DETERMINANTS OF DIETARY-INTAKE IN A SAMPLE OF WHITE AND MEXICAN-AMERICAN CHILDREN

Objective To assess the influence of several potential psychosocial de terminants on children's eating behavior. Participants Three hundred f ifty-one Mexican-American and non-Hispanic white children (mean age=4. 4 years old at baseline) participated in the San Diego Study of Childr en's Activity and Nutrition for up to 2.5 years. Methods Child's eatin g behavior was described by 3 dependent variables: total energy, perce ntage energy from fat, and sodium intake per 1,000 kcal. Dietary infor mation was collected 4 days a year using a 24-hour food intake record, which was a combination of direct observation and interviews with foo d preparers. The 35 predictor variables from child, parental, demograp hic, and environmental domains were collected by behavioral observatio n, interviewer-administered questionnaires, and physical measurements. Statistical analyses Bivariate and regression analyses via mixed line ar models were performed. Results Variables from the children's domain (such as skinfold thickness and weight) had the strongest association s with energy intake; parental variables (such as fat avoidance behavi or and prompts to increase children's food intake) were associated wit h child's percentage energy from fat and sodium intake. In regression analyses, parsimonious subsets of variables accounted for 46% of varia nce in energy intake (3 variables), 40% of the variance in percentage of energy from fat (4 variables), and 44% of variance in sodium intake per 1,000 kcal energy (1 variable) in between-subject variance compon ents. Conclusions Fat and sodium intake of children may be improved by improving parents' nutrition habits and by having parents encourage c hildren to eat a healthful diet. Few modifiable correlates of children 's energy intake were identified.