LAUSR

Aims To investigate the bidirectional longitudinal association between pain and depressive symptoms and explore whether gender modifies the association. Methods This study used data of 17,577 participants without depressive symptoms and 15,775 without pain at baseline from waves 1–8 (2002/2003 to 2016/2017) of the English Longitudinal Study of Aging (ELSA) and waves 1 to 3 [2011–2015] of the China Health Retirement Longitudinal Study (CHARLS). Cox regression models were performed at the cohort level to evaluate the potential longitudinal associations, and then random-effect meta-analyses were conducted to pool the results. The potential modifying effect was detected by Z-test. Results During 103,512 person-years of follow-up in participants without depressive symptoms, baseline pain intensity was associated with incident depressive symptoms. Compared with individuals who reported no pain at baseline, the pooled adjusted hazard ratio (HR) of incident depressive symptoms for participants with mild to moderate pain and for those with severe pain was 1.37 (95% CI: 1.22–1.55, p < 0.001) and 1.52 (95% CI: 1.34–1.73, p < 0.001), respectively. During 81,958 person-years of follow-up in participants without pain, baseline depressive symptoms were associated with a significantly higher incidence of pain, and the pooled adjusted HR of incident pain was 1.71 (95% CI: 1.60–1.82, p < 0.001). These associations were not modified by gender. Conclusions A bidirectional longitudinal association between pain and depressive symptoms was demonstrated, not modified by gender. Family doctors should be aware of the bidirectional association and advice individuals with pain or depressive symptoms to be screened for both kinds of symptoms.