Design: Systematic review and meta-analysis of 29 RCTs
Population: 1,913 adults with Parkinson disease and sleep disturbance
Primary outcome: Sleep quality, daytime sleepiness, insomnia and RBD measures
Effect: PSQI MD 2.66 (95% CI 1.63–3.69); EDS, insomnia, RBD non-significant

This systematic review pooled 29 randomised trials (1,913 participants) of drugs for sleep disturbance in Parkinson disease, analysed by phenotype. Only 17% of trials were at low risk of bias.

For overall sleep quality, drug treatment improved Pittsburgh Sleep Quality Index scores (mean difference 2.66, 95% CI 1.63–3.69; I² 71%), with melatonin-type chronobiotics and dopamine agonists contributing most. For excessive daytime sleepiness, insomnia and REM sleep behaviour disorder, pooled estimates were not significant and heterogeneity was so high (I² 91–99%) that the authors treated them as exploratory.

The practical conclusion is that sleep in Parkinson disease is common, disabling and poorly served by trials. Treatment should be matched to the specific problem — nocturnal akinesia, RBD, insomnia, daytime sleepiness — rather than a general 'sleep' drug.

Ask specifically about insomnia, dream enactment, daytime sleepiness and night-time immobility.
Consider nocturnal motor symptoms first — a bedtime controlled-release levodopa or agonist may help sleep more than a hypnotic.
Melatonin is a low-risk option for sleep quality and RBD.
Review sedating drugs and dopamine agonists as causes of daytime sleepiness.

Why it matters

It shows how little trial evidence underlies common prescribing for sleep in Parkinson disease.

The statistics, in plain English

A 2.7-point improvement on the 21-point PSQI is noticeable but modest, and the evidence is rated low certainty. When I² exceeds 90%, trials disagree so much that a pooled average is not meaningful.