Tobacco Documents Online

r R rEAg Yn ~z~ ~ TJ212 .Ayt,vcan JOUxN+wL oy Ei'aoc+ciouocx ti'oL 125. No. I Cyayrigiir : 198T by The Johns Hopkins Uai~enicy Scboollot Hypene and PubGc Health Arinted in U S:A: .All'rigt:o trxrved PASSIVE S'l-1OKLNG ANU I.UNG CANCER LN SWEDISH WOMEN GORAN PERSHaGEN,` Z'DLNEK HRtUBEC."' AND CHRISTER' SvENSSGN' Pershagen, G. (National Institute of' Environmental Medicine, P:O. Box 60208, S-10401 Stockholm, Sweden), Z Hrubec,,and C. Svensson. Passive smoking and lung canc>;rin Swedish women. Am'J Epidemio/'1987;125i17-2+t.. The relation between passiww smoking and lung cancer was examined l by means'of a case-control s2udy in a cohort of'27;4109 nonsmoking Swedish women identified from questionnaires mailed in ! 1961 and l 1963. A total l of 77 cases of - primary carcinoma of'the bronchus or'lung were found in a f ollow•up of the cohort through 1980. A new'questionnaine in 1984 providtr+dlinformation an smoking by study, subjects and'their spouses as well as on potentiai'coniounding factors. The study revealed a, relativei risk of 3.3, constituting: a statisticaily, sigraificant increase (p < 0.05) for squamous' cell and small cell carcinomas in women married'. to smokprs and a positive dose-responsa relaiionl No consistent effect could be seen ibr other histologic: types, indicating that, passive smoking is related primarily to those forms of, lung cancer, which show the highest relative risks in smokers: histoiogy„, lungI neoplasrns; smoking; tobacco smoke pollution Ia recent years there has been a gzowing, interest' in the health effects of environ- mental tobacco smoke.B!iologic monitoring has demonstrated th'atexposure to tobacco smoke constituents. may be aporeciable among passive smokers (1-4). Several stlnd- ies show that children with parents who sm-`e have an: increased risk of'bronehitis an+. pneunnoni'al and some data' also ialdi- cate changes of pulmon'ary function in adults and children exposed to environ- mentai tobacco smoke (5). A few epidemiologic studies have tteen published on passive'smoking and lung cam- ReceiYed for publication January '~ L986, and in final form May 6, L'986: ` Department of Epidemiology, Nacional [nsticute of'.E.avironmental Medicine, P:o. Bo:60;'08: S-104D,1 Stockholm. Sweden. (Reprint requests to Dr. GQran Pershagen.) ' Present address: Radiation Epidemiology, Branch. Nacional Cancer Institute, Bethesdh. _MD. This study was supported by a grant from the Swedish Cancer 3ociety. '1'3eauthors are graceful to Kristina Pannone. Na, tional institute', ofEnvironmentai Medicine, Lars Johnsson, ;fationaliCencral Bureau of Stacistics. and Birgitta Pershagen for help in data colleraon., cer (6-17)'. Some of these show increase+d' risks~ for nonsmokers married to smo'kers, but the results are not fiilly consistent. Most of the studies were not specifically designed to investigate effects of passih•ee smoking, and' there are va~rious potencial sources of' randkom, and' systematic errorss which make it dir"ficulC to interpret the fbn'dulgs. One aim of the present i'nvestiga~ tion was toi try to minumize such errors,especially with regard! tra the validity, of the. iaformationon exposure and'effects. VIATERI;BLLS _A.lID ?+tIETHOD'S' Srttdy'sub1ects This investigation is designed as a case- control study within a cohort of nonsmok- ing women. There are two sources for the cohort. Most of the subjects are,taken from a sample of about 35;iYQ(D men and women agedl 15-&5'years in the 1960 National Cen- sus of Sweden for whoffi~ tobacco smoking was investlgated by a questionnaire maL'ed in 1963. 'DDetailed descriptions of the sam- pling strategy and the questionnaire a~ce 17

