Tobacco Documents Online

- 1.7 692 o.3.b 9_5 9 State of the Art The Health Effects of Involunta Smoking1-3 SCOTT T, WEISS,4 IRA B. TA(iER, MARC SCHENKERJ and FRANK E. SPEIZER Introduction In the 15 years between the first Sur- geon General's Report on smoking and health in 1964 and the 1979 version of this report, over 20,000 articles explor- ing the effects of cigarette smoking on human health have been published (1). Fewer than 1% of these articles have d:alt with passive or involuntary smok- it:g. Given the large number of people involved, detection of even a small frac- tion of the population as suffering from adverse health effects of in- voluntary smoking in the United States could have important public health implications. This review explores recent data that relate involuntary cigarette smoke ex- posure to the occurrence of physiologic changes, symptoms, and diseases in nonsmoking adults and children. Health effects related to fetal exposure in utero, an area that has been more ex- tensively studied, will not be discussed. The interested reader is referred to sev- eral excellent recent reviews (1-3). Quantitation of Tobacco Smoke Exposure in Nonsmokers oluntary (passive) smoking is de- .:,cd as the exposure of nonsmokers to tobacco combustion products in the in- door environment. Analysis of the health effects of passive smoking re- quires not only some knowledge of [hc eonstituents of tobacco smoke, but most importantly, requires some quan- titation of tobacco smoke exposure. Tobacco smoke in the environment is derived from two sources: mainstream smoke and sidesfream smoke. Main- stream smoke emerges into the environ- ment after having been drawn through the cigarette, filtered by the smoker's own lungs, and then exhaled. Side- stream smoke arises from the burning end of the cigarette and enters directly into the environment. Different tem- peratures of combustion, filtration, SUMMARY Involunls amokinp /s defined es the expostwe of nonsmokers to labacco cembus- Ilon producb In the in oor environment. Involuntary .mokers .ro exposed to a quantitatively smaller and qualltatlv y different smoke exposure than active smokers. Quantltallon u1 axpo- suro Is particularly di/ icult In both physiologic and epldemiologic sludles. Acute physiologic studies heee doeument minlmal physiologic eharqes In healthy subJects. However, Indl9lduals with heart or lunq dlee se may be differenllally affected. A relatively large body of date rNates parenlal (pertltulerlY temaq clgsrefte smoking to the oeeurrenes of both aculs respiratory 111- nesses end chronie ro piratory symploms in children. The a1feG seems 10 be greatest early in life and cennul be as erated tram /n utero exposure. Da1a linking parental smoking to lower 19.01301 pulmonary fu ctlon ere etl cross-sectlonsl and less cencluslve. What ls apparent is that the megnituda nl the d act elfect of passive smoke exposure /s 1lkely to be relatively small (from t lo 5% reduction In m ximally oblained lung function level In exposed children). Dela on adults areInsu/llelentloallo toraquantltatlveesllmate.Thelmpanantelfeclsofpessivnsmnkeexpu- sure in childhood are olold. The slight reduction in pulmonary function level may predispose Individuals to Increase rlaks /rom environmentalseents later in Iila. In addition, having a parent who amokam substanli Ily Ineroases Ihe likelihood Ihat a child wlll become a smoker. Finally, two studies have Ilnked iu ceneer in nonsmokers to exposure to spouses' cigarette smoke. Further research Is needed to confirm these Ilndings. Invotuntery smoking may have important health eflects, aither dlroct Indirect, which deserve further study. M REV RESVIR pI519e3:12l:933-9/2 and amount of tob cco consumed all lead to marked dif rences in the con- centration of the co stituents of main- stream and sidestre m smoke (1, 4, 5). Many potentially t xic gas phase con- stituents are in hig r concentration in sidestream smoke t an in mainstream smoke (5) (table 1), and nearly 85% of smoke in a room res Its from sidestream smoke. Therefore, oth active and pas- sive smokers may e similarly exposed to sidestream smok although separat- ing out health ef ects of sidestrearn smoke from mains ream smoke in ac- tive smokers woul ,Je extremely diffi- cult. Obviously, m instream smoke is inhaled directly int thelungs and is di- luted only by the vo ume of air breathed in by the smoker 1 hen he/she inhales. Sidestream smoke is generally diluted in a considerably I rger volume of air. Thus, passive smo ers are exposed to a quantitatively sma ler and potentially qualitatively differ nt smoke exposure than active smoker. The quantification of the exposure of a passive smoker to these sidestream s oke constituents is often difficult. Fa ors such as the type and number of cigarettes burned, the -size of the room and ventilation rates, and residence times are all important variables in determining levels of expo- sure so that no single measurement can characterize constituents and exposure. A recent review of levels of indoor byproducts of tobacco smoke, which includes a discussion of the methodo- logic issues involved in making these measurements, is quite revealing in ' From the Channin@ I.aboratnr), Hrigham and Women's Hnspiial; the Charles A. Dana Research tnstitute and ihc Har;ard Thorndikc Laboratory of Beth Israel Hospdal; and Ihc Jnint Department of Medicina, Beth Israel and Brig- ham and Womcn's Hospitals and the Harvard Medical School. ' Supported in paa by Training tirantx No. HL05998, HL07427, and No. HL2252F from the Division of Lung Discase. National Hean, Lung and Blood [nstitute, National institutes of Health. ' Presented in part at the Annual Meeting of the American Thoracic Society, Detroit- Michi- gan. May 14, 1981. ' Recipient of Clinical investigator Award No. Hh00740 from the National Heart. Lung and Blood Institute. 2505487727 933

934 TABLE 1 SELECTED CONSTITUENTS OF CIGARETTE SMOKE-RATIO OF CONSI SIDESTREAM SMOKE (SS) TO MAINSTREAM SMOKE (MS)- TUENTS IN Gas Phase SS/MS Parliculale Phase Constituents -MS Ratio Constituents SSlMS S Ratio Carbon Dioxide 20 to 60 mg 8.1 Tar 1 to 40 mg 1.3 Carbon Monoxide 10 to 20 mg 2.5 Water 1 to 4 mg 2.4 Methane . 1.3 mg 3.1 Toluene 11 6 yg 5.6 A:etylene 27 pg 0.8 Phenol 20 tc 150 pg 2.6 Anmonia 80Ng 73.0 Methylnaphthalene .2 Mg 28.0 Hydrogen Cyanide 430 yg 0.25 Pyrene 50 b i200kg 3.6 Methylfuran 20 pg 3.4 Benzo(a)pyrene 20 tr ~ 40 pg 3.4 A:etonilrile 120 pg 3.9 Aniline 36 0 mg 30.0 Prtitllne 32 yg 10.0 Nicotine 1.0 Ic 25 mg 2.7 D methylnltrosamine 10 to 65 yg 52.0 2-Naphthylamine 2 mg 39.0 - Adapted rrom relerence 1, pp. 11-d pointing out the paucity of reliable in- formation available under realistic ex- posure conditions (4) (table 2). Of those constituents that have been mea- sured, few have been shown to be sig- nificantly elevated indoors as a result of cigarette smoking. However, those that hsve include nitrogen oxide, carbon monoxide, nicotine and respirable par- ticulates and, under special circum- stances, carbon dioxide. Nitrogen oxide is rapidly oxidized to nitrogen dioxide (NO,) in air and notably reaches equil- ibrium with outdoor levels of NO„ most of which is determined by auto- rc.obile exhaust. Nicotine that remains irdoors has been estimated to be equiv- alent to 0.04 "cigarette equivalents"/h ir terms of dose to a nonsmoker (4). The particulate concentration indoors clearly increases with increasing num- bers of smokers, although the back- ground level is determined by the out- door level. The conclusions from the few studies that actually measure ven- t'l.ation rates during exposure suggest tt.at under "normal" air circulation conditions, carbon monoxide (CO) levels will be relatively even modest reduction rates can lead to CO That is not surprising, sidestream smoke from of a cigarette releases wt of CO (8). Physiologic studies hl monoxide levels for bit ing of sidestream smo4 cause of its ease of mE the well studied relatic carbon monoxide and cr bin levels. Carbon monS finity for hemoglobinl. 