Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Aug 1.
Published in final edited form as: Diabetes Res Clin Pract. 2021 Jul 15;178:108948. doi: 10.1016/j.diabres.2021.108948

Differential Loss of β-cell Function in Youth vs. Adults Following Treatment Withdrawal in the Restoring Insulin Secretion (RISE) Study

KRISTINA M UTZSCHNEIDER 1,2, MARK T TRIPPUTI 3, ALEXANDRA KOZEDUB 1, ELENA BARENGOLTS 4, SONIA CAPRIO 5, MELANIE CREE-GREEN 6, SHARON L EDELSTEIN 3, LAURE EL GHORMLI 3, TAMARA S HANNON 7, KIEREN J MATHER 7, JERRY PALMER 1,2, KRISTEN J NADEAU 6; THE RISE CONSORTIUM
PMCID: PMC8628318  NIHMSID: NIHMS1729975  PMID: 34274407

Abstract

Aims:

To compare OGTT-derived estimates of β-cell function between youth and adults with impaired glucose tolerance (IGT) or recently diagnosed type 2 diabetes after treatment discontinuation in RISE.

Methods:

Youth (n=89) and adults (n=132) were randomized to 3 months glargine followed by 9 months metformin (G/M) or 12 months metformin (MET). Insulin sensitivity and β-cell responses were estimated from 3-hour OGTTs over 21 months. Linear mixed models tested for differences by time and age group within each treatment arm.

Results:

After treatment withdrawal, HbA1c increased in both youth and adults with a larger net increase in G/M youth vs. adults at 21 months. Among youth, β-cell function decreased starting at 12 months in G/M and 15 months in MET. Among adults, β-cell function remained relatively stable although insulin secretion rates decreased in G/M at 21 months. At 21 months vs. baseline β-cell function declined to a greater extent in youth vs. adults in both the G/M and MET treatment arms.

Conclusions:

After treatment withdrawal youth demonstrated progressive decline in β-cell function after stopping treatment with either G/M or MET. In contrast, β-cell function in adults remained stable despite an increase in HbA1c over time.

ClinicalTrials.gov Identifier:

NCT01779375 and NCT01779362 at clinicaltrials.gov

Keywords: β-cell function, impaired glucose tolerance, insulin secretion, type 2 diabetes, youth, metformin

1. INTRODUCTION

1.1 Type 2 diabetes in youth used to be rare, but now is increasing in incidence (1) and expected to increase three-fold over the next 40 years (2). The Treatment Options for Type 2 Diabetes in Adolescents and Youth (TODAY) Study found a rapid rate of decline in β-cell function despite treatment with metformin and/or rosiglitazone (3). Glycemic progression appears to run a more aggressive course in youth than adults before and after progression to type 2 diabetes (4), but matched longitudinal studies examining both youth and adults with identical protocols have never before been performed. The Restoring Insulin Secretion (RISE) Study is the first study to directly compare treatment responses between youth and adults using identical treatment arms as well as protocols to assess β-cell function (5). Baseline data from the RISE study demonstrated that youth secrete more insulin than adults despite similar degrees of obesity and dysglycemia, even after adjusting for the lower insulin sensitivity observed in youth (6–8), thus demonstrating that the capacity to secrete insulin is still robust in youth with IGT or recently diagnosed type 2 diabetes.

1.2 The premise of the RISE Studies was to determine if early treatment for 12 months could reverse the defect in β-cell function and alter the course of the disease by demonstrating persistent beneficial effects after treatment withdrawal. The RISE Studies were designed to provide two matching treatment arms in youth and adults, namely treatment with glargine for 3 months followed by metformin for 9 months or treatment with metformin alone for 12 months. Both treatment approaches were intended to off-load the stress on the β-cell in the hope that β-cell function would improve, and the natural course of the diseases process would be modified. At the time of the study, metformin and insulin were the only approved treatments for type 2 diabetes in youth.

1.3 Comparing estimates of β-cell function derived from hyperglycemic clamps in RISE, β-cell function in youth declined during treatment and 3 months after treatment withdrawal, while in adults transient improvement in β-cell function was observed at 12 months of treatment but was not sustained after treatment withdrawal (9). Here we report data from the oral glucose tolerance tests (OGTTs) performed in RISE. OGTTs were performed at multiple time points and over a longer period of follow-up than the hyperglycemic clamps (out to 21 months from baseline). This allowed the use of mathematical modeling of OGTT C-peptide and glucose data over the course of the study to determine at what time point β-cell function parameters started to decline. This also provided a longer time frame to compare changes in β-cell function between youth and adults. Additionally, the OGTT functions as a more integrated test by including the gut-pancreas axis and thus is more reflective of normal physiology than a hyperglycemic clamp. We hypothesized that estimates of β-cell function would decline in both groups over time, but earlier and to a greater degree in youth.

2. SUBJECTS, MATERIALS AND METHODS

2.1. Study Design

2.1.1 The RISE Study was a prospective, randomized, controlled clinical trial designed to determine if early treatment could reverse β-cell dysfunction in IGT and early type 2 diabetes. RISE included two medication studies, one in youth and one in adults, with shared standardized protocols for assessing β-cell function. This analysis compares results from the two matched medication arms that were identical in youth and adults: glargine followed by metformin (G/M) or metformin alone (MET). The RISE Study is registered under ClinicalTrials.gov identifiers: NCT01779375 and NCT01779362. RISE was approved at all study sites by the local investigational review boards and all participants gave written informed consent/assent consistent with the Declaration of Helsinki and the guidelines of each center’s institutional review board.

2.2. Screening and Eligibility Criteria

2.2.1 Overweight and obese youth (BMI ≥85th percentile for age) aged 10–19 years with pubertal development Tanner stage ≥II, and adults aged 20–65 years with a BMI ≥25 kg/m2 were screened if they were thought to be at high risk for IGT or type 2 diabetes. Screening consisted of a 75-gram, 2-hour OGTT and HbA1c, with eligibility requirements including a 2-hour glucose concentration ≥7.8 mmol/L and additional protocol-specific criteria for fasting glucose and HbA1c as detailed elsewhere (6). While adults were required to be drug-naive, youth with type 2 diabetes could qualify for the study if they had been taking metformin for less than 6 months as treatment for diabetes. Additional details on participant recruitment and eligibility criteria have been described elsewhere (5). Classification of glucose tolerance was based on American Diabetes Association OGTT criteria for fasting and 2-hour glucose (10).

