Physiology of Glycemic Recovery and Stabilization After Hyperinsulinemic Euglycemic Clamp in Healthy Subjects

Abstract

Background

The hyperinsulinemic euglycemic clamp (HEC) is the gold standard for measuring insulin sensitivity, but glycemic recovery and stabilization after the procedure have not been well studied. Here, we assessed the physiological determinants of postclamp recovery.

Methods

We analyzed data from 207 healthy subjects [102 African American (AA) and 105 European American (EA)] who underwent HEC in the Pathobiology of Prediabetes in a Biracial Cohort study. At the end of HEC, insulin infusion was stopped, and dextrose (20%) infusion was tapered and stopped when plasma glucose stabilized ≥20 mg/dL above the preclamp value (∼100 mg/dL). Glucose recovery time (GRT) was defined as the interval from cessation of insulin infusion to discontinuation of dextrose infusion. Insulin clearance was calculated under basal and clamp conditions.

Results

The mean (± SD) age and body mass index were 46.3 ± 9.96 years and 30.7 ± 8.43 kg/m², respectively. Plasma glucose (mg/dL) was 92.2 ± 6.26 preclamp and 124.2 ± 26.9 postclamp. The median GRT (minutes) was 65 (range, 30 to 270); mean GRT was 77.1 ± 42.7 (men: 82.9 ± 45.5; women: 74.4 ± 42.3; AA, 82.0 ± 49.6; EA, 72.3 ± 34.2; P > 0.1 for sex or race). The 90th percentile for GRT was 119 minutes. In regression models, significant predictors of GRT were age (P = 0.03), weight (P = 0.009), 2-hour plasma glucose (P = 0.0002), insulin sensitivity (P = 0.03), disposition index (P = 0.017), and basal insulin clearance (P = 0.02).

Conclusions

In our biracial cohort, glycemic recovery after hyperinsulinemic clamp was independent of sex or race but was significantly predicted by age, weight, and glucose tolerance and by insulin sensitivity, secretion, and clearance. We recommend that monitoring be maintained for ∼2 hours postclamp to ensure adequate glycemic stabilization.

After hyperinsulinemic euglycemic clamp, ∼50% of subjects achieve glycemic stability within 1 hour; glucose tolerance, insulin sensitivity, and secretion are major determinants of glycemic recovery.

Decreased insulin sensitivity (or insulin resistance) is a feature of a broad range of cardiometabolic disorders, including obesity, hypertension, polycystic ovary disease, and type 2 diabetes (1–3). Although several direct and indirect measures of insulin sensitivity have been described and are used widely in clinical research, the hyperinsulinemic euglycemic clamp (HEC) procedure (also referred to as the clamp method) has been the gold standard methodology for the direct assessment of insulin sensitivity in humans (4–8). The clamp method involves a constant intravenous infusion of insulin to achieve a supraphysiological plasma insulin concentration along with a variable-rate infusion of dextrose for ∼180 minutes. The rate of dextrose infusion is adjusted to maintain a euglycemic target (usually ∼100 mg/dL or 5.5 mmol/L) using frequent (every 5 to 10 minutes) bedside plasma glucose measurements. The supraphysiological levels of plasma insulin suppress hepatic glucose production, pancreatic islet insulin secretion, lipolysis, and proteolysis (9). These anabolic effects of hyperinsulinemia prevent the introduction of new endogenous glucose molecules through glycolysis or gluconeogenesis, thereby allowing the clamp operator to manipulate ambient plasma glucose levels by adjusting the dextrose infusion rate.

The formal clamp procedure ends after glucose infusion rates have been recorded at the end of the steady-state period (typically during the last 60 minutes). Thereafter, the insulin infusion is discontinued, and research subjects are offered a snack or meal. However, the dextrose infusion is continued, with monitoring of bedside glucose levels, until the operator determines that the study subject is stable enough for discontinuation of dextrose support and discharge from the research unit. Clinical research experience indicates considerable individual variation in the time required for postclamp monitoring and subsequent decision-making regarding cessation of dextrose support and final discharge. Factors that can potentially contribute to the variations in postclamp glucose recovery and stabilization times include continued glucose transport and metabolism, mediated by insulin molecules that were receptor-bound before cessation of insulin infusion; content of snack or meal ingested immediately after the clamp procedure; and individual differences in gastrointestinal absorption of ingested nutrients.

