Suppression of Endogenous Insulin Secretion by Euglycemic Hyperinsulinemia

Abstract

Context

The impact of insulin, particularly exogenous hyperinsulinemia, on insulin secretion in humans is debated.

Objective

We assessed the effects of exogenous hyperinsulinemia on insulin secretion and whether the response is altered in insulin resistance associated with obesity.

Methods

Insulin secretion rates (ISRs) during euglycemic hyperinsulinemic clamp studies (52 volunteers) were calculated using a model that employs plasma C-peptide concentrations. One study involved a 2-step insulin clamp and the other study was a single step insulin clamp. For both studies the goal was to achieve plasma glucose concentrations of 95 mg/dL during the clamp irrespective of fasting glucose concentrations. The percent change in ISR from fasting to the end of the insulin clamp interval was the main outcome. Linear regression and analysis of covariance were used to test for the effects of insulin on ISR and to test for group differences.

Results

ISR was greater in obese volunteers (P < .001) under fasting and hyperinsulinemic clamp conditions. The change in plasma glucose from baseline to the end of the insulin clamp interval was highly correlated with the change in ISR (r = 0.61, P < .001). From baseline to the end of the clamp we observed a 27% (SD 20) suppression of ISR. The participants who underwent a 2-step insulin clamp had greater suppression of ISR during the second step than the first step (P < .001). The proportional suppression of ISR during euglycemic hyperinsulinemia was not different between nonobese and obese groups (P = .19).

Conclusion

Hyperinsulinemia suppresses endogenous insulin secretion and the relative change in insulin secretion produced by exogenous insulin did not differ between nonobese and obese people.

Peripheral insulin resistance with its compensatory hyperinsulinemia can result in beta cell exhaustion and inadequate insulin secretion in the long-term (1). Although plasma glucose concentration is the primary driver of insulin secretion, there are several other proposed regulators (2). Beta cells have been found to insulin receptors that, when activated by insulin, change intracellular calcium concentrations, which regulates insulin secretion (3). However, the direction of the net effect of insulin stimulation on beta cells is inconclusive. While some investigators reported that insulin stimulated insulin secretion (3, 4) others described inhibition of beta cells (5, 6). In an animal model study, portal vein insulin concentrations remained stable in the face of increased arterial insulin concentrations induced by an exogenous insulin infusion despite, implying suppressed pancreatic insulin secretion (7). In another animal model study, disruption of the beta cell insulin receptors influenced glucose sensing and insulin production (8). However, there is no consensus on the effect of hyperinsulinemia on insulin secretion in humans (9-17).

Excess body fat can lead to insulin resistance, requiring greater insulin secretion to achieve euglycemia. Studies of the effects of exogenous hyperinsulinemia on beta cell function in vivo have shown no effect, suppression or stimulation of insulin secretion in obese adults (11-15). If the beta cells are resistant to the effects of insulin on its own secretion in obesity, this could explain altered islet responsiveness to stimuli that regulate insulin responses. Therefore, to understand the differences in insulin secretion in obesity, it is important to study the behavior of beta cells during hyperinsulinemia in this population.

To address this issue, we examined the effects of exogenous hyperinsulinemia on insulin secretion and whether the response is altered by obesity. Our hypotheses were (1) hyperinsulinemia suppresses insulin secretion; (2) obesity is associated with beta cell insulin resistance to hyperinsulinemia (eg, endogenous insulin secretion will be less suppressed under hyperinsulinemic conditions in obese than lean adults).

Material and Methods

The samples used to evaluate insulin secretion were from 2 different research protocols designed to study free fatty acid metabolism and were approved by the Mayo Clinic Institutional Review Board. The studies took place in the Clinical Research and Trials Unit (CRTU) at Mayo Clinic, Rochester, MN. All participants provided written informed consent, and the studies were conducted in accordance with the 1964 Declaration of Helsinki.

Clinical Protocol

Premenopausal women and age-matched men aged 18 to 55 years with stable weight (±1.5 kg) for at least 2 months before the study were included. Exclusion criteria included diabetes mellitus, hepatic or renal dysfunction, systemic inflammation, and the use of medications affecting insulin secretion, glucose, or lipid metabolism. In addition to standard anthropometric measures, fat-free mass, and total and regional (including visceral) fat mass were measured using a single-slice abdominal computed tomography image at the L2-3 interspace and dual energy x-ray absorptiometry (18). The body mass index criteria for nonobese and obese groups were 20 to 27 kg/m² and 30 to 37 kg/m², respectively.

