Hyperglycemia exacerbates neurologic damage in clinical and experimental central nervous system ischemia. The purpose of our study was to determine if insulin administration before significantly alters neurologic deficit and survival after ischemia using a newly developed rat cerebral ischemia model. One hour before the onset of ischemia, 40 200-300 -g Sprague-Dawley rats received intraperitoneal injections of either 1 ml normal saline or 0.4, 0.5, or 0.6 units regular insulin in 1 ml normal saline. Rats were then intubated and ventilated with 1-1.5% halothane. The aortic arch was exposed, and snares were placed on the innominate, left carotid, and left subclavian arteries. A 20-minute occlusion was begun, and anesthesia was discontinued. Baseline plasma glucose concentration was similar (p=0.48, Student's t test) in both groups, but it subsequently was significantly lower in the 0.4 unit insulin-treated group up to 4 hours after occlusion (p<0.0035, Student's t test). Neurologic deficit was scored on a 50-point scale (0=normal, 50=severe deficit) 1, 4, 18, and 24 hours after occlusion. In the 0.4 unit insulin-treated group the neurologic deficit score was significantly lower than in the salinetreated group 1, 4, 18, and 24 hours after occlusion (p<0.005, Student's t test). Survival was significantly higher (p=0.001) in the 0.4 unit insulin-treated (1.7 unit/kg dose) group than in the saline-treated group. No rats died when preocclusion plasma glucose concentration was between 65 and 175 mg/dl. We conclude that insulin can be used to maintain plasma glucose concentration within a normoglycemic to mildly hypoglycemic range, which is associated with decreased behavioral neurologic deficit and increased survival following temporary cerebral ischemia. Hyperglycemia or more extreme hypoglycemia (glucose concentration less than approximately 65 mg/dl) is detrimental in this model. (Stroke 1988;19:1411-1419) H yperglycemia has been shown to exacerbate neurologic deficit in many experimental models of central nervous system (CNS) hypoxic-ischemic injury.
-8 Clinical studies also indicate an adverse effect on outcome when CNS ischemia occurs in the setting of hyperglycemia. Diabetic patients experiencing a stroke have significantly worse outcomes than nondiabetic stroke patients. 9 Other studies have confirmed an increased mortality rate in acute stroke patients with fasting plasma glucose concentrations of >110 mg/dl. 10 These findings suggest that even relatively mild hyperglycemia, whether exogenously or endogenously induced, significantly worsens postischemic neurologic outcome. In the clinical setting, hyperglycemia resulting from exogenous glucose administration is potentially preventable but often inadvertently produced by administration of intravenous fluids. A more difficult problem arises in dealing with endogenous elevations of plasma glucose concentrations in patients at risk for ischemic neurologic injury. Examples include cardiac arrest patients and evolving stroke patients with concurrent hyperglycemia and...