Glucagon regulation of plasma ketone body concentration in human diabetes.

The present study was designed to test the hypothesis that physiological concentrations of glucagon may increase plasma ketone body concentration when sufficient free fatty acid substrate is available to support hepatic ketogenesis. Physiological elevations of plasma glucagon concentration were produced by a constant infusion of hormone, and increased plasma-free fatty acid availability was produced by simultaneous heparin injection to induce intravascular lipolysis. In the five insulin-dependent subjects studied, when plasma glucagon concentration remained at the normal basal level of 72+/-14 pg/ml during control saline infusion, the heparin-induced increase in free fatty acid availability resulted in approximately a 20% increase in plasma ketone body concentration. In contrast, when plasma glucagon concentration was elevated by hormone infusion to the physiological level of 215+/-35 pg/ml, the heparin-induced increases in free fatty acid availability resulted in approximately an 80% increase in plasma ketone body concentration. These results suggest that physiological elevations in plasma glucagon concentration may augment ketonemia in diabetic man when simultaneous elevations in plasma-free fatty acid arepresent.

The controversy concerning counterregulatory hormone secretion. A hypothesis for the prevention of diabetic ketoacidosis?

Diabetic ketoacidosis is characterized by an excess secretion of counterregulatory hormones (glucagon, catecholamines, cortisol, and growth hormone). Experimental evidence obtained in both diabetic man and animals suggests that elevation of the plasma concentration of these hormones is necessary to initiate excess hepatic production of ketone bodies. This increase in hepatic ketogenesis in concert with inability of peripheral tissues to completely utilize ketone bodies results in clinical ketoacidosis. This hypothesis would suggest that pharmacologic control of excess counterregulatory hormone secretion would be a rational therapeutic modality to prevent diabetic ketoacidosis.

Modulation of fatty acid metabolism by glucagon in man. IV. Effects of a physiologic hormone infusion in normal man.

This study was done to explore the role of physiologic elevations of glucagon concentration in plasma ketone body concentration in normal man. During the period of hormone elevation, plasma free fatty acids were pharmacologically elevated to ensure adequate free fatty acid substrate delivery to the liver to support hepatic ketogenesis. Eighty-minute infusions of glucagon resulted in a plasma hormone concentration of approximately 300 pg./ml. During the infusion, ketone bodies declined from their basal concentration and remained below basal for the duration of the infusion. An acute heparin-induced pharmacologic elevation of plasma free fatty acid concentration resulted in a transient rise in plasma ketone body concentration, but at no time did it attain the concentration observed during the control saline infusion. Plasma glucose concentration was not altered by glucagon infusion, but plasma insulin concentration rose by approximately 2.5 muU./ml. These results suggest that glucagon is not ketogenic in normal man as has been previously reported in insulin-deficient diabetics. The glucagon-induced rise in plasma insulin concentration may participate in the observed reduction in plasma ketone body concentration.

The regulation of plasma ketone body concentration by counterregulatory hormones in man.

Previous studies have attributed norepinephrine's ketogenic activity to its ability to mobilize peripheral fat stores. This study was designed to determine whether norepinephrine has ketogenic activity independent of its lipolytic effect in diabetic man. Six insulin-dependent diabetic subjects were infused with pathophysiologic concentrations of norepinephrine (0.08 microgram./kg./min.). As a control for norepinephrine's lipolytic effect, a separate heparin-induced free fatty acid generation study was performed on each subject. The results demonstrate, for the first time in man, that norepinephrine has ketogenic activity independent of its lipolytic effect. Furthermore, physiologic elevations of norepinephrine concentration were also demonstrated to increase plasma glucagon concentration. Our data are consistent with the possibility that the rise in concentration of glucagon may have a participated in the catecholamine-augmented ketogenesis.

Modulation of fatty acid metabolism by glucagon in man. II. Effects in insulin-deficient diabetics.

