Accurate identification and early warning of cognitive impairment after stroke

Reliable markers for predicting the progression of post-stroke cognitive impairment (PSCI) are currently lacking.

To identify predictive markers of PSCI in multiple dimensions and build an early warning model.

This was a prospective, observational cohort study. We enrolled 214 patients with acute ischaemic stroke (AIS) and followed up for 6 months, with cognitive function assessed at the end of the follow-up period. A total of 203 AIS patients completed follow-up. Data were analysed according to demographic, clinical, and imaging parameters. A nomogram model was established using binary logistic regression and evaluated using AUROC, internal Bootstrap calibration, and clinical decision curve analysis. Multiple linear regression was used to quantify the predictive ability of PSCI predictors.

Logistic regression indicated that age (1.10 [1.05-1.15, P < 0.001]), white matter hyperintensities (WMH) (2.68 [1.30-5.54, P = 0.008]) and NIHSS score (1.73 [1.32-2.25, P < 0.001]) were independent risk factors for PSCI in patients with AIS at 6 months. This nomogram model was robust and had a good potential utility. Multiple linear regression analysis showed that age, NIHSS score, and WMH severity were independently and inversely associated with MoCA score. After adjustment, each 1-year increase in age was associated with a 0.15-point lower MoCA score; each 1-point higher NIHSS score with a 0.40-point lower MoCA score; and each higher WMH severity grade with a 0.87-point lower MoCA score.

The nomogram model effectively identified stroke survivors at high risk of PSCI and may be a potential tool to guide clinicians in early stratification and decision-making. It must be emphasized that the proposed nomogram represents only an internally validated derivation model, which requires further validation.