How-To Guides

Amitriptyline, mirtazapine, and paroxetine were the antidepressants most associated with weight gain in this meta-analysis. Amitriptyline was significantly associated with weight gain during both acute treatment and maintenance treatment, mirtazapine was significantly associated with weight gain during acute treatment and maintained a similar trend over maintenance, and paroxetine shifted from marginal weight loss during acute treatment to weight gain over the medium-long term, with significant weight gain observed after 8 months.

During acute treatment, citalopram, fluoxetine, sertraline, venlafaxine, duloxetine, bupropion, and moclobemide were associated with weight loss to varying degrees. Over the longer term, only bupropion maintained a significant weight-lowering effect; the other antidepressants generally did not show a sustained effect on body weight during maintenance treatment. The authors also noted that fluoxetine's weight-loss effect appeared limited to the acute phase.

Yes. In this analysis, paroxetine was associated with a marginal weight loss during acute treatment but with weight gain over the medium-long term. A significant weight gain was observed after 8 months of treatment, making paroxetine the SSRI in this review most clearly associated with long-term weight increase.

Fluoxetine was associated with weight loss during acute treatment, but the effect did not appear to persist. The authors found that fluoxetine's apparent anorexigenic effect was limited to the acute phase, with no significant long-term weight effect seen during maintenance treatment. They noted that some restoration of weight may occur over time.

Bupropion stood out as the only antidepressant in this meta-analysis with a significant weight-lowering effect that persisted beyond the acute phase. While several SSRIs and SNRIs were associated with short-term weight loss, only bupropion maintained a significant effect on weight during longer-term treatment. The review therefore identified bupropion as clinically distinct from most other antidepressants with respect to body weight.

No. This meta-analysis found meaningful differences among SSRIs. Fluoxetine was associated with acute weight loss that appeared transient, paroxetine was associated with long-term weight gain, and citalopram and sertraline showed only small short-term weight loss without a significant maintenance effect. The authors concluded that SSRIs overall have a small effect on body weight, but fluoxetine and paroxetine differed from the rest in clinically important ways.

No significant effect of imipramine on body weight was found in this meta-analysis. The authors reported no significant body weight effect during either acute or maintenance treatment, even though imipramine is often clinically associated with weight gain. They described its weight effect as less clear than that of amitriptyline.

This was a systematic review and meta-analysis of 116 eligible studies identified from MEDLINE, ISI Web of Knowledge, and Cochrane databases through January 2009. The authors included adult patients receiving antidepressant monotherapy for at least 4 weeks and analyzed acute treatment separately from maintenance treatment, defining acute treatment as 4 to 12 weeks and maintenance treatment as more than 4 months. Because only 30 of the 116 trials were placebo controlled, non-placebo studies were compared with a virtual placebo sample derived from the weighted mean of all placebo samples.

The authors found that estimates from observational non-placebo studies were similar to those from randomized placebo-controlled studies for the drugs with enough placebo-controlled data to test. There were no significant differences for fluoxetine (P = .59), bupropion (P = .30), sertraline (P = .50), imipramine (P = .99), duloxetine (P = .15), paroxetine (P = .83), or amitriptyline (P = .98). On that basis, they concluded their method using virtual placebo comparisons produced reasonably consistent estimates, although they still discussed this as a limitation.

The main limitations were heterogeneity across studies, possible publication bias, and limited data for several individual drugs. Only 30 of 116 selected trials were placebo controlled, some drug-specific analyses were based on very few studies, and many trials did not consistently report weight change as a side effect. The authors also noted that they could not account for potentially important patient-level factors such as depressive severity, appetite loss, atypical features, premorbid weight, sex, or age.