Stroke Admission Hyperglycemia Linked to Worse Discharge Function

Key Takeaways

• Among adults with acute ischemic stroke treated at a comprehensive stroke center in Al Ain, United Arab Emirates, admission hyperglycemia was common and was associated with worse discharge function.
• Algorithm-defined inappropriate inpatient glycemic management was far more frequent in hyperglycemic than euglycemic admissions.
• Routine HbA1c testing identified previously unrecognized diabetes in patients without a prior diabetes diagnosis.
• Stress hyperglycemia ratio showed an unadjusted gradient in unfavorable discharge outcome, but that relationship did not remain independently associated after multivariable adjustment.
• No severe hypoglycemia was observed during hospitalization.

Elevated glucose at stroke presentation can reflect acute physiologic stress, previously unrecognized chronic dysglycemia, or both, leaving stroke teams to interpret its prognostic meaning while managing a neurologic emergency. In Al Ain, United Arab Emirates, that question also intersects with routine inpatient stroke-unit glucose management rather than a region-specific therapy, which gives the findings some relevance to North American practice despite the single-center setting. At Tawam Hospital, investigators examined how admission hyperglycemia related to discharge disability, in-hospital death, and early inpatient glycemic management during acute ischemic stroke admission.

In the acute ischemic stroke hyperglycemia cohort, investigators conducted a single-center retrospective study at Tawam Hospital, a comprehensive stroke center in Al Ain, United Arab Emirates, using the stroke registry to capture 218 consecutive adults admitted from July 1 through December 31, 2023, with 205 patients contributing to the primary admission-glucose analysis. Admission hyperglycemia was defined as capillary blood glucose greater than 7.8 mmol/L measured within 60 minutes of emergency department arrival. Co-primary outcomes were discharge modified Rankin Scale (mRS) scores of 3 to 6 and in-hospital death, while glycemic management was assessed every 6 hours across 13 time points from presentation to 72 hours. Supplementary analyses incorporated glycated hemoglobin (HbA1c) testing and stress hyperglycemia ratio (SHR), and the primary discharge model used a prespecified nine-covariate adjustment set that included age, sex, National Institutes of Health Stroke Scale (NIHSS) group, diabetes, prior stroke or transient ischemic attack, atrial fibrillation, heart failure, chronic kidney disease, and reperfusion therapy.

Admission hyperglycemia was independently associated with unfavorable discharge mRS, with adjusted OR 3.43 (95% CI 1.33–8.87; p=0.011). The crude outcome split also favored worse functional status in the hyperglycemic group. That association remained present across stroke-severity strata, with a Mantel–Haenszel pooled OR 3.47 (95% CI 1.48–8.10; p<0.001). Admission hyperglycemia was not independently associated with in-hospital death after parsimonious adjustment, and the mortality analysis was constrained by sparse events.

Inappropriate glycemic management occurred in 85.8% of hyperglycemic patients versus 11.1% of euglycemic patients, with adjusted OR 80.59 (95% CI 20.21–321.38; p<0.001). Routine HbA1c testing identified previously unrecognized diabetes in 18.4% of patients without known diabetes. The most common protocol failure was glucose at or above 10 mmol/L without same-timepoint sliding-scale insulin documentation, no severe hypoglycemia below 2.2 mmol/L was observed, glucose in the hyperglycemic group remained near or above target through 72 hours, and SHR showed a monotonic unadjusted gradient that did not remain independently associated after adjustment.

Because the cohort was observational and drawn from a single center, the findings support association rather than causation and may not generalize broadly outside this setting. Thirteen patients lacked admission glucose data, so investigators used complete-case analysis, and the mortality model was underpowered because few in-hospital deaths occurred. The very large inappropriate-management odds ratio reflected near-complete separation and is better read as a marker of a strong association than as a precise effect size. Subgroup patterns by reperfusion therapy or time to presentation and the longer-term modified Rankin Scale signal were exploratory.

The investigators reported that admission hyperglycemia tracked with worse discharge functional status after adjustment and also marked patients most likely to experience SHINE-aligned inpatient glycemic management gaps in this UAE stroke-center cohort. In this dataset, SHR did not clearly outperform absolute admission glucose after adjustment, while routine HbA1c testing frequently uncovered chronic dysglycemia and severe hypoglycemia was not observed. The overall pattern linked early bedside glucose status with both functional outcome and inpatient management patterns.

Clinician Questions

How was inappropriate glycemic management defined for acute ischemic stroke admissions in this cohort?

In adults with acute ischemic stroke at Tawam Hospital, glycemic management was assessed every 6 hours across 13 time points from presentation to 72 hours, and a patient was classified as receiving inappropriate management if any glucose value was 10 mmol/L or higher without same-timepoint sliding-scale insulin or if glucose remained 10 mmol/L or higher without dose escalation after insulin. For hypoglycemia-related inappropriate management, the methods defined the threshold as glucose below 2.2 mmol/L without documented corrective treatment, although the results later describe 13 protocol-failure episodes at glucose below 4 mmol/L without a documented response.

What did HbA1c testing show in stroke patients without a prior diabetes diagnosis?

Among acute ischemic stroke patients without known diabetes who underwent HbA1c testing, investigators found previously unrecognized diabetes and an even larger prediabetes group. Among hyperglycemic patients without known diabetes who had HbA1c measured, most testable cases were not consistent with normal chronic glycemia.

Did stress hyperglycemia ratio add prognostic information beyond absolute admission glucose in acute ischemic stroke?

In this acute ischemic stroke cohort, stress hyperglycemia ratio showed an unadjusted gradient in unfavorable discharge outcome across tertiles, but its association did not remain statistically significant after multivariable adjustment.