Growth hormone secretion in diabetic retinopathy

Summary The plasma HGH response to insulin-induced hypoglycemia (0.2 U/kg) and the 24-h plasma HGH pattern during a normal day have been studied in 14 non obese long-term insulin-dependent diabetics with proliferative retinopathy, mean age 39 ± 2 (ranging between 24 and 50 years). Plasma glucose and FFA were also determined. The results were compared with those of 18 normal subjects of similar age and weight. The mean plasma HGH response to insulin in retinopathic diabetics was slightly lower (with no significant differences) than in controls in whom hypoglycemia was induced with a smaller dose of insulin (0.1 U/kg). This pattern of plasma HGH could be related to the delayed plasma glucose fall observed in retinopathic diabetics in comparison to normal subjects, even if the HGH peak after insulin in both groups (18.61 ± 4.32 ng/ml in retinopathic diabetics, 27.43 ± 4.19 in controls) did not seem to be correlated to the degree of hypoglycemia, but rather to the age of the subjects. Plasma HGH pattern, studied with blood samples taken every three hrs during a normal day, did not reveal differences between the diabetics and controls. Plasma glucose, however, was higher in retinopathic diabetics than in controls in spite of the insulin treatment. These results show that in diabetic patients with retinopathy, increased HGH secretion does not occur in conditions of ordinary life or after insulin-induced hypoglycemia, although the HGH plasma levels observed in retinopathic diabetics could be considered too high in relation to the elevated blood glucose levels. Traduzione a cura degli AA.     

Plasma growth hormone response to insulin-induced hypoglycemia in diabetes mellitus

Summary Plasma human growth hormone (HGH) response to insulin-induced hypoglycemia was studied in patients with diabetes mellitus and age and sex matched normal subjects. The diabetics comprised two groups: (a) those with retinopathy, (b) those without retinopathy. Adequate hypoglycemia (50% or a greater fall in fasting blood glucose) was achieved by i.v. administration of insulin 0.1 U/kg body weight in normal subjects and 0.2 U/kg in diabetics. Mean fasting plasma HGH levels did not differ significantly between control subjects (1.39 ± 0.25 SEM, ng/ml), diabetics without retinopathy (1.55 ± 0.25) and diabetics with retinopathy (1.81 ± 0.6). There was a significant rise in plasma HGH, following insulin-induced hypoglycemia in all three groups. When mean Δ HGH or sum HGH value in different groups were compared, no significant difference was observed. However, mean HGH-hypoglycemia index (Δ HGH · 100/nadir blood glucose) was significantly lower in the two diabetic groups as compared to that obtained in control subjects, but it was identical in diabetics with or without retinopathy. In 32% of diabetic patients abnormal blood glucose and plasma HGH responses were observed.     

Dopaminergic stimulation of HGH in diabetes mellitus and in obesity

Summary The behavior of HGH in basal conditions and after L-Dopa infusion was studied in a group of patients with diabetic retinopathy, in 9 obese and 8 control subjects. In both diabetics and obese subjects, increases found in HGH plasma levels after administration of the drug were slighter than in the controls. On the basis of these results it can be concluded that in diabetics the availability of energetic substrates may modify the HGH response to the L-Dopa stimulus. In obesity, the possibility is considered of a reversible defect in the sensitivity of the dopaminergic receptors, induced by metabolic and endocrine factors.     

Growth hormone and cortisol secretion after propranolol--glucagon testing in the adult

Pituitary HGH response to propranolol-glucagon stimulation in a group of obese subjects (n = 45), patients with primary hypothyroidism (n = 13), and patients with hypopituitarism (n = 15) was compared to the response observed in normal subjects (n = 40). In normal and obese subjects, the magnitude of the HGH response to the same stimulus was compared with the results obtained with the insulin tolerance test. In addition, the cortisol response to propranolol-glucagon stimulation in normal subjects (n = 19) was compared to that obtained after insulin hypoglycemia. In the group of normal subjects, a positive HGH response after the propranolol-glucagon test (i.e., a peak value ≥ 9 ng/ml) was found in 38 of the 40 subjects tested. The mean HGH peak value (24.5 ± 2.1 ng/ml) was significantly lower than that (41.1 ± 2.8 ng/ml) on insulin testing (p < 0.001). In 45 overtly obese subjects, a mean peak value of 11.0 ± 1.3 ng/ml was reached, a significantly blunted response as compared to the group of normal subjects (p < 0.001). Again the HGH peak after propranolol-glucagon stimulation was significantly lower than the mean HGH response on insulin testing (18.7 ± 1.6 ng/ml) in a comparable group of obese individuals (p < 0.0025). A consistently blunted HGH response was observed in all the patients with pituitary insufficiency and in ten out of 13 patients with unsubstituted primary myxedema. The increase in plasma cortisol levels after propranolol-glucagon stimulation in normal subjects was comparable to the response after insulin hypoglycemia. It is concluded that the propranolol-glucagon test is a safe and reliable procedure, although normal “nonresponders” do exist, and that it represents a somewhat weaker stimulus for HGH release by the pituitary than postinsulin hypoglycemia. As plasma cortisol levels rise consistently, the test can also be used for the evaluation of the pituitary-adrenal axis.     

