From the Division of Cardiology and Department of Medicine, Harlem
Hospital Center (A.O., D.H., A.A., C.K.F.); the Divisions of Cardiology (D.H.)
and General Medicine (L.W., S.S.), Department of Medicine, Columbia University
College of Physicians and Surgeons; the Department of Statistics, Columbia
University (D.R.); and the Division of Epidemiology, Columbia University
School of Public Health (S.S.), New York, NY.
Correspondence to Dr Charles K. Francis, Department of Medicine, Room 14–401, Harlem Hospital Center, 506 Lenox Ave, New York, NY 10037.
Most of what is known about the effects of hypertension on mortality in
blacks versus whites is based on information derived from
epidemiological studies, clinical trials of drug treatment, or
anecdotal clinical experience. Autopsy data offer an opportunity to
compare the specific cardiovascular pathological
findings that underlie the racial differences described in mortality
statistics, but few autopsy studies addressing this issue have been
published. The study by Strong et al20 focused on
adolescents and young adults, while the study of Shirani et
al21 included only subjects over the age of 80
years. The large studies reported by Roeske et
al22 and Kitzman et al23
did not report the race of the subjects, whereas the study by Burke et
al24 focused on plaque morphology in 113 men, all
of whom had atherosclerotic coronary disease and died suddenly.
The Pathobiological Determinants of Atherosclerosis in
Youth (PDAY) study, which included autopsy data for 1532 subjects aged
15 to 34 years who died of external causes, addressed risk factors for
and the natural history of early atherosclerotic
lesions.25 26 Thus, existing autopsy studies have
not directly compared the underlying pathological causes of death in
blacks and whites who died of cardiovascular disease.
We reviewed the findings in all cases of cardiovascular
death autopsied in the New York City Medical Examiner's Office in
1991. Because of the higher prevalence of hypertension among US-born
blacks compared with whites,27 28 and the
well-established effectiveness of hypertension treatment in both blacks
and whites,29 we tested the hypothesis that
hypertensive vascular disease contributes to
cardiovascular mortality to a greater extent among
blacks than whites.
Autopsy and Clinical Cause of Death: Definition of
Terms
Atherosclerotic heart disease (ICD-9 codes 410 [acute myocardial
infarction], 411 [other acute and subacute forms of
ischemic heart disease], 414 [other forms of chronic
ischemic heart disease], and 440
[atherosclerosis]) was considered the autopsy cause
of death if the autopsy report described pathological evidence of
either myocardial necrosis or significant epicardial coronary
stenosis. Significant coronary stenosis was
defined as >50% narrowing of the cross-sectional luminal diameter of
the left main coronary artery or >75% narrowing of the
cross-sectional luminal diameter of one or more of the major epicardial
coronary arteries. The presence of recent myocardial infarction
was noted if the autopsy record indicated acute infarction within 7
days.
Hypertensive vascular disease (ICD-9 codes 401 [essential
hypertension], 402 [hypertensive heart disease], 403 [hypertensive
renal disease], and 404 [hypertensive heart and renal disease]) was
considered the autopsy cause of death if the autopsy report noted
myocyte hypertrophy in the absence of other heart disease
known to cause this lesion, including aortic valve stenosis or
hypertrophic cardiomyopathy, systemic arteriolar
changes consistent with hypertension, or microscopic evidence
of glomerular or arteriolar damage in the kidneys in the
absence of diabetes mellitus. Left ventricular
hypertrophy was not formally assessed at autopsy. If
atherosclerotic heart disease was present together with evidence of
hypertensive vascular disease, the autopsy cause of death was
considered to be hypertensive vascular disease if pathological evidence
of myocardial necrosis or significant coronary artery
stenosis was absent.
Congestive heart failure (ICD-9 codes 425
[cardiomyopathy] and 428 [heart failure]) was
considered the autopsy cause of death if the left ventricle or all the
heart chambers were dilated and/or if hepatic or pulmonary
congestion was noted in the autopsy report.