18 PERSbi:#GEa ET AL given elsewhere (13). The response rate among the women was 95.4 per cent. A total of 17,679 (66'.& per cent): of the women statedl that they had never smoked any formf of tobacco, and these are included in the present study. The secosn&sourceof subjectsis the "old' Swedish twin1 registerwlnicki containsabout 11,000 same-sex twin pairs born- between 1886 and 1925 (19),. The twins were iden- tified fsom bi=th certificates+ and a ques - tionnaire was mailed to them: iru 19611, pri- marily to determine zygosity and tobacco smoking, status. The response rate among the eii,gible female taia pairs. was ~ 85.1 per cent. In all, 9;7301 women (80.6 per cent) had never smoked; and they make up the rest of the study cohort. Cancer morbidity and mortality of the 27,409 women, in the study cohort were determined through, 1980 in the Swedish Cancer Itegister and the National Re;ister- on Causes of Deathi respectively. The qual- ity of the information in these registers is high for most cancer diagnoses (20). A total of 92 cases of tracheal, bronchial, lung, or pleural cancer were identified (lrtterna-, tionai Classij¢catioR of Diseases (ICD), Sev- enth Revision; eodes~ 162'-163)~ (21),. These subjects constitute the case series. Two control groups, each containing two controis ~ per case, were ailso selected from the study cohort. Coatrmllgs+oup 1 consisted of subjects who ~ were m ha~tc ed to their, re+ spective case on year of birtlt (t1 yeaT), Control group 2 included subjects who were matchedion year of birth as well as onl vital status at end of'follow.i,tp. The subjects in both control groups ~ were ~ selected at ran- dom from subjects who fiulfilled the match- ing criteria, with the exception that no woman could be used' as a control for her twin sister. The entire study group con- sisted of 460 subjects: ii8' cases and 232' controls from the 1'963 smoking,sample, as well as ~ 34 cases and 136' controls fsom, the twin register. Exposure i4ormation. There were two sources of exposure~ in- formation. First, as described above, data in the 1961 and 1963 questionnaires -•- -~.. used to define the, cohort from which cases and' controls originated The sec,_ ~~ source was a questionnaire mailed in to each, study subject or, if she was dea,-:. -;- the next-of-kin (excliadirag the husband,. - order to valiidnte the data on smoldmi well as to assess the exposure to en%vre^. mental' tobacco smoke froml husbands~ and parents. If a woman had been married trcreo than once, smoking was investigatecl o^: for the man with whom she had eohabi:~ ' the longest. Questions on, occupational residentiad history were also incliuded h-: questionnaire answers were incomplete. _~- ditional information was obtained' bv te; _. phone interview. The methodolo,y u_i=., next-of-kin to obtain data has been shornto provide exposure information of h~7-o ualitX (22-24). The residential history information *'ro` the 19$4 questionnaire included data o- addresses (parishes) and' types of houses _ which the study subjects had lived. A parL:L-1 was classified as urban if'90 per cent or more ot the popuiatiron lived in biniit-up areas according to the 19 ±0 National Cen- sus. One-family houses made of maner:?.i other than wood and wiih basements wero classiFied as dweilin~p presenting a: o eattr. risk of radon exposure. I;ndoor radon mea- surements show that the avera;e concea- trations in such houses are higher than in other common types of dwellln;sin Sweden (25). Statistical rrteth;ods Several methods uave been used in the statistical' analysis. The matching was re- tained in some anaiyses; and'~ maximum likelihood estimates~ of nelative risks (ap- prox'uinated' with odus~ ratios) and exact confidence iimtervals were computed ac- cord,ingto the method of Miet2inen (26'): Iis other analyses, the macching was dissolveci, and the relative risks and'confdence intzr- vals were estimated as suggested by titantei: and Haenszel (27) and' Cornneld (23), re- spectiveiy. The method'proposed by M antel (29) was used to test :inear trends :n these analyses: Besides the conventional strati-

