Neurologic damage following cardiac surgery continues to be an important cause of postoperative morbidity and loss of social functioning. Recent evidence has shown patients with diabetes are at risk for larger and more numerous brain infarctions following cardiac surgery and thus research into classification, management and outcome in patients with diabetes is important. At the present time about one third of patients having Cardiac Surgery have a history of diabetes, whereas 50-70% requires insulin in the operating room or later to control blood sugar. This research proposal will outline a comprehensive program for demonstration of genotype, phenotype and outcome in a group of diabetic patients undergoing cardiac surgery. We will undertake study of 530 patients over five years which will demonstrate current incidence of glucose intolerance and risk of poor outcome under optimum classification and management strategies. With accurate classification and optimal management, we would anticipate significant improvement in Neurologic, Neurobehavioral and Overall outcome coupled with fewer MRI brain infarctions at one week. This clinical study will take the form of a pilot program which could grow into a much larger and more comprehensive multi- institutional research program. Significant new information about the genetics, pathogenesis and clinical outcomes in patients with Diabetes Mellitus has appeared in the literature. Since neurologic and other clinical outcomes in diabetic patients having Cardiac Surgery are suboptimal when compared to nondiabetic patients clinical research is needed to provide guidelines and algorithms for surgical therapy. The research proposed in this grant application will carefully study the genetics and neurological outcomes of patients undergoing Cardiac Surgery randomly assigned to tight insulin management compared with traditional management protocols. ? ? ?