Sleep disturbance, neuro-immune markers, and depressive symptoms in older age: Conditional process analysis from the English Longitudinal Study of Aging (ELSA)

This study aimed to investigate the long-term association between subjective sleep disturbance and depressive symptoms in older adults, and ascertain whether this association is partially mediated by neuro-immune markers, while testing the moderation of sex using conditional process analyses, i.e., combinations of mediation and moderation analyses. We analysed data of 2124 participants aged 50 and above from the English Longitudinal Study of Ageing (ELSA) across three waves of data collection. Sleep disturbance was assessed in 2008/9, serum levels of high sensitivity C-reactive protein (hs-CRP), insulin like growth factor-1 (IGF-1), white blood cell (WBC) count in 2012/2013, and self-reported depressive symptoms in 2016/2017. After accounting for health-related and psychosocial confounders, results showed that sleep disturbance significantly predicted depressive symptoms at follow-up in the whole sample and hs-CRP in women, but not in men. Hs-CRP also predicted depressive symptoms only in women. Moreover, hs-CRP significantly mediated the association between sleep disturbance and depressive symptoms in women, but not in men. The association between sleep disturbance and IFG-1was marginal, and the latter did not mediate the association between sleep and depressive symptoms neither in women nor in men. Results on WBC count were all non-significant. In conclusion, sleep disturbance resulted significantly associated with depressive symptoms in a long-term follow-up. Moreover, ELSA data provide preliminary evidence that increased inflammation may mediate this association in women. Future longitudinal studies may advance the knowledge in the field considering the mediating role of proinflammatory cytokines such as interleukin-6