Glucose Level and Brain Infarction: A Prospective Case-Control Study and Prospective Study

Abstract

Hyperglycaemia in the acute phase of stroke has been established as a predictor of higher mortality. But recent data regarding active treatment of hyperglycaemia showed no clinical benefit suggesting that hyperglycaemia may not have a detrimental effect in brain infarction. Additional data are needed to resolve this uncertainty and identify patients at higher risk if any. A total of 477 adult Caucasian patients with brain infarction and 395 age- and sex-matched controls admitted at the same centres for nonneurological causes were recruited consecutively from 12 neurological centres in France. Electrocardiographic, carotid ultrasonography, and transcranial Doppler studies were performed. Blood was drawn in the morning from fasting subjects for glucose measurement. Functional outcome was measured on admission, at 10 days and at 6 months after the onset of stroke using the modified Rankin scale. Among 477 brain infarction patients and 395 hospitalised controls the adjusted mean (+/-SEM) glucose level was higher in cases (6.4+/-1.0 mmol/l) than in controls (6.0+/-1.01 mmol/l, P=0.006), with a significant heterogeneity across sexes. The fully adjusted odds ratio of brain infarction per 1-standard deviation increase in log-glucose level was 1.02 (95% confidence interval, 0.77-1.37) in men and 2.21 (95% confidence interval, 1.44-3.40) in women. Among the 477 brain infarction cases elevated admission glucose levels were associated with poor outcomes and higher poststroke mortality after adjustment for conventional vascular risk factors and infarct volume. These relationships were not modified by sex. Elevated admission glucose levels were associated with brain infarction in women only and with a higher 5-year mortality. Further investigation focusing on the impact of glucose level in different target population is needed to optimise glycaemic management in acute brain infarction patients.