Transsphenoidal surgery in hypersecreting pituitary tumors: endocrine follow-up (author's transl)

Among 18 patients suffering prolactin adenoma and 16 acromegalic patients, plasma levels of PRL and HGH respectively returned to normal vlues (PRL = 9 +/- 2 ng/ml HGH = 4.2 +/- 0.8 ng/ml) immediately after selective removal of the pituitary adenoma by transspheno?dal route. In all cases tumoral symptoms were reduced. Menses were recovered 36 +/- 5 days after surgery in all patients with prolactin adenoma. Pregnancy succeeded in nine of these women. Per-operative studies of HGH and PRL plasma variations proved to be of good prognostic value. The evaluation of the pituitary functions was performed before and after surgery. In no case pituitary deficiency occured after treatment. Furthermore the regulation of somatotropic and prolactin functions have been documented by the use of different provocative tests before and after treatment. They allow discussion of the pathogenecity of these hypersecreting tumors.     

Human growth hormone and cortisol response to insulin stimulation in aging.

The influence of age on plasma growth hormone (HGH) and cortisol response to i.v. insulin (0.1 U/kg of body weight) was evaluated in 32 healthy subjects whose ages ranged between 20 and 84 years. A significant reduction in HGH response to insulin was observed with aging. In the young (20-34 years), middel-aged (35-49 years), and elderly (53-84 years) groups, average HGH peaks were 46.51 +/- 7.37, 29.95 +/- 5.35, and 14.31 +/- 2.39 ng/ml while average HGH areas were 2.911 +/- 0.484, 1.654 +/- 0.316, and 0.699 +/- 0.149 mug-min, respectively. Since insulin's hypoglycemic effect became less rapid with aging, this could, in part, explain the progressive decline in the HGH response to insulin. This phenomen may also be attributed to histological changes occurring in the pituitary with aging. Moreover, cortisol response was similar to all three age groups. These findings suggest that, while HGH response to insulin is correlated with age, adrenal response does not show any important modifications with aging.     

Diabetic coma: Serum growth hormone before and during treatment

Summary Serum growth hormone values in 37 patients with diabetic ketoacidosis were 5.4±0.8 ng/ml (S.E.M.) in males and 6.7±1.1 ng/ml in females before treatment; while in five hyperosmolar non-ketotic patients the HGH concentration was 3.9±0.5 ng/ml. One hour after insulin 90% of patients showed a rise in HGH, to a mean of 33.7±9.8 ng/ml for males and 25.5±6.0 ng/ml for females in ketoacidosis; and to 27.1±9.9 ng/ml for hyperosmolar coma patients. The rise, which was transient, was inversely correlated with pretreatment plasma glucose, the l h plasma glucose concentration and plasma urea, and directly proportional to the % fall in blood glucose after 1 h. When the ketoacidosis patients were divided into two groups according to HGH response, those with a small response had the greater disturbances of plasma glucose, blood ketone bodies, blood lactate, plasma urea, blood pH, and blood pressure, the smaller 1 h fall in blood glucose, and the higher mortality. Thus the most severely ketoacidotic patients had the poorest growth hormone response. Growth hormone is probably of little importance as an insulin antagonist in diabetic coma.Presented in part at the Spring Meeting of the British Diabetic Association, York, April 1972.     

