The New England Journal of Medicine
e-mail icon  FREE NEJM E-TOC    HOME   |   SUBSCRIBE   |   CURRENT ISSUE   |   PAST ISSUES   |   COLLECTIONS   |    Advanced Search
Sign in | Get NEJM's E-Mail Table of Contents — Free | Subscribe
 
Original Article
Volume 329:73-78 July 8, 1993 Number 2
NextNext

Mortality Rates and Risk Factors for Coronary Disease in Black as Compared with White Men and Women
Julian E. Keil, Susan E. Sutherland, Rebecca G. Knapp, Daniel T. Lackland, Peter C. Gazes, and Herman A. Tyroler

 

This Article
-Abstract

Commentary
-Letters

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation
ABSTRACT

Background Currently recognized risk factors for coronary artery disease have been identified primarily from investigations of white populations. In this investigation, we estimated mortality rates for coronary disease and for any cause and identified risk factors for death from coronary disease among whites and blacks.

Methods Data collected over a 30-year period in the Charleston Heart Study were used to estimate mortality rates and quantify associations with risk factors assessed at the base-line examination in 1960 and 1961 of 653 white men, 333 black men, 741 white women, and 454 black women.

Results There were no significant racial differences in the rate ratios for death from coronary disease; however, women had significantly lower death rates than men. Over the 30-year period, the mortality rates for coronary disease per 1000 person-years were 5.2 for white men (95 percent confidence interval, 4.1 to 6.3), 4.6 for black men (3.0 to 6.2), 2.1 for white women (1.6 to 2.6), and 3.2 for black women (2.3 to 4.0). Significant, or nearly significant, predictors of mortality due to coronary disease were systolic blood pressure in all four groups; serum cholesterol level among white men, white women, and black women; and smoking among white men, white women, and black men. Although the difference was not statistically significant, the risk of death from coronary disease was consistently increased among diabetics in all four groups. A higher level of education was predictive of lower rates of death due to coronary disease among white men and black women. For all causes of death taken together, the rates for blacks were higher than the rates for whites. The presence of hypertension, a history of smoking, and a history of diabetes were significant or nearly significant predictors of mortality from any cause in all four groups.

Conclusions Although the rates of death from coronary disease were somewhat lower among black men than white men and higher among black women than white women, the black:white mortality rate ratios were not statistically significant, and the major risk factors for mortality from coronary disease were similar in blacks and whites in the 30-year follow-up of the Charleston Heart Study.


There has been lingering concern in the public health community that there may be racial differences in risk factors for coronary disease. The current recommendations for risk-factor modification were derived from numerous observational studies that followed the pioneering cardiovascular cohort studies conducted in Framingham, Massachusetts1,2. Although the results of these studies were generally consistent in identifying risk factors for cardiovascular diseases in racially or ethnically diverse populations, there was a paucity of community-based investigations in black populations.

Early reports suggested that the risks of cardiovascular diseases differed between blacks and whites: blacks were protected against coronary heart disease but had a much greater risk of stroke3,4. This paradox raised the question of why blacks did not have higher rates of coronary disease than whites, given their excess of hypertension5,6. More than 10 years ago Gillum7 suggested that the myth of blacks' immunity to coronary disease was at best a half-truth and at worst an obstacle to the understanding of coronary disease and to treatment for blacks. A similar concern led to the Charleston Heart Study8 and the Evans County Heart Study6 in 1960. Although these studies evaluated blood pressure, cholesterol level, smoking status, and other risk factors for coronary disease, they postulated that there were racial and social bases for the risk factors for coronary disease. Earlier findings9 indicated that there were nonsignificant differences in the rates of death from coronary disease between black men and white men in Charleston County, South Carolina. However, vital statistics and studies from Evans County, Georgia, indicated that the mortality rates for coronary disease were higher among blacks at younger ages but lower at advanced ages, resulting in lower rates of mortality due to coronary disease with prolonged periods of follow-up and advancing age.

