Depression as a Mediator and Social Participation as a Moderator in the Bidirectional Relationship Between Sleep Disorders and Pain: Dynamic Cohort Study (Preprint)

Abstract

Chronic pain, sleep disorders, and depression are major global health concerns. Recent studies have revealed a strong link between sleep disorders and pain, and each of them is bidirectionally correlated with depressive symptoms, suggesting a complex relationship between these conditions. Social participation has been identified as a potential moderator in this complex relationship, with implications for treatment. However, the complex interplay among sleep disorders, pain, depressive symptoms, and social participation in middle- and old-aged Asians remains unclear.

This study aimed to examine the bidirectional relationship between sleep disorders and pain in middle- and old-aged Chinese and measure the role of depression as a mediator and social participation as a moderator in this bidirectional relationship through a dynamic cohort study.

We used data from the China Health and Retirement Longitudinal Study across 5 years and included a total of 7998 middle- and old-aged people (≥45 years old) with complete data in 2011 (T1), 2015 (T2), and 2018 (T3). The cross-lag model was used to assess the interplay among sleep disorders, pain, depressive symptoms, and social participation. Depressive symptoms were assessed by the 10-item Centre for Epidemiological Studies Depression scale. Sleep disorders were assessed by a single-item sleep quality scale and nighttime sleep duration. The pain score was the sum of all pain locations reported. Social participation was measured using self-reported activity.

Our results showed significant cross-lagged effects of previous sleep disorders on subsequent pain at T2 (β=.141; <i>P</i>&lt;.001) and T3 (β=.117; <i>P</i>&lt;.001) and previous pain on subsequent poor sleep at T2 (β=.080; <i>P</i>&lt;.001) and T3 (β=.093; <i>P</i>&lt;.001). The indirect effects of previous sleep disorders on pain through depressive symptoms (β=.020; SE 0.004; <i>P</i>&lt;.001; effect size 21.98%), as well as previous pain on sleep disorders through depressive symptoms (β=.012; SE 0.002; <i>P</i>&lt;.001; effect size 20.69%), were significant across the 3 time intervals. Among participants with high levels of social participation, there were no statistically significant effects of previous sleep disorders on subsequent pain at T2 (β=.048; <i>P</i>=.15) and T3 (β=.085; <i>P</i>=.02), nor were there statistically significant effects of previous pain on subsequent sleep disorders at T2 (β=.037; <i>P</i>=.15) and T3 (β=.039; <i>P</i>=.24). Additionally, the mediating effects of depressive symptoms on the sleep disorders-to-pain pathway (<i>P</i>=.14) and the pain-to-sleep disorders pathway (<i>P</i>=.02) were no longer statistically significant.

There is a bidirectional relationship between sleep disorders and pain in middle- and old-aged Asians; depression plays a longitudinal mediating role in the bidirectional relationship between them; and social participation moderates the bidirectional relationship between them directly and indirectly by affecting depression. Future interventions may consider the complex relationship between these conditions and adopt a comprehensive treatment regime.