PAisNE SMOKING AND LUNG CA:vCER (Y SWEDEN f,j stsalyses, a conditional logistic regres- ~= snalysis' 000)~ was carrie& out in an ,0=,;tto co~ntrol, residualco'nfpunding; in thazisk estimates and'to study interactions. RESULTS A careful review of'the medical records ,f the 92 lung cancer cases showed that in, 2i,aecsses, the p~rimary: sitie- was nottheyronr3tus or lung (there were no primary cKheal'orpleura.l carcinomas), andinsia -_%ms the primary site was uncertain. Car- ;:noma of the breast, which occurred in fiti e c3:~,es, was the most common,cause of sec- ondary carcinomas. For 64 of the. 77 pr.i- marv carci;nomas of the bronchus or lung;, tba diagnoses were based' on histologic evi- dmc_- and for 12 diagnosis was based on ctitoIugy: In one case, an autopsy was per- aP7T1W'il but there was no histologic exami- nation. Tbe : distribution of histologic typesamon„the primary bronchial and lung car- cnomas, is shown in table 1. "I'he~ c:iassifiFcnrioa is based on' the information in thee medical records, particularly the pathologyy reports. ?.denocarcinoma is the most com- mon Jroup, cbnstituting 57.1 per cent of the tot.al. Squam'ous cell and small cell car- cinomas constitute 31-2 per cent~.. The av- erage ages at diabnosis:a.nd at death.fo;r the Wnolegroupof carcinomas are 69.0 and!69:6 yea¢s, respectively. I;n, the following; analy- sis, the squamouseell' and small cell carei- ~ (f 9 19 nomas are grouped together b"use thesee types have ;enerally shown the `.iichest rel- ative risks among smokers f3"I!. Table 2 shows the distribution of selected variables among t'he cases and the control groups. As a result of the matcl:ing criteria, the age distribution and vital status are similar for the cases and contz4 group 2. Irn concrol group 1, there is as`lift toward older ages, an& more subjects pvre alive at the endi of follow-up than in t::^e two other a oups• Questio'nnaiares were retur.:.,i tbr 90.2- 9fi.7 per cent of; the -studw ;:'- :-rcts in the different ;,rowps. Among the; ^: ^cyrespon~- dents, 63.-1 per cent were; ch::dren of the study subjects, 21.3 per cent aere brothers or sistars; and 110:3 per cent were other relative5. There were no differences in the type of proxy respondents berw,-en the case and~ control groups. All of theretnurned+ questio:::t3ires con- taiitediaLformation on smoking by the study subject and, with the e:tception of one sub- ject in each controll D oup; on whether she had been married and whether ^er husband had smoked. For the other ~.,-i:rstionnaiTe items; e.;:, smokiat, habits of -~:Lren2s, em- ployment, and resid'ential' hi_ tory, the iiz- ter~r.al nonresponse ratesran_,ed t"rorn 9.B- 32.6 per cent. The pe:rcen•.:, ;-s in table 2 are based on the~ number of st-pondents to each item. E'i;ht(1.8 percent)iof'the-~women for whom questionnaire infbrrrnat:u:n could be TABLE 1I obtained in: 1984had smoked' daily during iii;ropatJeology of prirnan• brnncJiial and'lung at lea5t two years. Four of these; h3d ca,c:nomas and meart ages at diagnoss and at death stooped beforera,tiswering the 1961 or the in a cohont:oj 27,109'norvmoiurtg,Swedish momen 1963 questionnail-e; and one had sLarted' iracnosis V o, ~ Squa.=ous cell }j caazcaoma 12 1 i:6 Smai'l ie13 carci- noma 12' 15.6 i j Ad6aricarrinoma +i' 57.1 [.argr cell carci- IIOtr1. J 6.5 Chher primary ca-.cinomas 4 5.2 Age lyears! Diaqnoain Death 683 "0:1 65.6 65,8 617' ;0'3 67~:9 68.0 'td Toca! 1-1 100:0 69:0 09:6 after that. Two women smoked 1-7 ci;a- rettes per day, and, one was a pipe smoker. These eight women were escluded in the subsequent analyses. There were no pro- nounced differences betwetn the o oups with regard to the percentz.~ of women who were married or the percentage who were marnedi to ~ smokers. For the remainder of the questionnaire items„ no consistent differences were seen between the ;roups; with the possible e5- ception of a tendency towsrd s larger per-