2.3. Interventions

2.3.1 Those randomized to G/M received glargine for 3 months followed by metformin alone for 9 months. While on glargine, participants were instructed to check their fasting blood glucose daily and the dose of glargine was titrated at least twice a week by phone (11) with a target fasting glucose of 4.4–5.0 mmol/L. Youth who were already taking metformin discontinued metformin when glargine was started. After 3 months of treatment with glargine, glargine was stopped and metformin was started or restarted, titrated to 1000 mg twice daily, and continued for 9 months. Those randomized to MET received metformin, titrated to 1000 mg twice daily, for 12 months. Study medication was stopped at 12 months and participants were followed off all glucose-lowering medications for an additional 9 months. Metabolic decompensation was defined as plasma glucose >300 mg/dl accompanied by significant symptoms (nausea, vomiting, dehydration) or ketosis. If the participant required insulin treatment for greater than 4 weeks, they were withdrawn from the study and censored at that time.

2.4. Anthropometric Measurements

Weight was measured using a calibrated electronic scale and height was measured without shoes using a stadiometer.

2.5. Oral Glucose Tolerance Test

2.5.1 Following a 10-h overnight fast, a 3-hour, 75-gram OGTT was performed. An indwelling intravenous catheter was placed for venous sampling. After 15 minutes of rest, venous blood samples were drawn 10 and 5 minutes prior to ingesting the glucose solution, which was consumed within 5 minutes. Additional blood samples were obtained 10, 20, 30, 60, 90, 120, 150, and 180 minutes after the start of glucose ingestion. If the participant was taking metformin, the last dose of metformin was taken the evening before the OGTT (11). OGTTs were performed at baseline prior to randomization, at 6 and 12 months while on treatment and at study months 15 and 21 (3 and 9 months off treatment).

2.5.2 Blood samples were collected and placed immediately on ice prior to being separated and frozen at −80°C. Samples were shipped to a central laboratory at the University of Washington for measurement of plasma glucose, C-peptide, and insulin.

2.6. Laboratory Assays

2.6.1 Blood samples were collected and processed according to standardized procedures and shipped for analysis to the RISE Study Central Biochemistry Laboratory (Northwest Lipid Metabolism and Diabetes Research Laboratories, University of Washington, Seattle, WA). Glucose was measured by the glucose hexokinase method on a c501 autoanalyzer (Roche, Indianapolis IN). C-peptide and insulin were measured by a two site immunoenzymometric assay performed on the Tosoh 2000 autoanalyzer (Tosoh Bioscience, Inc., South San Francisco, CA). Quality control samples were performed at regular intervals with low, medium, medium-high, and high concentrations with inter-assay coefficients of variation ≤2.0% for glucose, ≤4.3% for C-peptide, and ≤3.5% for insulin. HbA1c was measured by ion-exchange high performance chromatography on a TOSOH G8 analyzer (TOSOH Biosciences, Inc., South San Francisco, CA). The inter-assay CVs on low- and high-quality control samples were 1.9% and 1.0%, respectively. Further assay details have been published (6).

2.7. Calculations

2.7.1 Insulin Secretion Response Measures: Model parameters were estimated from OGTT glucose and C-peptide concentrations using C-peptide deconvolution (12), as previously described (13). β-cell measures were assessed from the OGTT using a model describing the relationship between insulin secretion rate (ISR) and glucose concentration as the sum of two components (13; 14). The first component represents the dependence of ISR on glucose concentration through a dose-response function relating the two variables. From the dose-response, β-cell glucose sensitivity (the slope), basal ISR and ISR at a fixed reference glucose concentration (6.5 mmol/L) were calculated. The integral of ISR during the whole test (total ISR) was also calculated. The dose-response is modulated by a potentiation factor, accounting for various mechanisms. The potentiation factor averages one during the test and expresses relative potentiation or inhibition of ISR; its excursion is quantified by the ratio between the time interval from 100–120 minutes relative to the basal time period 0–20 minutes (potentiation ratio). The second ISR component represents the dependence of ISR on the rate of change of glucose concentration and is determined by a single parameter (rate sensitivity), which is related to early insulin release (15). Only OGTTs with fasting values and at least four additional time points after glucose consumption were used to ensure successful identification of the model parameters. Modeling was performed using Matlab R2018b (Mathworks, Natick, MA) by one person trained in the technique and blinded to treatment assignment and age group.

2.7.2 The insulinogenic index (IGI) and C-peptide index (CPI) were calculated as the change in insulin or C-peptide from 0–30 minutes divided by the change in glucose over the same period (ΔI/ΔG and ΔCP/ΔG). Incremental area under the curve (iAUC) insulin (i), C-peptide (cp) and glucose (g) responses were calculated using the trapezoidal rule and iAUCi/iAUCg and iAUCcp/iAUCg computed.

2.7.3 Insulin Sensitivity: Oral glucose insulin sensitivity (OGIS) was quantified by modeling of OGTT glucose and insulin data (16). Insulin sensitivity was also estimated as 1/fasting insulin (1/FI).

2.8. Data Management and Statistical Analysis

2.8.1 Data included baseline and follow up visits (6, 12, 15 and 21 months). The 6- and 12-month timepoints represent on-treatment evaluations, and the 15- and 21-month timepoints represent 3- and 9-month post-treatment evaluations. Data were transformed where appropriate to approximate normality for statistical analysis.

