Pain–Depression Networks Based on Different Pain Resilience Levels Among Chinese Middle‐Aged and Elderly Individuals With Chronic Pain

Background Pain resilience profiles derived from item‐level responses and the corresponding pain–depression symptom networks within each profile remain unclear among Chinese middle‐aged and older adults with chronic pain. Methods A sample of 1045 Chinese middle‐aged and older adults with chronic pain completed self‐reported assessments of pain, depression, pain resilience and demographic characteristics. The Pain Resilience Scale (PRS) was used to assess pain resilience, including cognitive/affective positivity (CAP) and behavioural perseverance (BP). Latent profile analysis (LPA) was used to identify PRS subgroups based on item‐level responses. One‐way analysis of variance (ANOVA) method was used to compare overall pain, depression and pain resilience across the PRS subgroups. A network analysis approach was then conducted to identify the central symptoms in the pain–depression network for the whole sample and for each PRS subgroup. Results LPA identified four distinct profiles: low PRS‐CAP, low PRS‐BP; low PRS‐CAP, high PRS‐BP; high PRS‐CAP, low PRS‐BP; and high PRS‐CAP, high PRS‐BP. Results of ANOVA and network analysis supported the classification of four subgroups. In addition, the analysis of network models across different pain resilience levels revealed significant differences in network global strength ( p = 0.020) and edge weights ( M = 0.33, p = 0.040) between the lowest and highest PRS subgroups. Conclusions These findings suggest that pain resilience is heterogeneous among Chinese middle‐aged and older adults with chronic pain, and that pain–depression symptom networks differ across resilience profiles. Considering resilience profiles may help to inform more tailored assessment and intervention planning. Significance Statement This study identified four pain resilience profiles in Chinese middle‐aged and older adults with chronic pain and showed that pain–depression symptom networks differed across these profiles. By focusing on unique covariance, item‐level network analysis may reveal clinically or empirically informative associations that are less apparent at the total‐score level. These findings may help generate hypotheses about profile‐specific assessment and intervention approaches.