Clinical Hypoglycemia Research Articles

Hyperglycemia and Adverse Pregnancy Outcome Study: Neonatal Glycemia

The goal was to describe the temporal pattern of neonatal plasma glucose levels and associations with maternal glucose levels, cord serum C-peptide levels, and neonatal size and adiposity. A total of 17,094 mothers and infants were included in the Hyperglycemia and Adverse Pregnancy Outcome Study (15 centers in 9 countries). Mothers underwent a 75-g, 2-hour, oral glucose tolerance test (OGTT) at 24 to 32 weeks of gestation. Cord blood and neonatal blood samples were collected. Biochemical neonatal hypoglycemia was defined as glucose levels of <10th percentile (2.2 mmol/L). Clinically identified hypoglycemia was ascertained through medical record review and associations were assessed. Plasma glucose concentrations were stable during the first 5 hours after birth. Maternal glucose levels were weakly positively associated with biochemical neonatal hypoglycemia (odds ratios: 1.07-1.14 for 1-SD higher OGTT glucose levels). Frequency of neonatal hypoglycemia was higher with higher cord C-peptide levels (odds ratio: 11.6 for highest versus lowest C-peptide category). Larger and/or fatter infants were more likely to have hypoglycemia (P < .001), and infants with hypoglycemia tended to have a higher frequency of cord C-peptide levels of >90th percentile. Mean neonatal plasma glucose concentrations varied little in the first 5 hours after birth, which suggests normal postnatal adjustment. Biochemical and clinical hypoglycemia were weakly related to maternal OGTT glucose measurements but were strongly associated with elevated cord serum C-peptide levels. Larger and/or fatter infants were more likely to develop hypoglycemia and hyperinsulinemia. These relationships suggest physiologic relationships between maternal glycemia and fetal insulin production.

Gestational diabetes: risks, management, and treatment options

Gestational diabetes mellitus (GDM) is commonly defined as glucose intolerance first recognized during pregnancy. Diagnostic criteria for GDM have changed over the decades, and several definitions are currently used; recent recommendations may increase the prevalence of GDM to as high as one of five pregnancies. Perinatal complications associated with GDM include hypertensive disorders, preterm delivery, shoulder dystocia, stillbirths, clinical neonatal hypoglycemia, hyperbilirubinemia, and cesarean deliveries. Postpartum complications include obesity and impaired glucose tolerance in the offspring and diabetes and cardiovascular disease in the mothers. Management strategies increasingly emphasize optimal management of fetal growth and weight. Monitoring of glucose, fetal stress, and fetal weight through ultrasound combined with maternal weight management, medical nutritional therapy, physical activity, and pharmacotherapy can decrease comorbidities associated with GDM. Consensus is lacking on ideal glucose targets, degree of caloric restriction and content, algorithms for pharmacotherapy, and in particular, the use of oral medications and insulin analogs in lieu of human insulin. Postpartum glucose screening and initiation of healthy lifestyle behaviors, including exercise, adequate fruit and vegetable intake, breastfeeding, and contraception, are encouraged to decrease rates of future glucose intolerance in mothers and offspring.

Establishing consensus criteria for the diagnosis of diabetes in pregnancy following the HAPO study

The current diagnostic criteria for gestational diabetes mellitus are controversial because they lack correlation to maternal and perinatal outcome. The results of the hyperglycemia and adverse pregnancy outcome (HAPO) study demonstrate a linear association between increasing levels of fasting, 1- and 2-h plasma glucose post a 75 g oral glucose tolerance test to several significant outcome endpoints, such as birth weight above the 90th percentile, cord blood serum C-peptide level above the 90th percentile, primary cesarean delivery, clinical neonatal hypoglycemia, premature delivery, shoulder dystocia or birth injury, intensive neonatal care admission, hyperbilirubinemia, and preeclampsia. A consensus report by the IADPSG, based on a vigorous assessment of the HAPO results and other studies, recommended an endorsement of risk-based, internationally accepted criteria for the diagnosis and classification of diabetes in pregnancy. This review follows the steps from defining the problem to the endpoint of achieving a worldwide policy change.