2.8.2 Descriptive statistics include number and percent for categorical covariates and mean±SD for continuous covariates. Changes from baseline in measures of β-cell function were estimated from linear mixed models with a random intercept by clinic type with adjustment for race, sex, baseline BMI and diabetes status. Based on the hyperbolic relationship between insulin sensitivity and the early insulin response to glucose (17), the oral disposition index was calculated as ΔI/ΔG × 1/fasting insulin. Model-derived measures were adjusted for insulin sensitivity using time-varying OGIS. Directly calculated measures were adjusted for insulin sensitivity using time-varying 1/FI (except for the insulin disposition index model). The model for total ISR was further adjusted for iAUCg. Outcomes that were log transformed are presented as percent change from baseline in the geometric mean. Untransformed outcomes are presented as mean change. For rate sensitivity, a Box-Cox power transformation (lambda 1/3) was applied prior to testing.

2.8.3 The number of participants whose data was analyzed at each time point are delineated in the Consort Diagram (Figure 1). Missing data were not imputed. Statistical analyses were performed using R (R Core Team, 2017). Portions of the baseline data have been previously published (8).

Figure 1.

Figure 1.

Open in a new tab

Consort Diagram

3. RESULTS

3.1. Participant Characteristics

3.1.1 Of the 223 participants enrolled and randomized to G/M or MET, 91 were youth and 132 were adults. Two youth participants (one in each treatment arm) did not have baseline modeling data and were excluded from analyses. Enrollment and OGTT completion rates and reasons for missing data at each time point are detailed in Figure 1. Study withdrawal due to metabolic decompensation occurred in a total of 10 youth (7 in the G/M arm and 3 in the MET arm) and 8 adults (4 in the G/M arm and 4 in the MET arm) over the course of the study.

3.1.2 Baseline characteristics are presented in Table 1. There were racial/ethnic differences between youth and adults as well as a female preponderance in youth that was not present in adults. BMI was slightly lower in the adults vs. youth in the MET arm, but did not differ by age group in the G/M arm. HbA1c did not differ by age group, but fasting glucose was significantly lower and fasting insulin higher in youth vs. adults. The proportions of participants with diabetes and IGT were equally distributed by age and treatment group. Due to differences in inclusion criteria, 21 youth were on metformin prior to randomization, but they were equally distributed between the two treatment groups.

Table 1.

Comparison of Baseline Characteristics by Age and Treatment Group†

Glargine/Metformin (GM) Y vs A
p-value		 Metformin (Met) Y vs A
p-value		 GM vs Met p-value in youth GM vs Met p-value in adults
Youth (Y; N=43) Adult (A; N=67) Youth (Y; N=46) Adult (A; N=65)
DEMOGRAPHICS
Race/ethnicity < 0.001 < 0.001 0.413 0.606
 White 13 (30.2) 37 (55.2) 12 (26.1) 34 (52.3)
 Black 13 (30.2) 21 (31.3) 8 (17.4) 19 (29.2)
 Hispanic (any) 14 (32.6) 5 (7.5) 20 (43.5) 6 (9.2)
 Other‡ 3 (7.0) 4 (6.0) 6 (13.0) 6 (9.2)
Female 26 (60.5) 23 (34.3) 0.013 37 (80.4) 37 (56.9) 0.017 0.066 0.015
Age (years) 14.8 ± 2.0 53.5 ± 9.3 < 0.001 14.0 ± 2.1 55.2 ± 8.2 < 0.001 0.045 0.273
Body Mass Index (kg/m2) 36.6 ± 6.4 35.0 ± 5.9 0.212 37.2 ± 6.1 35.0 ± 5.1 0.046 0.630 0.960
GLYCEMIC MEASURES
HbA1c (%) 0.450 0.399 0.719 0.547
 % 5.7 ± 0.6 5.8 ± 0.3 5.7 ± 0.6 5.8 ± 0.4
 mmol/mol 39 ±7 40 ± 4 39 ± 6 39 ± 4
Fasting Glucose (mmol.L −1) 6.0 ± 0.8 6.2 ± 0.7 0.035 6.1 ± 1.1 6.2 ± 0.7 0.034 0.971 0.996
iAUCg (mmol.L −1) 633.9 ± 233.9 610.5 ± 260.4 0.731 620.5 ± 248.5 601.9 ± 248.0 0.643 0.829 0.761
Fasting Insulin (pmol.L −1) 270.7 ± 199.7 145.6 ± 155.5 < 0.001 268.4 ± 125.0 118.8 ± 61.5 < 0.001 0.311 0.785
IGT/diabetes Status at Screening 17 (39.5) 17 (25.4) 0.175 18 (39.1) 16 (24.6) 0.154 1.000 1.000
Metformin Use at Baseline 12 (27.9) 0 (0.0) < 0.001 11 (23.9) 0 (0.0) < 0.001 0.809
INSULIN SENSITIVITY
OGIS 3-hour (ml.min −1.m−2) 250.7 ± 76.7 294.7 ± 72.0 0.004 245.6 ± 72.0 290.9 ± 56.5 < 0.001 0.856 0.679
1/Fasting Insulin (pmol−1.L) 5.7 ± 4.4 10.5 ± 5.9 < 0.001 4.7 ± 2.5 11.0 ± 6.1 < 0.001 0.311 0.785
MODEL OUTCOMES
Basal ISR (pmol.min −1.m−2) 219.8 ± 87.9 165.7 ± 84.7 <0.001 244.1 ± 69.1 151.7 ± 50.4 < 0.001 0.585 0.053
ISR at 6.5 mmol/L (mmol.L −1) 304.0 ± 112.8 240.3 ± 91.9 0.004 374.3 ± 160.2 235.0 ± 118.1 < 0.001 0.029 0.726
Total ISR (nmol.m−2) 123.7 ± 35.7 97.1 ± 33.2 < 0.001 132.5 ± 48.0 96.7 ± 27.1 < 0.001 0.553 0.729
Potentiation Factor Ratio 1 1.2 ± 0.4 1.4 ± 0.5 0.040 1.2 ± 0.5 1.3 ± 0.4 0.148 0.913 0.735
Rate Sensitivity (mmol.L−1) 2166.3 ± 1480.8 997.2 ± 826.4 -- 2258.8 ± 1725.8 1125.0 ± 1170.2 -- -- --
 Rate Sensitivity, Power Transformed§ 33.8 ± 8.9 23.8 ± 10.5 < 0.001 32.8 ± 12.7 24.1 ± 12.1 < 0.001 0.886 0.753
Glucose Sensitivity (pmol.min −1.m −2.mM −1) 121.9 ± 57.8 95.0 ± 54.0 0.006 125.4 ± 74.5 109.5 ± 72.9 0.254 0.906 0.077
OGTT OUTCOMES
IGI (uU.mg −1) 224.5 ± 176.6 107.6 ± 96.4 <0.001 229.8 ± 174.8 116.3 ± 106.4 <0.001 0.945 0.437
Insulin disposition index (mmol−1.L) 1357.4 ± 991.7 1273.7 ± 826.6 0.846 1410.3 ± 1454.7 1438.6 ± 1186.7 0.153 0.381 0.532
CPI (ng.mg −1) 11.9 ± 6.8 7.4 ± 4.5 < 0.001 12.8 ± 7.6 8.2 ± 5.6 <0.001 0.610 0.270
iAUCi/iAUCg (pmol.mmol−1) 349.4 ± 272.3 225.9 ± 205.7 0.002 376.6 ± 273.8 202.8 ± 145.8 <0.001 0.627 0.704
iAUCcp/iAUCg (nmol.mmol−1) 1.0 ± 0.8 0.9 ± 0.6 0.170 1.0 ± 0.5 0.9 ± 0.5 0.085 0.415 0.687
Open in a new tab

