Search for a command to run...
ABSTRACT Background Previous evidence suggests that higher body size is associated with bipolar disorders, however, whether this association is causal remains uncertain. Interpretation is further complicated by heterogeneity across age, variation in clinical presentation, and potentially distinct underlying aetiologies. Aims To determine whether body size exerts heterogenous causal effects on bipolar disorder subtypes and symptom profiles. Methods By leveraging genetic instruments that differentiate effects at different life stages, summary–level univariable and multivariable Mendelian randomisation (MR) analyses were used to estimate how age–specific body size relates to adult psychiatric and symptomatic bipolar features; major depressive disorder (MDD), depressive symptom scores, subthreshold mania symptoms, bipolar disorder, bipolar type I and bipolar type II. Genetic instruments derived from genome–wide association studies (GWASs) for adult body mass index (BMI) (n= 681,275), childhood body size (n= 453,169) and mid-to-later life body size (n= 453,169) served as proxies for prepubertal and adult BMI measures. Results In univariable MR, higher genetically proxied adult BMI increased the odds of MDD (odds ratio (OR) = 1.13, 95% CI 1.09-1.16), subthreshold mania (OR = 1.09, 95% CI 1.0-1.19)), and depressive scores (β = 0.07, 95% CI 0.05-0.09). There was little evidence that childhood body size had an effect on any outcome. Robust evidence suggested bipolar disorder and MDD increased adult BMI in our reverse univariable analyses. Using multivariable MR, robust evidence indicated that increased adult body size after accounting for childhood body size increased the odds of MDD, subthreshold mania and depressive scores. Conclusions Body size may exert different causal effects on bipolar disorder depending on age and symptoms, with detrimental effects occurring during adulthood. Weaker evidence suggested varying effects across bipolar subtypes. Triangulation of findings and higher powered GWASs to detect symptom–specific genetic variants are required to explore whether body size contributes to distinct aetiologies across bipolar patients, informing the identification of novel and personalised treatment targets.