Renal threshold for glucose in non-insulin-dependent diabetic patients.

Measurement of glycosuria is still widely used for home monitoring of glycaemia control in non-insulin-dependent diabetes (NIDDM). This method has been criticized because the renal threshold for glucose (RTglu) varies between subjects. In order to evaluate the validity of RTglu by measuring corresponding measurements of blood and urine glucose in NIDDM patients, we studied the blood/urine glucose relationship in 24 NIDDM patients. RTglu estimated from 75 contemporary blood and urine glucose concentrations measured at home by each patient (h-RTglu) was compared with RTglu measured by a hyperglycaemic glucose clamp (c-RTglu). H-RTglu and c-RTglu, being 7.6 mmol/1 (range 5.5-12.4) and 10.3 mmol/1 (6.2-12.3) respectively (P < 0.005), were weakly correlated (R(S) = 0.35, P = 0.15). In conclusion, c-RTglu varies two-fold between NIDDM patients. RTglu detected by home monitored urine and blood glucose determinations underestimates the true RTglu, probably due to the splay phenomenon. However, the method for detection of RTglu used by us seems of clinical relevance, since it reflects the individual blood glucose level at which glucose is detectable in the urine.

Crystallographic studies of the interaction of cyclodextrin glycosyltransferase from Bacillus circulans strain 251 with natural substrates and products.

Asp-229, Glu-257, and Asp-328 constitute the catalytic residues in cyclodextrin glycosyl transferase from Bacillus circulans strain 251. Via site-directed mutagenesis constructed D229N, E257Q, and D328N mutant proteins showed a 4,000-60,000-fold reduction of cyclization activity. A D229N/E257Q double mutant showed a 700,000-fold reduction and was crystallized for use in soaking experiments with alpha-cyclodextrin. Crystal structures were determined of wild type CGTase soaked at elevated pH with alpha-cyclodextrin (resolution, 2.1 A) and maltoheptaose (2.4 A). In addition, structures at cryogenic temperature were solved of the unliganded enzyme (2.2 A) and of the D229N/E257Q mutant after soaking with alpha-cyclodextrin (2.6 A). In the crystals soaked in alpha-cyclodextrin and maltoheptaose, a maltotetraose molecule is observed to bind in the active site. Residue 229 is at hydrogen bonding distance from the C-6 hydroxyl group of the sugar, which after cleavage will contain the new reducing end. In the D229N/E257Q double mutant structure, two alpha-cyclodextrins are observed to replace two maltoses at the E-domain, thus providing structural information on product inhibition via binding to the enzyme's raw starch binding domain.

The effect of acute hyperglycemia on the plasma C-peptide response to intravenous glucagon or to a mixed meal in insulin-dependent diabetes mellitus.

The dose-response relationships between prestimulatory blood glucose concentration and the plasma C-peptide responses to stimulation with 1 mg of glucagon iv or a standard mixed meal were studied in 8 C-peptide positive patients with insulin-dependent diabetes mellitus. Hyperglycemia was maintained for 90 min before stimulation using a hyperglycemic clamp technique. Each test was performed at the steady state blood glucose levels approximately 5, approximately 12, and approximately 20 mmol/l. The glucose potentiation of the incremental plasma C-peptide area under the curve at the two levels of hyperglycemia in percent of the area at normoglycemia (median and range) was 343% (53-1053) (p less than 0.05) and 341% (267-1027) (p less than 0.05) after glucagon and 140% (76-227) (NS) and 251% (95-1700) (p less than 0.05) after the meal. The corresponding relative glucose potentiation of plasma C-peptide 6 min after stimulation with glucagon was 124% (100-427) (p less than 0.02) and 144% (100-209) (p less than 0.05). There was no significant difference in the degree of glucose potentiation at approximately 12 or approximately 20 mmol/l. Furthermore, there was no significant difference in the degree of glucose potentiation of the different estimated values of B-cell function. In conclusion, the plasma C-peptide responses to iv glucagon or to a standard test meal were markedly potentiated by acute hyperglycemia in insulin-dependent diabetes mellitus. No further potentiation was, however, obtained when the prestimulatory blood glucose concentration was raised above 12 mmol/l.(ABSTRACT TRUNCATED AT 250 WORDS)

Sulfonylureas. Why, which, and how?