The category "other cardiac causes" included ICD-9 codes 391 and
393 to 398 (acute and chronic rheumatic heart disease), 415 to 417
(acute and chronic pulmonary heart disease), 420
(pericarditis), 421 (endocarditis), 422 (myocarditis), 423 (other
pericardial diseases), 424 (other endocardial diseases), 426
(conduction disorders), 427 (cardiac dysrhythmias), 429 (ill-defined
heart disease), and 430 to 438 (cerebrovascular disease).
Clinical cause of death was determined by the investigators after
evaluating all available clinical and necropsy data, including heart
weight and descriptions of the events immediately preceding the death.
Myocardial infarction was considered the clinical cause of death based
on pathological evidence for recent myocardial necrosis and clinical
evidence of recent myocardial infarction. All 57 subjects who at
autopsy had evidence of recent myocardial infarction were classified in
terms of clinical cause of death as dying of myocardial infarction. One
additional subject had clinical evidence supporting recent myocardial
infarction and was classified in this category of clinical cause of
death.
Sudden cardiac death was specified as the clinical cause of death if an
unforeseen death occurred within 1 hour of onset of symptoms. Sudden
cardiac death was a clinical cause of death but not an autopsy cause of
death, and some subjects whose autopsy cause of death was
atherosclerotic heart disease, hypertensive vascular disease,
congestive heart failure, or other cardiac cause were classified as
sudden cardiac death in terms of clinical cause of death.
The criteria for congestive heart failure and "other cardiac
causes" were the same as for autopsy cause of death, as described
above. A single clinical cause of death could not be asssigned for 144
of the decedents (63 blacks and 81 whites). In such cases, clinical
cause of death was listed as "undetermined." The assigned
categories for all cases were reviewed by A.A. and A.O.
Statistical Methods
Atherosclerotic heart disease and hypertensive vascular disease were
the predominant autopsy causes of death among both blacks and whites
(Table 1
Analysis of the clinical causes of death showed that sudden
cardiac death was responsible for 274 of the 587 (47%) total deaths,
accounting for 156 of 314 (50%) of the deaths among blacks and 118 of
273 (43%) deaths among whites (Table 1
The predominance of hypertensive vascular disease as the autopsy cause
of death among blacks compared with whites was present among both
men and women (Table 2
The age-adjusted odds ratio (OR) for hypertensive vascular
disease as the autopsy cause of death for blacks versus whites was 2.2
(P<.001) among men and 3.1 (P=.01) among women
(Table 3
Mean heart weight was significantly greater in bivariate
analyses among black women compared with white women (460
versus 384 g; P<.001) (Table 4
Among the 274 subjects who died of sudden cardiac death, 234 had either
atherosclerotic heart disease or hypertensive vascular disease as the
underlying autopsy cause of death (Table 5
In considering whether the findings in this autopsy series can shed
light on the differential cardiovascular disease
mortality rates in blacks versus whites, a key issue is the basis for
selection of deaths for autopsy by the medical examiner. Because
selection was largely based on the circumstances of death, our case
series should not be viewed as a representative sample,
and the actual rates of specific cardiovascular causes
of death observed for blacks and whites should not be applied directly
to these populations as a whole. The finding that there were large
differences by race within this autopsy series in the relative rates of
hypertensive vascular disease and atherosclerotic heart disease does
suggest that this pattern may be present in the population as a
whole. While it is possible that factors leading to selection for
autopsy differed systematically between blacks and whites, this seems
unlikely because the Medical Examiner's Office was not aware of the
race of the decedent at the time it decided whether to take a case.
Selection biases are therefore unlikely to account for the large
differences between blacks and whites in the underlying pathology
responsible for the cardiovascular deaths in this
series. Our series included all adult autopsied
cardiovascular deaths in New York City in 1991,
excluding only those whose autopsy reports did not record their
race, whose race was recorded as neither black nor white, who were
morbidly obese, or who died under highly unusual circumstances.
The left ventricle constitutes the main component of the mean heart
weight, particularly in the hypertrophied
heart.31 The propensity of the left ventricle to
hypertrophy is strongly related to blood pressure
level.32 33 The relationship of left
ventricular hypertrophy to race and skin color
is controversial. Some studies have shown no relationship after taking
account of severity and duration of
hypertension,34 35 whereas others have found such
relationships.36 37 Left ventricular
hypertrophy defined
electrocardiographically33 and
echocardiographically38 was shown
in the Framingham Heart Study to be strongly associated with incident
cardiovascular events after adjustment for other
cardiovascular risk factors, including blood
pressure.33 The relationship in our study between
mean heart weight and autopsy-based classification of underlying cause
of death, with higher mean heart weight among those with hypertensive
vascular disease, supports the validity of this classification.