Premature discharge from the research unit after a hyperinsulinemic clamp procedure could expose subjects to the risk of hypoglycemia, especially after ambulation. Because most clamp procedures are performed during outpatient visits to the research center, many test subjects may drive automobiles or operate machinery upon return to work. These risk considerations are not merely theoretical: the occasional subject with postclamp hypoglycemia requiring prolonged dextrose infusion would be familiar to experienced researchers. However, despite the wide use of the HEC in clinical research, the subject of postclamp glycemic recovery has received little attention. To the best of our knowledge, there have been no prior studies specifically addressing that subject. Thus, our goal was to explore the roles of demographic, clinical, physiological, and behavioral variables as potential determinants of glucose recovery and stabilization after the HEC procedure. Specifically, we hypothesized that individual variations in insulin action and insulin clearance affect postclamp glucose recovery time (GRT).

Materials and Methods

Participants

We studied 207 subjects with evaluable data [age, 18 to 65 years; 102 African American (AA), 105 European American (EA); 142 women, 65 men] enrolled in the Pathobiology of Prediabetes in a Biracial Cohort study (10). All participants were healthy, normoglycemic individuals whose biological mother and/or father had a history of type 2 diabetes. At baseline, each participant underwent a 75-g oral glucose tolerance test (OGTT) and had normal fasting plasma glucose (FPG) (<100 mg/dL, 5.6 mmol/L) and/or normal glucose tolerance [2-hour plasma glucose (2hrPG) <140 mg/dL (7.8 mmol/L)], as previously described (10). Individuals with a known history of diabetes and those taking antidiabetic medications or any pharmacological agent that interferes with glucose metabolism or body weight were excluded from the study. The study was conducted at the University of Tennessee Health Science Center General Clinical Research Center (GCRC) after approval by the Institutional Review Board. All participants gave written informed consent prior to initiation of the study, which was conducted in accordance with the principles of the Declaration of Helsinki.

Assessments

Participants underwent prespecified Pathobiology of Prediabetes in a Biracial Cohort study assessments at the GCRC. Assessments staggered during quarterly visits in Year 1 included clinical examination, questionnaires on dietary and exercise habits, anthropometry, a standard OGTT, insulin secretion using frequently sampled intravenous glucose tolerance test, and measurement of insulin sensitivity by HEC (10). Height was measured in meters using a standard stadiometer, and weight was measured using a digital scale (WB-300 Plus; Tanita, Arlington Heights, IL). Body mass index (BMI) was calculated as weight in kilograms divided by the height in meter squared. Waist circumference was measured with a Gulick II tape measure. Habitual physical activity was recorded using the Modifiable Activity Questionnaire as previously described (11). The responses to the Modifiable Activity Questionnaire were converted to metabolic equivalent (MET) indicating the relative intensity of the reported activities and expressed as MET-hours/week (11).

HEC

Whole-body (principally skeletal muscle) insulin sensitivity was quantified using the HEC method of DeFronzo et al. (8). In brief, subjects avoided food overnight and were admitted to the GCRC as outpatients for the clamp procedure. Intravenous cannulas were inserted in both arms for insulin and dextrose infusion. Regular human insulin was infused continuously at a rate of 2 mU/kg/min (14.4 pmol/kg/min) for 180 minutes. Plasma glucose level was maintained (“clamped”) at ∼100 mg/dL (5.6 mmol/L) using variable-rate dextrose (20%) infusion. Arterialized blood samples were obtained every 10 minutes. The dextrose infusion rate during steady state (final 60 minutes of insulin infusion) was used to calculate insulin-stimulated glucose disposal. The latter was then divided by the steady-state plasma insulin level to obtain a measure of insulin sensitivity (Si-clamp) (8, 10).

