Search for a command to run...
Eemil Partinen,1– 4 Ari Ylikoski,5 Mariusz Sieminski,6 Madli Kaps,7 Markku Partinen,2,4 Christer Hublin8 1Sleep Disorder Outpatient Clinic, Department of Psychiatry, Helsinki University Central Hospital, Helsinki, Finland; 2Helsinki Sleep Clinic, Terveystalo Healthcare, Helsinki, Finland; 3Department of Neurology, Helsinki University Central Hospital, Helsinki, Finland; 4Department of Clinical Neurosciences, Clinicum, University of Helsinki, Helsinki, Finland; 5The Social Insurance Institution of Finland (KELA), Helsinki, Finland; 6Department of Emergency Medicine, Medical University of Gdansk, Gdansk, Poland; 7Department of Psychiatry, Helsinki University Central Hospital, Helsinki, Finland; 8Finnish Institute of Occupational Health, Helsinki, FinlandCorrespondence: Eemil Partinen, Email eemil.partinen@helsinki.fiAim: Sleep disturbances are a common non-motor symptoms in Parkinson’s disease (PD) and may occur already in the prodromal phase. In the general population, shorter sleep duration is linked to higher mortality, while longer sleep duration has been associated with an increased risk of developing PD. We examined the relationships between sleep duration, chronotype, and mortality in patients with PD.Methods: A total of 855 randomly selected patients with PD completed a structured questionnaire including self-reported sleep duration and chronotype. Mortality data were collected from the national death registry.Results: Altogether 435 subjects (238 men; 54.7%) were included. The mean follow-up time was 4.3 (0.3– 7.0) years, during which 99 participants had died. In a univariable model, longer sleep duration was associated with higher mortality [HR 1.31, 95% confidence interval (1.15– 1.49), P< 0.0001]. In a fully adjusted model, longer sleep duration [HR 1.17 (1.01– 1.35), P=0.035], older age, male gender, PD duration and depression were associated with higher mortality. BMI and high physical activity were associated with lower mortality. Although short sleep showed an association with mortality in univariable analyses, this association did not persist after multivariable adjustment [HR 0.82 (0.52– 1.30), P=0.401]. Chronotype was not associated with mortality. Long sleepers were older, had lower mean levodopa-dose, took more naps, and were sleepier than short sleepers.Conclusion: Self-reported longer sleep duration was associated with higher mortality in patients with Parkinson’s disease, whereas chronotype was not associated with mortality. This does not imply a protective effect of shorter sleep duration. Further studies are needed to clarify whether sleep duration reflects disease-related factors, behavioral patterns, or progression of Parkinson’s disease.Keywords: Parkinson’s disease, mortality, sleep duration, sleep apnea, restless legs, sleep length, chronotype, morningness, eveningness