Food choice and quantity of food consumption in relation to relevant variables: Genetic analyses in an older age sample

Recent studies have found evidence for patterns in eating behavior that may predispose individuals to health problems in later life. In most studies, a possible genetic influence on dietary patterns has been overlooked. This work explores the genetic and environmental factors in diet, health, and in selected variables that are relevant with respect to diet and/or health: body weight, physical exercise, socio-economic status (education and family income), alcohol consumption, smoking status, personality and depression. Subjects were an older age twin volunteer sample (1262 males and 3378 females), which was recruited through the AARP journal. Dietary information was obtained with the National Cancer Institute Diet History Questionnaire (Block et al., 1986). Factor analyses on food use, serving size, and consumption frequency measures from this questionnaire revealed two independent, meaningful factors. Food items with high fat, sugar or salt contents loaded high on factor1, while factor2 items reflected healthy eating behavior. Mean gender differences were found for these factor consumption patterns, with females reporting more factor2, and males more factor1 intake. Univariate genetic analyses indicated that genetic factors were important in the dietary factors (heritability estimates ranged from.29 to.42 for factor1 and.42 to.53 for factor2). There was evidence for family environmental effects on factor 1 food consumption in females, but not in males. The non-shared environmental components explained more variation than did genetic and family environmental factors. This suggested that nutritional education programs can play a role in a change towards healthier eating behavior. Genetic and environmental factors also contributed to the subjective health evaluation and the variables that are relevant to diet and health. The magnitude of the effects differed depending on the variable studied and the effects were mostly gender-specific. Multivariate models yielded evidence for genetic and family environmental influences that were shared by the dietary factors, health, and the relevant variables. Little of the specific environmental components was found to be shared between variables. This finding suggested that an environmental change in one health-related behavior will not necessarily effect another behavior.