EFFECT OF l-DOPA ON THE SECRETION OF PLASMA GROWTH HORMONE IN CHILDREN AND ADOLESCENTS

The stimulation of growth hormone release in children by L-dopa has been studied. An oral dose of 0.5 g L-dopa was administered to fourteen children with, and to fifteen children without, hypothalamic-pituitary insufficiency. In the control group, L-dopa induced a release of pituitary growth hormone, the peak of which occurred from 30 to 60 min after ingestion. Nine out of fifteen control subjects showed peak levels of plasma growth hormone greater than 8 ng/ml. None of the patients with hypothalamic-pituitary insufficiency showed levels greater than 5 ng/ml. In five out of six children with measurable amounts of plasma HGH, and in fourteen children with a lack of HGH, there was a good correlation between the HGH response after L-dopa, insulin hypoglycaemia and arginine infusion. It is concluded that the administration of one oral dose of L-dopa can be used as a provocative test of growth hormone secretion.     

Effect of aging on growth hormone,ACTH and cortisol response to insulin-induced hypoglycemia in type I diabetes

Summary The influence of age on plasma growth hormone (GH) response to i.v. insulin (0.2 U/kg body weight) was evaluated in clinically stable type I (insulin-dependent) diabetics divided into four age groups (range 18–80 years). ACTH and cortisol were also assayed in two groups of diabetics under and over 50 years of age. A significant reduction with aging in GH response to insulin was observed. On the contrary, the glucose fall was similar in all the groups. ACTH and cortisol responses to insulin were slightly decreased in the older diabetics. Since insulin-induced hypoglycemia was similar in all the age groups, the progressive decline with aging in the GH response to insulin may be attributed to age-related changes of the pituitary gland. The data on ACTH and cortisol are less striking. Our data, as a whole, confirm that growth hormone response to insulin-induced glucose fall is not critical in acute glucose counterregulation in insulin-dependent diabetics. In fact, in spite of a 4-fold difference in GH levels, there was no difference in the 2-h glycemic course after 0.2 U/kg of i.v. insulin between young and aged patients. When a group of 26 type I diabetics with proliferative retinopathy was compared with a group of age-matched type I diabetics without retinopathy and with 30 age-matched normal subjects (injected i.v. with 0.1 U/kg body weight of insulin), no difference was found in GH response to insulin, indicating that GH hypersecretion is not a characteristic finding of diabetic retinopathy. This study was supported by grant N. 83.02749.56 fromConsiglio Nazionale delle Ricerche.     

Long-term treatment with 2-Br-alpha-ergocryptine in acromegaly.

Thirty acromegalic subjects underwent chronic CB154 therapy (10-20 mg daily) for periods ranging from 3 months up to 2 years. In 18 out of 21 patients, who exhibited consistent HGH reduction following acute administration of the drug, there was also during chronic treatment, a suppression of the plasma HGH levels exceeding 50% of base line values, e.g. from mean daily values between 14-197 ng/ml (mean +/- SEM = 57.8 +/- 12.4 ng/ml pre-treatment) to 2-19 ng/ml (mean 8.3 +/- 1.2 ng/ml post-treatment). In 12 of the subjects who responsed to chronic CB154 treatment, the mean daily values of HGH were below 10 ng/ml. The suppression of plasma HGH was maintained unaltered throughout the whole course of therapy. In the 9 subjects, in whom no consistent HGH decrease was evidenced with acute CB154 administration, there was accordingly a minor or no suppression of HGH values during the chronic treatment. In 13 subjects, irrespective of the degree of their GH responses, the plasma prolactin levels were constantly inhibited by CB154; instead the drug failed to modify significantly the TRH or insulin-induced GH release. These changes in the hormonal parameters were paralleled by marked clinical amelioration and improvement of some of the metabolic alterations frequently encountered in acromegaly, e.g. reduced carbohydrate tolerance, increased insulin resistance, diminished fall of plasma phosphorus after insulin, decreased urinary excretion of phosphate, hyper-hydroxyprolinuria and hyper-calciuria. Collectively, these data demonstrate that CB154 thrapy is effective in reducing HGH hyper-secretion in many acromegalic patients during long-term treatment.     

