TABLE I.-Continued. Study Population Data collection method Measure of atherosclerosis Results Distribution by percentage of degree of atherosclerosis by smoking habits standardized for age (macroscopic study) Current cigarette smokers Degree of atherosclerosis NWW EX- smoked < pack l-2 packs 2+ packs Cigar/ cigarette regularly per day per day per day pipe smoker Lif&.ic (371 894 autopsies of males 20-79 at death in Yalta Interview with relatives Visual grading None or minimal 59.8 45.2 36.6 32.6 36.3 50.9 Slight 24.7 26.9 27.9 21.5 28.4 25.5 Moderate 10.2 16.2 16.0 16.5 21.0 12.6 Advanced 5.3 ii.7 lY.5 23.4 14.3 11.0 Total 100.0 loo.0 loo.0 1OQ.O 1Oil.o loo.0 Ratio of the extent of atherosclerotic lesions in the average coronary artery between nonsmokers and smokers Total Compli- athero- Fatty Fibrous cated Calcified Raised sclerosis streak plaque lesion lesion lesion Nonsmoker to heavy smoker Nonsmoker to smoker 1.0 1.1 1.0 1.5 0.6 1.0 1.0 1.1 1.1 1.0 0.6 1.0 TABLE l.-Continued. Data collection Measure of Study Population method atherosclerosis Results Schettler Autopsies of 89 Interview with Visual grading Stenosis et al. males aged 60-94 relatives (63) at death in Tokyo Smoking No Yes Total No 6 8 14 Yes, daily 8 67 75 Total 14 75 89 Rhoads et al. 109 sutop&s of Interview with AHA panel Mean coronary atherosclerosis grade versus selected attributes (56) Japanese American subject male8 born 1900- Regression coefficients 1919 who parti- cipated in Honolulu Examination variables Simple Multiple ' heart study Relative weight 1%) 0.031 J 0.025O Cigarettes/day 0.022* 0.024 3 Cholesterol(mg/dl) 0.011 3 0.0093 Triglycerides(mg/dlI o.Oo22 NS' Glucose(mg/dl) 0.004 2 NS' Hematirit 1%~ o.069z NS ' `Multiple regreaxon was done by B stepw.ee ehmmatmn procedure begmnmg wth the set of vanables shown: coefficwnts are for the final step. Multiple correlation (final step1~0.46 IN- 108l ~Slgnlficant at on5 level ~S1gmficant at 0 01 level `NS. vanable included ,n timt step. deleted an not slgmficant at 0.05 level. TABLE l.-Continued. Study Population Data collection method Measure of atherosclerosis ReSUlti 35-44 X 4.4 a.4 x 32.2 32.1 X 60.3 62.1 2 8 12 45-54 X 5.5 0.4 x 25.1 31.4 x 65.7 70.4 2 25 12 55-64 a.1 47 7.2 31.4 31.1 20.8 60.3 57.4 62.4 11 15 9 ' ATL as percent surface fatty streaks (F) plus raised lesions (Rl: FaF=F - (100 - R%E in percentage units explained in the text; X indicates subgroups having fewer than five members Holme et al. (29) 129 autopsies from Interview with Visual grading Correlation coefficient between number of cigarettes and raised lesions in the coronary 16,2CGC males aged subject arteries = ,039, not significant. 4M9 in Oslo prospective CHD stwiv Sternby (711 60 autopaiea of 703 Interview with Visual grading Smoking and stenosis or atherosclerosis in the left anterior descending coronary artery males in CHD subject study in Malmo. LAD Coronary SW&n Smoking category Number rased lesions artery stenosis Non 3 68 33 Ex 8 52 38 Light 18 45 39 Heavy 17 54 59 TABLE l.-Continued. Study Population Data collection method Measure of atherosclerosis Results Sorlie et al. ( 70) 139 autopsies of Interview with Visual grading Association of atherosclerosis in coronary arteries with antemortem characteristics: 9,824 Puerto Rican sub@ simple correlation coefficients (Puerto Rico heart health program) males aged 35-79 m a prospective study Correlation coefficients Characteristics measured at exam 1 Total (139) Rural (36) Urban ( 103) Systolic blood pressure Diastolic blood pressure Serum cholesterol Age, exam 1 Relative weight Physical activity Blood glucose Hematocrit Education level Income Cigarettes smoked Calories (24.hour recall) Starch (24.hour recall) Alcohol (24.hour recall) Total fata (24.hour recall) Triglycerides (fasting) Ventricular rate Vital capacity 0.22 0.07 0.30 0.26 0.09 0.30 0.42 0.59 0.38 0.01 0.32 4.08 0.21 JJ.15 0.25 -0.18 0.06 -0.22 0.20 4.04 0.21 0.14 0.36 0.12 0.14 a.40 0.24 0.16 4 17 0.16 4.16 xl.05 xl.22 a.14 a.43 407 -0.17 -0.29 -0.09 a.10 -9.10 a.13 -0.04 xl.53 0.03 0.23 0.49 0.19 0.13 0.20 0.08 -0.19 a.13 4.16 lesions were not identical, the findings from both of these large studies of autopsied men in the United States were remarkably similar. Both studies reported more extensive coronary atherosclero- sis among the cigarette smokers than among the nonsmokers, and for the major comparisons, with only rare exceptions, there was an orderly progression of least extensive lesions in nonsmokers, inter- mediate extent of lesions in light or moderate smokers, and most extensive lesions in heavy smokers. In the New Orleans study (741, lesions were measured not only by visual evaluation, but also by optical electronic scanning of radio- graphic images of the flattened arteries. The measurements of lesions from radiographs-relative mean coronary wall thickness