Minnesota Heart Health Program

BACKGROUND:

The rationale for prevention derives from the mass nature of adult cardiovascular diseases in affluent societies, the insidious development of the underlying processes, particularly hypertension and atherosclerosis, the high immediate out-of hospital mortality of coronary disease and stroke, and the long-term excessive risk after survival of an initial episode.

The potential for a preventive approach in cardiovascular diseases was based on the large differences in disease rates found between populations and the strong associations between individual risk factor levels and disease rates within high risk cultures. The congruence of these population observations with clinical evidence and with plausible mechanisms of cause, strengthened inference of their causal importance. Further, the safety and feasibility were well established of modifying individuals' cardiovascular risk characteristics and changing personal health behavior through medical and educational strategies. Finally, the dynamic nature of the changes in vital statistics on deaths reported from hypertension, stroke, and coronary heart disease (CHD) in many nations, at the rate of one to two percent a year and greater, both upward and downward, indicated their preventability, though none of these trends was adequately explained when the study began in 1980.

The rationale for a population or community-wide strategy, as in the Minnesota Heart Health Program, contrasted to a high-risk individual, medical approach, was based on all the above considerations, plus the demonstration that entire populations were at excess risk relative to others. Focusing solely on the portion of higher risk people among high risk cultures appeared to be a necessary medical part of a community-wide prevention approach, but it was insufficient and inefficient as a sole or major public health strategy. This was mainly because the bulk of attributable (excess) cases of cardiovascular disease came from the large central part of the population distribution of risk, not the tail. The socially learned behaviors which lead to the precursors of cardiovascular disease were also mass phenomena, requiring a population strategy of prevention and health promotion. Finally, concentration of preventive effort only among the high risk or only on adults tended to ignore the mass emergence of youth into early adulthood bearing the physical characteristics of excess risk and already having well-developed unhealthy behaviors. Therefore, a rational and effective public health strategy would appear to be one directed toward all ages and segments of the community, over a sustained period, with the ultimate objective to prevent elevated risk and risky behaviors in the first place.

DESIGN NARRATIVE:

Three pairs of communities were selected, each pair with one education and one control community. Communities were matched on size and distance from Minneapolis-St. Paul and pairs were similar in median income, education, health care, and media resources. The first pair, Mankato and Winona, were small agricultural communities of approximately 40,000 in population in 1980. The second pair, Fargo, North Dakota and Moorhead, Minnesota paired with Sioux Falls, South Dakota, were medium-sized urban centers of approximately 100,000 in population. Bloomington and Roseville, Minnesota were suburban centers in the Twin Cities metropolitan area with a population of approximately 80,000.

The education program began in Mankato in 1981 after two baseline surveys, in Fargo-Moorhead in 1982 after three baseline surveys, and in Bloomington in 1983 after four baseline surveys. Education activities continued in the three communities for five years concluding in the staged manner in which they began. Targeted risk factors included blood pressure, exercise habits, smoking, and blood cholesterol levels. Health messages were communicated by involvement of community leaders and organizations, media education, population-based risk factor screening and education, adult education classes, youth and parent education disseminated in schools, health professional's educa0tion, and community-wide risk factor education campaigns.

The effectiveness of the program was evaluated by annual population survey samples of cross-sections or cohorts and by morbidity and mortality surveillance. The annual risk factor survey measured community and individual change in risk factors and related behaviors in 25-74 year old persons living in each of the six sites. Annual surveys included between 300-500 persons and were population-based, random, neighborhood cluster samples of each town. Selected households were invited to participate. A home interview was conducted to collect data on health beliefs, attitudes, and behaviors, medical history, health message exposure, and demographic characterization. After the home interview, participants had additional risk factor measurements at a survey center. Data were collected on height, weight, blood pressure, serum total and HDL cholesterol, and serum thiocyanate. In each community, a 50 percent sample of subjects was assessed for dietary habits and the other 50 percent for physical activity.

Morbidity and mortality data on myocardial infarction and stroke were collected and analyzed for the Minnesota Heart Health Program areas and mortality data for all of Minnesota, North Dakota, and South Dakota. Computer classification algorithms were developed jointly with the Pawtucket Heart Health Program and the Stanford Five-City Multifactor Risk Reduction Program to allow pooling of data of the three studies. All hospitalized cases of myocardial infarction and stroke were investigated and cause of death was validated for cardiovascular disease deaths occurring out-of-hospital. Deaths occurring throughout Minnesota, North and South Dakota were catalogued. Hospital disease surveillance was carried out with the cooperation of 34 area hospitals. Mortality was ascertained by death certificate counts and from tapes supplied by the three State Health Departments. Morbidity and mortality data were compared between pooled education versus pooled comparison communities and mortality data were compared between these and the remaining areas of the three states.