Insulin bolus given by sprinkler needle: effect on absorption and glycaemic response to a meal.

Two studies were conducted investigating the effect of injecting short acting insulin subcutaneously by means of a sprinkler needle; this needle has 14 small holes in the wall but is sealed at the tip. In the first study absorption of 8 U iodine-125 labelled Actrapid HM injected subcutaneously at two separate sites in the abdominal wall was measured in 10 patients. One injection was given with the sprinkler needle and the other with a conventional needle. The initial dose absorbed during the first 30 minutes was significantly higher with the sprinkler needle. In the second study 10 U Actrapid HM was given to 11 other patients (all negative for C peptide and with low insulin binding antibody titres) on two separate days immediately before a standardised breakfast either by the sprinkler needle or by a conventional needle (random order). Plasma free insulin increased more rapidly and to higher concentrations with the sprinkler needle and the glycaemic response was considerably diminished. The sprinkler needle improves both the insulin absorption rate and the glycaemic response to a meal and may reduce the 30 minutes or so before meals that diabetics must inject to minimise postprandial hyperglycaemia.     

Twenty-four-hour metabolic profiles in diabetic children receiving insulin injections once or twice daily

Twenty-four-hour metabolic profiles were performed twice in each of 15 diabetic children, once when they were receiving single daily injections of insulin (Monotard plus Actrapid) and once on a twice-daily regimen (Semitard plus Actrapid). Before the study control was optimised at home on each regimen. There were no differences in overall 24-hour diabetic control on the two regimens as measured by mean blood glucose concentration, area under the blood glucose curve, M value, and 24-hour urinary glucose excretion. Hyperglycaemia after breakfast occurred on both regimens. Significant differences were noted before breakfast, when blood glucose and ketone concentrations were lower and plasma free insulin higher on the single-injection regimen, and after supper and during the night, when blood glucose values were lower on the two-injection regimen and associated with a rise in plasma free insulin after the evening injection. Once-daily injections provided insufficient circulating insulin after the evening meal, while twice-daily injections did not last through the night. Plasma C peptide, indicating residual endogenous insulin secretion, was just detectable in two children but easily detectable in four children, whose 24-hour diabetic control was significantly better than that in the remaining 11 children.Conclusions about the superiority of one insulin regimen over another must be based on specific differences in diabetic control. Both regimens studied achieved adequate control, and though neither provided physiological control specific modifications to the regimens could help to produce more normal profiles.     

Use of U-500 insulin in the treatment of severe insulin resistance

Background: Glycemic control is essential in the management of diabetes. However, many patients with diabetes are not achieving therapeutic targets, partly because they are receiving insufficient doses of insulin. This is particularly problematic in patients with severe insulin resistance, defined as insulin requirement >200 units/kg per day (>3 units/kg per day for pediatric patients). It is difficult to use U-100 forms of insulin at doses >200 units/kg per day because of the volume of insulin being administered subcutaneously. U-500, a concentrated form of insulin, may be useful in the treatment of these patients.Objective: Current practice regarding the use of U-500 insulin has been published elsewhere. This article presents an updated algorithm for the administration and dosing of U-500 insulin, based on clinical experience with severe forms of insulin resistance. Guidelines are provided for dose escalation of U-500 insulin.Methods: We reviewed the results of treatment with U-500 insulin in patients with severe insulin resistance. We analyzed the results, updated a pre-existing algorithm, provided additional practical information on the administration and dosing of U-500 insulin, and compared the cost of U-500 with that of U-100 insulin.Results: To date, we have treated 56 patients (age range, 9–54 years) with severe insulin resistance using U-500 insulin. Doses ranged from 1.5 to 566 units/kg per day. Based on the pharmacodynamic properties of U-500 insulin, this concentrated form must be administered and dosed differently than regular U-100 insulin. U-500 insulin cost more than U-100 insulin on a per-milliliter basis, but cost less in the end because of the lower volumes of insulin required and fewer syringes and pump cartridges needed to administer U-500 insulin.Conclusions: In our experience, U-500 insulin is a useful tool in the management of patients with severe insulin resistance. U-500 insulin alleviates the volume-related problems associated with U-100 insulin, making treatment with higher doses of insulin (≥200 units per day) more effective with U-500 insulin than with U-100 insulin.     