All participants received an isoenergetic diet (45% carbohydrate, 20% protein, and 35% fat) from the Mayo Clinic CRTU Metabolic Kitchen for 3 days before the study to assure standard energy and macronutrient intake (19). The evening before the study, participants were admitted to the CRTU, and an 18-gauge catheter was inserted into the forearm vein for infusions. Next morning, baseline blood samples were collected from an intravenous catheter placed in a retrograde fashion in a hand vein using the heated box technique. After fasting blood samples were collected, the volunteers received a primed, continuous insulin infusion according to their clamp scheme together with 50% dextrose infusion to maintain euglycemia. In Clamp A, all volunteers received the similar 1-step protocol (regular insulin infusion rate: 1 mIU*kg⁻¹*min⁻¹). The participants in Clamp B had 2 tandem steps of insulin infusions, as low dose (normal weight volunteers 0.25, and obese volunteers 0.5 mIU*kg⁻¹*min⁻¹) and high dose (normal weight volunteers 1, and obese volunteers 1.5 mIU*kg⁻¹*min⁻¹). Steady-state blood samples were collected for each step (Fig. 1). The goal plasma glucose for all participants was 95 mg/dL irrespective of the observed fasting plasma glucose (eg, we adjusted the glucose infusion rate to attempt to achieve that goal in volunteers with fasting plasma glucose concentrations above or below 95 mg/dL).

Figure 1.

Fifty-two (16 nonobese, 36 obese) participants were recruited. (A) Twenty-six participants (9 nonobese, 17 obese) underwent a single step hyperinsulinemic euglycemic clamp. (B) Twenty-six participants (7 nonobese, 19 obese) were recruited to a 2-step hyperinsulinemic euglycemic clamp.

Laboratory Measurements

Bedside plasma glucose concentrations were measured using the glucose oxidase method (Yellow Springs Instruments, OH, USA) at least every 10 minutes during euglycemic hyperinsulinemic clamp. Serum C-peptide concentrations were measured by electrochemiluminescence immunoassay method (Cobas e411, Roche Diagnostics, IN, USA) (RRID:AB_2909476). Plasma insulin concentrations were determined using an immunoassay autoanalyzer system (Unicel DxI 800, Beckman Coulter Inc., CA, USA) (RRID:AB_2756878).

Calculations and Statistical Analyses

Combined hepatic and muscle insulin sensitivity (Si) was calculated by dividing the steady-state glucose infusion rate (mg/kg fat-free mass ) by the increase from baseline in plasma insulin concentrations observed at the end of clamp.

Insulin secretion rates (ISRs) were calculated with the steady-state balance equation of ISR = C-peptide clearance • C-peptide concentration as previously suggested (20). Population-based formulas were used for C-peptide kinetics (21), and units were calculated using conversion factors (22).

Metabolic clearance rate of insulin (MCRI) was calculated by the formula of MCRI = ISR/plasma insulin concentration.

Variables are expressed as mean (SD) and median (interquartile range) based on normality. Comparisons between groups were performed with independent t-tests or Mann–Whitney U tests. Plasma insulin concentrations were log transformed to obtain a normal distribution for regression analyses, which were done using Pearson tests. Change in ISR from fasting to end clamp was described with percent change in ISR. Comparison of linear regression parameters between groups was performed by analysis of covariance. Probability values were 2-sided and considered statistically significant when P < .05. Statistical analyses were performed using BlueSky Statistics v7.10 (BlueSky Statistics LLC, Chicago, IL, USA). GraphPad Prism 9.5.1 and Biorender.com were used for visualization of the results and protocols.

Results

Samples from 52 (33 female, 19 male) participants from 2 different euglycemic hyperinsulinemic clamp studies were used to measure C-peptide concentrations at time points selected for stable glucose concentrations. In the obese group there were 3 volunteers with prediabetes and 2 volunteers with history of gestational diabetes. The subject characteristics, baseline and insulin clamp data are provided in Table 1. MCRI values were not different between obese and nonobese participants (P = .09).

Table 1.

Baseline and euglycemic hyperinsulinemic clamp characteristics of the participants