and percentage of the coronary artery intima involved with calcifica- tion-were consistently greater for the heavy smokers than for the nonsmokers. A variety of statistical analyses on smoking measures and atherosclerotic lesions were performed to determine the signifi- cance of the various differences and trends. These analyses con- firmed that the major differences between the heavy smokers and the nonsmokers in extent of raised atherosclerotic lesions (the sum of fibrous plaques, complicated lesions, and calcified lesions) were significant. A one-way multivariate analysis of nine atherosclerotic variables clearly indicated that there were statistically significant differences among the three categories of smokers (heavy, light to moderate, and nonsmokers) for mean coronary wall thickness, raised lesions in the coronary arteries, percentage of cases positive for fibrous plaques, percentage of cases positive for complicated lesions, and percentage of cases positive for calcified lesions, with lower values in nonsmokers and higher values in the heavy smokers. Pate1 et al. (53) evaluated this same data on smoking and atherosclerotic lesions to examine further the interrelationships with measures of obesity. The confounding effects of diseases such as hypertension and diabetes mellitus were controlled by excluding such cases from the analysis. The confounding effects of age and measures of smoking habits on the association between atherosclero- sis and obesity were controlled by multivariate regression analysis. This analysis disclosed an inverse relationship between smoking habits and obesity. There was also a weak positive association- when age and smoking were controlled for-between measures of obesity and mean coronary wall thickness and raised lesions in the coronary arteries among whites, but not among blacks. In the black men, again with age and smoking controlled in the analysis, a weak association between fatty streaks in the coronary arteries and obesity was found. This analysis confirmed the previously reported relationships between smoking habits and atherosclerosis, as mea- sured by mean coronary wall thickness, coronary calcifications, and raised lesions in the coronary arteries. 31 Since their first report in X65, Auerbach and his associates have investigated the relationship of cigarette smoking to microscopic findings in the coronary arteries (4). This study indicated that lesions were most extensive in cigarette smokers and confirmed earlier studies by Auerbach et al. (6) and Strong and Richards (74). The microscopic portion of the Auerbach et al. study (4) showed that fibrous thickening, atheroma, and calcifications of the coronary arteries all increased with increasing number of cigarettes smoked per day. They also found that the fibrous thickening of arteries increased in relation to the number of cigarettes smoked per day as the size of the artery decreased; i.e., it was least in the coronary arteries and greatest in the myocardial arteries. Lifsic (37) reported on the relationship of cigarette smoking to coronary atherosclerotic lesions based on the Yalta sample from the World Health Organization (87) autopsy study of five cities in Europe. Information on cigarette smoking was obtained by means of interviews with the subjects' near relatives. The prevalence and extent of atherosclerotic lesions were evaluated in autopsies of 865 men, aged 20 to 79 years, out of the 1,220 deaths occurring in Yalta residents of this age and sex group during the period of study. There was a positive association of smoking with the extent of coronary calcification; however, the author explained this association as being related to coexisting alcohol consumption and stated that smoking alone tended to be negatively associated with coronary calcification. The following paragraph from Lifiic's discussion provides additional, information from this report. There was little significant difference between smokers and nonsmok- ers in the prevalence and extent of atherosclerotic lesions in the coronary arteries. Thus, of the total of 210 comparisons of different indices of the prevalence and extent of atherosclerotic lesions between subgroups X and W, significant differences positively correlated with smoking were found in only 20. The tendency toward a positive correlation of coronary atherosclerosis with smoking was found mainly in subjects up to the age of 50, but after 60 the opposite tendency prevailed. These age peculiarities agreed with data from other studies showing that differences in the degree of atherosclerosis between smokers and nonsmokers. . . are more distinct below age 60. The author also mentioned a positive association between smoking and coronary calcification in "strenuous workers." A note added in proof to LifGic's article states, "Additional study of this