Postclamp monitoring

At precisely 180 minutes, insulin infusion was discontinued, and all subjects were offered a standard meal provided by the hospital kitchen. The dextrose infusion was continued, and plasma glucose levels were monitored every 10 minutes. If two consecutive bedside plasma glucose values remained ≥20 mg/dL higher than 100 mg/dL, dextrose infusion was decreased by ∼25%. If the plasma glucose value was <100 mg/dL, dextrose infusion was increased by 25% to 50%, depending on the exact value. The iterative process was continued until dextrose infusion could be tapered and discontinued. Study subjects were then discharged from the research center.

Definition of GRT

GRT was defined as the interval between cessation of insulin infusion at the end of the 180-minute clamp procedure and discontinuation of 20% dextrose infusion, after determination of glycemic stability.

Insulin secretion

Acute insulin response to glucose was assessed using the frequently sampled intravenous glucose tolerance test as previously described (10). The disposition index was calculated as the product of acute insulin response (pmol/L) and Si-clamp (μmol/kg/min/pmol/L).

Insulin clearance

Basal insulin clearance was calculated as the mean of the molar ratio of fasting C-peptide (pmol/L) to fasting insulin concentrations (pmol/L) in plasma specimens obtained at −30 minutes and 0 minutes before initiation of HEC (12). Clamp insulin clearance was calculated as the mean of the molar ratio of plasma C-peptide levels (pmol/L) to plasma insulin levels (pmol/L) at steady state (final 60 minutes) during HEC.

The metabolic clearance rate (MCR) of insulin was calculated by dividing the insulin infusion rate at steady state by the incremental insulin concentration above the basal level during HEC using the following equation (8, 13, 14):

$$\mathrm{MCR}=\frac{\left[2000\left(\mathrm{\mu U/g/min}\right)\times \mathrm{Weight}\left(\mathrm{kg}\right)\times \mathrm{Body\; surface\; area}\left({\mathrm{m}}^2\right)\right]}{\left[\left(\mathrm{Plasma\; insulin\; at\; 60}-\min \left(\mathrm{\mu U}/\text{mL}\right)–\mathrm{Fasting\; plasma\; insulin}\right)\left(\mathrm{\mu U}/\text{mL}\right)\right]}$$

Laboratory analyses

Plasma glucose was measured using the glucose oxidase method (Yellow Spring Instruments Co., Inc., Yellow Spring, OH). Plasma insulin and C-peptide levels were measured immunochemically using commercially available kits in our Endocrine Research Laboratory.

The homeostasis models of insulin resistance (HOMA-IR), an estimate of hepatic insulin sensitivity, was calculated using fasting glucose and insulin values (6).

Statistical analysis

Data are reported as means ± SD unless SEM is specified. The significance level was set as P < 0.05. Differences in continuous variables between defined groups were analyzed using t tests and ANOVA. The relationship between GRT and variables of interest (demographic, clinical, biochemical, and glucoregulatory parameters) was analyzed using linear regression and partial correlation coefficients. All analyses were performed using SAS statistical software, version 9.3 (SAS Institute Inc., Cary, NC).

Results

Baseline characteristics of study subjects

A total of 207 subjects (68.6% women; 102 AA, 105 EA) who underwent HEC and had complete data were included for the present analysis (Table 1). The mean age was 46.3 ± 9.96 years, and BMI was 30.7 ± 8.43 kg/m² for the entire cohort. AA subjects were younger, had higher mean weight and BMI, and had lower mean FPG compared with EA participants. Insulin sensitivity (Si-clamp) was nominally lower and insulin secretion significantly higher in AA compared with EA participants. Basal insulin clearance rates were similar, but AA participants also had lower MCR and clamp insulin clearance values (Table 1).

Table 1.