A dose range finding study of novel oral insulin (IN‐105) under fed conditions in type 2 diabetes mellitus subjects

The objective of the study was to establish the dose response of IN-105 tablets and explore a possible therapeutic window in type 2 diabetes subjects poorly controlled on metformin. The primary objective was to examine the effect of sequential single ascending doses of IN-105 on the plasma glucose concentration under fed conditions. All subjects received, sequentially, matching placebo, 10, 15, 20 and 30 mg IN-105 tablets in five consecutive periods. Tablets were administered 20 min prior to meal in all the periods. Plasma levels of immunoreactive insulin, C-peptide and glucose were measured up to 180 min from the time of dosing. The changes in postprandial glucose levels at 120 min in response to IN-105 administration were also compared against those of placebo. Changes in glucose from baseline (mean +/- s.d.) at 140 min (2 h postprandial) were 94.84 +/- 22.3, 79.45 +/- 43.00, 70.68 +/- 35.71, 63.47 +/- 42.75 and 53.06 +/- 47.27 mg/dL, respectively, and exhibited linear dose-response. The insulin C(max) values were found to be 50.8 +/- 26.0 mU/L for placebo, 100.3 +/- 66.7 with 10 mg IN-105, 177.69 +/- 150.3 with 15 mg IN-105, 246.2 +/- 245.2 with 20 mg IN-105 and 352.5 +/- 279.3 mU/L with 30 mg of IN-105. IN-105 absorption is proportional to the dose administered. The 2-h postprandial glucose excursion was reduced in a dose proportional manner. Circulating C-peptide levels were found to be suppressed in proportion to the IN-105 exposure. IN-105 reduces glucose excursion despite lower endogenous insulin secretion. IN-105 seems to have a wide therapeutic window as no clinical hypoglycaemia was observed at any of the doses studied.

Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study: associations with maternal body mass index

To determine whether higher maternal body mass index (BMI), independent of maternal glycaemia, is associated with adverse pregnancy outcomes. Observational cohort study. Fifteen centres in nine countries. Eligible pregnant women. A 75-g 2-hour oral glucose tolerance test (OGTT) was performed between 24 and 32 weeks of gestation in all participants. Maternal BMI was calculated from height and weight measured at the OGTT. Fetal adiposity was assessed using skinfold measurements and percentage of body fat was calculated. Associations between maternal BMI and pregnancy outcomes were assessed using multiple logistic regression analyses, with adjustment for potential confounders. Predefined primary outcomes were birthweight >90th percentile, primary caesarean section, clinical neonatal hypoglycaemia and cord serum C-peptide >90th percentile. Secondary outcomes included pre-eclampsia, preterm delivery (before 37 weeks) and percentage of body fat >90th percentile. Among 23 316 blinded participants, with control for maternal glycaemia and other potential confounders, higher maternal BMI was associated (odds ratio [95% confidence interval] for highest {> or =42.0 kg/m(2)} versus lowest {<22.6 kg/m(2)} BMI categories) with increased frequency of birthweight >90th percentile (3.52 [2.48-5.00]) and percentage of body fat >90th percentile (3.28 [2.28-4.71]), caesarean section (2.23 [1.66-2.99]), cord C-peptide >90th percentile (2.33 [1.58-3.43]) and pre-eclampsia (14.14 [9.44-21.17]). Preterm delivery was less frequent with higher BMI (0.48 [0.31-0.74]). Associations with fetal size tended to plateau in the highest maternal BMI categories. Higher maternal BMI, independent of maternal glycaemia, is strongly associated with increased frequency of pregnancy complications, in particular those related to excess fetal growth and adiposity and to pre-eclampsia.

Glucagon-Like Peptide-1 Receptor Imaging for Localization of Insulinomas

The surgical removal of insulinomas is hampered by difficulties to localize it using conventional radiological procedures. Recently these tumors were shown to exhibit a very high density of glucagon-like peptide-1 receptors (GLP-1R) in vitro that may be used as specific targets for in vivo receptor radiolabeling. The objective of the study was to test the 111In-labeled GLP-1R agonist 111In-DOTA-exendin-4 in localizing insulinomas using single photon emission computed tomography in combination with computed tomography images. This was a prospective open-label investigation. The study was conducted at three tertiary referral centers in Switzerland. Patients included six consecutive patients with proven clinical and biochemical endogenous hyperinsulinemic hypoglycemia. (111)In-DOTA-exendin-4 was administered iv at a dose of about 90 MBq (30 microg peptide) over 5 min. Whole-body planar images of the abdomen were performed at 20 min, 4 h, 23 h, 96 h, and up to 168 h after injection. After surgical removal of the insulinomas, GLP-1R expression was assessed in the tumor tissue in vitro by GLP-1R autoradiography. The detection rate of insulinomas was measured. In all six cases, the GLP-1R scans successfully detected the insulinomas identified using conventional methods in four cases. By using a gamma-probe intraoperatively, GLP-1R detection permitted a successful surgical removal of the tumors in all patients, diagnosed histopathologically as five pancreatic and one extrapancreatic insulinomas. In vitro GLP-1R autoradiography showed a high density of GLP-1R in all tested insulinomas. In vivo GLP-1R imaging is an innovative, noninvasive diagnostic approach that successfully localizes small insulinomas pre- and intraoperatively and that may in the future affect the strategy of insulinoma localization.

