Stanford Five-City Multifactor Risk Reduction Study

BACKGROUND:

The last fifty years of research into cardiovascular diseases have shown strong associations between individual risk factor levels and disease rates within and among many cultures. The potential for medical treatment of cardiovascular disease is limited by the insidious development of atherosclerosis and hypertension, the high immediate mortality of coronary disease and stroke, and the inadequacy of treatment for the advanced stages of cardiovascular disease. All of these factors, coupled with the predictable safety and general feasibility of individual risk factor reduction, provided the stimulus for developing community-wide strategies for the prevention of cardiovascular disease.

The Stanford Three Community Study, conducted between 1972 and 1975, and the Five City Project (FCP) were designed to test the feasibility and effectiveness of such community-wide preventive approaches. Both the Three Community Study and the Five City Project have shown that risk factors can be reduced on a community-wide level, particularly smoking rates and blood pressure levels. These studies were congruent with other contemporary studies, such as the Belgium Heart Disease Prevention Project and the North Karelia Project.

Despite these successes, however, there were additional questions. Would the reduction in risk be translated into a reduction in morbidity and mortality? The North Karelia Project was apparently associated with a fall in mortality as was the Belgium Heart Disease Prevention Program. However, the U.K. Heart Disease Prevention Project was not effective nor was the Multiple Risk Factor Intervention Trial in reducing mortality. It is an appropriate responsibility of the scientific community to inform the general public of the implications of this research and to give advice on the prudent preventive measures.

DESIGN NARRATIVE:

The study had a quasi-experimental design with two California cities, Salinas and Monterey, selected as intervention cities and three similar cities, San Luis Obispo, Modesto, and Santa Monica as control cities. The selection of cities was subject to several criteria including: location in Northern California; population exceeding 30,000 with a total of all five cities exceeding 300,000 to provide sufficient statistical power; ethnic, socioeconomic and demographic similarity; independence from other cities; no shared newspaper or electronic media between treatment and control cities and shared media markets between treatment cities; and independence of the two treatment cities. The education program began in 1980 after the completion of a baseline population survey and continued for eight years. Education methods included the use of television and radio, newspapers, books, pamphlets, and community programs developed through health departments, schools, voluntary organizations, hospitals and non-profit health service agencies.

The effectiveness of the program was evaluated by population surveys in Monterey, Salinas, Modesto, and San Luis Obispo but not Santa Monica and by a morbidity and mortality surveillance system. In the initial year of the survey in 1979-1980, approximately 600 people in each community visited the survey center. This comprised the first independent sample of the community. After an average of 15 months, these same individuals returned to the survey center for a follow-up evaluation. This cohort sample has been studied every other year for five surveys to determine the process of change in cardiovascular risk and related behaviors over time. Five randomly-selected independent samples have been surveyed to study cross-sections of the community over time. Questionnaires and dietary measures were the chief sources of data on health-related behaviors. The questionnaire contained components on: demographic measures; attitude and opinion measures; health knowledge assessment related to prevention of heart disease and stroke; stress behavior; diet, nutrition, and weight behavior; smoking behavior; physical activity behavior; communication media use; and medical history including medication use. A random 40 percent of the initial sample completed a 24-hour dietary recall which was re-administered in the fourth cohort survey and was also administered to a 50 percent random sample of the fourth independent survey. Biological measurements consisted of height and weight, blood pressure, resting heart rate, plasma thiocyanate, total cholesterol, triglycerides and cholesterol content of lipoprotein subfractions, expired air carbon dioxide, urinary sodium, potassium and creatinine, and an exercise test. Data were also collected on plasma apolipoprotein A-1 and B and on high density lipoprotein subfractions 2 and 3. A comprehensive dietary assessment was also conducted.

Community surveillance was conducted to allow the calculation of comparable, city-specific rates for total mortality, cardiovascular mortality, fatal myocardial infarction, nonfatal myocardial infarction, fatal and nonfatal stroke. Development of the community surveillance system was done in close collaboration with the Minnesota Heart Health Project and the Pawtucket Heart Health Project to insure comparability. Only those cardiovascular disease events which always resulted in hospitalization were monitored and included myocardial infarction and stroke. Fatal cases were identified from county death certificate files. Both fatal and nonfatal hospitalized events were investigated by abstracting the medical records onto a set of standard forms. All non-hospitalized deaths were investigated by contacting the decedent's family for medical history and circumstances of death and by contacting the decedent's physician for medical information. Data were monitored on the severity of coronary events and on the quality of care. An Acute Coronary Care form was designed in cooperation with the Minnesota surveillance project and with the MONICA project. The form was field tested in September 1985 and was implemented retroactively on 1980 and 1981 cases.

The study was renewed in 1992 to 1) extend morbidity and mortality surveillance by three years (to June 1994 for events through December 1992); 2) to extend analysis and publication of Five City Project (FCP) results and methods; and 3) to provide a small amount of incremental support to enable use of frozen plasma samples from earlier FCP population surveys to conduct a nested case-control study of the association of HDL subfractions and apolipoproteins with fatal and non-fatal cardiovascular disease. Extension of surveillance data collection was needed to insure the detection of any impact on morbidity and mortality of the significant risk factor changes that occurred in the treatment cities, compared to control cities, by 1986. The nested case-control study involved matching the survey and surveillance data sets to identify persons with documented events who participated in one of the five population surveys, identifying and contacting potential controls to be certain that they were free of CVD, and analysis of plasma samples frozen at -80 degrees centigrade for apolipoproteins A-I and B and HDL2,3 subfractions. This allowed multifactor assessment of the association between those factors and CVD in persons for whom many other variables were known.