Hemoglobin glycation index and mortality in chronic heart failure: A retrospective cohort study.

This study investigates the association between the hemoglobin glycation index (HGI) and 90-day and 365-day all-cause mortality (ACM) in hospitalized patients with chronic heart failure (CHF). A total of included 2964 hospitalized CHF patients from the Medical Information Mart for Intensive Care IV database were included. HGI was calculated as observed glycated hemoglobin (HbA1c) minus the cohort-specific predicted HbA1c estimated from a linear regression of HbA1c on fasting plasma glucose. We ascertained 90-day and 365-day ACM from database-recorded dates of death, counting events during the index hospitalization and after discharge within each window. Cox proportional hazards regression models were used to examine the association of HGI with 90-day and 365-day ACM. Restricted cubic spline (RCS) curves assessed for nonlinear relationships, while Kaplan-Meier (KM) survival curves compared survival differences among HGI groups. Subgroup analysis, the Boruta algorithm, and mediation analysis were employed to explore the underlying mechanisms. Logistic regression models were used for sensitivity analysis. Multivariate Cox regression analysis revealed that a higher HGI was significantly associated with decreased ACM at both 90 and 365 days in CHF patients. The KM survival curve demonstrated that patients in the lowest HGI quartile (Q1 group) had significantly lower survival rates. Sensitivity analysis further confirmed the effect of HGI, which was consistent across various subgroups. The Boruta algorithm identified HGI as an independent predictor of mortality. Mediation analysis indicated that the leukocyte count partially mediated the association between HGI and mortality. HGI is significantly and negatively associated with 90-day and 365-day IHM in patients with CHF. HGI was independently associated with better survival, an effect that is partially mediated by leukocytes.