i 20 PEALSNL>CuH.\ E'r' ?.L. TAz[E 2 Da.stnib'ution o/'se(ect.ed uariables among cases of lung concer and turo control grnt;xc nwtched /or year o/ oirnh. from a i coliort ,o( nor,smokir,g uwomen Vo. S, caxs Control group 1 Control group 2 cam Coctrol group 1 Control Froscp 2' Total 92 184 1'84 100 100 l.X1 Localization of primary tumor Bronchus or lung 77 83: 7 Other site,or uncertain 15 16.3 - age,at death,or at end of follow-up (years). 40-69 44' 38 93 478 70-i9 40 90 73 43:3, 4-6.9 - 80-91 8' 36 18 8J:' "r1.4 ~:3 Vital status at endiof7o11ow-up AUiM1e 6 121 10 3.4 rii.8' -. Dead 8~ 63 174 94.6 34.2 3;.6 Total questionnaire respondents 83 178 1'i5 90.2 96_% 95.1 S':noked'dailn+' 2 3 3 (2.4)' 11.7J (1.-) Marriedt' f i0 143' 1511 (84.3) (80:3) (e'b6.Si Niarried to smokert L 37 "r6' 1~7.'ll (4-4.6i /42:9) (ra3o Ao least,one parent smokert 12 30 21- (21.11i (21.4) ('.5:91 Employed outside homet' 33 13 62' (44.0) (463) (34.7i Lived in urban,area+ 39. i8 82 (60.9) (61.-4)1k (62.61 Lived in dwelling, presenting a, greater risk of'radon exposuret 11, 13' 9 (17.2) (10S)r, ( -.0) ' Numbers in parentheses correspond to percentages of total number of questio*+*9i+a respondents to each aem: +'Minimum duration of two years. :Ezposures occurring,aher the deeth of their respective case have been e:elUded'or connrols alive aothe endlof fbilow,up: centage of cases than of controls who livedd in dwellings presenting a o eater risk of radon exposure:A detailed, analysis of the occupations held by the cases and controls did not reveal any differences between the ;roups: The e eat majority of the occupa- tions were in the service sector and typical for womeni of' the age group under study, e$:, housemaiaL cook, seamstress, cl.eaner;: and n urse. In, the fo'lliowing, analyses, the 15 cases --ith primary sites other than the bronchus or lung have been excluded. Table 3;ves, in a matched analysis, the relative risks for primary careiiaouta of the bronchus or lung in women married to smokers. Never mar- ried women and women marnied' to non- smokers constitute the reference category. The resuJts are consistent for both contro'l groups. Pooling the control groups prp- duces a relative risk of 3.3 for squamous cell and sma11' celI'' carcinomas (95 per cent confidence interval (CI) = 1.1-11.4) asso- ciated with marriage to a smoker. Within this group, the relative rsks were increased' for both histologic types. The relative risks for the other histologic types and for the entire group are 0.8 (95 per cent CI = 0.4- 1.5) and 1.2 (95 per cent CI = 0:7-2.1),, respectively. Table 4 gives a dose-response analysis with regard to smoking by the husband. The matching was dissolved'in this analysis as well' as in table 5. T'Were is a positive trend in the relative risk for squamous celll and simall' clell carcinomas (x2 = 3.9), but not' for the ot:lier, histologic tyspes. The rel- ative risk in the highest expo5ure group; i.e., women with husbands who~ smoked more than 15 cigarettes per day or one pack e. I

P:4SSIVi SMOKING aND LIutiG CANC:R' i.`i 3wEID'EV TasLE 3 q,elrii;e:risks (RRi ) artd 95% confid'ence intervaLt (CI) lor pr,*,7tary a:rcinoma of the oroncitrc, or lung in nonamoking uromen rnarried to arrwkxrs with tmo evntml gr•~ups in a rr.atc.%ted analnais' 1 I Wsoloipc,typ. Ybf ot cases 21 Cbntrol qzoup I l't Cantrol,sroup 2; Both,control goups RR Cl RR CI Rdi CD yQuamous cell:or smailicell cueinomal 20 3:8' 1.1-16.9 3.4 0.8-20.1 3.3 1.1-11.4 ptlrer t,rpea 47 0.7 0:3'-1.6' 0:8 0.4-1.7 0.8 0.4-1.5 Tonl' 67' 12 0.6-2.2 L'_ 0.6-2.1 1.2 0.7-2.1 • Mever: married women andl women marriedl to nonsmokers constitute reference category. Maximum liwlihood estimates of relative risks and exact confidence intervals (26). t,Vtatched to cases on year of "oirth: Z Matched to cases on year of'hirth as well' as on vical status at end of follow.up: TAsLe a Rtlat~vnska (RR) and 95% confidence intervals (CI)'for primary carcirtomra o; the b'ron,chui or lang, in nonsmoking women in,relaoion to estimated'ezposure to tobacco smoke )'rom tite, husoand' {iataiogic cype \e+•er married or maried to a nonsmoker. Low rzposure co :,-aacco smoke of husbandt High expoaure:to tobacco smoke of nusaanda'' CSi- ;quare Va. of casza RR `lo. of cases RR C1 Vo. of casee RR, CI tor trer.n§ Squamous cell or small cell carcinoma ? 1l0 10 1.8' 1i 3 6.4 1.1-34.7 3.90 Other types 27 1.0 16! . 