Sudden death was the clinical cause of death in 274 of the 587 (47%)
of the total deaths in our study. This high frequency of sudden death
was not explained by acute cocaine or opiate use, since blood
toxicology was positive in only a small proportion of subjects. It was
also not explained by acute myocardial infarction or past myocardial
infarction, since pathological evidence of recent infarction was
present in only 57 subjects and past myocardial infarction in 46
additional subjects. Left ventricular
hypertrophy has been found to predispose to
ventricular arrhythmias39 40
and sudden death,41 and it is likely that
hypertensive heart disease and left ventricular
hypertrophy contributed to the high frequency of sudden
death, especially among blacks, in whom hypertensive vascular disease
was more common than atherosclerotic heart disease. The high proportion
of deaths that met clinical criteria for sudden death may also in part
reflect autopsy selection factors, since the Medical Examiner's Office
autopsies cases in which death is unexpected or unexplained.
National data show that the incidence of sudden death is greater among
blacks compared with whites or Hispanics.42 We
found that in whites who died of suddent death, atherosclerotic
coronary disease, presumably leading to acute or chronic
myocardial ischemia, was the predominant autopsy finding. In
contrast, among blacks who died of sudden death, hypertensive vascular
disease was more common at autopsy than atherosclerotic heart
disease.
Autopsy data from the Medical Examiner's Office have several
limitations. Documentation may be imprecise once it has been
ascertained that death was due to natural causes, so that attention to
subtle distinctions among etiologic causes may have varied from case to
case. The fact that several persons performed the autopsies may also
have contributed to variability in accuracy. The rigorous autopsy
techniques followed by hospital or research pathologists may not have
been uniformly or generally followed in all cases in this series. This
variability or imprecision in classification of specific
cardiovascular cause of death, however, is unlikely to
have been systematically different in blacks compared with whites.
Thus, possible misclassification of specific cause of death due to
inaccuracy would not explain the observed differences between blacks
and whites in our study, and, indeed, would lead to an underestimation
of the odds ratios. Another limitation is that clinical information on
many of the cases was sparse and probably unreliable. It is for this
reason that we based our analysis almost entirely on autopsy
findings, the main exception being the suddenness of death, a
variable that involves a minimum of subjectivity. Several studies
have reported that heart weight adjusted for body mass index is
superior to heart weight alone as an index of hypertensive left
ventricular
hypertrophy.43 44 We were unable to
calculate body mass index for our subjects because of lack of height
data, but we were able to adjust heart weight for body weight, as well
as for other covariates in our analysis.
Hypertension causes or contributes to mortality through damage of
organs other than the heart, in particular via its pathogenic role in
stroke.3 Our study focused on cardiac causes of
death based on the ICD-9 codes for the first cause of death assigned by
the Medical Examiner's Office. We therefore excluded
subarachnoid hemorrhage and other hypertension-related
strokes.
National data2 4 7 8 45 show that the greatest
excess cardiovascular disease mortality risk among
blacks is in younger age groups, and particularly among blacks in their
40s and 50s. While this excess risk is present among both men and
women, it is greater among black women, compared with white women, than
among black men, compared with white men. National data for 1991, the
year in which the subjects in our study died, show these same patterns,
with the greatest excesses in large core metropolitan areas compared
with large fringe metropolitan, medium/small metropolitan, urban
nonmetropolitan, or rural areas.45 These data are
based on ICD-9 diagnoses derived from death certificates, rather than
autopsy findings, and do not permit direct comparison of hypertensive
vascular disease with atherosclerotic heart disease as specific,
autopsy-validated causes of cardiovascular death.
Consistent with these national data, the black subjects in our
study were younger at time of death than whites, and this age
difference was present among women but not men.
The contribution of hypertension to the pathological processes
observed in this series of autopsies goes beyond those subjects whose
autopsy findings documented hypertensive vascular disease as the
underlying cause of death, since hypertension is also a major
contributing cause to atherosclerotic heart disease and congestive
heart failure.32 These three causes of
cardiovascular death accounted for more than 80% of
all the cardiovascular deaths in this autopsy series,
with other cardiac causes explaining only 16% of the total. Despite
the demonstrated effectiveness of antihypertensive medications in
lowering blood pressure and preventing the complications of
hypertension,29 national surveillance data
continue to show substantial gaps in detection, awareness, treatment,
and control.46 Rates of hypertension control are
lower among minority and socially disadvantaged
groups.47 48 Studies of patients with
hypertensive emergencies49 and severe
uncontrolled hypertension50 found that almost all
of these hypertensive persons had been previously treated. Lack of
access to primary care50 and financial
barriers51 are key problems for uninsured or
underinsured groups and groups with large out-of-pocket costs for
medications, although other studies have found low rates of
hypertension control even in fully insured healthcare
workers.52 The findings in this large autopsy
series underscore the importance of hypertension in the excess
cardiovascular mortality experienced by blacks in the
United States.
Received November 25, 1997;
first decision November 25, 1997;
accepted December 5, 1997.
© 1998 American Heart Association, Inc.
Scientific Contributions
Hypertensive Vascular Disease as a Cause of Death in Blacks Versus Whites
Autopsy Findings in 587 Adults
Abstract
Top
Abstract
Introduction
Methods
Results
Discussion
References
Abstract—Cardiovascular
disease is the major cause of excess mortality among urban US blacks,
but autopsy data comparing black-white differences in underlying
pathological causes of cardiovascular death are
lacking. We reviewed all 720 adult cases autopsied in 1991 in the New
York City Medical Examiner's Office in which the coded cause of death
was cardiovascular disease (International
Classification of Diseases, 9th Revision, codes 391, 393 to
398, 401 to 404, 410, 411, 414 to 417, 420 to 438, and 440 to 444).
After exclusion of 133 cases because race was missing or coded as other
than black or white, gender was not coded, or there was an unusual
circumstances of death or extreme obesity, 587 cases were available for
analysis. There were 314 black and 273 white subjects. Black
women were younger than white women at time of death (mean age, 54.7
versus 61.5 years; P<.001), whereas black and white men
did not differ in mean age at death. Hypertensive vascular disease was
the autopsy cause of death in 42% of blacks compared with 23% of
whites (P<.001). Conversely, atherosclerotic heart
disease was the autopsy cause of death in 64% of white subjects but
only 38% of blacks. These patterns were consistent in both
sexes and after adjustment for age. Hypertensive vascular disease was
far more common than atherosclerotic heart disease as the cause of
death at autopsy among blacks compared with whites in New York City,
whereas atherosclerotic heart disease was more common in whites. These
findings suggest that ineffective control of hypertension is a major
factor contributing to excess cardiovascular mortality
among urban blacks.
