Perceived stress is linked to heightened biomarkers of inflammation via diurnal cortisol in a national sample of adults

Abstract

Exposure to and perceptions of stress have been associated with altered systemic inflammation, but the intermediate processes by which stress links to inflammation are not fully understood. Diurnal cortisol slopes were examined as a pathway by which self-reported psychosocial stress is associated with inflammation [i.e., C-reactive protein (CRP), interleukin-6 (IL-6), fibrinogen, E-Selectin, and Intercellular Adhesion Molecule-1 (ICAM-1)] in a large sample of adults (the Midlife in the US study; N = 914; 55.9% female; aged 34-84 years). Structural equation modeling indicated that perceived psychological stress was associated with flattened diurnal cortisol slopes and flatter diurnal cortisol slopes were, in turn, associated with heightened inflammation in these cross-sectional analyses (index of indirect pathway, ω = 0.003, 95% CI [0.001, 0.004], ωSTD = 0.027; with covariates, ω = 0.001, [0.0002, 0.002], ωSTD = 0.011). A similar indirect effect was evident for self-reported traumatic life events (ω = 0.007, [0.004, 0.012], ωSTD = 0.030); however, inclusion of covariates (i.e., age, gender, race, ethnicity, body mass index, and other factors associated with physical health) accounted for this finding (ω = 0.001, [-0.001, 0.004], ωSTD = 0.005). These results support an allostatic load model of psychosomatic health, in which cortisol (along with other stress-responsive signaling molecules) is a necessary component for understanding links between stress exposure, perceived stress, and immune functioning.

Keywords: Allostatic load; Diurnal cortisol; Inflammation; Stress.

Copyright © 2021 Elsevier Inc. All rights reserved.

Figures

Figure 1.

Representation of the primary models for hypotheses H1 and H2 with unstandardized and standardized coefficients (B, BSTD), indices of indirect pathways (ω, ωSTD), and 95% confidence intervals. Values listed above each model are results from the analyses without covariates; values below each model are from the analyses with covariates (regression paths on cortisol and inflammation and covariance paths on stress/traumatic life events not pictured for covariate variables). Secondary models were constructed similarly, but with mean inflammation replaced by separate paths for each of the inflammation biomarkers (see Table 2).