Values are N (percent) or mean±SD as indicated.

Abbreviations: iAUCg, incremental area under the curve for glucose; OGIS, oral glucose insulin sensitivity; ISR, insulin secretion rate; OGTT, oral glucose tolerance test; IGI, insulinogenic index (ΔI30/ΔG30); Insulin oral disposition index (IGI*1/fasting insulin); CPI, C-peptide index (ΔC30/ΔG30); iAUCi/iAUCg, ratio of the incremental area under the curve for insulin (iAUCi) to that for glucose (iAUCg); iAUCcp/iAUCg, ratio of the incremental area under the curve for C-peptide (iAUCi) to that for glucose (iAUCg).

†

P-values for continuous variables are based on t-test or Wilcoxon test, as appropriate; P-values for categorical variables are based on chi-square or Fisher’s exact test, as appropriate.

‡

“Other” for race/ethnicity includes mixed, Asian, American Indian, and other.

§

A Box-Cox power transformation (lambda=1/3) was applied to rate sensitivity for testing, to approximate normality.

P-value no longer significant at the 0.05 significance level after adjustment for insulin sensitivity (OGIS 3-hour for model outcomes and 1/fasting insulin for OGTT outcomes), from ANOVA Type III F test.

3.2. Baseline Insulin Sensitivity and Insulin Secretion/Response Measures in Youth and Adults

3.2.1 Baseline unadjusted data for insulin sensitivity and β-cell function measures are provided in Table 1. Estimates of insulin sensitivity (OGIS and 1/FI) were significantly lower in youth vs. adults. Modeled measures of insulin secretion in the basal state (basal ISR and ISR at 6.5mmol/L glucose) and total ISR over the OGTT (tISR) and rate sensitivity were all higher in youth vs. adults and remained significant even after further adjustment for OGIS (data not shown). Glucose sensitivity was higher at baseline in youth vs. adults in the G/M arm but was no longer significant after adjustment for OGIS, and it did not differ between youth and adults in the MET arm (Table 1). The early insulin and C-peptide responses (IGI and CPI) and the integrated insulin response (iAUCi/iAUCg), which includes early and late insulin release and is impacted by insulin clearance, were higher in youth vs. adults at baseline. The oral disposition index did not differ significantly between youth and adults. After adjusting for insulin sensitivity, only the CPI in the G/M treatment arm remained significantly higher in youth. The integrated C-peptide response (iAUCcp/iAUCg) did not differ between youth and adults at baseline.

3.3. Glycemic Measures over Time

3.3.1 Changes in HbA1c and incremental AUC glucose (iAUCg) from baseline are depicted in Figure 2 with data on absolute values for HbA1c, fasting glucose and iAUCg provided in Supplemental Table 1 and change from baseline in Supplemental Table 2. In youth, HbA1c and fasting glucose increased progressively relative to baseline after treatment withdrawal, becoming significant in the G/M arm at study months 15 and 21 (3 and 9 months off study drug) with iAUCg significantly increased only at month 21. In youth in the MET arm these variables only became significantly higher at month 21 (9 months off study drug). In adults, HbA1c also increased after treatment withdrawal, becoming significant in the G/M arm at 21 months and in the MET arm at 15 and 21 months. Fasting glucose did not change significantly in adults, but iAUCg was higher at months 6 and 12 while on treatment in the G/M arm only. The net change in HbA1c from baseline was significantly higher at 21 months in youth compared to adults in the GM arm (p=0.013) but was not significantly different in the MET arm (p=0.15). In the G/M arm, fasting glucose was higher in youth vs. adults at 15 and 21 months, but iAUCg was higher in adults at 6 and 12 months, while in the MET arm both were higher in youth at 21 months (Supplemental Table 2).

Figure 2.

Figure 2.

Open in a new tab

Changes from baseline over time are depicted for each treatment arm for HbA1c (A and B), iAUCg (C and D), OGIS (E and F) and 1/FI (G and H). Youth (blue) and Adults (red) in the Glargine/Metformin arm are depicted in panels A, C, E and G and those in the Metformin arm in panels B, D, F and H. All values are model estimates and 95% CI after adjustment for race, sex, BMI, and baseline glucose tolerance category. Significant differences in changes in measures between youth and adults are denoted by an *. iAUCg = incremental area under the curve for glucose, OGIS = oral glucose insulin sensitivity, 1/FI = 1/fasting insulin.

