Objective: To analyze the change in health-related quality-of-life (HRQoL) in patients with Parkinson’s disease (PD) and to identify different classes of HRQoL decline. Methods: A longitudinal cohort study was performed to assess clinical parameters (unified PD rating scale, Beck Depression Inventory) and HRQoL data (EuroQol, Parkinson’s Disease Questionnaire [PDQ]-39) at baseline, 3, 6, 12, 24, and 36 months. A total of 145 patients with PD were consecutively recruited in the county of Northern Hessia, Germany, between January and June 2000. A latent growth mixture model was applied to analyse the heterogeneity in HRQoL trajectories. Results: We successfully applied latent mixture growth modeling in order to identify different classes of HRQoL trajectories in PD. Three growth models were developed and each resulted in a four-class model of distinct patterns using the generic EuroQol instruments’ outcomes (EuroQol-5 Dimensions and visual analogue scale) and the disease-specific PDQ- 39. The four classes were defined by individual trajectory characteristics. Classes one and two represented trajectories with moderate declines over 36 months, but with different initial intercepts. Class three consisted mainly of patients who passed away during the observation period and therefore had a large HRQoL decline. Class four was characterized by a low level of HRQoL at baseline and a significant subsequent decline. Conclusions: The findings provide a more elaborate understanding of the variability in HRQoL reduction in PD over time. The classification of different HRQoL subgroups may help to explain the response of PD patients to the natural history of the disease. Future research will enable the identification of HRQoL responder subgroups on different treatment regimens.

Objective: To analyze the change in health-related quality-of-life (HRQoL) in patients with Parkinson’s disease (PD) and to identify different classes of HRQoL decline. Methods: A longitudinal cohort study was performed to assess clinical parameters (unified PD rating scale, Beck Depression Inventory) and HRQoL data (EuroQol, Parkinson’s Disease Questionnaire [PDQ]-39) at baseline, 3, 6, 12, 24, and 36 months. A total of 145 patients with PD were consecutively recruited in the county of Northern Hessia, Germany, between January and June 2000. A latent growth mixture model was applied to analyse the heterogeneity in HRQoL trajectories. Results: We successfully applied latent mixture growth modeling in order to identify different classes of HRQoL trajectories in PD. Three growth models were developed and each resulted in a four-class model of distinct patterns using the generic EuroQol instruments’ outcomes (EuroQol-5 Dimensions and visual analogue scale) and the disease-specific PDQ- 39. The four classes were defined by individual trajectory characteristics. Classes one and two represented trajectories with moderate declines over 36 months, but with different initial intercepts. Class three consisted mainly of patients who passed away during the observation period and therefore had a large HRQoL decline. Class four was characterized by a low level of HRQoL at baseline and a significant subsequent decline. Conclusions: The findings provide a more elaborate understanding of the variability in HRQoL reduction in PD over time. The classification of different HRQoL subgroups may help to explain the response of PD patients to the natural history of the disease. Future research will enable the identification of HRQoL responder subgroups on different treatment regimens.