Purpose: To evaluate the extent to which levels of physical activity and nutrient intake aggregate in families, and secondarily, to assess the repeatability of these behavioral measures over a 5-year period.
Methods: Measurements were obtained in a population-based sample consisting of 1364 members of 42 large Mexican American families. Nutrient intake was assessed by a food frequency questionnaire validated for use in this population. Usual level of physical activity was estimated using a 7-day recall questionnaire.
Results: Correlations between baseline (obtained 1992-1995) and follow-up (obtained 1996 to 2000) measures of all behaviors were highly significant (p < 0.001), ranging from 0.24 for % of calories derived from fat to 0.44 for % of calories derived from alcohol. Familial effects, estimated using variance component methods, were stronger when modeled as a genetic heritability than as a shared household effect; as a heritability they accounted for a significant portion of the total variation of all traits (9% for physical activity levels, p < 0.05; and 13-26% for nutrient intake, p < 0.001 for all).
Conclusions: Measurements of physical activity and dietary behaviors in this population tracked over 5 years, and there was a significant degree of aggregation of these behaviors within families. Understanding the sources of these family effects may facilitate efforts to improve cardiovascular health.