3.4. Insulin Sensitivity over Time

3.4.1 Changes in insulin sensitivity relative to baseline are depicted in Figure 2 (OGIS E and F and 1/FI G and H) with data provided in Supplemental Tables 1 and 2. In youth, insulin sensitivity increased significantly while on treatment (OGIS: G/M at 6 and 12 months, MET at 12 months; 1/FI: G/M at 6 and 12 months, MET at 6 and 12 months). Off treatment, OGIS gradually returned to baseline by 21 months, but 1/FI remained significantly above baseline at 15 and 21 months in the G/M arm and at 21 months in the MET arm.

3.4.2 In adults, there were no significant changes in insulin sensitivity in the G/M treatment arm, even during active treatment while on metformin (Months 3–12). In the MET arm, insulin sensitivity increased significantly while on treatment (OGIS: 6 and 12 months; 1/FI: 6 and 12 months), then gradually returned to baseline at 21 months, i.e., 9 months after treatment withdrawal.

3.4.3 Significant differences between youth and adults in changes in insulin sensitivity relative to baseline were only observed in the G/M treatment arm (Figure 2, difference between youth and adults denoted by an *). Youth demonstrated significantly greater increases in OGIS at 12 and 15 months and in 1/FI at 12, 15 and 21 months.

3.5. Estimates of β-cell Function over Time in Youth

3.5.1 Changes in β-cell function relative to baseline are illustrated in Figure 3 (modeling measures) and Figure 4 (calculated measures) with data presented in Supplemental Tables 1 and 2. All statistical models were adjusted for changes in insulin sensitivity over time and tISR was further adjusted for iAUCg. In youth in the G/M treatment arm, tISR was significantly lower as early as 12 months while basal ISR@6.5 mmol glucose, tISR, glucose sensitivity and rate sensitivity were all significantly decreased after treatment withdrawal at 15 and 21 months (Figure 3). Early and integrated insulin and C-peptide responses were decreased in the G/M arm at 21 months while the iAUCi/iAUCg response was also decreased at 15 months (Figure 4). The composite oral disposition index trended to decrease at 21 monthsin the G/M arm, but this did not reach statistical significance (p=0.06). In the MET treatment arm, significant decreases in tISR and glucose sensitivity were noted at 15 months. At 21 months basal ISR, ISR at 6.5mmol/L glucose, tISR and rate sensitivity were decreased with a trend towards a decrease in glucose sensitivity (p=0.06) (Figure 3). In the MET arm, the CPI was decreased at 12, 15 and 21 months, and IGI and iAUC insulin and C-peptide responses were decreased only at 21 months (Figure 4). The disposition index did not change over time in the MET arm. Significant treatment-related differences in β-cell function measures in youth were only observed for glucose sensitivity at 21 months, with a greater decline in the G/M vs. MET arm (p=0.02). There were no significant changes in the potentiation factor ratio at any time point for any arm.

Figure 3.

Figure 3.

Open in a new tab

Changes from baseline over time are depicted for each treatment arm for each β-cell function measure derived from modelling. Youth (blue) and Adults (red) treated with Glargine/Metformin are depicted in panels A, C, G and E and those in the Metformin arm in panels B, D, F and H. All values are model estimates with 95% CI after adjustment for race, sex, BMI, baseline glucose tolerance category and time-varying OGIS, and additionally adjusted with time-varying iAUCg for tISR only. Significant differences in changes in measures between youth and adults are denoted by an *. ISR = insulin secretion rate, OGIS = oral glucose insulin sensitivity.

Figure 4.

Figure 4.

Open in a new tab

Changes from baseline over time are depicted for each treatment arm for each β-cell function measure calculated from the OGTT. Youth (blue) and Adults (red) treated with Glargine/Metformin are depicted in panels A, C, E, G and I and those in the Metformin arm in panels B, D, F, H and J. All values are model estimates with 95% CI after adjustment for race, sex, BMI, baseline glucose tolerance category and time-varying 1/FI (except for insulin DI model).Significant differences in changes in measures between youth and adults are denoted by an *. 1/FI = 1/fasting insulin, DI = disposition index, IGI = insulinogenic index, CPI = C-peptide index, iAUC = incremental area under the curve for insulin (i), C-peptide (cp) and glucose (g).

3.6. Estimates of β-cell Function over Time in Adults

3.6.1 In contrast to youth, there were fewer changes in measures of β-cell function in adults over time (Figures 3 and 4 and Supplemental Tables 1 and 2). In the G/M arm, basal ISR (12, 15 and 21 months) and tISR (15 and 21 months) declined significantly and remained lower at month 21 relative to baseline (Figure 3); there were no significant changes in the early insulin and C-peptide responses, but the iAUC insulin and C-peptide responses were decreased significantly at months 6, 12 and 15, but were no longer different from baseline at month 21 (Figure 4). The disposition index did not change over time in the G/M arm. In the MET arm, there were only transient changes with an increase in rate sensitivity at month 6 and decreases in tISR and glucose sensitivity at month 15 (Figure 3); there were no significant changes in calculated measures of β-cell function in the MET arm except for an increase in the disposition index while on treatment at month 6 (Figure 4). There were no treatment-related differences in β-cell function measures at any time point in adults (MET vs. G/M in adults, NS). There were no significant changes in the potentiation factor ratio at any time point in either treatment arm.

3.7. Differences in β-cell Function over Time between Youth and Adults

3.7.1 Changes in estimates of β-cell function relative to baseline were compared between youth and adults (See Supplemental Table 2 for analyses). In the G/M treatment arm, tISR decreased more in youth vs. adults at both 15 and 21 months, while in the MET treatment arm there was no significant difference over time. Glucose sensitivity decreased more in youth vs. adults at 21 months only in the G/M arm. In the G/M arm, rate sensitivity tended to decrease more in youth from 6–21 months (p=0.05–0.06) with the p value reaching statistical significance only at month 15 (p=0.04), while in the MET arm it decreased more at month 6 and month 21 (Figure 3). Of the calculated β-cell function OGTT indices, the IGI was significantly decreased in youth vs. adults at month 21 in both treatment arms, but the calculated disposition index did not differ between youth vs. adults in either treatment arm. The CPI was reduced in youth vs adults in the G/M arm at month 21 and in the MET arm at month 12 with a trend at month 21 (p=0.05, Figure 4). Integrated AUC indices did not differ between youth and adults at any time point.