Insulin treatment in children and adolescents with diabetes

Department of Pediatrics, Uddevalla Hospital, Uddevalla,SwedenCorresponding author:Hans-Jacob Bangstad, MDDepartment of PaediatricsOslo University HospitalUlleval˚Kirkeveien 166Oslo N-0407Norway.Tel: 0047-22118765;fax: 0047-22118663;e-mail: haba@uus.noConflicts of interest: RH has received lecture honoraria fromNovo Nordisk, Lilly, Sanofi-Aventis, Medtronic and Roche. PJ-Chas received lecture fees from Eli Lilly, Novo-Nordisk, Sanofi-Aventis, Abbott, Bayer, Medtronic and Roche.The remainingauthors have declared no potential conflicts.Editors of the ISPAD Clinical Practice Consensus Guide-lines 2009 Compendium: Ragnar Hanas, Kim Donaghue,Georgeanna Klingensmith, and Peter Swift.This article is a chapter in the

Environmental rather than genetic fetal overgrowth: defining the difference and hints for diagnosis and management

Environmental rather than genetic fetal overgrowth: defining the difference and hints for diagnosis and management

Unrecognized hypoglycemia due to maltodextrin interference with bedside glucometry

Glucometry is widely used to confirm or exclude hypoglycemia in patients with suggestive clinical findings. Nonglucose sugars may be detected by certain types of glucometers, causing false elevation of the glucometer analysis of the blood sugar. Since these other sugars are not functionally glucose and may even induce excess insulin release, clinical hypoglycemia may be missed. We report a 79-year-old man on enteral feeds containing maltodextrin, a glucose polymer, who had persistently high glucometer-measured blood glucose despite normal blood glucose measured by formal laboratory analysis. Excess insulin administration, based on the erroneous glucometer reading, may have caused unrecognized fatal clinical hypoglycemia. This has been reported following intravenous administration of related nonglucose sugars but not with enteral maltodextrin. Further study is required to confirm the effects of maltodextrin on glucometry. False elevation of blood glucose measured on certain point-of-care glucometers can occur following the oral administration of maltodextrin.

Neonatal glucose metabolism in offspring of mothers with varying degrees of hyperglycemia during pregnancy

The definition of neonatal hypoglycemia is controversial. Operational thresholds of blood glucose values at which intervention should be considered have been proposed. IDM and GDM infants frequently exhibit a pronounced drop of plasma glucose immediately after birth. This exaggerated physiological decline of glucose is transient and is seldom accompanied by suppressed lipolysis or clinical symptoms. It is generally attributed to hyperinsulinism elicited by maternal hyperglycemia. Alternative substrates for CNS i.e. lactate and astrocyte glycogen may explain lack of symptoms. Similarly low glucose values later on may cause clinical symptoms. Glucose production rates vary from attenuated to normal likely reflecting differences in maternal glycemic control. The HAPO study of around 25,000 non-diabetic pregnancies revealed strong associations between glucose values (75g OGTT) and increased fetal size and hyperinsulinemia at birth - findings adding strong support to the maternal hyperglycemia - fetal hypinsulinism theory. Mothers with the highest fasting glucose had infants with the highest frequency of clinical neonatal hypoglycaemia (4,6%).

Retrospective Study of Canine Insulinomas: Eight Cases (2005-2008)

This retrospective study examined 8 cases of pathologically diagnosed insulinoma. Surgical intervention was performed in all cases, once insulinoma was suspected based on unique clinical signs, hypoglycemia, and/or findings on ultrasonography. At the same time, insulin and glucose concentrations were measured before surgery for clinical diagnosis. Although all cases displayed typical clinical signs and hypoglycemia, insulin concentrations varied with 3 of 8 cases showing levels within the reference interval. In addition, to confirm the reliability of measured values, we submitted serum samples from 4 cases to two commercial veterinary laboratories. Results differed considerably between laboratories, with no apparent correlations between the two. In Laboratory A, 3 of 4 cases were above the reference interval, and 1 case was in the middle of the reference interval. Conversely, in Laboratory B, 3 of 4 cases were above the reference interval, and 1 case was below the reference interval. Split decisions regarding the diagnosis of insulinoma were seen for 2 of the 4 cases.