210 times that of oxyg~ nous production of ca bon monoxide at sea level leads to base ine levels of 0.4 to 0.6%u of carboxyhe Because blood carboxyl els in nonsmokers are this predicted level bas nous production alone, sources must account fo the difference. A rel mathematical model 1 structed to relate stead monoxide exposure to I low. However, ; in ventilation accumulation. given that the he burning end 11 over 200 ppm Ive used carbon tlogic monitor- e exposure be- asurement and mship between Irboxyhemoglo- Ixide has an af- approximately n (9). Endoge- moglobin (10). lemoglobin 1ev- roughly twice ed on endoge- environmental -a large part of atively precise las been con- y state carbon he rate of car- TABLE 2 WEISS. TAGER. SCHENKER. AND SPEIZER boxyhemoglobin formation (10, 11). The concentration of carbon monox- ide, the duration of exposure, and the alveolar ventilation are the most impor- tant variables (12). Stewart and associ- ates, using blood donors, found the median blood carboxyhemoglobin level for smokers and nonsmokers in selected populations to be 5.0 and 1.2%, respec- tively (13). This corresponds to steady state ambient CO concentrations of 3` ppm and 7 ppm, respectively (4). Thi lower level of exposure, although some- what higher than reported in table 2, is consistent with levels reported in prox- imity to motor vehicular traffic and is not an unreasonable estimate of what general background urban pollution levels would be. In fact, the current one hour national air quality standard for carbon monoxide is 9 ppm. Exposure levels to carbon monoxide are highly dependent on ventilation, occupancy, and smoking rates. Further- more, the half life of carboxyhemoglo- bin is approximately 4 hours. Thus, blood carboxyhemoglobin appears to be a useful biologic monitor of acute exposure to passive smoking, but it does not provide useful data for chron- ic exposure. To assess chronic exposure with some biologic marker would re- quire the ability to measure some ac- cumulating product of sidestream smoke. To date, these substances, in- cluding cotinine (14-18), thiocyanate (19, 20), and polonium-210 (21, 22), have been measured in active smokers. However, there are no convincing pub- lished data proving that any of these products accumulate with sufficient consistency to be useful as a biologic marker of chronic sidestream smoke exposure. In contrast to physiologic investiga- CONCENTRATIONS OF SEVERAL SIDESTREAM SMOKE CONSTI ENTS MEASURED UNDER REALISTIC CONDITIONS' Srbslance . Location Occupancy Ventilation Mean Nonsmoking Monitoring Levels Control Conditions Re/er ence Acroleln Bar 30 to 40 personsl50m' Open area 26 x 30 min samples 10 ppb Not given 6 Acrolein Caleteria 80 to 150l574m' 11 exchanges7h 24 x 30 min samples 6 ppb 5 ppb (non- 6 Nitrogen oxide Bar 30 to 40/5om' Open area 28 x 30 min samples 195 ppb smoking section) 44 ppb 6 Nitrogen dioxide Bar 30 to 40150m' Open area 28 x 30 min samples 21 ppb 48 ppb 6 Nilrosamines Bar Not given Nol given 3-h conlinuous 0.24 ng/L 0.005 nglL 5 Carbon monoxide Cafeleria 80 to 1507574m' 11 exchangeslh 24 x 30 min samples 1.2 ppm 0.4 ppm 6 Carbon monoxide Restaurant 60 to 100N4om' Open area 2B x 3g min samples 2.6 ppm 1.5 ppm 6 Ni.otine Restaurant Not given Not piven 2.5-h sample 5.2 ygfm' - 7 Ni:otine Cocktail lounge Not given Not glven 2.5-h sample 10-3 ygtm' - 7 Respirable Homes 2 smokers Not given 24h sample 70 MgBn' 21 pg7m' 8 Darticulates Respirable Homes 1 smoker Not given I 24h sample 37 pgfm• 21 Hg/m' 8 Jartlculale5 ' Adapled Irom re/erence 4, pp. 25e-8.

5":E OF THE APT; HEALTH EFFECTS OF IHVOLUHTANY SMOlIHO tions, epidemiologic studies have used the number of smokers in the home or in proximity of the working environ- rnent as the principal exposure variable. These relatively crude indexes ignore time spent with the smoker and the en- vironmental factors known to influence ambient smoke concentration as noted above. In summary, passive smoking research deals with an exposure that is qualita- ticely and quantitatively different from .-..at of active smoking. In addition, ::,lequate characterization of passive ex- posure in both epidemiologic and phys- iologic studies is substantially more dif- ficult. While the active smoker's total current cigarette consumption is rela- tively eesily quantitated, the lower dose and greater influence of ventilation and ambient environment for the passive smoker makes assessment of exposure rne of the most important methodo- ~gic issues in this research. Clearly, a biologic marker of chronic exposure that reflected amount of to- bacco product smoke to which persons who do not smoke were exposed would be a useful measure of passive expo- sure. Several investigators are actively pursuing this topic at present. In addi- tion, carefully formulated question- naires relating to passive smoking are also necessary, and may prove equally calid for assessing exposure. No single index has yet been accepted by all inves- tigators, and comparisons between studies remain difficult. Acute Elfects of Involuntary Smoking The acute response to passive smoke axposure has been investigated in both ormal subjects and in patients with trdiopulmonary disease-both by mptom questionnaire and by envi- ronmental chamber exposure followed by physiologic testing. Eye irritation is the most common complaint experi- enced by normal people acutely ex- posed to cigarette smoke. In one study, 69o/u of subjects reported ever experi- encing this symptom (23). Headache, nasal irritation, and cough also were re- ported by approximately one third of subjects in this and other investigations (24, 25). Allergic subjects report symp- toms of eye and nasal irritation with equal frequency to nonallergic subjects; thus, direct mucosal irritation rather than atopy may be the mechanism me- diating these symptoms (23, 25). Sever- al factors may alter the prevalence of irritant symptoms, including the amount of smoking, thd volved, the huni of ambient air, a lation (26). No I these irritant efP of increased sei have been report, stituent(s) respot is unknown. Pimm and co smoking adults sure chamber (2 levels of carbor mately 24 parl achieved in the , untary smoking. hemoglobin level 1% in subjects t but were signifi the study exposu volume curves, measured on a ments were mad, exercise under c posure condition vital capacity c with smoke exp and with exercist nitude of the cha crease in now in males. No other lung function wi Compared to of involuntary s mal subjects, inc ing cardiopulmG perience more pl acute exposure. tients with stab posed for 2 ho~ under condition room ventilatiohemoglobin co 1.26% and 1.30 sure during pe and poor room ly. With smoke moglobin levels good ventilation ventilation. Exe creased by 22% by 38%n in poorl smoke exposure. in heart rate a noted over bas during exercise posed to cigaret pain was repo segment depres grams was the s sure as without Dahms and a tients with bro normal subject size of the area in- dity and temperature id the extent of venti- )ngitudinal studies of cts (e.g., development aitivity or tolerance) d, and the smoke con- sible for these effects leagues exposed non- o smoke in an expo- ). Relatively constant monoxide (approxi- s per million) were hamber during invol- Peak blood carboxy- s were always less than efore smoke exposure antly greater during e. Lung volumes, flow and heart rate were 1 subjects. Measure- at rest and following ntrol and smoke