Association of hemoglobin glycation index with poor prognosis in critically ill patients with ischemic stroke: a cohort study.

Abstract

Background: The hemoglobin glycation index (HGI), quantifying interindividual hemoglobin glycation variability, predicts cardiovascular outcomes but lacks exploration in cerebrovascular prognosis. We evaluated its nonlinear association with mortality in critically ill ischemic stroke (IS) patients.

Methods: This retrospective cohort analyzed 2,035 ICU-admitted IS patients (MIMIC-IV database). HGI was computed as measured HbA1c minus regression-predicted HbA1c (fasting glucose-derived). Restricted cubic splines and multivariable Cox models assessed nonlinear mortality risks (30-/365-day).

Results: A U - shaped HGI - mortality relationship emerged. Compared to moderate HGI (Q3), low HGI (Q1) doubled the 30 - day mortality risk (adjusted HR = 2.00, 95% CI: 1.49-2.70) and increased the 365 - day risk by 59% (HR = 1.59, 95% CI: 1.27-1.99). High HGI (Q4) demonstrated a 38% increase in the 30 - day risk (HR = 1.38, 95% CI: 0.99-1.93) and a 24% increase in the 365 - day risk (HR = 1.24, 95% CI: 0.98-1.59). Continuous HGI was inversely correlated with mortality (30 - day HR = 0.86, 95% CI: 0.78-0.95, p = 0.003; 365 - day HR = 0.92, 95% CI: 0.86-0.99, p = 0.038). Threshold analysis identified inflection points (HGI = 0.58 for the 365 - day period; 1.299 for the 30 - day period), with mortality increasing beyond these thresholds.

Conclusions: HGI demonstrates a robust U-shaped association with mortality in critical IS, independent of glycemic status. Both extremes independently predict adverse outcomes, with low HGI conferring higher risk. HGI emerges as a novel prognostic biomarker, underscoring glycemic variability's clinical relevance beyond conventional metrics for IS risk stratification.