The potential role of glucagon in the pathogenesis of diabetic ketogenesis was examined by intravenous glucagon administration (1.0 mug/kg. body weight) in five insulin-dificient diabetic subjects. The metabolic response was defined by examining the changes in the plasma concentration of insulin, glucose, free fatty acids (FFA), triglycerides, betahydroxybutyrate, and acetoacetate in comparison to those changes occuring in five nondiabetic control subjects tested under identical experimental conditions. Our results demonstrate that in the absence of endogenous insulin secretion, exogenous glucagon administration results in an enhanced ketonemia indipendent of the rise of FFA substrate. This augmented ketogenic response to glucagon in diabetic subjects was not reflected in the dynamics of glucose regulation, with similar changes in blood glucose concentration occurring in both diabetic and nondiabetic populations. These data demonstrate that glucagon may potentiate ketonemia in insulin-deficient diabetics, although the mechanism mediating this augmentation is not defined. Since the rise in FFA substrate was indistinguishable in the diabetic as compared to the control group, increased hepatic conversion of FFA substrate into ketone bodies is suggested.

Modulation of fatty acid metabolism by glucagon in man. I. Effects in normal subjects.

This investigation was designed to examine the contribution of glucagon to the regulation of free fatty acid (FFA) metabolism in man. The acute effects of exogenous glucagon upon the concentration of plasma FFA and its metabolites: acetoacetic acid (AcAc), beta hydroxybutyric acid (BOH), and triglyceride (TG) were assessed in five normal male subjects. The threshold of response was determined by administering intravenous glucagon in a graded dose-response fashion, andassessing the magnitude of change in FFA and its metabolites. A prompt rise in FFA concentration occurred after glucagon injection, and was associated with a synchronous elevation in both AcAc and BOH concentration in plasma. The rise in FFA concentration was maximal with the glucagon dosage of 0.5 mug./kg. with no greater FFA response seen athigher hormone dosage. In contrast, the rise in plasma ketones was linearly related to thedosage of glucagon but not to the corresponding rise in FFA concentration. The behavior of plasma triglyceride concentration following glucagon challenge also contrasted with the linear rise in ketone concentration. At the lowest dosage of glucagon, TG concentration rose transiently, while at the highest two dosages the concentration of TG was reduced. The transitional dosage of hormone at which no change in TG concentration was observed was approximately 0.5 mug./kg. Comparing our in vivo data with the conclusions reached from published liver perfusion studies, it can be suggested that in the dosages employed, glucagon acutely elevates plasma FFA concentration and modulates their metabolism by augmenting conversion to AcAc and BOH while tending to reduce conversion to TG in man.

Modulation of fatty acid metabolism by glucagon in man. III. Role of pharmacologic limitation of FFA availability.

This study was undertaken to examine the contribution of plasma free fatty acid availability upon the regulation of ketone body and endogenous triglyceride concentration in man. This substrate-product relationship was examined both in the basal state and during hormonally induced ketogenic stimulation. Five insulin-deficient diabetics receiving a fixed dose of exogenous insulin were studied after a twelve-hour fast and twenty-four hours after their last therapeutic insulin injection. Reduction in basal free fatty acid concentration was induced with two weeks of clofibrate administration, and hormonal ketogenic stimulation was induced with glucagon administration. A highly significant correlation was observed between the basal free fatty acid level and the basal ketone body concentration. This substrate-product relationship persisted throughout hormonally induced ketogenic stimulation, suggesting that the basal free fatty acid concentration is a major determinant of the plasma ketone body concentration in man. In contrast, alterations in basal free fatty acid concentration were not accompanied by consistent changes in plasma triglyceride concentration, suggesting that plasma free fatty acid concentration may not be the principal determinant of the endogenous triglyceride concentration in clofibrate-treated diabetic man.

Bactericidal properties of commercial U.S.P. formulated insulin.

This study examined the suitability of commercial U.S.P. insulin as a bacterial growth medium. Purified port insulin (Iletin II, Eli Lilly & Co., Indianapolis, Indiana) was inoculated with approximately 500,000 bacteria per ml of: (1) Staphylococcus epidermidis, (2) Staphylococcus aureus, or (3) Escherichia coli. All three bacterial cultures were sterilized by the insulin within 24 h at 37 degrees C. However, Staph epidermidis was the most sensitive and Escherichia coli was least sensitive to the bactericidal effects of commercial U.S.P. insulin. Components of commercial U.S.P. insulin were then examined for their bactericidal activity against Escherichia coli. Phenol (230 mg/dl), glycerol (1.6 d/dl), and zinc cations (4.0, 2.0, and 1.0 mg/dl) demonstrated bactericidal activity, whereas dialyzed insulin demonstrated minimal effects. We conclude that insulin contaminated with 5 X 10(5) bacteria commonly found on the skin will self-sterilize within 24 h. This effect is secondary to the additives placed in the insulin and not to the insulin itself.

