- Email: [email protected]
Cigarette smoking and coronary heart disease in blacks: Comparison to whites in a prospective study
Cigarette smoking and coronary in blacks: Comparison to whites prospective study
heart disease in a
The American Cancer Society’s prospective study of 1 million Americans was analyzed to determine whether coronary heart disease (CHD) mortality rates by cigarette smoking in blacks differed from those in whites. More than 22,000 blacks were followed up for nearly 200,000 person-years in the 12-year study, 1960-1972. A total of 642 black males and 467 black females who died of CHD at age 40 years and older were included. CtlD mortality ratios by number of cigarettes smoked were about the same at given smoking levels in black and white males and were slightly lower in black than in white females. The combined effects of a history of high blood pressure and other cardiovascular disease with smoking elevated the risk of dying of CHD, particularly in persons under age 65 years at the start of the study. In black females under 65 years of age who smoked cigarettes and had a history of these diseases, the CHD rate was 7.3 times higher than in white females who did not smoke and had no history of these diseases. The effect of other smoking variables and of exercise and overweight on the differences in rates between blacks and whites is also discussed. (AM HEART J 106:602, 1964.)
Lawrence Garfinkel,
M.A. New York, N.Y.
Coronary heart disease (CHD) mortality is somewhat higher in black males than in white males through age 64 years, and in black females it is higher’through age 74 years. In older ages it is higher in whites than blacks. There has been a good deal of speculation about which of the risk factors for CHD may be responsible for the drop in mortality rates in both sexes and both races over the past 10 to 15 years.’ Gillum and Grant? have pointed to the need for prospective study data for blacks, particularly on smoking habits, blood pressure, and blood lipid levels, to shed light on the differences in risk factors for CHD. METHODS
The American Cancer Society (ACS) prospective study was conducted from 1959-1972. Sixty-eight thousand volunteers in 25 states collected more than 1 million confidential questionnaires from friends, relatives, and neighbors and traced them over a period of 13 years. The questionnaire contained a large number of questions on exposures, history of disease, and life-styles, including questions on smoking habits. Every year the volunteers reported on the status (alive or dead) of the study subjects From the Department Society.
of Epidemiology
Reprint requests: L. Garfinkel, tics, American Cancer Society,
802
Department 4 W. 35th
and
Statistics,
American
Cancer
of Epidemiology and StatisSt., New York, NY 10001.
and any changes of address. Death certificates were obtained from state health departments for subjects reported to have died. Every other year for the first 6 years of the study and again in 1972, the persons enrolled in the study completed short supplementary questionnaires that included questions on current smoking habits. More than 98% of the subjects originally enrolled in the study were traced after 12 years, with little or no difference in the tracing rates by age, sex, and race. Causes of death were obtained from death certificates coded by the Seventh Revision of the International Statistical Classification of Diseases, Injuries and Causes of Death (ICD-7). C&egories 420.0, 420.1 and 420.2 were used for CHD aa the underlying cause of death. Death certificates were coded by ACS clerks, and when there was doubt about the underlying cause of death, the certificate was reviewed by a supervisory nosologist. No attempt was made to write back to the certifying physician to verify the diagnosis, although we did ascertain the primary site and the basis of diagnosis for cancer cases. About 25,000 persons enrolled were black; 20,655, or 2.2 % , were age 40 years or older. This analysis covers the 12-year mortality from CHD among the blacks enrolled in the study and compares these data to the mortality experience of the white population in the study. Expected deaths are based on white CHD death rates by sex and 5-year age group at the start of the study. Since these cohort death rates are not comparable to the attained-age death rates published by national vital statistics sources, the actual death rates are not shown.
“Olwne Number
108 3, Part 2
Cigarette
smoking
and CHD in blacks
803
I. Observed and expected* deaths from CHD in blacks compared to whites by ageand sex
Table
Oh. deaths
Exp. deaths
Ratio
50-59 60-69 70-79 80+
65 198 189 152 38
99.8 245.0 242.7 188.8 49.3
0.65 0.81 0.78 0.81 0.77
(0.503, (0.700, (0.672, (0.682, (0.545,
Total
642
825.6
0.78
(0.719, 0.840)
40-49 50-59 60-69
63 127 150
25.8 76.7
2.44 1.66
(1.876, (1.380,
70-79
109
139.6 147.9
38 487
Age Sex
group
95% confid. limits
Male 40-49
0.830) 0.929) 0.898) 0.944) 1.058)
Female
80+
Total
1.07
(0.909, 1.261)
64.6
0.74 0.59
(0.605, (0.416,
0.889) 0.807)
459.6
1.07
(0.978,
1.171)
Obs. = Observed; Exp. = expected; Confid. = confidence. *Expected deaths are computed based on white CHD death rates by 5-year age-sex groups at the start of the study multiplied risk by the black population over the 12-year period of the study.