material by individually matched case-control analyses revealed a marked trend toward a positive association between smoking and atherosclerotic raised lesions in the coronary arteries" (82). Thus, while the author's abstract does not indicate an important relationship of smoking and coronary atherosclerosis, there are findings in the study that do 32 indicate significant relationships between smoking and coronary atherosclerosis, especially in the younger subjects. A subsequent study by Vikhert et al. (83) on material from five cities in the U.S.S.R. evaluated the effect of nutritional status and tobacco smoking on atherosclerotic changes in the coronary arteries as measured by a visual planimetric method. This material was also utilized for a WHO-sponsored epidemiological study of atherosclero- sis (87). The vessels examined were from 430 men 40 to 69 years of age. The major analyses concerning tobacco smoking were made from 313 male heavy smokers and 82 nonsmokers. The investigators studied both manual workers and white-collar workers and found that tobacco smoking in combination with overnourishment had a much more positive effect on the development of coronary athero- sclerosis in white-collar workers than in the manual workers. Prospective epidemiologic studies of cardiovascular disease with autopsy followup provide additional information concerning the relationship of smoking to atherosclerotic lesions in the artery wall. The epidemiological studies in Oslo, Puerto Rico, and Honolulu are characterized by careful documentation of selected major risk factors, including cigarette smoking habits during life, and by standardized evaluation of atherosclerotic lesions at autopsy (29, 56, 70). Each of these three studies reported findings on the relationship of CHD risk factors to atherosclerotic lesions in more than 100 autopsies of deceased men who had been part of larger cohorts that had been examined and followed during life. In addition, a smaller study from Malmo, Sweden, had some of the same features as these larger studies (71). The Oslo, Malmo, and Puerto Rico studies used identical methods for grading the extent of atherosclerotic lesions. These prospective epidemiologic studies with autopsy followup are in general agreement concerning the relationship of serum cholesterol levels and blood pressure to the extent of atherosclerotic lesions in the coronary vasculature. The findings concerning the relationship of cigarette smoking to the extent of coronary atherosclerosis are not uniform. The Honolulu study (56) showed a significant relationship between smoking habits and extent of coronary atherosclerosis. The Oslo study (29) did not show a significant relationship between cigarette smoking and coronary atherosclerosis. The Puerto Rico study (70) also did not show a significant relationship between smoking and the extent of coronary athersclerosis. A somewhat similar study from Japan by Hatano and Matsuzaki (26) indicated a significant relationship between cigarette smoking and coronary artery stenosis. Thus, there is some inconsistency concerning the association between cigarette smoking habits and coronary athero- sclerosis in the prospective epidemiologic studies with autopsy followup. 33 In considering this entire body of evidence, however, the prepon- derance of evidence suggests that cigarette smoking has an effect on the development of atherosclerotic lesions in the coronary artery wall in the U.S. population, and that its effect is not limited to those events immediately surrounding the occlusive episode. Small Arteries in the Myocardium Table 2 reviews those studies that. have examined the relationship between cigarette smoking and lesions of the arterioles within the myocardium. Auerbach et al. (7) found a relationship between smoking habits and thickening of the walls of the arterioles and small arteries of the myocardium. Auerbach et al. (4) also performed a microscopic study of coronary artery lesions in autopsied men in relation to previous smoking histories. In the microscopic portion of this study, fibrous thickening, atheroma, and calcification increased with an increased number of cigarettes smoked per day. Moderate to advanced hyaline thickening of the arterioles in the myocardium was strongly related to smoking. It was found in 98.6 percent of the autopsied subjects with a two pack per day smoking habit and not found in the group of subjects who never smoked regularly. Naeye and Truong (51) reported essentially similar alterations in the intramyocardial arteries, which developed more rapidly in cigarette smokers than in nonsmokers. The Aorta Those studies that provide autopsy and other evidence for the relationship between cigarette smoking and atherosclerosis of the aorta are summarized in Table 3. Wilens and Plair (85) found significantly more severe