CARDIOVASCULAR DISEASES. CARDIOVASCULAR DISEASES Introduction While the mortality and morbidity rates of coronary heart disease (acute myocardial infarction and chronic ischemic heart disease) (CHD) are lower for women than men, CHD still repre- sents the major cause of death among women in the U.S. In 1976 the United States recorded 284,055 female deaths as attributa- ble to this cause (Table 2). The difference in mortality rates between the sexes is more marked for acute myocardial infarc- tion, with males of all ages experiencing 189 deaths and females 111 deaths per 100,000 (Table 1). Observed differences by sex in susceptibility to coronary heart disease are not fully understood but appear to be affected by multiple specific risk factors within any demographic group. McGill and Stern have recently provided an extensive review of sex differences in susceptibility to atherosclerosis in humans and in experimental animals, including an analysis of factors known to predispose to atherosclerosis and its dependent dis- eases (25). Mortality Rates In the United States, the National Center for Health Statis- tics has reported mortality rates from acute myocardial infarc- tion and chronic ischemic heart disease classified by age, sex, and race, for the years 1968 and 1976 (Tables 1-3) (33). These tables show that mortality rates for acute myocardial infarction among adults up to age 64 are highest for white men and are succeeded by progressively lower rates for other men, other women, and finally, white women. Mortality rates for chronic ischemic heart diseases vary. The rates for white men are sec- ond to those for other men and close to those for nonwhite women; again, however, rates for white women are by far the lowest. Both white and nonwhite women show consistently lower rates until extreme old age. However, the differences nar- row markedly in age in comparison with those in young adult- hood and middle life (Table 1). Male-to-female mortality ratios for acute myocardial infarc- tion among adults in their 30’s and 40’s are approximately 5 to 6 for whites and 2 to 3 for nonwhites; among adults in their 70’s and 80’s, they are roughly 1.6 and 1.4. The actual number of deaths involved is very large; their distribution by age, sex, and race is shown in Table 2. Between 1968 and 1976, a striking decline occurred in the acute myocardial infarction mortality rate for men and women of all ages and races. These are shown 79 08 TABLE 1.— Death rates* for acute myocardial infarction and chronic ischemic heart di groups, by color and sex; United States, 1968-1976 sease for specified age Total White All Other Both Both Both Year and age sexes Male Female sexes Male Female sexes Male Female 1976 Acute myocardial infarction Allages ............,. 148.8 189.0 110.8 158.7 202.2 117.3 84.0 100.3 69.0 25-34 years ............. 2.8 4.6 1.1 2.6 4.3 0.9 4.2 6.4 2.3 35-44 years ............. 27.0 46.2 8.8 26.6 46.1 7.6 30.4 47.5 10.3 45-54 years ............. 111.7 186.9 41.3 111.8 190.1 37.7 111.2 159.8 68.9 55-64 years ............. 309.5 490.3 147.2 312.2 501.1 142.1 283.2 386.5 194.8 65-74 years ............. 660.1 989.8 406.8 674.5 1,024.7 406.5 524.6 667.9 409.9 75-84 years ............. 1,328.0 1,806.7 1,035.7 1,364.8 1,881.4 1,054.3 917.0 1,061.1 813.0 85 years and over ........ 2,038.0 2,564.7 1,790.3 2,135.0 2,709.6 1,869.9 1,126.5 1,369.1 990.1 1968 Allages ............. 185.4 243.0 130.6 195.9 258.0 136.7 109.5 133.2 87.7 25-34 years ............. 4.6 7.2 2.2 4.1 6.5 1.7 8.7 13.1 5.0 35-44 years ............. 42.3 70.9 15.2 40.3 69.6 12.1 57.9 81.6 37.9 45-54 years ............. 158.5 267.1 56.8 157.6 270.4 51.3 166.6 236.2 105.3 55-64 years ............, 420.8 668.3 197.1 423.9 684.3 188.4 390.5 512.5 281.0 65-74 years ............, 900.5 1,315.0 574.1 919.8 1,360.8 574.4 706.7 870.1 571.2 75-84 years ............. 1,687.1 2,228.4 1,316.5 1,732.1 2,306.5 1,342.8 1,103.1 1,291.4 961.1 85 years and over ........ 2,911.8 3,570.7 2,553.0 3,012.9 3,715.3 2,637.8 1,782.4 2,163.4 1,526.2 18 TABLE 1.— Death rates* for acute myocardial infarction and chronic ischemic heart disease for specified age groups, by color and sex; United States, 1968—1976—(Continued) Total White All Other Both Both’ Both Year and age sexes Male Female sexes Male Female sexes Male Female 1976 _ Chronic ischemic heart disease Allages ............. 150.2 153.5 147.0 155.5 157.7 153.4 115.4 125.4 106.4 25-34 years ............. 1.6 2.4 0.8 1.2 1.9 0.5 4.2 6.1 2.5 35-44 years ............. 12.8 20.3 5.6 10.6 17.5 3.9 27.5 41.0 16.3 45-54 years ............, 57.7 90.9 26.7 50.4 82.6 20.1 116.1 160.7 77.4 55-64 years ............. 173.3 258.5 96.8 159.5 244.3 83.2 302.2 396.1 222.0 65-74 years ............. 487.4 674.8 343.4 467.8 660.5 320.4 672.1 805.8 565.2 75~—84 years ............. 1,621.5 1,947.4 1,422.6 1,626.0 1,968.0 1,420.4 1,572.0 1,742.7 1,448.8 85 years and over ........ 4,647.4 4,945.8 4,507.0 4,859.8 5,208.0 4,699.1 2,650.8 2,782.4 2,576.9 1968 Allages ............. 150.6 156.3 145.1 153.1 158.3 148.2 132.0 141.6 123.3 25-34 years ............. 1.6 2.3 11 1.0 1.6 0.4 6.2 7.2 5.3 '-44 years... oo .....,. 13.6 20.5 TA 10.4 17.0 4.0 38.8 49.8 29.5 45-54 years ............. 57.0 85.6 30.2 47.5 76.0 20.7 142.6 175.8 113.3 55-64 years ............, 190.6 273.4 115.7 169.2 253.4 93.0 393.1 468.6 334.8 65-74 years ............, 590.4 769.1 449.7 560.6 742.8 417.9 889.5 1,025.0 777.2 75-84 years ............. 1,826.0 2,075.5 1,655.3 1,833.9 2,093.7 1,657.8 1,724.6 1,858.1 1,628.0 85 years and over ........ 