0.8 0.4--1! "o 4 2.4 0:6-817' 0.03 Total 34 1.0 26 1L0 0!6-1.3 7 3.2' 1!0-9:5 1.4'5 ' Age-standardized reiative risk estimates (27) and approximate confidence intervals (23). t HGsband'smoking up co 15 cigarettes per day or one, pack !3D z) of pipe tobacco perwee4 or any amounc during Less chan 30 years of,marraage. I Husband smoking more chaa 155 ci;arettes per day or one pac'z:oi pioe cobacco per •xee's during 30'years of marriage or mo re: , § Test for linear trend (29). of pipe tobacco per week during 30 years of marriage or more, is 3.2 (95 per cent CI = 1.0-9.5) for all histologic types combined. Table 5 shows the influence of parental smoking on the risk of'primary carcinoma of the bronchus or lung, cmntroiling for smoking by the husband. There is no con- sistent evidence of an etfect, a'nd the 95 per cent confidence intervala for the relative risks in women with at :east one smoucing. parent encompass 1.0 for both histologic groups. These results must be interpreted with caution in view of the lack of infor- mation on parental smoking habits for 24 per cent of the questionnaire respondents. The results of the conditional logistic regression analysis, which included cases. ' 11 and macc::ed control5 with information on all variabies, were consistent wit'h the re- sults of t`:e stratified analyses. There was no, impor'ant confou.^.ding of the associa- tion between smoking by the husband and squamous cell and small cell carcinomas by occuoatiom by living, in 5ouses with a greater risk of'radrJn exposure, or by iiving, in urban areas. -None of the relative risks associated with these factors deviated sig- nificantly from 1.0 upon statistical testing. For all h:stologic types taken together, the rQlative :sks and 95 per cent confidence intervals associated with marriage to aa smoker and: with living in a house present- ing, a greater risk of radon exposure were 1.? (95 per cent CI = 0.6--?:6) and 1.4 (95 A

22 PERSHAG&N ET AL T.suE 5 Rzl.atiue rv~_, (RR) and 93 ~! confidence intervals fCTJ for primarti carcinoma o('rhe bronchua'or llcng,in nonsmokinP uomen in relation to smoking,habits oJ' parents" FNistolope type ' aoth parents Atleast one smoking nonsmokers parent Vo: of RR N1o: of RR Cl cases cases Squamous cell or small cell carcinoma 0 .0~ ' .9~ .5-6.2 Other types 28 1.01 3 016 01-1.9 Totai 38 1.0 9 1-0 0:4-? 3' ' Niantel-r[aenszel estimates of relative risks (27) standardized ior age and smoking of husband with appronimate confidence intervals' (28). per cent CT = 0-4-5.4L respectively. For women who had' been,married to a smoker and& who had lived in a house presenting a greater risk of'radon exposure the relative risk was ?.5 (95' per cent CI = 0.8-8.5); suggesting a positive interaction, between the two variables. DISCUSS1ioNThe results of our study indicate that exposure to environmental tobacco' smoke is, related to an increased risk of those histologic types of lung cancer which showw the highest relative risks in smokers. This is in, general agreement with the findings of Trichopoulos et' al. (G)„ Garfinkel et al. (i16), and Koo et al! (16), although, thesee authors looked at somewhat different car- cinoma types and,/or used other definitions of'exposure: It would be of interest to see an analysis of the risks for different tiisto- logic types in the other published studies on passive smoking and 1'ung cancer, espe cial0y those with an aporeciable number of cases, as well as in subsequent studies on this topic. Combining the published epideznuologic studies provides a weighted average relative risk of lung cancer of 1-5 associated with marriage to a smoker (5). The results of the present studyy are consistent with this esti- mate. A 50 per cent increase in risk does not seem unreasonable in view of exposuri. estimates among passive smokers (5, 32': and the excess risks of between 100 anc: 900 per cent for smokers in the lowe>•' esposure categorx, as al rule 1-9 cigaret,e - per day, in the major cohorts studied (li, 33=39). It should be noted that relati• ~ risks for squamous cell and, smaIl cell car- cinomas would' be expected' to be even higher, i!e:, ifthe'ca_ce groupis not"diluted` with adenocarcinomas or other types with weaker association to smoking. Severall sources of random and sv.