Cluster‐Defined Response Patterns Following Pain Neuroscience Education With Physical Activity in Older Women With Chronic Low Back Pain: A Quasi‐Randomized Exploratory Study

ABSTRACT Background and Aims To explore patterns of treatment response to pain neuroscience education combined with physical activity (PNE‐PA) and to examine baseline characteristics associated with these patterns in older women with chronic low back pain (CLBP). Methods In a community‐based quasi‐randomized trial, older women with CLBP were analyzed in a PNE‐PA program ( n = 43) or a biomechanical (BM) comparison group ( n = 13). Physical function, physical activity (daily step counts by pedometer), pain intensity, disability, and pain‐related psychological outcomes were assessed at baseline and 12 weeks. Between‐group effects were examined using rank‐based analysis of covariance adjusting for baseline values, age, and body mass index. Within the PNE‐PA arm, k‐means clustering on changes in 30‐s chair‐stand performance and pain catastrophizing (PCS) was used to identify response profiles, with baseline characteristics compared across clusters. Results Compared with BM, PNE‐PA produced larger adjusted improvements in chair‐stand repetitions, daily step counts, and PCS scores; no significant between‐group differences were observed for other outcomes. Three clusters were identified: minimal responders ( n = 15), physical‐function responders ( n = 18), and psychological responders ( n = 10). Physical‐function responders showed greater chair‐stand gains with lower baseline lower‐limb function; psychological responders exhibited larger PCS reductions and lower baseline physical performance, together with higher PCS scores. Cluster separation was modest, and the gap statistic favored a single cluster, indicating descriptive, hypothesis‐generating patterns. Conclusion Older women with CLBP receiving PNE‐PA showed heterogeneous changes in physical function and pain‐related cognitions. Lower baseline mobility and higher catastrophizing were associated with larger domain‐specific improvements, although these associations may partly reflect regression to the mean. Identifying baseline physical and psychological profiles may inform more targeted group‐based programs for older adults with CLBP.