The role of immune cell phenotypes and metabolites in bipolar disorder risk: a Mendelian randomization-based mediation analysis
摘要
Bipolar disorder (BD) is a multifactorial psychiatric disorder with contributions from genetic susceptibility and environmental exposures. Recent evidence suggests immune dysfunction and metabolic disturbances may contribute to BD, but the causal roles and mediation mechanisms remain ambiguous.
MethodsTwo-sample MR analyses were conducted leveraging publicly accessible GWAS summary data to investigate causal associations among immune cell phenotypes, circulating metabolites, and BD risk.Immune trait data included 731 phenotypes; metabolite data comprised 486 blood metabolites. BD outcome data (n = 413,466) were sourced from the IEU OpenGWAS database (ID: ieu-b-5110). We applied inverse-variance weighted (IVW), MR-Egger, weighted median, and mode-based methods. Bidirectional MR, two-step MR (TSMR), and multivariable MR (MVMR) were used to assess direct and mediated effects.
ResultsA significant association was observed between BD and distinct B cell phenotypic profiles. The proportion of naive-mature B cells was inversely associated with risk (β = -0.091, SE = 0.035, P = 0.008), while CD24 + CD27 + B cells showed a positive association (β = 0.093, SE = 0.033, P = 0.005). Carnitine and its derivatives (e.g., acetylcarnitine) were linked to reduced BD risk, whereas maleate, alpha-hydroxyisovalerate, and phospholipid derivatives were positively associated with risk. Naive-mature B cells were positively linked to alpha-hydroxyisovalerate levels. The mediation analysis demonstrated a statistically significant indirect pathway. (β = 0.00269, P = 0.046) with a mediated proportion of -2.95%.
ConclusionOur results demonstrate genetic underpinnings of the causal relationships between specific immune cell phenotypes, blood metabolites, and BD pathogenesis. A potential immune-metabolic pathway was identified, where naive-mature B cells affect BD risk via alpha-hydroxyisovalerate. These findings offer mechanistic insight and suggest novel targets for biomarker development and therapeutic intervention.