4. Of the studies in which dose-specific prevalence rates were examined, strong dose-response relationships between cigarette smoking and symptoms of COPD were generally demonstrated. 5. The relationship between cigarette smoking and COPD mor- tality has been demonstrated in the United States, Canada, Great Britain, and Ireland ; strong associations between cigarette smok- ing and COPD morbidity have been shown in the United States, Canada, England, Australia, Finland, Sweden, France, Belgium, Hungary, and Japan. 6. Epidemiologic evidence from many countries indicates that, for both sexes, symptomatic and asymptomatic cigarette smok- ers have greater impairment of pulmonary function than do nonsmokers.1 7. Previous evidence indicates that cessation of smoking re- sults in lower death rates from COPD, improved pulmonary func- tion, and a decrease in the prevalence of pulmonary symptoms. 8. Prospective and cross-sectional analyses of data reveal that pipe and cigar smokers have higher mortality rates from chronic bronchitis and emphysema than do nonsmokers, but lower rates than those of cigarette smokers. Pipe and cigar smokers have a higher prevalence of respiratory symptoms than do nonsmokers. The limited data on pulmonary function studies in pipe and cigar smokers are, thus far, inconclusive. 9. Available data suggest that although air pollution may con- tribute to the prevalence of symptoms of respiratory disease, cigarette smoking is far more important in producing respira- tory disease. Cigarette smoking and air pollution may interact to produce higher rates of puImonary disease than are seen with either factor alone. 10. Certain occupational exposures result in an increased inci- dence of COPD, but the relationship is not as strong as for ciga- rette smoking. The combination of certain occupational hazards and cigarette smoking has been observed, in many studies, to result in additive effects on morbidity from COPD. Exposures to cotton fiber, asbestos, and coal dust, in particular, appear to act in concert with cigarette smoking in the development of pul- monary disease. The role cigarette smoking plays in the develop- ment of coal workers' pneumoconiosis is unclear at present. ' In these studies, the degree of the relationship between smoking and im- paired pulmonary function was found to be dependent on the sensitivity of the particular pulmonary function test utilized to detect pnImonary obstruc- tion and/or small airways disease, the age, sex, occupation, place of residence, general state of health, and intensity of the smoking habit of the population examined. 76 11. A genetically determined protease-deficiency (alpha,- antitrypsin deficiency), inherited as an autosomal recessive trait, is found as a homozygous deficiency in approximately 1 in 3,600 people and as a heterozygous deficiency in approximately 5 to 8 percent of the population. Those with the homozygous deficiency have an increased prevalence of pulmonary emphysema. It is not clear whether cigarette smoking is an important contributor to the premature development of emphysema in people with the homozygous or heterozygous deficiency states. It is also Unknown whether nonsmoking heterozygotes are at a greater risk of developing emphysema than nonsmokers or smokers with normal alpha,-antitrypsin activity. 12. Data from most studies implicate cigarette smoking as an important factor in increasing the risk of developing post- operative pulmonary complications. 13. Some data suggest that cigarette smoking may increase the risk of development of spontaneous pneumothorax. 14. Data from pathologic and autopsy studies have demon- strated a dose-response effect of cigarette smoking on the severity of emphysema ; pipe and cigar smokers have degrees of emphy- sema intermediate between those of nonsmokers and cigarette smokers. 1.5. Goblet ceI1 density and distention, alveolar septal rupture, thickened bronchial epithelium, and mucous gland hypertrophy have been shown at autopsy to be more common in cigarette smokers than in nonsmokers. 