Although controversies remain as to the usefulness of sulfonylureas, most evidence is in favor of their use in many if not patients with non-insulin-dependent diabetes mellitus. When used properly, sulfonylureas improve insulin secretion and action, and these effects may be maintained for years. If combined with hypocaloric dietary regulation, rapid- and short-acting sulfonylureas may help patients reach and maintain euglycemia without provoking chronic hyperinsulinemia or weight increase. There is no evidence that sulfonylurea treatment causes beta-cell exhaustion; instead, the antihyperglycemic effect helps improve beta-cell function. Sulfonylurea "failures" are often dietary failures or may be due to late introduction of these drugs, i.e., when beta-cell function is already attenuated. Desensitization of the insulinotropic effect of sulfonylureas may occur but might be avoided by discontinuous (less than 24 h/day) sulfonylurea exposure, i.e., once-daily administration of a short-acting sulfonylurea in a moderate dose. The most important adverse effect of sulfonylureas is long-lasting hypoglycemia, which may lead to permanent neurological damage and even death. This is mainly seen in elderly subjects who are exposed to some intercurrent event, e.g., acute energy deprivation or a drug interaction, i.e., aspirin. Long-acting sulfonylureas may be more likely to promote long-lasting hypoglycemia. The dose-response relationships of sulfonylureas have been poorly investigated, and sulfonylurea therapy should always be initiated and maintained at the lowest possible dose.

Acute actions of sulfonylurea drugs during long-term treatment of NIDDM.

The acute effect of sulfonylurea drugs during long-term treatment was evaluated in two separate studies. In the first study, the levels of plasma glucose, insulin, and C-peptide were measured after intake of 10 mg glyburide alone or together with a standardized mixed meal in 10 non-insulin-dependent diabetes mellitus (NIDDM) patients treated with 10-20 mg glyburide/day for greater than 2 yr. There was no acute effect of glyburide on the insulin and C-peptide responses to the meal or during continued fasting, indicating the absence of an acute insulinotropic action of glyburide during chronic treatment. The glucose increase after the meal was also unchanged, but a significant glucose reduction was found after glyburide intake during continued fasting, suggesting a sustained acute extrapancreatic (hepatic) effect. In the second study, the diurnal glycemic excursions in 8 NIDDM patients chronically and continuously (24 h/day) exposed to glipizide (2.5-7.5 mg 3 times/day) were similar when the drug was taken 30 min before each of the three main meals and when taken immediately before the meals. It is concluded that there is no acute insulinotropic action of sulfonylurea drugs during chronic continuous exposure, whereas an acute extrapancreatic action may prevail. During such exposure, there seems to be no clinical benefit in taking a sulfonylurea 30 min before rather than at the start of a meal.

[Non-insulin-dependent diabetes mellitus. Diagnosis and treatment. An interpretation]. / Ikke-insulinkraevende diabetes mellitus. Diagnostik og behanding. En klaringsrapport.

There are approximately 100,000 non-insulin-dependent diabetics in the Danish population and the incidence appears to be increasing. As the disease is complicated by a series of arteriosclerotic manifestations together with hypertension and eye changes, it presents great problems in the primary and also in the secondary sector. The Danish Association for Internal Medicine has, on this basis, prepared an explanatory report with the main object of establishing guidelines for treatment of the disease both medically and by organisation. The report contains a review of the patho-physiological conditions with emphasis on resistance to insulin and the basic cellular defect. The connections with arteriosclerosis and hypertension are also emphasized. The significance of the diet in the treatment of the condition is also emphasized as approximately 80% of the patients are overweight. Guidelines for treatment with sulphonurea, biguanides and insulin are presented. In order to provide optimal treatment and control of non-insulin-dependent diabetics and to provide them a better quality of life closer cooperation between general practitioners and diabetic clinics is recommended. In particular, it appears to be desirable to involve diabetic clinics at the commencement of the disease when the need for training is great.

Metabolic effect of islet B-cell function in insulin-treated diabetes.