= tematic errorshave ro be considered, in thr interpretation of the findings. In contras,, to eariier studies on passive smoking and'lung cancer, the present study has a"double check" on the smoking status of all studti- subjects. Data were obtained' fromi the 190"'11 and 1!963~ questionnaires that were used tcrn define the cohort as well as from the 198-' questionnaire. Our results indicate that misclassification of nonsmokers was a mi- nor problem and that failure to take thiss problem iilto account would not severely bias the association berween passive smok- ing and, lung cancer: This is' supported by the findings of other Swedish studies. whichl show a high quality of questionnaire information on smoking; both when the data were obtained from the subjects them- selves and when data were obtained from next-of-kin (I2'2; 23). U-sing, smoking bv the husband as the only measure of exposure to environmental tobacco smoke will resLlt in misclassifica- tions in the exposure assessment. To the extent that such miscl4ssifications are un- related' to the disease in question, this would tend' to reduce any true associationn between passive smoki,n; and lung cancer. The similar percentages of exposedipersans among the cases, excluding squamous cell and small cell carcinomas, and the two control groups suggest that errors in the reporting, did not affect the cases and con- trols differently. This lends further support to the associationl with smoki:ag,of the hus- bands, which was noted for squamous cell and small cell carcinomlli s orsl'y: Obviously,

PASS1«'6VK'1'IGI1G A1+1© LUNG CA_\CER 1N aN.cDE ! 23' 3 is unlikely that the nezt-of-kin rtapon- 3mts were aware of the histologic subtypes &2rsosed for the cases. Our results show that poor quality of the &qMosis may be a problem in studies of I=g cancer in female nonsmokers. Second- ay pulmonary carcinomas or carcinomas sith unknown primary, sites appeared' in aboutone-siath of the cases reported,in the caacerandJor cause of death reg!isters:'T'his is in close agreement with the findings of Garfinkel (8), which were based on death certificate diagnoses in the United 3ttates.. If secondary tumors are not excluded Prom the case series, the; relative risks associated with any factor that causes primarily lung muciinomas are likely to be underestimated. Aa noted previously, the analysis may be further strengthened by separati.ng; differ- ent histologic types. Besides : the quality rn'f the exposure and. dia.gnostic infornation, the validity of our atudy is also affected by the control of confoundzng; factors. The, association be- tweea passive smoking and~ ll,¢ng cancer of the squamous cell and small cell types was not confounded' by occupation, urbaniza- tion, or living in houses with a greater risk of radon exposure. nor were any of these factors associated with a clear increase in risk when passive smoking was controlled. These Findings shoulid' be interpreted with some caution in view of the interna1 non- response on the questionnaire for items other than smoking of the study subjects and their spouses. It is, however, improba- ble that uncontrolled confounding by the factors under study explains relative risks of'the magnitude observed; as well as the positive dose-respon5erelations. No infor- mationwas, obtained on, intakeof' food items that may affect the lung cancer risk. Analysis of all the lung cancer cases sub• gested a positive interaction between mar- riage to a sffioker and living in dwellings presentingia greater risk of radon exposure, i-e:, one-family houses made of material other than wood and'with a basement. Iin- creased risks of ltnng cancer associate& with living in such houses have been observed previously (40--f2); but our study also pro- vides data on exposure to environmental