16. Experimental data on humans have demonstrated that inhalation of cigarette smoke results in acute impairment of cer- tain parameters of pulmonary function. Overall pulmonary clear- ance, ciliary function, and aIveolar macrophage function have been found to be impaired in smokers as compared to nonsmok- ers. Some recent data suggest that acute heavy cigarette smoking with deep inhalation may result in increased pulmonary clear- ance. 17. In animal studies, in vivo and in vitro exposures to whole cigarette smoke (CWS) and several of its components have re- suited in impairment in overall pulmonary clearance, ciliary function, and alveolar macrophage function. 18. Experimental data on humans and animals presented in the past suggest that cigarette smoke may impair the function of the pulmonary surfactant system. Most of the studies reviewed in the last year confirmed the knowledge of the relationship between cigarette smoking and hronchopulmonary disease, A listing of these studies appears in a 77 separate section of the Supplemental Bibliography. A number of studies extended the knowledge of the association between ciga- rette smoking and bronchopulmonary disease, but several studies presented data which were either partially or wholly inconsistent with the known relationships ; these two types of studies are re- viewed below. EPIDEMIOLOGIC STUDIES Smoking and COPD There have been relatively few studies designed to evaluate the association between cigarette consumption and the prevalence of chronic obstructive pulmonary disease (COPD) in elderly popu- lations. In a random cross-sectional study of 487 men and women between the ages of 62 and 90, living in Edinburgh and registered with a practicing physician, Milne and Williamson (BP 45, 46) reported that over 73 percent of the women had never smoked compared with 7.9 percent of the men ; 62 percent of the men were current smokers (71 percent of whom inhaled), while only 18 percent of the women were current smokers (50 percent in- halers). In both men and women, a higher percentage of smokers had persistent cough and sputum production than nonsmokers (P <.OOl for men and P <.Ol for women), but twice the pro- portion of male smokers had these symptoms than women smok- ers. A dose-response relationship was demonstrated, since a higher percentage of heavy smokers had these symptoms than lighter smokers (P c.01). In men, 12.4 percent of the smokers had persistent cough, sputum, and a recent chest illness; none of the nonsmokers had this combination. For men, significant dif- ferences in histories of wheezing and dyspnea were found be- tween smokers and nonsmokers. For women, a significant differ- ence between smokers and nonsmokers was demonstrated only for wheezing (P <.05). The authors found that the FEV% (FEV/VC) was below 60 in 32 percent of the men who smoked compared to 6.7 percent of the nonsmokers (P <.05). For women, the figures were 9.4 percent and 3.9 percent. This differ- ence was not statistically significant. In a cross-sectional study of 300 men and women aged 65 and over in Glasgow, Scotland, Caird and Akhtar (BP 6) found that among women chronic bronchitis was reported by 11 percent of nonsmokers, 13 percent of light smokers, and 50 percent of heavy smokers. For men, a dose-response relationship was shown for light and heavy smokers, but the small numbers of non- 78 smokers ( 5 nonsmokers ; 2 with chronic bronchitis) limit the conclusions which can be drawn from the data. In a retrospective study of 5,433 men aged 40 and over who were current smokers, Rimington (BP 55) examined the relation- ship between the pattern of smoking and the prevalence of chronic bronchitis. He found that for each level of daily con- sumption of cigarettes, chronic bronchitis was more prevalent among those smokers who kept their cigarettes in their mouths during the entire period of smoking ("droopers") than among those smokers who removed their cigarettes from their mouths between puffs (normals) (table 1). For all levels of consumption, there was a significantly higher prevalence of chronic bronchitis among "droopers" than among normal smokers (P <.OOl). When these values were age-standardized (this was necessary because there was both a higher incidence of bronchitis and a higher percentage of droopers in men over 60 years of age), there was still a higher prevalence of chronic bronchitis among the "droopers" than among the normals, but the statistical signifi- cance of this difference was not presented, nor could it be calcu- lated from the data given. In an analysis of data from Bosnia and Hercegovina in Yugo- slavia, Zarkovic (BP 73) reported dose-response relationships be- tween depth of cigarette smoke inhalation and prevalence rates for chronic bronchitis, pulmonary emphysema, asthma, car pul- monale, and clinical and laboratory signs of obstructive lung disease. TABLE l.