Comparative efficacy and safety of liraglutide versus metformin, naltrexone/bupropion, and phentermine-topiramate in psychiatric patients.

BACKGROUND: Psychiatric patients have a high risk of obesity, frequently due to psychotropic medication-induced weight gain. However, real-world comparative data on antiobesity medications (AOMs) in this population remain rare. OBJECTIVES: To compare short-term weight-loss efficacy, adverse events (AEs), and early discontinuation (ED) among psychiatric outpatients taking liraglutide (LIRA), naltrexone/bupropion, phentermine-topiramate (PT), or metformin (MET). DESIGN: Retrospective observational cohort study. METHODS: We conducted a 12-week retrospective chart review of 117 psychiatric outpatients with International Classification of Diseases, 10th Revision, F01-F99 diagnoses. Percent weight change over time was analyzed using linear mixed-effects models. AEs and ED were compared across treatment groups. RESULTS: Compared with MET, LIRA was associated with a greater percent weight reduction (estimate -3.45%, 95% confidence interval (CI) -5.35 to -1.55,  < 0.001), with a significant treatment-by-time interaction at 12 weeks ( = 0.019). Female sex and full-time employment were associated with attenuated weight loss, and the number of concomitant psychotropic medications with moderate weight-gain risk showed a trend toward greater weight reduction ( = 0.066). No significant differences were observed in AE incidence across AOMs. ED rates differed by drug type ( = 0.017), being lowest in the MET group (39.1%) and highest in the PT group (72.2%). CONCLUSION: In this real-world psychiatric cohort, LIRA was associated with greater short-term weight loss than MET without an increased observed frequency of AEs. ED rates varied across AOMs. These findings should be interpreted cautiously, given the observational design and short follow-up period, and require confirmation in larger, long-term studies.