Associations between the novel endothelial activation and stress index and mortality in depressed adults: mediation analysis of systemic inflammation in a large NHANES sample.

Abstract

Background: Endothelial dysfunction contributes to poor outcomes in depression. However, the prognostic significance of the endothelial activation and stress index (EASIX), a novel marker for endothelial dysfunction, for mortality risk in depression remains unknown. This study aimed to: (1) evaluate the prognostic value of EASIX for mortality risk in US adults with clinically significant depressive symptoms, and (2) explore the potential mediating role of systemic inflammation using multiple hematologic indices.

Methods: We included 1717 adults with clinically significant depressive symptoms (PHQ-9 ≥ 10) from NHANES 2005-2018 after applying predefined exclusion criteria, with mortality follow-up through December 2019. Nonlinear associations were assessed using RCS, and survival differences were evaluated with Kaplan-Meier curves. Cox proportional hazards models estimated hazard ratios (HRs). We additionally performed supplementary weighted analyses. Exploratory mediation analyses examined the neutrophil-to-lymphocyte ratio (NLR), eosinophil-to-lymphocyte ratio (ELR), monocyte-to-lymphocyte ratio (MLR), and the systemic inflammation response index (SIRI). Additionally, we conducted sensitivity analyses excluding early deaths (≤ 2 years) and participants with CKD, abnormal platelet counts, or extreme LDH levels.

Results: Over a median follow-up of 83 months (IQR, 46-125 months), 251 all-cause and 56 cardiovascular mortalities occurred. Participants in the highest EASIX quartile showed higher mortality rates. RCS analyses indicated a mostly linear association for all-cause mortality and a nonlinear relationship for cardiovascular mortality. In fully adjusted Cox analyses, participants in the highest EASIX quartile had a 1.82-fold higher risk of all-cause mortality (95% CI: 1.13-2.95). For cardiovascular mortality, the point estimate indicated an increased risk (HR = 1.26); however, it was not statistically significant (95% CI: 0.47-3.41). Sensitivity analyses confirmed robustness, with significant associations persisting after excluding early deaths (HR 2.25, 95% CI 1.34-3.78) and after excluding participants with conditions affecting EASIX components (HR 1.93, 95% CI 1.12-3.34). Exploratory mediation analyses showed that systemic inflammation explained a significant portion of the EASIX-mortality link, with mediated proportions ranging from 7.3% (SIRI) to 29.2% (MLR) for all-cause mortality and from 6.3% (SIRI) to 26.8% (MLR) for cardiovascular mortality.

Conclusion: Elevated EASIX is strongly associated with increased all-cause mortality in individuals with clinically significant depressive symptoms, with systemic inflammation mediating a substantial portion of this risk.