Insulin suppresses its own secretion in vivo.

This study addressed the controversial question of whether a negative-insulin-feedback loop exists in vivo. We utilized prehepatic insulin production, calculated by computerized deconvolution analysis of peripheral C-peptide concentration, as a measure of endogenous insulin secretion. Prehepatic insulin production was determined in 10 normal men who randomly underwent a control study and two additional studies involving different insulin infusion rates that achieved circulating insulin concentrations within the physiologic range during euglycemic clamps. The results demonstrate a dose-dependent suppression of prehepatic insulin production from 5.8 +/- 1.4 mU/min during the control study to 4.0 +/- 1.2 and 3.2 +/- 0.9 mU/min during plasma insulin levels of 34 +/- 4 and 61 +/- 6 microU/ml, respectively (P less than .05). Therefore, in contrast to recently reported results in vitro, insulin inhibits its own secretion in humans.

Subcutaneous peritoneal access device for type I diabetic patients nonresponsive to subcutaneous insulin.

Three type I diabetic patients nonresponsive to subcutaneous insulin were implanted with a subcutaneous peritoneal access device. In these patients, multiple subcutaneous injections had been unable to prevent recurrent hospital admissions for diabetic ketoacidosis. The patients were responsive to intravenous insulin but had limited accessible peripheral veins. Complications of thrombosis and/or septicemia from permanent central venous catheters prevented the long-term use of this route. The peritoneal access device was implanted subcutaneously adjacent to the umbilicus with its insulin delivery catheter terminating in the peritoneal space. Transcutaneous injection of insulin into the subcutaneous access port resulted in the same quantity of insulin entering the peritoneal space. Using a mixture of regular and protamine zinc insulin in a ration of 1:1 resulted in acute increases in plasma free insulin concentration with meals and a declining background level postprandially. All peritoneal access devices have been functioning well for at least 2 mo and in one of the implanted diabetic subjects, it has been in continuous use for 5 mo with no evidence of peritonitis or resistance to peritoneal insulin. These results suggest that a subcutaneous peritoneal access device may provide an alterative insulin delivery route for patients who are nonresponsive to subcutaneous insulin injections.

Lispro insulin: benefits and limitations.

Lispro insulin is the first synthetic insulin analog commercially available for the treatment of diabetes. Its rapid absorption and short half-life results in both advantages and disadvantages to the patient with diabetes. Understanding the physiology of normal insulin secretion and the pathophysiology of diabetes is paramount to optimizing Lispro insulin in patients with diabetes.

Pathogenesis of diabetic ketoacidosis: a reappraisal.

This study reviews the pathogenic hormonal abnormalities (insulin deficiency and stress hormone excess) in diabetic ketoacidosis. The data both supporting and negating a primary role for insulin deficiency in the pathogenesis of diabetic ketoacidosis are examined. Evidence implicating excess stress hormone secretion as a necessary event in the development of severe metabolic decompensation is discussed. The data suggest that diabetic ketoacidosis may be prevented by correcting either the relative deficiency of insulin or the excess secreation of one or a combination of the stress hormones. Studies supporting a primary role for insulin deficiency in the pathogenesis of diabetic ketoacidosis include the beneficial therapeutic response to insulin administration in ketoacidosis, development of ketoacidosis; and (3) stress hormone excess is necessary for fulminant ketoacidosis to be manifested.s following insulin withdrawal from diabetic man and animals, and hypoglycemic and hypoketonemic effects of insulin. Studies negating a primary role for insulin deficiency in ketoacidosis include the "normal" plasma insulin concentration in the majority of ketoacidotic cases, delayed onset of ketoacidosis after insulin withdrawal from diabetic man, and lack of hypolipolytic and hypoketonemic effect of insulin without prior stress hormone adipocyte and hepatocyte stimulation. Evidence that stress hormones (glucagon, catecholamines, cortisol, and growth hormone) contribute to the metabolic decompensation of ketoacidosis includes: (1) in all cases of ketoacidosis, at least one stress hormone is always elevated; (2) pharmacologic blockade of each of the stress hormones reduces the rate and/or frequency of metabolic decompensation in diabetic man; (3) removal of the pituitary and/or the adrenal gland in diabetic animals completely prevents the development of ketoacidosis after insulin withdrawal; and (4) administration of each of the four stress hormones under appropriate conditions induces metabolic decompensation in diabetic man with "normal" circulating levels of plasma insulin concentration. From these studies, the following conclusions are supported: (1) absolute insulin deficiency is an unusual cause of ketoacidosis; (2) the presence of relative insulin deficiency is necessary for the development of ketoacidosis; and (3) stress hormone excess is necessary for fulminant ketoacidosis to be manifested.

