Progression in Parkinson's disease: variation in motor and non-motor symptoms severity and predictors of decline in cognition, motor function, disability, and health-related quality of life as assessed by two different methods

Background: Parkinson's disease (PD) is multi-symptom disease with variable progression. Objectives: We performed a longitudinal study to address the evolution of motor symptoms (MS) and non-motor symptoms (NMS), predictors of motor-, cognitive-, disability-, and health-related quality of life (HRQL) status and the relative usefullness of a battery of separate NMS scales (BSS) versus the Non-Motor Symptom Scale (NMSS). Methods: Seventy-two patients were assessed at baseline and 4 years later with the NMSS and BSS. We assessed the following outcomes: cognition (Montreal Cognitive Assessment scale [MoCA]), disability (Unified Parkinson's Disease Rating Scale Part II [UPDRS II], Schwab and England [S&E]), motor dysfunction (Unified Parkinson's Disease Rating Scale Part III [UPDRS III], Hoehn and Yahr [HY]), and HRQL (EuroQol [EQ] EQ-vertical visual analogue scale [VAS] and EQ-Index). Statistical analysis included a comparison between scales scores at both time points and multivariate regression analysis to calculate the impact of each baseline symptom in outcomes. NMSS and BSS were introduced in separate models. Results: NMSS Domain 4: perception/hallucinations, Parkinson's Psychosis Questionnaire, Apathy Scale, NMSS Domain 7: urinary, S&E, UPDRS II, HY, and MoCA scores worsened significantly. Dementia increased to a 4-year prevalence of 39.8%. In the multivariate model using BSS, cognitive state variation was significantly predicted by baseline HY, EQ-Index, and S&E. Using the NMSS, MoCA change was significantly associated with NMSS Domain 4: perceptions/hallucination score, cognitive status with UPDRS III score, HRQL with NMSS Domain 4: perception/hallucinations score, and S&E. Conclusion: Our study suggests that NMS progress heterogeneously, BSS approach being more sensitive to change than NMSS. The multivariate analysis has shown that S&E and NMSS Domain 4: perception/hallucinations scores are the stronger predictors of HRQL and cognitive dysfunction variation, favoring NMSS over the BSS approach.