5,523.6 5,636.6 5,468.4 5,695.3 5,831.8 5,629.4 3,605.9 3,736.6 5 518.0 *Rates are deaths per 100,000 population. For acute myocardial infarction, rates are based on deaths assigned to category number 410 of the Eighth Revision of the International Classification of Diseases, adapted for use in the United States, adopted in 1965, and for chronic ischemic heart disease, to category number 412 of this revision SOURCE: Rosenberg, H.M. (33). 0o bo TABLE 2.— Number of deaths* for acute myocardial infarction and chronic ischemic heart disease for specified age groups, by color and sex; United States, 1968 and 1976 Total White All other Both Both Both Year and age sexes Male Female sexes Male Female sexes Male Female 1976 Acute myocardial infarction Allages ............. 319,477 197,429 122,048 295,613 183,820 111,793 23,864 18,609 10,255 25-34 years ............. 890 718 172 720 598 122 170 120 50 35-44 years ............. 6,223 5,182 1,041 5,838 4,558 780 885 624 261 45-54 years ............,. 26,405 21,361 5,044 23,479 19,407 4,072 2,926 1,954 972 55-64 years ............,. 62,091 46,516 15,575 56,623 43,072 13,551 5,468 3,444 2,024 65-74 years ............. 93,695 61,038 32,657 86,566 57,004 29,562 7,129 4,034 3,095 75-84 years ............. 89,969 46,395 43,574 84,852 43,912 40,940 5,117 2,483 2,634 85 years and over ........ 40,068 16,132 23,936 37,939 15,201 22,738 2,129 931 1,198 1968 Allages ............. 369,610 236,017 183,593 342,999 220,517 122,482 26,611 15,500 11,111 25-34 years ............. 1,099 838 261 846 664 182 253 174 19 35-44 years ............. 9,980 8,132 1,848 8,412 7,122 1,290 1,563 1,010 558 45-54 years ............. 36,032 29,368 6,664 32,261 26,860 5,401 3,771 2,508 1,263 55-64 years ............. 76,108 57,387 18,721 69,504 53,287 16,217 6,604 4,100 2,504 65-74 years ............. 109,672 70,564 39,108 101,863 66,205 35,658 7,809 4,359 3,450 75~84 years” ............. 100,312 53,838 46,474 95,613 51,436 44,177 4,699 2,402 2,297 85 years and over ........ 36,135 15,711 20,424 34,317 14,824 19,493 1,818 887 931 1976 Chronic ischemic heart disease Allages ............. 322,382 160,375 162,007 289,572 148,372 146,200 32,810 17,003 15,807 25-34 years ............. 502 381 121 332 266 66 170 115 55 35-44 years ............. 2,937 2,273 664 2,137 1,734 403 800 539 261 &8 age groups, by color and sex; United States, 1968 and 1976—(Continued) Total White All other Both Both Both Year and age sexes Male Female sexes Male Female sexes Male Female 45-54 years .........000. 13,649 10,391 3,258 10,593 8,426 2,167 3,056 1,965 1,091 55-64 years .....ss eee eee 34,765 24,525 10,240 28,929 20,996 7,933 5,836 3,529 2,307 65-74 years ...cec eee eee 69,176 41,612 27,564 60,042 36,745 23,297 9,134 4,867 4,267 75-84 years ......... eee 109,860 50,010 59,850 101,088 45,932 55,156 8,772 4,078 4,694 85 years and over ........ 91,368 31,109 60,259 86,358 29,217 57,141 5,010 1,892 3,118 1968 All ages .....eeceeeee 300,216 151,815 148,401 268,124 135,333 132,791 32,092 16,482 15,610 25-34 years 2.2... eee eee 390 262 128 211 166 45 179 96 83 35-44 years ............. 3,212 2,350 862 2,162 1,734 428 1,050 616 434 45-54 years .......-..4.. 12,953 9,412 3,541 9,727 7,545 2,182 3,226 1,867 1,859 55-64 years ...........4. 34,475 238,481 10,994 27,743 19,732 8,011 6,732 3,749 2,983 65-74 years ...........0. 71,905 41,270 30,635 62,076 36,135 24,941 9,829 5,135 4,694 75-84 years ........ 0 108,576 50,145 58,431 101,229 46,689 54,540 7,347 3,456 3,891 85 years and over ........ 68,548 24,801 43,747 64,870 23,269 41,601 3,678 1,532 2,146 *Number of deaths due to acute myocardial infarction are those assigned to category number 410 of the Eighth Revision of the International Classification of Diseases, adapted for use in the United States, adopted in 1965; and for chronic ischemic heart disease to category number 412 of this revision SOURCE: Rosenberg, H.M. (33). as percent changes in rate in Table 3. The percent change has been larger at younger ages (Tables 2 and 3). The changes for chronic ischemic heart disease are similar but less dramatic (Table 3). Atherosclerosis Differences in heart attack mortality rates among men and women parallel pathology data concerning atherosclerotic plaques of the coronary arteries. The International Atherosclerosis Project systematically collected autopsy obser- vations on persons from 14 geographic locations and 19 ethnic groups in different parts of the world, and found that women from 11 of the 19 groups, when compared to their male counter- parts, had as much or even more aortic atherosclerosis. Men over age 39 had more raised plaques in their coronary arteries than women (24). These findings indicate that the occurrence of coronary plaques was parallel to heart attack rates, but that the occur- rence of aortic lesions was not. Coronary plaque severity had a male-to-female ratio of 1.61 among whites and of 1.14 among blacks. Studies of a white population in Sweden (40) and of west- ern Europeans from five locations (18) demonstrate similar find- ings: a clear excess of coronary atherosclerosis among men and a similar severity of aortic atherosclerosis among men com- pared to women. Autopsy studies thus show a selective liability of the male coronary arterial bed for atherosclerosis, as compared to the female, especially among white men but also among men of other races. The pathological findings are congruent with the clinical data on heart attack mortality rates. Autopsy studies also show that, among men or women with manifest coronary heart disease, women patients have roughly the same preva- lence of advanced atherosclerotic lesions of the coronaries as men (41). These data suggest that the amount of atherosclerosis necessary to precipitate a heart attack is the same, on the aver- age, in both sexes. This generalization about the amount of coronary atherosclerosis appears to hold for heart attacks at younger and older ages, for recent and old infarcts, and coro- nary occlusion without infarct, and for stenosis, as well as for complicated and calcified lesions and raised plaques in the coro- nary arteries (41). It should be noted that the grading of atherosclerosis at au- topsy is not a simple matter because there are several types of lesions and several ways of evaluating or measuring them. Moreover, the development of the different sorts of lesions is 84 TABLE 3.