This study investigated whether the nurturing hypothesis - that breastfeeding serves as a proxy for family socio-economic characteristics and parenting behaviours - accounts for the association of breastfeeding with children's academic abilities. Data used were from the Child Development Supplement of the Panel Study of Income Dynamics, which followed up a cohort of 3563 children aged 0-12 in 1997. Structural equation modelling simultaneously regressed outcome variables, including three test scores of academic ability and two subscales of behaviour problems, on the presence and duration of breastfeeding, family socio-economic characteristics, parenting behaviours and covariates. Breastfeeding was strongly related to all three tests scores but had no relationships with behaviour problems. The adjusted mean differences in the Letter-Word Identification, Passage Comprehension) and Applied Problems test scores between breastfed and non-breastfed children were 5.14 [95% confidence interval (CI): 3.14, 7.14], 3.46 (95% CI: 1.67, 5.26) and 4.24 (95% CI: 2.43, 6.04), respectively. Both socio-economic characteristics and parenting behaviours were related to higher academic test scores and were associated with a lower prevalence of externalising and internalising behaviour problems. The associations of breastfeeding with behaviour problems are divergent from those of socio-economic characteristics and parenting behaviours. The divergence suggests that breastfeeding may not be a proxy of socio-economic characteristics and parenting behaviours, as proposed by the nurturing hypothesis. The mechanism of breastfeeding benefits is likely to be different from those by which family socio-economic background and parenting practices exert their effects. Greater clarity in understanding the mechanisms behind breastfeeding benefits will facilitate the development of policies and programs that maximise breastfeeding's impact.
Keywords: academic ability; behaviour problems; breastfeeding; child development.
© 2015 John Wiley & Sons Ltd.