Original Research Reports The Association of Lifetime Psychiatric Dlness and Increased Retinopathy in Patients With Type I Diabetes Mellitus S. TZIPORAH COHEN, M.D., GARRY WELCH, PH.D. ALAN M. JACOBSON, M.D., MARy DE GROOT, ED.M. JACQUELINE SAMSON, PH.D. Forty-nine patients with Type I diabetes mellitus were assessed to examine the relation- ship between lifetime prevalence of psychiatric illness and retinopathy severity. The subjects with a history of psychiatric illness had significantly worse retinopathy than the subjects without psychiatric illness. Eighty-nine percent of the subjects with severe nonproliferative retinopathy or proliferative retinopathy had a history of psychiatric illness, predominantly affective illness. In addition, the subjects with a history of psy- chiatric illness had significantly higher current glycohemoglobin levels than those with no psychiatric history. This study:SO findings suggest that psychiatric illness may be a riskfactor for development of retinopathy in Type I diabetic patients. (Psychosomatics 1997; 38:98-108) S everal studies have reported that rates of psychiatric illnesses, most specifically de- pression, are more prevalent in patients with Type I diabetes mellitus than in the general population. ' -6ln addition, there is now evidence from cross-sectional studies that comorbid de- pression is associated with poor glycemic con- troI. 2 • 6 . 7 For example, Lustman and colleagues 2 found that lifetime diagnosis of a psychiatric disorder in a group of subjects was associated with higher levels of glycosylated hemoglo- bin, compared with the subjects without a Received June IS. 1995; revised August 10. 1995; accepted October 13. 1995. From the Mental Health Depart- ment of the Joslin Diabetes Center. Harvard Medical School. and Mclean Hospital. Boston. Massachusetts. Address re- print requests to Dr. Jacobson. Joslin Diabetes Center, One Joslin Place. Boston. MA 02215. Copyright C 1997 The Academy of Psychosomatic Medicine. 98 psychiatrichistory, and that this association was accounted for by the subgroup of patients with major depression. Similarly, Robinson and col- leagues 7 found that the patients with a recent history (within I month) of major depressive disorder or borderline depression experienced worse blood-glucose control than the patients without any history of psychiatric illness. In a recent study by our group, the patients with Type I diabetes who had a lifetime history of major depression had worse current glycemic control, as measured by HbA" than the Type I patients with no history of Axis I psychiatric disorders. 6 In addition, the patients with the most recent episode of depression had the high- est HbA, levels. The influence of poor glycemic control on the development and progression of diabetes complications is now well established. Epi- demiologic studies have consistently shown an PSYCHOSOMATICS