— Percent change* between 1968 and 1976 in death rates for acute myocardial infarction and chronic ischemic heart diseases for specified age groups, by color and sex: United States Total White All Other Both Both Both Age Sexes Male Female Sexes Male Female Sexes Male Female Acute myocardial infarction Allages ............, ~19.7 —22.2 ~15.2 -19.0 —21.6 -14.2 -23.3 ~24.7 —21.3 25-34 years ............. ~39.1 ~—36.1 -50.0 —36.6 —33.8 ~47.1 -51.7 -51.1 -54.0 35-44 years ............. —36.2 ~34.8 ~42.1 ~34.0 —33.8 ~37.2 -47.5 ~41.8 -57.0 45-54 years ............. —29.5 ~30.0 -27.3 ~—29.1 ~-29.7 ~26.5 —-33.3 —32.3 -34.6 55-64 years ............, —26.4 —26.6 ~25.3 —26.4 ~26.8 —-24.6 -27.5 —24.6 ~30.7 65-74 years ............. —26.7 —24.7 ~29.1 —26.7 —24.7 —29.2 ~25.8 ~-23.2 —28.2 75-84 years ............. -21.3 ~18.9 —21.3 —21.2 ~-18.4 -21.5 -16.9 -17.8 —-15.4 85 years and over ........ —80.0 ~28.2 —29.9 ~29.1 —27.1 ~29.1 —-36.8 ~36.7 ~35.1 Chronic ischemic heart diseases Allages ............. -0.3 -1.8 1.3 1.6 -0.4 3.5 ~12.6 ~11.4 -13.7 25-34 years ............. 4.3 ~27.3 20.0 18.8 25.0 -82.3 ~15.3 -52.8 35-44 years ............, -5.9 -1.0 —21.1 1.9 2.9 —2.5 —29.1 -17.7 —44,7 45-54 years ............. 1.2 6.2 -11.6 6.1 8.7 -2.3 -19.6 ~8.6 -31.7 55-64 years ............. -9.1 —5.4 -16.3 ~5.7 -3.6 —10.5 —24.1 -15.5 ~33.7 65-74 years ............, -17.4 —-12.3 ~23.6 -16.6 —11.1 —23.3 —24.4 ~21.4 —27.3 75-84 years ............, ~11.2 -6.2 —14.1 ~-11.3 -6.0 -14.3 -8.8 ~6.2 -11.0 85 years and over ........ -15.9 ~12.3 -17.6 ~14.7 -10.7 -16.5 —26.5 -25.5 —26.8 *Percent changes are based on rates per 100,000 population. For 1968 and 1976, rates for acute myocardial infarction are based on deaths assigned to category number 410 of the Eighth Revision of the International Classification of Diseases, adapted for use in the 2 United States, adopted in 1965, and for chronic ischemic heart disease, on category number 412 of this revision SOURCE: Rosenberg, H.M. (33). not necessarily parallel. Sternby provides a useful discussion of issues in the grading of atherosclerosis (40). Nevertheless, the major studies noted above provide strong evidence that women have less coronary atherosclerosis on the average than men of the same age in the same population Risk Factors Factors present in individuals which correlate with future liability to disease are risk factors for that disease. In the case of heart attack, for example, it has been shown that age, male sex, cigarette smoking, hypertension, elevated blood cholesterol, and several other conditions are positively and independently associated with the probability of heart attack. The level of high-density lipoprotein cholesterol in the serum has a negative correlation with heart attack; that is, higher levels are protec- tive. The various risk factors have been identified for both men and women and have been shown on multivariate analysis to be independent. A combination of risk factors is synergistic, pro- ducing an associated risk greater than the simple sum of the individual risks. Although the data for women are much less extensive than for men, they indicate that cigarette smoking is a major risk factor for heart attack in women. The Effect of Smoking ATHEROSCLEROSIS There is little autopsy information about the amount of atherosclerosis in women smokers. Sackett and his associates reported on aortic atherosclerosis among both men and women: of their 450 female subjects, 309 were nonsmokers, 52 smoked less than a half pack per day, and 89 smoked more (34). Mean, age-adjusted aortic atherosclerosis was found to increase in conjunction with the amount and duration of smoking. A study of the intramyocardial arteries and arterioles of the heart in 13 women and 21 men who were nonsmokers, and 16 women and 27 men who were smokers, indicated that prolifera- tive lesions in intramyocardial arteries were more advanced relative to age in smokers than nonsmokers. It was also found that subendocardial arterioles were thickened in smokers. A separate analysis by sex was not performed, but the authors remarked that the lesions developed as rapidly and as exten- sively in women as in men in both smoking and nonsmoking groups (28). Studies of the severity of atherosclerotic plaques in the ar- teries of women who smoked in comparison with those who did 86 48 TABLE 4.—Coronary heart disease mortality ratios related to smoking — prospective study Author, Number and Follow- Number year, type of Data up of country populations collection (years) deaths Cigarettes/day Age Variation Hammond 358,584 Questionnaire 6 14,819 M F Males and males and follow-up NS ........ 1.00 1.00 40-49 50-59 60-69 70-79 Garfinkel, 445,875 of death certi- 1-9 ........ 1.27 0.81 1969, females age cate 10-19 ...... 1.00 1.22 NS ...... 1.00 1.00 1.00 1.00 U.S.A. 40-70 at 20-30 ...... 1.75 1.52 19 0... 1.00 1.50 1.48 1.14 entry. >40 ....... 1.77 0.61 10-19 ... 239 2.13 1.82 1.41 20-30 ... 3.76 2.40 1.91 1.49 >40 20... 3.51 2.79 171 1.47 Females 40-49 50-59 60-69 70-79 NS ...... 