Thirteen trials (around 3,600 patients, range six to 1,176) were included in the review. Several trials had associated publications of sub-studies. Three trials were non-randomised open studies, five trials used low risk methods for sequence generation and methods in six trials were unclear. Risk of biases due to inadequate blinding were generally low (except in the open studies). All trials had low risk of bias from incomplete outcome data but judgements on selective reporting were unclear for most trials.
Rasagiline monotherapy 1mg/day in early stages of the disease significantly reduced the UPDRS score when compared with placebo (mean difference -3.06, 95% CI -3.81 to -2.31; three studies; Ι²=3%); a similar result was seen when pooling the same three studies for the 2mg/day data. Rasagiline also reduced the UPDRS score in trials with a delayed-start design (mean difference -0.89, 95% CI -1.78 to 0.00; Ι²=0%; two studies) although there was significant heterogeneity of effect when the analyses were subgrouped by dose.
Compared with placebo, 1mg/day of rasagiline plus levodopa significantly reduced off-time (mean difference -0.93 hours, 95% CI -1.17 to -0.69; three studies; Ι²=0%). A smaller reduction was seen in the one study that used a dose of 0.5mg/day.