	Nonobese (n = 16)	Obese (n = 36)	P
Female/Male	10/6	23/13	.92
Age, years	30 (24-35)	36 (26-43)	.20
BMI (kg/m2)	23.5 (22.6-24.9)	33.5 (31.2-35.6)
Body fat (%)	31(25-35)	46 (40-48)	<.001
Visceral fat mass (kg)	0.9 (0.6-1.6)	4.4 (3-5.9)	<.001
Abdominal SC fat mass (kg)	9.5 (8.6-10.8)	22.6 (19.3-24.7)	<.001
Mean fasting plasma glucose (mg/dL)	87 (6)	91 (5)	.017
Mean end-clamp plasma glucose (mg/dL)	91 (6)	92 (5)	.61
Mean Δglucose (mg/dL)	4 (7)	1 (7)	.13
Fasting plasma insulin (mIU/L)	4.1 (3.6-5.3)	10.3 (6.6-12.9)	<.001
End-clamp plasma insulin (mIU/L)	45.7 (43.1-53.4)	104 (79.6-122.5)	<.001
Fasting C-peptide (nmol/L)	0.52 (0.15)	0.96 (0.30)	<.001
End clamp C-peptide (nmol/L)	0.38 (0.29-0.45)	0.71(0.44-0.95)	<.001
Si (mg/kg FFM)	0.29 (0.26-0.34)	0.11 (0.09-0.14)	<.001
MCRI (L/min)	4.98 (4.76-5.29)	4.35 (3.71-5.05)	.09
Mean fasting ISR (mIU/min)	21.3 (6.4)	45.3 (14.5)	<.001
End clamp ISR	16.4 (12.1-19.2)	35.2 (21.1-42.5)	<.001
Mean ΔISR(%)	(−22) (19)	(−30) (20)	.19

Because BMI is not a random variable these values are not subject to statistical testing. The P value for the first row refers to the proportional representation of males and females between the non-obese and obese volunteer groups.

Abbreviations: Δglucose, delta plasma glucose; ΔISR(%), delta percentage of insulin secretion rate; BMI, body mass index; FFM, fat-free mass; ISR, insulin secretion rate; MCRI, metabolic clearance rate of insulin; SC, subcutaneous; Si, insulin sensitivity.

The change in plasma glucose concentrations (Δglucose) from prior to the insulin clamp to the end of the clamp was 2 (SD 7) mg/dL. There were no differences in Δglucose between obese and nonobese participants (P = .13, Fig. 2A).

Figure 2.

Fasting and end-clamp values of (A) plasma glucose concentrations (mg/dL) (B) plasma insulin concentrations (mIU/L) (C) calculated insulin secretion rates (mIU/min) in nonobese and obese participants.

Mean fasting plasma C-peptide concentrations were 0.83 (SD 0.33) and the mean end clamp C-peptide concentrations were 0.6 (SD 0.3) nmol/L. The achieved plasma insulin and C-peptide concentrations, as well as ISR at the end of insulin clamp were greater in obese than in nonobese participants (P < .001 for all).

From baseline to the end of the euglycemic, hyperinsulinemic clamp, a 27% (SD 20) suppression of ISR was found with no difference between sexes (P = .57). There was a similar degree of suppression of C-peptide concentrations from beginning to end of clamp. The proportional suppression of ISR during euglycemic hyperinsulinemia was not different between nonobese and obese groups (P = .19).

We used correlation analyses to test for factors statistically linked to the change in ISR. The percentage change in ISR was correlated only with the change of plasma glucose concentrations from baseline to the clamp interval during which the measures were made (r = 0.61, P < .001) (Fig. 3A). By simple regression analyses, the achieved plasma insulin concentrations and Si were not significantly correlated (P = .86 and .34, respectively).

Figure 3.

(A) Regression analysis between percentage of delta insulin secretion rate (ΔISR%) and delta plasma glucose (Δglucose) (mg/dL), and covariance analyses of this correlation. ΔISR%= (−30.5) + (1.7 • Δglucose). (B) between low [ΔISR%= (−13.3) + (2.8 • Δglucose)] and high dose [ΔISR%= (−25.7) + (1.9 • Δglucose)] insulin infusion steps of Clamp B. (C) between nonobese and obese participants.

Because plasma glucose concentration was the only factor that we found to be correlated with the ISR changes, and because even small changes had a strong influence on ISR, we used regression analysis to examine glucose independent ISR during hyperinsulinemia. The model predicted a 31% suppression of ISR during hyperinsulinemia with a change in glucose concentrations of 0 mg/dL (Fig. 3A). The ISR at the end of the high insulin dose interval was suppressed compared with the ISR at the end of the low insulin dose interval (P < .001, Fig. 3B) in participants who underwent the 2-step insulin clamp (Fig. 3B) after controlling for the confounding effects of glucose concentrations (analysis of covariance). The relationship between ISR and Δglucose was also similar between nonobese and obese groups (analysis of covariance, Fig. 3C).

Discussion

We tested the hypothesis that the suppression of endogenous insulin secretion by exogenous insulin is reduced in obese compared with nonobese people. Because we found that insulin secretion was remarkably sensitive to small changes in plasma glucose concentrations, we needed to test this hypothesis by assessing the effects of exogenous insulin on ISR, independent of changes in plasma glucose concentrations. We found that ISR was suppressed during an insulin infusion when plasma insulin concentrations were intentionally increased and plasma glucose remained within the euglycemic range. Although adults with obesity and metabolic impairments had greater insulin secretion during hyperinsulinemia, the proportional suppression was not different from nonobese participants.

