Pharmacokinetic and glucodynamic variability: assessment of insulin glargine, NPH insulin and insulin ultralente in healthy volunteers using a euglycaemic clamp technique

Aims/hypothesis This single-dose, double-blind, randomised, parallel-group study evaluated the reproducibility in systemic exposure and glucodynamic effect of insulin glargine, NPH insulin (NPH) and insulin ultralente (ultralente) using the manually adjusted euglycaemic clamp technique.Methods In total, 36 healthy volunteers received two consecutive s.c. injections (0.4 IU/kg) of glargine, NPH or ultralente with a wash-out period of 7 days between treatments.Results In healthy volunteers, glargine presented well-reproduced flat concentration profiles and no pronounced peaks in activity. NPH, by contrast, showed well-defined peaks in concentration and glucose disposal, while ultralente had highly variable profiles. Within-subject variability (ANOVA) for insulin exposure over 24 h was 15% for glargine and 19% for NPH, compared with 67% for ultralente (p<0.05, glargine and NPH vs ultralente). The 49% within-subject variability in total glucose disposal (glucose infusion rate [GIR]-AUC_{0–24 h}) with ultralente was about twice as large as the 22% with NPH (p<0.05), but was intermediate with glargine at 31% (p=NS). By contrast, variability in the diurnal time-action profile (SD of diurnal day-to-day differences in GIR) for glargine was 30% (p<0.05) and 50% (p<0.05) less than with NPH and ultralente, respectively. No serious adverse events were reported.Conclusions/interpretation Although representing insulins of different profiles, glargine and NPH showed a high and similar reproducibility of total absorption and glucodynamic effect, whereas ultralente proved to have poor reproducibility. However, while NPH yields peaks in concentration and activity, glargine shows flat and non-fluctuating profiles resulting in less variation in day-to-day 24-h activity.     

The Influence of Insulin Antibodies on the Pharmacokinetics of NPH Insulin in Patients with Type 1 Diabetes treated with Human Insulin

The influence of insulin binding antibodies on the pharmacokinetics of NPH insulin was studied in Type 1 diabetic patients on human insulin. Insulin-antibody binding (B_o) was measured during a screening procedure in 155 Type 1 diabetic patients. In 36 patients, B_o was <1.5%, and in 38 patients B_o was >10.0%. Of these, 6 patients, group 1 (B_o < 1.5%) and 8 patients, group 2 (B_o > 10.0%), respectively, subsequently participated in a pharmacokinetic study. Free insulin and the glucose infusion rate were measured using a euglycaemic clamp after subcutaneous injection of NPH insulin (0.4 U kg^?1). The areas under the curve (AUC) of free insulin concentration were smaller for group 2 (p = 0.01) than for group 1 (212.2 ± 22.0 vs 316.8 ± 25.3 mU l^?1 h). The AUCs of the glucose infusion rate were also smaller for group 2 (p < 0.05) than for group 1 (2.50 ± 0.32 vs 3.58 ± 0.36 g kg^?1). A significant negative correlation exists between the AUCs for free insulin concentration and insulin-antibody binding B_o (r=0.76, p = 0.001). The daily insulin dosage was higher in group 2 (p = 0.02) than in group 1 (0.66 ± 0.03 vs 0.53 ± 0.03 U kg^?1). We conclude that insulin antibodies influence the pharmacokinetics of NPH human insulin. The demonstrable influence on the kinetics of free insulin and glucose utilization leads to a slight increase in daily total insulin requirements.     

A comparison of semisynthetic human NPH insulin and porcine NPH insulin in the treatment of insulin-dependent diabetes mellitus

Twenty-two insulin-dependent diabetic patients participated in a double-blind, cross-over study, where treatment with semisynthetic human NPH insulin (Novo Industri) was compared with porcine NPH insulin (Nordisk). Each treatment period lasted 8 weeks. Blood glucose level, glycosylated haemoglobin, insulin requirements, and frequency of hypoglycaemic events were compared. No difference was found in 24-hour blood glucose profiles. Fasting blood glucose level was 8.3 mmol/l during treatment with human insulin and 8.7 mmol/l during treatment with porcine insulin (p less than 0.1). Mean HbA1c was 7.7% at the end of study compared to 9.5% at baseline (p less than 0.01), but this decline in HbA1c was independent of the treatment regimen. Forty-six hypoglycaemic events occurred during treatment with human insulin compared to 39 events during treatment with porcine insulin. No difference was found regarding insulin requirements during the study. It is concluded that semisynthetic human NPH insulin is indistinguishable from porcine NPH insulin with respect to 24-hour blood glucose profile, HbA1c level and insulin dose requirements.     

