Clinical Question

In patients with acute ischemic stroke, is glycemic control with insulin effective in decreasing death and disability?

Evidence-Based Answer

Administering insulin to maintain glycemic control does not improve the rates of mortality or dependency, or the final neurologic outcome after an acute ischemic stroke. Additionally, in randomized controlled trials, patients in the intervention groups (those under tight glycemic control) had a 25-fold higher incidence of symptomatic hypoglycemia. (Strength of Recommendation: A, based on consistent, good-quality, patient-oriented evidence.)

Practice Pointers

Hyperglycemia occurs in up to two-thirds of patients with acute ischemic stroke and predicts increased stroke mortality, independent of age and stroke severity. Compared with their normoglycemic counterparts, patients with hyperglycemic stroke are more than two times as likely to be dead at 90 days after stroke onset. Although some maintain that hyperglycemia is a physiologic response to the stress of an acute ischemic stroke, others posit that it augments stroke injury by enhancing cortical intracellular acidosis. Acidosis promotes glutamate release, which leads to cortical depression and necrosis of tissue in the ischemic penumbra, the potentially salvageable area surrounding the infarct.

The purpose of this Cochrane review was to determine whether controlling hyperglycemia during acute ischemic stroke influenced patient outcomes. The authors found seven eligible trials involving 1,296 participants with and without diabetes mellitus. Participants in the intervention groups were given insulin infusions with or without mealtime subcutaneous insulin, with varying tight glucose targets (range = 70 to 110 mg per dL [3.89 to 6.11 mmol per L] to 90 to 144 mg per dL [5.00 to 7.99 mmol per L]). Patients in the control groups were given insulin infusion for a loose target of 70 to 200 mg per dL (3.89 to 11.10 mmol per L), subcutaneous insulin as needed for glucose levels in excess of 180 to 306 mg per dL (9.99 to 16.98 mmol per L), placebo, or no treatment.

The mean glycemic level was significantly lower in the intervention groups (mean difference = –8.3 mg per dL; 95% confidence interval [CI], –12.1 to –4.5 mg per dL). However, there were no differences among the groups in dependence (defined as being severely dependent on others in activities of daily living) or deaths at 30 or 90 days, or final neurologic deficit (as measured by the National Institutes of Health Stroke Scale and the European Stroke Scale). Stratifying for studies with higher and lower numbers of patients with diabetes did not alter this result. There was an increased risk of symptomatic hypoglycemia in the treatment groups (odds ratio = 25.9; 95% CI, 9.2 to 72.7).

The 2007 American Heart Association (AHA) clinical practice guideline on the early management of ischemic stroke offers a consensus opinion that “The minimum threshold [to treat hyperglycemia]…likely was too high, and lower serum glucose concentrations (possibly > 140 to 185 mg/dL) probably should trigger administration of insulin.”¹ However, the AHA guideline was published before the release of the negative results of the Glucose Insulin in Stroke Trial (the largest trial to date), and a growing body of literature has since demonstrated that tight glycemic control in critically ill patients does not provide benefit and may promote harm.² Although stroke-related hyperglycemia is associated with greater brain injury, poorer functional outcomes, and increased mortality, intensive therapy to lower blood glucose levels does not appear to improve functional outcomes or mortality rates.