We studied the relationship between endogenous insulin secretion and fasting levels of plasma free fatty acids (FFA), plasma acetoacetate plus plasma 3-hydroxybutyrate (total ketone bodies), blood glucose, and HbA1 in 132 diabetic outpatients treated with conventional insulin regimens. Patients were divided into four groups according to plasma C-peptide concentration after intravenous stimulation with glucagon: one group with C-peptide stimulation less than 0.06 nmol/l, one group with C-peptide stimulation 0.06- less than 0.32 nmol/l, one group with C-peptide stimulation 0.32- less than 0.60 nmol/l, and one group with C-peptide stimulation greater than 0.60 nmol/l. According to clinical criteria the prevalence of insulin-dependent diabetes mellitus was approximately 90% in patients with C-peptide stimulation less than 0.32 nmol/l, approximately 25% in patients with C-peptide stimulation from 0.32- less than 0.60 nmol/l, and approximately 10% in patients with C-peptide stimulation greater than 0.60 nmol/l. All metabolic variables were significantly higher in patients without detectable C-peptide in plasma when compared to values found in patients with C-peptide stimulation from 0.06- less than 0.32 nmol/l. These two patient groups also had similar peripheral plasma free insulin levels and were comparable according to age, sex, and body mass index.(ABSTRACT TRUNCATED AT 250 WORDS)

Sensitivity and reproducibility of urinary C-peptide as estimate of islet B-cell function in insulin-treated diabetes.

The aims of the present study were to evaluate the ability of urinary C-peptide determination to demonstrate presence of residual insulin secretion, and to evaluate the reproducibility of urinary C-peptide excretion in 125 insulin-treated diabetic patients. C-peptide was determined in two consecutive 24-h urine specimens and related to plasma C-peptide 6 min after the intravenous injection of 1 mg glucagon. The detection limit of C-peptide in plasma was defined analytically (greater than or equal to 0.02 nmol l-1) and from pancreatectomized patients (greater than or equal to 0.06 nmol l-1), and in urine only analytically (greater than or equal to 0.1 nmol l-1). If the analytical detection limit of plasma C-peptide was used as indicator of residual insulin secretion, islet B-cell function was preserved in all patients. In patients with stimulated plasma C-peptide levels from 0.02- less than 0.06 nmol l-1 no increase was found in plasma C-peptide values after stimulation with glucagon. This unresponsiveness of islet B-cells is in good agreement with the existence of a biological detection limit of C-peptide in plasma of 0.06 nmol l-1. Using this biological plasma C-peptide detection limit, 49 of 125 patients were without residual insulin secretion. In contrast to this, only 7 patients were diagnosed as C-peptide nonsecretors using the analytical detection limit of urinary C-peptide. Eighty-four per cent of patients considered to have Type 1 (insulin-dependent) diabetes with a duration of diabetes of more than 15 years had detectable C-peptide in the urine.(ABSTRACT TRUNCATED AT 250 WORDS)

Fasting plasma C-peptide, glucagon stimulated plasma C-peptide, and urinary C-peptide in relation to clinical type of diabetes.

Many patients with Type 2 (non-insulin-dependent) diabetes mellitus are treated with insulin in order to control hyperglycaemia. We studied fasting plasma C-peptide, glucagon stimulated plasma C-peptide, and 24 h urinary C-peptide in relation to clinical type of diabetes in 132 insulin treated diabetic subjects. Patients were classified clinically as Type 1 (insulin-dependent) diabetic subjects in the presence of at least two of the following criteria: 1) significant ketonuria, 2) insulin treatment started within one year after diagnosis, 3) age of diagnosis less than or equal to 40 years, and 4) weight below 110% of ideal weight of the same age and sex. Eighty patients were classified as Type 1 and 52 as Type 2 diabetic subjects. A second classification of patients into 6 C-peptide classes was then performed. Class I consisted of patients without islet B-cell function. Class II-VI had preserved islet B-cell function and were separated according to the 20%, 40%, 60% and 80% C-peptide percentiles. The two classifications of patients were compared by calculating the prevalence of clinical Type 1 and Type 2 diabetes in each of the C-peptide classes. This analysis showed that patients with a fasting plasma C-peptide value less than 0.20 nmol/l, a glucagon stimulated plasma C-peptide value less than 0.32 nmol/l, and a urinary C-peptide value less than 3.1 nmol/l, or less than 0.54 nmol/mmol creatinine/24 h, or less than 5.4 nmol/24 h mainly were Type 1 diabetic patients; while patients with C-peptide levels above these values mainly were Type 2.(ABSTRACT TRUNCATED AT 250 WORDS)

Sulphonylurea antidiabetic drugs. An update of their clinical pharmacology and rational therapeutic use.

