BACKGROUND:

Mortality from coronary heart disease has declined over 40 percent since the late 1960s. Possible explanations for the decline include the observed trends in risk factors such as changes in rates of hypertension, dyslipidemia, and glucose intolerance; lifestyle changes such as improvements in diet and decreases in the levels of cigarette smoking; and better detection and treatment of heart disease. The specific effects of risk factor, lifestyle, or detection and treatment changes on the declining population rates of coronary heart disease morbidity and mortality were unknown in 1990. Secular trend analyses were needed to elucidate the relative contribution of these factors, particularly dietary variables, on population rates of coronary heart disease.

DESIGN NARRATIVE:

Multiple regression was used to assess the effect of dietary variables on changes in serum cholesterol between 1957-1960 and 1966-1969 in a sample of 200 men for whom repeated measurements existed in the Framingham cohort. Analysis of covariance was used to assess influence of dietary intake on differences in serum cholesterol levels between independent samples of men studied in 1957-1960 and 1966-1969 in the Framingham cohort, and independent samples of women studied between 1957-1960 and 1984-1988 in the Framingham cohort and Offspring cohort, and independent samples of men studied between 1966-1969 and 1984-1988 in the Framingham cohort and Offspring cohorts respectively.