tobacco smoKe. Our findings are consistent with an interaction between~tobacco smoke and; radon daughters similar to the one obsemmedi.n, uraniumminers (43) and in smokers lRving in dwellings with- a;reater risk of radon exposure (z1)1 Iit is also of interest to note that the radorl- daughter concentration has been shown to increasee considerably as a result of attachment to aerosol particles in rooms filled with to- bacco smoke (44). In coatciusion,, our results indicate that, exposure to environmental t~obacco smoke~ is related pritaarily, to those forrss of iung cancer which show the highest relative risks in smokers. The resuits are internally consistentand~ in'generalagseetnent with other studies: Our findinpare of scientif c inte'rest: and have public health implica- tions, althoughitis obvioustha't tungcan- cer ih~ passive smokers is a rar° phenome• noni The accumulatin;,evidence in children and4dults shows that serious health effectss can probably result from, heav,y, eexposure to environmentalitobacoo sznoke.l'..isshould encourage further research, including, both exposure assessments and etiolo;c studies. REF.ERsNCES 1. ^e.erabend C. H:gge.^.bocram T. R::ssel N1ArI., Nicocine concentrations in, urine ~nd saliva of >noaers and non-smokers. Br i.1 l98'?: 23.1:SQU2-4. -?. Jar%is 4iJ. Rcsszl' NfLAH. Feyerabecd C: aEsorp+ tion of nicorine andicarbon monox,ide irom passive smoining under natural conditions of exposure. Thorax 198;3:313:329-33. 3~: Greenberg R.Ji, Haley 'J; Etzel FY. et al: J+teasur- ing the erposure of infants to cobacco smoke: V Wnglid! Vted' :934:310:1Q75'-23. 4. Matsukura S. Taninaoo T. KitanoN. eoal. Effects of environcsental tobacco smoke on _rinary cotin- ine ~xcrecion in non-smokes. N Engli J Yled, 15+?4:311':323-32_ 5: Per-nagen G. Review of epid'emioto7r in reiacionn to passive smokin2. Arch Tocico! :3upp7)i 1986: 9:0,3:-;1, 6., Hiiravama T. \on-smoking wives of heavy smok+ ers have a higher risk of luTg;cancer a,study from Japan., 9r Med J L9a'1;_°'32•f83-.i. 7: Trichopoulos ©I Kaiandidi' A. Soarros G.. ec, aL, L.ung, cancer and passive smoki.^.;; !nr J Cancer 1981,27:1-14. 3. Carunket L'I"une,trensis in lungcsncer nortafiry

- 24 PERSHAGEM. E'r AL among non-smokers and'a note on passive smok- ing; JNCI 1981;66:1061-6. 9. Chan WC. FvngSC. Lung cancer in non-smokers in Hong Kong, lin: Grundmann E„ edi Cancer campaign. Vol 6 Stuttgart: Gustav Fischer Verlag, 1982:i99-'?02: 10. Correa P. Pickle LW, Fontham E, et ±L Passive smoking and lung cancer. Lancet 19&3'.2i595-7. ' 11. Trichopoulos D; Kalandidi A. Sparros L Lung cancer and passive smoking; conclusion of' Greek study. Lancet 1983,^:6~-41-8: 12. G'tllis CR, Hole DJi Hawthorne Wlt, ec aL The effect of envinonmental tobacco smoke in two urban communities in , the west of Scotland. Eur J Respir Dis 1984:65 (suppl 133):'121-6. 13. Kabat GC, Wvnder EL. Lung cancer in non- smokers. Cancer: 1984:53:1214-21. 14: Koo LC, Ho JHC, Saw D. Ia passit:e smoking an adde& risk, factor for lung cancer in Chinesee women?'J Exp Clin Cancer Res 1984:3:'377-53. 15. Garfinkel L, Auerbach 0, Joubert L Involuntary smoking and lung cancer. a case-concrol study.. JNCI 1985: M463-9: 16. Koo LC; Ho di$C::Lee N. Ananalysis of sometisk factors for lung caacer in Hong Kong. Inc J Cancer 1985:35:149-55. M Wu AH. Henderson BE;,Pike MC, ec aL Smoking and other risk factors' for lung cancer in women. J NC I 1985: 74: 7 4 7-5 t. 13: Cederlbf R, Friberg L. Hrubec Z; et aL The"i,ela- taionship of smoking,and some social covariables to mortality and' cancer morbidity. Stockholm: Department of' Environmental Hy,giene; Karolin- ska Institute. 1975. 19. Cederlof,R~. The twin method in epidemiological studies on chronic disease. Stockholm: Depart- ment of Environmental Hygiene. Kamlinska , In- stitute, and' Department of Sociologs,, University of Stockholm: i9fr6. 20: Mattson B! Cancer registration,in Sweden: studies on completeness and validity of incidence and mortality registxrs: Stockholm: Department of' Oncology and Cancer Epidemiology; Karolinska Institute. 