Insulin action and pharmacokinetics in insulin treated diabetics during the third trimester of pregnancy

Insulin action and pharmacokinetics were compared, using the euglycaemic insulin clamp technique, in seven insulin-treated diabetics during the third trimester of pregnancy and one to three weeks post-partum. At an insulin infusion rate of 1 mU/kg/min, insulin mediated glucose disposal was significantly greater (p less than 0.02) following delivery (1.194 +/- 0.138 mmol/m2/min) than in pregnancy (0.761 +/- 0.072 mmol/m2/min) and the rate of decline in insulin mediated glucose disposal, at the end of the insulin infusion, was significantly greater (p less than 0.02) following delivery (24.78 +/- 4.22 mumol/m2/min2) than in pregnancy (15.17 +/- 2.00 mumol/m2/min2). The metabolic clearance rate, distribution space and pharmacological half-life of insulin were not significantly altered by pregnancy. These findings show that the third trimester of pregnancy is associated with steady state insulin resistance accompanied by a reduced rate of insulin deactivation, but normal insulin pharmacokinetics.     

Basal activity profiles of NPH and [Nɛ-palmitoyl Lys (B29)] human insulins in subjects with IDDM

Summary [N^e-palmitoyl Lys (B29)] human insulin is a fatty acid-acylated derivative of insulin with extended action compared to unmodified insulin when infused intravenously (i. v.) secondary to its binding to circulating albumin. The duration and activity profile of the acylated (A) and NPH (B) insulins were assessed following subcutaneous (s. c.) doses of (A) 6 nmol/kg and (B) 1.2 nmol/kg (equivalent to 0.2 U/kg) in 9 subjects with IDDM. After overnight i.v infusion of regular human insulin, morning glucose was (A) 6.9 ± 0.1 and (B) 6.8 ± 0.1 mmol/l. After the s. c. injection, i. v. human insulin or glucose was infused to maintain near-basal glycaemia and tracer glucose to assess hepatic glucose production (HGP). An activity profile was deduced for each study by expressing the glucose infusion rate at each time point, as a fraction (%) of the basal (measured) HGP, and the i. v. insulin infusion rate as a fraction (%) of the basal requirement. The two fractions are combined by adding the fractional glucose infusion rate and subtracting the fractional insulin infusion rate. Infusion rates of i. v. insulin in the morning were (A) 0.96 ± 0.096 and (B) 1.22 ± 0.09 pmol · kg^–1· min^–1. After insulin injection, i.v insulin requirements decreased and were below 10 % of basal between 100 and 150 min. A constant activity profile of 0 % represents a perfect substitution of the basal i. v. insulin infusion by the s. c. dose. The actual profile is defined by deviations from this (above) and was –17 ± 11, 7 ± 10, –9 ± 6 and –18 ± 18 % for [N^e-palmitoyl Lys (B29)] human insulin and 17 ± 12, 5 ± 6, –9 ± 15, 22 ± 18 % for NPH insulin at 3, 6, 9 and 12 h after s. c. injection. HGP was similar for the two insulins, demonstrating similar metabolic actions and profiles both peripherally and at the liver. [Diabetologia (1998) 41: 116–120] Received: 6 August 1997 and in revised form: 13 October 1997     

Absorption kinetics and action profiles of mixtures of short-and intermediate-acting insulins

Summary The effects of mixing short- and intermediate-acting insulins (lente and NPH) on plasma insulin levels and action profiles, assessed by the euglycaemic clamp technique, were studied in 10 volunteers. Four protocols were used: (1) comparison between two semi-synthetic human soluble insulins in seven subjects (0.22 IU/kg); (2) assessment of insulin levels and action profiles of lente insulin in six subjects and of NPH insulin in five subjects (0.33 IU/kg); (3) comparison between mixtures of soluble with lente insulin and soluble with NPH insulin, administered immediately after mixing, in eight subjects (0.55IU/kg, 40% short-acting); (4) same mixtures, administered 2 days after preparation, in seven subjects. No differences in insulin levels and action profiles during the first 4 h after injection were found between both short-acting insulins and the soluble + NPH insulin mixtures. After the administration of NPH insulin, plasma insulin levels rose slightly faster in comparison with lente insulin, with no significant differences between the action profiles for either insulin. Onset of action was delayed after soluble + lente insulin, both when administered immediately after mixing and to a greater extent when stored for 2 days before administration. After the latter procedure, the onset of action was markedly retarded and only slightly faster than after lente insulin alone.We conclude, therefore, that mixing soluble with NPH insulin in a ratio of 2:3 does not affect the absorption kinetics of soluble insulin, whereas the onset of action is delayed when soluble is combined in the syringe with lente insulin, even when administered immediately after mixing.     

