Introduction: Pain is a common and complex symptom in Parkinson's disease. The underlying mechanisms and longitudinal patterns are not well understood, which impedes therapeutic decision making. The objectives of this study were to characterize longitudinal pain trajectories, identify clusters (subgroups) with similar patterns, and examine associations with sociodemographic and clinical characteristics.
Methods: Latent class growth analysis was applied to 16,863 people with Parkinson's disease stratified by early (N = 8612; <3 years), mid (N = 6181; 3-10 years) and later (N = 2070; >10 years) disease duration over ∼4.5 years (2017-2021) using the Fox Insight Data Exploration Network, to discern clusters of individuals with similar longitudinal patterns of self-reported pain. Associations were evaluated between cluster membership and sociodemographic and clinical factors.
Results: Across the disease duration strata, five clusters were identified. The clusters ranged from none to moderate pain, with a small cluster of subjects with severe pain. The percentage of subjects with moderate (early = 17.3%, mid = 24.2%, later = 34.4%) and severe (early = 2.3%, mid = 4.4%, later = 6.5%) pain at baseline increased across disease duration groups. The trajectories tended to be variable or slightly worsening in the early duration group, more stable in the mid duration group, and improving in the later duration group. Across strata, the clusters with moderate to severe pain were associated with more severe impairment, depression, anxiety and arthritis, higher body mass index, lower income, and lower education.
Conclusion: This latent class growth analysis, applied to people with Parkinson's disease, provides a template for using self-reported outcomes to improve our understanding of pain trajectories.
Keywords: Heterogeneity; Latent class growth analysis; Pain; Parkinson's disease; Patient reported outcomes; Trajectories.
Copyright © 2022 The Authors. Published by Elsevier Ltd.. All rights reserved.