sclerosis of the aorta in cigarette smokers than in nonsmokers. Sackett and Winkelstein (61) reported that elderly cigarette smokers had signifi- cantly higher rates of aortic calcification, detected on chest X-ray, than did nonsmokers. Sackett et al (601, in an autopsy study, found a significant relationship between the use of cigarettes and the severity of aortic atherosclerosis. An interim report by Strong et al. (75) concluded that atherosclerotic involvement of aortas was greatest in heavy smokers and least in nonsmokers among autopsied men in New Orleans. Most of these studies, reviewed in the 1971 Report of the Surgeon General The Health Consequences of Smoking (801, indicate that differences between heavy cigarette smokers and nonsmokers are particularly great in young individuals, and that heavy smokers have increased surface involvement with fibrous plaques or more advanced atherosclerotic lesions. Since the 1971 review, a study of smoking and atherosclerosis in deceased men in New Orleans has been completed. Several reports 34 TABLE 2.-Autopsy studies of atherosclerosis involving small arteries in the myocardium Study Population Measure of atherosclerosis ReSUlta Auerbach et al (7, 1,164 males Records and Biopsy of Grade of thickness of walls of arterioles' autopsied family myocardium at VA Number of men Percentage of men Grade Grade Grade Grade Grade Grade Age Smoking Total 0 1 2. 3 Total 0 1 2. 3 - (45 None Cigar-pipe ctte3 1-19 Ctte 2&39 ctte 40+ 45-59 None Cigar-pipe ctte l-19 ctte 2cL39 Ctte 40+ 60-69 None Cigar-pipe ctte 1-19 ctte 2&39 ctte 40+ 70+ None Cigar-pipe ctte 1-19 ctte 20-39 ctte 40+ 22 2 4 - 50 1 a5 4 29 - 15 1 13 - 33 - 99 - 50 56 4 35 92 193 67 - 32 2 40 30 - 46 - 9 - 19 1 loo.0 1 3 100.0 31 18 lc0.0 35 46 100.0 10 19 loo.0 12 2 loo.0 8 5 lc0.0 17 16 loo.0 35 64 loo.0 11 39 1Oc.o 36 16 loo.0 22 13 loo.0 44 4.3 100.0 58 135 1WJ.O 21 66 loo.0 18 12 100.0 19 21 loo.0 12 18 1lx.o 12 34 100.0 3 6 loo.0 9.1 66.4 4.5 25.0* 75.0' 2.0 62.0 36.0 4.7 41.2 54.1 - 34.5 65.5 6.7 80.0 13.3 61.5 38.5 - 51.5 48.5 - 35.4 64.6 22.0 78.0 7.1 64.3 26.6 - 629 37.1 47.0 52.2 - 30.1 69.9 - 24.1 75.9 6.3 56.2 37.5 47.5 52.5 - 40.0 60.0 - 26.1 73.9 - 33.3' 66.71 `In the right ventrmdar wall of 1,020 men by age and smoking habits `Percentage8 baeed on less than 10 cases `Ctte indicates cigarettes. TABLE 2.-Continued. Study Population Smoking data BO"TlX Measure of atherosclerosis ResUlts Auerbach et al. (4 1,056 males autopsied at VA Relatives and records Microscopic examination Distribution by percentage of degree of fibrous thickening of myocardial arteries, subepicardial arteries. and hyaline thickening of myocardial arterioles (microscopic myocardial study), by smoking habit standardized by age Current cigarette smokers Demee of findings Never smoked Ex ~1 pack 1-2 packs z+packB Cigar/ cigarette regularly per day per day per day pipe smokers - Myocardial arteries None or minimal slight Moderate Advanced Total S&epicardial arteries None or minimal Slight Moderate Advanced Total 97.3 24.1 2.7 62.2 +16 12.3 - 1.4 100.0 loo.0 74.7 17.5 2.4 1.4 24.9 56.8 35.1 32.7 0.4 19.0 28.8 23.8 - 6.7 33.7 42.1 loo.0 100.0 100.0 100.0 2.9 1.1 22.0 32.0 37.1 29.6 70.6 63.2 39.0 45.0 6.7 4.2 21.0 24.3 0.7 0.6 loo.0 100.0 100.0 1rKl.o 21.5 26.2 64.8 60.5 9.9 11.6 3.8 1.7 loo.0 loo.0 Myocardial arterioles None or minimal Slight Moderate Advanced Total 92.0 2.1 - - - 3.2 8.0 28.7 2.2 1.4 39.6 40.8 - 20.8 9.6 7.9 19.6 19.1 - 40.4 68.2 90.7 40.8 36.9 100.0 100.0 loo.0 loo.0 loo.0 100.0 TABLE 3.-Autopsy studies of atherosclerosis involving the aorta Study Smoking data Measure of Population source atherosclerosis Results Wilens and Plair WI 989 consecutive necropsea at NY VA hospital Patient chart Visual grading Percent of subjects by smoking status and atherosclerosis Severity of Non- Pipe/ sclerosis smoker Heavy Moderate Light cigar Other Number 161 199 288 152 70 119 Percent above average 9.9 25.1 26.4 19.1 10 10.9 Percent average 60.2 61.3 62.5 63.2 60 63.0 Percent below average 29.6 13.6 11.1 17.6 30 26.1 Strong and Richards (74) 1,320 Interview Visual grading Mean percent of intimal surface of abdominal aorta involved with raised lesions autopsies with relatives of males Average number of cigarettes smoked per day aged 25-64 at death Age and race 0 l-24 25+ White males 25-34 35.44 4554 5.