1.00 1.00 1.00 1.00 19) ....., 1.31 1.15 1.04 0.74 10-19 ... 2.04 2.37 1.79 0.98 20-30 ... 3.62 2.69 2.00 1.27 >40 20... +3.31 3.73 +2.02 Based on 5-9 deaths NS = nonsmokers, M = males, F = females SOURCE: U.S. Public Health Service (44,45). not smoke involve too few subjects to be satisfactory. Inves. tigating the relationship of these arterial lesions and cigarette smoking in women is fundamental to understanding the occur- rence of heart attack and other ischemic diseases. CORONARY HEART DISEASE Coronary heart disease (acute myocardial infarction and chronic ischemic heart disease) occurs with greater frequency in smoking than in nonsmoking women. The prospective study of Hammond and Garfinkel, published in 1969, included data on approximately 446,000 women between the ages of 40 and 79 (10). The inerease in mortality ratios in conjunction with in- creasing numbers of cigarettes smoked per day for various ages is shown below in Table 4 (43,44), Mortality ratios were higher for younger ages and lower for older ages. The one-pack-a-day smoker’s risk of death from heart attack was approximately twice that of the nonsmoker. The prospective data of Shapiro and colleagues are based on a population of 120,000 men and women (36). Using a sampling factor of about one-thirtieth, they examined 4,301 women at risk of a first myocardial infarction between the years 1962 and 1964. The smokers compared with nonsmokers had roughly twice as many rapidly fatal heart at- tacks and heart attacks that were not fatal within 48 hours. The ratio was approximately 2.9 among younger women aged 45 to 54 and 1.8 for the subjects aged 55 to 64. Heavy smokers had higher ratios, but the data did not permit a detailed study of dose relationships or of the experience of female ex-smokers. A recent study examined the cause-specific mortality of 6,194 British women physicians over the period 1951 to 1973 (6). Table 5 presents the results of this study in conjunction with the pre- viously published results among male physicians during the same period (7). The clear association of cigarette smoking and ischemic heart disease previously described in males was con- firmed in female physicians. For women who reported smoking 15 or more cigarettes per day, mortality due to ischemic heart disease was more than double that of nonsmokers. Although the results demonstrated a similar effect of smok- ing in the development of ischemic heart disease in both male and female physicians, the association of smoking with heart disease was less striking in women physicians. Ischemic heart disease was less prominent as a proportional cause of death in this population of women than in male colleagues (16 percent vs. 32 percent of all deaths). Ischemic heart disease mortality was only 26 percent higher for all ever-smoked women than for never-smoked women. However, for females who smoked heav- 88 68 TABLE 5.— Death from ischemic heart disease and smoking habits when last asked, British physicians 1951-1973 Annual Death Rate per 100,000 Persons Standardized for Age X? Number Current Smokers - Dose Per Day Nonsmokers Total of vs. Popul. Deaths Nonsmokers Ex-smokers 1-14 15-24 > 25 others Trend Women 6194 179 138 126 132 304 292 wee 21.14* (number of cigarettes) Men 34,440 3191 413 533 501 598 677 22.59% 53.56* (any tobacco—grams) (1 gram = 1 cigarette) *P<0.001. SOURCE: Doll, R. (6,7). ily (= 25 cigarettes per day), the relative risk of death from ischemic heart disease was 2.2, a finding consistent with that demonstrated in males, who had a relative risk of 1.6. In such studies, standardization for amount smoked daily by each of the sexes does not, however, correct for differences in age at initiation of smoking and degree of inhalation. This fact greatly complicates comparison of the magnitude of biologic ef- fect in the two sexes. This “cohort effect” (i.e., unmeasured but documented dissimilarities in total smoking experience) may lead to an erroneous interpretation that cigarette smoking is less damaging to women than to men. This issue cannot be re- solved until studies examine the effect of smoking in more re- cent cohorts of women whose lifetime smoking behavior is more similar to that of men. Among 26,467 Swedish women observed during a 10-year period, the risk of developing fatal coronary heart disease was significantly higher among smokers than nonsmokers (50). The relative risk was 1.9 at ages 40 to 49 and 1.3 at ages 50 to 59. An extensive mortality study in Japan also reported a highly signif- icant increase in deaths from ischemic heart disease among female smokers, with a mortality ratio for smokers of 1.6 (29). Coronary heart disease morbidity data are available on women from prospective studies in Framingham, Mas- sachusetts, Tecumseh, Michigan, and the greater New York areas. The Tecumseh data of 1967 do not show a relationship of such morbidity with smoking (Table 6) (8). The Framingham Heart Study found an increased risk for women smokers, but the associations were weak (19,20). The study of Shapiro and colleagues considered both mortal- ity and morbidity (36). It reported separately on deaths within 48 hours of onset and on all definite myocardial infarctions after that time interval. Using this classification, the incidence of coronary heart disease among women smokers was distinctly higher than it was among nonsmokers. While there is some variability in the strength of this associa- tion, the data from the various prospective studies of mortality and morbidity from coronary heart disease establish smoking as a positive correlate, or risk factor, for women. However, the risk ratios tend to be smaller than for men at a given level of cigarette consumption in all age groups. This trend may result from the different smoking patterns reported by men and women who smoke the same number of cigarettes per day (6,7,25). Men generally begin smoking at an earlier age and have thus smoked for a longer time period than women. Men also inhale more often than women and are more likely to smoke more than half of a cigarette. These smoking styles would ex- 90 16 TABLE 6.