Antepartum and intra-partum insulin management of type 1 and type 2 diabetic women: Impact on clinically significant neonatal hypoglycemia

Objective Assess efficacy of intrapartum intravenous (i.v.) insulin and effect of antepartum and intrapartum diabetic control on various measures of clinical neonatal hypoglycemia. Design Retrospective chart review. Setting Regional Diabetes in Pregnancy Clinic. Results Maternal hypoglycemia occurred intrapartum in 56% of subjects. Mean delivery blood glucose with intravenous insulin use was 6.7 mmol/l. Capillary blood glucose <2.2 mmol/l occurred in 69% of neonates and 44% received intravenous glucose. Maternal delivery glucose >6.5 mmol/l correlated with occurrence of neonatal glucose <2.2 mmol/l but not requirement for i.v. glucose or NICU admission. A third trimester HbA1c >6.5% had a stronger association with NICU admission and i.v. glucose requirement. Conclusions Blood glucose at delivery >6.5 mmol/l predicts neonatal hypoglycemia but does not correlate with severity. Chronic maternal hyperglycemia, reflected by pre-delivery HbA1c, predicts severe fetal hyperinsulinism and requirement for aggressive intervention. This stresses the importance of a target third trimester HbA1c of <6.5%.

Continuous Subcutaneous Insulin Infusion and Multiple Dose of Insulin Regimen Display Similar Patterns of Blood Glucose Excursions in Pediatric Type 1 Diabetes

Continuous subcutaneous insulin infusion (CSII) is believed to decrease glycemic variability and clinical hypoglycemia compared with the multiple daily insulin (MDI) regimen. To compare the indices of glycemic instability between CSII and MDI, we analyzed the continuous glucose monitoring system (CGMS) (Medtronic MiniMed, Northridge, CA) profiles of a group of children with type 1 diabetes mellitus with history of frequent blood glucose (BG) fluctuations and hypoglycemia. Data from 14 (nine girls, five boys) patients (3.9-16.8 years old) on CSII and 14 age- and sex-matched (nine girls, five boys) patients (3.9-16.0 years old) on MDI with similar glycemic control (hemoglobin A1c: 7.9 +/- 1.0% vs. 7.9 +/- 1.5%) and body mass index (BMI) (20.1 +/- 4.3 vs. 19.9 +/- 4.1 kg/m(2)) were evaluated by the CGMS. Mean BG (MBG), absolute means of daily differences (MODD), mean amplitude of glycemic excursion (MAGE), and number of hypoglycemic events (BG <60 mg/dL) for 48 h were calculated. The MBG, MODD, MAGE, and number and mean duration of hypoglycemia events in the CSII group were similar to those in the MDI group. The MAGE had an inverse correlation with age (CSII: r (2) = 0.52, P = 0.003; MDI: r (2) = 0.29, P < 0.04) and BMI (CSII: r (2) = 0.38, P < 0.02; MDI: r (2) = 0.71, P < 0.0002). However, there was a positive relationship between MAGE and bolus:basal insulin ratio in the CSII (r (2) = 0.28, P < 0.05) and MDI (r (2) = 0.33, P < 0.03) groups. Also, the MAGE had a positive correlation with frequency of hypoglycemic events in the CSII (r (2) = 0.44, P < 0.01) and MDI (r (2) = 0.35, P < 0.03) groups. The CSII and MDI regimens in children and adolescents with comparable glycemic control displayed similar patterns of glycemic excursions, implying that factors influencing glycemic instability in pediatric type 1 diabetes mellitus appear to be independent of treatment modality.