4. DISCUSSION

4.1 Results from the RISE Study, where youth and adults in matched treatment arms were compared, demonstrate a progressive decline in β-cell function in response to an oral glucose challenge and worsening of HbA1c in youth, while β-cell function in adults remained relatively stable over 21 months. The results presented here expand on previously published hyperglycemic clamp findings from RISE (9; 11; 18) by evaluating OGTT responses at multiple time points out to 21 months, when participants were off medications for 9 months. Further, we used modeling to assess insulin secretion rates and estimates of β-cell sensitivity to a more physiologically relevant stimulus of orally delivered glucose. In youth, the G/M treatment arm had a greater decline in the sensitivity of the β-cell to secrete insulin in response to glucose (glucose sensitivity) and an earlier decline in ISRs than the MET treatment arm. These results suggest that an initial 3 months of treatment with glargine insulin followed by metformin may be inferior to treatment with metformin alone. Despite the relative stability of measures of β-cell function in adults, it is important to point out that HbA1c was stable while on treatment but increased significantly after treatment withdrawal in both youth and adults. The RISE Study illustrates the more progressive nature of the disease in youth and the inability of the two treatment approaches to arrest and/or slow progression of the underlying pathology in youth or reverse it in adults.

4.2 We can assess the results of RISE in the context of other large treatment trials, although most large treatment trials do not include a wash-out period with extended follow-up. The TODAY study in youth with new-onset type 2 diabetes demonstrated a progressive decline in β-cell function despite ongoing treatment with either metformin, metformin plus rosiglitazone or metformin plus lifestyle changes (3), which is consistent with our results from RISE, but now extend to youth with type 2 diabetes of shorter duration and with IGT. Type 2 diabetes is a heterogenous disease and it is plausible that dysglycemia presenting in youth represents a more severe phenotype of progressive β-cell dysfunction and/or more severe insulin resistance that therefor manifests earlier than adults. In adults with treatment-naïve type 2 diabetes enrolled in A Diabetes Outcome Progression Trial (ADOPT), early improvement in β-cell function and stability over time was demonstrated (19), but 21% of those treated with metformin had monotherapy failure at 5 years (compared to 15% with rosiglitazone and 34% with glyburide) (20), suggesting that ongoing treatment is likely necessary but not 100% effective at preventing progression. Treatment with metformin in the Diabetes Prevention Program failed to prevent progression from IGT to type 2 diabetes completely but did decrease the risk by 31% (21); this effect was attenuated to 25% after a short 1–2 week wash-out period (22). Using multiple different measures of β-cell function derived from OGTTs, our data demonstrate stability of β-cell function after a 9-month wash-out period in adults with IGT or treatment-naïve type 2 diabetes, although HbA1c increased once metformin was stopped. The increase in HbA1c despite stable β-cell function in adults could be related in part to an increase in endogenous glucose production once metformin was discontinued (23; 24).

4.3 It has been hypothesized that resting the β-cell can restore β-cell function (25). The glargine followed by metformin intervention was selected based on data showing that short-term intensive treatment with insulin via continuous subcutaneous insulin infusion or multiple daily injections of insulin early in the course of type 2 diabetes could lead to remission and improved β-cell function (26; 27). Reversing glucotoxicity on the β-cell by aiming to normalize glucose levels was one of the premises of the Outcome Reduction with Initial Glargine Intervention (ORIGIN) Trial. In ORIGIN, incident diabetes was evaluated in adults with IGT or impaired fasting glucose at baseline. Compared to placebo, those treated with glargine once daily, aimed at reducing fasting glucose to 5.3 mmol/L or less, had lower incident diabetes, even after insulin glargine was stopped for 100 days (28). However, in the RISE Study, despite achieving mean fasting blood glucose levels of 5.2±0.7 mmol/L in youth (11) and 5.2±0.4 mmol/L in adults (18) at the end of 3 months followed by metformin, glucose sensitivity declined at 21 months relative to baseline to a greater extent in youth in G/M compared to those treated with metformin alone. In adults, ISRs were significantly lower relative to baseline starting at 12 months in G/M while stable in MET, but there were no significant treatment differences in measures of β-cell function. Lack of a benefit of insulin treatment in the RISE Study could be related to the short duration of insulin treatment in RISE (3 months) compared to years in ORIGIN. Additionally, previous studies that showed benefit in terms of remission and improved β-cell function (26; 27) utilized both basal and prandial insulin while the RISE Study employed basal glargine insulin alone. Thus, it is possible that targeting post-prandial glucose excursions may be necessary to fully rest the β-cell. Finally, RISE allowed for those with early or newly diagnosed type 2 diabetes and perhaps the intervention came too late to reverse the process.

4.4 The first signal of worsening β-cell function in youth was a decrease in total insulin secretion during the OGTT observed as early as 12 months while still on active treatment in the G/M arm and at 15 months in the MET arm, consistent with decreases in steady state C-peptide concentrations during the hyperglycemic clamp at these time points (11). Changes in rate sensitivity and glucose sensitivity were also picked up starting at 15 months. By month 21, all estimates of β-cell function were diminished in youth, including the IGI, which is typically used to assess β-cell function when OGTTs are utilized. In adults, only changes in ISR were noted in the G/M group. These results suggest that use of C-peptide to model insulin secretion and measures of β-cell function, such as glucose sensitivity and rate sensitivity, may provide earlier signals of β-cell failure. While data from the hyperglycemic clamps in RISE also demonstrated a decline in C-peptide responses at 12 and 15 months in youth (11), these procedures are time consuming, costly, and increase participant burden. Use of orally stimulated C-peptide measures with deconvolution techniques to estimate insulin secretion may be a good alternative.