The purpose of this study was to compare the 30-year coronary mortality experience of blacks and whites in the Charleston Heart Study and to examine the predictors of coronary disease mortality in this cohort. Since the certification of causes of death may be less reliable for older people, we also examined the rates and predictors of death from any cause.

Methods

The Charleston Heart Study, a population-based prospective cohort study begun in 1960, contained a random sample of blacks and whites 35 years of age and older. The study cohort consisted of 2181 subjects and represented an overall response rate of 84 percent to the population-based sampling plan. Details of the sampling procedure and other procedures have been published previously8,9. Except for Table 1, all the tables in this paper show data on participants 35 to 74 years of age.

View this table:
[in this window]
[in a new window]
 
Table 1. Base-Line Characteristics and Vital Status of the Charleston Heart Study Cohort.

 
The methods of measurement of base-line (1960 and 1961) levels of most variables have been reported previously8. Variables measured at base line included the first blood-pressure measurement taken in the seated position, serum cholesterol level determined by the direct-cholesterol method of Zlatkis et al.,10 body-mass index (the weight in kilograms divided by the square of the height in meters), years of education, cigarette-smoking status (dichotomized as current vs. never or past), and diabetes status ascertained on the basis of medical history. The number of years of education was used as a surrogate for socioeconomic status.

The causes of death were derived from nosologists' coding of the underlying causes of death. Death certificates for the period from 1960 through 1967 (which used codes from the seventh edition of the International Classification of Diseases [ICD]) were recoded according to the eighth edition of the ICD. The ICD codes used for coronary disease were 410 through 413 for the period from 1960 through 1978 and 410 through 414 (ICD, ninth edition) for the period from 1979 through 1990.

The prevalence of coronary heart disease at base line was about 3 percent among black men, black women, and white women and 8 percent among white men. Persons who were lost to follow-up were not excluded from the analyses, but the data on these subjects were censored at the time of the last contact. Mortality rates for coronary disease or for any cause were calculated with the SAS program Personyrs11 and expressed as age-adjusted rates per 1000 person-years of observation. Survival curves for coronary heart disease were plotted according to race and sex for the 30-year period by the Kaplan-Meier method12,13. The overall survival curves were compared for black men and white men and for black women and white women with the log-rank test. The black:white mortality ratios were derived from Cox proportional-hazards regression analyses, first by including only age and race in the model and then by including age, race, and the cardiovascular risk factors. The proportional-hazards model was used to identify predictors of survival with respect to death from coronary disease and all causes14. The putative risk factors in the regression analyses were age, systolic blood pressure, cholesterol level, body-mass index, smoking status, and diabetes status. The assumptions (of proportional hazard over time) made in the proportional-hazards regression analyses were met for race in men and women according to goodness-of-fit tests and according to visual comparison in men. The relative risk for continuous variables was calculated on the basis of an increase of 1 SD in the values. Participants with coronary heart disease at base line were included; the exclusion of such patients at base line made essentially no difference in the estimates of relative risk. All P values were calculated from two-tailed tests.

Results

Characteristics at Base Line and Vital Status in 1990

The base-line characteristics (1960 and 1961) and the vital status (1960 through 1990) of the study participants are shown in Table 1. By the end of 1990, the vital status of 98 percent of the white participants and 99 percent of the black participants was known.

Comparison of Mortality between Blacks and Whites

Table 2 shows the age-adjusted mortality rates for coronary heart disease and for any cause of death in the four groups defined by race and sex. After 30 years of observation, the black:white coronary mortality rate ratio was 0.9 (P = 0.33) for men and 1.2 (P = 0.20) for women (Table 3).

View this table:
[in this window]
[in a new window]
 
Table 2. Age-Adjusted Rates of Mortality from Coronary Heart Disease and from Any Cause.

 
View this table:
[in this window]
[in a new window]
 
Table 3. Black:White Mortality Rate Ratios for Coronary Heart Disease and All Causes of Death, Adjusted for Age and for Other Risk Factors.