ex- s. Flow at 2507c of the ecreased significantly sure at rest in males in females. The mag- nge was small: 7% de- males and 1407o in fe- consistent changes in re observed. he physiologic effects oking noted in nor- ividuals with preexist- nary disease may ex- ofound effects during ronow studied 10 pa- e angina pectoris ex- rs to cigarette smoke s of good and poor (28). The carboxy- centrations averaged o before smoke expo- 6ods simulating good ventilation, respective- ~ exposure, carboxyhe- rose to 1.7707a with and 2.28"/a with poor cise time to angina de- in well ventilated and ventilated rooms with A 10 to 15% increase d blood pressure was line at both rest and vhen subjects were ex- te smoke. At the time ed, the degree of ST ion oh electrocardio- tme with smoke expo- exposure. sociates studied 10 pa- ichial asthma and 10 passively exposed to 2505487729 smoke in an environmental cham (29)- Pulmonary function was not ured at 15-min intervals for I It a: smoke exposure. Blood carboxyher globin levels were measured beforet after the 1-h exposure. Carboxyher globin levels in subjects with asthma creased from 0.82 to 1.20%. In non subjects, the increase was from 0.62 1.06%a. The increases in carboxyher globin in the two study groups were significantly different, and the It corresponds to an ambient CO conc tration of 15 to 20 ppm. Asthmatic s jects had a decrease in forced vital pacity (FVC), forced expiratory volu in one second (FEV,), and mid-mt mum expiratory flow rate (MMEF) t level significantly different from t} pre-exposure values. The decreases asthmatic subjects were present at min but worsened over the course the hour to approximately 75%0 of pre-exposure values. Normal subjt had no change in pulmonary funct with this exposure. In this study, subjects were blinded as to the exposure and were lected because of complaints ab, smoke sensitivity. Because psychok cal factors are known to produce bn chospasm in asthmatics (30), such f tors potentially could explain the sults of Dahms' study. Shephard i coworkers (31), in a very similar exp. ment, subjected 14 asthmatic subjt to a 2-h cigarette smoke exposure i closed room (14.6 m'). The increa carbon monoxide levels (24 ppm) w similar to those predicted in the sti of Dahms and associates. No blc carboxyhemoglobin levels were me ured. Subjects were randomized s blinded to sham (no smoking) t smoke exposure and tested on two s arate occasions. Data were expressec percent change from the sham ex sure. No significant changes in FVC FEV, were observed between sham e smoke exposure periods. The limited data on normal men e women suggest that low level cigan exposure has minimal physiologic fects. However, even for normal s jects, the irritant effect and nuisanct such exposure may be considerable. tients with preexisting pulmonary ease, in particular asthmatic subje. may or may not have significant syr toms precipitated by low dose cigart smoke exposure. In spite of the lack of change in c boxyhemoglobin levels in these as matics, it is clear that other consti

836 ents of sidestream smoke that were not measured in these studies could have ir- ritating effects in some subjects. Clear- ly, some asthmatics are sensitive to low le~els of other highly soluble irritants su,:h as sulphur dioxide (SO,) (32). Thus, other gases, or particulate sub- stances from passive smoking impact- ing on the larger airways of asthmatics, could produce similar effects. However, further studies will be necessary to es- tablish these effects and the nature of the putative constituents. In contrast to th.: unclear situation in asthmatic sub- jects, acute passive exposure to cig- arette smoke does lower exercise time to chest pain under experimental condi- tions in patients with known stable an- gina. Whether such effects can occur in ncnexperimental situations remains to be documented. Chronic Elfects of Involuntary Smoking Early Childhood Illnesses Bronchitis/pneumonia and other lower respiratory illnesses are significantly more common in the first year of life in children who have one or two smoking parents (table 3). Harlap and col- leagues studied 10,672 births in Israel between 1965 and 1968 and observed that infants whose mothers said they smoked (as determined at an antenatal vi:;it) experienced a 27.5%u greater hos- pi:al admission rate for pneumonia and bronchitis than children of nonsmok- ing mothers (33). In addition, they demonstrated a dose-response relation- ship between the amount of maternal smoking and hospital admission for these conditions. Unfortunately, these data may be confound d by prenatal effects in that the molh s were report- ing antenatal smoking, not smoking during the first year of ife Thus it is unclear whether the res iratory illness experience reflects pre atal or post- natal maternal smoking or both. British investigators studying live births between 1963 an 1965 in Lon- don also observed an increased fre- quency of bronchitis an pneumonia in the first year of life ass iated with in- voluntary smoking that did not carry over to years 2 to 5. Thi effect was in- dependent of parents' wn symptoms and increased with t e amount of smoking by parents (34). However, bronchitis/pneumonia lso increased with an increased num er of siblings and this was not contr led for in the analysis. Fergusson and ass iates studied 1,265 New Zealand chil ren from birth to age three years (35). They demon- strated an increase in b th bronchitis/ pneumonia and lower espiratory ill- ness during the first 2 ears of life in children whose mother smoked. Cor- recting for maternal a e, family size, and socioeconomic stat is did not af- fect the linear relations ip between the degree of maternal sm king and the rate of respiratory illn ss. This effect declined with increasi g age of the child. Leeder and colleagu s, studying a British cohort of childre born between 1963 and 1965, demon rated that pa- rental cigarette smoki g was signifi- cantly associated with )ronchitis and pneumonia during the f rst year of life. A dose response associ ition persisted after correcting for pare tal respiratory TABLE 3 symptoms, sex of the child, number of siblings, and a history of respiratory ill- ness in the siblings (36). Pullan and Hay (37) studied children who were hospitalized with documented respiratory syncytial virus (RSV) infec- tion in infancy. They found a signifi- cant difference in the smoking habits of mothers at the time of the infection, compared to children hospitalized for other illnesses -including respiratory diseases for which RSV infection war not documented. These children re- ported an excess occurrence of wheeze and asthma and lower levels of pulmo- nary function, which persisted to age ten. The authors could not separate the possibility that the infection caused damage that persisted and affected the maturation of the lung or that these children were already more susceptible to severe RSV infection, possibly by a passive smoking effect. Respirnrory Symproln and Illness in Older Children Studies of children 5 to 20 yr of age from several different countries (table 4) have shown a positive relationship between parental cigarette smoking and the frequency of acute respiratory ill- ness (38), chronic cough, phlegm, and/or persistent wheeze (39-42), ton- sillectomy used as an index of severe re- current respiratory illness (43), and days in bed from respiratory illness (44). Some of the studies may be con- founded by an increased reporting of symptoms in the child by parents who smoke and have symptoms (38, 43), or by the child's own smoking habits (38, 39, 44), and not all studies show statis- tical significance (39, 41). A consistent EARLY CHILDHOOD ILLNESS AND INVOLP NTARY CIGARETTE SMOKING Atthor Subjects Findings H uiap and Davies 433) 10,6)2 birlhs, 1965 lo Hospitalized for bronchitisl Colley (39)' 1968, West Jerusa- pneumonia in first year lem, Israel of life 2.205 births, 1963 to Ouestionnaire on bronchitis! 1965. London, England pneumonia in first year of life Fergusson and 1,265 births, 4 months, Ouestionnaires on doctor or h associates (35) 1977. Christchurch, pilaf visits for bronchnis/pn New Zealand nla; check by hospital recor Assessment at 4 months, 1, ' These data are conslEered in a more expanded analysis Cy Leedar and co,vo.Xen (te). WEiSS, TANER, SCHENKER, AND gpEIZER Illness Ratesl100 by cigarettes per day 0 1 to 10 11 10 20 20. Comments 9.5 10.8 16.2 ---~31.7 Smoking history obtained l.6 10.4 111 15.2 = anlenalally - ma lernal smoking only Asymptomatic parenls. 10.3 15.1 14.5 23.2 = Symptomatic parents N . E11 Neilhcr controlled for 0 r.0 12.8 13.4 Maternal number ot siblings nor sex Of smokers Combined effect sig. Ul A me only nificant for maternal ~ s. 7.0 4.6 8.8 Paternal smoking first year 3 yr. only of Ilte only W 0

