The longitudinal relationship between changes in wellbeing and inflammatory markers: Are associations independent of depression?

Objectives: There is a large literature linking inflammation with mental ill health, but a much smaller literature focusing on mental wellbeing. Specifically, it remains unclear whether mental wellbeing is longitudinally and independently associated with inflammation or only via associated changes in mental ill health. Methods: This study used data from 8780 adults aged 50 + in the English Longitudinal Study of Ageing. Hedonic wellbeing (both positive affect and life satisfaction) and eudemonic wellbeing (self-realisation and control-autonomy) were measured at data collection waves 2 (2004/05), 4 (2008/09) and 6 (20012/13), along with measures of C-reactive protein (CRP), fibrinogen and white blood cells (WBC). Fixed effects modelling was performed to identify the longitudinal relationship between wellbeing and inflammation, adjusting for timevarying mental ill health and other identified confounders. Results: Both measured aspects of hedonic wellbeing were associated with lower WBC count, independent of mental ill health. For life satisfaction, this relationship was explained by confounders, whilst for positive affect it persisted. Both measured aspects of eudemonic wellbeing were associated with lower CRP, fibrinogen and WBC, independent of mental ill health. For control-autonomy, this relationship was explained by confounders, whilst for self-realisation it persisted. Results were present in both men and women, although more strongly in men, and were robust to a range of sensitivity analyses. Conclusions: This study builds on the strong literature showing a relationship between mental ill health and inflammation by showing that there is also an apparently independent relationship between mental wellbeing, in particular eudemonic wellbeing, and inflammation that is unexplained by socio-economic or other time-constant factors and in some instances persists independent of time-varying confounders.