Exploratory associations of composite inflammatory-immune markers with 28-day and 1-year all-cause mortality in patients with sepsis: a retrospective cohort study

Aims Inflammation and immune dysregulation are central to sepsis pathobiology, but the prognostic value of a broad panel of composite inflammatory-immune indices remains incompletely defined. We evaluated the associations of 18 composite inflammatory-immune markers with 28-day and 1-year all-cause mortality in sepsis patients. Methods This single-center retrospective cohort study included 396 adult patients admitted to the intensive care unit (ICU) with sepsis. Eighteen inflammatory-immune indices were calculated from blood tests obtained within the first 24 h after ICU admission. Per-1-standard deviation (SD) multivariable Cox models were used to evaluate each marker’s association with 28-day and 1-year all-cause mortality. Subgroup analyses were performed by age (<75 vs. ≥75 years), sex, and sepsis-associated acute kidney injury status, with multiplicative interaction terms. Restricted cubic spline (RCS) analyses were used to assess nonlinearity. Results Among 396 patients, 174 (43.9%) died within 28 days and 230 (58.1%) died within 1 year. In the primary per-1-SD multivariable Cox analyses, higher C-reactive protein-to-lymphocyte ratio (CLR) and platelet-to-lymphocyte ratio (PLR) were associated with higher 28-day mortality (CLR: HR 1.313, 95% CI 1.145–1.505; P<0.001; PLR: HR 1.253, 95% CI 1.114–1.410; P<0.001) and 1-year mortality (CLR: HR 1.258, 95% CI 1.104–1.435; P<0.001; PLR: HR 1.193, 95% CI 1.071–1.329; P = 0.001). Higher lymphocyte-to-HDL-C ratio (LHR) was inversely associated with both endpoints (28-day: HR 0.759, 95% CI 0.626–0.920; P = 0.005; 1-year: HR 0.819, 95% CI 0.703–0.954; P = 0.010). Other markers showed endpoint-dependent or otherwise variable associations. Most formal interaction tests did not indicate heterogeneity by age, sex, or sepsis-associated acute kidney injury (SA-AKI) status; nominal interactions were observed for pan-immune-inflammation value with age for both endpoints and with SA-AKI status for 1-year mortality. RCS analyses showed nonlinear associations of log 10 -transformed PLR and CLR with both endpoints and an approximately linear inverse association of log 10 -transformed LHR with 28-day mortality. Conclusions Higher CLR and PLR and lower LHR were independently associated with both short- and long-term mortality in patients with sepsis. However, the findings were derived from a single-center retrospective cohort without independent validation and do not establish clinical prediction utility. Further multicenter studies are required to validate these markers before they can be considered for clinical risk stratification.