4.5 The strengths of this study lie in the matched treatment arms to directly compare youth and adults with impaired glucose tolerance and early type 2 diabetes, a rigorous protocol design, repeat OGTTs with use of C-peptide and modeling to assess β-cell function and an extended wash-out period to assess for longer-term effects of the intervention. Previous RISE analyses have detailed that compared to adults, youth are more insulin resistant and have greater insulin and C-peptide responses to both intravenous and oral glucose (6–8). While youth and adults differed in several baseline characteristics including sex and race/ethnicity, which were adjusted for in the statistical analysis, they had similar BMI and changes in insulin sensitivity were similar and transient. Limitations include the relatively small sample size of the study and the short duration of basal insulin treatment in the G/M arms. It is possible that a longer period or more intensive insulin treatment with basal/bolus therapy may have had a more beneficial effect, although weight gain occurred with 3 months of insulin treatment in the youth that was never lost, which might have worsened with longer duration of insulin treatment. Additionally, entry criteria for youth allowed treatment on metformin which applied to 25% of youth in the study. This could have biased baseline results in youth to more favorable glucose profiles and thus enhanced any decrement over time when examined off treatment. However, excluding those youth on metformin at baseline did not markedly change the results of the study.

4.6 These data reinforce the progressive nature of the pathology underlying type 2 diabetes and the differences in this progression between youth and adults who were comparable in terms of obesity and glycemia. While insulin sensitivity improved transiently in both youth and adults with treatment, it was not different from baseline at month 21 after treatment wash-out and thus is likely not the major factor driving the observed decline in β-cell function in youth. Even though estimates of β-cell function remained stable in adults, HbA1c increased, so other factors beyond β-cell function are also operative in terms of glycemic control in adults. Unfortunately, treatments studied in youth to date have not reversed the decline in β-cell function and more research into newer treatment options is needed.

Supplementary Material

Supplemental Appendix

ACKNOWLEDGEMENTS

The RISE Consortium acknowledges the support and input of the RISE Data and Safety Monitoring Board and Barbara Linder, the NIDDK Program Official for RISE, and Peter Savage, who served as the Scientific Officer for RISE prior to his retirement. The Consortium is also grateful to the participants who, by volunteering, are furthering the ability to reduce the burden of diabetes.

RISE is supported by grants from the National Institutes of Health (U01DK-094406, U01DK-094430, U01DK-094431, U01DK-094438, U01DK-094467, P30DK-017047, P30DK-020595, P30DK-045735, P30DK-097512, UL1TR-000430, UL1TR-001082, UL1TR-001108, UL1TR-001855, UL1TR-001857, UL1TR-001858, UL1TR-001863), the Department of Veterans Affairs. Additional financial and material support from the American Diabetes Association, Allergan Corporation, Apollo Endosurgery, Abbott Laboratories, and Novo Nordisk A/S is gratefully acknowledged. The content of this publication is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. A complete list of Centers, investigators, and staff can be found in the Appendix.

Abbreviations:

CPI

C-peptide index

FI

fasting insulin

G/M

glargine followed by metformin treatment arm

IGT

impaired glucose tolerance

iAUC

incremental area under the curve

ISR

insulin secretion rate

IGI

insulinogenic index

MET

metformin treatment arm

OGIS

oral glucose insulin sensitivity

OGTT

oral glucose tolerance test

tISR

total insulin secretion rate

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declarations of Interest: KJM is currently employed by Eli Lilly and Co. No other potential conflicts of interest relevant to this article were reported.

DATA SHARING

In accordance with the NIH Public Access Policy, we continue to provide all manuscripts to PubMed Central including this manuscript. The RISE Consortium has provided the protocols to the public through its public website (https://www.risestudy.org). The RISE Consortium abides by the NIDDK data sharing policy and implementation guidance as required by the NIH/NIDDK (HTTPS://WWW.NIDDKREPOSITORY.ORG/STUDIES/RISE).