Normalization of plasma insulin profiles in diabetic subjects with programmed insulin delivery.

Normalization of plasma glucose concentration with subcutaneous injections of insulin has been difficult. With intravenous insulin delivery, normalization of plasma glucose concentration in adult-onset diabetic patients has been achieved when their plasma insulin concentration was normalized. In juvenile-onset diabetic subjects totally lacking endogenous insulin, the effect of normalization of the plasma insulin concentration has not yet been reported. In the present studies, the plasma insulin profile was normalized in four brittle, insulin-dependent diabetic subjects throughout three meals (breakfast, lunch, and supper). Plasma free insulin concentration was assayed after precipitation of endogenous insulin antibodies with polyethylene glycol. Acute meal-related increases in plasma free insulin concentration were achieved with a programmable intravenous insulin delivery system. When plasma insulin profiles were normalized in insulin-dependent diabetic subjects lacking endogenous insulin secretion, an improvement in meal-related glucose excursions was observed. However, complete normalization of plasma glucose concentration was not achieved, suggesting that factors other than plasma insulin concentration modulate carbohydrate homeostasis in brittle diabetes. These factors may include portal vein hypoinsulinemia during peripheral vein insulin infusion, stress hormone concentrations, and tissue insulin resistance.

Meal composition is a determinant of lispro-induced hypoglycemia in IDDM.

OBJECTIVE: Lispro is a newly FDA-approved analog of human insulin that will be widely used in patients with IDDM. This insulin, however, may have an increased potential for hypoglycemia because of its very rapid subcutaneous absorption, especially in a setting of decreased carbohydrate intake. Using a short-term prospective randomized parallel group-study design, we studied the incidence of hypoglycemia when lispro was given before breakfast compared with regular human insulin. Since carbohydrate intake is a determinant of postprandial glycemia, we administered three isocaloric meals characterized by low, average, and high carbohydrate content. RESEARCH DESIGN AND METHODS: Two groups of six IDDM subjects were randomized to receive 0.15 U/kg of lispro or regular human insulin subcutaneously before the ingestion of three 500-kcal breakfast meals of differing carbohydrate content on separate days. Lispro was administered at mealtime, and regular insulin was administered 30 min before mealtime. RESULTS: Postprandial plasma glucose concentrations were decreased in the lispro group compared with the regular-insulin group for all three meal types (P < 0.05), and hypoglycemia developed more frequently and rapidly in the lispro group, compared with the regular-insulin group by survival analysis. Additionally, peak insulin concentrations were higher (P < 0.001) and peaked more rapidly (P < 0.05) in the lispro group, compared with the regular-insulin group. CONCLUSIONS: We conclude that lispro has a tendency for early postprandial hypoglycemia compared with regular insulin in the setting of reduced carbohydrate intake. This fact should be told to patients who decide to switch from regular insulin to lispro. Health care professionals should instruct their IDDM patients to monitor glucose levels frequently after switching to lispro since adjustments in their carbohydrate intake and/or their lispro dosage may be necessary to avoid hypoglycemia.

Diabetic glucose control: matching plasma insulin concentration to dietary and stress hyperglycemia.

Optimal management of the diabetic patient includes normalization of plasma glucose concentration. Attainment of this goal is difficult because both food and stress result in acute elevations of blood glucose that cannot be matched with a single subcutaneous injection of NPH insulin. This paper examines the currently available methods for delivery of insulin to the diabetic subject and the degree of metabolic control attained. It suggests that optimal diabetic control will be achieved only when newer methods of insulin delivery are available to the clinician that match plasma insulin requirements to the simultaneous plasma glucose concentration.

A clinical algorithm to determine the etiology of brittle diabetes.

