BackgroundAlthough some data exist describing the use of topiramate in patients with antipsychotic-induced obesity, direct comparision with metformin is limited. The purpose of this study was to explore the effectiveness and safety of concomitant topiramate on antipsychotic-induced obesity and compare with metformin.Methods62 stabilized outpatients with antipsychotic-induced obesity were randomized to 16-week treatment group of topiramate or metformin. The patients’ weight, body mass index (BMI), waist-hip ratio and side effects were assessed and compared. Intention-to-treat and completer analyses were performed. Meanwhile, Covariance analysis was conducted to control the impact of the significant difference in BMI when comparison between the two groups.ResultsTwo groups had comparable characteristics though the difference in baseline BMI was significant. Intention-to-treat analyses: The random missing values were replaced using the last observation carried forward method when intention-to-treat analyses. Compared with the baseline, the weight, BMI and waist-hip ratio with topiramate markedly decreased at each follow-up, whereas, only waist-hip ratio with metformin significantly decreased at 4-week. Compared with metformin,only weight and BMI with topiramate significantly decreased at 4-week, and all weight, BMI and waist-hip ratio at week 8–16 also remarkably declined. Completer analyses: Compared with the baseline, the weight, BMI and waist-hip ratio with topiramate at week 4–16 markedly decreased, whereas, only waist-hip ratio with metformin significantly decreased at 4-week. Compared with metformin, all BMI with topiramate markedly decreased at week 4–16, moreover weigh and waist-hip ratio also notably lowered at 8-week. No significant differences in adverse events were found between the two groups.ConclusionsTopiramate, similar to metformin in reducing obesity as previously reported, also significantly reduced body weight, BMI and waist-hip ratio in patients with antipsychotic-induced obesity and demonstrated well tolerance in psychiatric patients.The trial was registered at http://www.chictr.org.cn, and the number was ChiCTR-IPR-17013122.