Admission glucose level in relation to mortality and morbidity outcome in 252 stroke patients.

In a prospective study to correlate admission glucose level with neurologic outcome in stroke, 252 acute stroke patients without prior disability and admitted within 24 hours of onset of ictus were assessed. The stroke was classified into one of three types--cortical infarct, lacunar infarct, or intracerebral hemorrhage--by clinical, computed tomographic, and necropsy findings. Fifty-one diabetic patients were excluded from the entire cohort to form a nondiabetic category for analysis. We found that admission glucose level showed a significantly higher degree of correlation with mortality and morbidity (measured as arm function, leg function, and activities of daily living) when cortical (n = 118) and lacunar (n = 58) infarcts were pooled compared with when they were assessed separately. For intracerebral hemorrhage (n = 76), admission glucose level correlated with mortality but not morbidity. This trend persisted despite exclusion of diabetic patients. These results are consistent with previous observations of a correlation between a high admission glucose level and the severity of stroke. The importance of segregating cortical from lacunar infarcts, two groups with a different natural history and prognosis, in any future analysis is emphasized.