The effect of intrahippocampal insulin microinjection on spatial learning and memory

Insulin is best known for its action on peripheral target tissues such as the adipocyte, muscle and liver to regulate glucose homeostasis. Insulin and its receptor are found in specific area of CNS with a variety of region-specific functions different from its direct glucose regulation in the periphery. The hippocampus and cerebral cortex distributed insulin/insulin receptor has been shown to be involved in brain cognitive functions. Previous studies about the effect of insulin on memory are controversial. In the present study, the effect of insulin microinjection into CA1 region of rat hippocampus on water maze performance has been investigated. Insulin had a discrepant effect dose dependently. The spatial learning and memory were impaired with lower dose of insulin, had not changed with intermediate doses, while they improved with higher doses. These results suggest that insulin may have a dose-dependent effect on spatial learning and memory.     

Characterization of coho salmon (Oncorhynchus kisutch) insulin

Insulin has been isolated from islet tissue of coho salmon (Oncorhynchus kisutch) by gel filtration and HPLC and the complete amino acid sequence has been determined. The sequence differs from bovine insulin at 14 sites but all interchanges are conservative from the viewpoint of preservation of conformation. A comparison of insulin sequences from other fish is presented. Salmon insulin cross-reacts very weakly with antiserum to bovine insulin and vice versa. A completely homologous radioimmunoassay has been developed and used to estimate the insulin in salmon islet tissue and in plasma. The hypoglycemic effect of salmon insulin in salmon was more pronounced and persisted longer than that caused by identical doses of bovine insulin.     

When a unit of insulin is not a unit: Detemir dosing and insulin cost in type 2 diabetes mellitus

Background: Increasing acceptance of basal-bolus insulin therapy for the control of diabetes mellitus (DM) has led to newer formulations of basal insulin analogues. The newest one is detemir.Objectives: Clinical evidence suggests that patients with type 2 DM require higher doses of detemir than other basal insulins to achieve equivalent glycemic control. This study examines evidence for greater dosing requirements and the implications of higher doses on the cost of insulin treatment.Methods: We performed a MEDLINE search for randomized, prospective studies comparing detemir with other basal insulins in patients with type 2 DM that were published in English between January 2000 and November 2008. The mean daily doses of basal and bolus insulin and the mean total daily insulin doses were determined. Overall weighted mean doses of the insulins were used to estimate the mean total daily insulin doses required for a 100-kg patient, and published 2008 US retail prices were used to estimate the retail costs of basal-bolus and basal-only insulin regimens.Results: Seven trials involving 3311 patients were identified in the literature search. The mean total daily insulin dose was 0.80 unit/kg for detemir-based regimens and 0.58 unit/kg for comparison regimens. For basal-bolus regimens, the estimated retail cost of the mean total daily insulin dose was $11.24 for detemir-based regimens compared with $8.99 for glargine-based regimens and $6.41 for neutral protamine Hagedorn (NPH)-based insulins. For basal-only regimens, the estimated retail cost of the mean total daily insulin dose was $8.23 for detemir compared with $5.19 for glargine and $2.35 for NPH.Conclusions: It is important for health care providers and patients to know that patients with type 2 DM may require substantially higher doses of detemir than other basal insulins. This should be considered when titrating the dose as well as in cost-benefit analyses of detemir versus other insulins.     

Chronic Intracerebroventricular Infusion of Insulin Failed to Alter Brain Insulin-Binding Sites, Food Intake, and Body Weight

The present study was performed to explore the role of exogenous insulin in CSF in the control of energy balance in the rat. For this purpose, adult male Sprague-Dawley rats carrying an indwelling cannula in the right lateral cerebral ventricle were infused for a maximum of 10 days with insulin (Actrapid) at various rates (starting at 0, 45, 85, 170, and 600 ng/day) or anti-insulin antibody (IgG fraction; diluted 1:10 wt/vol) with an osmotic minipump. All those treatments did not modify the growing rates; neither total daily food intake nor the circadian rhythm of food intake was further modified. The chronic insulin infusion starting at 600 ng/day resulted in a chronic significant increase in CSF insulin levels without changing the plasma insulin level. It failed to alter specific insulin binding sites to Triton X-100 solubilized microsomal membranes from various brain areas (cerebral cortex, olfactory bulbs, and lateral and medial hypothalami) at the end of the 5- or 10-day period of insulin infusion. Purification of insulin receptors on a wheat germ agglutinin did not reveal any further effect of insulin. From these results, it seems unlikely that the input to the brain insulin-effector systems could arise from CSF insulin.     