%64 Black males 2534 3M4 4&54 55J34 1 7 7 14 33 44 33 52 56 46 63 71 4 7 9 6 20 28 14 37 45 26 51 56 g TABLE $-Continued. Study Smoking data Measure of Population source atherosclerosis Results Sackett and 590 Winkelstein white male (61) admissions to Powell Park Mem- orial In- stitute in 1955 Patient Chest X-ray The relationship between smoking and calcification of the thoracic aorta questionnaire for calcification in thoracic aorta Nonsmokers Smokers Percent with Percent with Probability Age group Number calcification Number calcification value 1 Totals 50-59 61 13 131 11 0.4 192 60-69 90 16 124 26 0.2 214 70 and over 116 37 63 54 0.02 164 Totals 267 25 323 26 - 590 Age-adjusted percent - 22 - 30 - - `Chi-square of independence, twotcded. The relationship between amount smoked and calcification of the thoracic aorta Age group m-59 60-69 70 and over Totals Nonsmokers Light smokers Heavy smokers Percent with Percent with Percent with Number calcification Number calcification Number calcification Totals 61 13 104 9 27 22 192 90 16 107 24 17 35 214 116 37 63 56 5 40 164 267 26 274 29 49 27 590 Age-adjusted percent - 22 - 29 - 32 - TABLE S.-Continued. Study Smoking data Measure of Population Bource atherosclerosis Results Sackett et al. (60) 1,019 Standardized Visual grading Mean ag?+adjusted atherosclerosis ridits versus gr aded use of cigarettes and alcohol consecutive interview with on a numerical autopsies of patient on scale Alcohol Cigarettes white admiwion patienta &/day None 1 :! pack 1'2 pack+ None ,351 ,466 ,498 0.5-1.5 ,424 ,570 .56l3 1.6+ ,426 ,526 ,589 Strong et al. Autopies Interview Visual grading Mean percentage of intimal surface of aorta involved with raised lesions by age, race, and average rate (75) of 741 with relativea and optical of cigarette smoking in the last 10 years of life males 20-64 Bcanning year8 at death Cigarettes pe r day Age and race White males 3-4 4654 55-64 Black males 35-44 45-54 55-64 0 l-24 25+ 16 35 49 29 52 54 48 66 70 3 22 24 12 38 50 21 50 49 $ TABLE 3 -4`ontinued. Study Smoking data Measure of Population source atherosclerosts Results Litsrc (37) 865 autopsies Relatives and records Visual gradmg Prevalence of atherosclerotic lesions in the abdominal aorta in different subgroups (percentage) of males aged 2&79 at death in Yalta Smokrng group Never Lrght Heavy Fatty streak 96.5 92.8 90.7 Fibrous plaque 96.0 96.5 97.5 Complicated lesion 43.0 53.4 60.9 Calcified lesion 25.0 42.3 57.3 Extent of atherosclerotic lesions (percentage of surface) in the abdominal aorta Smoking Fatty group streak Never 7.0 Light 6.1 Heavy 5.8 Fibrous plaque 28.1 31.6 26.5 Complicated lesion 2.0 5.1 4.1 Calcified lesion 1.2 2.2 3.4 TABLE 3.-Continued. Study Smoking data Measure of Population source atherosclerosis Results Rhoads et al. (56-l 124 Japanese American males autop sied as part of the Honolulu heart study Interview with subject Visual by AHA panel method Correlation coefficients among selected autopsy and examination variables' Aorta (N = 124). atherosclerosis grade Age at death Examination variables Height (cm) Relative wt. (%) Cigarettes/day Cholesterol (mg/dl) Triglycerides (mg/dl) Uric acid (mg/dl) Glucose (mg/dl) Hematocrit (%I Vital capacity (liters) Alcohol (pm, Systolic pressure (mm Hg) Diastolic pressure (mm Hgl Mean coronary grade Aorta grade 0.30 3 -0.12 a.10 0.14 0.243 0.14 -0.052 0.15 a.03 X1.23~ -o.08z 0.29 3 0.05 0.503 (96)' I N= number of specimens. sS~gniiicant at 0.05 level a Significant at 0.01 level. `When a correlation coeffXent is based on less than 95 percent of the ~pecmwns available (because of missing data), the number of observations is indicated in parentheses. There were 96 autopsies with both aorta and coronary vesel grades available. 13 wth coronary only. and 28 with aorta only TABLE 3.-Continued. Study Smoking data Measure of Population SO"FX? atherosclerosis ReSUlta Auerbach 1,412 males Family Visual grading Percentage of selected findings by smoking habits and Garfinkel autopsied (5) at VA hos- Percentage of cases' pita1 "Current" cigarette smokers Findings Never smoked 1 pack l-2 pack3 2+ packs Cigar Excigarette regularlv per day per day per day or pipe smoker Thoracic aorta Many or diffuse distribution of plaques Moderate or advanced ulceration Moderate or advanced calcification Thrombus present Abdominal aorta Many or diffuse distribution of plaques Moderate or advanced ulceration Moderate or advanced calcification Thrombus present 16 26 41 37 27 29 4 6 14 12 10 8 56 63 74 74 53 67 4 7 14 11 11 9 28 54 68 79 46 53 7 19 27 27 13 22 63 76 84 88 74 81 7 23 23 31 14 23 lPercentages are adjusted to distribution by age group of all men in study. TABLE 3.-Continued. Study Smoking data Measure of Population source atherosclerosis Results Sternby (71) 60 Interview Visual grading Smoking and atherosclerosis of the aorta autopsies with subject from 703 Raised lesions in the males enrol- Smoking category Number abdominal aorta led in a CHD study Non 3 26 in Malmo. ET. 8 43 Sweden Light 18 53 Heaw 7 83 Smoking and atherosclerosis m peripheral arteries Femoral artery Lower leg artery Smoking Iliac artery Raised Sterosis RalSed Stenosis category N Raised lesions lesions (701 lesions (70) Non 3 17 20 0 2 0 EX 8 18 43 33 18 22 Light 18 29 29 6 3 11 Heavy 17 50 50 35 12 41 g TABLE S.