# Combined value of triglyceride-glucose index and non-high-density lipoprotein cholesterol in predicting early cognitive impairment after acute ischemic stroke

## TL;DR

High levels of triglyceride-glucose index and non-HDL cholesterol predict early cognitive decline after stroke, with combined use improving prediction accuracy.

## Contribution

This study identifies the combined predictive value of TyG index and non-HDL-C for early cognitive impairment after acute ischemic stroke.

## Key findings

- Elevated TyG index and non-HDL-C levels are independently linked to cognitive impairment after stroke.
- Combining TyG index and non-HDL-C improves diagnostic accuracy for predicting cognitive impairment.
- Both markers show strong predictive value regardless of diabetes status in stroke patients.

## Abstract

Acute ischemic stroke (AIS) represents a significant global cause of mortality and long-term disability. Cognitive impairment frequently occurs among AIS patients, adversely affecting their functional outcomes. Identifying modifiable risk factors linked to cognitive dysfunction after stroke is thus critical for effective prevention and targeted therapeutic interventions.

This study explored the relationship between serum triglyceride-glucose (TyG) index and non-high-density lipoprotein cholesterol (non-HDL-C) levels and early cognitive impairment in AIS patients.

The Neurology Department of Anhui Medical University’s Fourth Affiliated Hospital recruited a total of 235 individuals diagnosed with AIS between September 2023 and January 2025. Patients served as a cognitive impairment group (n = 135) and a control group (n = 100). Furthermore, participants were dichotomized according to diabetic status, and the predictive value of the TyG index and non-HDL cholesterol for cognitive impairment following acute ischemic stroke was evaluated in these subgroups. The Montreal Cognitive Assessment (MoCA) was used to evaluate cognitive ability at seven days post-stroke; a score below 26 indicated impairment. After identifying independent risk variables for cognitive impairment using logistic regression analysis, the diagnostic value of these factors was determined using receiver operating characteristic (ROC) curve analysis.

The cognitive impairment group exhibited significantly elevated TyG index and non-HDL-C serum levels (p < 0.001). Patients with cognitive impairment were older, had less educational attainment, higher NIHSS scores, and reduced MoCA scores (p < 0.05). Additionally, glycemic indicators (FPG, HbA1c, TyG index) and lipid markers (TC, non-HDL-C, LDL-C, TG) were markedly elevated, while HDL-C was reduced among cognitively impaired individuals (p < 0.05). Patients in the high-TyG group displayed substantially increased glycemic parameters, lipid profiles, and higher diabetes prevalence (p < 0.05). Univariate logistic regression revealed each unit rise in TyG index and non-HDL-C (all p < 0.05) significantly elevated the risk of cognitive impairment. Both parameters negatively correlated with MoCA scores (both p < 0.001). The rise in non-HDL-C levels correlated with the increase in the TyG index (p < 0.001), which may indicate that both factors act in a coordinated manner within shared metabolic pathways. The combined predictive model incorporating both TyG index and non-HDL-C exhibited superior diagnostic performance (p < 0.001). Regardless of diabetic status, both the TyG index and non-HDL-C demonstrated significant predictive value for post-AIS cognitive impairment. Their combination provided incremental predictive information beyond either marker alone (p < 0.001).

Elevated serum levels of TyG index and non-HDL-C independently predict early cognitive impairment in AIS patients, with their combination significantly improving predictive accuracy. These results suggest potential benefits from early metabolic interventions to enhance cognitive recovery post-stroke.