Impact of Insulin and MetforminVersusMetformin Alone on b-CellFunction in Youth With ImpairedGlucose Tolerance or RecentlyDiagnosed Type 2 DiabetesDiabetes Care 2018;41:1717–1725 | https://doi.org/10.2337/dc18-0787

OBJECTIVE

Pediatric type 2 diabetes prevalence is increasing, with b-cell dysfunction key inits pathogenesis. The RISE Pediatric Medication Study compared two approachesdglargine followed by metformin and metformin alonedin preserving or improvingb-cell function in youth with impaired glucose tolerance (IGT) or recently diagnosedtype 2 diabetes during and after therapy withdrawal.

RESEARCH DESIGN AND METHODS

Ninety-one pubertal, overweight/obese 10–19-year-old youth with IGT (60%)or type 2 diabetes of <6 months duration (40%) were randomized to either 3 monthsof insulin glargine with a target glucose of 4.4–5.0 mmol/L followed by 9months ofmetformin or to 12 months of metformin alone. b-Cell function (insulin sensitivitypaired with b-cell responses) was assessed by hyperglycemic clamp at baseline,12 months (on treatment), and 15 months (3 months off treatment).

RESULTS

No significant differences were observed between treatment groups at baseline,12 months, or 15 months in b-cell function, BMI percentile, HbA1c, fasting glucose,or oral glucose tolerance test 2-h glucose results. In both treatment groups, clamp-measured b-cell function was significantly lower at 12 and 15 months versusbaseline. HbA1c fell transiently at 6 months within both groups. BMI was higherin the glargine followed by metformin versus metformin alone group between3 and 9 months. Only 5% of participants discontinued the interventions, andboth treatments were well tolerated.

CONCLUSIONS

In youth with IGT or recently diagnosed type 2 diabetes, neither 3 months of glarginefollowed by 9 months of metformin nor 12 months of metformin alone haltedthe progressive deterioration of b-cell function. Alternate approaches to preserveb-cell function in youth are needed.

The SEARCH for Diabetes in Youth (SEARCH) epidemiological study has highlighted acontinued increase in the incidence of type 2 diabetes in adolescents aged 10–19years, particularly among ethnic minority groups. The Treatment Options forType 2 Diabetes in Adolescents and Youth (TODAY) study demonstrated that in

RISE Coordinating Center, Rockville, MD

Corresponding author: Sharon L. Edelstein, rise@bsc.gwu.edu.

Received 10 April 2018 and accepted 13 May2018.

Clinical trial reg. no. NCT01779375, clinicaltrials.gov.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc18-0787/-/DC1.

*A complete list of the RISE Consortium Inves-tigators can be found in the Supplementary Dataonline.

© 2018 by the American Diabetes Association.Readers may use this article as long as the workis properly cited, the use is educational and notfor profit, and the work is not altered. More infor-mation is available at http://www.diabetesjournals.org/content/license.

See accompanying articles, pp. 1560,1696, and 1707.

The RISE Consortium*

Diabetes Care Volume 41, August 2018 1717

PATH

OPHYSIO

LOGY/CO

MPLIC

ATIO

NS

50% of youth with type 2 diabetes,initial assigned treatment failed, re-quiring the initiation of insulin ther-apy after an average follow-up of 3.9years, which is in keepingwith faster dis-ease progression than reported in adults(1–3). In TODAY, lower b-cell functionat baseline as well as a greater decline inb-cell function over time predicted worseglycemic control and treatment failure(4). Thus, interventions to preserve orimprove b-cell function in youth arecritically needed.Metformin and insulin each has been

shown to improve b-cell function inadults with impaired glucose tolerance(IGT) or recent-onset type 2 diabetes. TheDiabetes Prevention Program (DPP) dem-onstrated that metformin improves b-cellfunction and reduces diabetes progres-sion by 31% over 3 years in adults withIGT (5,6).Weng et al. (7) showed in adultswith newly diagnosed type 2 diabetesthat,2 weeks of intensive insulin ther-apy improves and maintains b-cell func-tion, resulting in prolonged remissionfrom requiring diabetes medication.Whether such interventions have similarbenefits in youth is unclear. The Re-storing Insulin Secretion (RISE) PediatricMedication Study was designed as aproof-of-principle trial to test in youthacross a continuum of b-cell dysfunctionfrom IGT to mild, recently diagnosedtype 2 diabetes the hypothesis that initialshort-term treatment with insulin glarginefor 3 months followed by metformin for9 months would preserve or improve b-cellfunction compared with metformin alone,with a sustained effect after withdrawalof therapy.

