Background: Regulated public childcare must follow nutrition and physical activity guidelines, but the impact of public childcare on childhood adiposity is unclear.
Objectives: To estimate the effects of universal preschool childcare on children's BMI in elementary school in Quebec, Canada, and whether the effects differed in children from more or less advantaged families.
Methods: For 1657 children enrolled in the Quebec Longitudinal Study of Child Development (1998-2010), BMI z-scores (BMIz) from 6 to 13 years were regressed on the childcare used from 2 to 5 years, adjusted for pre-childcare variables. Average treatment effects were estimated using the Bayesian multilevel linear regression and g-computation for four childcare profiles: 1) parental care or full-time care (35 hours/week) in a 2) centre-based, 3) regulated home-based or 4) unregulated home-based arrangement.
Results: Had all participants attended centre-based care, mean BMIz in kindergarten would have been 0.38 (95% credible interval [CrI] 0.23, 0.52), which was 0.40 (95% CrI 0.14, 0.65) SD higher than regulated home-based, 0.20 (95% CrI -0.04, 0.43) SD higher than unregulated home-based and 0.36 (95% CrI 0.11, 0.60) SD higher than parental care. By 12 years, mean BMIz had increased for all childcare profiles, but differences between childcare profiles had diminished.
Conclusions: Although centre-based childcare was associated with an earlier rise in BMI, compared with informal care, it had no large, enduring effect, overall, or for less advantaged children, in particular.
Keywords: Bayesian analysis; body mass index; child; child care; child health; child preschool; epidemiologic effect modifier; linear models; public policy.
© 2021 John Wiley & Sons Ltd.