DESIGN NARRATIVE:

Longitudinal analyses were conducted on the associations between diet and cardiovascular and cardiopulmonary disease risk in approximately 3,800 Offspring men and women over eight years of follow-up (from 1984 to present). These analyses considered key nutrients as well as newly defined dietary patterns in Framingham men and women. Cross sectional analyses of diet and plasma lipoprotein relationships in Offspring men and women at two examinations (Exam 3, 1984-88; and Exam 5, 1991-94) were also conducted. All analyses evaluated the total (direct and indirect) and joint effects of dietary nutrients. Where appropriate, analyses controlled for plasma nutrient levels (folacin, vitamins B6 and B12 and homocysteine). The influence of potential confounding variables, effect modifiers, and intervening variables that lie in the causal pathway were examined, as appropriate to the statistical model, including traditional cardiovascular disease risk factors and apoE isoform status.

Several existing Framingham data sets were utilized, including comprehensive data sets on nutrient intake, cardiovascular disease risk factors, clinically verified cardiovascular and cardiopulmonary endpoints and other covariates. The Offspring Exam 3 and 5 nutrition data sets included multiple day assessments of dietary intake which improved the ability to detect relationships of interest. The investigators completed the Framingham Food Habit Questionnaire data set for Exam 5; this instrument included important covariates for the analyses (including vitamin supplement use).