OBJECTIVE: To determine the role of growth hormone (GH) in the development of diabetic retinopathy. RESEARCH DESIGN AND METHODS: Medical records of 1,423 patients who had undergone insulin tolerance tests (1976-1991) at the Mayo Clinic were examined, and diabetic subjects were identified as either GH-deficient (GH increment after hypoglycemia < 5 micrograms/L and peak < 10 micrograms/L) or GH-sufficient. Prevalence of retinopathy was determined in these cases and in a cohort group of diabetic subjects selected to match the GH-deficient cases. These control patients (32 cases) were selected from medical records of individuals who had received medical care at Mayo during the same interval but who had not undergone insulin tolerance testing. RESULTS: Twenty-four patients with diabetes were identified, of whom 16 were GH-deficient and 8 GH-sufficient. Despite comparable age, duration of diabetes, and metabolic control, the prevalence of diabetic retinopathy in the GH-deficient group (2 of 16; 12.5%) was less (P < 0.05) than that observed in the GH-sufficient group (5 of 8; 62.5%). Prevalence in the GH-deficient group also was lower than that observed in the cohort control group (15 of 32, 47%). CONCLUSIONS: These data strongly suggest that GH contributes to the development of diabetic retinopathy in humans.
OBJECTIVE: To determine the role of growth hormone (GH) in the development of diabetic retinopathy. RESEARCH DESIGN AND METHODS: Medical records of 1,423 patients who had undergone insulin tolerance tests (1976-1991) at the Mayo Clinic were examined, and diabetic subjects were identified as either GH-deficient (GH increment after hypoglycemia < 5 micrograms/L and peak < 10 micrograms/L) or GH-sufficient. Prevalence of retinopathy was determined in these cases and in a cohort group of diabetic subjects selected to match the GH-deficient cases. These control patients (32 cases) were selected from medical records of individuals who had received medical care at Mayo during the same interval but who had not undergone insulin tolerance testing. RESULTS: Twenty-four patients with diabetes were identified, of whom 16 were GH-deficient and 8 GH-sufficient. Despite comparable age, duration of diabetes, and metabolic control, the prevalence of diabetic retinopathy in the GH-deficient group (2 of 16; 12.5%) was less (P < 0.05) than that observed in the GH-sufficient group (5 of 8; 62.5%). Prevalence in the GH-deficient group also was lower than that observed in the cohort control group (15 of 32, 47%). CONCLUSIONS: These data strongly suggest that GH contributes to the development of diabetic retinopathy in humans.