Nanisme hypophysaire familial par déficit isolé en somathormone (3 cas)

The authors report the case of three sisters aged 73, 75 and 81 years with dwarfism. They had normal body proportions, underwent normal puberty, had normal secondary sexual characteristics and one had normal reproduction capacity. Two were overt diabetics and the third had latent diabetes.Low HGH levels under basal conditions and after arginine infusion was suggestive of selective HGH deficiency, other hormonal parameters being normal. The insulin response to arginine was moderate.One of the patients died. Post mortem examination showed no peripheral or coronary atherosclerosis but severe cerebral atheroma associated with cortical and subcortical atrophy. Given the age of the patient, the genital organs were normal. The endocrine glands were also normal except the hypophysis, the number of somatotrophic cells of which was significantly reduced. No diabetic angiopathy was demontrated.The genetic study showed an autosomal recessive inheritance, and therefore the condition was a sex-linked familial dwarfism due to selective HGH deficiency. HGH deficiency would appear to give rise to diabetes after a variable period of time and would also appear to protect against diabetic microangiopathy. These observations suggest that HGH plays a role in the mechanism of diabetic microangiopathy in patients with insulin deficiency and/or hyperglycemia.     

Effect of streptozotocin diabetes and insulin replacement on growth hormone in rats

The effects of insulin deprivation on the growth rate, plasma and pituitary growth hormone (GH) and GH synthesis were investigated in male Wistar rats. Diabetes was induced by administration of streptozotocin (STZ), 7 mg/100 g bw, and plasma and pituitary GH levels were measured by means of a specific radioimmunoassay. GH synthesis was determined in pituitaries by the in vitro incorporation of [3H] leucine into specific immunoprecipitates. The body weight and the pituitary GH content of normally developing rats showed an almost linear increase throughout the observation period, whereas diabetic rats stopped growing immediately after receiving STZ, and remained smaller than age-paired controls. Pituitary GH content remained within the control range through the 5 days following STZ administration and thereafter decreased reaching 10% of control values by the 30th day. Furthermore, pituitaries from diabetic rats incorporated [3H] leucine into r-GH at a greatly reduced rate, which could explain the diminished r-GH storage in pituitaries of diabetic rats. Plasma GH concentrations remained within the normal range for 10 days after STZ, thereafter plasma GH were markedly reduced. Insulin treatment prevented the metabolic changes, and restored normal levels of plasma and pituitary GH in diabetic rats. These findings indicate that diabetes, in rat, is characterized by an inhibitory effect on GH secretion, probably via a diminished GH synthesis by the pituitary gland.     

Role of ghrelin in streptozotocin-induced diabetic hyperphagia

Ghrelin, an endogenous ligand for the growth hormone (GH) secretagogue receptor, was originally purified from the rat stomach. We have previously reported that central administration of ghrelin increases food intake and body weight. To investigate the role of ghrelin in the hyperphagic response to uncontrolled diabetes, adult male rats were studied 14 days after administration of streptozotocin (STZ) or vehicle. STZ-treated diabetic rats were markedly hyperphagic. This hyperphagia was accompanied by hyperglycemia, hypoinsulinemia, and reduced plasma GH levels. Treatment of diabetic rats with insulin reversed these changes. Plasma ghrelin concentrations in untreated diabetic rats were significantly higher than in control rats and were normalized by insulin treatment. The ghrelin gene expression in the stomach was also higher in STZ diabetic rats than in control rats, but this difference was not significant. In contrast, plasma leptin was markedly reduced in STZ diabetic rats. This reduction in plasma leptin levels was reversed by insulin treatment. In addition, hypothalamic NPY mRNA levels were increased in STZ-treated diabetic rats and were reversed by insulin treatment. Furthermore, the hyperphagia was partially reversed by the administration of a ghrelin-receptor antagonist. Therefore, we conclude that the elevated plasma ghrelin levels, along with decreased plasma leptin levels, could contribute to the diabetic hyperphagia in part by increasing hypothalamic NPY. This is the first report to show the pathophysiological significance of ghrelin in diabetes.     

Daily production and metabolic clearance of growth hormone in juvenile diabetes mellitus

Summary Daily production (PR) of human growth hormone (HGH) was calculated in patients with juvenile diabetes and control subjects by determining metabolic clearance rate (MCR) of¹³¹I HGH, at equilibrium, and mean endogenous HGH levels throughout a 24 h day. Half hourly sampling or a constant withdrawal pump were used to obtain an integrated mean endogenous HGH level. MCR (liters/day) was significantly reduced in all diabetic subjects both in absolute terms (96 ± 15 vs 274 ± 37) and relative to surface area (62 ± 8 vs 171 ± 21) (p < 0.01). Mean HGH levels were 8.4 ng/ml in the diabetics and 5.5 ng/ml in age matched controls. Daily HGH PR in the diabetic subjects (339 to 1365 g/day) did not exceed values in the control subjects (1005–1426 g/day). The results indicate that the elevated plasma HGH levels and increased HGH response to stimuli observed in diabetes, reflect reduced metabolic clearance, rather than increased pituitary secretion.     