Brittle diabetes is a subset of insulin-dependent diabetes mellitus for which multiple causes have been suggested. In its most severe form, brittle diabetes is incapacitating, preventing gainful employment and a normal lifestyle. Although some brittle diabetic individuals will significantly improve by intensive insulin therapy and education, many others remain unable to function normally because of recurrent episodes of hyperglycemia and hypoglycemia. We studied 30 incapacitated brittle diabetic subjects and developed an efficient algorithmic approach to determine the etiology of brittleness. Central to our diagnostic algorithm was the glucose response to 0.1 U/kg insulin administered subcutaneously and intravenously. If this response was normal, then psychosocial evaluations were completed, including psycholinguistic and health psychological testing. Other parameters affecting blood glucose concentration were also assessed, such as gastric motility, counterregulatory hormones, and, most important, patient compliance with prescribed regimens. However, if an "abnormal" glucose response to the insulin challenge tests was observed, the location of the insulin resistance was identified as being subcutaneous, intravascular, or at the peripheral tissue. Using our diagnostic algorithm, the identification of the etiology of brittleness in 29 of the 30 referred patients was possible. Thus, the purpose of an algorithmic approach to diagnosis is not only to avoid unnecessary testing, but also to determine the correct etiology of the brittle diabetes to determine appropriate therapy.

The etiology of incapacitating, brittle diabetes.

Incapacitated brittle diabetic subjects are a small subset of insulin-dependent diabetic individuals who are unable to maintain a normal lifestyle because of frequent disruptions secondary to severe hyperglycemic and/or hypoglycemic episodes. Thirty incapacitated patients were referred for evaluation because the cause of their diabetic instability could not be determined by their personal physicians despite extensive patient training in correct diabetes management, frequent hospitalizations for observation, and multiple diagnostic testing. From the 30 patients, a diagnostic algorithm was developed (described in the companion article) from which the etiology of brittle diabetes could be established in 29. This article provides the clinical characteristics of each of the 30 patients, a description of the etiologic categories of brittle diabetes, and the clinical follow-up from the time that the etiologic diagnosis was established and treatment recommended. Although extensive medical records were sent with each patient, without prospective objective testing under rigidly controlled conditions, the correct etiologic diagnosis would not have been evident from the clinical presentation of the patient. Of equal importance in identifying the etiology of brittle diabetes was the acceptance and cooperation of the referring physician in providing close follow-up and repeat insulin challenge testing when necessary. In this referred patient population, eight subjects had factitious disease, eight were malingering, seven had communication deficits, two had gastroparesis, two had systemic insulin resistance, two had miscellaneous causes of brittle diabetes, and one patient remained undiagnosed.(ABSTRACT TRUNCATED AT 250 WORDS)

Comparative efficacy of a once-daily controlled-release formulation of glipizide and immediate-release glipizide in patients with NIDDM.

OBJECTIVE: To compare the efficacy and safety of controlled-release glipizide (glipizide-GITS [gastrointestinal therapeutic system]) and immediate-release glipizide in patients with non-insulin-dependent diabetes mellitus (NIDDM). RESEARCH DESIGN AND METHODS: In a multicenter, open-label, randomized, two-way crossover study, 132 patients with NIDDM received daily doses of 5, 20, or 40 mg of either glipizide-GITS or immediate-release glipizide for 8 weeks followed by 8 weeks of the alternate formulation. Plasma glucose, serum insulin, C-peptide, and plasma glipizide levels were measured at fasting and post-Sustacal challenge at the end of 1 and 8 weeks of each treatment phase. HbA1c was measured at the end of weeks 7 and 8 of each treatment phase. RESULTS: Both formulations of glipizide yielded similar mean HbA1c values. However, mean fasting plasma glucose (FPG) levels were significantly lower with glipizide-GITS treatment than with immediate-release glipizide at the end of week 1 (11.0 vs. 11.6 mmol/l; P < 0.01) and at the end of the 8-week treatment phase (10.9 vs. 11.7 mmol/l; P < 0.001). Fasting insulin and C-peptide levels were lower after 5 mg glipizide-GITS vs. immediate-release glipizide. Glucose responses to Sustacal were similar after both formulations of glipizide; however, serum insulin (P < 0.01) and C-peptide responses (P < 0.05) were lower with glipizide-GITS than with immediate-release glipizide treatment at the end of the 8-week treatment phase. Mean plasma glipizide concentrations were stable by the end of week 1, and the concentrations increased proportionately with dose. Once-daily Glipizide-GITS provided effective mean glipizide concentrations (> 50 ng/ml) 24 h after dosing, even at the lowest (5 mg) dose level. Both formulations were well tolerated. CONCLUSIONS: Glipizide-GITS was significantly more effective than immediate-release glipizide in reducing FPG levels. Both formulations reduced postprandial plasma glucose levels equally; however, glipizide-GITS exerted its control in the presence of lower plasma glipizide concentrations in addition to significantly lower insulin and C-peptide levels. This suggests that glipizide-GITS improves insulin sensitivity.