Apart from the amelioration of symptoms, a major aim of the treatment of non-insulin-dependent diabetes mellitus (NIDDM, type 2 diabetes) should be the prevention of cardiovascular complications. These are associated with the chronic hyperglycaemia that is characteristic of NIDDM, and the risk of complications is already increased in subjects with impaired glucose tolerance (IGT). For these reasons, and because hyperglycaemia appears to be a self-perpetuating condition, treatment should be introduced as early as possible and should be aimed at normalisation of blood glucose. To enable early detection and intervention, screening is necessary. As diet regulation alone rarely suffices to normalise blood glucose, addition of sulphonylurea drugs is indicated in many cases. If introduced in the IGT phase, sulphonylureas drugs combined with diet regulation may postpone the development of IGT to manifest NIDDM, and may reduce the increased risk of cardiovascular morbidity and mortality. Sulphonylureas stimulate insulin release, possibly via interaction with receptors in the pancreatic B cells. In addition, such treatment enhances the reduced insulin action. This might be a primary effect but is also a consequence of the increased access to insulin and the subsequent reduction of hyperglycaemia. Sulphonylureas may enhance insulin availability by reducing insulin clearance. Effects on blood lipids are probably secondary phenomena. Fast and short acting sulphonylureas may improve the impaired meal-induced acute insulin release. If combined with weight-reducing diet regulation and introduced early, such treatment can maintain (near) normal blood glucose levels and an improved insulin action for several years without increasing basal insulin secretion, without chronic hyperinsulinaemia, and without weight increase. If not combined with diet regulation, sulphonylurea therapy is likely to fail. If introduced when NIDDM is advanced, the efficacy of these drugs is limited, with secondary failures developing at a rate of 5 to 10% per year. Continuous (24-hour-a-day) exposure to drug treatment could possibly desensitise the B cell to sulphonylurea stimulation. 'Second-generation' sulphonylurea drugs have a higher potency than 'first-generation' drugs, but this need not signify a greater clinical efficacy. The effect of several of these drugs may be increased if they are ingested half an hour before meal(s). Short acting sulphonylureas may be safer than long acting ones, which seem more likely to cause long lasting and fatal hypoglycaemia, at least in elderly patients.(ABSTRACT TRUNCATED AT 400 WORDS)

The beta-cell response to glucagon and mixed meal stimulation in non-insulin dependent diabetes.

The aim of this study was to evaluate the correlations of the C-peptide and insulin responses after stimulation with glucagon intravenously as well as the 24-h urinary excretion of C-peptide to the C-peptide response to a standard mixed meal in 30 patients with non-insulin dependent diabetes mellitus (NIDDM). Fasting plasma C-peptide as well as the C-peptide and insulin responses to glucagon, showed similar but only modest correlations with the C-peptide response to the meal. Urinary C-peptide showed no correlation with the C-peptide response to the meal, but correlated modestly with fasting plasma C-peptide (r = 0.55, p less than 0.01). The C-peptide and insulin responses after meal stimulation correlated modestly inversely with HbA1. In conclusion, measurement of C-peptide in fasting state, as well as measurements of C-peptide and insulin after glucagon stimulation, only modestly predict the C-peptide response to physiologic stimulation in NIDDM. Twenty-four-hour urinary C-peptide excretion does not predict this response. Patients with NIDDM seem to show a better metabolic control if they have a more pronounced beta-cell response to physiologic stimulation.

Classification of newly diagnosed diabetic patients as insulin-requiring or non-insulin-requiring based on clinical and biochemical variables.

In a prospective study of 41 consecutively referred newly diagnosed diabetic patients, we evaluated the predictive value of fasting and glucagon-stimulated C-peptide values, ketonuria, age, and body weight in the classification of subjects as insulin-requiring (IR) or non-insulin-requiring (NIR). The patients were followed up for greater than or equal to 12 mo and classified as NIR if adequate glycemic control could be achieved without insulin (i.e., fasting plasma glucose less than 8 mM and no glycosuria). Patients who needed insulin to obtain this status were classified as IR. We found that all subjects with plasma C-peptide values greater than 0.60 nM 6 min after intravenous glucagon were NIR, whereas all IR subjects together with 3 NIR subjects had C-peptide values below this limit. All NIR subjects but 1 had fasting C-peptide values greater than 0.30 nM, and all IR subjects but 1 had C-peptide values below this limit. Seventy-five percent of the subjects could be correctly classified by use of age and percent desirable body weight. Thus, all subjects greater than 40 yr old and greater than 100% ideal body weight were NIR, and all subjects below both these limits were IR. Ketonuria was found in 10 of 12 IR subjects and in 10 of 29 NIR subjects. We conclude that 1) 75% of the subjects could be correctly classified by use of age and percent desirable body weight only and 2) C-peptide measurements are useful in the classification of newly diagnosed diabetes, whereas presence of ketonuria is of limited value.