Insulinothérapie : insuline ou analogues ? Injection ou perfusion ? Boucle ouverte ou boucle fermée ?

Insulin has been purified, humanized and then synthesized by microorganisms. It is mandatory to be able to use insulin, whose kinetics and reproducibility allow glycemia near to normal without increasing hypoglycemia. Use of insulin analogs allows a slight improvement in glycemic control and decrease hypoglycemia frequency. Flexibility of treatment is also improved. “Continuous subcutaneous insulin infusion” (CSII) using rapid analogs mimics physiologic insulin secretion. Major indications are: high HbA1c despite well-managed basal-bolus regimen, severe hypoglycemia, brittle diabetes or “dawn phenomenon”. Children, adolescents as well as pregnancy are also good indications. “Continuous intraperitoneal insulin infusion” major interest is the predominant absorption via the portal system. Kinetic is comparable to rapid analogs delivered subcutaneously. The dramatic reduction of severe hypoglycemic events has been related to good reproducibility of insulin absorption and restoration of glucagon response. Hypoglycemia prone type 1 diabetic patients, uncontrolled with well-managed CSII as well as subcutaneous insulin resistance are the major indications. The association of optimized insulin therapy to “real time continuous glucose monitoring” allows better doses adaptation. Alarms can be set to avoid glycemic excursions and thus severe hypoglycemia. Using these devices, HbA1c is significantly improved without any increase in hypoglycemic events. These devices are one of the steps towards the “closed-loop insulin delivery” concept. Restoration of missing beta-cell function by an automated, glucose-modulated, insulin-delivery system would allow near normal glycemia without the risk of hypoglycemia. First studies show a good regulation of interprandial glycemia; prandial doses seem more difficult to assess. Nevertheless the “holy grail” might be closer than we think.     

Role of 17beta-estradiol administration on insulin sensitivity in the rat: implications for the insulin receptor

The role of 17beta-estradiol in the early steps of insulin action is only partially known, although its effect on glucose homeostasis has been reported. In this paper, we attempt to prove the influence of 17beta-estradiol on the insulin receptor of ovariectomized rats treated with different hormonal doses. Our results show that high doses of estradiol impair insulin sensitivity while low doses improve it. We think that these results are the consequence of changes at a molecular level, because high doses of estradiol produced lower expression of the insulin receptor gene, lower content of this receptor in target tissues, and lower phosphorylation of insulin receptor in these tissues. However, low doses of estradiol seem to produce just the opposite. The possible existence of consensus response elements in the insulin receptor gene promoter to estradiol could be controlling the expression of this gene, this control being dose and timing dependent. Moreover, we cannot discard a possible effect of estradiol on the activity of protein tyrosine phosphatases, and therefore, on the activity of the insulin receptor. These new findings improve knowledge about the possible risk for insulin resistance in women taking oral contraceptives or receiving hormonal replacement therapy around the menopause, but could also open the door towards the possible utilization of 17beta-estradiol in some diabetes cases.     

Effect of treatment with different doses of 17-beta-estradiol on the insulin receptor

The mechanism for the development of insulin resistance in normal pregnancy is complex and is associated with serum levels of sex hormones. However, the influence of these hormones on the early steps of insulin action has not been extensively studied, although the potentially beneficial effect of estradiol on glucose homeostasis has been reported. In this paper, we attempted to determine the effect of 17-beta-estradiol on the insulin receptor of ovariectomized rats treated with different doses of hormones. Our results showed a tissue-dependent response to estradiol. We found that low doses of estradiol increased the amount of insulin receptors in liver and muscle on days 6 and 11 of treatment but not in adipose tissue, and after 16 days only the muscle responsed in this way. On the other hand, high doses of estradiol significantly decreased the amount of insulin receptors, at least in muscle and adipose tissue. We believe that the low concentrations of 17-beta-estradiol (similar to early pregnancy) could be responsible for the increase in insulin sensitivity by increasing the amount of insulin receptors in peripheral tissues. When the hormone levels were high (similar to late pregnancy) the amount of insulin receptors decreased in peripheral tissues, and insulin sensitivity is diminished just as in late pregnancy. The specific molecular mechanism for this action is as yet unknown.     

