Pleiotropic effects drive correlation between body mass index and cortical myelination

Background Epidemiological studies have reported significant associations between obesity and neurocognitive decline. Understanding these associations will require deeper analyses of how body mass index (BMI) and brain structure are related. Here we explore the extent to which shared genetic factors (pleiotropy) govern the association between BMI and cortical myelination. Methods Statistical models of bivariate heritability were applied to structural MR image data from a cohort of monozyogotic and dizygotic twins. Estimates of phenotypic and genetic correlation between BMI and cortical myelination were derived. A co-twin control design based on monozygotic twins was used to test the hypothesis of a causal relationship between BMI and myelination. The variation in the genetic correlation across the cortex was compared with the average statistical enrichment of genes associated with obesity derived from data from the Allen brain atlas. Results Statistically significant phenotypic and genetic correlation between BMI and cortical myelination was observed across the cortex. Taking the heritability of each trait into account, approximately 80% of the phenotypic correlation between the traits was accounted for by shared genetic factors. Intra-pair differences between traits in monozygotic twins failed to support a causal relationship. Moreover, variation in genetic correlation across the cortex was significantly associated with the statistical enrichment of genes related to obesity. Conclusions These results support the hypothesis that pleiotropic effects drive the association between BMI and cortical myelination. This observation may help to explain the co-occurrence of obesity in neurocognitive decline and mental health disorders characterized by changes in myelination and oligodendrocyte function.