What are the causal effects of breastfeeding on IQ, obesity and blood pressure? Evidence from comparing high-income with middle-income cohorts

• DOI: 10.1093/ije/dyr020
• Publication Date: 2011-02-25T01:47:54Z
• Authors: Marie-Jo A Brion; Debbie A Lawlor; Alicia Matijasevich; Bernardo Horta; Luciana Anselmi; Cora L Araújo; Ana Maria B Menezes; Cesar G Victora; George Davey Smith
• Keywords: Longitudinal Study

Background A novel approach is explored for improving causal inference in observational studies by comparing cohorts from high-income with low- or middle-income countries (LMIC), where confounding structures differ. This applied to assessing effects of breastfeeding on child blood pressure (BP), body mass index (BMI) and intelligence quotient (IQ). Methods Standardized approaches the structure socio-economic position were British Avon Longitudinal Study Parents Children (ALSPAC) (N ≃ 5000) Brazilian Pelotas 1993 1000). was used improve regarding associations BP, BMI IQ. Analyses extended include results a meta-analysis five LMICs 10 000) compared randomized trial promotion. Findings Although higher strongly associated ALSPAC, there little such patterning Pelotas. In lower IQ, adjusted confounders, but directions expected if due socioeconomic patterning. contrast, Pelotas, not BP Differences observed between ALSPAC LMIC line those robust evidence heterogeneity detected associations. Trial data supported conclusions inferred cross-cohort comparisons, which provided IQ BMI. Conclusion While reported are likely reflect residual confounding, may have Comparing populations differing can be studies.