Potential of stem cell-derived exosomes to regenerate β islets through Pdx-1 dependent mechanism in a rat model of type 1 diabetes

Type 1 diabetes, has been recognized as an autoimmune disease. Like other immunological conditions, regulation of immune response is a key strategy to control the autoimmunity in diabetic patients. Mesenchymal stem cells have been shown to have a distinct potential in modulating the immune reactions. However, treatment with stem cells is combined with concerns about safety issues. To overcome these concerns, in this study, we have utilized the regenerative potential of exosomes isolated from menstrual blood-derived mesenchymal stem cells to restore the β-cell mass and insulin production in type 1 diabetes. Exosomes are nanovesicles containing various cargos involved in cellular communications. Streptozotocin was used to induce islet destruction and diabetes in male Wistar rats. Then, exosomes were intravenously injected into animals at different time points and in a single or repeated therapeutic doses. After about 6 weeks, animals were euthanized and the pancreas was analyzed for the presence of the regenerated β islets as well as the insulin secretion. The non-fasting blood glucose and the serum insulin level were also monitored during the study. Our results represented that menstrual blood-derived mesenchymal stem cell-derived exosomes enhance the β-cell mass and insulin production in the pancreas of diabetic animals that received repeated doses of exosomes. Immunohistochemistry analysis also confirmed the presence of insulin in the islets of treated animals. Further investigations proposed that exosomes induce the islet regeneration through pancreatic and duodenal homeobox 1 pathway. The exosome tracking also revealed the homing of injected exosomes to the pancreas.     

Treatment of Severe Diabetes Mellitus by Insulin Infusion

A new and simple form of insulin therapy for diabetic hyperglycaemia and ketoacidosis has been developed using a continuous intravenous infusion of insulin at a rate of 2·4 U/hr to maintain serum insulin concentration at physiological levels. This rate raises the mean serum insulin to 83 μU/ml and has a therapeutic effect which is not augmented by higher infusion rates. The response to such low doses of insulin indicates a need for a reappraisal of currently held theories about insulin resistance in diabetic ketoacidosis. In 11 diabetic patients with a mean plasma glucose of 514 mg/100 ml this therapy produced continuous falls in plasma glucose at a mean rate of 75 mg/100 ml/hr, and 10 out of 11 patients recovered within eight hours. This form of therapy is simple to institute, not complicated by hypoglycaemia, and avoids the confusion and empiricism of previously described forms of therapy.     

Right thyroid hemiagenesis with adenoma and hyperplasia of parathyroid glands -case report

The uncertainties regarding dose similarities between basal long-acting insulin analogues remain. Recent real-world studies indicate dose similarities between insulin detemir and insulin glargine, but further studies are still warranted. The aim of this study was to compare real-life daily doses of insulin detemir and insulin glargine in type 2 diabetes patients when administered once daily. We analysed 536 patient cases from general practice (63%) and endocrinological outpatient clinics (37%). A self-administered questionnaire completed by the treating physician was used to obtain data on patient characteristics (gender, age, weight, height, latest HbA1c-value), daily doses, administration of and number of years treated with insulin detemir and insulin glargine, concomitant insulin use and use of non-insulin anti-diabetic medication. Both bivariate analyses and multivariate regression analyses were applied to examine whether there were differences in the daily doses of insulin detemir and insulin glargine. There was no significant difference in the mean daily doses of insulin detemir (0.414 U/kg) and insulin glargine (0.416 U/kg) (p = 0.4341). In multivariate regression analyses, age and BMI had a significant influence on daily insulin dose with the dose increasing 0.003 U/kg (p = 0.0375) and 0.008 U/kg (p = 0.0003) with every 1 increment in age and BMI, respectively. Dose similarities between insulin detemir and insulin glargine were seen in type 2 diabetes patients when administered once daily. Thus, the use of insulin detemir and insulin glargine is not associated with different medical costs if the price and treating algorithm are similar.     

Insulin-induced hypoglycaemia in an accident and emergency department: the tip of an iceberg?