—Coronary heart disease morbidity as related to smoking Author, Number Follow- Number year, and type of Data up of country population collection years! incidents? Cigarettes/day® Pipes, cigars Epstein, 6,568 male Initial medical 4 96 male, 92 Males Males 1967, and female examination female 40-59 60 and over 40-59 U.S.A. residents of and repeat CHD inelud- NS ...........005 1.00 (1) 1.00 (7) SM ........... 1.80 (2) Tecumseh, follow-up ing deaths, | Op, 6.383 (10) 1.27 (11) Mich. examinations. angina, and Cigarettes ....... 5.20 (36) 1.90 (23) 60 and over myocardial SM ........--- 0.80 (6) infarctions Females 40-59 60 and over NS ...... eee eee 1.00 (21) 1.00 (47) EX ..........005. 0.89 (3) 1.31 (5) Cigarettes ....... 1.02 (14) 0.42 (2) ‘Reexamination of patients was spread over 142-6 year period, but data are reported in terms of 4-year incidence rates. 2Actual number of CHD incidents derived from data on incidence and total in smoking class. 3Risk ratios— actual number of CHD incidents shown in parentheses. SM = smokers, NS = nonsmokers, EX = ex-smokers. SOURCE: U.S. Public Health Service (45). pose men to a larger dose of smoke per cigarette and a larger lifetime amount than that experienced by women. Case control and retrospective studies of women who have had heart attacks have suggested an increased incidence of heart attack among smokers. For example, a case control study of 55 women who had heart attacks before age 50 (an uncommon event in women) found that 89 percent were smokers in contrast to 55 percent in a control group without myocardial infarction. Heavy smokers (35 or more cigarettes per day) had an estimated myocardial infarction rate approximately 20 times that of the nonsmokers. As far as possible, women using oral contracep- tives and those with other identifiable risk factors were excluded from the study (37). Spain and his associates conducted a retrospective autopsy study of women who had died suddenly of coronary heart dis- ease and compared this verified diagnosis to the women’s smok- ing habits as reported by the closest living relative (38). Only witnessed sudden deaths were included in the data. Compari- sons were made between women who had died of coronary heart disease and women who died suddenly of causes other than heart attack. It was found that 62 percent of the women suffer- ing sudden cardiac death were heavy smokers in contrast with only 28 percent of the control group. For those who smoked heav- ily, the mean age at death was 19 years younger than that of nonsmokers; lighter smokers died at an intermediate mean age. In a retrospective study emphasizing psychosocial variables, Talbott and associates reported on 64 white women who died suddenly of arteriosclerotic heart disease (42). They found that women who died suddenly smoked more cigarettes than the comparison group. The relative risk for those smoking more than a pack a day compared with those smoking less than a pack a day was 3.9 (p<.004). Smoking, as well as other risk factors, raises the already somewhat higher risk of myocardial infarction among women who use oral contraceptives. During the child-bearing years, the use of oral contraceptives doubles the risk of myocardial infarction; women who both smoke and use oral contraceptives have approximately 10 times the risk of women who neither smoke nor use oral contraceptives (14). These issues are consid- ered below in a separate section. Cessation of Smoking and “Tar” and Nicotine Content of Cigarettes Existing data are inadequate to determine the effect of smok- ing cessation on the incidence of coronary heart disease in 92 women. Hammond and associates have reported that mortality rates from coronary heart disease were lower in women who smoked low-“tar” and low-nicotine cigarettes (as sold in the 1960s) than in those who smoked medium level products, and still lower than for those who smoked high-“tar” and high- nicotine products; even so, the mortality rate for those women smoking low-“tar”, low-nicotine products was significantly higher than that of nonsmokers (11). Evidence considered below suggests that stopping smoking is beneficial in the treatment of women suffering from peripheral vascular disease. ANGINA PECTORIS The Framingham Heart Study reported that there was a posi- tive association between smoking and angina pectoris among men but not among women (20). In an extensive study con- ducted in New York City, Shapiro and colleagues reported a positive association between the development of angina pec- toris and smoking among men and a nonsignificant positive trend among women (37). Among patients with angina pectoris, smoking lowers the exercise threshold for the onset of angina (46). Only male patients have been studied thus far; equivalent data apparently have not been published for women with an- gina and angiographically proven coronary atherosclerosis. CEREBROVASCULAR DISEASE The incidence of stroke as a manifestation of cerebrovascular disease appears to be somewhat greater in men than in women, but the difference is small (21,30,43). In an autopsy assessment of cerebrovascular atherosclerosis, Sternby reported more atherosclerosis of the common carotid artery and the carotid sinus in men than women. There was also more intracranial atherosclerosis of certain vessels in men than women. However, using the area-grading method, no sex dif- ference was found in total intracranial atherosclerosis (40). The International Atherosclerosis Project also reported a slight ex- cess of cerebrovascular atherosclerosis among males (24). On the whole, the available pathological evidence suggests a minor increase in cerebrovascular atherosclerosis among men in com- parison with women, although some studies fail to confirm this conclusion (see 40). It is not clear whether smoking is a risk factor among women for the development of atherothrombotic stroke. Kannel has discussed the issue and the current literature in some detail (19). The Framingham Heart Study has reported a dose-related 93 ¥6 TABLE 7.