UC San Diego

OBJECTIVE: To examine the association of sleep disturbance with Parkinson disease (PD) during 10+ years of follow-up among postmenopausal women, 50 to 79 years of age at baseline. METHODS: Longitudinal data on 130,502 study-eligible women (mean ± standard deviation baseline age = 63.16 ± 7.20 y) from the Womens Health Initiative Clinical Trials and Womens Health Initiative Observational Study were analyzed. The cohort was followed for 15.88 ± 6.50 years, yielding 2,829 (2.17%) PD cases. Sleep disturbance (habitual sleep duration, insomnia symptoms, obstructive sleep apnea risk factors, sleep aids among those with WHI Insomnia Rating Scale scores (WHIIRS) > 9) was measured at baseline and one follow-up time by September 12, 2005. Cox proportional hazards models evaluated relationships controlling for sociodemographic, lifestyle, and health characteristics. RESULTS: PD was significantly associated with long sleep duration (≥9 h) versus a benchmark of 7 to 8 hours (hazard ratio [HR] = 1.296, 95% confidence interval [CI]: 1.153-1.456), WHIIRS (>9 vs ≤9) (HR = 1.114, 95% CI:1.023-1.214), and use of sleep aids (yes vs no) (HR = 1.332, 95% CI:1.153-1.539) among those with WHIIRS > 9. Compared with 7 to 8 hours, short (<7 h) sleep duration was unrelated to PD. Finally, the presence of obstructive sleep apnea risk factors was not associated with PD. CONCLUSIONS: Among postmenopausal women, sleep disturbance was associated with approximately 10% to 30% increased PD risk after ∼16 years follow-up. Prospective cohort studies with objective exposures and adjudicated outcomes that include men and women of diverse backgrounds are required to confirm and extend these findings.