Table
3.124) 1.970)
by person-years
exposed to
II. Smoking habits* by sex and race Smoking habits
Never smoked Pipe or cigar Ex-cigarette
only
Current cigarette Less than 10 a day lo-19 a day 20-39 a day 40+ a day Other Total *Percentages
have been adjusted
White male (%)
White female (%)
21.9 11.4
68.5 0.5 5.4 25.1 6.4 7.9
17.0 45.3 5.0 8.8 26.1 5.4 4.4 100.0 for age on the age distribution
10.0 0.8 0.5
100.0 of the total study population
RESULTS
Six hundred forty-two black males and 487 black females 40 years of age and older died of CHD in the 12-year period, which included 197,826 person-years of exposure to risk. Compared to deaths of whites in similar &year age groups at the start of the study, the observed-to-expected mortality ratio in black males was 0.78; in black females, it was 1.07. The black/white mortality ratio by age group did not change much in males; in females it dropped considerably w.ith advancing age from 2.44 in females age 40 to 49 years at the start of the study to 0.59 in females 80 years of age and older (Table I). The smoking habits in this analysis were classified according to the data on the first questionnaire in i959 (Table II): About 22% of white males and 24% of the black males reported that they had never smoked, and more than 68% of white females and
Black male 0~) 23.9 15.3 8.2 47.3 14.2 15.4 16.7 1.0 5.3 100.0
Black female (5%) 72.3 1.5 3.7 20.0 10.9 5.5 3.5 0.1 2.5 100.0
at the start of the study.
72% of black females had never smoked. White males and white females who currently smoked cigarettes smoked more heavily than the blacks. Thirty-one percent of white males and 18 % of black males smoked 20 or more cigarettes a day, and 11% of white females vs 3.6% of black females smoked 20 or more cigarettes a day. White and black males were classified by their lifetime history of smoking, and the numbers of deaths from CHD were compared to the expected numbers, based on the rates by &year age groups of the males who had never smoked regularly. Pipe and cigar smokers who never smoked cigarettes had a mortality ratio (adjusted for age) 8% higher than males who never smoked, ex-cigarette smokers about 25% higher, and current cigarette smokers about 65% to 75% higher. There was little difference in these ratios in white and black males.
September,
804
American
Garfinkel
NEVER SMOKED
w NE4
w 0 I-S
w lo-Is" CIGARETTES
Fig.
w 2O+B
i-9
1984 Journal
20 +
IO-19 CIGARETTES
Heart
A DAY
Fig. 3. CHD mortality ratios for males by amount of
cigarette smoking per day in subjects with no history of
A DAY
seriousdisease,by race.
1. CHD mortality ratios for males by amount of
cigarette smoking per day, by race.
l-9
SMOKED
---+
IO-IS
20+-
CIGARETTES A DAY 2. CHD mortality ratios for females by amount of cigarette smoking per day, by race.
Fig.
The CHD ratio in white females who were excigarette smokers was 21% higher than in females who had never smoked, and the ratio was 70% higher in current smokers-about the same pattern as was seen in white males. The CHD ratio in black females who were ex-cigarette smokers was 80% higher (based on 12 expected deaths) and 50% higher among the current cigarette smokers. Fig. 1 shows the mortality ratios for CHD in current cigarette smokers for white and black males classified by the number of cigarettes smoked per day compared to males who never smoked. The mortality ratio in white males rose from 1.20 in males who smoked 1 to 9 cigarettes a day to 1.59 in those who smoked 10 to 19 a day and to 1.94 in those who smoked 20 or more cigarettes a day. In black males the increases were about the same: 1.48, 1.59, and 1.92, respectively. In females the mortality ratios for CHD rose more steeply by amount of smoking in whites than in
blacks (Fig. 2). In whites, mortality was 11% higher in smokers of 1 to 9 cigarettes a day, 82% higher in smokers of 10 to 19 a day, and 116% higher in smokers of 20 cigarettes or more a day. The black females’ mortality ratios were based on 43 deaths in the group that had smoked 1 to 9 cigarettes a day, 19 deaths in the group that had smoked 10 to 19 cigarettes a day, and 15 deaths in the group that had smoked 20 or more cigarettes a day. The differences in amount between smoking groups were not statistically significant (p > 0.05). Subjects in the study were classified according to their report of “serious disease”-a history of heart disease, stroke, or high blood pressure at the start of the study. Report of a history of high blood pressure was by far the most common of these three diseases reported. History of one or more of these diseases increased with age, with more blacks reporting a history of “serious disease” than whites of the same sex at each age group (Table III). Overall, 21% of black males and 17 % of white males, as well as 32 % of black females vs 20% of white females, reported such a history. Black males with a history of serious disease had a CHD rate 85% higher than black males without such a history; in black females the rate was 76% higher. Fig. 3 shows CHD mortality ratios in males with no history of “serious disease” by number of cigarettes smoked per day. These analyses were undertaken to investigate the effect of smoking not confounded by history of serious diseases. The death rates for these subjects are from 30% to 90% of the value of the CHD rates for all subjects, depending on age group. Mortality ratios of cigarette smokers among white males are slightly higher with each level of smoking than they are for ail white males regardless of history of disease (Fig. 1). The ratios for black males were lower than for all black males,
Volume Number
108 3, Part 2
Table
Cigarette smoking and CHD in blacks
III. Percent of subjects reporting history of serious disease* at start of study, by sex and race
Age group (yr)
White male (%)
40-49 50-59 60-69 70-79 tso+ Total
10.5 15.3 23.3 29.2 32.2 16.8
White
female f%)
Black
10.2 18.4 30.5 41.4 47.8 20.1
male (%o)
Black
13.1 18.6 31.2 36.9 46.6 21.3
female (“0) 21.4 32.4 45.8 51.0 51.0 32.2
*History of heart disease, stroke, or high blood pressure. Of those reporting one or more of these three diseases, high blood pressure was reported white males, 85% of white females, 83% of black males, and 92% of black females.