s'-,rE OF THE ART: HEALTH EFFECTS OF INVOLUNTARY SMOKING RESPIRATORY SYMPTOMS IN CHILDREN I Respiratory Symplon Authors Subjects or Illness Colley and 2,426 children aged Chronic cough by queslio associates (34) 5 lo 14. England completed by parent Bland and 3,105, 12 lo 13 yr Cough during day or at nlg coworkers (40) old children who did not admit to Morning cough . ever smoking clga- rettes, England Cameron and 158, 6 to 9 yr old associates (38) children, parents completed tele- phone queslion- naire. U.S.A. Said and 3,920 10 to 20 yr coworkers (43) old children. France L,•yvitz and 1.252 children less aurrows (41) than 15 yr old, U.S.A. Weiss and 650, 5 to 9 yr old associates (42) children, U.S.A. Ware and 8,528 children 51o 9 coworkers (50) yr of age with 2 parents of known smoking status in 6 U.S. cities Respiratory Illness with r¢l activity and/or medical consultatlon in last year Tonsillectomy and/or adem tomy, generally before al indicalor of frequent resl lory tract Infecllon Persistent cough, phlegm, wheeze, or asthma, broni trouble, or emphysema Persistent cough and phie6 Persistent wheeze Respiratory illness last yea Persistent cough Persistent wheeze Ootlge 451) 628 3rd and 4th Wheeze grade children in Phtegm 2-parenl house- Cough holds. Ouestion- naire response of parents. U.S.A. finding in all reported data is an in- crease in symptoms with an increased n;-lmber of smoking parents in the me. --.\'heezing symptoms and asthma sodes hace heen sludied Icss fre- L:uently and nilh Iess consistent results. O'Connell and Logan (45) identified 37 asthmatic children who ltere "both- ered" by parental ciearette smoke. Var- ents of 20 of ihc children stopped smoking and 18 of 20 (900-o) of Ihe chil- dren had an improvement in swnptoms. The control group consisted of 15 chil- dren (2 were not followed up) whose parents did not stop smokin_c. Only 4 of 15 (27(Vo) of thesc children inlproved. In addition to possible bias in the selec- tion of cases and in the reporting of •>mptoms, subjective criteria for im- hrovement and an unclear duration of follow-up 17aw this study. British worker hort, demonstra dence of wheezi among nonasthn 2 parents who sr examined bY log tal smoking was dictor of occurn future occurrend subgroup ofthe~ asymptomatic pq leagues (46) werd ABLE 4 RELATION TO INVOLUNTARY SMOKE EXPOSURE Rates per 10D by No. of Smoking Parents 0 1 2 Comment naire 15.6 17.7 22.2 Trend significanl. Possible that symptoms parents could result In reporting bias. Active smoking In children could also bp results. Bias unlikely to explain full ette of trend. it 16.4 19.0 23.5 Self reported symptoms and smoking hish collected simultaneously from children. 1.5 2.8 2.9 Difference belween morning and daytim cough suggested as different diseases. However, could be difference In exposur in that exposure more likely daytime rather than when asleep. ricted 1.3 7.4 Illness reporting not verified. Not clear ho, reporting adult was related to child. Iidec- 28.2 41.4 50.9 Sell reporting by children. Not clear that `e 5 as smoking habits of parents at time of Iira- reporting directly relate to exposure roxim l r a t 10 li :hial m y pp a + yea s ear er. e Trend but gradlent smoking no signi across catego ficant ries Higher rates in symptomatic households v trends persisting but no1 significant. 1.7 2.7 3.4 Trend not significant. 18 6.8 11.8 Significanl lrentl. 12.9 13.7 14.8 Adjusted for age, sex, and city cohort 7.7 8.4 1D.6 effecls. Significant trends. 9.9 11.0 13.1 27.6 27.9 40.0 All trends significant. Some of effect 6.4 109 12.0 might relate to parental symptoms; 14.6 23.0 27.8 however, not likely 10 influence trends. i, studying a birth co- ed an increased inci- tg over a 5-yr period atic children who had ioked. However, when stic regression, paren- not a significant pre- nce of wheeze or the ~ of asthma (40). In a ohort, 861 children of rents, Leeder and col- unable to show a sig- nifican( trend in asthma-wheeze symp- toms by increa ing level of parental smoking over a : of 650 children f and associates (4 trend in the r( chronic wheezinl smoking; the rat and 11.8%, for parents. Althoui -yr period. In a study to 10 yr of age, Weiss 12) showed a significant ~portcd prevalence of with current parental es were 1.85%p, 6.85%p 0, 1, and 2 smoking ;h the data given are for all households, when the anal: was restricted to those househc where neither parent reported syr toms, the results were identical, t suggesting that in this population, : nificant reporting bias is not respoa ble for the observed results. Gr macher and coworkers (47) studied I populations of children, newborn tc yr of age. They found a significant sociation between parental reportin): children's asthma and maternal sm ing. Maternal smoking alone was as ciated with approximately 20% of asthma. The effect persisted when and sex of the child, allergies, and ft ily income and education were c trolled for in the analysis. No com was attempted for the children's c smoking habits. In addition, the pot tial for an increased reporting of syr toms in children of symptomatic r 2505487731

938 ents exists in these data (47). Other population-based studies (41, 48, 49) have failed to confirm these obser- vations. Lebowitz and Burrows (41), in a group of 463 current smoking and never smoking households with chil- dren below age 15, found trends for a variety of symptoms including wheeze most days in the direction of excess rates in households with smokers; how- ever, these rate differences did not achieve statistical significance. In the same study among 849 households with older children and adults there were es- sentially no differences for any symp- tom prevalence between current smok- ing and never smoking household members. In the general population study of Schilling and coworkers (49), in spite of reporting no association be- tween wheeze and involuntary smok- ing, the number of children available for analysis was really too small to ade- quately assess this question. In a preliminary report from one of the largest studies currently under way, Speizer and associates (48) reported no association of persistent wheeze with the presence of some smoking in the household in approximately 8,000 chil- dren 6 to 11 yr of age in 6 communities. Subsequent analyses of these same cohorts, with the addition of approxi- mately 2,000 more children and a more detailed assessment of the smoking be- havior of each parent, revealed a persis- tent relationship that increased with the amount of maternal smoking and was only modestly affected by taking into account the parents' own symptoms (50). Dodge (51), studying 3rd and 4th grade children, found that although