The Decline: Trends in Coronary Heart Disease and Stroke Mortality
It is difficult today to evoke the dramatic impact on the scientific community of the 1977 New England Journal of Medicine editorial by medical practitioner, Weldon Walker, which first called wide attention to the substantial decline occurring in death rates due to coronary heart disease (CHD) among the U.S. population. (1) Epidemiologists and officials concerned with health and vital statistics were embarrassed at having missed it; grumbling that “it takes X years to define a true trend among the vagaries of annual death rates, etc.”
But the decline in age-adjusted CHD death rates was a fact. It could be seen in data from the mid-1960s, and, though its beginnings in California women had been reported early, the author of the report, Nemat Borhani, had not ballyhooed the observation and it had not been widely recognized by other experts.
In an enterprising 1978 response to this new awareness, and presumably to explore whether its research program had influenced the decline, the National Heart and Lung Institute convened a special conference to which leading researchers in CVD epidemiology were invited to ponder the issues: Was the decline real and, if so, what were its components, and what might be the contributing factors?
Conference participants met, deliberated, and concluded broadly that the decline was real and that both medical care and risk factor reduction had contributed to it, but that many questions remained about the nature and underlying causes of the trends. Plans were made for surveillance research in the U.S. and abroad. Three such research-based monitoring programs provided much of the documentation, analysis, and explanation of the trends, establishing the reality, nature, and importance of the decline in coronary death rates: the Minnesota Heart Survey (MHS), the WHO MONICA Project, and ARIC (Atherosclerosis Research in Communities). Studies in Massachusetts and in Olmsted County, Minnesota contributed.
Coronary Heart Disease (CHD) rates
Researchers at the University of Minnesota already had set out in 1977 to examine systematically the components of the decline in CVD death rates with a proposal to NIH for the Minnesota Heart Survey (MHS). This project, which continues almost three decades later, undertook surveillance of validated deaths and hospitalizations due to CVD, trends in medical care and CVD risk factor levels in cluster samples of the seven-county Greater Twin Cities area, combined with methodolological pursuits needed to improve long-term surveillance.
The MHS compared rates for deaths occurring in-hospital to those occurring in the community to estimate the relative contribution of each to total coronary (CHD) mortality. In doing this it was assumed that a decline in out-of-hospital and sudden coronary death rates reflects predominantly a reduction in new cases of CHD and thus is mainly attributable to preventive efforts and reduced risk factor levels in the population. On the other hand, declines of in-hospital deaths were assumed due mainly to improvements in medical care.
Early results indicated that from 1970 until 1986, CHD death rates declined overall while deaths outside hospital declined at a greater rate than those in-hospital, implying that public health efforts during that period had a relatively greater impact on CHD death rates than did improvements in medical care. Similar results were found in other communities.
Results taken together suggest that the decrease in CVD death rates during the 1970s and ‘80s in the U.S. and abroad was due both to a change in health behavior and risk factor levels and to cardiac care, but predominantly to the former, or public health efforts. In contrast, during the 1990s and the early part of the 21st century, death rates for CHD continued to decline but at a slower rate than in the previous twenty years. Moreover, CHD deaths in-hospital began to decline at a greater rate than those outside hospital, suggesting that improvements in medical care had a relatively greater effect than did public health efforts during the recent period. Similar results were found in U.S. communities studied in the Atherosclerosis Risk in Communities (ARIC) study from 1987 until 1996.
Stroke
The picture of stroke death rates is different, in that the decline started early in the 20th century but became more rapid during the 1970s and ‘80s, followed by a distinct leveling off in the 1990s. The decline began a more dramatically steep descent in 1972, coincident with introduction of the National High Blood Pressure Education and Control Program from NHLI. This involved wide screening and education for high blood pressure with referral for medical treatment.
The steeper slope of the decline in U.S. stroke mortality at that early time was thought to be due to enhanced anti-hypertensive medical treatment and control, independent of public health efforts at altering diet and physical activity behaviors nationally. The sum of these changes, resulting in a downward shift in the average level of blood pressure in the population, was not, however, explainable by use of medications among the “treatable” upper fractiles of pressure level. For example, the WHO international monitoring study, MONICA, showed that the downward trend in average blood pressure and death rates could be found at every segment of the pressure distribution, including the 20th, 50th, and 80th percentiles, indicating that medical therapy was probably not the primary factor in the overall international blood pressure decline of the period. It remains unclear whether the plateau of stroke death rates first seen in the late 1980s in the U.S. is due to a “ceiling effect” of medical treatment, that is, reaching a limit of hypertension control achievable in those having access to good care, or due to “backsliding” in prevention efforts to reduce risk factor levels across the board. (Henry Blackburn)
Reference
1. Walker, W.J. 1977. Changing United States lifestyle and declining vascular mortality: cause or coincidence? (Editorial) New England Journal of Medicine 297 (3): 163-165.
Correspondence