Action profiles of fast onset insulin analogues

Summary Recombinant DNA technology allows the production of insulin analogues with faster absorption rates from subcutaneous tissue as compared to conventional human regular insulin. We report the time-action profiles of 12 U subcutaneously injected insulin analogues (B9Asp + B27Glu or B10Asp) as evaluated against human regular insulin by means of the euglycaemic clamp technique (blood glucose 5.0 mmol/l) in healthy men. After injection of 12 U of either insulin preparation identical values were found for maximal insulin action (maximal glucose infusion rate, time to peak action), total amount of glucose infused as well as area under the curve of glucose infusion rate. Half-maximal glucose infusion rate was reached significantly earlier after injection of modified insulins (mean ± SD 38 ±7 and 43±5 min) as compared to regular insulin (56 ±14 min, p < 0.01). Forty-five min after injection of both insulin analogues glucose infusion rate had increased by 7.4±1.8 or 6.1 ±1.8mg·kg^–·min^–, reflecting 83 ±27 or 67 ±15% of maximal regular insulin action. In conclusion, the two tested insulin analogues showed similar action profiles, but a significantly faster onset of action as compared to regular insulin.     

Absorption of NPH (isophane) insulin in resting diabetic patients: evidence for subcutaneous injection in the thigh as the preferred site.

The absorption kinetics of NPH (isophane) insulin injected subcutaneously into the abdominal wall and subcutaneously (SC) and intramuscularly (IM) into the thigh was studied in 11 Type 1 diabetic patients. The thickness of the subcutaneous adipose tissue layer was measured by ultrasound. NPH (isophane) insulin injected IM into the thigh was absorbed faster than NPH insulin injected SC into the thigh (T50%, IM 8.0 +/- 0.6 h and SC 10.3 +/- 0.7 h, p less than 0.05). No difference in T50% values was found for injection into the abdominal wall (9.7 +/- 1.2h) compared with the thigh. The mean absorption rate from 1.5 to 13.5 h after injection was higher after injection IM into the thigh (6.4 +/- 0.3% of initial dose injected absorbed per h) than after SC injection into the thigh (5.2 +/- 0.3% h-1) and SC into the abdominal wall (5.1 +/- 0.3% h-1) (p less than 0.01). The most constant absorption rate was obtained after SC injection into the thigh (within-study day CV of the mean absorption rate 19.9 +/- 3.2% vs 34.4 +/- 3.2% after IM injection into the thigh and 27.1 +/- 4.9% after SC injection into the abdominal wall (p less than 0.02]. The study provides further evidence that the subcutaneous tissue of the thigh is the preferred injection site for NPH insulin.     

Morphology of palpably abnormal injection sites and effects on absorption of isophane(NPH) insulin

The absorption of isophane(NPH) insulin (Human Insulatard) from palpably abnormal thigh insulin injection sites was determined in 10 C-peptide negative diabetic patients. Absorption was compared with a control day study when the insulin was injected into normal thigh. Standard meals were given 30 and 240 min after the injection. Within 2 weeks the tissue morphology and adipose tissue depths at both injection sites were assessed by ultrasound scanning. Absorption of isophane(NPH) insulin was markedly defective from the abnormal compared with the normal injection sites (area under the free insulin curve to 10 h 115 +/- 15 vs 188 +/- 21 mU l-1 h; p less than 0.01). The area under the blood glucose curve from 270 min to the end of the study at 600 min was significantly greater after injection into the palpably abnormal injection site compared with normal thigh (80.4 +/- 5.2 vs 61.2 +/- 7.0 mmol l-1 h; p less than 0.05) representing a 22% improvement in blood glucose control on the normal injection site afternoon. The depth of abnormal injection site tissue was significantly greater than the depth of adipose tissue at the control site (17 +/- 6 vs 5 +/- 4 mm; p less than 0.001) and considerable disruption of the normal anatomy observed. These results demonstrate defective absorption of isophane(NPH) insulin from palpably abnormal injection sites and describe the morphology of the abnormal tissue.     

