The inflammation-depression link: How social networks buffer or exacerbate risk

Major depressive disorder (MDD) is a prevalent mental disorder, and low social support and high strain could impact its long-term symptom severity. Increased inflammation, marked by C-reactive protein (CRP) and fibrinogen, has also been correlated with more MDD symptoms. However, the inflammation-MDD symptom association might vary by social support dimensions. The current study thus examined how social support dimensions moderated the inflammation-MDD severity correlation. Community adults ( N = 1,054) with and without MDD provided plasma samples to measure CRP and fibrinogen levels and completed self-reports of perceived support and strain from family, friends, and partners at Wave 1 (W1). MDD severity was assessed at W1 and Wave 2 (W2, nine-year follow-up). Multiple linear regressions and generalized additive modeling (GAM) assessed how W1 social support dimensions and inflammation levels interacted to predict W2 MDD severity, controlling for clinical and sociodemographic covariates. Increased W1 fibrinogen predicted higher W2 MDD severity in participants with lower (vs. higher) W1 social support and higher (vs. lower) social strain (|standardized β| = 0.18–2.31 vs. 0.01–0.03). Further, increased CRP predicted more MDD symptoms in participants with higher (vs. lower) social strain (|β| = 0.24–0.26 vs. 0.15–0.16). These significant interaction findings were identical in linear and GAM models that accommodate non-linear associations. Conclusions : Results suggested that increased proinflammatory activity indexed by CRP and fibrinogen levels could predict nine-year MDD severity under social strains, consistent with the social signal transduction theory. Improving social support and decreasing social strain might buffer inflammation-related depression. • Social strain was linked to greater nine-year inflammation-MDD severity connection. • Fibrinogen interacted with social support and strain in predicting nine-year MDD. • CRP levels interacted with social strain in predicting long-term MDD severity. • GAM approach replicated similar linear and non-linear variable associations. • The social signal transduction theory might inform MDD prevention approaches.