symptoms including wheeze were re- lated both to the presence of symptoms in the parents and the number of smok- ers in the households, after excluding the potential effect of reporting bias by symptomatic parents, the gradient of the wheeze effect persisted. Thus, the relationship of chronic wheezing to in- voluntary smoking, although not uni- formly confirmed, is likely to be real. In summary, several studies suggest important increases in severe respirato- ry illnesses in very young (less than 2 years old) children of smoking parents. Young children may represent a more susceptible population for adverse ef- fects of involuntary smoking than older children and adults. The amount of time spent with active smokers, partic- ularly by children under 2 yr of age with smoking mothers, may be the im- portant factor. How influences this risk is ic respiratory sympt n utero exposure nknown. Chron- 0 ms in older chil- e) also are linked e exposure. Few dren (5 to 20 yr of a to involuntary smo data are available on sure necessary to pro the implication of th future lung growth he level of expo- uce symptoms or se symptoms for nd development. No data currently ar available on the relationship of pas ive smoking to other putative risk fa such as atopy, res and increased levels tors for wheezing ratory infection, f airways respon- siveness; nor are suf icient data avail- able to estimate ho these early ex- posures affect the ccurrence of re- spiratory disease 1 er in life. The characteristics of the hild who may be susceptible to this ty e of exposure are unknown. However, he data are suffi- ciently consistent to uggest that pedi- atricians should utinely inquire about smoking habit of parents when caring for children with chronic or recurrent respiratory ymptoms and ill- nesses. It would al be prudent to advise parents of chi ren who are suf- fering from recurrent Ilnesses or persis- tent wheeze or asth a not to smoke. Pulmonary Functio and Involuntary Smok ng Relatively few studi have been pub- lished relating routi e spirometric in- dexes in children and adults to involun- tary smoking in the me or workplace (table 5). In childre almost an equal number of studies n be found that find significant as ciation between passive smoking an level of function as do not. In some s dies, there seems to be a dose respons relationship (42, 52); i.e., the greate the number of smokers in the home, the lower the level of function. When alyzed by multi- ple regression techni ues, however, ma- ternal smoking habit have the greatest impact (42, 44, 53) and also suggest a dose response elationship (50). Younger children see to be more ad- verscly affected than Ider children (42, 52), and clearly there is an added effect if children themselv s smoke (52). In contrast, several inv out in warm, dry ar States fail to find t pulmonary function parental smoking ha All of these studi having quantitative e smoke exposure. In U.S., the actual amo posure to children m tigations carried as of the United e association of in children with its (49, 54). s suffer from not timates of passive ome areas of the nt of indoor ex- y be significantly WEI55, TADER, SCNENKER, pND SFEIZER less, either because these children are in fact outdoors more, or because the ven- tilation rates in these climates are higher than in generally colder areas. Thus, real differences in exposure to children in these variously reported studies may be occurring. Even less data exist for adults. White and Froeb reported on 2,100 asymptomatic adults drawn from :: population about to enter a physica fitness program. Studying FEV, an, MMEF as percent of predicted, the} demonstrated statistically significant decreases in these tests in nonsmokers exposed to tobacco smoke in the work environment compared with nonex- posed workers (55). The decrement was comparable to that seen in smokers inhaling I to 10 cigarettes per day. However, the absolute magnitude of the difference in mean levels of func- tion in the smoke-exposed and unex- posed groups was quite small: 160 ml (5.5%) for FEV, and 465 ml/s (13.5%) for MMEF. Carbon monoxide levels were measured in the workplace and ranged from 3.1 to 25.8 ppm. Their measurement of carbon monoxide levels in the workplace correlated with worker histories and level of pulmo- nary function. Potential biases could have affected the results of this study. The population was self-selected; re- sponse was related to current work- place exposure and did not account for persons who changed jobs, and ex- smokers seemed to have been left out. Comstock and associates examined 1,724 subjects drawn from two separate studies in Washington County, Mary- land (56). They found no statistically significant greater risk of an FEV, less than 80%a of predicted in male non- smokers exposed to wives' cigarette smoke at home. Nonsmoking .vomen with smoking husbands were not avail- able for analysis in this study. Schilling and coworkers (49) were also unable to find an effect of this passive smoking in adults. Both these latter studies included adults in their samples who were rela- tively young and generally would not have had an opportunity for long-term passive exposure in adult life. This point was brought out by the recently reported large study from France (57). Kaufmann and associates (57) reported from a 7-city investigation in which a total of 7,818 adults were studied. In a subsample of 1,985 nonsmoking wo- men 25 to 59 yr of age, in which 58%a Were exposed to a smoking husband,

$T.TE OF THE ART: HEALTH EFFECTS OF INVOLUNTARY SMOKING TAB E 5 PULMONARY FUNCTION IN CHILDREN AND A ULTS EXPOSED TO INVOLUNTARY SMOKING Population and Author Age Range Schilling (49) 816 children 7 to 17 yr of age In CT and SC Tager (52) 444 children 5 to 19 yr of age In East Boston, MA R'.-as (42) 65o chlldren 5 to 9 yr of age in East Boston, MA Vedal and asso 4,000 children 6 to 13 clales (sub years of age milled 10 Am Rev Respir Dis) Lebowitz (41) 271 households with complete smoking hislorles of both parents and pulmonary func- tion of children 6+, Tucson, AZ Dodge (51) 558 children 8 to 10 yr of age in AZ Hasselbladl (53) 16,689 children 5 to 17 yr of age. 7 geographic regions in U.S. Speizer (481 8,120 children 6 lo 10 years in 6 U.S. cities Ware (50) 10,000 children 6 to 11 yr of age in 6 U.S. cities White (55) 2.100 adults in San Diego, CA Comstock (56) 1,724 adults in Washington County, MD Kauffman (57) 7,818 adults in 7 cities in France, selected subgroups there was a significant difference in level of MMEF between truly non- smoking women and women of com- parable ages exposed to passive smok- ing. This effect did not become apparent until age 40. The changes were small and although not adjusted for differ- ences in body size, lend credence to the possible effect of long-term exposure in adult life. The physiologic and clinical signifi- cance of these small changes in pulmo- nary function is unclear. In addition, none of these studies, with the excep- tion of that of White and Froeb (55), attempts to characterize exposure more definitively than the number of smok- ers in the home. All of the variables Pulmonary Function Measure FEV, as percent predicted MMEF In standard deviation units MMEF In standard deviatlon units FEV,,., FVC, Vmax,,, Vmas,,, Vmax.. FEV„ FVC, Vmax..r, Vmax,r tlerlved fro MEF,V curves, ex-pressed as standa d deviation units FEV, by age change FEV,1H' per year FEV,,, as percent predicted FVC and FEV, as percenl predicted FEV, and FVC FVC, FEV„ and MMI~ as percent predict d FEV, as percent predicted FEV„ FVC, and MMEF Outcome No effect of parental smoking Significant effect of parental smoking Signiflcant effecl of parental smoking FVC positively associated flows, negatively