Different insulin concentrations in resuspended vs. unsuspended NPH insulin: Practical aspects of subcutaneous injection in patients with diabetes

Aims

This study measured the insulin concentration (Ins_[C]) of NPH insulin in vials and cartridges from different companies after either resuspension (R+) or not (R?; in the clear/cloudy phases of unsuspended NPH).

Study of porcine and human isophane (NPH) insulins in normal subjects

Summary The plasma glucose, C-peptide and insulin responses to subcutaneously administered highly purified porcine, semi-synthetic and biosynthetic human isophane (NPH) insulin and diluting medium as control in normal male subjects were evaluated. Porcine and semi-synthetic human NPH insulins were administered at two dose levels of 0.15 and 0.30 U/kg body weight and biosynthetic human NPH at 0.15 U/kg body weight only. At the low dose level the three insulin preparations resulted in a similar maximal hypoglycaemic effect within 3–5 h after administration. However, over the remainder of the 11 h post-injection period, the plasma glucose level was lower after semi-synthetic human insulin. In contrast, at the 0.30 U/kg dose level, there was no difference in the early or late hypoglycaemic response between porcine and semi-synthetic human NPH insulins of equivalent pharmaceutical formulation. The clinical relevance of these findings needs further evaluation. The data suggest that for the intermediate-acting NPH insulin preparations, both the species of insulin, nature and quantity of the retarding protein and their subsequent interaction may determine their time-action characteristics.     

NPH insulin administration by means of a pen injector

In an open randomized cross-over study 50 patients with insulin-dependent diabetes were allocated to 3 months of treatment with NPH insulin either by means of a pen injector (Insuject-X) or by conventional syringes. The needle of the NPH pen injector was removed immediately after use to avoid possible leakage of solvent. Ambulatory control was assessed every 6 weeks, including blood sampling (HbA1c and insulin antibodies) and recording of hypoglycaemia. NPH insulin containers were collected for insulin potency measurement by HPLC. A seven-point blood glucose profile was performed fortnightly by means of home blood glucose monitoring. At the end of the 6 months a questionnaire was completed. No significant changes occurred in HbA1c (difference 0.1 +/- 0.7 (SD)%), blood glucose profile, or the incidence of hypoglycaemic episodes on the two regimens. The concentration of NPH insulin in the containers remained constant. All but two of the patients preferred to continue to use the pen injector. This NPH pen injector is a reliable and efficacious tool which may also prove more convenient for the patients.     

Comparison of the effects on glucose and lipid metabolism of equipotent doses of insulin detemir and NPH insulin with a 16-h euglycaemic clamp

Aims/hypothesis The association of insulin detemir with non-esterified fatty acid binding sites on albumin may limit its transfer from the circulation into the extravascular extracellular space in adipose tissue and muscle, due to the capillary endothelial cell barrier. In the liver, the open sinusoids may expose hepatocytes to insulin detemir, enabling it to have a greater effect in the liver than in peripheral tissues.Methods We investigated the effects of equipotent doses of insulin detemir and NPH insulin on hepatic glucose rate of appearance (R_a), peripheral glucose rate of disposal (R_d) and glycerol R_a (a measure of lipolysis) using stable isotope techniques. We also investigated the effects of these insulins on NEFA concentrations in seven healthy volunteers during a 16-h euglycaemic clamp. A higher dose of insulin detemir was also studied.Results There was no difference in the glucose infusion profile between insulin detemir and NPH. Insulin detemir had a greater effect on mean suppression of glucose R_a (mean difference 0.24 mg kg^–1 min^–1; CI 0.09–0.39; p<0.01), and minimum glucose R_a, with minimum low dose detemir –0.10±0.15 mg·kg^–1·min^–1 and minimum NPH 0.17±0.10 mg·kg^–1·min^–1 (p<0.02). However, it had a lesser effect on mean suppression of NEFA concentrations (mean difference –0.10 mmol/l; CI –0.03 to –0.17; ANOVA, p<0.02) than NPH. The effect of insulin detemir on glucose R_d and glycerol R_a was not different from NPH. Following high-dose detemir, total glucose infused and maximum glucose R_d were higher (p<0.02, p<0.03) and plasma NEFA concentrations lower (p<0.01) than with low-dose determir.Conclusions/interpretation This study suggests that insulin detemir, when compared to NPH insulin, has a greater effect on the liver than on peripheral tissues and thus has the potential to restore the physiological insulin gradient.     