-Number, percentage, and age-standardized percentage of chronic bronchitics among S,4$8 cigarette smoking male volunteers for mass radiography, aged 40 and older, by amount and method of smoking Cigarettea per day 1-a IO-19 eo+ All D. N. D. N. D. N. D. N. Number of volunteers ______ 60 681 134 1,839 266 2,568 460 4,978 Number of chronic bronchitics ______ 22 160 56 552 113 971 191 1,673 Percentage chronic bronchitics _ _ _ _ _ _ 36.6 25.8 41.8 30.0 42.4 37.5 41.5' 33.6* Age-standardized percentape of chronic bronchitics _ _ _ _ _ _ 33.9 26.0 41.1 32.1 44.1 41.1 41.6 35.1 `P <.ool. D. = "drooDing" cigarette smokera. N. = normal cigarette smokers %NJRCE: Riminstos. J. (BP 55). 79 Olziihutag, et al. (BP 50) studied the prevalance of chronic bronchitis in Mongolia and found no association between ciga- rette smoking and chronic bronchitis in urban women, and a neg- ative association in rural women. These authors found close asso- ciations between chronic bronchitis and smoking in men. The authors pointed out that chronic bronchitis increased in fre- quency with age. Sherman, et al. (BP 58) conducted a study in Detroit on 489 working men and women, among whom 459 were employed in the auto industry. All subjects were referred to one physician for evaluation for workmen's compensation. The authors concluded that their data challenged "the traditional view (held) by . . . much of the medical profession that workers' lung and heart diseases are largely caused by cigarettes rather than by work- place poisons." These investigators studied various occupational exposures within the auto industry and found that both in ex- posed and unexposed working populations, approximately the same percentages of cigarette smokers and nonsmokers suffered from bronchitis, emphysema, and heart disease. Imprecise smok- ing histories and the absence of adjustment for several potentially confounding variables limit the conclusions which can be drawn from these data. The Eflects of Smoking on Pulmonury Function in Patients with COPD In a retrospective study of 41 hospitalized cigarette smokers with a diagnosis of pulmonary emphysema, Lepine and Myre (BP 37) found dose-response relationships between number of daily cigarettes smoked and years of dyspnea, years of cough, and impairment of the maximum expiratory flow rate (MEF). No dose-response relationships were found for the presence of car pulmonale by ECG, X-ray evidence of cardiomegaly, impair- ment of carbon monoxide diffusion, functional residual capacity, arterial blood gas abnormalities, or the ratio of residual volume t.o total lung capacity (RV/TLC) . In a retrospective analysis of pulmonary function tests (PFTs) of 140 patients with emphysema, chronic bronchitis, or both, Kass, et al. (BP 31) found no correlation between the severity of impairment of pulmonary function tests and the amount or duration of cigarette smoking. The Effects of Smoking on Pulmonary Function in Healthy Populations Grimes and Hanes (BP 24) studied 1,059 employees of a Iarge insurance company and found that cigarette smoking was associ- 80 ated with decreases in FVC and FEV, for all age groups in men. In women, the younger ex-smokers had higher values on pulmo- nary function testing than the nonsmokers. Higgins and Keller (BP 26), utilizing data obtained from the Tecumseh Study, did note differences in FVC, FEV,, FEVJVC, and MEF,,% between smokers and nonsmokers for both sexes and between smokers of greater than and less than 20 cigarettes per day. In this study, smokers of either sex had lower mean FVC, FEV,, FEVJFVC, MEF 508, average flow during the middle half of expiration (MMEF 25-i5s), average flow between 0.2 and 1.2 liters of expira- tion (MMEF 0.2.2L), and peak expiratory flow rate (PEF) than nonsmokers, and all these values decreased with increasing tobac- CO consumption. Krumholz and Hedrick (BP 35) studied pulmonary function in 91 cigarette smoking and 136 nonsmoking "healthy" male execu- tives, aged 35 to 64. They found significant impairment in the smokers for VC (P <.Ol), FEV, (P <.OOl), FEV% (P <.OOl), FEW 25-75~~ (P <.OOl), Raw (airway resistance) (P <.05), MVV (P <.05), RV/TLC (P <.05), CO diffusion (DXO) (P <.OOl), and D&O/TLC% (P <.OOl). Mean lung volumes were the same in the two groups except for RV/TLC. The methods of selection of patients for this study were not detailed. Brooks and Waller (BP 2), in a study of 2,703 people attending. a public health exhibition, found a nonsignificant difference in peak flow rates between smokers and nonsmokers age 45 and over; no differences were demonstrated for the younger than 45 age groups. The authors pointed out a number of biases which limit the conclusions which may be drawn from these data. Coleman, et al. (BP 11) investigated the maximal oxygen con- sumption (physical work capacity) of 78 members of the Texas Tech University faculty and found no difference in this value between smokers and nonsmokers. However, as the authors pointed out, the mean age of the smokers was seven years less than that of the nonsmokers, and the daily activity level of the smokers was also greater than that of the nonsmokers. The combination of these two effects may have partially accounted for the lack of difference in maximal physical work capacity be-.- tween the smokers and nonsmokers in this study population. In a cross-sectional study of men and women from the Western Highlands District and Trobriand Islands in New Guinea, Wool- cock, et al. (BP 70) found a greater decrease of FVC, FEV,, and PEF with age in men who smoked compared to nonsmokers (no P value reported). No such differences were found for women for FEV, and PEF. Woolcock, et al. (BP 69) also reported that in this same group 81 of New Guineans smoking was not strongly associated with cough on a single examination in the Western Highlands Dis- trict (WHD) population, but was strongly associated in the Trobriand Islands (TI) population. The authors stated, though, that the TI population smoked western cigarettes, whereas the WHD population smoked predominantly home-grown tobacco rolled in newspaper and smoked as cigars. The Roles of Smoking and Pollution in the Development of COPD Cigarette smoking is the predominant factor in the develop- ment of chronic nonspecific respiratory disease (CNRD), but there have been few prospective studies on the interaction be- tween air pollution and cigarette consumption as risk factors in the development of chronic nonspecific respiratory diseases. In an analysis of the initial data from a prospective study of Bos- ton policemen, Speizer and Ferris (BP 61) found that a higher percentage of men in three of four smoking categories who worked in areas of heavy traffic had chronic nonspecific respira- tory disease compared with men who worked in the outskirts of Boston (figure 1) .I In general, for each of the four traffic exposure *Criteria for diagnosis of CN&D were those established by the British Medical Research Council Bronchitis Committee (1966). Nonsmokan UL l-24 25+ Ex-smokers Current Number of Cigarettes Smoked Daily FKWBE l.-Prevalence of chronic nonspecific respiratory disease by cigarette smoking habits and trafIic exposure. SOURCE 82 Spelzer, F. E., Ferris, B. G., Jr. (BP 6ZJ. ,.stegories, the prevalence of CNRD was greater among ex- smokers than nonsmokers, and greater among current cigarette smokers than among either ex-smokers or nonsmokers (table 2). conversely, the prevalence of CNRD in current smokers appeared to be related to the number of years of traffic exposure ; those men with few years of such exposure had approximately the same incidence as those who worked in the outskirts. In the analysis of this reIatively homogenous group of men, it appears that "traffic pollution" and cigarette smoking may be acting in concert to increase the risk of deveIoping chronic respiratory disease. TABLE 2.-?%?vatence of chronii: nonspecijic respiratory disease grouped by current cigarette categories and trafic exposure -~ In TmUic %%E I-10 11-20 20+ Total No. Men 3-r Never smoked ___ 46' 11.6 11.1 26.0 16.7 Ex-smoker ______ 86 30.3 27.3 19.0 28.6 Current cigarette smoker _ _ _ _ _ _ _ _ 137 49.2 44.1 67.7 Total _ _ _ _ _ _ _ __ _ _ 268 36.1 36.2 39.2 o ????? ?????? utecor~ includes 12 men who I've smoked pipe mnd cigars. SOURCE: Spcieer. F. E. Feti. B. G.. Jr. (BP 81). 