Background Fluoxetine is a serotonin reuptake inhibitor indicated for major depression. It is also thought to affect weight control: this seems to happen through appetite changes resulting in decreased food intake and normalisation of unusual eating behaviours. However, the benefit-risk ratio of this off-label medication is unclear. Objectives To assess the effects of fluoxetine for overweight or obese adults. Search methods We searched the Cochrane Library, MEDLINE, Embase, LILACS, the ICTRP Search Portal and ClinicalTrials.gov and World Health Organization (WHO) ICTRP Search Portal. The last date of the search was December 2018 for all databases, to which we applied no language restrictions. Selection criteria We included randomised controlled trials (RCTs) comparing the administration of fluoxetine versus placebo, other anti-obesity agents, non-pharmacological therapy or no treatment in overweight or obese adults without depression, mental illness or abnormal eating patterns. Data collection and analysis Two review authors independently screened abstracts and titles for relevance. Screening for inclusion, data extraction and risk of bias assessment was performed by one author and checked by the second. We assessed trials for the overall certainty of the evidence using the GRADE instrument. For additional information we contacted trial authors by email. We performed random-effects meta-analyses and calculated the risk ratio (RR) with 95% confidence intervals (95% CI) for dichotomous outcomes and the mean difference (MD) with 95% CI for continuous outcomes. Main results We identified 1036 records, scrutinized 52 full-text articles and included 19 completed RCTs (one trial is awaiting assessment). A total of 2216 participants entered the trials, 1280 participants were randomly assigned to fluoxetine (60 mg/d, 40 mg/d, 20 mg/d and 10 mg/d) and 936 participants were randomly assigned to various comparison groups (placebo; the anti-obesity agents diethylpropion, fenproporex, mazindol, sibutramine, metformin, fenfluramine, dexfenfluramine, fluvoxamine, 5-hydroxy-tryptophan; no treatment; and omega-3 gel). Within the 19 RCTs there were 56 trial arms. Fifteen trials were parallel RCTs and four were cross-over RCTs. The participants in the included trials were followed up for periods between three weeks and one year. The certainty of the evidence was low or very low: the majority of trials had a high risk of bias in one or more of the risk of bias domains. For our main comparison group - fluoxetine versus placebo - and across all fluoxetine dosages and durations of treatment, the MD was -2.7 kg (95% CI -4 to -1.4; P < 0.001;10 trials, 956 participants; low-certainty evidence). The 95% prediction interval ranged between -7.1 kg and 1.7 kg. The MD in body mass index (BMI) reduction across all fluoxetine dosages compared with placebo was -1.1 kg/m' (95% CI -3.7 to 1.4; 3 trials, 97 participants; very low certainty evidence). Only nine placebo-controlled trials reported adverse events. A total of 399 out of 627 participants (63.6%) receiving fluoxetine compared with 352 out of 626 participants (56.2%) receiving placebo experienced an adverse event. Random-effects meta-analysis showed an increase in the risk of having at least one adverse event of any type in the fluoxetine groups compared with placebo (RR 1.18, 95% CI 0.99 to 1.42; P = 0.07; 9 trials, 1253 participants; low-certainty evidence). The 95% prediction interval ranged between 0.74 and 1.88. Following fluoxetine treatment the adverse events of dizziness, drowsiness, fatigue, insomnia and nausea were observed approximately twice as often compared to placebo. A total of 15 out of 197 participants (7.6%) receiving fluoxetine compared with 12 out of 196 participants (6.1%) receiving placebo experienced depression. The RR across all fluoxetine doses compared with placebo was 1.20 (95% CI 0.57 to 2.52; P = 0.62; 3 trials, 393 participants; very low certainty evidence). All-cause mortality, health-related quality of life and socioeconomic effects were not reported. The comparisons of fluoxetine with other anti-obesity agents (3 trials, 234 participants), omega-3 gel (1 trial, 48 participants) and no treatment (1 trial, 60 participants) showed inconclusive results (very tow certainty evidence). Authors' conclusions Low-certainty evidence suggests that off-label fluoxetine may decrease weight compared with placebo. However, low-certainty evidence suggests an increase in the risk for dizziness, drowsiness, fatigue, insomnia and nausea following fluoxetine treatment.