— Deaths from cerebrovascular disease related to smoking Number of Number deaths due Author, and type underlying to year, of popu- Data Follow-up CVD as Mortality country lation collection years cause ratios Hammond 358,584 Questionnaire 6 4,099 Age and males and follow- Cigarettes/day 40-49 50-59 60-69 70-79 Garfinkel, 445,875 up of death Males 1969, females certificate Never smoked 1.00 1.00 1.00 1.00 U.S.A. 40-79 years 1-9 2... ee. eee 2.79 1.95 1.30 0.95 of age at 10-19 ......... 1.14 1.48 +1.44 0.92 entry. 20-380) ......... 2.21 2.08 1.62 1.22 >40 ....ceeeeee 1.64 2.40 1.72 +0.68 Females Never smoked 1.00 1.00 1.00 1.00 L-9 Lecce cee eee eee 1.50 1.26 1.26 0.83 10-19... cece eee eee 2.60 2.70 2.15 +0.57 20-30 wo... eee eee 2.90 2.67 1.83 1.28 > 4D Lecce cece eee +5.70+3.52 — — SOURCE: U.S. Public Health Service (44,45), correlation between the incidence of atherothrombotic stroke and cigarette smoking in men but not in women. The extensive prospective study of Hammond and Garfinkel, which involved almost 446,000 women and recorded 1,905 deaths from cere- brovascular disease during a six-year period, found that smok- ing was a positive correlate for such mortality (10); in both men and women, the mortality ratio was increased by roughly 2 or 2.5 times (Table 7) (44,45). That some of these deaths may have involved subarachnoid hemorrhage rather than brain infarction, is suggested by a re- cent report that found the incidence of subarachnoid hemor- rhage to be positively associated with smoking for both men and women (2). The relative risk for men was 3.9 and for women, 3.7. The association appeared to relate to hemorrhage from rup- tured cerebral aneurysms rather than to other conditions that may give rise to subarachnoid hemorrhage. A synergism be- tween smoking and the use of oral contraceptives and sub- arachnoid hemorrhage is noted below (31). The Japanese study cited in the discussion of ischemic heart disease has also re- ported on 366 deaths from cerebrovascular disease among women who smoked (29). The risk ratios for subarachnoid hemorrhage and cerebral hemorrhage were both significantly increased among women smokers (p<.001) as was the risk rate for the category, “other forms of cerebrovascular disease” (p<.05). ARTERIOSCLEROTIC PERIPHERAL VASCULAR DISEASE Clinicians have noted that arteriosclerotic peripheral vascu- lar disease is more common in men than women. Sternby has reported from autopsy studies that men generally have some- what more atherosclerosis of the femoral and pelvic arteries than women (40). Kannel has reviewed the relationship of smoking to the inci- dence of arteriosclerotic peripheral vascular disease (19). In the Framingham Heart Study the incidence of peripheral vascular disease was increased among smokers of both sexes; cigarette smoking was as strong an independent risk factor in women as in men. Heavy smokers had a threefold increased incidence. Weiss studied 245 women with arteriosclerotic peripheral vascular disease (49). Ex-smokers who had not smoked for 5 years or more had nearly a normal risk ratio of 1.06; those who had not smoked for the last 1 to 5 years had a risk of 1.70; continuing smokers of less than a pack a day, 5.15; pack a day smokers, 11.53; and those smoking more than a pack a day, 15.56 (relative to nonsmokers, 1.00). The increased risk was particu- 95 larly associated with proximal (aortoiliac) disease, and there was less association with distal (femoropopliteal) disease. Age- standardized relative risk ratios for those smoking a pack a day were 30.06 for proximal and combined proximal and distal dis- ease and 6.32 for distal disease alone. A retrospective study of 217 patients who underwent arterial reconstructive procedures of various kinds for peripheral vascu- lar disease has been reported by Myers and colleagues (27). Diabetics were excluded from the report. There were 164 male and 53 female patients. The late patency rate of the vascular reconstruction was followed for 1 to 4 years. The authors re- ported that the number of cigarettes smoked before surgery did not influence the outcome, but cessation of smoking after surgery had a favorable impact. There were no significant dif- ferences in outcome between men and women. The patency rate 4 years after aortofemoral surgery was 90 percent in those who smoked five or fewer cigarettes per day after surgery and 75 percent in those who smoked a greater amount. Following femoropopliteal reconstruction, the 2-year patency rates were 95 percent for those who stopped smoking, 75 percent for those smoking as many as 15 cigarettes per day, and 65 percent for those who continued to smoke more than 15 cigarettes per day. AORTIC ANEURYSM Studies have not been reported for women with respect to atherosclerotic aortic aneurysm and smoking. Deaths for women are about one-fifth those for men (10). HYPERTENSION