Whether exogenous hyperinsulinemia suppresses endogenous secretion is controversial; some authors reported no change in secretion (9, 10, 16, 17, 23, 24), whereas others described suppression during hyperinsulinemia (11-15, 25-29). These different results may arise from variations in other modifiers of insulin secretion. Our findings are consistent with those who report suppression of ISR in response to exogenous insulin administration. Although some investigators have reported the association between changes in insulin secretion and plasma glucose during hyperinsulinemic clamps (26, 30), to our knowledge, this is the first study to define that hyperinsulinemia suppresses insulin secretion independent of changes in plasma glucose.

Plasma insulin concentrations remained greater in the obese than nonobese volunteers at the end of the hyperinsulinemic clamp studies. While part of the explanation for the greater hyperinsulinemia might be a result of higher insulin infusion rates, the greater ISR in adults with obesity also contributes. Previous investigators have reported that, in response to insulin infusions, there is greater suppression of insulin secretion in obese (11) or lean people (12-15), we observed similar proportional suppression of ISR in both groups after adjusting of plasma glucose changes. This argues against insulin resistance with regards to ISR of beta cells in obesity.

We found that plasma insulin concentrations did not predict any of the variance in suppression of insulin secretion, which is in the line with previous reports (25, 30). It might be that the full ability of exogenous insulin to suppress endogenous insulin secretion occurs at relatively low insulin concentrations, in which case we would not observe a greater suppression in those with higher insulin concentrations. However, this would not explain why we observed additional, glucose-independent suppression of ISR during the second, higher insulin infusion rate in those volunteers undergoing a 2-step insulin clamp. The disconnect between these 2 observations may be related to the time of exposure to exogenous insulin. If there is a delay in the time to achieved suppression of ISR by exogenous insulin, the decrease in C-peptide concentrations reflective of the suppression would be delayed due to its long half-life. Thus, the greater suppression of ISR with the second, increased insulin infusion rates in volunteers who underwent the 2-step hyperinsulinemic clamp could be related to the duration, not the dose of exogenous insulin.

The main strength of this study is the quantitative framework for insulin secretion during exogenous hyperinsulinemia. Our methodology enabled us to account for fluctuations in insulin secretions that were attributable to changes in plasma glucose concentrations. A limitation in our study was the goal to achieve similar plasma glucose concentrations in all participants rather than maintaining their original fasting glucose concentrations. Although we were able to account for the effects of plasma glucose changes on ISR, an ideal study for understanding the effect of hyperinsulinemia would be to sustain glucose concentrations (within 1 mg/dL) to those found under fasting conditions for each volunteer. However, maintaining this degree of stability is almost technically impossible for most insulin clamp studies, especially with higher insulin doses. Secondly, the calculations of ISR rely on steady-state assumptions in our study. Although the total duration of both clamp protocols is more than 7 times the half-life of C-peptide (31), the first step of Clamp B had a shorter duration. Therefore, when we compare different steps of Clamp B, we cannot rule out the likelihood that the longer duration of exogenous hyperinsulinemia could have led to more suppression. Nonetheless, our findings suggest a new avenue for future research.

Conclusion

Hyperinsulinemia suppresses endogenous insulin secretion, independent of plasma glucose concentration. Slight changes in blood glucose concentrations drive changes in insulin secretion even during “euglycemia.” Adults with obesity have greater insulin secretion during hyperinsulinemia, which correlates with insulin resistance and excessive body fat. The percentage of ISR suppression in obese individuals has similar characteristics as in nonobese individuals. Future euglycemic hyperinsulinemic clamp studies with different durations and the design of achieving isoglycemia will provide a more definitive picture of the extent to which hyperinsulinemia suppresses pancreatic beta cell secretion.

Abbreviations

ISR

insulin secretion rate

MCRI

metabolic clearance rate of insulin

Si

insulin sensitivity

Funding

This work was supported by National Institutes of Health (Grant Numbers: 5T32DK007352-44, DK40484 and DK45343).

Author Contributions

N.D.K.: Conceptualization, methodology, data collection, analysis, interpretation, writing—original draft. Y.S.: methodology, data collection, writing—review and editing. M.C.L.: methodology, data collection, analysis, interpretation, writing—review and editing. A.V.: Conceptualization, methodology, interpretation, writing—review and editing. M.D.J.: Conceptualization, Validation, Methodology, Supervision, Funding acquisition, Writing—review and editing.

Disclosures

The authors declare that no competing interests exist.

Data Availability

The datasets generated and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.