LAUSR

OBJECTIVE To assess the association between REM sleep behavior disorder (RBD) and other determinants and incident impulse control disorder behaviors (ICBs) in patients with early Parkinson's disease (PD) using longitudinal data from the Parkinson's Progression Markers Initiative (PPMI). METHODS 401 newly-diagnosed PD patients were prospectively evaluated at baseline (BL), month 6, and annually for 5 years. Probable RBD (pRBD) was assessed with the RBD Screening Questionnaire and dichotomized using a cut-off value ≥6. The association of BL and time-dependent (TD) pRBD and other covariates with the development of ICB symptoms, was evaluated using Cox proportional hazards regression and general estimating equations logistic regression. Models considered adjustment for age, sex, MDS-UPDRS III, Geriatric Depression Scale (GDS-15), RBD medication use, total levodopa equivalent daily dose (LEDD), and dopamine agonist (DA) and antidepressant medication use. RESULTS Both baseline pRBD and TD pRBD were not associated with an increased risk for incident ICB symptoms after adjustment for covariates (adjusted HR=1.17, p=0.458 and HR=1.27, p=0.257 respectively). In a modified-TD pRBD model (i.e., considering subjects as pRBD onward from the first time point with RBDSQ score ≥6), the risk for incident ICB symptoms was higher in pRBD in unadjusted (HR=1.48, p=0.038), but not adjusted (HR=1.29, p=0.202) models. TD DA use (HR=1.64, p=0.039), TD GDS-15 score (HR=1.12, p<0.001), and male sex (year 3: HR=2.10,p=0.009; year 4: HR=3.04,p=0.006; year 5: HR=4.40,p=0.007) were associated with increased ICB symptom risk. INTERPRETATION pRBD is not clearly associated with ICB symptom development in early PD, in contrast to DA use, depression, and male sex. This article is protected by copyright. All rights reserved.