Smoking is not associated with an increased prevalence of essential hypertension in men or women (39). However, smoking does combine with hypertension (and other risk factors) as a risk factor for heart attack, synergistically compounding the risk. Two recent case control studies of rapidly progressive, severe or malignant hypertension have found that there is an overrep- resentation of smokers among patients with this uncommon phase of hypertension (3,13). In one study of 82 patients who developed malignant hypertension, 67 were smokers. Thirty- three of those were women. In the study, 77 percent of the female patients with malignant hypertension smoked, and only about 44 percent of those with essential hypertension and of the general female population smoked. The difference is highly sig- nificant. A similar and parallel study of 48 patients with malig- nant hypertension contained 33 men and 15 women; 25 men (76 96 percent) and 8 women (53 percent) were smokers compared with 44 percent and 80 percent, respectively, of a group of 44 men and 44 women with nonmalignant hypertension. The difference is significant for men but does not reach significance for women. VENOUS THROMBOSIS The section of the 1979 Surgeon General’s Report dealing with venous thrombosis noted a case control study by Vessey and Doll of 84 women who had venous thromboembolism (45). There was no significant relationship to smoking, although there was a trend (p=0.08) reasonably attributable to chance (46). Simi- larly, Lawson, Davidson, and Jick reported no association with smoking among 60 premenopausal women who used oral con- traceptives and who had uncomplicated venous thromboem- bolism (22). The issue is reopened, however, by a recent paper derived from the Walnut Creek Contraceptive Drug Study. The authors analyzed 38 cases of venous thromboembolic events among the approximately 16,700 women followed in the study. These women were matched with 8,174 controls from the same cohort, providing each case with 61 to 559 comparison subjects. The relative risk of cigarette smoking was 2.6 with a one-sided p value of less than 0.01. On multivariate analysis, the smoking effect was independent and remained significant. Of the 17 idiopathic cases of thromboembolic disease, 65 percent occurred in smokers, while 33 percent of the controls were smokers. The relative risk for smokers was 4.2. Both smoking and oral con- traceptive use were independent risk factors for venous throm- boembolic disease in this cohort of women (32). The same section of the 1979 Surgeon General’s Report noted a controversy about whether smokers who suffered myocardial infarction had a relative protective effect from leg vein throm- bosis in the immediate post infarction period (45). The authors did not provide an analysis for each sex. A recent investigation of women undergoing gynecologic op- erations has studied the incidence of deep vein thrombosis of the leg in relation to smoking. In the prospective study of 231 women, their smoking habits during the month before the oper- ation were determined. The occurrence of deep vein thrombosis (DVT) was assessed by the radioactive fibrinogen technique, with routine scans on the first, third, and sixth postoperative days. Of the 231 patients, 99 smoked and 132 did not smoke. Eight of the smokers (8.1 percent) and 29 of the nonsmokers (22 percent) developed DVT. Following an analysis of other factors, the authors concluded that smoking provided an apparent “‘pro- 97 tective” effect against postoperative DVT, based on the fact that smokers constituted only 21 percent of the patients with DVT. They also noted that the women who developed DVT weighed more than those who did not and that smokers who developed CVT were more overweight than nonsmokers with DVT (5). In a continuing prospective study of the relationship of blood clotting and blood thrombogenic properties to ischemic heart disease, Meade and associates have reported on a number of blood coagulation variables and their relationship to smoking among 1,426 men and 638 women in England (26). Forty-three percent of the men and 36 percent of the women were smokers. Smoking was not found to have an effect in women on factors V or VII, fibrinogen, fibrinolytic activity, antithrombin ITI, platelet adhesiveness, or platelet count. Smoking decreased fib- rinolytic activity in men and decreased factor VIII activity in both men and women. Oral contraceptive users were found to show an increase in fibrinolytic activity only if the women were nonsmokers. : HIGH-DENSITY LIPOPROTEIN High-density lipoprotein (HDL) is a protein complex that transports cholesterol in the blood. A higher level of HDL is correlated with a reduced risk of heart attack. It has been ob- served that women who smoke have lower levels of HDL than expected (1,4,9). Oral Contraceptive Use, Smoking, and Cardiovascular Disease The association of oral contraceptive use and an increased incidence of certain cardiovascular disorders has attracted much interest. Smoking has emerged as a strong synergistic risk factor, and an additional study has focused on smoking as an independent risk factor. The effects of smoking and of estrogen and progestin con- traceptives on the level of high-density lipoprotein in women have been studied by Bradley and associates. They measured serum HDL among almost 5,000 women between the ages of 21 and 62 (4). They reported that the use of oral estrogens raised the level of HDL significantly above the level in nonusers while progestin use lowered it. Combination drugs tended to change the HDL level according to their relative estrogen-progestin formulation. The