Association of the glucose-to-lymphocyte ratio with short-term mortality in critically ill patients with documented mitral valve disease: A retrospective cohort study
Zhenyu Yuan, Lei Zhao, Hong Geng, Qiang Ji, Kun Zuo, Yuanzhi Luo, Guanli Liang, Gengxu HeObjective
The glucose-to-lymphocyte ratio (GLR) integrates metabolic stress and lymphocyte depletion, but its prognostic relevance in critically ill patients with mitral valve disease (MVD) is uncertain. We examined the association between early GLR and short-term all-cause mortality.
Methods
We used MIMIC-IV version 3.0 to identify adults with a documented MVD diagnosis and calculable GLR from the first valid glucose and absolute lymphocyte measurements within 24 hours of ICU admission. Primary analyses used a 24-hour landmark and modeled log2-transformed GLR continuously. Cox regression, restricted cubic splines, multiple imputation, timing and phenotype sensitivity analyses, paired DeLong tests, bootstrap-corrected C-index, calibration, and decision-curve analysis were performed.
Results
Among 1,538 patients with calculable GLR, seven died by the 24-hour landmark, leaving 1,531 patients; 148 (9.7%) died by day 28 and 203 (13.3%) by day 90. In the primary complete-case model (n=1,245), each doubling of GLR was associated with higher 28-day mortality (HR 1.43, 95% CI 1.22-1.66) and 90-day mortality (HR 1.30, 95% CI 1.14-1.49; both P<0.001). Spline analyses showed no evidence of nonlinearity, and multiple-imputation and major sensitivity analyses were directionally consistent. GLR discriminated mortality better than glucose alone but not significantly better than lymphocyte count. Adding GLR to age, sex, SOFA, and CCI increased optimism-corrected C-index by 0.030 at 28 days and 0.023 at 90 days, with good calibration and modest net benefit.
Conclusions
Higher early GLR was independently associated with 28-day and 90-day mortality in critically ill patients with documented MVD and provided modest adjunctive prognostic information.