Table
805
in 62% of
IV. CHD mortality ratios in subjects classifiedby presenceor absenceof a history of seriousdiseaseand by
smoking
Sex
Age
(yr)
White - smoking -history
Black - smoking - history
Black - smoking + history
Black + smoking - history
Black + smoking + history
Male Total Under 65 65 and over
1.00 1.00 1.00
0.93 0.92 0.94
1.27 1.33 1.24
1.21* 1.39* 0.98
2.20* 3.57* 1.52*
Total Under 65 65 and over
1.00 1.00 1.00
1.17* 1.99* 0.83
1.72* 3.48* 1.26*
2.45* 3.68* 0.95
3.64* 7.26* 1.42
Female
*Low value of mortality
ratio of 95% confidence
limit is greater
than 1.00.
regardless of history, and lower than for white males at each level of smoking, but the differences were not statistically significant (p > 0.05). The CHD mortality ratios for white females without serious disease by amount of smoking is also higher than for all females (Fig. 4). The ratios for black females without serious disease are also somewhat higher than for all females regardless of history, rising to a ratio of 2.44 for smokers of 20 or more cigarettes a day. Table IV shows the individual and combined effects of cigarette smoking and history of serious disease in black males and females. Expected values are based on 5-year cohort death rates for white males and white females who did not smoke and had no history of serious disease. These rates were used, instead of black rates, for greater stability of expected values. An asterisk indicates that the low value of the 95 % confidence limits for the mortality ratio was greater than 1.00. Compared with white males without serious disease who never smoked, black males who did not smoke and had no serious disease had a CHD mortality ratio of 0.93. Black males who did not smoke but had a history of serious disease had a rate 27% higher than comparable white males with no history. The rate for black males who smoked but had
no serious disease was 21% higher, but those who both smoked and reported serious disease had a mortality ratio of 2.20. Most of the effect of smoking and of history of disease was in black males under the age of 65 years. Among black nonsmoking males, the mortality ratios were about the same in those under and over the age of 65 years at the start of the study. In black males who smoked but had no history of serious disease, however, the entire effect was in those under age 65 years-mortality ratio of 1.39 in males under age 65 and 0.98 in males age 65 years and older. In black males who both smoked and had a history of serious disease, the mortality ratio rose to 3.57 compared to whites who neither smoked nor had a history of serious disease and was 1.52 in those age 65 years and older. In black females the differences were even greater. Compared to white females with no history of serious disease who never smoked, black females with similar characteristics had a CHD rate 17% higher. Black females who did not smoke but had a history of serious disease had a CHD rate 72% higher; the CHD rate for those who smoked but had no history of serious disease was 145% higher, and for those with serious disease who also smoked it was 264% higher. As with black males, most of this
September,
806
Garfinkel
American
rette smokers were smoking
CiSARETTES
A DAY
Fig. 4. CHD mortality ratios for females by amount of cigarette smoking per day in subjects with no history of seriousdisease,by race.
effect, was in younger black females. The CHD mortality ratio in black nonsmoking females under the age of 65 years with no history of serious disease was twice that of comparable white females. In younger black females who either smoked but had no serious disease or who had serious disease but were nonsmokers, mortality ratios were about 3.5. Black females under age 65 years who both smoked and had a serious disease had a mortality ratio of 7.26. CHD mortality ratios in black females over age 65 years were very similar to those of black males of the same ages. DISCUSSION
There are other smoking variables, in addition to the number of cigarettes smoked per day, that can affect, the CHD rate. Males in the ACS study were classified by the number of cigarettes smoked per day at the start of the study. Over the course of the study, large numbers of cigarette smokers quit smoking. A previous analysis showed that the CHD mortality ratios in former smokers of 20 or more cigarettes a day decreased steadily according to the number of years they had quit smoking. After 10 years, it dropped to only 25% greater than in males who had never smoked, and after 20 years the rate was about the same as for males who had never smoked.3 In a matched-groups analysis, males and females who had switched to cigarettes with relatively low tar and nicotine content had a death rate from CHD that was about 14% lower overall than did those still smoking cigarettes with relatively high tar and nicotine.4 The period of this study (1960-1972) was a time in which smoking habits were changing rapidly, with large numbers of people quitting smoking and continuing smokers switching to cigarettes with lower tar and nicotine. By 1982, more than 65% of ciga-
1984
Heart Journal
