[Multi-level mixed effect model study on the association between dietary environmental risk factors and obesity among Chinese children and adolescents].

OBJECTIVE

To analyze the influence of dietary environmental factors in 12 provinces of China on obesity among children and adolescents aged 7-17 in 2011.

METHODS

Based on the dietary environmental and physical examination data of China Health and Nutrition Survey 2011(Multi-stage stratified cluster random sampling method, 3 day 24 hour retrospective method, individual-household-community questionnaire survey and physical examination method such as height and weight measurement were used), 1416 children and adolescents aged 7-17(male 50. 71%; the age group of 7-10 years old accounted for 43. 50%, that of 11-14 years old 37. 43%, and that of 15-17 years old 19. 07%) with complete data were selected as the research subjects, and combined with the dietary environmental data of the corresponding provinces from National Statistical Yearbook 2011, a three-level(province-community-individual) logistic mixed effect(random intercept) model was constructed to binary outcome variables of obesity or not, the effect of dietary environments on obesity in children and adolescents was analyzed, and the characteristics of risk factors were determined.

RESULTS

Considering that the P values of the two-level empty model of "province-individual"and"community-individual" constructed separately were P<0. 005 and 0. 050<P<0. 100 respectively when the random effects of layering being tested, this study fitted the three-level(province-community-individual) logistic random intercept mixed effect model. The model analysis showed that after adjusting the impact of other indicators at various levels, the provincial GDP(OR=1. 00006, P=0. 001, 95% CI1. 00003-1. 00010) and the provincial per capita daily consumption of edible oil(OR=1. 18317, P<0. 001, 95% CI 1. 10054-1. 27200) were positively correlated with obesity aged 7-17. Taking the age group of 7-10 years old as the reference group, the two groups of 11-14 years old and 15-17 years old were negatively correlated with obesity, with parameter estimates of(OR=0. 59518, P=0. 049, 95% CI(0. 35465-0. 99885) and(OR=0. 13225, P<0. 001, 95% CI0. 04716-0. 37084) respectively. In addition, the estimated parameters of the logistic three-level random intercept prediction model for obesity in children and adolescents aged 7-17 years were obtained: intercept(-2. 23457, P<0. 001, 95% CI-2. 59581--1. 87334]); provincial GDP(0. 00003, P=0. 007, 95% CI 0. 00001-0. 00005); per capita edible oil intake in province(0. 10642, P<0. 001, 95% CI 0. 06365-0. 14919); taking age group 7-10 years old as a reference, the group of 11-14 years old(-0. 56777, P=0. 007, 95% CI-0. 97837--0. 15717), the group of 15-17 years old(-1. 36679, P<0. 001, 95% CI-2. 02890--0. 70469).

CONCLUSION

This study suggests that in addition to individual factors, obesity of children and adolescents aged 7-17 years is also affected by the dietary environmental factors at provincial level, among which provincial GDP and per capita daily edible oil consumption have significant positive impacts. In terms of individuals, the age between 7 and 10 years before puberty is a highly risky period for childhood and adolescent obesity.