# U shaped relationship between serum glucose potassium ratio and mortality in critically ill patients with toxic encephalopathy

## TL;DR

This study found that both very low and very high serum glucose-potassium ratios are linked to higher mortality in critically ill patients with toxic encephalopathy.

## Contribution

The study reveals a U-shaped relationship between glucose-potassium ratio and mortality in toxic encephalopathy patients.

## Key findings

- Both low and high GPR levels are associated with increased 28-day and 90-day mortality in TE patients.
- The U-shaped relationship was confirmed with inflection points showing significant hazard ratios on both ends.
- Sensitivity analysis supported the stability of the observed U-shaped association.

## Abstract

The link between serum glucose-potassium ratio (GPR) and mortality in critically ill toxic encephalopathy (TE) patients is not well defined. This study has aimed to clarify the effect of GPR fluctuations on TE mortality. A total of 3,462 TE patients with TE requiring ICU care were selected from the Medical Information Mart Intensive Care (MIMIC-IV) database. Patients were categorized into three groups based on GPR tertiles: Tertile 1 (n = 1154, range 0.103–1.442), Tertile 2 (n = 1153, range 1.444–1.966), and Tertile 3 (n = 1155, range 1.967–12.937). The primary outcomes studied were 28-day and 90-day all-cause mortality (ACM). To analyze the relationship between GPR and outcomes, we employed Cox regression models adjusted for multiple covariates and restricted cubic splines to explore the potential non-linear association. The 3,462 - patient cohort had a mean age of 67.7 ± 16.6 years, with 58.2% male. The 28-day and 90-day ACM were 21.9% and 31.2%, respectively. Multivariate adjusted analysis showed no overall GPR–ACM correlation at 28 and 90 d. Regarding different groups, with T2 as the reference group (Ref), for 28-day ACM, the adjusted hazard ratio (HR) of the T1 was 1.20 (95% Confidence Interval [CI]: 1.00–1.44, p ≡ P = 0.049), and that of T3 group was 1.22 (95% CI: 1.01–1.47, P = 0.035). For 90 - day ACM, the adjusted HR of the T1 was 1.19 (95% CI: 1.02–1.39, P = 0.023), and the T3 was 1.20 (95% CI: 1.03–1.40). The correlation between the GPR lesvel and ACM was U-shaped association. The left and right - hand side effect sizes at the inflection point (1.65) were 0.472 (HR: 0.472, 95% CI 0.306–0.728, P < 0.001) and 1.127 (HR: 1.127, 95% CI 1.032–1.229, P = 0.0075). Sensitivity analysis was stable. Our findings have revealed a U-shaped relationship between GPR levels and ACM in critically ill patients with TE. Close attention should therefore be paid to this issue in order to improve patient care.

The online version contains supplementary material available at 10.1038/s41598-025-12496-4.