64.8 39.0 These authors also measured pulmonary function in this co- hort of policemen (BP 62) and found correlations between im- pairment of FEV, and lifetime cigarette smoking for all the men (P <.OOl). Statistically significant correlations between im- pairment of flow volume relationships at 50 and 25 percent lung volume and current cigarette consumption (P <.05 and <.OOl), and lifetime cigarette consumption (P <.Ol and <.OOl) were found for the outskirt station officers, but not for the. traffic officers, although the heavier smokers among them did demon- strate impairment of these parameters compared to the non- smokers and ex-smokers. The data also revealed that the heavier smokers with the longest exposure to traffic had the greatest impairment of flow-volume relationships at 50 percent (and 25 percent) vital capacity, again suggesting the synergistic action of air pollution and cigarette smoking in producing obstructive pulmonary disease. 33 The Relationship Between Cigarette Smoking and Small Airways Disease The role of small airways disease in the pathogenesis of COPD has come under close scrutiny in recent years. Results from several studies indicate that the resistance of airways less than 2 mm. internal diameter contributes little to the total measurable pulmonary resistance, and that considerable obstruction of these small airways may be present before changes in the total pul- monary resistance are recorded (BP 41) . Several techniques have been developed to detect the presence of small airways disease, but some of these are technically difficult, expensive, and imprac- tical for large-scale screening. The measurement of dynamic compliance was one of the first techniques used to demonstrate disease of the small airways (BP 71). Patients with small air- ways disease demonstrate frequency dependent decreases in dy- namic compliance compared to controls. More recently, the meas- urement of closing volume (CV) has been used as a technically easier and less expensive method for the assessment of small air- ways function. The theoretical basis of these methods in the as- sessment of small airways disease is described in many recent publications (BP 3, 4, 5, IO, 20, 23, 25, 28, 34, 36, 40, 41, 42, 43, 63, 66, 71). It is currently unclear whether those subjects with evidence of small airways disease are particularly susceptible to the development of clinically identifiable forms of COPD. McCarthy, et al. (BP 40) measured closing volumes in 112 subjects by the single-breath argon gas bolus method. Closing volume increased in a linear fashion with respect to age. Of the 66 nonsmokers, no subjects had closing volumes greater or less than 2 SDS from the mean normal values, whereas 26 of 39 cigarette smokers (7 smokers yere excluded because of grossly abnormal ventilation distribution as measured by the argon tech. nique) had closing volumes greater than 2 SDS above the mean (figure 2). This difference in closing volume was highly signifi- cant (P <.OOl) and indicated a higher prevalence of small air- ways disease in the group of smokers. Of 14 smokers with abnormalities of standard pulmonary function tests, 13 were symptomatic and all but one had abnormal closing volumes. Of note was that of 17 asymptomatic smokers, 9 had abnormally high closing volumes. Although none of the smokers had sought medical attention, 29 of the 46 smokers had chronic bronchitis, and had, on the whole, higher closing volumes than the asymp- tomatic smokers. In a separate pubhcation, McCarthy and Craig (BP 39) re- ported that 15 percent of a group of 91 asymptomatic female smokers in Manitoba had abnormally high closing volumes (CV), 34 60 I I I :- Smokers Seated . Symptomatic 0 Asymptomatic 10 50 30 40 Age (Yea6 60 70 FIGURE 2.-Relationship between "closing rolume" and age in 39 smokers. (Thirty-two smokers with normal conventional lung function data and seven smokers whose data are identified by numbers.) Average relationships 22 SD between "closing volume" and age in sixty-six nonsmokers are also shown. Solid circles indicate smokers who, according to the questionnaire used, had simple chronic bronchitis, and open circles indicate smokers who were as.vmp- tomatic. Note that in nine asymptomatic smokers the "closing volume" was above the normal limits. SOURCE : i\icCartbp, D. S., et al. (BP JO). in contrast to the `72 percent of 46 male smokers in London (BP 48) who had abnormally high closing volumes. None of the female nonsmokers had any CV abnormalities. The authors sug- gested that differences in pollution exposure of the London and