Effect of octreotide on insulin requirement, hepatic glucose production, growth hormone, glucagon and c-peptide levels in type 2 diabetic patients with chronic renal failure or normal renal function

Aim of this study was to investigate whether octreotide, a synthetic somatostatin analogue that inhibits growth hormone, insulin and glucagon secretion and improves glycaemic control in insulin dependent diabetic patients was able to exert similar effects in insulin treated type 2 diabetic patients with chronic renal failure who have high plasma glucagon levels. For this purpose saline or octreotide was randomly administered by continuous subcutaneous infusion (100 mcg/daily) in addition to usual insulin treatment for 5 days to six type 2 insulin treated diabetic patients with chronic renal failure and to six type 2 patients with normal renal function, as a control group. At day 3 of insulin plus saline or insulin plus octreotide treatment, total glucose uptake and hepatic glucose production (HGP) were investigated during an euglycemic clamp; at day 5 GH, glucagon and C-peptide plasma levels were evaluated. Octreotide treatment lowered endogenous insulin secretion (evaluated by C-Peptide levels assay), GH and glucagon in all patients, but caused a significant reduction of daily insulin requirement (32 +/- 14 I.U. vs 41 +/- 19 I.U., P<0.02) only in patients with renal failure. HGP was significantly (P<0.05) lowered in patients with renal failure but glucose uptake remained unchanged. The lowering effect of octreotide on insulin requirement in diabetic patients with renal failure in spite of the contemporaneous inhibition on insulin secretion could be explained on the basis of the greater reduction of glucagon levels which are very elevated in these patients as compared to patients with normal renal function. The lowering of glucagon could decrease HGP and, consequently, insulin requirement.     

Glyburide in non-insulin-dependent diabetes. Its therapeutic effect in patients with disease poorly controlled by insulin alone

Glyburide, a second-generation sulfonylurea compound, was combined with insulin to evaluate its therapeutic effectiveness in 14 patients with non-insulin-dependent diabetes mellitus (NIDDM), poorly controlled by insulin alone. Patients were studied before and three months after the addition of glyburide to their insulin program. Fasting plasma glucose concentration fell an average of 57 mg/dL, associated with an approximately 25% reduction in postprandial glucose response. Therapeutic responses varied widely from patient to patient; the greatest improvement in diabetic control was seen in heavier patients, who had retained the ability to secrete insulin in response to meals and who were not excessively insulin resistant. The glyburide-induced fall in plasma glucose concentration was associated with improvements in both insulin secretion and insulin action, but only the enhanced insulin action correlated with the reduction in fasting and postprandial glucose levels. Thus, diabetic control was significantly improved by glyburide. Combined insulin-sulfonylurea therapy may be useful in the treatment of NIDDM that cannot be easily controlled with either agent alone.     

Effects of insulin on plasma renin activity, plasma atrial natriuretic peptide and body fluid volume in diabetes mellitus

The effects of insulin treatment on plasma renin activity (PRA), plasma atrial natriuretic peptide (ANP) and body fluid volume were studied in 16 hospitalized patients with insulin-independent diabetes mellitus. Parameters were recorded for 2 days during treatment by diet alone and for 3 weeks after starting insulin. Blood samples were obtained weekly from 9 patients for the measurement of fasting plasma glucose, hematocrit, PRA and plasma ANP. A 24-hr urine sample was collected to determine the urinary excretion of glucose and sodium. In a separate group of 7 patients, plasma volume and extracellular fluid volume were determined by the Evans blue and sodium thiocyanate dilution tests, respectively. In the group of 9 diabetic patients, significant (p less than 0.05) reductions in fasting plasma glucose, hematocrit and the urinary excretion of sodium and glucose were seen with insulin treatment. PRA fell significantly (p less than 0.05) from 5.2 +/- 1.2 ng/ml/hr (mean +/- SEM) on the control days to 2.3 +/- 0.5 on the 21st day after starting treatment. Plasma levels of ANP averaged 35 +/- 5 pg/ml on the control days and these did not change significantly. In the other group of 7 patients, both plasma volume and extracellular fluid volume increased significantly (p less than 0.05) with insulin treatment. A sodium-retaining effect of insulin and a decrease in osmotic diuresis may have increased the body fluid volume and caused the fall in PRA. Thus, a vasodilatory action of insulin may assist in compensation for the increase in body fluid volume, preventing a rise in plasma ANP levels.     

