Pharmacological interventions for antipsychotic-induced weight gain in schizophrenia: A network meta-analysis

Abstract

Objective

Antipsychotic-induced weight gain (AIWG) is a significant but frequently neglected adverse effect of first- and second-generation antipsychotic therapy, which may lead to cardiovascular disturbances. The present network meta-analysis (NMA) was conducted to evaluate and compare the effects of available treatment options in antipsychotic-induced weight gain (AIWG).

Methods

The data was extracted from 68 relevant clinical trials after a literature search on MEDLINE/PubMed, Embase, Scopus, Cochrane databases and clinical trial registries. Random-effects Bayesian NMA was done to pool the effects across the interventions for the change in body weight from baseline. A network graph was built, a consistency model was run, node split analysis was performed, treatments were ranked as per the SUCRA score and meta-regression was done for the duration of therapy, baseline body weight and treatment strategy as the predictor variables. Finally, the results were sorted based on the certainty of evidence.

Results

The drugs showing significant reduction in body weight in order of magnitude of effect size include sibutramine 10 mg (−8.0 kg; −16. to −0.21), metformin 750 mg + lifestyle modification (−7.5 kg; −12 to −2.8), topiramate 200 mg (−7 kg; −10 to −3.4), metformin 750 mg (−5.7 kg; −9.3 to −2.1), topiramate 100 mg (−5.7 kg; −8.8 to −2.5), topiramate 50 mg (−5.2 kg; −10 to −0.57), liraglutide 1.8 mg (−5.2 kg; −10., −0.080), sibutramine 15 mg (−4.5 kg; −8.9 to −0.59), nizatidine 300 mg (−3.0 kg; −5.9 to −0.23) and metformin 1000 mg (−2.3 kg; −4.6 to −0.0046). There was no effect of duration of follow-up, baseline body weight and, preventive versus therapeutic strategy on weight reduction in AIWG.

Conclusion

Metformin 750 mg with lifestyle modification was the most effective treatment for AIWG, followed by topiramate 200 mg, metformin 750 mg, and topiramate 100 mg with moderate certainty of evidence.