Metabolic effects of pioglitazone in combination with insulin in patients with type 2 diabetes mellitus whose disease is not adequately controlled with insulin therapy: Results of a six-month, randomized, double-blind, prospective, multicenter, parallel-group study

Type 2 diabetes mellitus (DM) is a progressive disease. Initial therapy begins with dietary and lifestyle modifications. However, as the disease progresses, glycemic control becomes more difficult to attain, often requiring > or =1 oral antihyperglycemic medication (OAM), and finally the addition of insulin to the OAMs and insulin monotherapy. This study was designed to determine the effect of pioglitazone 30 mg plus insulin (PIO + INS) versus placebo plus insulin (PLB + INS) on glycemic control, the serum lipid profile, and selected cardiovascular risk factors in patients with type 2 DM whose disease was inadequately controlled with insulin therapy alone despite efforts to intensify such treatment. This was a 6-month, randomized, double-blind, prospective, multicenter, placebo-controlled, parallel-group study. Patients with type 2 DM and a glycosylated hemoglobin (HbA(1c)) value > or =7.5% who were using insulin (with or without OAMs) entered a 3-month insulin intensification phase to achieve blood glucose targets with insulin monotherapy. After insulin intensification, those patients with HbA(1c) values > or =7.0% were randomized to PIO + INS or PLB + INS. The primary end point was the change in HbA(1c) from baseline. Cardiovascular risk markers (highly sensitive C-reactive protein [hs CRP] and plasminogen activator inhibitor-1 [PAI-1]) were measured at baseline and end point. Of the 289 patients randomized to treatment (mean [SD] age, 58.9 [7.1] years; 164 women, 125 men), 142 received PIO + INS and 147 received PLB + INS. A total of 263 patients completed the study. After 6 months, PIO + INS reduced mean HbA(1c) (-0.69%; P < 0.002) and mean fasting plasma glucose ([FPG] -1.45 mmol/L; P < 0.002) from baseline. PLB + INS produced no significant changes in HbA(1c) or FPG. The between-treatment differences for HbA(1c) (-0.55%; P < 0.002) and FPG (-1.80 mmol/L; P < 0.002) occurred despite a reduction of insulin dose in the PIO + INS group from baseline (-0.16 U/d . kg; P < 0.002). Significant between-group differences were observed for high-density lipoprotein cholesterol (0.13 mM; P < 0.002), triglycerides (ratio of geometric mean [PIO/PLB], 0.871; P < 0.01), atherogenic index of plasma (-0.11; P < 0.002), PAI-1 (-5.10 U/mL; P < 0.001), and hs CRP (-1.47 mg/L; P < 0.05). The rate of clinical and biochemical hypoglycemia (blood glucose <2.8 mmol/L) did not differ statistically between treatment groups, but reported incidences of subjective hypoglycemia occurred more often with PIO + INS than with PLB + INS (90 vs 75; P < 0.05). Edema was more common with PIO + INS than with PLB + INS (20 vs 5 instances, respectively), as was gain (mean [SEM]) in body weight (4.05 [4.03] vs 0.20 [2.92] kg, respectively). Adding pioglitazone to insulin in these study patients with type 2 DM whose disease was inadequately controlled with insulin monotherapy further improved their glycemic control.

Changes in cardiac repolarization during clinical episodes of nocturnal hypoglycaemia in adults with Type 1 diabetes.

Experimental hypoglycaemia leads to abnormal cardiac repolarization manifest by a lengthened QT interval and caused by adrenergic stimulation. However it is less clear whether spontaneous clinical episodes lead to similar changes. We have therefore measured cardiac ventricular repolarization and counterregulatory responses in patients with Type 1 diabetes during hypoglycaemic and euglycaemic nights. We studied 22 patients with Type 1 diabetes (mean age 40.4+/-17.2 years, duration of diabetes 17.2+/-9.3 years, HbA1c 8.2+/-1.2% overnight). Measurements were taken hourly of blood glucose, plasma potassium, catecholamines and high resolution electrocardiograms. Hypoglycaemia (blood glucose level <2.5 mmol/l) occurred on 7 of the 22 nights. During overnight hypoglycaemia, QTc interval increased by 27 ms (+/-15) above baseline, compared with 9 ms (+/-19) during nights with no nocturnal hypoglycaemia (p=0.034, 95%CI 2, 35). Adrenaline increased by 0.33 nmol/l (+/-0.21) above baseline during hypoglycaemia, compared with -0.05 nmol/l (+/-0.08) during euglycaemia (p=0.001, 95%CI 0.19, 0.56 nmol/l). There was no significant difference between potassium, and noradrenaline concentrations between the two groups. QTc interval lengthens significantly during spontaneous nocturnal hypoglycaemia. Increases are generally less than those observed during experimental hypoglycaemia and could reflect attenuated sympathoadrenal responses during clinical episodes. The clinical relevance of these changes is uncertain but is consistent with the hypothesis that clinical hypoglycaemia can cause abnormal cardiac repolarization and an attendant risk of cardiac arrhythmia.