Insulin action and insulin secretion in identical twins with MODY. Evidence for defects in both insulin action and secretion.

To evaluate the pathogenetic mechanisms responsible for development of diabetes in the genetically inherited disease maturity-onset diabetes of the young (MODY), we have investigated a pair of identical twins (19 yr old) from a MODY family. One twin had nondiabetic fasting plasma glucose values but impaired glucose tolerance (IGT), whereas the other suffered from frank diabetes (fasting plasma glucose 12.5 mM). Differences in insulin secretion pattern and/or insulin action between the twins is supposed to be responsible for development of hyperglycemia in MODY. On the other hand, identical defects in insulin secretion and action in the twins may point to the primary genetic defect in MODY. Therefore, our aim was to investigate insulin secretion and insulin action in the twins to find these differences and similarities. We found that fasting plasma insulin and C-peptide values were slightly increased in the twins, whereas the responses of insulin and C-peptide to oral glucose tolerance tests (OGTT) and meals were similar in the twins and within normal range. The insulin responses to OGTT were, however, lower than expected from the glucose values, indicating a beta-cell defect. Despite elevated plasma insulin levels, basal hepatic glucose output (HGO) was normal in the IGT twin but increased by 75% in the diabetic twin. The maximally inhibitory effect of insulin on HGO, when estimated at euglycemia, was normal in the IGT twin but reduced by 60% in the diabetic twin. Furthermore, the maximal insulin-mediated glucose uptake in peripheral tissues was reduced by 40% in the diabetic twin.(ABSTRACT TRUNCATED AT 250 WORDS)

Insulin requirement in non-insulin-dependent diabetes mellitus: relation to simple tests of islet B-cell function and insulin sensitivity.

Evaluation of simple tests of islet B-cell function and insulin sensitivity as predictors of metabolic control was performed during 3 months of insulin withdrawal in 25 insulin-treated diabetic subjects. All patients had a glucagon stimulated plasma C-peptide concentration above 0.33 nmol/l and a fasting plasma C-peptide concentration above 0.20 nmol/l a few days before insulin withdrawal. Insulin sensitivity was measured as the glucose disappearance rate (k) during an intravenous insulin tolerance test. Two patients were considered insulin-requiring due to high fasting blood glucose levels (greater than 20 mmol/l) and two patients due to an increase in glycosylated haemoglobin of more than 1.1% (greater than approximately 3SD) in combination with weight loss. None of the remaining patients had a significant increase in glycosylated haemoglobin. An inverse correlation was found between stimulated C-peptide levels and insulin sensitivity (r = 0.41, p less than 0.05). Fasting and stimulated C-peptide concentrations of 0.40 and 0.70 nmol/l, respectively, separated non-insulin-requiring patients from a group consisting of both insulin- and non-insulin-requiring patients. At these C-peptide levels the predictive value of a positive test was 100% while the predictive value of a negative test was as low as 33% or 27% depending on whether fasting or stimulated C-peptide concentration was used. Including the k value in the prediction only increased the predictive values of negative tests to 40% and 33%, respectively.

Interaction of ethanol and glipizide in humans.

The hypoglycemic effect of 2.5 mg glipizide and the potentiation of this effect by ethanol were studied in 10 normal-weight nondiabetic subjects. The reductions in blood glucose concentrations were similar in time of onset and extent (2 mM) whether glipizide was taken alone or in combination with ethanol. However, the return of blood glucose toward fasting level was delayed by ethanol. Beta-Cell secretory activity, evaluated from the concentrations of insulin and C-peptide, was unchanged by ethanol. The serum glipizide concentrations were reproducible within subjects, whereas there was a considerable interindividual variation. This heterogeneity in the rise in glipizide concentration was strongly correlated with blood glucose fall and insulin secretion. Thus, ethanol can prolong but does not augment the hypoglycemia induced by glipizide. The heterogeneity in glipizide concentration seems to be caused by an interindividual variation in kinetics.