 
When age-specific mortality rates were examined for the 30-year observation period (data not shown), none of the white women and black women were found to have died of coronary disease at 35 to 44 years of age. Not until the age of 65 was there a significant increase in the mortality rate among the women. The mortality rate was higher among black men at the ages of 35 to 44 and 55 to 64 than among white men. However, there were no statistically significant differences in age-specific coronary mortality rates between either the two groups of men or the two groups of women.

For all causes of death (Table 2), the age-adjusted rate was significantly higher for black women than white women. The rate for black men was not significantly higher than the rate for white men. The excess deaths among black women were attributed to diabetes and to causes listed as ill-defined or unspecified. The excess mortality from any cause among black men was due to stroke and ill-defined or unspecified causes. When the rates of mortality from any cause were examined in 10-year age-specific strata, black men had the highest rates, followed by white men, black women, and white women.

The black:white age-adjusted mortality rate ratios for coronary heart disease (0.9 in men and 1.2 in women), when also adjusted for cardiovascular disease risk factors, decreased to 0.7 in men and 0.9 in women (Table 3). The black:white ratio for mortality from any cause among men was 1.2 when adjusted only for age and 1.0 after adjustment for all risk factors; the values for women were 1.5 and 1.1, respectively.

Kaplan-Meier survival analysis indicated that white men had the highest probability of death from coronary disease and white women the lowest over the 30-year observation period (Figure 1). Black men and women had similar survival curves for the first 17 years, after which there was a divergence, with the curve for black men more closely paralleling the curve for white men but with a greater probability of survival. There were no statistically significant differences in the overall survival curves between black men and white men (P = 0.27 by the log-rank test) or between black women and white women (P = 0.17).


View larger version (14K):
[in this window]
[in a new window]
 
Figure 1. Kaplan-Meier Estimates of the Probability of Death from Coronary Heart Disease in the Charleston Heart Study, 1960 through 1990.

The log-rank test revealed no significant differences in survival between black men and white men (P = 0.27) or black women and white women (P = 0.17).

 
Predictors of Coronary Mortality

Systolic blood pressure, whether adjusted for age or age and other risk factors, was a significant predictor of coronary mortality across all four groups (Table 4). The effect of body-mass index on the relative risk was not significant in any group, and a higher level of education was a significant and protective factor (relative risk, 0.79) only for white men. Additional sex-specific analyses indicated no significant interactions of race with any risk factors (data not shown). The relative risk of coronary disease in the presence of elevated cholesterol levels (a change of 48.4 mg per deciliter) and after adjustment for age was highest for white women (1.34), black women (1.32), and white men (1.13); each of these values was significant or nearly significant.

View this table:
[in this window]
[in a new window]
 
Table 4. Relative Risk of Death from Coronary Heart Disease, 1960 through 1990, after Adjustment for Age and for Other Risk Factors.

 
A history of cigarette smoking in 1960 was associated with a relative risk of 3.12 in black men, 1.94 in white women, and 1.60 in white men. Smoking was not a predictor of death from coronary heart disease in black women. A history of diabetes adjusted for all the risk factors was associated with a nonsignificant elevation in the risk of coronary disease in all four groups.

Predictors of Death from Any Cause

The relative risk of death from any cause in the presence of elevated levels of systolic blood pressure or serum cholesterol levels, and a history of cigarette smoking or diabetes, was significant or nearly significant in all four groups. The exceptions were for serum cholesterol level among men, diabetes status among white men, and smoking status among black women (Table 5). Body-mass index was not a significant predictor of death in any group. A higher level of education was a significant (protective) predictor among white men (relative risk, 0.85) and black women (relative risk, 0.79).

View this table:
[in this window]
[in a new window]
 
Table 5. Relative Risk of Death from Any Cause, 1960 through 1990, after Adjustment for Age and for Other Risk Factors.