Manitoba study groups might, in part, account for the differ- ences in prevalence of the CV abnormalities. In a study of pulmonary function of subjects voluntarily re- porting to an emphysema screening center, Buist ,et al. (BP 5) reported that 6 percent of the nonsmokers, 35 percent of the current cigarette smokers, and 23 percent of the ex-smokers had abnormal CV/VC ratios. In each decade from age 20 to 79, more smokers and ex-smokers had abnormal CV/VC ratios than non- smokers (figure 3). The daily consumption of cigarettes was re- lated to CV abnormalities in a dose-response relationship for men (figure 4). Among the women, those with a daily consump- tion of less than 10 cigarettes per day had significantly lower 86 n Nonsmqkem (284) Smokers (524) 0 n=7 7 3 43 61 20 30 91 26 32 126 51 63 143 60 66 65 54 32 11 21 11 3 <20 20-29 30-39 40-49 !lo-59 60-69 70-79 =a0 Ace (yearn) F'IOUBE 3.-Prevalence of abnormal closing volume/vital capacity ratios in non- smokers, cigarette smokers, and ex-smokers by age decades. SOUBCl : Buirt, A. S., et at (BP 6). CV/VC ratios than those smoking more than this amount (P <.05) ; but overall, no dose-response relationship was demon- strated (figure 5). 100 SO 60 E 0 2 2 40 20 0 < 10 Cigarettas/Day n=21 10-20 Cigamttar/Day n=136 2-O Cigarettes/Day n= 175 >40 Cigarettes/Day n=19 Symptomatic 26.3 FIQIJBE 4.-Comparison of the prevalence of respiratory symptoms and pulmonary function abnormalities in male smokers according to their daily cigarette consumption. cc -Closing capacity TLC -Total lung capacity cv --Closing volume vc -Vital cepacity FEvl --One-second forced expiratory volume SOUBCE : Buist, A. 9.. et al. (BP 5). The Interactions Between Cigarette Smoking and the Genetic Susceptibility to the Development of COPD Mittman, et al. (BP 47, 47, 49) reported on the interaction between cigarette smoking and the genetic susceptibility to de- velopment of chronic obstructive pulmonary disease (the alpha,- antitrypsin deficiency state). These authors described the 8'7 100 a0 E 8 60 k P 40 20 0 100 m < 10 Cigarettes/Day n=19 IO-20 Cigarettes/Day n=75 20-40 Cigarettes/Day "~77 > 40 Cigarettes/Day n=3 FIMJRE Zi.~omParison of the prevalence of respiratory symptoms and p& monarY function abnormalities in female smokers according to their daily cigarette consumption. CC -Closing capacity TLC -Total lung capacity cv -Closing volume vc -Vital capacity FE& --One-second forced expiretory volume SOURCE : Buid, A. S.. et aL (BP 5). polymorphic (multiple gene) system of protease inhibition (Pi) by alphal-antitrypsin (AAT), and listed some of the partial and severe deficiency states of this enzyme system. In a series of 170 consecutive patients with a diagnosis of COPD admitted to the City of Hope Medical Center who had no previously known his- tory of AAT deficiency, 40 patients (24 percent) demonstrated some type of AAT deficiency. This was a significantly higher 88 20 10 2 0 2 8 I3 E 1 >4Q : e E 30 t E s ; 20 L 10 0 I 0 I I I I I I I I 15 30 45 50 75 90 105 120 Pack Years percentage than was found in a control group of the Norwegian population, which is known to have a high incidence of this enzyme deficiency (P <.OOl). The lifetime cigarette consump- tion of the population of patients with emphysema who had an intermediate degree of AAT deficiency was significantly less than those emphysema patients with a normal phenotype (PiMM) (P <.05) (figure 6), suggesting a possible interaction between smoking and the genetic abnormality. The data imply that a greater degree of exposure to tobacco was required to produce FIOIJEE B.-The distribution of smoking histories in men with bronchitis and/or emphysema. Patients grouped by phenotype; Pi a patients above, those with intermediate MT deficiency below. Each bar depicts the fraction of patients reporting smoking histories in the ranges shown. SOURCE : Mlttman, C., et aL (BP 48). 