Hypoglycaemia in Diabetes

<h2>Abstract</h2> Most cases of clinical hypoglycaemia are insulin-induced episodes in patients with diabetes. Hypoglycaemia is probably the major factor preventing insulin-treated patients from achieving the glucose targets needed to prevent diabetic complications. The incidence of hypoglycaemia reflects the inadequacy of current methods of insulin delivery, which lead to inappropriately high insulin concentrations, particularly some hours after eating and at night.

Management of Feline Diabetes Mellitus

Up to one quarter of diabetic cats can be well controlled with oral hypoglycemic drugs, although at least 75% require insulin therapy. Most available insulins provide good clinical control but only moderate glycemic control. Because mild to moderate hyperglycemia is well tolerated by cats receiving insulin but hypoglycemia can be life threatening, conservative insulin dosing is recommended. Clinical signs and water intake indicate whether a dose adjustment is required, but serial blood glucose measurements are usually needed to determine the direction of the adjustment. Starting doses of 0.3 to 0.5 IU/kg administered twice daily (rounded down to the nearest whole unit) are usually safe. Dose adjustments should not exceed 1 IU per cat every 2 to 4 weeks unless clinical hypoglycemia has occurred. Cats with clinical hypoglycemia need to be reassessed to see if they are in remission. If not, a 50% to 75% reduction in dose is advised. Approximately 30% of cats go into diabetic remission 1 to 4 months after an adequate treatment protocol is instituted.

Current understanding of feline diabetes: Part 2, treatment

When treating diabetic cats, the primary aim is to control clinical signs without causing clinical hypoglycaemia. Secondary goals are to maximise the chances of attaining diabetic remission and to minimise the risk of complications due to chronic hyperglycaemia. A treatment plan that is convenient for the owner is important for compliance. Underweight or overweight diabetic cats should be fed with the aim of normalising bodyweight. Current evidence suggests that non-obese diabetic cats can be fed ad libitum. The oral hypoglycaemic drug glipizide is well established as a treatment for about a third of diabetic cats, which have residual beta cell function. Preliminary studies on other oral agents such as vanadium salts, metformin, and troglitazone indicate a potential use in some diabetic cats. Insulin treatment remains the treatment of choice for the majority of diabetic cats. Choice of insulin, dose rates and monitoring of treatment are discussed.

Neoplasms Causing Nonhyperinsulinemic Hypoglycemia.

Non-islet cell tumor hypoglycemia (NICTH) is uncommon. Many of the tumors associated with NICTH are mesenchymal tumors, although carcinomas are also involved in some cases. High serum levels of insulin-like growth factor II (IGF-II) have been associated with NICTH. Analysis of 4 pituitary tumors, 2 adrenocortical tumors, 42 solitary fibrous tumors, and 23 other mesenchymal tumors for IGF-II protein and mRNA showed that most mesenchymal tumors expressed IGF-II protein and mRNA, although only 4 of 48 patients had associated hypoglycemia. Tumor size was related to IGF-II production. Tumors less than 5 cm were usually negative for IGF-II mRNA, whereas 92.3% of tumors greater than 9 cm were positive for IGF-II mRNA. These results show that IGF-II mRNA and protein can be readily detected in many tumors, even when the tumors are not associated with clinical hypoglycemia. The expression and production of this growth factor cannot accurately predict patients with clinical evidence of hypoglycemia.

Diabetic hypoglycaemia.

Hypoglycaemia is a major factor preventing insulin-treated patients from achieving normoglycaemia. This reflects the inadequacy of current insulin treatment, which causes high insulin concentrations in the post-absorptive period. Physiological defences to hypoglycaemia include autonomic activation, which limits the fall in glucose level and causes symptoms, alerting patients to an impending episode. Many patients develop defective responses and hypoglycaemia unawareness after longstanding disease or with tight glycaemic control and are then prone to severe attacks. This may be the result of repeated hypoglycaemic episodes, which by altering cerebral glucose uptake, disturb the mechanisms that activate the central response to hypoglycaemia. Preventing further hypoglycaemia can partially reverse these defects and restore symptomatic awareness. Clinical hypoglycaemia has also been implicated in the 'dead in bed' syndrome and in chronic cognitive impairment. The problem of hypoglycaemia will eventually be solved by better insulin delivery and non-invasive glucose meters, but until then, more focused education may have a more substantial impact.