 
Discussion

This 30-year observational study of blacks and whites from a southeastern area of the United States showed that there were no statistically significant differences between blacks and whites in coronary mortality rates, although there were significant differences in coronary mortality rate ratios between men and women. Black women fared better than black men, and white women better than white men. There were no significant differences in survival patterns between blacks and whites over the 30-year period.

The rate of mortality due to coronary disease among black men was slightly lower than the rate among white men (4.6 vs. 5.2 per 1000 person-years). This difference, however, was not statistically significant. The mortality ratio for blacks as compared with whites was slightly less than 1 whether we considered the age-adjusted ratio of 0.9 or the ratio of 0.7 adjusted for age and other risk factors (Table 3). The upper limit of the ratio after adjustment for age and other risk factors was only slightly greater than 1. The rate among black men may be understated because more blacks than whites died outside the hospital or because there was insufficient information to determine the cause of death for more blacks. For example, during the period from 1960 through 1990, 30 of 36 deaths whose causes were listed as ill-defined or unspecified occurred in blacks. However, the 20-year black:white coronary disease mortality ratio among men was 0.9 in Evans County,15 as compared with the 30-year ratio of 0.9 in the Charleston Heart Study. These strikingly similar, independent findings from two Southern community-based studies suggest that black men may be relatively protected from coronary disease.

The findings of lower rates of mortality from coronary heart disease among black men in these two closed-cohort studies are not discordant with national statistics, which indicate consistently higher rates for white men than for black men for the period from 1968 through 198616. Gillum17 examined black:white mortality ratios for coronary disease in the United States in 1986 to determine age-specific coronary mortality. The ratio was 1.6 for people 35 to 44 years of age, 1.2 for those 45 to 54 years of age, 1.0 for those 55 to 64 years of age, 0.9 for those 65 to 74 years of age, 0.8 for those 75 to 84 years of age, and 0.9 for those >= 85 years of age, reflecting a mortality crossover among older people.

On the basis of risk-factor profiles, it would seem reasonable for blacks to have rates of mortality from coronary disease that equaled or exceeded those for whites. Although the total cholesterol levels in blacks in our study were lower than those in whites (Table 1), blood pressures were markedly higher in blacks than whites. The proportion of smokers was similar among black men and white men; only 27 percent of black women, as compared with 40 percent of white women, reported smoking cigarettes. On average, blacks had lower levels of education than whites. The prevalence of diabetes among blacks may be understated because, in 1960, blacks in the study area had less access to medical care than whites, and thus we were unable to determine whether they had a history of diabetes. Black men and white men had essentially the same distribution of values for body-mass index. The mean body-mass index of black women was significantly higher than that of white women (27.3 vs. 24.4).

For all causes of death, the rate was highest among black men (18.9 per 1000 person-years), but it was not significantly higher than the rate (15.3 per 1000 person-years) among white men. The rate of death from any cause was significantly and substantially lower among women than among men, but the rate among black women was significantly higher than the rate among white women (12.7 vs. 7.3 per 1000 person-years).

The finding of elevated blood pressure, elevated serum cholesterol level, and a history of cigarette smoking as risk factors for coronary disease in white men and white women was expected. The lower level of association of diabetes and death from coronary heart disease in white men may be explained by an increased control of diabetes. The finding of a similar spectrum of risk factors among blacks and whites is important because it is derived from 30 years of observation of a closed cohort. After a 20-year study of black men in Evans County, Georgia, Tyroler et al.15 reported that systolic blood pressure and cigarette smoking were significant predictors of coronary mortality. Our study confirms these earlier findings, but presents evidence that there are similarities between blacks and whites in other risk factors, such as cholesterol level and diabetes.

It is interesting that single assessments of blood pressure, serum cholesterol level, smoking status, and diabetes status made 30 years ago are predictors of mortality, even though in the past decade there has been considerable emphasis on modifying these risk factors.

A lower level of education was a significant predictor of coronary disease and mortality from any cause only among white men and black women. The absence of such an effect among black men and white women needs additional investigation.

