Distinct profiles of white matter diffusivity in obesity and bipolar disorders: An ENIGMA bipolar disorders study in...

Bipolar disorders (BD) are associated with a wide range of brain white matter abnormalities, but results vary across studies. This heterogeneity likely relates to effects of additional clinical/diagnostic variables, including medications and obesity, on white matter in BD. In this mega-analysis, we studied the links between relevant clinical variables and four diffusivity measures.

We obtained diffusion tensor imaging-derived fractional anisotropy (FA), mean, axial, and radial diffusivity (MD, AD, RD) values from 910 individuals with BD and 1256 healthy controls across 16 research sites in the ENIGMA-BD Working Group. We used principal component analysis to detect patterns of covariance across white matter tracts and studied associations between multivariate patterns of white matter alterations and demographic and clinical variables, including body mass index (BMI), medications, and diagnostic information, using mixed regression models.

Higher BMI was associated with lower PC1 FA and higher PC1 RD; BD was associated with higher PC1 MD, while treatment with antiepileptic medications was associated with lower PC1 FA, higher PC1 MD, and PC1 RD.

We found FA and MD are weakly associated, providing complementary information, and there are unique associations between specific diffusivity measures and clinical predictors. Treatment with antiepileptics may inflate the statistical effects of BD on white matter. By using all diffusivity measures and accounting for additional clinical factors, we can gain new insights into brain correlates of severe mental illness.