cigarettes with low tar and nicotine-classified as those having less than 15 mg tar and 0.8 mg nicotine. About 49% of the males and 33% of the females who smoked cigarettes at the start of the study reported on the 1972 questionnaire that they had quit, and this factor, together with the switch to the cigarettes with lower tar and nicotine, must have had some effect on CHD mortality rates. However, when Kleinman et aL5 attempted to determine how much of the decrease in CHD mortality could be attributed to changes in the number of smokers and in the number of cigarettes smoked per day, they could not account for a substantial part of the decline. Although blood lipids were not determined in this study, other risk factors for CHD were explored in another analysis of this data base. CHD rates in males and females reporting no exercise were higher than those reporting heavy exercise. The degree of the relationship rose with age, and in the age group of persons 70 to 79 years old, it was 2.85 in males and 3.53 in females who said they took no exercise, compared to those with heavy exercise.3 Since only 4% of blacks reported that they took no exercise, a separate analysis was not made. It should be noted that 24% of black males and 17% of black females reported heavy exercise in work or play (compared to 15% and 9% of whites, respectively). This should tend to make CHD rates in blacks relatively lower. Males who were 40% or more overweight, compared to males of the same 5-year age group and the same height, had a death rate from CHD 95 % higher than males of average weight.‘j The comparable mortality ratio in females 40% overweight was 2.07. This risk was only slightly increased in males who smoked 20 or more cigarettes a day compared to nonsmokers of average weight (mortality ratio, 2.29). Since 5% of black males and 4% of white males were 30% or more overweight, increased mortality associated with overweight cannot account for a large portion of the differences in CHD rates in black and white males. In obese females (40% or more overweight) who smoked 20 or more cigarettes a day, however, the CHD mortality rate rose to 7.18 times the rate for females of average weight who never smoked, and this could be an important factor because 17% of black females and 4% of white females were 30% or more overweight. We did not investigate the relationship of smoking habits in black males and females, controlled for degree of exercise and obesity, because the numbers of deaths were too small to be meaningful. The black/white ratio for CHD deaths shown in Table I shows a negative relationship, rather than
Volume Number
108 3, Part 2
Cigarette smoking and CHD in blacks
the positive relationship seen in U.S. national data. It is not clear why the relationship is reversed. The mortality ratios in blacks are based on large-enough numbers (as shown by the confidence limits) so that the discrepancy cannot be explained by sampling variations. It may be that the blacks in this study were of a higher socioeconomic class than in the rest of the country, or they may differ with respect to other risk factors. In general, the persons enrolled in the study had higher educational levels than shown in census figures. Illiterates, migrant and institutionalized populations, and persons living in inner cities were underrepresented in this study. Nancy LaVerda assisted in the processing of the data in this study.
REFERENCES 1.
2. 3. 4. 5. 6.
Havlik RJ, Feinleib M, editors: Proceedings of the Conference on the Decline in Coronary Heart Disease Mortality. National Institutes of Health, May 1979. Gillum RF, Grant CT: Coronary heart disease in black nonulations. II. Risk factors. AM HEARTJ lD4:839. 1982. Hammond EC, Garfinkel L: Coronary heart disease, stroke, and aortic aneurysm: Factors in the etiology. Arch Environ Health 19:167, 1969. Hammond EC, Garfinkel L, Seidman H, Lew EA: “Tar” and nicotine content of cigarette smoke in relation to death rates. Environ Res 12:263, 1976. Kleinman JC, Feldman JJ, Monk MA: The effects of changes in smoking habits on coronary heart disease mortality. Am J Public He-&h 69:795, 1979. Lew EA, Garfinkel L: Variations in mortality by weight among 750,000 men and women. J Chronic Dis 32:563, 1979.
Black-white differences in plasma levels of apolipoproteins: The Evans County Heart Study Evans County black males had lower ischemic heart disease (IHD) prevalence, incidence, and mortality than white males. High-density lipoprotein (HDL) cholesterol was lower in IHD cases than in subjects without IHD. HDL cholesterol and apoiipoprotein A-i (Apo A-i) were higher and low-density lipoprotein (LDL) cholesterol, very low-density lipoprotein (VLDL) cholesterol, and Apo C-ii were lower in black than white males. Of the black-white male HDL cholesterol difference, 22% was statistically explained by Apo A-i. Controlling for Apo C-ii reduced the black-white differences in total cholesterol 87%, LDL cholesterol 44%, VLDL cholesterol 83%, and total triglyceride 83%. There were negative associations between Apo A-i and age, Queteiet index, and cigarettes smoked; the association between Apo A-i and alcohol was positive. Only body mass index and race were strong correlates of Apo C-ii. The ratios of Apo A-i to Apo A-ii and of HDL cholesterol to Apo A-ii were htgher in black than white males with adjustment for age, body mass, and cigarette and alcohol consumption. Thus black-white differences in total lipids, lipoprotein lipids, and lipoprotein apoproteins were observed, indicating a relatively antiatherogenic profile in black males only partially explained by known correlates. (AM HEART J 108:807, 1984.)
Gerard0 Heiss, M.D., Gustav Schonfeld, M.D., Jeffrey L. Johnson, M.S., Siegfried Heyden, M.D., Curtis G. Hames, M.D., and H. A. Tyroler, M.D. Chapel Hill, and Durham, N.C., St. Louis, MO., and Claxton, Ga.
From the University of North Carolina School of Public Health, Department of Epidemiology, Chapel Hill, the Lipid Research Clinic, Washington University School of Medicine, St. Louis, the Duke University Epidemiology Department, Durham, and the Evans County Heart Study, Hames’ Clinic, Claxton. Reprint Patient
requests: Registry,
Gerard0 Heiss, M.D., Lipid Research Clinic, NCNB Plaza 322A, Chapel Hill, NC 27514.