RESEARCH DESIGN AND METHODS

Study ProtocolThe RISE Pediatric Medication Studywas a four-center, randomized, open-label clinical trial funded by the NationalInstitute of Diabetes and Digestive andKidneyDiseases. The rationale andmeth-ods have been described in detail else-where (8–10); additional specifics areprovided in the Supplementary Data.Each center’s institutional review boardapproved the protocol. Written parentalconsent and child assent were obtainedconsistent with the Declaration of Helsinkiand each center’s institutional reviewboard guidelines. The protocol can befoundat https://rise.bsc.gwu.edu/web/rise/collaborators.

ParticipantsA summary of study visits is shown inSupplementary Fig. 1, and a summary ofparticipants enrolled and their rates ofcompletion of study visits is shown in aCONSORT (Consolidated Standards of Re-porting Trials) diagram (SupplementaryFig. 2). Enrollment occurred betweenJuly 2013 and April 2016. Youth aged10–19 years with a BMI $85th percen-tile for age and sex but,50 kg/m2, IGTor recently diagnosed (,6 months dura-tion) type 2 diabetes, and Tanner stage$II (females, Tanner stage $II breastdevelopment; and males, testicular vol-ume .3mL) were eligible. Screening in-cluded a history and physical examination,laboratory tests, and a 75-g oral glucosetolerance test (OGTT). Eligibility criteriaincluded the simultaneous presence ofall the following: negative GAD and IA2autoantibodies, fasting glucose$5 mmol/L,OGTT 2-h glucose $7.8 mmol/L, andHbA1c #8.0% (64 mmol/mol) if drugnaıve. In youth with type 2 diabetes al-ready taking metformin, eligibility wasmodified to HbA1c #7.5% (58 mmol/mol)if on metformin for ,3 months and#7.0% (53 mmol/mol) if on metforminfor 3–6 months.

Final eligibility required $80% com-pliance with 3 weeks of placebo medi-cation (or study-provided metformin ifalready taking metformin) and atten-dance at scheduled run-in visits. Eligi-ble participants underwent a baselinehyperglycemic clamp and OGTT followedby random 1:1 treatment assignment bystudy site, stratified by glycemic status.

InterventionsThe glargine followed by metformingroup received 3 months of insulin glar-gine titrated at least twice weekly by apreset algorithm (Supplementary StudyMethods) to achieve a fasting glucoseof 4.4–5.0 mmol/L on the basis of dailyself-monitored blood glucose (SMBG)(Freestyle Lite, Abbott Laboratories, LakeBluff, IL), which was followed immediatelyby 9 months of metformin (titrated to1,000 mg twice daily). Youth alreadytaking metformin and randomized toglargine therapy discontinued activemetformin at the time of initiation ofglargine treatment. The metformin groupreceived metformin alone (titrated to1,000 mg twice daily) for 12 months.

Participants took their assigned inter-vention for 12 months, after which it

was withdrawn, and follow-up continuedwhile off study medication for an additional3 months to determine the 15-monthprimary outcome of treatment durability.Anthropometrics, HbA1c, and safety eval-uations were performed every 3 months,and hyperglycemic clamps and OGTTswere repeated at 12 and 15 months(8–10) (Supplementary Fig. 1). Labora-tory assessments were performed in acentral laboratory at the University ofWashington as summarized below anddescribed in greater detail elsewhere (9).

Medication adverse effects, symptomsof hypoglycemia and hyperglycemia, andadherence (returned medication bottlepill count, returned insulin pen residualvolume) were assessed at all study visits. Adetailed algorithm was used for address-ing hypoglycemia, hyperglycemia, andacute metabolic decompensation at eachvisit (Supplementary Study Methods). Ifprotocol-specified HbA1c safety thresh-olds were exceeded at any visit, outcomeassessments were performed wheneverpossible. The participant was then with-drawn from the study and referred foradditional diabetes treatment as appropri-ate. Safety was monitored by an indepen-dent data and safety monitoring board.

Procedures and CalculationsA two-step hyperglycemic clamp (11.1mmol/L, .25 mmol/L) paired with theinsulin secretagogue arginine at thesecond step was performed at baseline,12 months, and 15 months as previouslydescribed (8–10). If the participant wason metformin, the last dose of metforminwas taken the evening before the hy-perglycemic clamp. The hyperglycemicclamp was used to quantify simulta-neously insulin sensitivity (M/I) and threeb-cell response measures: 1) steady-state (second-phase) C-peptide, 2) acuteC-peptide response to arginine atmaximalglycemic potentiation (ACPRmax), and 3)acute (first-phase) C-peptide responseto glucose (ACPRg). M/I was calculatedas the mean glucose infusion rate (M) at100, 110, and 120 min of the clampdivided by the corresponding mean steady-state plasma insulin concentration (I)(11–13). