Death Rates from Coronary Disease -- Progress and a Puzzling Paradox

In the late 1940s, recognition of coronary heart disease (CHD) as the leading cause of death in the United States prompted efforts to identify factors promoting its occurrence, in the belief that the resultant insights would advance knowledge of the pathophysiology of atherosclerotic cardiovascular disease and pave the way for its prevention. Hypertension, hypercholesterolemia, and cigarette smoking were identified by the early 1960s as principal risk factors for CHD. (1,2) More recently, clinical trials of antihypertensive drugs and lipid-lowering therapies, along with investigations of smoking cessation, have documented the benefits of treating these risk factors and established beyond any doubt their causal roles in promoting CHD. (3)

National guidelines for the treatment of hypertension and hypercholesterolemia have evolved to keep pace with the growing body of evidence from basic science, animal models, observational studies, and clinical trials. Updated guidelines have redefined the levels of blood pressure and serum cholesterol that warrant treatment, as well as the goals of therapy. (4,5) National programs directed at these risk factors have contributed to substantial improvements in awareness, treatment, and control of hypertension and hypercholesterolemia and to reductions in cigarette smoking. Data from the second and third National Health and Nutrition Examination Surveys (covering 1976 through 1980 and 1988 through 1994, respectively) for adults 20 to 74 years of age reveal a 40 percent decline in the prevalence of hypertension (from 39 percent to 23 percent) and a 28 percent decrease in the prevalence of hypercholesterolemia (from 26 percent to 19 percent). (6) Similarly, from 1979 to 1994, cigarette smoking among adults 18 years of age or older declined by 25 percent, from 33 percent to 25 percent. (6) These favorable trends have been observed for both sexes and for blacks as well as whites.

In a parallel manner, enormous advances have been made in the management of acute myocardial infarction and in the long-term treatment of CHD. Such advances include the increasing use of aspirin, thrombolytic therapy, beta-blockers, angiotensin-converting-enzyme inhibitors, and lipid-lowering therapy, as well as improvements in revascularization procedures. (7)

Over the past 30 years, mortality from CHD has declined by more than 50 percent (Figure 1). (8,10) This downward trend is continuing, and the decrease has occurred among men and women, blacks and whites, and both middle-aged and older persons. From 1990 to 1996 the average annual decline in mortality from CHD was 2.6 percent. (8,10) This decline is probably attributable to advances in both primary and secondary prevention. A concomitant decline in case fatality rates for hospitalized patients with acute myocardial infarction also has been observed (Figure 1), (9,10) presumably resulting from improvements in secondary prevention, although a reduction in the severity of myocardial infarction in response to primary-prevention efforts could also lower the case fatality rate. Whereas extensive data regarding mortality and case fatality rates for CHD are available, information on contemporary trends in the incidence of CHD is more limited.

In this issue of the Journal, investigators from the Atherosclerosis Risk in Communities (ARIC) study report on trends in mortality due to CHD and in the incidence of myocardial infarction from 1987 through 1994, on the basis of a review of 3023 deaths due to CHD and 8572 hospitalizations for myocardial infarction in four communities of varying size. (11) Rosamond et al. report a decline in mortality from CHD that amounts to 28 percent among men and 31 percent among women. This reduction was accompanied by declines in case fatality rates for myocardial infarction and in rates of reinfarction.

The declines in mortality due to CHD and in case fatality rates for myocardial infarction in the sample were similar to national trends during the same period (Figure 1) and to trends in mortality reported elsewhere. (7,12) Yet the ARIC data on mortality due to CHD have advantages over national mortality statistics, which are based on death certificates and subject to greater error due to misclassification, because Rosamond et al. validated the cause of death through an intensive process of review and adjudication. Although theirs is not a national sample, the diverse nature of the communities studied provides stronger evidence of the magnitude of the decline in mortality from CHD in the United States in recent years and offers gratifying testimony to the effect of secondary prevention on the public health.

In contrast to the declines in mortality from CHD and in case fatality and recurrence rates, the incidence of myocardial infarction was unchanged or increased slightly in the ARIC sample. At first glance, the divergent trends in mortality from CHD and the incidence of myocardial infarction would appear to be in conflict. However, incidence and mortality need not necessarily track in the same direction. It is possible for trends in these two statistics to move in opposite directions if an increase in incidence coincides with a decline in case fatality rates resulting from the occurrence of less severe infarctions or improvements in acute or long-term management.

A more puzzling paradox is the fact that no decline in the incidence of myocardial infarction was observed in the ARIC sample during a period when the prevalence of causal risk factors was reduced. This apparent conflict may be explained in several ways. First, the data of Rosamond et al., if representative and unbiased, may reflect a failure of primary prevention on a national level. This explanation is not likely, given the declines in the prevalence of hypertension, hypercholesterolemia, and smoking in our country and the relation of these risk factors to the incidence of myocardial infarction. (13) It is also at odds with recent estimates that 25 percent of the decline in mortality from CHD is attributable to primary prevention. (14) Second, the findings of the ARIC study are based on only four communities and may not be representative of national trends in the incidence of myocardial infarction. Earlier studies have reported declines in the incidence of myocardial infarction. In most cases, however, these changes were small, and they were not always statistically significant. (7,12) Third, and most probable, although rates of myocardial infarction have declined, the declines may not be reflected in the ARIC data for a variety of complex reasons that influence all surveillance studies of trends in the incidence of myocardial infarction.

Greater public awareness of the symptoms of myocardial infarction may have resulted in more frequent hospitalization and the diagnosis of events that formerly went unrecognized. Better emergency response services may have permitted the survival of patients with severe myocardial infarctions that previously would have been fatal before patients could reach the hospital and thus would have gone undetected. The education of physicians may have yielded higher rates of admission and more diagnostic testing (notably electrocardiography and assays for biochemical markers of infarction) in recent years. Improved laboratory tests for the detection of myocardial infarction may have shifted patients with unstable angina toward a diagnosis of myocardial infarction, thereby raising the incidence. During the period from 1987 through 1994, the tests used for the diagnosis of myocardial infarction moved away from reliance on lactate dehydrogenase (LDH) and total creatinine kinase (CK) levels to measurement of their more sensitive isoenzymes LDH1 and CK-MB. Even the assays for CK-MB changed considerably over the same period, from qualitative assessment to semiquantitative electrophoresis, immunoinhibition, and more recently, measurement of CK-MB mass. The introduction of new and far more sensitive biochemical markers of myocyte necrosis will affect current and future rates of myocardial infarction. Two additional changes that can falsely raise the apparent incidence of myocardial infarction, but that were largely controlled for by the methods used by Rosamond et al., are changes in the codes in the International Classification of Diseases in the late 1980s and the effects of Medicare reimbursement based on diagnosis-related groups. (15)

We have cause to celebrate the steep reductions in mortality from CHD that have occurred over the past 30 years. These declines are best explained by the joint contributions of primary and secondary prevention. Several factors, however, hamper our ability to make accurate assessments of trends in morbidity from CHD; paradoxically, some apparently adverse trends are a consequence of our successes in education and treatment. Despite important advances in treatment and prevention, CHD remains the single most common cause of death in the United States. It is essential that we continue our efforts to reduce morbidity and mortality from CHD through both primary and secondary prevention.

Daniel Levy, M.D.
National Heart, Lung, and Blood Institute
Framingham, MA 01702

Thomas J. Thom
National Heart, Lung, and Blood Institute
Bethesda, MD 20892

Text

References