average HDL concentration was reduced by smoking. Among nonsmoking women the HDL concentration was 63.7 + 16.8 mg/dl. This was reduced by 2.2 mg/dl for those smoking half a pack per day; and by 7.3 mg/d] for those smoking 98 one or more packs per day. A reduction in the HDL level among women who smoked was also reported from Holland. This study found an independent negative association with the HDL level among oral contraceptive users (1). It has been reported from long-term studies that women using oral contraception have a two to threefold statistically significant increase in risk of venous thromboembolic disease when compared to those using other forms of contraception (47). This study concluded that smoking did not significantly in- crease the incidence of venous thromboembolism (46). By con- trast, the Walnut Creek Study reported that smoking contrib- uted to venous thromboembolism among both users and nonus- ers of oral contraceptives (32). Conclusions about the effect of smoking on venous thromboembolic phenomena, therefore, must be regarded as uncertain at this time since there are few relevant studies and they provide somewhat contrary conclu- sions. In 1973, the Collaborative Group for the Study of Stroke in Young Women estimated that the relative risk of cerebral is- chemia or thrombosis was approximately nine times greater for women who use oral contraceptives than for those who do not. A detailed analysis of smoking was not presented, but one of the study’s striking findings was the high proportion of women with stroke who currently or at some time smoked cigarettes regu- larly (73.8 percent), compared with smoking rates of 43.4 percent among neighborhood controls aged 17 to 44. The study also found an increase in hemorrhagic strokes among white women. Almost half of the hemorrhagic strokes were attributable to bleeding from congenital aneurysms leading to subarachnoid hemorrhage (5). Recently an association between smoking and aneurysmal subarachnoid hemorrhage in both men and women has been documented (2). The Walnut Creek Contraceptive Drug Study reported that in a cohort of approximately 16,700 women, the risk of sub- arachnoid hemorrhage for smokers was 5.7 times that of nonsmokers; the risk for oral contraceptive users was 6.5 times that of nonusers; and the relative risk for women who used both cigarettes and oral contraceptives was 22 times as great. Past users of oral contraceptives also had an increase in relative risk, but an analysis of risk was not possible because of the small number of cases (31). The risk of myocardial infarction in women is increased by cigarette smoking and by the use of oral contraceptives, it is compounded when both are used together. For example, Mann and associates reported a retrospective study of 63 women below the age of 45 with acute myocardial infarction. The pro- 99 portion of heart attack patients who had used oral contracep- tives in the previous months was significantly higher than ex- pected. The relative risk for myocardial infarction among “women smoking 25 or more cigarettes per day was 11.3 times greater than that among nonsmokers. Moreover, there was evi- dence for synergism of the two risks (23). Jick, et al. reported a case control study of 107 women under age 46 who were discharged from the hospital after suffering nonfatal, acute myocardial infarctions (15,16,17). The annual risk of nonfatal myocardial infarction (MI) among healthy women aged 39 to 45 who both smoked and used estrogens for noncontraceptive purposes was approximately 1 in 750. They noted that although an acute myocardial infarction is uncom- mon in healthy young women, the risk appears to be substantial in women over the age of 38 who both use estrogens and smoke cigarettes (17). In this same study, a relative risk of 14 was reported for oral contraceptive users compared with nonusers (90 percent confi- dence limits of relative risk from 5.5 to 37) (16). In women smok- ing more than 25 cigarettes per day the relative risk rose to 34 times that of women who were both nonusers and nonsmokers. While the number of subjects was small, the authors calculated that for women exposed to either oral contraceptives or smok- ing, but not both, the annual age-specific risks for nonfatal MI were roughly 1 per 190,000 at ages 27 to 37; 1 per 47,000 at ages 38 to 40; 1 per 23,000 at ages 40 to 43; and 1 per 16,000 at ages 44 and 45. If, however, both cigarettes and oral contraceptives are used, the annual age-specific risk is estimated to be much higher and the respective risks become 1 in 8,400; 1 in 920, 1 in 540, and 1 in 250. The authors report that a dose-response rela- tionship exists between smoking and risk among their popula- tion of female myocardial infarction patients, such that smok- ing 1 to 14 cigarettes per day carried a relative risk of nonfatal myocardial infarction of 9.2; 15 to 25 cigarettes of 7.9; and 26 or more cigarettes of 21, relative to those who never smoked (15). In another recent study of 234 pre-menopausal women.who had suffered a first myocardial infarction and 1,742 control pa- | tients drawn from the hospital population, Shapiro and his co- workers found an association between recent oral contraceptive. use and smoking (35). They found no evidence that past use of oral contraceptives was related to heart attack or that heightened risk was associated with increased duration of use of the oral contraceptives. For nonsmokers who used oral con- traceptives, the rate of myocardial infarction increased fourfold compared to nonusers and nonsmokers; in those women who smoked 25 or more cigarettes a day but did not use oral con- 100