Baseline Characteristics of the Study Subjects

	EA	AA	P Value
Number	105	102
Age, y	48.7 ± 9.91	43.9 ± 9.46	0.0003
BMI, kg/m2	29.2 ± 8.77	32.1 ± 7.81	0.008
Weight, kg	81.4 ± 19.62	90.97 ± 23.52	0.001
Waist circumference, cm	92.8 ± 14.87	96.97 ± 16.42	0.05
MAQ, MET-h/wk	23.6 ± 31.9	21.6 ± 35.8	0.65
FPG, mg/dL	93.4 ± 5.6	91.0 ± 6.65	0.004
2hrPG, mg/dL	125 ± 23.22	126 ± 29.1	0.68
Final postclamp PG, mg/dL	126 ± 27.4	122 ± 26.24	0.29
Glycemic recovery time, min	71.1 ± 29.8	73.3 ± 33.2	0.63
HOMA-IR	1.80 ± 1.48	1.83 ± 1.61	0.89
Si-clamp	0.143 ± 0.004	0.126 ± 0.005	0.08
AIR, pmol/L	367 ± 234	640 ± 541	<0.0001
Disposition index	47.9 ± 30.4	65.2 ± 51.8	0.005
Basal insulin clearance	15.5 ± 7.05	13.9 ± 8.76	0.19
Clamp insulin clearance	1.48 ± 0.64	1.05 ± 0.48	<0.0001
MCR, mL/kg/m2/min	3324 ± 1346	3177 ± 1372	0.44

Abbreviations: AIR, acute insulin response; MAQ (MET-h/wk), modifiable activity expressed in MET as MET h per wk; Si-clamp, insulin sensitivity measured with HEC. To convert the values for glucose to millimoles per liter, multiply by 0.0555.

Postclamp glycemic profile

At the end of the clamp procedure and at cessation of insulin infusion, study subjects were gradually weaned off dextrose infusion based on ambient plasma glucose levels. The final plasma glucose level at the time of discontinuation of dextrose infusion was 124.2 ± 26.9 mg/dL for the entire cohort. The final plasma glucose values were similar in men vs women (123.9 ± 25.8 mg/dL in men vs 124.3 ± 27.4 mg/dL in women; P > 0.5) and in AA vs EA participants (122.1 ± 26.2 mg/dL in AA vs 126.1 ± 27.4 mg/dL in EA; P > 0.5).

GRT

GRT measured the interval from cessation of insulin infusion to the time of discontinuation of dextrose infusion. There was a considerable variation in the GRT, with a range of 30 minutes to 270 minutes among the study subjects (Fig. 1A). The median GRT was 65 minutes, and the interquartile range was 36.5 minutes. Analysis of percentile distribution of the GRT in our study population showed that 90% of subjects had achieved postclamp glucose stabilization by 119 minutes (Fig. 1B). The mean GRT was 77.1 ± 42.7 minutes for the entire group, without significant differences by sex (men, 82.9 ± 45.5 minutes; women, 74.4 ± 42.3 minutes; P = 0.12) or ethnicity (AA, 82.0 ± 49.6 minutes; EA, 72.3 ± 34.2 minutes; P = 0.11).

Figure 1.

(A) Scatter diagram and (B) percentile distribution of glycemic recovery times among 207 subjects without diabetes after cessation of insulin infusion at the conclusion of the HEC procedure. Grid lines (from bottom to top) represent the 10th, 25th, 50th, 75th, and 90th percentiles of the distribution of glycemic recovery time.

Correlates of GRT

Using linear regression, we tested the association of GRT with age, sex, race/ethnicity, weight, waist circumference, FPG, 2hrPG, habitual physical activity (MET-hours/week), HOMA-IR, insulin sensitivity (Si-clamp), and insulin clearance. There was no significant association between GRT and race, sex, waist size, or FPG (Table 2). In univariate models, the significant predictors of postclamp GRT were age (r = −0.15, P = 0.03), weight (r = 0.18, P = 0.009), 2hrPG (r = −0.25, P = 0.0002), basal insulin clearance (r = 0.18, P = 0.02), whole-body insulin sensitivity (Si-clamp, r = 0.19, P = 0.03), HOMA-IR [a measure of insulin action in the liver (r = −0.17, P = 0.01)], and insulin secretion corrected for ambient insulin sensitivity (i.e., disposition index) (r = 0.18, P = 0.01) (Fig. 2 and Table 2). The association between habitual physical activity (MET-hours/week) and GRT was of borderline significance (r = 0.14, P = 0.052). Two measures of insulin clearance obtained during the HEC (clamp insulin clearance and MCR) exhibited no significant association with GRT, although ethnic disparity was noted in clamp insulin clearance (15) (Table 2).