MTT is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

REFERENCES

  • 1.Mayer-Davis EJ, Dabelea D, Lawrence JM. Incidence Trends of Type 1 and Type 2 Diabetes among Youths, 2002–2012. N Engl J Med 2017;377:301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Imperatore G, Boyle JP, Thompson TJ, Case D, Dabelea D, Hamman RF, Lawrence JM, Liese AD, Liu LL, Mayer-Davis EJ, Rodriguez BL, Standiford D, Group SfDiYS. Projections of type 1 and type 2 diabetes burden in the U.S. population aged <20 years through 2050: dynamic modeling of incidence, mortality, and population growth. Diabetes Care 2012;35:2515–2520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Today Study Group. Effects of metformin, metformin plus rosiglitazone, and metformin plus lifestyle on insulin sensitivity and beta-cell function in TODAY. Diabetes Care 2013;36:1749–1757 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hannon TS, Arslanian SA. The changing face of diabetes in youth: lessons learned from studies of type 2 diabetes. Ann N Y Acad Sci 2015;1353:113–137 [DOI] [PubMed] [Google Scholar]
  • 5.RISE Consortium. Restoring Insulin Secretion (RISE): design of studies of beta-cell preservation in prediabetes and early type 2 diabetes across the life span. Diabetes Care 2014;37:780–788 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.RISE Consortium. Metabolic Contrasts Between Youth and Adults With Impaired Glucose Tolerance or Recently Diagnosed Type 2 Diabetes: I. Observations Using the Hyperglycemic Clamp. Diabetes Care 2018;41:1696–1706 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.RISE Consortium. Metabolic Contrasts Between Youth and Adults With Impaired Glucose Tolerance or Recently Diagnosed Type 2 Diabetes: II. Observations Using the Oral Glucose Tolerance Test. Diabetes Care 2018;41:1707–1716 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Utzschneider KM, Tripputi MT, Kozedub A, Mather KJ, Nadeau KJ, Edelstein SL, Hannon TS, Arslanian SA, Cree-Green M, Buchanan TA, Caprio S, Mari A, Consortium R. Beta-cells in youth with impaired glucose tolerance or early type 2 diabetes secrete more insulin and are more responsive than in adults. Pediatr Diabetes 2020;21:1421–1429 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.RISE Consortium. Effects of Treatment of Impaired Glucose Tolerance or Recently Diagnosed Type 2 Diabetes With Metformin Alone or in Combination With Insulin Glargine on beta-Cell Function: Comparison of Responses In Youth And Adults. Diabetes 2019;68:1670–1680 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 2003;26 Suppl 1:S5–20 [DOI] [PubMed] [Google Scholar]
  • 11.RISE Consortium. Impact of Insulin and Metformin Versus Metformin Alone on beta-Cell Function in Youth With Impaired Glucose Tolerance or Recently Diagnosed Type 2 Diabetes. Diabetes Care 2018;41:1717–1725 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Van Cauter E, Mestrez F, Sturis J, Polonsky KS. Estimation of insulin secretion rates from C-peptide levels. Comparison of individual and standard kinetic parameters for C-peptide clearance. Diabetes 1992;41:368–377 [DOI] [PubMed] [Google Scholar]
  • 13.Mari A, Tura A, Gastaldelli A, Ferrannini E. Assessing insulin secretion by modeling in multiple-meal tests: role of potentiation. Diabetes 2002;51 Suppl 1:S221–226 [DOI] [PubMed] [Google Scholar]
  • 14.Mari A, Schmitz O, Gastaldelli A, Oestergaard T, Nyholm B, Ferrannini E. Meal and oral glucose tests for assessment of β-cell function: modeling analysis in normal subjects. Am J Physiol Endocrinol Metab 2002;283:E1159–1166 [DOI] [PubMed] [Google Scholar]
  • 15.Mari A, Ferrannini E. Beta-cell function assessment from modelling of oral tests: an effective approach. Diabetes Obes Metab 2008;10 Suppl 4:77–87 [DOI] [PubMed] [Google Scholar]
  • 16.Mari A, Pacini G, Murphy E, Ludvik B, Nolan JJ. A model-based method for assessing insulin sensitivity from the oral glucose tolerance test. Diabetes Care 2001;24:539–548 [DOI] [PubMed] [Google Scholar]
  • 17.Utzschneider KM, Prigeon RL, Faulenbach M, Tong J, Carr DB, Boyko EJ, Leonetti DL, McNeely MJ, Fujimoto WY, Kahn SE. Justification for Use of an Oral Disposition Index to Assess β-cell Function Across Glucose Tolerance Categories. Diabetes 2008;57 A46 [Google Scholar]
  • 18.RISE Consortium. Lack of Durable Improvements in beta-Cell Function Following Withdrawal of Pharmacological Interventions in Adults With Impaired Glucose Tolerance or Recently Diagnosed Type 2 Diabetes. Diabetes Care 2019;42:1742–1751 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kahn SE, Lachin JM, Zinman B, Haffner SM, Aftring RP, Paul G, Kravitz BG, Herman WH, Viberti G, Holman RR. Effects of rosiglitazone, glyburide, and metformin on β-cell function and insulin sensitivity in ADOPT. Diabetes 2011;60:1552–1560 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Kahn SE, Haffner SM, Heise MA, Herman WH, Holman RR, Jones NP, Kravitz BG, Lachin JM, O’Neill MC, Zinman B, Viberti G. Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy. N Engl J Med 2006;355:2427–2443 [DOI] [PubMed] [Google Scholar]
  • 21.Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, Nathan DM. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002;346:393–403 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Diabetes Prevention Program Research Group. Effects of withdrawal from metformin on the development of diabetes in the diabetes prevention program. Diabetes Care 2003;26:977–980 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Cusi K, Consoli A, DeFronzo RA. Metabolic effects of metformin on glucose and lactate metabolism in noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1996;81:4059–4067 [DOI] [PubMed] [Google Scholar]
  • 24.DeFronzo RA, Barzilai N, Simonson DC. Mechanism of metformin action in obese and lean noninsulin-dependent diabetic subjects. J Clin Endocrinol Metab 1991;73:1294–1301 [DOI] [PubMed] [Google Scholar]
  • 25.van Raalte DH, Verchere CB. Improving glycaemic control in type 2 diabetes: Stimulate insulin secretion or provide beta-cell rest? Diabetes Obes Metab 2017;19:1205–1213 [DOI] [PubMed] [Google Scholar]
  • 26.Weng J, Li Y, Xu W, Shi L, Zhang Q, Zhu D, Hu Y, Zhou Z, Yan X, Tian H, Ran X, Luo Z, Xian J, Yan L, Li F, Zeng L, Chen Y, Yang L, Yan S, Liu J, Li M, Fu Z, Cheng H. Effect of intensive insulin therapy on beta-cell function and glycaemic control in patients with newly diagnosed type 2 diabetes: a multicentre randomised parallel-group trial. Lancet 2008;371:1753–1760 [DOI] [PubMed] [Google Scholar]
  • 27.Chon S, Rhee SY, Ahn KJ, Baik SH, Park Y, Nam MS, Lee KW, Yoo SJ, Koh G, Lee DH, Kim YS, Woo JT, investigators Ks. Long-term effects on glycaemic control and beta-cell preservation of early intensive treatment in patients with newly diagnosed type 2 diabetes: A multicentre randomized trial. Diabetes Obes Metab 2018;20:1121–1130 [DOI] [PubMed] [Google Scholar]
  • 28.Origin Trial Investigators, Gerstein HC, Bosch J, Dagenais GR, Diaz R, Jung H, Maggioni AP, Pogue J, Probstfield J, Ramachandran A, Riddle MC, Ryden LE, Yusuf S. Basal insulin and cardiovascular and other outcomes in dysglycemia. N Engl J Med 2012;367:319–328 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Appendix
NIHMS1729975-supplement-Supplemental_Appendix.docx (82.9KB, docx)

Data Availability Statement

In accordance with the NIH Public Access Policy, we continue to provide all manuscripts to PubMed Central including this manuscript. The RISE Consortium has provided the protocols to the public through its public website (https://www.risestudy.org). The RISE Consortium abides by the NIDDK data sharing policy and implementation guidance as required by the NIH/NIDDK (HTTPS://WWW.NIDDKREPOSITORY.ORG/STUDIES/RISE).

MTT is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

RESOURCES