Richard Samuel Surwit
Professor Emeritus in Psychiatry and Behavioral Sciences

Our research program is focused the role of stress, depression and personality variables in the etiology and treatment of diabetes mellitus. We have recently finished a large clinical trial of stress-management in the treatment of type 2 diabetes that demonstrated that a brief, group, stress management intervention can improve long-term glycemic control in a large sample of patients with type 2 diabetes. In addition we have two other projects currently ongoing. The first is concerned with the impact of a cognitive behavioral program in the treatment of diabetes. Extensive literature documents increased incidence of depression in patients with diabetes mellitus. However, the degree to which the treatment of depression impacts on diabetes control is not clear. Some investigators have found a strong association between depression and diabetes control, while others have found a week assocation. A review of existing literature as well as preliminary data from our laboratory suggest that this discrepancy may be due, in part from failure of many studies to clearly differentiate type 1 and type type 2 diabetes in their patient samples. The strongest association between depression and diabetes control has been reported in studies of patients with type 1 diabetes. Preliminary data from our group suggests that the Beck Depression Inventory (BDI) scores may be more significantly related to hemoglobin A1c (HbA1c) in type 1, than in type 2, diabetes mellitus. Studies assessing the effects of pharmacologic antidepressant therapy on diabetes have been confounded by the combined use of both diagnostic categories as well as by the direct metabolic effects of most anitidepressant drugs. An emerging literature on the use of Cognitive Behavior Therapy (CBT) suggests that improving affect through behavioral intervention can improve diabetes control, particularly in type 1 patients. The overall aim of this project is to determine if CBT differentially improves glucose control in type 1 and type 2 diabetes patients. We hypothesize that CBT will produce a greater reduction in HbA1c in type 1 diabetes than in type 2 diabetes and that CBT-induced improvement in HbA1c is mediated by an improvement in depression. We will measure the effects of CBT on changes in HbA1c and daily blood glucose in a sample of 150 depressed patients with type 1 and type 2 diabetes over one year. To evaluate the mechanism by which CBT-induced changes in depression affect blood glucose, we will determine the role of cortisol and the role of diabetes self-care behaviors as mediating variables. Changes in cortisol and self-care are predicated to impact blood glucose levels to a greater extent in type 1 diabetes because these individuals are metabolically more sensitive to any variation.

The second major focus of our research program is to evaluate whether the apparent differential relationship between the personality construct of hostility and glucose metabolism in blacks and whites might contribute to the racial disparity in the prevalence of 2 diabetes. African-Americans are at increased risk for developing type 2 diabetes, but the reasons for this racial disparity are poorly understood. We are exploring our previous finding of a differential relationship of hostility to glucose metabolism in African-Americans and Caucasians and determine the underlying behavioral and/or physiologic mechanisms of these relationships. Two studies are being undertaken. First, we are assessing the multi-factorial nature of hostility in 100 African American men, 100 African American women, 100 Caucasian men, and 100 Caucasian women. These measures will be examined in relation to fasting and two-hour post-prandial glucose and insulin and hemoglobin A1c (HbA1c) on all subjects. This study should allow us to further understand the personality variables most directly related to glucose metabolism and to investigate the possibility that race and sex interact in determining this relationship. We also plan to further explore our preliminary findings that hostility is related to insulin sensitivity and elevated insulin levels in Caucasians and to fasting glucose in African-Americans. We will assess glucose-stimulated insulin release, glucose disposal, hepatic glucose production and glucose effectiveness utilizing an intravenous glucose tolerance test with labeled glucose in a subset of our sample. In addition, three sets of mediators will be examined: measures of hypothalamic-pituitary-adrenal (HPA) axis activity, measures of body mass and fat distribution, and measures caloric intake and exercise habits. It is hypothesized that hostility is related to diminished insulin sensitivity, but not to a defect in insulin secretion in Caucasians and that this relationship is mediated by BMI and the behavioral variables that impact on BMI. Furthermore, we hypothesize that hostility is related to increased hepatic glucose production and diminished insulin secretion in African-Americans and these relationships are mediated by neuroendocrine factors, but not by BMI. Confirmation of this hypothesis would confirm that African-Americans are more vulnerable to the metabolic effects of hostility and could partially explain the racial disparity in diabetes prevalence.

Finally, our laboratory continues to work on defining the mechanisms by which the C57BL/6J (B6) mouse develops diet-induced obesity and diabetes. This mouse will remain lean and euglycemic if it is raised on a low fat diet. However, when allowed access to a high fat diet, the B6 mouse will become obese and develop an analog of type 2 diabetes, even if caloric intake is matched to that of mice reared on a low fat diet.

Current Appointments & Affiliations

Contact Information

  • 3rd Floor Duke South Room 3087, Box 3862, Durham, NC 27710
  • Box 3862 Med Ctr, Durham, NC 27710

Some information on this profile has been compiled automatically from Duke databases and external sources. (Our About page explains how this works.) If you see a problem with the information, please write to Scholars@Duke and let us know. We will reply promptly.