Purpose: To investigate the effect of garlic on total cholesterol level in persons with elevated levels by conducting a meta-analysis of randomized, double-blind, placebo-controlled trials.

Data sources: Systematic literature searches were conducted on the MEDLINE, EMBASE, BIOSIS, Cochrane Library, AMED, and CISCOM databases. Manufacturers of commercial garlic preparations and experts in the field were asked about published or unpublished trials.

Study selection: Selected trials were required to state that they were randomized, double-blind, and placebo-controlled; use garlic monopreparations; include persons with mean total cholesterol levels of at least 5.17 mmol/L (200 mg/dL); and report total cholesterol level as an end point. There were no language restrictions.

Data extraction: Two reviewers, blinded to key identifiers of each paper, independently extracted data in a standardized manner and assessed methodologic quality by using the Jadad scale. Discrepancies were settled through discussion.

Data synthesis: In the 13 trials included in the meta-analysis, garlic reduced total cholesterol level from baseline significantly more than placebo (P < 0.01); the weighted mean difference was -0.41 mmol/L (95% CI, -0.66 to -0.15 mmol/L) (-15.7 mg/dL [CI, -25.6 to -5. 7 mg/dL]). Six diet-controlled trials with the highest scores for methodologic quality revealed a nonsignificant difference between garlic and placebo groups; the weighted mean difference was -0.11 mmol/L (CI, -0.30 to 0.08 mmol/L) (-4.3 mg/dL [CI, -11.7 to 3.1 mg/dL]).

Conclusions: The available data suggest that garlic is superior to placebo in reducing total cholesterol levels. However, the size of the effect is modest, and the robustness of the effect is debatable. The use of garlic for hypercholesterolemia is therefore of questionable value.