GLP-1 Receptor Agonists and Risk of Kidney Replacement Therapy and Health Care Utilization in Bipolar Disorder With Chronic Kidney Disease.

Objective: To examine whether glucagon-like peptide-1 receptor agonist (GLP-1 RA) use is associated with reduced risk of kidney replacement therapy (KRT) and health care utilization (HCU) compared with active comparators in adults with bipolar disorder (BD) and chronic kidney disease (CKD; stage ≤3). Methods: Using TriNetX, we conducted a real-world, new-user active-comparator cohort study among adults with BD+CKD initiating GLP-1 RA or dipeptidyl peptidase-4 (DPP-4) inhibitor therapy (primary comparator), with a secondary analysis versus renin-angiotensin-aldosterone system (RAAS) inhibitor users. Propensity score matching (PSM) balanced baseline characteristics. Cox proportional hazards regression estimated adjusted hazard ratios (aHR) for KRT and HCU over 5-year follow-up. Subgroup analyses examined effect modification by sex, race, ethnicity, obesity, and psychotropic medication use. Results: Compared with DPP-4 inhibitor use, GLP-1 RA use was associated with lower observed risk of KRT (aHR 0.56, 95% CI 0.50-0.64) and HCU (aHR 0.74, 95% CI 0.71-0.78). Following PSM (n=3,908 per group), 5-year KRT-free survival was 87.93% versus 84.63% (P<.001); HCU-free survival did not differ significantly (P=.89). Versus RAAS inhibitors, GLP-1 RA use was similarly associated with lower observed risk of KRT (aHR 0.66) and HCU (aHR 0.62); after PSM (n=2,539 per group), KRT-free survival was higher (94.68% vs. 89.77%; P<.001), while HCU-free survival did not differ significantly (P=.15). Benefits were consistent across sex, race, and obesity status and across psychotropic subgroups, including lithium. Conclusions: GLP-1 RA use was associated with lower observed risk of KRT and HCU compared with DPP-4 inhibitors, corroborated for KRT in a secondary comparison against RAAS inhibitors. These exploratory findings warrant further investigation, including prospective studies, to evaluate whether this association reflects a causal effect.