LAUSR

AIMS We aimed to investigate the effect of glycemic variability in the acute stage of stroke on the development of post-stroke cognitive impairment (PSCI). METHODS Patients who underwent blood glucose tests at least five times within 7 days after acute ischemic stroke were included. Factors related to glycemic variability (standard deviation (SD), coefficient of variance (CV), and mean absolute glucose (MAG)) were calculated; neuropsychological assessments were administered 3 months after stroke. PSCI was defined as a score of less than -2 SDs for age-, sex-, and education-adjusted means in at least one cognitive domain. RESULTS A total of 354 patients were enrolled. PSCI was identified in 74 (20.9%) subjects. In the diabetic group (n = 87), MAG was a significant predictor for PSCI (adjusted OR, 1.94; 95% CI, 1.11-3.42); however, it was not significant in the non-diabetic group, although PSCI exhibited an increasing tendency within higher SD and MAG tertiles. Moreover, hyperglycemia demonstrated a detrimental effect on PSCI, regardless of diabetes status; this effect did not appear in poorly-controlled diabetic patients with HbA1c ≥ 8.0%. CONCLUSIONS Glycemic variability and hyperglycemia during acute ischemic stroke were identified as novel predictors for PSCI. Although this result is not evidence of a causal relationship, our study suggests that monitoring glycemic index and controlling its variability during the acute phase of ischemic stroke may help to prevent poor cognitive outcomes.