Reproducibility of beta-cell function estimates in non-insulin-dependent diabetes mellitus.

We evaluated the reproducibility of different estimates of endogenous insulin secretion in 30 patients with non-insulin-dependent diabetes mellitus (NIDDM). Fasting blood glucose concentration was similar on the 2 days of study. The coefficients of variation of fasting plasma C-peptide, plasma C-peptide 6 min after the injection of 1 mg i.v. glucagon, and the increment in plasma C-peptide after glucagon were 16.0, 14.8 and 24.1%, respectively. The coefficients of variation of the corresponding plasma insulin values were 19.2, 24.8, and 34.8%, respectively. The coefficient of variation of 24-h urinary C-peptide excretion was 22.1%. Because fasting plasma C-peptide correlated closely with plasma C-peptide 6 min after glucagon (test 1: r = .70, P less than .01; test 2: r = .76, P less than .01), it seems that these two values can be used equally well as assessment of beta-cell function in NIDDM. In conclusion, fasting plasma insulin, fasting plasma C-peptide, and plasma C-peptide 6 min after glucagon stimulation showed a similar and acceptable degree of reproducibility. Plasma insulin 6 min after glucagon and increments in plasma insulin and C-peptide, as well as urinary C-peptide, seem to be less reproducible.

Correlations between fasting plasma C-peptide, glucagon-stimulated plasma C-peptide, and urinary C-peptide in insulin-treated diabetics.

This study correlated fasting plasma C-peptide (CP), plasma CP 6 min after stimulation with 1 mg glucagon i.v., and the mean of three 24-h urinary excretions of C-peptide (UCP)/creatinine in 132 insulin-treated diabetics. Patients were divided into three groups: group 1, stimulated CP less than 0.06 nM (n = 51); group 2, stimulated CP 0.06-0.60 nM (n = 48); and group 3, stimulated CP greater than 0.60 nM (n = 33). In all patients fasting CP was closely correlated to stimulated CP (r = .988, P less than .001), whereas the correlations between UCP and both fasting CP (r = .904, P less than .001) and stimulated CP r = .902, P less than .001) were slightly less pronounced. The associations between UCP and both fasting CP (r = .716, P less than .001) and stimulated CP (r = .731, P less than .001) were modest in group 2, and even more so in group 3 (r = .557, P less than .001 and r = .641, P less than .001, respectively). In conclusion, fasting CP is closely correlated to glucagon-stimulated CP in insulin-treated diabetics and can probably be used equally well in the assessment of beta-cell function. The associations between UCP and both fasting and glucagon-stimulated CP are less pronounced, especially in patients with well-preserved beta-cell function.

Prevalence of fasting hyperglycemia and known non-insulin-dependent diabetes mellitus classified by plasma C-peptide: Fredericia survey of subjects 60-74 yr old.

A Danish population of 5699 individuals (60-74 yr old) was screened by fasting blood glucose (FBG) and interviewed about known diabetes. The distribution of FBG in individuals not known to have diabetes showed no sex difference or significant variation with age. Fasting hyperglycemia (FH), defined as FBG greater than or equal to mM in subjects without a history of diabetes, was found in 1.7% of men and women. Known diabetes (KD) had a prevalence of 3.9 and 5.0% in men and women, respectively. The prevalence rates of FH and KD increased significantly with age. In the two subgroups, plasma C-peptide was measured after overnight fasting and subsequently 6 min after an intravenous injection of glucagon. Based on the distribution of the C-peptide concentrations in non-insulin-treated KD subjects, lower limits for non-insulin-dependent diabetes mellitus (NIDDM) of 0.30 pmol/ml for fasting C-peptide and 0.60 pmol/ml for stimulated C-peptide were arbitrarily chosen. According to these cutoff points, only 38.5% of KD subjects treated with insulin had insulin-dependent diabetes mellitus, corresponding to 9.3% of all KD subjects. After exclusion of these patients, the prevalence of recognized NIDDM was 3.5% in men and 4.5% in women. All FH subjects except one had C-peptide values in the NIDDM interval. A close agreement between fasting and glucagon-stimulated C-peptide was seen. In epidemiological studies with an expected high prevalence of NIDDM, we propose to use fasting C-peptide for classification of patients with insulin-treated diabetes.