.png){kind=icon}
.png){kind=icon}
Return
Multidimensional pain patterns in Parkinson’s disease: a longitudinal mobile health study
.png){kind=icon}
OA
AI
E
A
A
F
M
C
F
S
R
DOI：10.1007/s00702-026-03250-9.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Pain is a non-motor symptom of Parkinson’s disease (PD). In clinical practice, it is often reduced to single metrics, such as intensity, omitting important aspects of its complexity. Mobile health tools enable repeated recording and multidimensional analysis of pain. To examine how pain intensity, duration, and spatial distribution differentiate patient profiles using longitudinal mobile health data. We analyzed 20,971 daily records from 68 PD patients over 14 months. Patients reported pain intensity, episode duration, and pain location. Data were aggregated at the individual level to avoid pseudo-replication. Profiles were identified using unsupervised clustering (HDBSCAN with KMeans fallback). Associations between pain dimensions were assessed using Spearman correlation. Two profiles were identified: a high-burden group (n = 17) and a low-burden group (n = 50). One additional cluster (n = 1) was excluded due to instability. Differences between groups were greater for cumulative pain duration (6.7-fold) and spatial extent (9–11-fold) than for mean intensity (2.4-fold). Mean pain intensity was moderately correlated with cumulative duration (ρ = 0.66) and spatial extent (ρ = 0.60 anterior; ρ = 0.44 posterior), whereas mean episode duration showed weak and non-significant associations with other pain metrics. Recording frequency was not associated with cluster membership (r = 0.16, p = 0.20), and both groups showed sustained engagement. Pain intensity alone may not adequately reflect overall burden in PD. Considering temporal persistence and spatial distribution provides a more complete view of patient experience and supports a multidimensional approach to assessment.
Keywords:
Parkinson’s disease
Chronic pain
Mobile health
Unsupervised learning
Pain distribution
Journal
IF:
4
Papers:
4.8K
Citations:
8.5K
Organization
Cited Papers