associated No effect of parental smoking No effect of parental smoking Significant effect of mothei s smoking but not father's smoking No effect for FEV, or FVC FVC positively associated with smoking, FEV, negatively associated Significant effect of office exposure to Involuntary smoke No effect of wives' smoking on husbands' pulmonary function Significant effect In wives of smoking husbands in afi measures. Only for MMEF in husbands of smoking wives such as ventilation room size, number of rooms in the ho e, duration of con- tact with the active smoke, and number of cigarettes smok d could significant- ly influence total e posure. Differences in these exposure ariables and inade- quate characteriza on of exposure may explain the diffe nces in these study results. All curren data (table 4) are cross-sectional, an the relationship to long-term change in lung function, either growth in c ildren or decline in adults, is unclear. Most investigat rs would agree that the natural histor of the development of chronic obstru tive airways disease, particularly as re ted to active ciga- rette smoking, is relatively slow but Comments 939 No control for slbship size or correlation of siblings' pulmonary function. Child's lunction first adjusled for parent's function; this adjustment may have marked any effect of parental smoking. Analysis controlled for sibship size and correlation of slblings' pulmonary funcllon Analysis conlrolled for sibship size and correlation of siblinga' pulmonary funclion Flows dose response with amount smoked by mother Suggestion that real differences in Indoor levels of exposure compared to more northern climates may be occurring Potential bias In participation rates. Crosssectional data not conlrotled lot children's height. Annual change in FEV,1H• at ages 8, 9, 11 conals. tently greater In nonsmoking house, holds than 2-parent smoking house- holds; however, statislical test not significant. Large number of children excluded because of invalid pulmonary function data or missing parental smoking data A recent analysis of this cohort demom strated an effecl for FVC and FEV, (54) FEY, dose response with amount smoked by mother Potential bias in selection. Only assessed current cigerette smoke exposure Includes adults 20+ Not adjusted for height. Dose response lo amount of husbands' smoking for MMEF in wives. No effect below age 40. 2505487733 unrelenting frocess. The disease be comes manifest in mid-adult life anc progresses, as long as the individua continues to smoke, in an unrelentinj fashion despite modern modes of ther apy. The epidemiologic data sugges that the major difference betweer adults destined to develop severe ob structive disease and patients who pre sent the obstruction, is a subjectivi matter of perception of time of onse of disability by the individual patien (58). Because impressive active smok ing histories are generally present, i seems unlikely that adult nonsmoker with no other risk factors for chroni. respiratory disease are likely to sustait -enough damage to their lungs fron

940 involuntary smoking to lead to sympto- matic air-flow obstruction. However, there may be groups of subjects at high risk. For example, adults with asthma or with other evidence of increased levels of airways responsiveness may be more susceptible to exposures. More fundamentally, causes of asthma and increased levels of airway responsive- ness are largely unknown, and the rela- tionship of these disorders to passive smoking is unclear-either in terms of direct precipitating causes or in increas- ing responsiveness to other environ- mental agents. In summary, several studies have demonstrated with reasonable certainty that passive smoking affects lung func- tion, particularly in young children. The degree to which this effect in child- hood is one of the important risk deter- minants in identifying those persons who will develop chronic airways dis- ease in adult life is unknown. The con- cept of multiple childhood risk factors predisposing to chronic respiratory dis- ease in adult life is actively being pur- sued by a number of investigators at the present time. Small changes in pulmo- nary function produced in childhood as a result of passive exposure to cigarette smoke or other putative risk factors could affect the rate of attainment of maximum lung size in adult life. The direct effect of involuntary smoking in childhood on maximum obtained lung growth is likely to result in less than a 5%u reduction in lung size. TFis reduction is not likely to produce a detectable health consequence. Even if ;uch an effect, not yet substantiated in adults, were to occur over an adult's lifetime, it would not in and of itself produce significant airways obstruc- tion. On the other hand, the best pre- dictor of who is likely to develop ob- structive airways disease is the level of pulmonary function measured at an earlier point in adult life (59). Thus, even modest reductions of pulmonary function may be an indicator of who is at risk for developing obstructive dis- eaee when exposed more directly to per- sonal (e.g., cigarette smoking) or other em,ironmcntal (e.g., occupational or general air pollution) agents. Further exploration of these hypoth- eses will require long-term commitment of both fiscal and personnel resources to follow cohorts of children with well characterized exposures through the period of maximum lung growth and beyond to a point when active exposure to ciga people with documente involuntary exposure at controls can be assessed From the public heal an indirect effect of inv sure to cigarette smoke r mind. There is no questi rect effect on pulmonaactive smoking can ret mined in young teenager a minimal number of cig (Tager [B, personal co Furthermore, there is lil parental smoking is a s dictor of children takin smoking (60). Thus, chil ing parents are twice a come smokers themse therefore at risk of d effects. : the effects of rette smoke in d exposure to id appropriate th perspective, aluntary expo- nust be kept in on that the di- -y function of dily be deter- t smoking only arettes per day rnmunication). tie doubt that ignificant pre- g up cigarette dren of smok- I likely to be- ves, and are irect smoking Lung Cancer and Involuotary Smoking Recent studies from Japa (62), and the United St suggested that wives exl husband's cigarette smok increased risk of lung c< Hirayama studied 91,' ing married women. Th follow-up (1966 to 197S deaths from lung cancer Diagnosis was establishet tificate, but the histologi the lung cancers wer4 r The lung cancer incider 15.8/100,000 person-yea with nonsmoking hus 100,000 for women wh smoked I pack per d~ 100,000 for women wh smoked more than I pacl relative risk of lung canc whose husbands smoket day was found to be 1.6 women with nonsmokii Women whose husbands than I pack per day had~ of 2.1 compared to wom bands did not smoke. Fo, women who actively smo' of lung cancer 2.1 time that of all involuntary s women, and the involunexposed women had a ris cer 1.7 times greater tha women with nonsmoki (61). Although the study cized because of the met to summarize the strati n (6)), Greece tes (63) have osed to their e may have an ncer. 40 nonsmok- e 13 years of ) yielded 346 in the cohort. by death cer- cell types of ot identified. ce rates were s in women ands, 24.4/ ose husbands `y, and 29.6/ bse husbands PPLPL per day. The er for women ~ I pack per hcompared to g husbands. smoked more a relative risk Fn whose hus- r comparison, ced had a risk greater than noke-exposed ary smoking- k of lung can- i nonsmoking ng husbands tas been criti- odology used ied data, the WEISS. TAGEn, SCHENKEa, AND SPEIZER technical flaws probably do not invali- date the outcome (63-66). A second