89 emphysema in those patients who did not have a genetic predis. position than in those with the genetic defect. The authors con- cluded that any degree of AAT deficiency makes an individual more susceptible to the effects of smoking. The same authors have also examined 144 people with partial AAT deficiencies who were apparently healthy and compared them with 100 con- trols matched for age, sex, and smoking history (BP 48). They found that 25 of the 62 smokers with partial AAT deficiency (40 percent) had abnormalities of pulmonary function tests sug- gestive of obstruction, while 7 of 47 smokers in the control group (15 percent) demonstrated such abnormalities. This differ- ence was statistically significant (P c.05). Hutchison, et al. (BP 27) studied 28 patients with pulmonary emphysema, 8 of whom were homozygous deficient for alpha,- antitrypsin. Although the annual consumption of tobacco up to the age of onset of dyspnea was equal in the deficient and non- deficient group of patients, total lifetime tobacco consumption was significantly Iess among the AAT deficient patients than among the nondeficients (P c.01). AI1 8 AAT deficient patients were smokers. Although there was no significant difference in the incidence or age of onset of chronic bronchitis between the two groups, the AAT deficient group of patients developed exertional dyspnea 12 years earlier than the nondeficients (P <.OOl). These data suggest a synergistic effect of cigarette smoking on the development of pulmonary emphysema in those patients with homozygous deficiency of alpha,-antitrypsin. Colley, et al. (BP 12) analyzed a cohort of 3,899 persons born in the last week of March 1946 in England, Scotland, and Wales and found that irrespective of a history of lower respiratory tract illness before the age of two, the smokers had a greater prevalence of symptoms of winter cough at age 20 than the nonsmokers (table 3). The authors argued that cigarette smoking, by age 20, is a far more important factor in the development of respiratory disease than is a history of lower respiratory tract illness. The results of this study are suggestive evidence against the hypothesis of a purely constitutional susceptibility to the development of respiratory diseases independent of tobacco ex- posure. The Effect of Smoking on the Development of BuUous Disease 04 the Lungs Stoloff and Victor (BP 64) reviewed 44,887 outpatient photo- fluorograms seen in the Philadelphia Central Mass X-ray~Unit from 1969 to 1970, and found 59 men and one woman with bul- lous disease of the lung. Smoking information was available on 90 TABLE 3.-Prevalence (percent) of cough o2q.1 or night in both sexes in winter by air pollution index, social class, cigarette making, and history of chest illness under two years of age.* (Figures in parentheses are population.) Air poktion index HistorY Of 7-17 18-28 cigarette Chest illness under SocisI dell9 Social claes smoking 2 Years of age 1+2 a-l-4 1+2 a-t4 Never No chest illness _ 4.7 (344) 6.7 (369) 4.7 (277) 6.6 (212) smoked One or more chest illnesses _ _ _ _ _ _ 12.3 (67) 8.3 (108) 8.3 (84) 10.8 (102) Present No chest illness _ 11.2 (214) 12.6 (326) 14.1 (192) 16.7 (261) smoker One or more chest illnesses _____. 16.4 (66) 11.8 (102) 12.3 (73) 22.2 (144) *Excluding 980 pzmone-that ie. examokem and thwe whose history of cigarette smoking. social class, air pollution index, cheet illneaa under 2 yearn of age. and history of cough day or nipht not known. SOURCE: Cdey, J. IL T.. et IL (BP 12). 51 of the men. There were no nonsmokers among the 51 cases (P <.OOl). In nonwhite and white men under age 45 and in nonwhites greater than 45 years old, the rates of this disease increased for each progressively higher level of daily cigarette consumption (table 4). When men without known possible or probable occupational hazards were studied, dose-response rela- tionships were again demonstrated in the nonwhite population, inclusive of all age groups (table 5). The absence of dose- response relations in whites older than 45 may be at least par- tially explained by the small numbers of cases of bullous disease of the lung found in whites (19 of the 51 cases). The authors stated that the data "are consistent with the hypothesis that cigarette smoke is capable of causing alveolar septal rupture . . ." and, hence, bullous disease of the lung. TABLE 4.-Estimated rates of bullous disease of the lung per 1,000 men by age, race, and cigarette smoking habits. yg2 $x Nonsmoker <1 Pack/Day 1 Pack+/Day WM 26-44 0 0 1.0 ____________________-- 46f _______________________ 0 4.6 2.9 NWM 26$ 9.9 __________________----- 0 4.3 46+ ___________________-___ 0 4.1 13.0 SOURCE: StolofY. I. IA, Victor. S. B. (BP 64). 91 TABLE 5.-Estimated rates of bullous disease of the lung per 1,000 men with no demonstrable occupational huzccrd (c.luss 1). Nonsmoker