Central
For the goals of this symposium we have summarized and integrated two bodies of information: (1) a series of empirical studies of lipids, lipoproteins, and apolipoproteins performed among black and white participants in the Evans County Heart Study 807
heart disease in a
The American Cancer Society’s prospective study of 1 million Americans was analyzed to determine whether coronary heart disease (CHD) mortality rates by cigarette smoking in blacks differed from those in whites. More than 22,000 blacks were followed up for nearly 200,000 person-years in the 12-year study, 1960-1972. A total of 642 black males and 467 black females who died of CHD at age 40 years and older were included. CtlD mortality ratios by number of cigarettes smoked were about the same at given smoking levels in black and white males and were slightly lower in black than in white females. The combined effects of a history of high blood pressure and other cardiovascular disease with smoking elevated the risk of dying of CHD, particularly in persons under age 65 years at the start of the study. In black females under 65 years of age who smoked cigarettes and had a history of these diseases, the CHD rate was 7.3 times higher than in white females who did not smoke and had no history of these diseases. The effect of other smoking variables and of exercise and overweight on the differences in rates between blacks and whites is also discussed. (AM HEART J 106:602, 1964.)
Lawrence Garfinkel,
M.A. New York, N.Y.
Coronary heart disease (CHD) mortality is somewhat higher in black males than in white males through age 64 years, and in black females it is higher’through age 74 years. In older ages it is higher in whites than blacks. There has been a good deal of speculation about which of the risk factors for CHD may be responsible for the drop in mortality rates in both sexes and both races over the past 10 to 15 years.’ Gillum and Grant? have pointed to the need for prospective study data for blacks, particularly on smoking habits, blood pressure, and blood lipid levels, to shed light on the differences in risk factors for CHD. METHODS
The American Cancer Society (ACS) prospective study was conducted from 1959-1972. Sixty-eight thousand volunteers in 25 states collected more than 1 million confidential questionnaires from friends, relatives, and neighbors and traced them over a period of 13 years. The questionnaire contained a large number of questions on exposures, history of disease, and life-styles, including questions on smoking habits. Every year the volunteers reported on the status (alive or dead) of the study subjects From the Department Society.
of Epidemiology
Reprint requests: L. Garfinkel, tics, American Cancer Society,
802
Department 4 W. 35th
and
Statistics,
American
Cancer
of Epidemiology and StatisSt., New York, NY 10001.
and any changes of address. Death certificates were obtained from state health departments for subjects reported to have died. Every other year for the first 6 years of the study and again in 1972, the persons enrolled in the study completed short supplementary questionnaires that included questions on current smoking habits. More than 98% of the subjects originally enrolled in the study were traced after 12 years, with little or no difference in the tracing rates by age, sex, and race. Causes of death were obtained from death certificates coded by the Seventh Revision of the International Statistical Classification of Diseases, Injuries and Causes of Death (ICD-7). C&egories 420.0, 420.1 and 420.2 were used for CHD aa the underlying cause of death. Death certificates were coded by ACS clerks, and when there was doubt about the underlying cause of death, the certificate was reviewed by a supervisory nosologist. No attempt was made to write back to the certifying physician to verify the diagnosis, although we did ascertain the primary site and the basis of diagnosis for cancer cases. About 25,000 persons enrolled were black; 20,655, or 2.2 % , were age 40 years or older. This analysis covers the 12-year mortality from CHD among the blacks enrolled in the study and compares these data to the mortality experience of the white population in the study. Expected deaths are based on white CHD death rates by sex and 5-year age group at the start of the study. Since these cohort death rates are not comparable to the attained-age death rates published by national vital statistics sources, the actual death rates are not shown.
“Olwne Number
108 3, Part 2
Cigarette
smoking
and CHD in blacks
803
I. Observed and expected* deaths from CHD in blacks compared to whites by ageand sex
Table
Oh. deaths
Exp. deaths
Ratio
50-59 60-69 70-79 80+
65 198 189 152 38
99.8 245.0 242.7 188.8 49.3
0.65 0.81 0.78 0.81 0.77
(0.503, (0.700, (0.672, (0.682, (0.545,
Total
642
825.6
0.78
(0.719, 0.840)
40-49 50-59 60-69
63 127 150
25.8 76.7
2.44 1.66
(1.876, (1.380,
70-79
109
139.6 147.9
38 487
Age Sex
group
95% confid. limits
Male 40-49
0.830) 0.929) 0.898) 0.944) 1.058)
Female
80+
Total
1.07
(0.909, 1.261)
64.6
0.74 0.59
(0.605, (0.416,
0.889) 0.807)
459.6
1.07
(0.978,
1.171)
Obs. = Observed; Exp. = expected; Confid. = confidence. *Expected deaths are computed based on white CHD death rates by 5-year age-sex groups at the start of the study multiplied risk by the black population over the 12-year period of the study.