-Continued. Study Population Smoking data Measure of source atherosclerosis Results Tracy et al. Autopsies Interview Visual exam Means of observed minus expected raised-among-lesions GE), fatty streaks among flat surfaces (FaF), (77) of 1.380 with relatives all types of lesions tATL1, and number of cases (N) by age, race, and cause of death according to white and smoking category I, abdominal aorta black males aged 2564 O-E Faf ATL at death Age 0 l-24 25+ 0 1-24 25+ 0 l-24 25+ White basal 2534 3544 45-54 5.5-64 White CHD 25-34 35-44 45-54 55-64 Black basal 2634 3%44 4554 55-64 Black CHD 24-34 35-44 45.54 5564 -3.7 3.9 0.6 25.3 32.1 36.6 26.4 36.6 34.7 -7.3 5.0 12.7 22.8 30.8 35.5 27.9 48.1 64.7 77 6.1 3.6 21.3 28.9 31.4 47.3 57.3 65.8 -0.6 3.0 6.7 33.7 33.1 35.5 56.5 68.6 76.2 X X X X X X X x x -18.7 15.3 6.8 43.2 28.2 39.2 55.3 67.9 68.0 8.0 3.6 1.3 25.7 40.8 33.1 500 81.4 73.4 4.5 5.4 2.2 44.3 37.7 40.1 77.7 82.8 83.7 5.3 -3.5 -3.8 28.6 32.8 36.8 30.7 34.9 38.6 -16.9 -3.1 -3.9 26.5 31.8 33.0 27.8 43.9 45.6 -15.7 -7.3 -5.4 25.0 32.7 37.4 33.8 47.2 60.0 -9.5 a.3 -2.8 29.7 31.6 32.1 43.9 61.6 55.5 X 6.3 X X 39.7 X X 44.6 X X -2.7 9.3 X 28.0 29.9 X 55.1 61.8 X 4.6 4.8 X 34.2 386 X 72.4 73.0 -14.4 -2.8 2.5 41.9 35.8 40.7 62.5 66.2 81.4 ' ATL as % fatty streaks (F) plus raised lesions (RI; FaF=F : (100 - R); O-E in percentage units explained in TABLE 3.-Continued. Study Smoking data Measure of Population 8oUrC.2 athercscleroeie ReSUlts the text. X mdvxtes subgroups having fewer than live members Sorlie et al. ( 70) 139 autopsies of 9.824 Interview Visual with subject evaluation Awxiation of atherosclerosis in aorta with antemortem characteristics. simple correlation coefficients (Puerto Rican heart health program) Puerto Rican males Correlation coefficients aged 35-79 Characteristics measured at exam 1 Total (120) Rural (31) Urban (89) Systolic blood pressure Diastolic blood pressure Serum cholesterol Age, exam 1 Relative weight Physical activity Blood glucose Hematcvxit Education Income Cigarettes smoked Calories (24.hour recall) Starch (24-hour recall) Alcohol (24.hour recall) Total fats (24.hour recall) Triglycerides (fasting) Ventricular rate Vital capacity 0.25 0.27 0.24 0.19 0.29 0.16 0.29 0.38 0.28 0.31 0.39 0.29 4.08 -0.22 4.06 -0.18 -lx21 -0.14 0.14 0.05 0.17 0.23 0.33 0.21 -0.08 a.23 a.03 0.01 -0.01 4.01 032 0.37 0.31 -0.24 -0`5.5 4 12 -0.19 a.45 -0.07 -0.18 4.39 AI 18 -0.19 4.49 a.11 0.11 0.53 0.04 0.07 0.11 0.05 -0.29 -0.28 -0.29 based on the findings in that study, as well as various interpretations of those findings, have been published. Strong and Richards (74) reported the basic findings on the association of cigarette smoking and aortic atherosclerosis in 1,320 autopsied men in New Orleans, 25 to 64 years of age. Aortic lesions were evaluated visually in coded specimens and objectively by analysis of radiographs. Interviewers obtained estimates of cigarette smoking habits of the deceased men from surviving relatives. Data were compared for black men and white men, and also were analyzed in groups according to the presence or absence of diseases thought to be associated with smoking or with coronary heart disease (emphysema, lung cancer, myocardial infarction, hypertension, diabetes mellitus, stroke, etc.). Atherosclerotic involvement of the aorta was greatest in heavy smokers and least in nonsmokers for both races in the total sample, as well as in the basal group (those cases least influenced by the bias of autopsy selection). The lesions were measured not only by visual evaluation, but also by optical electronic scanning of radiographic images of flattened arteries. Atherosclerotic lesions in the abdominal aorta were more extensive in the heavy smokers than in the nonsmokers, and there was an orderly trend of increased lesions with increased smoking. In general, the magnitude of difference in extent of lesions between nonsmokers and heavy smokers was greater in the abdominal aorta than in the coronary arteries. A variety of statistical analyses of smoking measures and atheroscle- rotic lesions was applied to determine the significance of the various differences and trends. All of the analyses confirmed that the differences between the heavy smokers and the nonsmokers in extent of raised atherosclerotic lesions were significant. A one-way multivariate analysis of nine atherosclerotic variables indicated that there were statistically significant differences among the three categories of smokers (heavy, light to moderate, and nonsmokers) for lesions in the abdominal aorta. Following the initial report of Strong and Richards (74), there were three additional publications from this study. Two of these were directed toward interpretation of findings in regard to the effect of cigarette smoking on fatty streaks (the earliest grossly visible lesions of atherosclerosis) and raised atherosclerotic lesions (the more advanced stage