Our data suggest that the major coronary heart disease risk factors predict mortality among blacks as well as whites. Otten et al.18 have shown in a national probability sample that smoking, systolic blood pressure, cholesterol level, body-mass index, and diabetes explain some of the excess mortality from any cause among blacks. They also reported that 38 percent of the excess mortality among blacks, as compared with whites, can be explained by family income. As Ragland and Brand19 have suggested, with the use of a long follow-up period, it can be inferred that the predictive variables are robust. Our findings are consistent with others inferring that the standard, major risk factors for coronary heart disease are predictive for both blacks and whites. However, the risk of coronary disease may vary between blacks and whites. In the two cohort studies of community-based samples in the southeastern United States, the risk appeared somewhat higher for white men than black men and for black women than white women at comparable levels of the standard cardiovascular risk factors. This suggests that other determinants of coronary disease, such as high-density lipoprotein cholesterol level20 and fibrinolytic activity, reported as differentially distributed across these populations, may have resulted in different absolute levels of coronary heart disease. The failure to identify a "significant" level of risk for risk factors may be due to a lack of statistical power to detect an effect or to the true lack of an association between risk factor and coronary heart disease mortality. It is obvious that there is not a linear effect between every variable and mortality. The cholesterol level21 and the body-mass index may require quadratic or cubic terms specific for race and sex to explain their relation with mortality properly.

Other potentially biasing factors, such as errors in death-certificate diagnoses and lack of response to the initial request for participation, have to be considered. The lack of complete comparability between the eighth and ninth editions of the ICD codes for coronary disease could influence the differences in rates between blacks and whites. Repeated contacts by the investigators with participants during the observation period may have also had an unintended interventional effect. Larger-scale studies including blacks must be advocated to overcome the limitations of past or ongoing studies. The findings regarding cardiovascular risk factors in blacks and whites are based on limited numbers in studies carried out in one region of the United States (i.e., the Southeast), and the results therefore are not necessarily generalizable to the whole country. However, the available cohort data indicate that, while we await the results of additional observational studies and clinical trials, we should pay equal attention to controlling the standard major risk factors in blacks and in whites.

Supported by a grant (R01 HL31397) from the National Heart, Lung, and Blood Institute.

We are indebted to the late Dr. Edwin Boyle, Jr., founder of the Charleston Heart Study, for his contributions to the implementation of this study.


Source Information

From the Charleston Heart Study, Medical University of South Carolina, Charleston (J.E.K., S.E.S., R.G.K., D.T.L., P.C.G.), and the Department of Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill (H.A.T.).

Address reprint requests to Dr. Keil at the Charleston Heart Study, Rm. 908 Harborview Office Towers, Medical University of South Carolina, 171 Ashley Ave., Charleston, SC 29425-2239.

