Background: Diabetes is an increasingly important risk factor for ischemic stroke and worsens stroke prognosis. Yet a large proportion of stroke patients who are eventually diabetic are undiagnosed. Therefore, it is important to have accurate assessment of glycaemia in stroke patients.
Design: Secondary outcome analysis of a randomized controlled trial focussing on parameters of glucose metabolism and detection of diabetes and prediabetes in patients with acute ischemic stroke (AIS).
Methods: a total of 130 consecutively admitted patients with AIS without previously known type 2 diabetes mellitus (T2DM) were screened for diabetes or prediabetes as part of secondary outcome analysis of a randomized controlled trial that tested lifestyle intervention to prevent post-stroke cognitive decline. Patients had the oral glucose tolerance test (OGTT) and glycated hemoglobin (HbA1c) measurements in the second week after stroke onset and after one year. The detection rates of diabetes and prediabetes based on the OGTT or HbA1c values were compared.
Results: By any of the applied tests at the second week after stroke onset 62 of 130 patients (48%) had prediabetes or T2DM. Seventy-five patients had results from both tests available, the OGTT and HbA1c; according to the OGTT 40 (53.3%) patients had normal glucose metabolism, 33 (44%) had prediabetes, two (2.7%) T2DM. In 50 (66.7%) patients the HbA1c results were normal, 24 (32%) in the prediabetic and one (1.3%) in the diabetic range . The detection rate for disorders of glucose metabolism was 10% higher (absolute difference; relative difference 29%) with the OGTT compared with HbA1c. After one year the detection rate for prediabetes or T2DM was 7% higher with the OGTT (26% relative difference). The study intervention led to a more favourable evolution of glycemic status after one year.
Conclusions: Many patients with acute stroke have unrecognised diabetes or prediabetes. The OGTT is a more sensitive screening tool than HbA1c for the detection of previously unrecognised glycemic disorders in patients with acute stroke with an at least 25% relative difference in detection rate. Therefore, an OGTT should be performed in all patients with stroke with no history of diabetes. Measuring HbA1c alone will lead to a false assurance of normal glycemic status in many stroke patients.