Protocol: Using N-of-1 tests to identify responders to melatonin for sleep disturbance in Parkinson's disease

Abstract

Background40% of Parkinson's Disease (PD) sufferers experience insomnia, impacting health and quality of life for patients and family members, especially carers. There is little evidence that current treatments are effective. ObjectivesTo determine the effectiveness of melatonin in reducing insomnia in 44 individuals with PD using N-of-1 trials. To aggregate group data to arrive at population estimates of effectiveness (measured by improvements in PDSS-2) and safety (measured by adverse events) of melatonin in improving insomnia in PD. To assess the feasibility of offering N-of-1 trials for insomnia in PD. MethodologyParticipants will receive either immediate-release melatonin or placebo in random order in 3 paired two-week treatment periods (12 weeks total). Based on their response in a two-week run-in period on 3 mg daily, they will trial either 3 mg or 6 mg. Patients will keep daily sleep diaries and wear a MotionWatch throughout. After the trial patients will discuss their individual report with their doctor, which provides direct feedback about effectiveness and safety of melatonin for them. Statistical methodsWe will analyse N-of-1 tests 1) individually: effects of melatonin on PDSS-2 and safety will be reported; and 2) aggregated across individual N-of-1 studies, combined using a Bayesian multilevel random effects model, which will account for repeated measures on individuals over time, and will return posterior estimates of overall treatment effect, and effect in each individual. Clinical Trial Registration numberACTRN12617001103358.