Optimal administration of lispro insulin in hyperglycemic type 1 diabetes.

OBJECTIVE: Lispro is a new rapidly absorbed insulin analog. At present, there are no recommendations for the optimal injection time of lispro insulin in hyperglycemic patients. In contrast to normoglycemic patients with diabetes, we hypothesized that injection of lispro insulin 15-30 min before meal ingestion would improve postprandial glucose excursion in hyperglycemic diabetic subjects. RESEARCH DESIGN AND METHODS: In 48 randomized overnight studies, 12 healthy adult type 1 diabetic patients received lispro insulin 0.15 U/kg admixed with human ultralente 0.2 U/kg (as background insulin) subcutaneously at minutes (-30, -15, 0, and +15) relative to the ingestion of an American Diabetes Association breakfast of 8.6 kcal/kg. Pre-breakfast hyperglycemia of 10.2 +/- 0.2 mmol/l was established before the study by continuous overnight infusion of intravenous insulin, which was stopped 30 min before lispro insulin injection. Glucose and insulin levels were measured every 30 min for 5 h after breakfast. RESULTS: Results demonstrated that postprandial glucose excursion was reduced when lispro insulin was administered 15 or 30 min before the meal compared with lispro insulin injected at the meal (P < 0.002). The postprandial glucose excursion (millimoles per liter per hour) was -6.4 +/- 3 for the -30-min group, -5.1 +/- 2.9 for the -15-min group, 3.4 +/- 4.1 for the 0-min group, and 5.7 +/- 4.4 for the +15-min group. Although injecting lispro insulin at 30 min before the meal resulted in a significant reduction in postprandial glycemia, it was accompanied by loss of glucose control at 4 h postmeal in two subjects. CONCLUSIONS: Optimization of lispro insulin in hyperglycemic patients requires timing of the insulin injection at least 15 min before the meal.

Low-dose ethanol predisposes elderly fasted patients with type 2 diabetes to sulfonylurea-induced low blood glucose.

OBJECTIVE: It has previously been demonstrated that the risk of hypoglycemia is low among otherwise healthy elderly fasted patients with type 2 diabetes taking oral sulfonylurea medications. Nevertheless, these agents do cause hypoglycemia in clinical practice, suggesting that accompanying factors must typically be present for hypoglycemia to occur. Ethanol is one putative risk factor that has not been evaluated as a mechanism for low blood glucose among sulfonylurea users. We hypothesized that low concentrations of ethanol would reduce blood glucose concentrations in elderly type 2 diabetic patients receiving sulfonylureas during a short-term fast. RESEARCH DESIGN AND METHODS: A total of 10 type 2 diabetic patients, aged 68 +/- 3 years and receiving 20 mg glyburide daily, participated in a prospective double-blind placebo-controlled in-patient study consisting of two 24-h fasts at least 1 week apart. During hours 14 and 15 of the fasting studies, subjects received intravenous infusions of either 4.35 mmol.kg-1.h-1 ethanol (equivalent to one or two alcoholic beverages) or saline placebo in random order. Ethanol, plasma glucose, insulin, and counterregulatory hormones were assessed very 30-60 min during the final 10 h of the fast. RESULTS: Blood ethanol levels peaked at 17 +/- 2 mmol/l (the lower legal limit of intoxication in New Mexico) during the ethanol study. Plasma glucose concentrations did not differ at baseline (placebo 8.5 +/- 1.8 vs. ethanol 8.7 +/- 1.7 mmol/l; P = 0.50), but nadir plasma glucose was lower after the ethanol infusion compared with placebo (4.4 +/- 1.2 vs. 5.0 +/- 1.4 mmol/l; P = 0.01), and the absolute decline in plasma glucose was also greater during the ethanol study than the placebo study (4.7 +/- 0.9 vs. 3.6 +/- 1.2 mmol/l; P = 0.01). Counterregulatory hormone levels were increased during the ethanol study and nonesterified fatty acid concentrations were suppressed compared with the placebo study. CONCLUSIONS: Low doses of ethanol predispose fasted elderly type 2 diabetic patients to low blood glucose during a short-term fast. This may be one of several mechanisms by which sulfonylurea-induced hypoglycemia occurs in elderly patients.