCrved.

PtNCEeI CxPOSCU FERSO:+S ,~ ~ 0 1 ACTIVE SMOKERS HORSMOKERS. EXPOSED T(f OTNER PEOPIE'S SMOKE ~ T-T• ~ I n I f~'1 I-1TI""T-P:o -~ nT :Tr arrhr. -- Tf!£ ~ , r i. s e r, ro ~r n W s r r~e r DF ---o' 00.Y AREA UNDER TRE CURVE: t"• PEREEnT DF EXPOSED N-TIME (PERSON NUURS) Fu;. M. Twenty-tuur hour time patarrn uf activr anJ pa..ivc ,nm6ing. P[RECNI E%POSEU PERSaYS I 'GOIl --{ a • BOT EXPOSED NEVER SnOKER • B • EXPOSED NEVER SMDRER mv- GROUP B GROUP A r e ~ o a u n u ro y n . e =o a, tr n x TIPE DF DAT Fuo. 'l. fwenty-tirur hour timC pattcrn ul Pa„iVr ,nwking in two .uhgruups of ;r pralimtnarv Jeli- nuwun. 722 TI BU 31718 ! T 20 i= ,5t. 10 % Si a

Lli'rZLL ANI) JOHNSON 20% I ~ ~ I i ~I , ---- - -- - - - - I I I I -ia vr.m un„ .cwn" .a wrc a. 3 ESTIMATE 2'?, M-TIME ([OUNTtNG NUIE 1) I -----4~ I I 1' 4 LOwEN t.tMiT (COUMt1MG NuCE 3) 0 5%. ~ I I I I I ~2 ESTIMATE 1(COUnTInG aULE 2) 1- 1L - 25 1 36-tb t6-55 56 - 65 YEARS OF AGE Fut. 11. Percentage exposed M-time (person hours) tbr women (TM). The percentage of maximum exposure time (M-time)' is also intluenced by age and sex. Figure Il) shows the means fur tive age groups in man. The upper limit of exposed time is highest between the ages uf 26 and 35 years (just above '_U` ; ut time) and is luwe,t in the oldest group (below The upper limit tur severz exposure lies well below 5<< in all age groups. The concept ot' "upper limit of exposure" was chosen because if a person states that he or she was exposed to other peopie's smoke during a certain huur, this does nut autuutatically mean that exposure occurred over the whole hour. Uncorrecttal calculations hasutl on counting rule I' are therefore certainly the upper limit ot'the time ut'actual exposure. In cuntrast, M-time calculated by using counting rule 3' could serve as an estimate t't+r the lower limit of exposure. t3ut, this might be too conservative. Counting rule 2' lies in between and leads to an estimate of about one-third M-time (as cal-ulatztl by rule 1). It might well be that the actual exposure may best be represented by this estimate of one-third M- time. In men, it averages around 5`%,. decreasing with age. In women (Fig. 11) it is even lower, antd it does not seem to be intluencctl by age. When all age groups arc taken together (Fig. 12). the exposure to other people's smoke is slightly higher in nonsmoking men than in nonsmoking women. M-Time tT"')-Person hour-an hour in which a person reports exposure. Cuunung rules: a, = U il't'7^,1, < K Rule I-K = I Kule 2 - K = 2 a„ = I ili' !^,t, K ltule, 3}-+ K_ _} The inJiviJual'. \cure is calculated as the sum of the expressed hours. Thi% ,um can he viewed as the exposed 14-tune ul' a prnun 7'%1 1 17"1 1 =`' i- I 8ii). It ;an alhu he rcpres,rd lts a percentage ut'the total time under utTservattuil l'r M-umc of a per.unt. TI BU 31720 T ^ i 5 % . `. Figure smukini: _ cumulati%, trihution t is itlentic~t tur gruup-, to our thr, groups wit suns whet Pract ic: zyually ',t. matiun ct. In urJu approxui) exposed exposed t of passivc: ! It' we reported : ,ure" l l - lation 11, avC1'a_L'C. yuesttunY ^ Hirayair the wutllCn '

Lli'rZLL ANI) JOHNSON 20% I ~ ~ I i ~I , ---- - -- - - - - I I I I -ia vr.m un„ .cwn" .a wrc a. 3 ESTIMATE 2'?, M-TIME ([OUNTtNG NUIE 1) I -----4~ I I 1' 4 LOwEN t.tMiT (COUMt1MG NuCE 3) 0 5%. ~ I I I I I ~2 ESTIMATE 1(COUnTInG aULE 2) 1- 1L - 25 1 36-tb t6-55 56 - 65 YEARS OF AGE Fut. 11. Percentage exposed M-time (person hours) tbr women (TM). The percentage of maximum exposure time (M-time)' is also intluenced by age and sex. Figure Il) shows the means fur tive age groups in man. The upper limit of exposed time is highest between the ages uf 26 and 35 years (just above '_U` ; ut time) and is luwe,t in the oldest group (below The upper limit tur severz exposure lies well below 5<< in all age groups. The concept ot' "upper limit of exposure" was chosen because if a person states that he or she was exposed to other peopie's smoke during a certain huur, this does nut autuutatically mean that exposure occurred over the whole hour. Uncorrecttal calculations hasutl on counting rule I' are therefore certainly the upper limit ot'the time ut'actual exposure. In cuntrast, M-time calculated by using counting rule 3' could serve as an estimate t't+r the lower limit of exposure. t3ut, this might be too conservative. Counting rule 2' lies in between and leads to an estimate of about one-third M-time (as cal-ulatztl by rule 1). It might well be that the actual exposure may best be represented by this estimate of one-third M- time. In men, it averages around 5`%,. decreasing with age. In women (Fig. 11) it is even lower, antd it does not seem to be intluencctl by age. When all age groups arc taken together (Fig. 12). the exposure to other people's smoke is slightly higher in nonsmoking men than in nonsmoking women. M-Time tT"')-Person hour-an hour in which a person reports exposure. Cuunung rules: a, = U il't'7^,1, < K Rule I-K = I Kule 2 - K = 2 a„ = I ili' !^,t, K ltule, 3}-+ K_ _} The inJ