MRI-derived body composition in bipolar and psychotic disorders

ObjectiveBipolar (BD) and psychotic disorders (PD) are severe mental disorders (SMDs) associated with higher cardiometabolic disease burden. Prior studies of body composition in SMDs are inconclusive. Magnetic resonance imaging (MRI) offers detailed insights into body composition and may improve our understanding of cardiometabolic disease risk in patients with SMDs. MethodsWe derived body composition measures from MRI of patients with BD (n=36, 63.9% females, mean age=36.6{+/-}10.9 years), PD (n=27, 33.3% females, mean age=31.7{+/-}8.9 years), and controls (n=251, 78.8% females, mean age=40.7{+/-}12.1 years). Using linear models we tested for case-control differences in body mass index (BMI) and body composition in all patients and split by diagnostic group (BD and PD), accounting for age, sex, and scanner, and subsequently adding BMI. Next, we tested for age- and sex-adjusted associations between body composition and antipsychotic dose in patients. ResultsAnalyses showed significantly higher BMI (BD: Cohens d=0.70; PD: d=0.65), abdominal subcutaneous fat (BD: d=0.78; PD: d=0.71), visceral fat (BD: d=0.76; PD=0.68), liver fat (BD: d=0.53; PD: n.s.), and thigh muscle fat infiltration (BD: d=0.80; PD: d=0.87) in cases compared to controls. When accounting for BMI, muscle fat infiltration remained significantly higher (BD: d=0.54; PD: d=0.68) and thigh muscle volume was lower (BD: d=-0.65; PD: d=-0.63). There were no significant associations of antipsychotic dose with BMI or body composition. ConclusionsThe observed case-control differences suggest that individuals with BD and PD, at the group level, exhibit higher regional fat accumulation and lower muscle volume, possibly contributing to the heightened risk for cardiometabolic comorbidity in SMDs.