2306-P: Determinants of Overweight and Obesity in 8-Year-Old Children in ATS Insubria: A Cumulative Logit Model

Introduction and Objective: ATS Insubria, the local health authority in Como and Varese, using data from “OKkio alla Salute” a nutritioral surveillance system in primary school, developed a predictive model to estimate the probability of children falling into different weight categories. The model studied associations with potential risk factors like parental BMI, dietary habits and physical activity. Methods: The children were categorised according to their BMI as normal weight, overweight or obese. A cumulative logit (proportional odds) regression model was estimated for ordinal outcomes. The coefficients were interpreted as odds, thereby providing insight into how each factor influences the risk of having a BMI in the higher range. Results: The use of AIC in model selection has led to the identification of parental BMI, sex, the number of days per week on which the child partakes in a minimum of one hour of active play outside of school and the frequency of vegetable consumption (excluding potatoes) as explanatory variables. The results of the estimation process are reported in Table 1. Conclusion: Children whose lifestyle does not adhere to CREA guidelines for healthy eating and exercise, as well as having at least an overweight parent, are more likely to be at least overweight themselves. Prevention strategies should focus on improving urban health through nutrition, movement, and environmental changes to combat chronic diseases, as proposed by “Cities for Better Health” program.