Effect of storage temperature of insulin on pharmacokinetics and pharmacodynamics of insulin mixtures injected subcutaneously in subjects with Type 1 (insulin-dependent) diabetes mellitus

Summary These studies were undertaken to assess the influence of storage temperature of insulin vials on pharmacokinetics and pharmacodynamics of a mixture of lente insulin (Monotard HM) and regular insulin (Actrapid HM) injected subcutaneously. Seven subjects with Type 1 (insulin-dependent) diabetes mellitus were studied twice after overnight normalization of plasma glucose. A mixture of lente insulin (0.22 U/kg) and regular insulin (0.11 U/kg) was prepared from insulin vials kept either refrigerated (4 °C) or at room temperature (18 °C) and injected subcutaneoulsy (abdomen). Euglycaemia was maintained for the following 16 h by glucose infusion at variable rate. With refrigerated insulin, the plasma free insulin peak was greater (53±5 versus 45±6 mU/l) and occurred earlier (2.5±0.2 versus 6±0.3 h), and the glucose infusion rate showed a greater (16.5±1.2 versus 14.5±0.9 mol·kg^–1·min^–1) and earlier peak (3.2±0.2 versus 6±0.4 h) as compared to that occurring with the non-refrigerated insulin (p<0.05). However, 6 h after insulin injection, both plasma free insulin and glucose infusion rate were 30% lower with the mixture of refrigerated as compared to that of non-refrigerated insulin (p<0.05). In contrast, when NPH-insulin (Protaphane HM) was mixed with regular insulin and injected in 4 out of the 7 diabetic patients, the storage temperature of insulin vials had no effect on the pharmacokinetics and pharmacodynamics of the mixture. Thus, the storage temperature of insulin vials profoundly influences the effects of the mixture lente/regular insulin, but does not affect the pharmacokinetics and pharmacodynamics of the mixture NPH/regular insulin.     

The effect of mixing human soluble and human crystalline zinc-suspension insulin: plasma insulin and blood glucose profiles after subcutaneous injection

The effect of mixing human soluble insulin with two newly developed formulations of human crystalline zinc-suspension insulin was studied in two groups of 8 normal and 6 diabetic subjects. Mixing Human Actrapid with Human Ultratard and Humulin-S with Humulin-Zn, for 60 s before subcutaneous injection, significantly blunted the rise in free insulin levels (p less than 0.05) and the onset of action of the short-acting insulins (p less than 0.01). The loss of solubility that occurs on mixing these insulins, with the consequent loss of the rapid acting component, may diminish their therapeutic usefulness in the control of post-prandial blood glucose.     

A comparison of the activity and disposal of semi-synthetic human insulin and porcine insulin in normal man by the glucose clamp technique

Summary The activity of semi-synthetic human insulin has been compared with porcine insulin in normal man using an euglycaemic glucose clamp at two different insulin infusion rates. In a two hour infusion insulin levels plateaued for both types of insulin at 44–48 mU/l (infusion rate 0.05 U kg body weight^-1 h^-1) and 22–24 mU/l (0.02 U kg^-1 h^-1), giving identical metabolic clearance rates. The glucose delivery required to maintain euglycaemia in the second hour of insulin infusion was 13.9±2.1 g (mean±SEM) and 14.7±1.5 g (NS) at the lower dose for porcine and human insulins respectively, and 27.1±2.5 and 28.0±2.9 g (NS) at the higher dose. The potency ratio for human, compared with porcine, insulin was 1.06 ±0.12. No differences were seen in the time of onset of action of the insulins, serum half-life or distribution space. The responses of blood lactate, pyruvate, alanine, glycerol and 3-hydroxybutyrate were identical. No untoward reactions occurred. The activity and disposal of this semi-synthetic human insulin are indistinguishable from porcine insulin in normal euglycaemic man.     

The variability in the action of unmodified insulin is more dependent on changes in tissue insulin sensitivity than on insulin absorption

Eight normal subjects were studied twice for 360 min after the subcutaneous injection of unmodified insulin (0.15 U kg-1) during euglycaemic clamps. Insulin absorption was assessed by both the area under the insulin-time curve above baseline (AUC) and time course of absorption (time to 25% and 50% of total AUC). Insulin action was measured as the amount of glucose infused. The maximal serum insulin concentration was 0.27 +/- 0.02 (+/- SE) nmol l-1 at 112 +/- 10 min. Fifty percent of total glucose infused occurred at 218 +/- 7 min. The maximal glucose infusion rate was 5.11 +/- 0.70 mg kg-1 min-1 and occurred at 256 +/- 12 min. Intrasubject coefficients of variation (CV) for total insulin AUC (11.2%), time to 25% of maximum AUC (12.1%) and time to 50% of maximum AUC (10.2%) were considerably lower than that for total insulin action (22.6%). Total insulin AUC did not correlate with total glucose utilization (r = 0.06, NS). We conclude that when glucose concentrations are maintained by euglycaemic clamps the peak of unmodified insulin action is later and the duration longer than traditionally recognized, insulin AUC does not predict insulin action, and the higher variability of insulin action compared with the indices of absorption suggests that day-to-day changes in tissue insulin sensitivity contribute more to the variability in insulin action than changes in absorption.     