In one year a prospective survey in a large accident and emergency department identified 204 admissions of adults with severe hypoglycaemia, 200 in insulin-treated patients. Ninety-six had one admission while 34 others were admitted on 104 occasions. Of the 130 patients, 111 attended diabetic clinics in Nottingham, forming 9% of a known clinic population of 1229 on insulin treatment. Since many other episodes of hypoglycaemia were presumably treated outside hospital, 9% a year is a minimum estimate of the incidence of severe hypoglycaemia in our area. The mean insulin dose was 1.2 units/kilogram/day for those admitted twice or more and 0.9 U/kg/day for those admitted once; these doses were significantly higher than those of an age-matched clinic population. A year after the latest admission with hypoglycaemia, the mean insulin dose in the group with two or more admissions had fallen to 0.8 U/kg/day, suggesting that over-treatment had been an important causal factor. A similarly high incidence has been reported in other studies, and we believe that it is due mainly to the inadequacy of conventional subcutaneous insulin treatment.     

A monomer-dimer model explains the results of radiation inactivation: binding characteristics of insulin receptor purified from human placenta

The technique of radiation inactivation has been used on highly purified human placental insulin receptor in order to determine the functional molecular size responsible for the insulin binding and to evaluate the "affinity regulator" hypothesis, which has been proposed to explain the increase in specific insulin binding to rat liver membranes observed at low radiation doses [Harmon, J. T., Hedo, J. A., & Kahn, C. R. (1983) J. Biol. Chem. 258, 6875-6881]. Three different types of inactivation curves were observed: (1) biphasic with an enhanced binding activity after exposure to low radiation doses, (2) nonlinear with no change in binding activity after exposure to low radiation doses, and (3) linear with a loss in the binding activity with increasing radiation exposures. A monomer-dimer model was the simplest model that best described the three types of radiation inactivation curves observed. The model predicts that an increase in insulin binding activity would result after exposure to low radiation doses when the initial dimer/monomer ratio is equal to or greater than 1 and a monomer is more active than a dimer. The monomer size of the binding activity was estimated to be 227,000 daltons by this model. This value most likely reflects the size of the monomeric alpha beta form. To substantiate this model, the purified receptor was fractionated by Sepharose CL-6B chromatography. The insulin binding profile of this column indicated two peaks.(ABSTRACT TRUNCATED AT 250 WORDS)     

A Bridge to Insulin Pump Therapy: Twice-Daily Regimen with NPH and Detemir Insulins During Initial Treatment of Youth with Type 1 Diabetes Mellitus

ObjectiveTo describe clinical outcomes in youth with new-onset type 1 diabetes mellitus (TlDM) treated with a modified, twice-daily regimen of a mixture of NPH insulin and rapid-acting insulin analogue at breakfast and separate injections of rapid-acting insulin analogue and insulin detemir at dinner.MethodsOur clinic database was used to describe changes in insulin doses, hemoglobin A_1c (A1C) levels, and frequency of severe hypoglycemia during the first year of therapy in young patients with T1DM diagnosed between September 2006 and April 2009. Data are presented as median values (25%, 75%).ResultsOverall, 108 patients (62 girls; mean age, 10.0 ± 0.4 years) were eligible for inclusion. Total daily insulin doses at 3, 6, and 12 months were 0.6 (0.4, 0.8), 0.7 (0.4, 0.9), and 0.8 (0.6, 0.9) U/kg, respectively. A1C levels were 9.8% (8.5%, 10.8%) at 2 weeks (baseline). Of the 108 patients, 19 had switched to insulin pump therapy by 3 months and 49 had switched by 12 months after initial diagnosis of T1DM. The 49 pump-treated patients had an A1C of 6.9% (6.6%, 7.3%), whereas the 59 injection- treated patients had an A1C of 7.2% (6.7%, 7.7%) by 12 months. There were only 6 severe hypoglycemic events in 5 patients; none occurred during the first 3 months, none occurred during the night, and all occurred in patients receiving insulin injection treatment.ConclusionA twice-daily insulin regimen that uses insulin detemir for overnight basal replacement and morning NPH insulin to avoid lunch and afternoon snack doses is an effective initial treatment for young patients with new-onset T1DM that can provide a smooth transition to intensive basal/bolus insulin pump therapy. (Endocr Pract. 2011;17:862-866)     

Is Dynamic Autocrine Insulin Signaling Possible? A Mathematical Model Predicts Picomolar Concentrations of Extracellular Monomeric Insulin within Human Pancreatic Islets