study that investigated the relationship between lung cancer and passive smoking was performed by Tri- chopoulos and associates (62). A case control design was used to study 40 lung cancer cases in never-smoking women admitted to an Athens hospital between 1978 and 1980. The contro! subjects were women admitted to a ho~ pital for orthopedic disorders. Nor smoking women with smoking hus bands were 2.4 times more likely to have lung cancer than nonsmoking women whose husbands did not smoke. The risk increased to 3.4 times for wives whose husbands smoked more than 1 pack per day. A similar relationship held if the data were expressed in terms of the total lifetime cigarette consump- tion of the husband. Garfinkle examined lung cancer mortality rates in the United States using data from the American Cancer Society's prospective study and from the Dorn Study of Veterans (63). He ex- amined lung cancer mortality in 176,739 nonsmoking women who were married to men with various smoking histories. Compared to unexposed women, women with husbands who smoked less than a pack a day had a relative risk of lung cancer of only 1.3. Women of hus- bands who smoked greater than one pack per day had a relative risk of lung cancer of only 1.1 compared to unex- posed women. Neither of these results were statistically significant. The results were unchanged when adjustment was performed for age, race, education, residence, and husband's occupation (63). Several methodologic differences exist between the American and the Japanese studies. A potentially impor- tant factor is that the American Cancer Society study lacked smoking data on 72.9% of the husbands of nonsmoking women in comparison to only 27.7% in the Japanese study (64). This lack of information may have created biases in the data. A greater number of working women, larger homes and a higher di- vorce rate in the United States all could serve to potentially account for the di6 ferences in results between these studies (64). Previous estimates would have esti- mated the attributable risk of lung can- cer due to passive smoking to be 30v/u greater in nonsmokers exposed regular-

STA1 E OF THE ART: HEALTH EFFECTS OF INVOLUNTARY SMOKING ly to passive smoke compared to non- smokers not exposed (67). In fact, all three studies are consistent with such an effect. Conclusion While current data document acute ir- ritant effects and minimal physiologic changes in normal adults, the data for subjects with cardiopulmonary disease, p-.r:.icularly asthma, are conflicting. .'-,:;:itionally, better data on naturally c:._urring exposures and the relation- ship of such exposure to symptoms will be necessary to reach any conclusions about the medical importance of pas- sive smoking in those with cardiopul- monary disease Effects of parental smoking, particu- larly maternal smoking, on childhood respiratory illness experience have been dc:umented in several studies. The in- ft:.-nnce of in utero exposure on subse- qucnt postnatal lung function is un- i known and could influence these results. 1"-Further work is necessary to relate the effects of parental smoking to other childhood risk factors (notably infec- tion, airways responsiveness, and atopy) and to changes in level of pulmonary function. jThe child may be a more vulnerable host than the adult or the increased symptoms seen in children may reflect the amount of time spent with smoking parents. Obviously, small children are less likely to be exposed to `other potentially irritating pollutanl~ Longitudinal studies of children are necessary to assess the contribution of early passive smoke exposure to the subsequent development of symptoms rillnesses and to determine whether :se illnesses or exposures lead to an . teased susceptibility of developing structive airways disease. The prob- iem is made more complicated by the long-term nature of these studies and the tendency of children whose parents smoke to become smokers themselves. Given the magnitude of the physiologic changes in adults observed cross-sec- tionally, involuntary smoking alone is unlikely to have important effects on adult decline in pulmonary function. Certain adults with other potential risk factors, such as airways responsiveness or atopy, may be at risk, and these spe- cial population groups should be furth- ~r investigated. Involuntary exposure to cigarette smoke represents the classic low-dose exposure problem such, carefully d adequate exposup sary to address tt above. In collec data, in addition other potential il such as NO, fror formaldehyde frot foam insulation ( products from I stoves (68) also investigation. Wit upwards of 80%a indoors (69) and rising energy cost proved insulation exchanges, the in likely to become a tant health conce, potentially long ts Adc Since the submissior additional report (Tat A, Rosner B, Spe Study of Maternal7 Function in Childr le 209: 699-703) has that in children of rate of increase in adolescent growth i dren of nonsmokin Reft 1. U.S. Public Heal' health. A public heah Department of Heahl Public Health $ervici Education and WN (PHS) 79-50066 19791 2. Abel EL. Smokin view of effects on gr offspring. Hum Biol 3. weinbergerSE, W In: Burrosv GN, Ferri cations during pregna phia: WB Saunders, I 4. Sterling TD, Dimi, byproduct levels of to view of the literature. , 1982; 32:250-9. 5. Brunnemann KB, The influence of lob mosphcre IJ. Volatile main and sidcstream lion to indoor polluti joint symposium on New Orleans, LA 1 Sociely 1978; 876-80. 6. Weber A, Fische smoking in experime viron Res 1979; 20:20 7. Hinds WC, First nicotine and tobacco Med 1975; 292:844-5 g. Spengler JD, Doc 9. in epidemiology. As son JM, Ferris BC. Long term measurements Signed studies with respuable sulfates and panicles inside and ot data will be neces- side homes. Atmos Environ 1981; 15:23-30. questions outlined 9. Comroe 1. Physiology of respiration. CI cago, IL: Chicago Year Book Medical Publisha ing these exposure 1974, p. 192. Ito cigarette smoke, 10. Jones RM, Fagan R. Carboxyhemoglobin door air pollutants nonsmokers: a mathematical model. Arch E viron Health 1975; 30:184-9. gas stoves (48 50) , , urea formaldehyde 8), and combustion replace and wood will require further h the estimate that of our time is spent the knowledge that s are leading to im- and less frequent air 7oor environment is i increasingly impor- -n for the near and rm future, endwn i of this manuscript, an ;er IB, Weiss ST, Munoz zer FE. Longitudinal imoking in Pulmonary ll. Coburn RF, Fonter RE. Kane PB. Considl ations of the physiologic variables that determi the blood carboxyhemoglobin concentration man. 3 Clin Invest 1965; 44:1899-910. 12. Stewart RD. The effect of carbon monoxi. on humans. I Occup Med 1976; 18:304-9. 13. Stewart RD. Baretta ED, Platte LR, er , Carlwxyhemoglobin levels in American blor donors. JAMA 1974; 229:1187-96. 14. Matsukura S, Sakamoto N, Seino Tamada T. Matsuyama H. Muranaka H. Co tinine exeretion and daily cigarette smoking habituated smokers. Ctin Pharmacol Ther 197 25:555-61. 15. Langone JJ, Gjuka HB, Vunakis HV. Nic tine and its metabolites. Radioimmunoassays f nicotine and cotinine. Biochemistry 1973; 1 5025-30. I6. Williams CL, Eng A, Botvin GJ, Hill Wynder EL. Validation of student's self report cigarette smoking status with plasma cotinl levels. Am J Public Health 1979; 69:1272-4. 7. N Engl 7 Med 1983; 17. Feyerabend C, Russell MAH. Rapid g; ippeared, which shows liquid chromatographic determination of c smoking mothers, the tinine in biological Duids. Analyst 1980; 105:99 'EV, and MMF during Joot. slowed relative to chil- 18. Russell \IAH, Sutton SR. Feyerabend , mothers. Vese)' Cl. Long term switching to low-tar Ic rences h Service. Smoking and service rerview:1979. U.S. , Education and Welfare; Department of Health, are Publication Number II:6-24. g during pregnancy: a re- owth and development of 1980; 52:593-625. iss ST. Pulmonary disease. TF, eds, Medical compli- icy. 2nd edition. Philadel- 981. pp. 405-34. h H, Kobayashi D. Indoor sacco smoke: a critical re- Air Pollm Control Assoc nicotine cigarettes. Br y Addict 1982; 77:145! 