Influence of exogenous insulin on C-peptide levels in subjects with type 2 diabetes

Aim:The aim of this study was to determine whether the influence of insulin therapy on fasting and stimulated C-peptide levels in type 2 diabetic subjects is due to plasma glucose reduction or a direct effect of exogenous insulin.Methods:Plasma glucose and serum C-peptide levels were determined before and after IV injection of 1 mg glucagon on three separate days in 21 type 2 diabetic subjects. Day 1: without pharmacological treatment and fasting plasma glucose >11.1 mmol/L; day 2: fasting plasma glucose 4.4–7.8 mmol/L, 1 h after withdrawing intravenous regular insulin infusion; day 3: fasting plasma glucose 4.4–7.8 mmol/L with bed-time NPH insulin.Results:Fasting and glucagon stimulated C-peptide levels were higher on day 1 than days 2 and 3. Fasting, but not stimulated C-peptide levels, were lower on day 3 than day 2. These differences were not appeared when the percentage of C-peptide increment or the C-peptide/glucose ratio were compared in the three days.Conclusions:Blood glucose reduction instead of exogenous insulin is responsible for the C-peptide decrease during insulin therapy in type 2 diabetic subjects.     

Growth hormone response to insulin and to arginine in patients with familial hypercholesterolaemia.

Human growth hormone (HGH) response to i.v. insulin (0.1 U/kg body weight) and arginine infusion (25 g of L-arginine for 30 min) was studied in 9 patients (5 males and 4 females) with primary familial hypercholesterolaemia and belonging to 4 families. Mean age was 28 +/- 2 years (range 18-36) and body weight was less than 105% of ideal body weight. Glucose tolerance and insulin response to oral glucose were normal in all patients. HGH release after insulin and after arginine was slightly increased as compared to 21 normal controls, but the differences were not significant. Insulin and glucagon response to arginine in these patients was within the normal range. Plasma glucose and free fatty acids were normal after both insulin and arginine. Moreover, no significant correlation was found between fasting cholesterol and HGH peaks after insulin and after arginine, nor between cholesterol and insulin and glucagon responses. Despite marked hyperlipidaemia, HGH-deficient patients examined by other authors never present signs of atherosclerotic disease. Our data suggest that HGH, in the presence of elevated cholesterol levels, might play an important role in the development of atherosclerotic lesions.     

Continuous subcutaneous octreotide infusion markedly suppresses IGF-I levels whilst only partially suppressing GH secretion in diabetics with retinopathy

The response to GH releasing hormone (GHRH 1-29) and 24-h serum GH and IGF-I levels were measured in 9 insulin-dependent diabetics with retinopathy and 6 normal volunteers before and after different treatment regimens with octreotide, a long-acting somatostatin analogue. Octreotide, 50 micrograms by sc injection, completely suppressed GHRH-stimulated GH release in both groups. Thrice daily sc injections for up to 20 weeks were associated with variable plasma octreotide levels and failed completely to suppress GH secretion in either the patients or the normal controls. Three days of continuous sc pump infusion (500 micrograms/24-h) resulted in consistently high plasma octreotide levels and completely suppressed 24-h GH in 4 normal subjects, whilst treatment for up to 16 weeks only partially suppressed GH levels in 6 patients (AUC mU.l-1.h-1; 209 +/- 81 vs 121 +/- 82; P = 0.01). Mean +/- SD IGF-I levels (micrograms/l) in the patients (but not controls) were suppressed into the hypopituitary range by median 6 weeks (range 2-16) pump administration (203 +/- 62 vs 60 +/- 25; P = 0.02). Pump treatment achieved total GH suppression in normal subjects; diabetics with retinopathy seem more resistant to the GH suppressing effects of the drug. However, the reduction of serum IGF-I with prolonged treatment may be of clinical value in arresting the progress of diabetic retinopathy.