Two placebo-controlled crossover randomised controlled trials (RCTs) were included (n=19 and n=24, respectively).
The included studies exhibited several methodological limitations: small sample size, short duration of treatment, older age of study populations and the use of means to report AIMS scores.
One crossover RCT (19 patients) found no statistically significant difference between baseline and end-of-treatment AIMS scores (change of less than 1 point with each treatment), or between baseline and end of washout, for either melatonin or placebo.
The second crossover RCT (24 patients randomised, two withdrew before the start of treatment) found that 10 mg/day melatonin significantly decreased the AIMS score compared with placebo: decrease of 2.45 points with melatonin versus 0.77 points with placebo (P=0.001). Melatonin significantly increased the proportion of patients with a decrease in AIMS score of more than 3 points versus placebo (7 patients with melatonin versus 1 patient with placebo, P<0.001).
No adverse effects were reported in either RCT, and there were no withdrawals due to adverse effects.