Table 2.

Linear Regression of Glycemic Recovery Time vs Selected Variables

	r	F	P Value
Age, y	−0.15	4.30	0.03
Weight, kg	0.18	5.25	0.009
Waist, cm	0.05	0.43	0.65
FPG, mg/dL	0.04	0.185	0.67
2hrPG, mg/dL	−0.25	13.4	0.0002
MAQ, MET-h/wk	0.14	3.79	0.052
HOMA-IR	−0.17	5.15	0.01
Si-clamp	0.16	4.61	0.03
Disposition Index	0.18	6.01	0.015
Basal insulin clearance	0.18	5.88	0.01
Clamp insulin clearance	0.06	0.55	0.46
MCR, mL/kg/m2/min	0.05	0.49	0.48

Abbreviations: 2hrPG, 2-h plasma glucose during OGTT; MAQ (MET-h/wk), modifiable activity expressed in MET as MET h per wk; Si-clamp, insulin sensitivity measured with HEC.

Figure 2.

(A) GRT in relation to age, (B) 2hrPG during OGTT, (C) basal insulin clearance, (D) insulin sensitivity by HEC, (E) HOMA-IR, and (F) disposition index. The disposition index is a measure of insulin secretion, adjusted for ambient insulin sensitivity.

Stratification of glycemic recovery time intervals

To further understand the heterogeneity of individual postclamp GRT, we stratified participants into three groups: early recovery (GRT ≤60 minutes), intermediate recovery (GRT 61 to 120 minutes), and late recovery (GRT >120 minutes) (Table 3). Approximately 45% of participants achieved glycemic recovery and stabilization within 60 minutes postclamp, 47% did so within 120 minutes, and 8% required prolonged dextrose infusion for at least 2 hours. ANOVA revealed that participants who experienced late glycemic recovery and stabilization after the clamp procedure were more glucose tolerant (lower values for 2hrPG), were more insulin sensitive (lower HOMA-IR and higher Si-clamp values), and had greater insulin secretion (disposition index) compared with those whose blood glucose levels stabilized over an earlier or intermediate time frame (Table 3). There were no significant differences in age, weight, BMI, waist size, or insulin clearance values among participants from the three GRT strata (Table 3).

Table 3.

Characteristics of Study Subjects by Post-Clamp GRT Intervals

	Early Recovery (≤60 min)	Intermediate Recovery (61–120 min)	Late Recovery (>120 min)	P Value
Number	93	97	17
GRT, min	49.4 ± 8.20	82.8 ± 17.0	197 ± 41.1	<0.0001
Age, y	47.6 ± 10.2	45.2 ± 10.1	43.1 ± 7.69	0.46
BMI, kg/m2	30.4 ± 6.96	30.6 ± 10.2	31.5 ± 6.87	0.97
Weight, kg	84.4 ± 19.2	85.9 ± 25.3	94.7 ± 21.1	0.40
Waist, cm	94.6 ± 14.3	94.1 ± 17.5	98.4 ± 16.0	0.65
FPG, mg/dL	92.1 ± 7.04	92.0 ± 5.86	92.5 ± 4.95	0.89
2hPG, mg/dL	133 ± 30.2	122 ± 21.4	112 ± 19.4	0.04
HOMA-IR	2.20 ± 1.76	1.41 ± 1.19	1.56 ± 1.63	0.02
Si-clamp	0.124 ± 0.067	0.144 ± 0.067	0.161 ± 0.059	0.04
MAQ, MET-h/wk	20.9 ± 35.1	24.0 ± 34.7	31.4 ± 33.5	0.77
AIR, μU/mL	77.3 ± 55.2	85.0 ± 80.1	118 ± 99.7	0.25
Disposition index	50.3 ± 39.1	60.6 ± 44.8	83.5 ± 50.9	0.005
Basal insulin clearance	13.9 ± 6.68	15.7 ± 7.98	16.0 ± 14.1	0.14
Clamp insulin vlearance	1.22 ± 0.55	1.29 ± 0.65	1.43 ± 0.67	0.63
MCR, mL/kg/m2/min	3076 ± 1178	3360 ± 1440	3814 ± 1841	0.11

