Effects of neighbourhood income on reported body mass index: an eight year longitudinal study of Canadian children

Background: This study investigates the effects of neighbourhood income on children's Body Mass Index (BMI) from childhood (ages 2-3) to early adolescence (ages 10-11) using longitudinal data. Methods: Five cycles of data from the Canadian National Longitudinal Survey of Children and Youth are analyzed for a sub-sample of children (n = 2152) aged 2-3 at baseline (1994) and assessed at two year intervals to 2002. Body mass index percentiles are based on height/weight estimates reported by proxy respondents (child's person most knowledgeable). Family and neighbourhood factors were assessed at baseline. The prevalence of neighbourhood low income was obtained from the 1996 Census and divided into three categories from 'most poor' to 'least poor'. Longitudinal modelling techniques were applied to the data. Results: After controlling for individual/family factors (age, sex, income, education, family structure) living in the 'most poor' neighbourhood was associated with increasing BMI percentile (1.46, 95% CI 0.16 to 2.75) over time compared to a 'middle' income neighbourhood. Living in an urban (vs. rural) neighbourhood was associated with a decreased BMI percentile (-3.57, 95% CI -6.38 to -0.76) across all time periods. Conclusion: These findings provide evidence that effects of neighbourhood disadvantage on children's BMI occur between childhood and early adolescence and suggest that policies should target the conditions of childhood, including the neighbourhood environment.