This secondary cohort analysis assessed whether preexisting diabetes mellitus or atherosclerotic cardiovascular disease (ASCVD) modify the relationship between baseline soluble tumor necrosis factor receptor 1 (sTNFR1) and 90-day mortality in adult patients with sepsis. The primary finding was that the positive association between higher sTNFR1 and mortality observed in patients without diabetes or ASCVD was markedly attenuated among those with these comorbidities, suggesting comorbidity-dependent differences in biomarker–outcome relationships.
The analysis used data from the Acetaminophen and Ascorbate in Sepsis: Targeted Therapy to Enhance Recovery research materials. The analytic cohort comprised 466 adults hospitalized with sepsis and either respiratory failure or vasopressor-dependent shock who had baseline sTNFR1 measurements available.
Of the 466 patients, 243 (52.1%) had a diagnosis of diabetes or ASCVD as captured by Charlson comorbidity categories. Although illness severity and overall mortality were similar between patients with and without these comorbidities, median sTNFR1 concentrations differed: patients with diabetes or ASCVD had higher median sTNFR1 (6547 pg/mL) compared with those without (4883 pg/mL).
Investigators tested for interaction between sTNFR1 and presence of diabetes or ASCVD on 90-day mortality. They used Cox proportional hazards regression with restricted cubic splines for sTNFR1 to allow for non-linear associations. Models were adjusted for illness severity, age (entered in tertiles), and sex. Interaction testing produced p-values to assess whether the sTNFR1–mortality association differed by comorbidity status.
Among patients without diabetes or ASCVD, increasing sTNFR1 was associated with higher 90-day mortality, with a reported hazard ratio (HR) of 2.25 per standard deviation increase in sTNFR1 (95% CI, 1.76–2.88). In contrast, among patients with diabetes or ASCVD the association was substantially attenuated. The relationship in this group showed a plateau in mortality risk at higher sTNFR1 values—approximately above 12,000 pg/mL—with an overall HR of 1.04 (95% CI, 0.83–1.30). The interaction between sTNFR1 and diabetes/ASCVD on mortality was statistically significant (interaction p < 0.001), indicating meaningful effect modification by these comorbid conditions.
The investigators evaluated whether the observed interaction was specific to sTNFR1 or reflected a broader pattern across inflammatory or endothelial biomarkers. They report that the interaction with comorbidity status was specific to sTNFR1; similar effect modification was not observed for interleukin-6 (IL-6) or angiopoietin-2.
The study also examined whether obesity or hypertension without coexisting diabetes or ASCVD modified the association between sTNFR1 and mortality. No effect modification was present for patients with obesity or hypertension in the absence of diabetes or ASCVD, suggesting the attenuation phenomenon was specific to the diabetes/ASCVD comorbidity cluster.
These results suggest that preexisting cardiometabolic disease states—specifically diabetes and ASCVD—may change the prognostic meaning of elevated sTNFR1 in sepsis. Higher baseline sTNFR1 among patients with these comorbidities did not translate into the same incremental mortality risk seen in patients without these conditions, and the concentration–risk curve plateaued at very high sTNFR1 values. The authors highlight that comorbid conditions can alter the biological relationship between a biomarker and outcome, which has implications for using biomarkers like sTNFR1 in risk stratification and for understanding sepsis pathophysiology across heterogeneous patient subgroups.
This report is a secondary cohort analysis of an existing research resource. The abstract provides details about sample size, comorbidity definitions (Charlson categories), covariates included in adjusted models, and the statistical approach (Cox regression with restricted cubic splines). Additional limitations, detailed methods, and full reporting elements beyond what is contained in the abstract were not reported in the source provided here.
In this cohort of 466 adults with sepsis and respiratory failure or vasopressor-dependent shock, the association between baseline sTNFR1 and 90-day mortality was strong among patients without diabetes or ASCVD but was attenuated among those with these comorbidities. The interaction was specific to sTNFR1 and not observed for IL-6 or angiopoietin-2. These findings indicate that preexisting diabetes or ASCVD can modify biomarker–outcome relationships in sepsis, with potential consequences for biomarker-driven risk prediction and mechanistic research.