Abbreviations: Si-clamp, insulin sensitivity measured with HEC. To convert the values for glucose to millimoles per liter, multiply by 0.0555.

Discussion

HEC has been the gold standard method for assessing whole-body insulin sensitivity in humans for nearly 40 years (8). Typically, study subjects arrive for the test after avoiding food overnight and receive supraphysiological infusion of insulin for several hours. Although measurements of insulin sensitivity using the clamp method have been reported from centers around the world, few reports have focused on safety issues. Study subjects who consent to undergo HEC seldom experience significant harm besides the burdens of venipuncture and frequent blood collections over a prolonged period. Because insulin infusion is discontinued at the end of the clamp, the procedure is generally considered safe. As a further precaution, investigators typically run the dextrose infusion for some time after the insulin infusion has been discontinued and monitor blood glucose levels to determine the appropriate time for discharging study subjects.

However, residual insulin in the plasma compartment and receptor-bound insulin prior to cessation of insulin infusion continue to exert biological effects, including stimulation of glucose transport and metabolism. Thus, the ultimate time of dissipation of insulin action could vary significantly among individuals. Similarly, gastric emptying, digestion, and absorption of nutrients ingested immediately after cessation of insulin infusion are expected to display individual variation. As a result of these variations in the dissipation of insulin action and delivery of nutrients, prolonged dextrose infusion may be necessary to prevent or correct postclamp hypoglycemia in some individuals. To our knowledge, the factors that control or modulate the time to full glycemic recovery and stabilization after exposure to HEC conditions have not been fully determined. Some putative contributors to postclamp glycemic recovery and stabilization include the concentration of bioactive insulin in the circulation, insulin clearance (via hepatic extraction, insulin degrading enzymes, and other pathways), the extent of binding of insulin to receptors and postreceptor signaling, and tissue sensitivity to insulin action. Additional factors include the rate of recovery or rebound of endogenous insulin secretion and hepatic glucose production after their suppression during the hyperinsulinemic clamp.

In the present report, we systematically studied glycemic recovery after a standard HEC procedure in a large, diverse group of healthy subjects. Our findings indicate a ninefold variation (30 to 270 minutes) in postclamp GRT among otherwise healthy individuals. Nearly 50% of participants achieved glycemic stabilization with 1 hour postclamp, and the vast majority (∼90%) did so within 2 hours. We observed that GRT was inversely related to age and directly related to insulin sensitivity and insulin secretion, such that longer periods of dextrose infusion were required to achieve glycemic stabilization in younger insulin-sensitive subjects with higher disposition index compared with older insulin-resistant subjects with lower disposition index. The association between GRT and insulin sensitivity was evident using either Si-clamp or HOMA-IR as the measure of insulin sensitivity. Because HOMA-IR is a surrogate measure of hepatic insulin resistance, whereas Si-clamp assesses whole-body (principally, skeletal muscle) insulin sensitivity, our findings indicate that both hepatic and skeletal muscle insulin resistance may be protective of postclamp hypoglycemia. Our finding that higher 2hrPG during OGTT (evidence of glucose intolerance) predicted shorter GRT also is supportive of the aforementioned relationship between insulin resistance and postclamp glycemic recovery.