Key Words: atherosclerosis • autopsy • cardiovascular diseases • race
Introduction
Top
Abstract
Introduction
Methods
Results
Discussion
References
Cardiovascular
disease remains the leading cause of death for both blacks and whites
in the United States,1 but substantial
differences in cardiovascular disease mortality rates
between blacks and whites have been
described.2 3 4 In particular, the interaction of
poverty with race appears to play an important role in contributing to
excess cardiovascular mortality among blacks living in
impoverished urban areas, with the greatest increase in risk among
younger age groups.5 6 7 Moreover, recent studies
show adverse trends for cardiovascular disease
mortality in African Americans compared with whites, with widening of
the differential.2 8 Recent data also show a
greatly increased risk of cardiovascular mortality
among African Americans born in the United States compared with
Caribbean-born blacks,9 and it has been suggested
that these differences are mediated in part by acculturation and
accompanying differences in health-related behaviors that influence
cardiovascular risk.10 Several
hypotheses have been put forward to explain differences between US-born
blacks and whites, including different patterns of risk
factors.11 12 In addition, it has been noted that
blacks have lower rates than whites of utilization of high-technology
cardiology services including coronary
angiography, angioplasty, and
revascularization,13 14 15 16 as
well as other cardiovascular
treatments.17 18 19 It has also been suggested that
fatal cardiovascular disease in blacks involves less
atherosclerosis of the epicardial coronary
arteries than among whites.7
Methods
Top
Abstract
Introduction
Methods
Results
Discussion
References
Data Source and Population
Data were derived from a consecutive series of autopsy
records maintained in the New York City's Medical Examiner's
Office. The autopsy series was drawn from the 72 421 deaths occurring
within the city's geographic boundaries during 1991. Of these, 21 564
were reported to the Medical Examiner's Office, and autopsies were
performed on 7162 of the decedents (9.8% of recorded deaths). All
records in which the decedent was 20 years or older and in which
the cause of death was coded as cardiovascular in
origin (conditions corresponding to International Classification
of Diseases, 9th Revision [ICD-9]30 codes:
391, 393 to 398, 401 to 404, 410, 411, 414 to 417, 420 to 438, and 440
to 444) were reviewed by two of us (A.O. and A.A.). There were 720
cases that met these criteria. Of these, 128 individuals were excluded
from our sample because their race, as coded by the Medical Examiner's
Office, was either missing, mixed, or other than black or white. Five
other cases were excluded: 1 because gender was not indicated, 2
because of extreme obesity, and 2 because their deaths occurred under
highly unusual circumstances (one during electric-shock therapy, the
other from complications of childbirth), leaving a final analytic
sample of 587. Individuals lacking data on heart weight (n=20), body
weight (n=27), and clinical cause of death (n=144) were excluded from
analyses that included those variables but were included in
all other analyses.
Autopsy cause of death was defined by the first cause of death
listed by the medical examiner; this was identified as the principal
cause of death. Only one autopsy cause of death was assigned per case.
Autopsy causes of death were grouped by the investigators into
atherosclerotic heart disease, hypertensive vascular disease,
congestive heart failure, and "other cardiac causes," according to
explicit criteria.
Descriptive statistics were developed to characterize the sample
by race, age, gender, mean heart weight, and autopsy and clinical
causes of death and to examine autopsy and clinical causes of death
according to race and gender. Logistic regression was used to calculate
age-adjusted odds ratios for the most frequent causes of death in
blacks versus whites. Differences in heart weights according to race,
gender, and cause of death were compared using Student's t
tests. Differences in heart weight were also examined using multiple
linear regression analysis in which age, gender, body weight,
and race were entered simultaneously within each stratum of
autopsy or clinical cause of death. Two-tailed values of
P>.05 were considered nonsignificant.
Results
Top
Abstract
Introduction
Methods
Results
Discussion
References
The ages of the 587 study subjects at time of death ranged from 22
to 95 years. Black men did not differ from white men in mean age at
time of death (51.7 versus 51.2 years; P=NS) (Figure
,
A), but black women were significantly
younger at time of death (mean age 54.7 versus 61.5 years;
P<.001) (Figure, B). There were 417 men and 170 women in
this autopsy series, but the proportions of males and females did not
differ significantly by race (69% males among blacks versus 72% among
whites) (Table 1).
View larger version (45K):
[in a new window]
Figure 1. Age distributions of male (A) and female (B) study
subjects by race.
View this table:
[in a new window]
Table 1. Characteristics of Sample Population 
). However, hypertensive vascular disease was approximately
twice as frequent among blacks compared with whites (42% versus 23%;
P<.001) as the autopsy cause of death. In contrast,
atherosclerotic heart disease was more frequent among whites compared
with blacks (64% versus 38%). ). Blood toxicology findings
were available for 485 (86%) of the subjects. Toxicology results were
positive for cocaine in 14 of the 267 (5.2%) blacks and 1 of 218
(0.5%) whites in whom these data were available and for opiates in 12
(4.5%) blacks and 8 (3.7%) whites. ). Hypertensive
vascular disease was the pathological cause of death among 41% of
black males compared with 24% of white males (P<.001) and
44% of black females compared with 19% of white females
(P<.001). Atherosclerotic heart disease was the
pathological cause of death among 64% of white males compared with
39% of black males and 64% of white females compared with 36% of
black females.
View this table:
[in a new window]
Table 2. Autopsy and Clinical Causes of Death for Men (n=417)
and Women (n=170) ). Blacks were correspondingly
less likely than whites to have atherosclerotic heart disease as the
autopsy cause of death (age-adjusted OR, 0.4 for men,
P<.001; age-adjusted OR, 0.4 for women, P<.01).
The age-adjusted ORs comparing clinical causes of death for blacks
versus whites indicated that recent myocardial infarction was less
frequent among blacks versus whites for both men (age-adjusted OR, 0.5;
P<.05) and women (age-adjusted OR, 0.2;
P<.05).
View this table:
[in a new window]
Table 3. Age-Adjusted Odds Ratios1
and 95% Confidence
Intervals for Black vs White Autopsy and Clinical Causes of Death ) but not among black men compared with
white men. Mean heart weight was greater among both blacks (536 versus
416 g; P<.001) and whites (487 versus 432 g;
P<.01) whose autopsy cause of death was hypertension
compared with those with atherosclerotic heart disease. Comparison of
black versus white mean heart weight stratified by autopsy cause of
death showed no significant differences except among subjects with
hypertensive vascular disease as the autopsy cause of death, where
blacks had higher mean heart weight than whites (536 versus 487 g,
P<.05), strongly suggesting that the overall increased mean
heart weight among blacks was explained by the greater prevalence and
severity of hypertension. Comparison of black versus white mean heart
weight stratified by clinical cause of death (lower panel of Table 4)
showed no significant differences between blacks and whites. Multiple
linear regression analyses of mean heart weight in relation to
age, gender, body weight, and race were conducted within each stratum
of autopsy and clinical cause of death. In these analyses, mean
heart weight was associated with race (P=.046) among
subjects with hypertensive vascular disease as the autopsy cause of
death, but not in any of the other strata, consistent with the
bivariate findings.
View this table:
[in a new window]
Table 4. Autopsy Heart Weights in Grams ), whereas the remainder had congestive
heart failure or other cardiac causes as the autopsy cause of death. In
this subgroup of 234 subjects, among blacks atherosclerotic heart
disease was the autopsy cause of death in 42% versus hypertensive
vascular disease in 58%, whereas among whites these proportions were
68% and 32% (P<.001). This pattern was present among
both men and women (Table 5).
View this table:
[in a new window]
Table 5. Autopsy Causes of Death in Blacks vs Whites Among
234 Subjects Who Died of Sudden Cardiac Death
Discussion
Top
Abstract
Introduction
Methods
Results
Discussion
References
The main finding of this study is that hypertensive vascular
disease was substantially more common than atherosclerotic heart
disease as the underlying cause of death among black adults dying of
cardiovascular disease and coming to autopsy in the New
York City Medical Examiner's Office in 1991. Among comparable white
adults, this pattern was reversed, and atherosclerotic heart disease
was considerably more frequent than hypertensive vascular disease as
the underlying cause of cardiovascular death. These
race-specific patterns were consistent in both men and women
and persisted after adjustment for age. Mean heart weight, reflecting
left ventricular hypertrophy, was increased in
both blacks and whites dying of hypertensive vascular disease, compared
with other cardiovascular causes, but was not
associated with race after the underlying cause of death and other
covariates were taken into account.
Acknowledgments
We wish to acknowledge the help and cooperation of the Medical
Examiner's Office of the New York City Department of Health.
References
Top
Abstract
Introduction
Methods
Results
Discussion
References
This article has been cited by other articles:
 |
|
 |
|
  T. C. Lee, P. G. O'Malley, I. Feuerstein, and A. J. Taylor The prevalence and severity of coronaryartery calcification on coronary arterycomputed tomography in black and white subjects J. Am. Coll. Cardiol., January 1, 2003; 41(1): 39 - 44. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. G. McMahon, K. Fujioka, B. N. Singh, C. M. Mendel, E. Rowe, K. Rolston, F. Johnson, and A. D. Mooradian Efficacy and Safety of Sibutramine in Obese White and African American Patients With Hypertension: A 1-Year, Double-blind, Placebo-Controlled, Multicenter Trial Arch Intern Med, July 24, 2000; 160(14): 2185 - 2191. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||