Hemoglobin glycation index predicts mortality in intracerebral hemorrhage: a nonlinear threshold effect

The Hemoglobin Glycation Index (HGI), integrating acute and chronic glycemic dysregulation, may reflect metabolic state in critical illness. We aimed to evaluate whether HGI predicts 30-day mortality in intracerebral hemorrhage (ICH) patients and identify potential threshold effects.

We analyzed 1351 ICH patients from the MIMIC-IV database. HGI was calculated as the difference between observed and predicted hemoglobin A1c (HbA1c), derived from admission glucose. Patients were stratified into HGI quartiles (Q1-Q4). The primary outcome was 30-day mortality, assessed using Cox regression models adjusted for demographics, comorbidities, and severity scores. Nonlinear relationships were evaluated with restricted cubic splines and threshold effect analysis.

The cohort exhibited a nonlinear, L-shaped association between HGI and mortality (P for nonlinearity = 0.0008). Compared to Q4 (highest HGI), Q1 (lowest HGI) had a 53% higher mortality risk (adjusted HR 0.47, 95% CI 0.32-0.68, P < 0.001). A threshold effect was identified at HGI = 0.78: below this value, each unit increase in HGI reduced mortality by 24% (HR 0.76, 95% CI 0.68-0.85, P < 0.001), while no significant association existed above it (HR 1.11, P = 0.319). Subgroup analyses confirmed robustness across age, diabetes status, and neurological severity, with a pronounced protective effect in males (P for interaction = 0.038).

Low HGI independently predicts increased 30-day mortality in ICH, with a critical threshold at 0.78. HGI may represent a potential biomarker for metabolic risk stratification in neurocritical care, offering insights beyond conventional glycemic metrics.