Insulin Secretion and Insulin Sensitivity in Relation to Fasting Glucose in Healthy Subjects

The study group included 148 nondiabetic women, aged 52–70 years (mean ± SD 62.2 ± 2.3 years), who were recruited from a health screening study in Malmö on the basis of normal glucose tolerance (18). All subjects were healthy, and none were taking any medication known to affect carbohydrate metabolism. None of the subjects had a first-degree relative with type 2 diabetes. All women had glucose levels determined after an overnight fast at three different occasions with at least 1 week in between.

Insulin secretion was determined with the intravenous arginine stimulation at three glucose levels (fasting, 15 mmol/l, and >25 mmol/l) as previously described (15,16). After an overnight fast, intravenous catheters were inserted into antecubital veins in both arms. One arm was used for the infusion of glucose and the other arm for intermittent sampling. The sampling catheter was kept patent by slow infusion of 0.9% saline when not in use. Baseline samples were taken at −5 and −2 min. A maximally stimulating dose of arginine hydrochloride (5 g) was then injected intravenously over 45 s. Samples were taken at 2, 3, 4, and 5 min. A variable rate 20% glucose infusion was then initiated to raise and maintain blood glucose at 15 mmol/l. Blood glucose was determined every 5 min bedside and the glucose infusion adjusted to reach the desired blood glucose level in 20–25 min. New baseline samples were taken, and then arginine (5 g) was again injected and new 2-, 3-, 4-, and 5-min samples were taken. A 2.5-h resting period was allowed, after which new baseline samples were obtained and a high-speed (900 ml/h) 20% glucose infusion during 25–30 min was used to raise blood glucose to >25 mmol/l, as determined bedside. At this blood glucose level, new baseline samples were taken and arginine (5 g) was injected followed by final 2-, 3-, 4-, and 5-min samples.

Insulin sensitivity was determined with the euglycemic-hyperinsulinemic clamp (17). After an overnight fast, intravenous catheters were inserted into antecubital veins in both arms. One arm was used for infusion of glucose and insulin. The contralateral arm was used for intermittent sampling, and the catheter was kept patent with slow infusion of 0.9% saline. A primed constant infusion of insulin (100 units/ml Actrapid; Novo Nordisk, Bagsvaerd, Denmark) with a constant infusion rate of 0.28 nmol · m⁻² body surface · min⁻¹ was started. After 4 min, a variable rate 20% glucose infusion was added and its infusion rate was adjusted manually throughout the clamp procedure to maintain the blood glucose level at 5.0 mmol/l. Blood glucose was determined every 5 min. Samples for analysis of the achieved insulin concentrations were taken at 60 and 120 min.

Blood glucose concentration was determined bedside by the glucose dehydrogenase technique with a Hemocue (Hemocue, Ängelholm, Sweden) during the euglycemic-hyperinsulinemic clamp and with an Accutrend (Boehringer Mannheim Scandinavia, Bromma, Sweden) during the arginine test. Blood samples for insulin and glucose from the arginine study and for insulin from the clamp study were frozen immediately, and serum was frozen at −20°C. Serum insulin was analyzed with a double-antibody radioimmunoassay technique with the use of guinea pig anti-human insulin antibodies, mono-¹²⁵I-tyr-human insulin, and human insulin standard (Linco, St. Charles, MO). Plasma glucose concentrations were analyzed using the glucose oxidase method. All samples were analyzed in duplicate.

Data are presented as means ± SE, unless otherwise stated. The subjects were divided into quartiles for fasting glucose, as determined by the mean of the three measures (n = 37 in each group; range <4.32, 4.33–4.84, 4.85–5.22, and 5.23–6.1 mmol/l). Mean glucose in the four groups was 4.2 ± 0.1, 4.6 ± 0.1, 5.0 ± 0.1, and 5.6 ± 0.1 mmol/l, respectively. For the determination of insulin secretion, the acute insulin response (AIR) to arginine was determined as the mean of the 2- to 5-min samples minus the mean prestimulus insulin concentration at fasting (AIR₁), at 15 mmol/l (AIR₂), and at >28 mmol/l (AIR₃) glucose. Insulin sensitivity was calculated as the glucose infusion rate per kilogram body weight during the 2nd h divided by the mean of the insulin levels at 60 and 120 min during the clamp (i.e., nmol glucose · kg body wt⁻¹ · min⁻¹ per pmol insulin/l). Differences between the four quartiles of fasting glucose were determined by ANOVA with Bonferroni post hoc analysis. Stepwise forward linear multiple regression was used to assess the independent effect of several variables.