1984. ?21. International Classification of' Diseases. Seventh 96evision_ Geneva: Worddl Health Or;anization. 1957. 22. Pershagen G;Aaedson 0. Avalidacionofquestion- naire information on occupational exposure and smoking. Scand J Work Envirnn Health 1982; t3:2". 23. Damber L. Lung cancer, in males: an eoidemiolog• ical study in northern Sweden with snecial,regard to smoking and occvpacion. U1nes; Sweden:, De- partment of Oncology, Gmei University„198a''. 24. Pershagen G. Lung cancer mortality among men living near an ar~enic-emitting, smelter. Am J Epidemiol' 1985:12'?:6&tJ-94: 25. Fa1k R: Study of a met,hod, to estimate the expo- sure to radon daughters in dwellings. In: Pilot studies and theoretical considerations on the pos• sibility to study the n laoion between radon daugh- ter exposure in dwellings and'i lung cancer. f'!n Swedisn/. Stocknolm: National Institute of'Envi- ronmenoal .Iftedicine: 1982.. 26. MiettinenI0: Estimation of'relacive risk from in- dividually matched series. BiomecTics 1'9 -.d}.°_6:Z5- 86. , 27. Mancel Y, Haenszel W. Statistical aspects of' the analysis of data from~ retrospective studies of dis, ease. JtiCI 19i9i2"_'719-~3:, 28. Cornfield, J. A statistical problem arising from retrospectiti-e studies. ln: Proceedings of the thirdl Berkei'ev ;vmoosium on mat4ematicail statistics's andlprobabiiicy. Berkeley. CA: University of Cal- ifornia,Ptess. C956:, 29. Mantel `:, Chi-iquare tests with one degree of freedom: extensions of,the 5tantel-Haensael pro- cedure. Am:Statistical :Assoc J 1963;58c690-700. 30: Breslow N. Day N. Statistical methods in cancer research-the analysis of case-control studies. Lyon: IARC Scientific Publications, 1!980!, 31. Kreyberg L'. Histological lung,caneer types: a mor- pholo;ical and!biologicai correlation. Acta Pathol' YGic obioi Scand U95?:Supol. LiZ:I-9'?: 32. Vutuc C. Lung cancer risic, and' passive smok'ing: quantitaticr aspects. ZbllBak2 Hyg I'Abc,Orig,B 1963:177:90-.5. 33. Hammond EC, Horn D:, Smoking and death rates-report on forry-four montlis of,follow-up of 181i„783 men: II. Death rates by cause. JAVI:a ' 958:166:1 _' 4-1308. 34. Hammond EC. Smoking in relation to the death rates of one~million meniand women. :Eatl Cancer Inst:Monogr f96a":19ii2~,"-204. 35. Best EWR; A Canadian, study of smoking andd health. Ottawa:, DepartWenc of vetional Health and Welfare.,t966. 36. Kaitn, H-a. i.ae Dora study of smoking and mor- tality among US veterans: report on eight and one-half years of observauon: Vat!' Cancer Inst Monogr 1966:19i1-125'. 37: Weir JM. Dunn JE. Smoking and mottality:, a, prospective study, Cancer 1950:25:D05--1'2' 38. Hirayama T. Prospective saudies' on cancer epi- demiology based on census population in Japan. In: Cancer epideraiology, and'environmental fac- tors: Vol'~ ;. .amsterdam: Escerpta 4fedica, 1975:26-35. 39. Doll R„ Peto: R. Cigarrotte smoking and bronchial carcinoma: dose and oime relhtior.ships among re;ular smokers and iifedong non«;mokers. d' Epi- detniol Community Heaitn '.9S-t:3?:30:3-0. 40. A'xe!son 0. Ediing C. iilin,, H. Lung cancer and' nesidency-a casarreferent study oni the possible impact of'exposure to radon and its daughters in dwellings:, Soand J Work =aviron Health 1979: 5:10115: 41. Edling,C, Kling H; aselson 0. Radon in homes- a possible, cause of lung eancer: Scand J Work Environ Health 198a;10:'20-;4. 42. Pershagan G Damber L. i alk R. Exposure to radon in dwellings and lung;csncer-a pilotstudy: In:, Pcoceedir.gs of the T:urd .nternacional Con- ference on Indoor .air Qualily and Climate. Vol 2. Soockholm: S-edish Council for Building Re- search: 198a::3-3: 43. Archer ~.a. Wagoner JK. Lundin FE. Uranium mining and,cigareste smoking effects on man. J Occup Med 1973:1a:204-11. 44. Bergman H. Edling C. Axelson 0: Indoor radon daughter concer.t rations ar.d passive smoking:, ln: Ptoceedings, of the Third! rncernationai Confer- ence on Indoor Air Quality and Climate. Voll 2:, Stockholm: Swedish Council (or Built5ng, Re- search 1984::.9-54: