Analysis of family influencing factors of dietary behavior pattern of children and adolescents

Abstract

Objective: To identify dietary patterns among children and adolescents aged from 7 to 18 and the associations between these patterns and family characteristics. Methods: A stratified cluster sampling method was used. Data was collected on 2 438 students and their parents through physical examinations and questionnaires. Students were from 16 schools (4 urban primary schools, 4 rural primary schools, 4 urban middle schools, and 4 rural middle schools) in Fangshan district, Beijing. Dietary patterns were derived by factor analysis. Rank sum tests and Pearson correlation analysis were used to analyze the correlations between family characteristics and the scores on dietary patterns. Generalized linear mixed models were used to examine the associations between family characteristics and dietary patterns, for univariate analyses. Results: Two dietary patterns were identified: the risk pattern and the protective pattern. Results from the univariate analyses showed that maternal BMI was associated with the risk pattern (P=0.011). All factors, including the only-child, parental education level, monthly household income, paternal age at birth and maternal BMI, were related to the protective pattern (all P<0.05) except for the paternal BMI. After adjusting for gender, age, locations of residence (urban-rural) and BMI z-score, children with a lower parental education level and higher monthly household income were more likely to adhere to the risk pattern (β=-0.10, 95%CI:-0.16- -0.04; β=0.07, 95%CI: 0.02-0.12, respectively). For the protective pattern, the scores were positively associated with parental education level (β=0.08, 95%CI: 0.02-0.14), monthly household income (β=0.06, 95%CI: 0.02-0.11) and maternal age at birth (β=0.02, 95%CI: 0.00-0.03). Children from the one-child families were more likely to adhere to the protective pattern (β=-0.13, 95%CI: -0.22- -0.03). Conclusions: Differences of dietary behaviors were seen among children and adolescents from families with different characteristics. Protective patterns for children from families with lower parental education, lower monthly household income, lower maternal age at birth or multiple children etc. should be promoted. Risk patterns of children with lower parental educational or higher monthly household income also need to be corrected.