Similar progression of diabetic retinopathy with insulin glargine and neutral protamine Hagedorn (NPH) insulin in patients with type 2 diabetes: a long-term, randomised, open-label study

Aims/hypothesis  This long-term study was designed to further characterise the retinal safety profile of insulin glargine and human neutral protamine Hagedorn (NPH) insulin in patients with type 2 diabetes mellitus. Methods  An open-label, 5 year, randomised (1:1), multicentre, stratified, parallel-group study conducted in the USA and Canada enrolled individuals with type 2 diabetes and either no or non-proliferative retinopathy (less than severe; Early Treatment Diabetic Retinopathy Study [ETDRS] level less than 53 in both eyes) who were treated with oral hypoglycaemic agents (OHAs) alone, insulin alone or OHAs with insulin for ≥3 months prior to study entry and a baseline HbA_1c level of 6.0–12.0%. Patients were randomised by the investigator according to the centralised interactive voice response system to receive twice-daily NPH insulin (n = 509) or once-daily basal insulin glargine (n = 515). The investigator was not blinded to the treatment group to which each participant had been assigned. The main objective of this study was to compare the progression of diabetic retinopathy between treatment groups by analysing the percentage of patients with three or more step progression in the ETDRS retinopathy patient-level severity scale after treatment with either basal insulin. Masked, centralised grading of seven-field stereoscopic fundus photographs was used. Results  Similarly sustained glycaemic control was observed in both the insulin glargine and NPH insulin treatment groups. Despite a slightly greater severity of diabetic retinopathy for the insulin glargine group at baseline, three or more step progression in ETDRS score from baseline to end-of-study was similar between treatment groups (14.2% [53/374] of insulin glargine-treated patients vs 15.7% [57/363] of NPH-treated patients); the difference in the incidence of progression was −1.98% (95% CI −7.02, 3.06%). Other measures of retinopathy—the development of proliferative diabetic retinopathy and progression to clinically significant macular oedema—occurred to a similar degree in both treatment groups. No other safety issues, such as unexpected adverse events for either insulin emerged during the 5 year study. However, NPH insulin treatment was associated with a higher incidence of severe hypoglycaemia compared with insulin glargine. Conclusions/interpretation  This study shows no evidence of a greater risk of the development or progression of diabetic retinopathy with insulin glargine vs NPH insulin treatment in patients with type 2 diabetes mellitus. Trial registration  ClinicalTrials.gov NCT00174824 Funding  This study was sponsored by sanofi-aventis. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorised users.     

Exercise augments the absorption of isophane (NPH) insulin

In order to determine the effect of exercise on the rate of absorption of an isophane (NPH) insulin, 7 normal men were studied on two separate occasions using the euglycaemic clamp technique. On one day subjects undertook 60 min of exercise on a treadmill (5 km h-1, 5 degrees slope) 180 min after injection of human isophane (NPH) insulin (0.25 U kg-1), while on the control day they remained at rest throughout the 420 min study. Serum insulin concentrations rose in parallel after injection achieving similar values at 180 min on exercise and control days (13.7 +/- 1.2 (+/- SE) vs 12.0 +/- 1.5 mU l-1; NS), respectively. After the onset of exercise, insulin concentrations rose markedly to a peak of 29.8 +/- 2.2 mU l-1 after 50 min of exercise (p less than 0.001), compared with a slight fall on the control day to 10.4 +/- 1.7 mU l-1 (NS). At the end of the exercise period, serum insulin concentrations returned rapidly to control day values. Glucose infusion requirements rose significantly during the exercise period from 2.8 +/- 0.5 mg kg-1 min-1 to a peak at 235 min of 11.1 +/- 1.2 mg kg-1 min-1 (p less than 0.001), compared with an increase on the control day from 2.0 +/- 0.6 to 2.5 +/- 0.6 mg kg-1 min-1 (NS) over the same period. These results demonstrate an enhanced rate of absorption of an isophane (NPH) insulin during exercise in normal subjects.