Genetically predicted breastfeeding propensity and brain microstructural morphology: Insights from genome-wide association study summary statistics and neurogenomics.

BACKGROUND: Breastfeeding provides nutrients and bioactive components relevant to infant development, but observational associations with brain outcomes may be influenced by socioeconomic, maternal, and family-level factors. Mendelian randomization (MR) offers a genetic framework for evaluating associations by using genetic variants as proxies for breastfeeding propensity; it does not directly measure the effect of actual breastfeeding behavior. OBJECTIVES: This study investigated associations between genetically predicted breastfeeding propensity and brain imaging-derived phenotypes (IDPs). METHODS: We performed a two-sample MR analysis using genome-wide association study (GWAS) summary statistics from 352,094 European participants. Genetically predicted breastfeeding propensity served as the exposure, and 393 brain IDPs from the UK Biobank served as outcomes. Double Machine Learning (DML), using random forest models and 5-fold cross-fitting, was applied as a complementary validation approach. Heterogeneity and pleiotropy analyses were conducted to assess the robustness of the instrumental-variable estimates. RESULTS: At the nominal significance level, positive MR estimates were observed between genetically predicted breastfeeding propensity and the left anterior thalamic radiation fractional anisotropy measure (IDP dMRI ProbtrackX FA atr l: Beta = 2.787, p = 0.045), the right lingual area (Beta = 2.032, p = 0.022), and the right superior temporal area (Beta = 1.719, p = 0.011). Inverse estimates were observed for measures including left mean cortical thickness (Beta = -2.266, p = 0.014). DML provided complementary evidence for associations involving the subcallosal gyrus and precentral thickness. Statistical power exceeded 80% for the primary outcomes; however, nominal findings that did not meet the Bonferroni threshold should be considered exploratory. CONCLUSIONS: This study provides genetic evidence of region-specific associations between genetically predicted breastfeeding propensity and brain structural phenotypes, including an association with higher fractional anisotropy in the left anterior thalamic radiation. Associations involving regions implicated in language, visual processing, and cortical maturation do not establish corresponding cognitive or functional effects. The findings are hypothesis-generating and require replication in independent and longitudinal datasets before direct behavioral or public health implications can be inferred.