Backgrounds:
In primary prevention trials, tight glycemic control (below glycated hemoglobin level 6.5%), could not reduce macrovascular events. Cardiovascular mortality may increase due to myocardial infarction or arrhythmias triggered by serious hypoglycemia. In ischemic stroke, the stroke event of patients with tight long-term regulation of plasma glucose may associate with cardioembolism etiology and worse clinical outcome.
Methods:
From our stroke registry, we selected stroke patients with a history of diabetes, who were classified into two groups using glycated hemoglobin level 6.5%, for comparing stroke subtypes, clinical characteristics, and outcome. Good outcome was defined as modified Rankin scale 0, 1, and 2 at 3 months. One-year cardiovascular event and mortality were assessed for long-term outcome.
Results: 
Among 1,147 patients, mean age was 66.4 years (SD 10.7) and male was 719 (62.7%). Tight group (230, 19.7%) had more likely to have older age, hypertension, previous stroke, atrial fibrillation, better lipid profile, and lower BMI. Cardioembolism subtype was more common in tight group (25.7% vs.19.5%, p=0.05). Good outcome was 65.3% (vs. 74.9%, p=0.005) and one-year mortality was higher in tight group (19.1% vs. 10.8%, p<0.001). Fatal cardiovascular event was also higher (13.5% vs. 8.4%, p=0.026). After adjusting for compounding factors, tight group was an independent factor for one-year mortality (OR 1.59, 95% CI 1.04–2.38, p=0.028).
Conclusion:
In diabetes, long-term tight control of hyperglycemia may negatively affect early outcome and one-year mortality after a stroke event. Stroke patients with a low glycated hemoglobin level should be carefully evaluated and managed for better cardiovascular outcome.

Influence of tight glycemic control on ischemic stroke outcome in patients with diabetes mellitus.