of the atherosclerotic process). The other study was directed toward the interrelations of obesity, smoking, and athero- sclerotic lesions in these same cases. The original report by Strong and Richards (74) indicated that raised lesions, the more advanced lesions, were greater in heavy smokers than in nonsmokers. They also reported statistically significant differences for fatty streaks in the abdominal aorta and for fatty streaks in the coronary arteries, with the highest values in the nonsmokers and lowest values in the heavy smokers. The well- 46 recognized problem of evaluating fatty streaks when more advanced lesions of atherosclerosis are present made it difficult to interpret the findings on fatty streaks. Pate1 et al. (54) approached this problem by using a simple two-parameter model of fatty streaks arising from a normal intimal surface at a constant rate and with subsequent conversion to raised lesions at a constant rate. They concluded that in the abdominal aorta, smoking enhances the formation of fatty streaks as well as the subsequent conversions to more advanced lesions, and in the coronary arteries, smoking seems only to enhance the conversion of fatty streaks to fibrous plaques. Tracy et al. (77) evaluated the same data from the New Orleans study on smoking and atherosclerotic lesions. They approached the problem using a different model: N = F -+ R, where N denotes normal intima, F denotes fatty streaks, and R denotes raised lesions. In this model, class A causes are viewed as promoting the process from beginning to end, while class B agents act at the first or the second step, but not at both. Their analysis and interpretation suggest that cigarette smoking has a large class B effect. They concluded that the target tissue of smoking is the fatty streak, and the slowly progressing or regressing fatty streak (formed alike in smokers and nonsmokers) is caused to progress more rapidly or to cease to regress by smoking. Both of these studies, Pate1 et al. (54) and Tracy et al. (73, agree that smoking has a role in the progr&ion of fatty streaks to a more advanced stage of the atherosclerq$c process. Auerbach and Garfinkel in 1980 (5) published findings on smoking habits and atherosclerotic lesions in over 1,400 aortas collected at autopsy from male patients. The extent of atherosclerotic lesions (plaques, ulcerations, and calcification) increased with number of cigarettes smoked, and was also greater in excigarette smokers and pipe smokers than in nonsmokers. The findings were more striking in the abdominal aorta than in the thoracic aorta. Aortic aneurysms were found eight times more frequently among those who smoked one to two packs of cigarettes per day than in nonsmokers. LifSc (37) reported on the relationship of cigarette smoking to aortic lesions based on the Yalta sample from the World Health Organization (WHO) autopsy study of five cities in Europe (87). Information on cigarette smoking was obtained by means of inter- views with the subjects' near relatives. The prevalence and extent of atherosclerotic lesions were evaluated in autopsies of 865 men, aged 20 to 79 years, out of 1,220 deaths occurring in Yalta residents of this age and sex group during the period of study. There were significant positive relationships between smoking and the extent of fibrous plaques, complicated lesions, and calcified lesions in the abdominal aorta. 47 Aortic atherosclerosis has also been evaluated using autopsy followup of prospective epidemiologic studies of cardiovascular disease. Epidemiological studies in Puerto Rico and Honolulu documented selected risk factors, including cigarette smoking habits, during life and had standardized evaluation of atherosclerotic lesions at autopsy (56, 70). Each of these studies reported findings on the relationship of risk factors and aortic atherosclerotic lesions in more than 100 deceased men from large cohorts that had been examined and followed during life. A smaller study from Malmo, Sweden, had some of the same features as these larger studies (71). All of these studies found a significant positive relationship between cigarette smoking and aortic atherosclerosis. The prospective epidemiologic studies with autopsy followup confirmed the relationship between smoking and atherosclerotic aortic lesions found in earlier autopsy studies. The preponderance of evidence suggests that cigarette smoking aggravates or accelerates aortic atherosclerosis, and this effect on atherosclerosis may be more pronounced in the aorta than in the coronary arteries. Cerebral Vasculature The relationship between cigarette smoking and atherosclerosis in the cerebral vasculature has not been extensively evaluated. Two studies that have examined this question are summarized in Table 4. Sternby (71) reported that cigarette smokers had more extensive raised