Clinical Question: Which dietary supplements have evidence of promoting sustained weight loss?
Setting: Outpatient (any)
Study Design: Systematic review
Synopsis: The authors of this review searched for double-blind randomized trials, systematic reviews, and meta-analyses of alternative medicine products used to promote weight loss. They also searched medical journals and examined reference lists of included studies. They did not limit articles to any language, which is a benefit because many of these studies are published in languages other than English. They identified five systematic reviews and meta-analyses and 25 additional randomized trials not included in the reviews. Overall, the studies of the various products were of short duration (i.e., one to three months), whereas studies of at least six months to one year are needed to establish benefit.
In the short-term studies, the authors found that an ayurvedic herbal preparation (ingredients not specified) caused an approximately 8 kg (17 lb, 10 oz) weight loss, which was significantly higher than placebo. According to the studies, chitosan did not have any effect over placebo. Chromium picolinate was ineffective in a meta-analysis of 10 studies. Ephedra (now removed from the U.S. market) resulted in a 1 kg (2 lb, 3 oz) weight loss over placebo. Garcinia cambogia, a source of hydroxycitric acid, produced mixed results across several studies. Glucomannan produced an effect greater than placebo in a single study. The fiber products guar gum and psyllium were found to be ineffective. Hydroxymethylbutyrate, or pyruvate, which is used by bodybuilders, may have an effect on increasing lean body mass. Yerba maté, a source of caffeine, has little proof of effectiveness, and yohimbe has conflicting evidence when used with a low-calorie diet.
Bottom Line: The evidence is weak that any of the commonly used alternative products are effective for reducing weight in moderately overweight individuals. None of the products has been studied for longer than three months. (Level of Evidence: 1a)