Eur J Med Res. 2026 Jan 31. doi: 10.1186/s40001-026-03916-w. Online ahead of print.
ABSTRACT
BACKGROUND: Insulin resistance and chronic inflammation are closely associated with cognitive impairments. This study systematically investigates the relationship between biomarkers of insulin resistance/chronic inflammation and cognitive dysfunction in patients with non-disabling ischemic cerebrovascular events (NICE).
METHODS: We collected demographic information and clinical data from 236 patients with NICE. Based on Montreal Cognitive Assessment (MoCA) scores, participants were categorized into normal cognitive function (NCF) and VCI groups. Propensity score matching (PSM) was applied to balance baseline characteristics. Differences in chronic inflammatory markers and insulin resistance levels were compared between groups. LASSO regression was used to identify independent risk factors, while restricted cubic spline (RCS) analysis was performed to validate dose-response relationships. A nomogram model was constructed using LASSO-selected predictors, and its performance was evaluated by ROC curves, calibration plots, and decision curve analysis (DCA). Internal validation was performed through simple cross-validation, with both accuracy and Kappa statistics reported.
RESULTS: Among 236 NICE patients, 115 (48.73%) were diagnosed with VCI. Following propensity score matching, the VCI group exhibited significantly higher levels of insulin resistance and chronic inflammation compared to the NCF group. LASSO regression identified the metabolic score for insulin resistance (METS-IR) as an independent risk factor for cognitive impairment (OR = 1.11, 95% CI: 1.06-1.17). RCS confirmed a linear negative correlation between METS-IR and MoCA scores (P for overall = 0.014, P for non-linear = 0.715). Mediation analysis revealed that the systemic Immune-Inflammation Index (SII) partially mediated the association between METS-IR and MoCA scores. The nomogram model demonstrated good discrimination (AUC = 0.78, 95% CI: 0.72-0.83), with calibration plots showing high consistency between predicted and observed probabilities (Hosmer-Lemeshow test P = 0.718). DCA confirmed a favorable clinical net benefit. Cross-validation results demonstrated favorable model accuracy and consistency (accuracy = 0.71, Kappa value = 0.43).
CONCLUSIONS: Cognitive impairment in NICE patients is strongly associated with elevated insulin resistance and chronic inflammation. METS-IR exhibits a linear negative association with cognitive function, serving as an independent risk predictor. The constructed nomogram provides a reliable tool for early VCI detection with robust discrimination and calibration. Notably, SII partially mediates the association between METS-IR and cognition, highlighting inflammatory pathways as a candidate target for future interventional studies.
PMID:41620791 | DOI:10.1186/s40001-026-03916-w