Table
3.124) 1.970)
by person-years
exposed to
II. Smoking habits* by sex and race Smoking habits
Never smoked Pipe or cigar Ex-cigarette
only
Current cigarette Less than 10 a day lo-19 a day 20-39 a day 40+ a day Other Total *Percentages
have been adjusted
White male (%)
White female (%)
21.9 11.4
68.5 0.5 5.4 25.1 6.4 7.9
17.0 45.3 5.0 8.8 26.1 5.4 4.4 100.0 for age on the age distribution
10.0 0.8 0.5
100.0 of the total study population
RESULTS
Six hundred forty-two black males and 487 black females 40 years of age and older died of CHD in the 12-year period, which included 197,826 person-years of exposure to risk. Compared to deaths of whites in similar &year age groups at the start of the study, the observed-to-expected mortality ratio in black males was 0.78; in black females, it was 1.07. The black/white mortality ratio by age group did not change much in males; in females it dropped considerably w.ith advancing age from 2.44 in females age 40 to 49 years at the start of the study to 0.59 in females 80 years of age and older (Table I). The smoking habits in this analysis were classified according to the data on the first questionnaire in i959 (Table II): About 22% of white males and 24% of the black males reported that they had never smoked, and more than 68% of white females and
Black male 0~) 23.9 15.3 8.2 47.3 14.2 15.4 16.7 1.0 5.3 100.0
Black female (5%) 72.3 1.5 3.7 20.0 10.9 5.5 3.5 0.1 2.5 100.0
at the start of the study.
72% of black females had never smoked. White males and white females who currently smoked cigarettes smoked more heavily than the blacks. Thirty-one percent of white males and 18 % of black males smoked 20 or more cigarettes a day, and 11% of white females vs 3.6% of black females smoked 20 or more cigarettes a day. White and black males were classified by their lifetime history of smoking, and the numbers of deaths from CHD were compared to the expected numbers, based on the rates by &year age groups of the males who had never smoked regularly. Pipe and cigar smokers who never smoked cigarettes had a mortality ratio (adjusted for age) 8% higher than males who never smoked, ex-cigarette smokers about 25% higher, and current cigarette smokers about 65% to 75% higher. There was little difference in these ratios in white and black males.
September,
804
American
Garfinkel
NEVER SMOKED
w NE4
w 0 I-S
w lo-Is" CIGARETTES
Fig.
w 2O+B
i-9
1984 Journal
20 +
IO-19 CIGARETTES
Heart
A DAY
Fig. 3. CHD mortality ratios for males by amount of
cigarette smoking per day in subjects with no history of
A DAY
seriousdisease,by race.
1. CHD mortality ratios for males by amount of
cigarette smoking per day, by race.
l-9
SMOKED
---+
IO-IS
20+-
CIGARETTES A DAY 2. CHD mortality ratios for females by amount of cigarette smoking per day, by race.
Fig.
The CHD ratio in white females who were excigarette smokers was 21% higher than in females who had never smoked, and the ratio was 70% higher in current smokers-about the same pattern as was seen in white males. The CHD ratio in black females who were ex-cigarette smokers was 80% higher (based on 12 expected deaths) and 50% higher among the current cigarette smokers. Fig. 1 shows the mortality ratios for CHD in current cigarette smokers for white and black males classified by the number of cigarettes smoked per day compared to males who never smoked. The mortality ratio in white males rose from 1.20 in males who smoked 1 to 9 cigarettes a day to 1.59 in those who smoked 10 to 19 a day and to 1.94 in those who smoked 20 or more cigarettes a day. In black males the increases were about the same: 1.48, 1.59, and 1.92, respectively. In females the mortality ratios for CHD rose more steeply by amount of smoking in whites than in
blacks (Fig. 2). In whites, mortality was 11% higher in smokers of 1 to 9 cigarettes a day, 82% higher in smokers of 10 to 19 a day, and 116% higher in smokers of 20 cigarettes or more a day. The black females’ mortality ratios were based on 43 deaths in the group that had smoked 1 to 9 cigarettes a day, 19 deaths in the group that had smoked 10 to 19 cigarettes a day, and 15 deaths in the group that had smoked 20 or more cigarettes a day. The differences in amount between smoking groups were not statistically significant (p > 0.05). Subjects in the study were classified according to their report of “serious disease”-a history of heart disease, stroke, or high blood pressure at the start of the study. Report of a history of high blood pressure was by far the most common of these three diseases reported. History of one or more of these diseases increased with age, with more blacks reporting a history of “serious disease” than whites of the same sex at each age group (Table III). Overall, 21% of black males and 17 % of white males, as well as 32 % of black females vs 20% of white females, reported such a history. Black males with a history of serious disease had a CHD rate 85% higher than black males without such a history; in black females the rate was 76% higher. Fig. 3 shows CHD mortality ratios in males with no history of “serious disease” by number of cigarettes smoked per day. These analyses were undertaken to investigate the effect of smoking not confounded by history of serious diseases. The death rates for these subjects are from 30% to 90% of the value of the CHD rates for all subjects, depending on age group. Mortality ratios of cigarette smokers among white males are slightly higher with each level of smoking than they are for ail white males regardless of history of disease (Fig. 1). The ratios for black males were lower than for all black males,
Volume Number
108 3, Part 2
Table
Cigarette smoking and CHD in blacks
III. Percent of subjects reporting history of serious disease* at start of study, by sex and race
Age group (yr)
White male (%)
40-49 50-59 60-69 70-79 tso+ Total
10.5 15.3 23.3 29.2 32.2 16.8
White
female f%)
Black
10.2 18.4 30.5 41.4 47.8 20.1
male (%o)
Black
13.1 18.6 31.2 36.9 46.6 21.3
female (“0) 21.4 32.4 45.8 51.0 51.0 32.2
*History of heart disease, stroke, or high blood pressure. Of those reporting one or more of these three diseases, high blood pressure was reported white males, 85% of white females, 83% of black males, and 92% of black females.
