Admission triglyceride–glucose index and 3-month post-stroke cognitive impairment in young adults with first-ever ischemic stroke: a retrospective cohort study

Post-stroke cognitive impairment (PSCI) is common after young-onset ischemic stroke, but simple admission biomarkers remain limited. The triglyceride-glucose (TyG) index is a routinely available surrogate marker of insulin resistance. We conducted a single-center retrospective cohort study using a prospectively maintained stroke registry. Consecutive adults aged 18 to 45 years admitted within 72 h of first-ever acute ischemic stroke between January 2022 and December 2025 were screened. Fasting triglycerides and plasma glucose were measured on the morning after admission, and TyG was calculated as ln[(triglyceride mg/dL x fasting plasma glucose mg/dL)/2]. Cognitive function was assessed at 3 months with the Beijing version of the Montreal Cognitive Assessment (MoCA). PSCI was defined as MoCA <26. Multivariable logistic regression, restricted cubic splines, bootstrap-corrected receiver operating characteristic analysis, and DeLong testing were used. Among 203 eligible patients, 139 (68.5%) had PSCI at 3 months. TyG increased stepwise across cognitive severity strata (8.42 ± 0.34 to 9.98 ± 0.44; P < 0.001) and correlated inversely with MoCA score (Spearman rho = − 0.729; P < 0.001). After adjustment for demographic, vascular, metabolic, biochemical, and stroke-severity covariates, each 1-standard deviation increase in TyG was associated with PSCI (adjusted odds ratio 2.38; 95% confidence interval 1.61 to 3.52; P < 0.001). The bootstrap-corrected area under the curve was 0.852 for TyG alone and 0.898 for TyG plus admission NIHSS (DeLong P = 0.018). Admission TyG was independently associated with 3-month PSCI in young adults with first-ever ischemic stroke. These findings support TyG as a pragmatic candidate marker requiring prospective external validation.