Abstract WP376: Stress-Mediated Hyperglycemia or Undiagnosed Diabetes?

Abstract

Background: “Stress mediated” hyperglycemia is often assumed to drive initial hyperglycemia during stroke admission. We sought to examine hemoglobin A1c (HbA1c) in relation to initial serum glucose. Methods: An observational design was used to examine glucose data from prospectively admitted acute stroke patients at 4 Comprehensive Stroke Centers. Data were entered and analyzed in SPSS using descriptive statistics, student t-tests, and Mann Whitney U. Results: 1226 patients were studied; 50% male, 63% White, 34% African American (AA); age averaged 69±13.5 years, and 93% were ischemic strokes. DM was known on admission in 37%. Admission HbA1c was 5.8±.7 in non-DM patients, compared to 7.8±1.8 in known DM patients (p<0.001; 95% CI=1.7-2.1). Patients with admission glucose >180mg/dL had significantly higher HbA1c (9.0±1.9) compared to lower glucose values (6.1±1.1, median 5.9; p<0.001; 95% CI=2.5-3.2). Importantly, of those with an admission glucose >180mg/dL, 94% met the diagnosis of DM by HbA1c, although 13% had not been previously diagnosed. Discharge mRS was significantly worse at a median of 4 (unable to walk independently) vs. 3 (p=0.001) in patients with admission glucose >180mg/dL, despite having equivalent median admission NIHSS of 3 in each group (p=NS). Conclusions: Admission glucose >180mg/dL should be recognized as a strong indicator of DM diagnosis and not be assumed to be stress mediated. Stroke nurses are encouraged to lead the early diagnosis/management of diabetes.