19. Bottoms SF, Kuhnet BR, Kuhnet PM, Ret AL. Maternal passive smoking and fetal sen thiocyanate levels. Am J Obstet Gynecol 19F 164:787-91. 20. Cohen JD, Bartsch GE. A comparison t tween carbom yhemoglobin and serum thiocs nate dcterminations as indicators of cigaro smoking. Am 1 Public Health 1980; 70:284-8 21. Radford EP Jr. Hunt UR. Polonium-21Q volatile radioelement in cigarettes. Science 19( 143:247-9. 22 Little J, .McGendy RB. Measurement of pl tonium-210 in human blood. Nature 1966; 2 g42-3_ 23. Speer F. Tobacco and ihe nonsmoker. study of subjective symptoms. Arch Envin Health 1968; 16:443-6. 24. Weber A, Jermini C, Grandjean E. Irritati 'Adams JD, Hoffman D. effects on man of air pollution due to cigarc Leco smoke on indoor at- smoke. Am J Public Health 1976; 66:672-6. nd specific nitrosamines in 25. Slavin RG, Hertz M. Indoor air pollution smoke and their contribu- study of the Thirtieth Annual Meeting of t )n. Proceedings of the 4th American Academy of Allergy. Paper present environmental pollutants, at the American Academy of Allergy, Thirtie 977. American Chemical Annual Meeting, San Diego, California, Fe ruary 15-19, 1975. T, Grandjean E. Passive 26. Johansson CR. Tobacco smoke in room a at and fall conditions. En- an experimental investigation of perception a. i-14. irritasing effects. Building Service Engineer 19~ MW. Concentrations of 43:254-62. n public places. N Engt J 27. Pimm PE, Silverman F. Physiologic effn of acute passive exposure to cigarette smol ey DW, Thurer WA, Wolf- Arch Environ Health 1978; 33:201-13. ik I 2505487735

28. Aronow WS. Effect of passive smoking on angina pectoris. N Eng) J Med 1978; 299:21-4. 29. Dahms TE, Bolin JF. Slavin RG. Passive smoking: effects on bronchial asthma. Chest 1981; 80:530-4. 30. McFaddcn ER Jr, Luparello T, Lyons HA, er al. The mechanisms of action of suggestion in the induction of acute asthma. Psychosom Med 1964 31:734-8. 31. Shephard RJ, Collins R, Silverman F. Pas- sfve exposure of asthmatic subjects to cigarette smoke. Environ Res 1979; 20:392-402. 32. Sheppard D, Saisho A, Nadel JA, Boushey HA. Exercise increases sulphur dioxide induced bronchoconstriction in asthmatic subjects. Am Rev Respir Dis 1981; 123:486-91. 33. Harlap S, Davies AM. lnfant admissions to hospital and maternal smoking. Lancet 1974; 1: 529-32. 34. Colley JRT, Holland W W, Corkhill Ri, In- fluence of passive smoking and parental phlegm on pneumonia and bronchitis in early childhood. Lancet. 1974; 2:1031-4. 35. Fergusson DM, Horwood LJ, Shannon FT, raylor B. Parental smoking and lower respiratory Ilness in the first three years of life. J Epidemiot Community Health 1981; 35a180-4. 36. Leeder SR, Corkhill R. Irsvig LM, Holland W W. Collcy lRL Influence of family factors on :he incidence of lower respiratory illness during :he first year of life. Br ] Prev Soc Med 19761 30: 203-12. 37. Pullan CR, Hey EN. Wheezing, asthma and '>ulmonary dysfunction 10 years after infection with respiratory syncytial virus in infancy. Br Med J 1982; 284:1665-9. .18. Cameron P, Kostin 1S, Zaks 1>4, el aL The health of smokers' and nonsmokers' children. J Allergy Clin Immunol 1969; 43:336-t1. 39. Colley fRT Respiratory symptoms in chil- dren and parental smoking and phlegm produc- tion. Br Med J 1974; 2:201-4. <I0. Bland hf, Bewley BR, Pollard V, Banks MH. Effect of children's and parents' smoking on re- spiratory symptoms. Arch Dis Child 1978; 53: 100-5. 41. Lebowirz Mf). Burrows B. Respiratory symptoms related to smoking habits of family adults. Chest 1976; 69:48-50. 42. Weiss ST, Tager IR, S Persistent wheeze: its relati illness, cigarette smoking a function in a population sa Rev Respir Dis 1980; 122:6 43. Said G, Zalokar J, Parental smoking related t tonsillectomy in children. J ity Health 1918; 32:97-101 e{zer FE, Rosner B. nship to respiratory level of pulmonary ple of children. Am 7-707. louch 3, Patois E. adenoidectomy and pidemiol Commun- 44. Bonhaim GS, Wilson I4W. Children's health in families with cigarette s Health 1981; 71:290-3. 45. O'Connell EJ, Logan in childhood asthma. Ann 46. Leeder SR. Corkhill R WW. Influence of family during the first five years Med 1976; 30:213-8. 47, Gortmacher SL. Walk Ruch-Ross H. Parental smo childhood asthma. Am I 72:574-9. 48. Speizer FE, Ferris BI ler 1. Respiratory disease function in children associ sure. Am Rev Respir Dis 19 49. Schilling RSF, Istai A Schoenberg JR Bouhuys disease and smoking in fami 1977; 106:274-83. 50. Ware 1H, Spiro A IfI, FE, Spengler 1D, Ferris BG. B. Parental smoking Ilergy 1974; 3:142-5. Irwig LM, Holland aors and wheezing life. Br J Prev Soc r DK, Jacobs RH, ing and the risk of ublic Health 1982; hop YMM, Speng- es and pulmonary ed with NO, expo- 0; 121:3-10. Hirs SL, Beck ST, nction, respimtpry, es. Am J Epidemiol ockery DW, Speizer espiratory illness in children exposed to ambien air pollution, cig- arette smoke, and gas stoves 'n six cites. Am Rev Respir Dis (submltted for p blication). 51. Dodge R. The effccts of indoor pollution on Arizona children. Arch Envi on Health 1982; 37: 151-5. 52. Tager IB, Weiss ST. Ro$ner B, Speizer FE. Effect of parental cigarette oking on the pul- monary function of childre Am 3 Epidemiol 1974 110:15-26. 53. Hassdbladt U. Humble{ CG, Graham MG. Anderson HS. Indoor envtt onmental determi- nants of tung function in ch Idren. Am Rev Re- spir Dis 1981; 123:479-85. 54. Lebowitz MD, Armet D Knudson R. The effect of passive smoking on ulmonary function in children. Environ tnternat nal 1982; 8:371-3. WEISS, TA6ER, ECNENKER, AND aPE12ER 55. White JR. Froeb WF. Small-airways dys- function in nonsmokers chronically exposed to tobacco smoke. N Engl J Med 1980; 302:720-3. 56. Comstock GW, Meyer MB, Helsing KJ, Tockman MS. Respiratory effects of household exposure to tobacco smoke and gas cooking. Am Rev Respir Dis 1981; 124:143-8. 57. Kauffmann F, Tessier JF, Oriol P. Adult pas- sive smoking in the home environment: a risk fac- tor for chronic airflow limitation. Am I Epide- miol 1983; 117:269-80. 58. Speizer FE, Tager IB. Epidemiology o` chronic mucus hypersecretion and obstructiafrways disease. Epidemiology Reviews 1979; 124-41. 59. Fletcher CM, Peto R. Tinker CM, er a!. Nat- ural history of chronic bronchitis and emphy- sema. Oxford; Oxford University Press, 1976, pp. 1-147. 60. United States Department of Health, Edu- cation and Welfara Teenage smoking, national patterns of cigarette smoking, ages 12 through 111, in 1972 and 1974. DHEW Publication No. (NIH) 76-931. Bethesda, MD: National Institutes of Health, 1976:1-25. 61, Hirayama T. Non-smoking wives of heav> smokers have a higher risk of lung cancer: ~ study from Japan. Br Med 1 1981; 282:183-5. 62. Trichopoulos D- Kalandidi A. Sparros L, MacMahon B. Lung cancer and passive smoking. Int J Cancer 1981; 27:1-4. 63- Garfinkle L. Time trends in lung cancer mortality among non-smokers and a note on pas- sive smoking. I Nat Cancer Soc 1981; 66:1061-6. 64. Hirayama T Letter to the editor. Br Med J 1981; 283:916-7. 65. Harris JE, DuMouchel WH. Lerter to the editor. Br Med J 1981; 283:915. 66. Mantel N. Letter to the editor. Br Med 1 1981; 283:914-5. 67. t.eeder SR. Breathing other peoples'smoke. Br Med J 1978; 2:453-4. 68. Frank R, Itbow'itz MD. The risk of staying in. Am Rev Respir Dis 39914 521 -2. 69. Van der Lende R. Health aspects related to indoor air pollution. Inr 3 Epidemiol 198tr 9: t95-7.