Stress hyperglycemia ratio and short-term mortality in critically ill patients with respiratory diseases: a systematic review and dose-response meta-analysis

ObjectiveTo systematically evaluate the association between the HbA1c-adjusted stress hyperglycemia ratio (SHR) and short-term mortality in patients with severe respiratory diseases and characterize its linear and nonlinear dose-response relationships.MethodsPubMed, Embase, China National Knowledge Infrastructure, Wanfang, VIP, and the Chinese Biomedical Literature Database were searched for observational studies evaluating SHR and short-term mortality, including in-hospital, intensive care unit (ICU), 28-day, or 30-day mortality. Odds ratios (ORs) and hazard ratios (HRs) were pooled separately using random-effects models. Sensitivity analyses, linear dose-response analyses, restricted cubic spline analyses, and exploratory analyses of low and high SHR were performed.ResultsEleven studies were included. Four studies contributed to the OR-based primary meta-analysis. Short-term mortality tended to be higher in the highest SHR category than in the study-specific reference category, but the association was not statistically significant (OR = 2.72, 95% CI: 0.74–9.93, P = 0.091; I2 = 86.4%). Three studies contributed to the HR-based primary meta-analysis, which showed a significant association between higher SHR and increased short-term mortality risk (HR = 1.87, 95% CI: 1.71–2.05, P = 0.001; I2 = 0%). In HR studies, each 0.1-unit increase in SHR was associated with an approximately 5% increase in short-term mortality risk (HR = 1.05, 95% CI: 1.02–1.08, P = 0.023). OR studies showed a similar direction of association but did not reach statistical significance. Restricted cubic spline analyses found no clear evidence of nonlinearity. Exploratory analyses suggested higher risks at both low and high SHR levels, but the estimates were imprecise.ConclusionsCurrent evidence suggests that higher SHR is associated with increased short-term mortality risk in patients with severe respiratory diseases. HR-based evidence was more consistent and supported a positive linear dose-response relationship. OR-based findings pointed in the same direction but showed substantial heterogeneity and did not reach statistical significance. The available evidence is insufficient to confirm a clear nonlinear pattern. Given the limited number of studies and wide confidence intervals, the apparent increase in risk at both low and high SHR levels should be interpreted as exploratory and does not yet support a stable U-shaped or J-shaped dose-response pattern.Systematic Review Registrationhttps://www.crd.york.ac.uk/PROSPERO/view/CRD420261418259, PROSPERO CRD420261418259.