Receptor-bound insulin molecules continue to mediate glucose transport after discontinuation of exogenous insulin infusion at the end of the formal clamp procedure. The exact duration of the postclamp persistence of insulin action is unknown but could be ∼30 minutes or longer (16). Furthermore, free insulin molecules in the circulation remain ligands for receptor-activated glucose transport. Thus, an important variable in postclamp insulin dynamics would be the efficiency of insulin clearance and degradation. Using the standard approach, we calculated insulin clearance from the molar ratio of C-peptide to insulin concentration in basal (preclamp) plasma specimens and at steady state during HEC (12). A third measure, the MCR of insulin, was derived from the insulin infusion rate and ambient incremental plasma insulin during the HEC (8, 13, 14). Of these measures, only basal insulin clearance was significantly associated with postclamp GRT. Not surprisingly, the other two measures of insulin clearance obtained during the clamp procedure did not predict postclamp GRT.

Because the liver is the predominant site for insulin metabolism in humans, clearance values derived from the molar ratio of C-peptide and insulin largely reflect hepatic insulin extraction (12). Accordingly, data for insulin clearance obtained from insulin and C-peptide measurements in preclamp plasma specimens reflect basal hepatic insulin extraction efficiency, which is expected to be restored after cessation of hyperinsulinemic clamp. Thus, the finding that individuals with higher basal insulin clearance (i.e., greater hepatic insulin extraction) required longer times for postclamp glycemic recovery is physiologically congruent. Although hepatic insulin extraction exhibits marked individual and ethnic variations (15, 17–19), intrahepatic insulin promotes glycogen storage and inhibits gluconeogenesis and glycogenolysis, well-known effects that would depress ambient plasma glucose levels during the postclamp period (20).

The univariate models showed weak correlations between GRT and individual predictor variables, which suggests that unmeasured variables contributed to the variance in GRT. Factors related to postclamp meal consumption may be among these unmeasured variables. The standard practice is to offer study subjects a meal upon cessation of insulin infusion at the end of the HEC. However, the amount of food consumed as well as the time required for gastric emptying, digestion, and absorption of ingested nutrients could vary considerably among individuals. Depending on the liquid and solid composition of the ingested meal, gastric emptying half-time could be as long as 40 to 50 minutes, and additional time would be required to digest and absorb nutrients (21). Thus, ingested meals contribute to the variance in GRT; however, meals may not be a reliable source of glycemic support during initial 1- to 2-hour postclamp period due to variations in absorption kinetics. When our study participants were stratified according to early (≤60 minutes), intermediate (60 to 120 minutes), or late (>120 minutes) postclamp glycemic recovery time interval, we observed that early recovery was associated with lower glucose tolerance, insulin resistance, and lower disposition index, whereas delayed glycemic recovery was associated with better glucose tolerance, higher insulin sensitivity, and greater disposition index. Thus, among healthy individuals who had been fasting and had completed 3 hours of the HEC procedure, early recovery and stabilization of blood glucose occurred in those who had lower insulin sensitivity and secretion profile. Such a metabolic profile is physiologically congruent because it would favor increased endogenous glucose production and decreased peripheral glucose utilization.

In summary, using data from a biracial cohort of 207 healthy subjects who underwent the HEC procedure at our center, we have described the physiology of postclamp glycemic recovery and stabilization. Approximately 90% of subjects achieved stable autochthonous normoglycemia within ∼120 minutes after cessation of insulin infusion. The critical physiological mechanisms that modulate glycemic recovery and stabilization during the vulnerable period between cessation of insulin infusion and adequate delivery of nutrients from ingested meals include resumption of hepatic glucose production, innate glucose tolerance, insulin sensitivity and secretion, and basal insulin clearance. Our findings stress the need for parenteral dextrose support and close monitoring in subjects completing the HEC procedure until there is clear evidence of glycemic recovery and stabilization after tapering the dextrose infusion.

Glossary

Abbreviations:

2hrPG

2-hour plasma glucose

AA

African American

BMI

body mass index

EA

European American

FPG

fasting plasma glucose

GCRC

University of Tennessee Health Science Center General Clinical Research Center

GRT

glucose recovery time

HEC

hyperinsulinemic euglycemic clamp

HOMA-IR

homeostasis model of insulin resistance

MCR

metabolic clearance rate

MET

metabolic equivalent

OGTT

oral glucose tolerance test