lesions in the basilar artery than had nonsmokers. This study was based on 60 autopsy subjects from 703 men born in 1914 who participated in a study of cardiovascular disease in Malmo, Sweden. Holme et al. (29) reported a positive correlation coefficient between raised lesions in the cerebral vessels and the number of cigarettes smoked; this relationship was not statistically significant, however. The limited amount of information available on the relationship between cigarette smoking and atherosclerosis in the cerebral vasculature does not allow a clear conclusion to be drawn at this time. Pathophysiologic Mechanisms of Tobacco Smoke Studies of Components of Tobacco Smoke The possible pathophysiologic mechanisms for the atherogenic influence of cigarette smoking were reviewed in the 1971 Report of the Surgeon General The Health Consequences of Smoking (80). The major components of cigarette smoke considered in that review were nicotine and carbon monoxide. Numerous investigators have studied the effect of nicotine administration, either subcutaneously or intravenously, upon experimentally induced changes in the aorta and coronary arteries of animals. When administered alone, nicotine 48 TABLE 4.-Autopsy studies of atherosclerosis involving cerebral vasulation Study Population Smoking data 8ource Measure of atherosclerosis Result.6 Sternby (711 60 autoplied subjects from 703 male8 in CHD study in Malmo. Sweden Interview with subject Visual inspection Smoking and atherosclerosis m the basilar arteries Smoking category Number Basilar artery raised lesions Non 3 1 EX 8 6 Light 18 3 Heavy 17 7 Holme et al. (29) 129 autopsies out of 16,200 men aged 4049 in Oslo CHD study Interview with subject Visual grading Correlation coefticlent between raised lesions in the cere- bra1 vessels and number of cigarettes smoked per day is 0.090 (not statistically signifxantl. induces certain degenerative or necrotic changes in the arterial wall, but these are characteristically medial changes rather than the intimal changes that characterize atherosclerosis. When nicotine is administered in combination with a high cholesterol diet, it seems to aggravate arterial damage, according to a preponderance of studies. Some studies, however, do not report this synergism between cholesterol feeding and nicotine (16, 84). Schievelbein and associates (66) reported the effect of long-term nicotine exposure on the development of arteriosclerosis in rabbits. They administered nicotine to rabbits not being fed an atherogenic diet. All animals had arteriosclerotic lesions in the aorta and coronary arteries at the end of the experiment, but there was no difference between the control group and the experimental animals administered nicotine. They reviewed the experiments of several authors who studied nicotine and their own animal experiments and concluded that the evidence did not establish a causative role for nicotine in the etiology of arteriosclerosis. A recent report by Liu et al. (38) on experimental arterial lesions in rhesus monkeys with various combinations of dietary hypercho- lesterolemia, hypervitaminosis D(2), and nicotine indicated that the combination of these three factors produced high scores for various measures of arteriosclerotic changes in aorta, coronary, and limb arteries of the monkeys. When the factors were administered singly, however, very little arterial disease was demonstrated over the period of the experiment. The group with all three factors was the only group with significant coronary arteriosclerosis as well as complicated lesions of the arteries of extremities. Dooyse et al. (15) reported the effects of chronic oral consumption of nicotine on the rabbit aortic endothelium. They found that fasting serum levels of glucose, triglyceride, total cholesterol, and LDL cholesterol were elevated in nicotine-treated rabbits as compared with controls. They found no significant differences between the experimental group and the controls for leukocyte, erythrocyte, and platelet counts, or for hematocrit and hemoglobin. Endothelial cells from the aortic arch of the nicotine-treated animals showed exten- sive changes, such as increased cytoplasmic silver deposition, in- creased formation of microvilli, and numerous focal areas of "ruffled" endothelium. The authors concluded that nicotine adminis- tered orally to rabbits has a demonstrable morphologic effect on endothelial cells in the aortic arch. While the evidence for and against a primary role for nicotine in the development or acceleration of atherosclerosis is not conclusive, nicotine is certainly one of the components of tobacco smoke for which there are both some supporting data and a rational conceptu- alization for a role in the pathogenesis of atherosclerotic lesions. There is little doubt that nicotine alone or in combination with other 50