Table 1 shows the characteristics of the 148 women who participated in the study and the characteristics of the four quartile subgroups of fasting glucose. Figure 1 shows the distribution of fasting glucose.

When arginine was injected intravenously after an overnight fast, there was a sharp and rapid increase in circulating insulin in all subjects (Fig. 2). Glucose was then administered, which raised plasma glucose to 15 mmol/l. A new arginine administration resulted again in a rapid and marked increase in circulating insulin. Finally, glucose was administered again, which raised plasma glucose to >28 mmol/l. When arginine was injected, there was a marked further increase in insulin levels. AIR₃ is considered the maximal possible insulin response under these conditions (15).

The subjects were divided into four quartiles depending on their fasting glucose. Figure 3 shows the AIR₁, AIR₃, and insulin sensitivity in the respective quartile groups. AIR₁ increased by increasing fasting glucose. AIR₁ in subjects in the two quartile groups with the highest fasting glucose was significantly higher than AIR₁ in the two groups with the lowest fasting glucose (P = 0.004). In the quartile group with the highest fasting glucose (mean 5.6 mmol/l), AIR₁ was 445 ± 47 mmol/l vs. only 285 ± 23 pmol/l in the subjects in the lowest quartile (4.2 mmol/l). It is thus apparent that a glucose threshold for augmenting arginine-stimulated insulin secretion may exist between 4.6 and 5.0 mmol/l, although the data do not allow for a more specific estimation of a threshold. Similarly, the maximal insulin secretion (AIR₃) was significantly higher in the quartile with the highest fasting glucose (P = 0.012). In contrast, AIR₂ did not show a relation to fasting glucose and was not significantly different between the different groups. Insulin sensitivity, as determined by the euglycemic clamp technique, was not different between the groups of the three lowest glucose quartiles but was lower in the subjects with the highest fasting glucose than in the other three groups (P = 0.026).

Insulin secretion negatively correlated to insulin sensitivity. The correlations after logarithmic transformation of the data were for AIR₁: r = −0.60 (P < 0.001), AIR₂: r = −0.46 (P < 0.001), and AIR₃: r = −0.51 (P < 0.001). Figure 3D shows the relation between insulin sensitivity and AIR₃ in the four quartile subgroups of fasting glucose. It is seen that the quartile group with the highest glucose showed a clear increase in insulin secretion in association with the reduced insulin sensitivity.

It has been documented that insulin resistance is adapted by a compensatory increase in insulin secretion (1–6). This concept is of importance for the understanding of the key role of the islet β-cells for the development of impaired glucose tolerance and type 2 diabetes. If the β-cell compensation to insulin resistance fails, glucose homeostasis deranges, which will result in impaired glucose tolerance and type 2 diabetes. In this study, we examined whether β-cell dysfunction and insulin resistance exist in healthy subjects with high, but still normal, fasting glucose. Insulin secretion was evaluated by the glucose-dependent arginine stimulation test and insulin sensitivity was determined from the euglycemic-hyperinsulinemic clamp technique, and the study was undertaken in a large number of healthy postmenopausal women. From the glucose-dependent arginine stimulation test, three different insulin responses to arginine are estimated. The first response (AIR₁) is the insulin response to arginine at fasting glucose. The second response (AIR₂) is the response to arginine when glucose has been raised to 15 mmol/l. Finally, the AIR₃ is the maximal β-cell response when a combined arginine/glucose challenge is given. In accordance with previous reports (6), we found that AIR₁, AIR₂, and AIR₃ were negatively related to insulin sensitivity. Therefore, the islet capacity to respond to exogenous arginine and glucose negatively relate to insulin sensitivity.

