Case Control Polysomnographic Studies of Sleep Disorders in Parkinson's Disease

BACKGROUND: The relationship between a number of primary sleep disorders and Parkinson's disease (PD) is still debated. There are limited case control polysomnographic studies in PD and most of these study sample sizes are small. METHODOLOGY/FINDINGS: We conducted one of the largest case-control studies involving overnight polysomnographic evaluation, with prospective recruitment of unselected Parkinson's disease patients and healthy controls from an Asian population. The cases were recruited from the specialized movement disorder outpatient clinics in a tertiary referral center, and controls from the same geographical locations. All subjects underwent an overnight polysomnographic study and a multiple sleep latency test. A total of 124 subjects including 56 patients and 68 controls frequency-matched for age and sex were included. Multivariate analysis revealed that patients had significantly shorter total sleep time than controls (p = 0.01), lower sleep efficiency (p = 0.001) and increased REM latency (p = 0.007). In patients, multivariate analysis showed that reduced total sleep time was significantly associated with increased age (p = 0.001) and increased levodopa dose (p = 0.032). The mean Insomnia Severity Index was higher in PD patients (9.0±7.1) compared to controls (3.3±3.9, p<0.001). The mean Epworth Sleepiness Scale score was higher in PD patients (9.3±5.9 vs. 5.7±4.8, p<0.001). Nocturnal arousals, obstructive sleep apnea, periodic leg movements and objective abnormal sleepiness were not increased in our patients. CONCLUSIONS/SIGNIFICANCE: Our case-control polysomnographic study, the first-ever performed in an Asian population, revealed altered sleep architecture and reduced sleep in PD patients compared to controls. Reduced total sleep time was associated with increased age and levodopa dose. However, nocturnal arousals, primary sleep disorders and abnormal sleepiness were not increased in our PD patients suggesting that ethnic/genetic differences may be a factor in the pathophysiology of these conditions.