References

  1. Gordon T, Kannel WB. Multiple risk functions for predicting coronary heart disease: the concept, accuracy, and application. Am Heart J 1982;103:1031-1039. [CrossRef][Medline]
  2. Dawber TR. The Framingham Study: the epidemiology of atherosclerotic disease. Cambridge, Mass.: Harvard University Press, 1980.
  3. Peery TM, Langsam SM. Incidence of fatal cardiovascular disease in Charleston, S.C. Arch Intern Med 1939;64:971-987. [Free Full Text]
  4. Cassel J, Heyden S, Bartel AG, et al. Incidence of coronary heart disease by ethnic group, social class, and sex. Arch Intern Med 1971;128:901-906. [Free Full Text]
  5. Comstock GW. An epidemiologic study of blood pressure levels in a biracial community in the southern United States. Am J Hyg 1957;65:271-315.
  6. McDonough JR, Hames CG, Stulb SC, Garrison GE. Coronary heart disease among negroes and whites in Evans County, Georgia. J Chronic Dis 1965;18:443-468. [CrossRef][Medline]
  7. Gillum RF. Coronary heart disease in black populations. I. Mortality and morbidity. Am Heart J 1982;104:839-851. [CrossRef][Medline]
  8. Boyle E Jr, Griffey WP Jr, Nichaman MZ, Talbert CR Jr. An epidemiologic study of hypertension among racial groups of Charleston County, South Carolina: the Charleston Heart Study, phase II. In: Stamler J, Stamler R, Pullman TN, eds. The epidemiology of hypertension: proceedings of an international symposium. New York: Grune & Stratton, 1967:193-203.
  9. Keil JE, Loadholt CB, Weinrich MC, Sandifer SH, Boyle E Jr. Incidence of coronary heart disease in blacks in Charleston, South Carolina. Am Heart J 1984;108:779-786. [CrossRef][Medline]
  10. Zlatkis A, Zak B, Boyle AJ. A new method for the direct determination of serum cholesterol. J Lab Clin Med 1953;41:486-492. [Medline]
  11. Bergstrahl EJ, Offord KP, Kosanke JL, Augustine GA. PERSONYRS: a SAS procedure for person year analyses. In: Proceedings of the Eleventh Annual Conference of the SAS User's Group International, Atlanta, February 9-12, 1986. Cary, N.C.: SAS Institute, 1986:868-73.
  12. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457-81.
  13. So Y. PROC PHREG: a procedure for Cox's proportional hazards regression analysis. In: Proceedings of the SAS User's Group International Conference, New Orleans, February 17-20, 1991. Cary, N.C.: SAS Institute, 1991:1-52.
  14. Cox DR. Regression models and life-tables. J R Stat Soc [B] 1972;34:187-220.
  15. Tyroler HA, Knowles MG, Wing SB, et al. Ischemic heart disease risk factors and twenty-year mortality in middle-age Evans County black males. Am Heart J 1984;108:738-746. [CrossRef][Medline]
  16. Cooper RS, Ghali JK. Coronary heart disease: black-white differences. In: Saunders E, Brest AN, eds. Cardiovascular diseases in blacks. Vol. 21. No. 3 of Cardiovascular clinics. Philadelphia: F.A. Davis, 1991:205-25.
  17. Gillum RF. Cardiovascular disease in the United States: an epidemiologic overview. In: Saunders E, Brest AN, eds. Cardiovascular diseases in blacks. Vol. 21. No. 3 of Cardiovascular clinics. Philadelphia: F.A. Davis, 1991:3-16.
  18. Otten MW Jr, Teutsch SM, Williamson DF, Marks JS. The effects of known risk factors on the excess mortality of black adults in the United States. JAMA 1990;263:845-850. [Free Full Text]
  19. Ragland DR, Brand RJ. Coronary heart disease mortality in the Western Collaborative Group Study: follow-up experience of 22 years. Am J Epidemiol 1988;127:462-475. [Free Full Text]
  20. Gillum RF, Grant CT. Coronary heart disease in black populations. II. Risk factors. Am Heart J 1982;104:852-864. [CrossRef][Medline]
  21. Knapp RG, Sutherland SE, Keil JE, Rust PF, Lackland DT. A comparison of the effects of cholesterol on CHD mortality in black and white women: twenty-eight years of follow-up in the Charleston Heart Study. J Clin Epidemiol 1992;45:1119-1129. [CrossRef][Medline]

 

This Article
-Abstract

Commentary
-Letters

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation

Related Letters:

Explaining the Connection between Privilege and Health
Muller H. E., Jackson R. H., Davis T. C., Lonsdale D., Huber F. C., Pincus T., Callahan L. F., Roberg N. B., Angell M.
Extract | Full Text  
N Engl J Med 1994; 330:139-140, Jan 13, 1994. Correspondence

This article has been cited by other articles:



HOME  |  SUBSCRIBE  |  SEARCH  |  CURRENT ISSUE  |  PAST ISSUES  |  COLLECTIONS  |  PRIVACY  |  TERMS OF USE  |  HELP  |  beta.nejm.org

Comments and questions? Please contact us.

The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved.