Ten RCTs (n=3,320 participants) were included in the review. The sample sizes ranged from 22 to 1,273 participants. All the included trials reported concealment of randomisation and blinding of health care providers, data collectors and outcome assessors. Two trials were stopped prematurely by the sponsors.
The participants treated with topiramate lost a mean of 5.34kg (95% CI -6.12 to -4.56) of additional weight compared with participants who received placebo, regardless of dose and duration of treatment. There was substantial heterogeneity present (I2=75.5%).
The likelihood of significant weight loss was higher in the topiramate group compared with placebo; higher proportions of participants who received topiramate achieved weight loss of 5% or more (OR 6.02, 95% CI 4.81 to 7.53; NNT 2.63; I2=31.1%) and 10% or more of baseline body weight (OR 7.16, 95% CI 5.48 to 9.36; NNT 3.7; I2=17.8%).
The results of meta-regression analyses found that the effect of topiramate on weight loss increased with treatment duration and dosage.
Participants with diabetes who were treated with topiramate lost a mean of 5.25 kg (95% CI -6.66 to -3.85 kg; four trials) of additional weight compared with placebo-treated participants, but substantial heterogeneity was observed for this outcome (I2=78.3%).
In the analysis of adverse events, participants treated with topiramate had an increased risk of paraesthesia (OR 8.70, 95% CI 6.90 to 11.0; I2=58.5%) and of adverse events leading to withdrawal from treatment compared with placebo-treated participants (OR 1.94, 95% CI 1.64 to 2.29; I2=0; NNH 13.7 patients). There were higher risks of taste perversion (OR 8.61, 95% CI 5.35 to 13.87; I2=25.9%), psychomotor impairment (OR 7.82, 95% CI 3.71 to 16.46; I2=0) and hypoaesthesia (OR 4.51, 95% CI 2.76 to 7.40; I2=0) and various other adverse events in the topiramate group.
There was no evidence of publication bias from the appraisal of funnel plots or the Egger's test.