It previously has been shown that the variables obtained by the glucose-dependent arginine stimulation test are reduced in subjects with type 2 diabetes (19) as well in those with impaired glucose tolerance (11,20). All subjects in the present study had normal fasting glucose (i.e., <6.1 mmol/l). We found a higher AIR₁ (i.e., the insulin response to intravenous arginine at fasting glucose in the subjects was augmented by higher fasting glucose within the normal limits). The higher AIR₁ in subjects with fasting glucose >∼5 mmol/l shows how sensitive the β-cells are to minute changes in fasting glucose in healthy subjects. An apparent glucose threshold for arginine-stimulated insulin secretion in humans seems to exist between 4.6 and 5.0 mmol/l. A previous study in a small number of individuals showed that subjects with high (but still normal) fasting glucose had a tendency of a lower insulin response to intravenous glucose than subjects with lower fasting glucose (21). This would suggest that the insulin response to intravenous arginine, in contrast to the response to glucose, is preserved when fasting glucose is raised within the normal range (i.e., impaired sensitivity to glucose is an earlier phenomenon during development of type 2 diabetes).

We also found that the insulin response during the glucose-dependent arginine stimulation test at >28 mmol/l glucose is the maximal insulin response (AIR₃) and was increased in subjects in the quartile with the highest fasting glucose. Although not demonstrated experimentally, it is possible that this is due to increased β-cell mass. A potential mediator of this increased AIR₃ would therefore be the higher fasting glucose, since glucose is a powerful stimulator of β-cell growth as shown in experimental animals (rev. in 22).

Since insulin secretion is negatively related to insulin sensitivity, measures of insulin secretion need to be related to insulin sensitivity for a correct judgment of β-cell function. We found that insulin sensitivity was not different between the three groups with the lowest fasting glucose but was lower in subjects in the highest quartile of fasting glucose. This suggests that insulin resistance seems to exist in subjects with fasting glucose >∼5.6 mmol/l. It is thus of interest that both AIR₁ and AIR₃ were augmented in this subgroup compared with the others. Figure 3 illustrates the relation between insulin sensitivity and AIR₃ in the four subgroups with different fasting glucose. Hence, in subjects with low insulin sensitivity, insulin secretion is higher, and this is evident in the subjects with the highest fasting glucose, although these subjects still have fasting glucose levels within the normal range. It is therefore a possibility that the slightly higher glucose in these subjects mediates the higher insulin secretion.

In contrast to the augmented AIR₁ and AIR₃ in the subjects with the highest fasting glucose, there was no significant difference in AIR₂ between the groups. These data might suggest a defective sensitivity of the β-cells for the action of glucose to augment arginine-induced insulin secretion at the level of ∼15 mmol/l. However, this needs to be examined in more detail. A previous study (13) reported from oral glucose tolerance data in obese subjects that β-cell function had deteriorated by 70% in subjects with high, but still normal, 2-h glucose levels. It should be emphasized, however, that the glucose sensitivity of β-cell function during an oral glucose tolerance test includes the release and action of gut incretins and possibly also the action of the autonomic nervous system in addition to β-cell capacity per se, whereas the response to intravenous arginine and glucose is governed by β-cell function. This may explain the severe defect in β-cell response during oral glucose tolerance tests in subjects with high normal glucose, since these subjects might have reduced release of the gut incretin glucagon-like peptide-1, in analogue to reduced glucagon-like peptide-1 response to meal ingestion in subjects with impaired glucose tolerance (22).

In conclusion, this study has shown that reduced insulin sensitivity is seen in subjects with high, but normal, fasting glucose, that the rapid insulin response to intravenous bolus of arginine is dependent on the fasting glucose level with a glucose threshold of ∼4.6 mmol/l, and that in subjects with fasting glucose levels >5.6 mmol/l there is an upregulation of the maximal insulin response to arginine and glucose. Based on these findings, we suggest that basal and maximal insulin secretion is upregulated by reduced insulin sensitivity and that a high, but still normal, fasting glucose may contribute to this regulation.

The study was supported by Swedish Research Council (grant no. 6834), the Swedish Diabetes Association, Region Skåne, and the Faculty of Medicine, Lund University.

B.A. is grateful to Lilian Bengtsson and Margaretha Persson for expert assistance.