Insulin signaling is essential for -cell survival and proliferation in vivo. Insulin also has potent mitogenic and anti-apoptotic actions on cultured -cells, with maximum effect in the high picomolar range and diminishing effect at high nanomolar doses. In order to understand whether these effects of insulin are constitutive or can be subjected to physiological modulation, it is essential to estimate the extracellular concentration of monomeric insulin within an intact islet. Unfortunately, the in vivo concentration of insulin monomers within the islet cannot be measured directly with current technology. Here, we present the first mathematical model designed to estimate the levels of monomeric insulin within the islet extracellular space. Insulin is released as insoluble crystals that exhibit a delayed dissociation into hexamers, dimers, and eventually monomers, which only then can act as signaling ligands. The rates at which different forms of insulin dissolve in vivo have been estimated from studies of peripheral insulin injection sites. We used this and other information to formulate a mathematical model to estimate the local insulin concentration within a single islet as a function of glucose. Model parameters were estimated from existing literature. Components of the model were validated using experimental data, if available. Model analysis predicted that the majority of monomeric insulin in the islet is that which has been returned from the periphery, and the concentration of intra-islet monomeric insulin varies from 50–300 pM when glucose is in the physiological range. Thus, our results suggest that the local concentration of monomeric insulin within the islet is in the picomolar ‘sweet spot’ range of insulin doses that activate the insulin receptor and have the most potent effects on -cells in vitro. Together with experimental data, these estimations support the concept that autocrine/paracrine insulin signalling within the islet is dynamic, rather than constitutive and saturated.     

The difficult choice of treatment for poorly controlled maturity onset diabetes: tablets or insulin?

Patients with maturity onset diabetes that is poorly controlled on maximal doses of oral hypoglycaemic agents are difficult to treat. A prospective randomised crossover study was performed in 58 predominantly non-obese patients on maximal doses of glibenclamide or metformin, or both, to find out if insulin would improve control or well being. The patients were given daily injections of up to 48 units of highly purified porcine lente insulin. Glycaemic control was improved by 15% or more in only 18 patients; 14 others felt better but their diabetes was no better controlled. Those whose control was improved by insulin could not be distinguished by age, duration of diabetes, body mass index, or their own treatment preference. C peptide concentrations, however, did help predict the response to insulin, the fasting C peptide to glucose ratio being considerably lower in those patients whose control was better on insulin. These findings suggest that a simple insulin regimen does not necessarily lead to better glycaemic control in maturity onset diabetes. Nevertheless, a trial of insulin is often justified since it poses few practical difficulties and makes some patients feel better even if their control is not improved. A more complex regimen might improve control in more cases, but it might also be less acceptable to older patients.     

Glycemic Outcomes of Hospitalized Patients on Ambulatory Humulin-R U-500 Insulin

ObjectiveManaging hospitalized patients on ambulatory U-500 insulin is challenging because of limited guidance on how to safely adjust insulin doses during admission. We sought to evaluate glycemic outcomes in relation to inpatient insulin doses in patients receiving U-500 prior to hospitalization.MethodsRetrospective study of hospitalized patients on ambulatory U-500 seen consecutively from January 2015 to December 2019. Primary outcomes were inpatient hypoglycemia, hyperglycemia, and normoglycemia at different insulin dosages expressed as weight-based (unit/kg/d) inpatient total daily dose (TDD) and ratio of inpatient to outpatient TDD.ResultsWe identified 66 admissions of 46 unique patients. The median (interquartile range) body mass index was 41.0 kg/m² (35.1, 46.8), home TDD 212 units (120, 300), and home insulin dose 1.6 units/kg/d (1.1, 2.2). The median (interquartile range) inpatient insulin dose was 0.7 unit/kg/d (0.3, 1.0) and the ratio of inpatient to outpatient TDD was 0.4 (0.2, 0.8). Hyperglycemia persisted throughout the hospitalization. For the outcomes of hyperglycemia and normoglycemia, we found no association between increased levels of insulin dosages. For the outcome of hypoglycemia, significantly higher odds were observed when non-fasting patients received an inpatient TDD that was either > 40% of their home TDD or > 0.6 unit/kg/d of insulin.ConclusionPatients on ambulatory U-500 have significant hyperglycemia during admission. Inpatient insulin doses of 40% of home TDD or ≤ 0.6 unit/kg were not associated with increased hypoglycemia risk. Further prospective studies are needed to determine effective doses in these high-risk patients.