Seven trials were included in the review (282 participants; calculated as 236 from Table 1, with 45 crossover participants included twice, producing 281 total; ranging from 11 twice to 34 twice or 20 to 57). There were five parallel-group trials, one double-arm crossover trial, and one single-arm crossover trial. All of the trials had low risk of bias for selective reporting. Most had an unclear risk of bias for most other domains, and a low risk of other bias.
No statistically significant differences between treatment and control groups were observed for the mean changes in plasma concentrations of total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, and triglycerides.
The results were similar in post hoc subgroup analyses for coronary heart disease risk (high or low), and treatment duration (three months or more, or less than three months). The removal of one trial, in a sensitivity analysis, showed a significant reduction of total cholesterol plasma concentration in the treatment groups compared with controls (WMD -14.58, 95% CI -28.47 to -0.69; four trials; Ι²=0). Meta-regression revealed no significant dose-response associations for the effects of resveratrol on plasma lipid concentrations.
Compared with controls, a statistically significant greater increase in plasma concentration of free fatty acids was observed with resveratrol (WMD 0.09, 95% CI 0.02 to 0.16; three trials; Ι²=0).
Some evidence of publication bias was found (reported fully in the paper).