20 septembre 2021
info:eu-repo/semantics/openAccess
Lidia Panico et al., « International differences in gradients in early childhood overweight and obesity: the role of maternal employment and formal childcare attendance », Archined : l'archive ouverte de l'INED, ID : 10670/1.wd27k4
Socio-economic differences in BMI/overweight can already be observed from early childhood. There are important cross-country differences in such gradients in later childhood and adulthood, but few studies have assessed whether such international variation is already evident from early childhood. To explain gradients in childhood BMI/overweight, most research has focused on parental behaviours and characteristics, while ignoring other important spheres of a child’s early environment such as their mode of childcare. Access to childcare often varies by socio-economic background, and formal collective childcare appears to be positive for some, but not all, child outcomes; we know less about its role in children’s physical well-being, and especially BMI/overweight. The role of childcare in explaining BMI/overweight gaps might vary widely across countries, given differences in access, quality of care, heterogeneity in quality across groups, etc. In this paper, we explore the variation in gaps in children’s BMI and overweight by parental education across six high income countries, at 3 years old; and assess whether differential attendance to childcare in early life accounts for some of these gaps Preliminary results for France suggest that only the most disadvantaged groups (as measured by the parents’ combined educational level) have an increased risk of overweight at age 3, compared to the most advantaged. Socio-demographic household characteristics (income, family structure, nativity, mother’s age), parental behaviour (smoking during pregnancy, ever breastfed), and child characteristics (birthweight, prematurity) decrease coefficients by about a quarter, while remaining statistically significant. For France, childcare attendance and intensity do not appear to explain observed gaps in BMI/overweight. Substantive results were similar whether a linear BMI variable or a dummy for overweight/obesity is used. Harmonized analyses from the UK, Netherlands, Germany, US, and Japan will assess to what extent these results are generalizable across rich countries.