# Association of leuko-glycemic index with mortality in critically ill stroke patients: analysis from the MIMIC-IV database and an institutional cohort

## TL;DR

A new study finds that a blood marker called the leuko-glycemic index is strongly linked to higher mortality in critically ill stroke patients.

## Contribution

This is the first study to evaluate the leuko-glycemic index as a prognostic biomarker in critically ill stroke patients using two independent cohorts.

## Key findings

- Patients in the highest quartile of LGI had over threefold higher 28-day mortality risk compared to the lowest quartile.
- LGI was also associated with higher in-hospital and 365-day mortality in both study cohorts.
- The risk of mortality increased steeply at lower to moderate LGI levels and then more gradually at higher levels.

## Abstract

The leuko-glycemic index (LGI), combining white blood cell count and blood glucose, reflects systemic inflammation and stress hyperglycemia. Its prognostic value in critically ill stroke patients remains unclear. This study evaluated the association between LGI and mortality using the MIMIC-IV database and an independent institutional cohort.

Patients with acute stroke were identified from MIMIC-IV (version 3.1) and a neurosurgical intensive care unit cohort. LGI was calculated as white blood cell count (×109/L) × blood glucose (mg/dL)/1,000 using measurements within the first 24 h after ICU admission and categorized into quartiles. The primary outcome was 28-day mortality. Secondary outcomes included in-hospital mortality in both cohorts and 365-day mortality in MIMIC-IV. Multivariable Cox models were used to assess the association between LGI and mortality; in MIMIC-IV, spline-based analyses further explored the dose–response pattern, and subgroup analyses were performed.

A total of 5,267 patients from MIMIC-IV and 424 from the institutional cohort were included. In MIMIC-IV, after multivariable adjustment, patients in the highest LGI quartile had more than threefold higher risk of 28-day mortality compared with the lowest quartile (hazard ratio [HR] 3.259, 95% confidence interval [CI] 2.568–4.138; p for trend < 0.001) and higher 365-day mortality (HR 2.703, 95% CI 2.203–3.317; p for trend < 0.001). LGI was also positively associated with 28-day mortality in the institutional cohort (adjusted HR 1.351, 95% CI 1.090–1.675), and in-hospital mortality rates were likewise highest in the top LGI quartile. Exploratory spline analyses in MIMIC-IV showed that risk increased steeply from low to moderate LGI and then rose more gradually at higher levels. Subgroup analyses suggested broadly consistent positive associations across most clinical strata.

LGI was significantly associated with mortality in critically ill stroke patients across two independent cohorts and may serve as a simple, readily available biomarker to aid early risk stratification.