Table
805
in 62% of
IV. CHD mortality ratios in subjects classifiedby presenceor absenceof a history of seriousdiseaseand by
smoking
Sex
Age
(yr)
White - smoking -history
Black - smoking - history
Black - smoking + history
Black + smoking - history
Black + smoking + history
Male Total Under 65 65 and over
1.00 1.00 1.00
0.93 0.92 0.94
1.27 1.33 1.24
1.21* 1.39* 0.98
2.20* 3.57* 1.52*
Total Under 65 65 and over
1.00 1.00 1.00
1.17* 1.99* 0.83
1.72* 3.48* 1.26*
2.45* 3.68* 0.95
3.64* 7.26* 1.42
Female
*Low value of mortality
ratio of 95% confidence
limit is greater
than 1.00.
regardless of history, and lower than for white males at each level of smoking, but the differences were not statistically significant (p > 0.05). The CHD mortality ratios for white females without serious disease by amount of smoking is also higher than for all females (Fig. 4). The ratios for black females without serious disease are also somewhat higher than for all females regardless of history, rising to a ratio of 2.44 for smokers of 20 or more cigarettes a day. Table IV shows the individual and combined effects of cigarette smoking and history of serious disease in black males and females. Expected values are based on 5-year cohort death rates for white males and white females who did not smoke and had no history of serious disease. These rates were used, instead of black rates, for greater stability of expected values. An asterisk indicates that the low value of the 95 % confidence limits for the mortality ratio was greater than 1.00. Compared with white males without serious disease who never smoked, black males who did not smoke and had no serious disease had a CHD mortality ratio of 0.93. Black males who did not smoke but had a history of serious disease had a rate 27% higher than comparable white males with no history. The rate for black males who smoked but had
no serious disease was 21% higher, but those who both smoked and reported serious disease had a mortality ratio of 2.20. Most of the effect of smoking and of history of disease was in black males under the age of 65 years. Among black nonsmoking males, the mortality ratios were about the same in those under and over the age of 65 years at the start of the study. In black males who smoked but had no history of serious disease, however, the entire effect was in those under age 65 years-mortality ratio of 1.39 in males under age 65 and 0.98 in males age 65 years and older. In black males who both smoked and had a history of serious disease, the mortality ratio rose to 3.57 compared to whites who neither smoked nor had a history of serious disease and was 1.52 in those age 65 years and older. In black females the differences were even greater. Compared to white females with no history of serious disease who never smoked, black females with similar characteristics had a CHD rate 17% higher. Black females who did not smoke but had a history of serious disease had a CHD rate 72% higher; the CHD rate for those who smoked but had no history of serious disease was 145% higher, and for those with serious disease who also smoked it was 264% higher. As with black males, most of this
September,
806
Garfinkel
American
rette smokers were smoking
CiSARETTES
A DAY
Fig. 4. CHD mortality ratios for females by amount of cigarette smoking per day in subjects with no history of seriousdisease,by race.
effect, was in younger black females. The CHD mortality ratio in black nonsmoking females under the age of 65 years with no history of serious disease was twice that of comparable white females. In younger black females who either smoked but had no serious disease or who had serious disease but were nonsmokers, mortality ratios were about 3.5. Black females under age 65 years who both smoked and had a serious disease had a mortality ratio of 7.26. CHD mortality ratios in black females over age 65 years were very similar to those of black males of the same ages. DISCUSSION
There are other smoking variables, in addition to the number of cigarettes smoked per day, that can affect, the CHD rate. Males in the ACS study were classified by the number of cigarettes smoked per day at the start of the study. Over the course of the study, large numbers of cigarette smokers quit smoking. A previous analysis showed that the CHD mortality ratios in former smokers of 20 or more cigarettes a day decreased steadily according to the number of years they had quit smoking. After 10 years, it dropped to only 25% greater than in males who had never smoked, and after 20 years the rate was about the same as for males who had never smoked.3 In a matched-groups analysis, males and females who had switched to cigarettes with relatively low tar and nicotine content had a death rate from CHD that was about 14% lower overall than did those still smoking cigarettes with relatively high tar and nicotine.4 The period of this study (1960-1972) was a time in which smoking habits were changing rapidly, with large numbers of people quitting smoking and continuing smokers switching to cigarettes with lower tar and nicotine. By 1982, more than 65% of ciga-
1984
Heart Journal
cigarettes with low tar and nicotine-classified as those having less than 15 mg tar and 0.8 mg nicotine. About 49% of the males and 33% of the females who smoked cigarettes at the start of the study reported on the 1972 questionnaire that they had quit, and this factor, together with the switch to the cigarettes with lower tar and nicotine, must have had some effect on CHD mortality rates. However, when Kleinman et aL5 attempted to determine how much of the decrease in CHD mortality could be attributed to changes in the number of smokers and in the number of cigarettes smoked per day, they could not account for a substantial part of the decline. Although blood lipids were not determined in this study, other risk factors for CHD were explored in another analysis of this data base. CHD rates in males and females reporting no exercise were higher than those reporting heavy exercise. The degree of the relationship rose with age, and in the age group of persons 70 to 79 years old, it was 2.85 in males and 3.53 in females who said they took no exercise, compared to those with heavy exercise.3 Since only 4% of blacks reported that they took no exercise, a separate analysis was not made. It should be noted that 24% of black males and 17% of black females reported heavy exercise in work or play (compared to 15% and 9% of whites, respectively). This should tend to make CHD rates in blacks relatively lower. Males who were 40% or more overweight, compared to males of the same 5-year age group and the same height, had a death rate from CHD 95 % higher than males of average weight.‘j The comparable mortality ratio in females 40% overweight was 2.07. This risk was only slightly increased in males who smoked 20 or more cigarettes a day compared to nonsmokers of average weight (mortality ratio, 2.29). Since 5% of black males and 4% of white males were 30% or more overweight, increased mortality associated with overweight cannot account for a large portion of the differences in CHD rates in black and white males. In obese females (40% or more overweight) who smoked 20 or more cigarettes a day, however, the CHD mortality rate rose to 7.18 times the rate for females of average weight who never smoked, and this could be an important factor because 17% of black females and 4% of white females were 30% or more overweight. We did not investigate the relationship of smoking habits in black males and females, controlled for degree of exercise and obesity, because the numbers of deaths were too small to be meaningful. The black/white ratio for CHD deaths shown in Table I shows a negative relationship, rather than
Volume Number
108 3, Part 2
Cigarette smoking and CHD in blacks
the positive relationship seen in U.S. national data. It is not clear why the relationship is reversed. The mortality ratios in blacks are based on large-enough numbers (as shown by the confidence limits) so that the discrepancy cannot be explained by sampling variations. It may be that the blacks in this study were of a higher socioeconomic class than in the rest of the country, or they may differ with respect to other risk factors. In general, the persons enrolled in the study had higher educational levels than shown in census figures. Illiterates, migrant and institutionalized populations, and persons living in inner cities were underrepresented in this study. Nancy LaVerda assisted in the processing of the data in this study.
REFERENCES 1.
2. 3. 4. 5. 6.
Havlik RJ, Feinleib M, editors: Proceedings of the Conference on the Decline in Coronary Heart Disease Mortality. National Institutes of Health, May 1979. Gillum RF, Grant CT: Coronary heart disease in black nonulations. II. Risk factors. AM HEARTJ lD4:839. 1982. Hammond EC, Garfinkel L: Coronary heart disease, stroke, and aortic aneurysm: Factors in the etiology. Arch Environ Health 19:167, 1969. Hammond EC, Garfinkel L, Seidman H, Lew EA: “Tar” and nicotine content of cigarette smoke in relation to death rates. Environ Res 12:263, 1976. Kleinman JC, Feldman JJ, Monk MA: The effects of changes in smoking habits on coronary heart disease mortality. Am J Public He-&h 69:795, 1979. Lew EA, Garfinkel L: Variations in mortality by weight among 750,000 men and women. J Chronic Dis 32:563, 1979.
Black-white differences in plasma levels of apolipoproteins: The Evans County Heart Study Evans County black males had lower ischemic heart disease (IHD) prevalence, incidence, and mortality than white males. High-density lipoprotein (HDL) cholesterol was lower in IHD cases than in subjects without IHD. HDL cholesterol and apoiipoprotein A-i (Apo A-i) were higher and low-density lipoprotein (LDL) cholesterol, very low-density lipoprotein (VLDL) cholesterol, and Apo C-ii were lower in black than white males. Of the black-white male HDL cholesterol difference, 22% was statistically explained by Apo A-i. Controlling for Apo C-ii reduced the black-white differences in total cholesterol 87%, LDL cholesterol 44%, VLDL cholesterol 83%, and total triglyceride 83%. There were negative associations between Apo A-i and age, Queteiet index, and cigarettes smoked; the association between Apo A-i and alcohol was positive. Only body mass index and race were strong correlates of Apo C-ii. The ratios of Apo A-i to Apo A-ii and of HDL cholesterol to Apo A-ii were htgher in black than white males with adjustment for age, body mass, and cigarette and alcohol consumption. Thus black-white differences in total lipids, lipoprotein lipids, and lipoprotein apoproteins were observed, indicating a relatively antiatherogenic profile in black males only partially explained by known correlates. (AM HEART J 108:807, 1984.)
Gerard0 Heiss, M.D., Gustav Schonfeld, M.D., Jeffrey L. Johnson, M.S., Siegfried Heyden, M.D., Curtis G. Hames, M.D., and H. A. Tyroler, M.D. Chapel Hill, and Durham, N.C., St. Louis, MO., and Claxton, Ga.
From the University of North Carolina School of Public Health, Department of Epidemiology, Chapel Hill, the Lipid Research Clinic, Washington University School of Medicine, St. Louis, the Duke University Epidemiology Department, Durham, and the Evans County Heart Study, Hames’ Clinic, Claxton. Reprint Patient
requests: Registry,
Gerard0 Heiss, M.D., Lipid Research Clinic, NCNB Plaza 322A, Chapel Hill, NC 27514.
Central
For the goals of this symposium we have summarized and integrated two bodies of information: (1) a series of empirical studies of lipids, lipoproteins, and apolipoproteins performed among black and white participants in the Evans County Heart Study 807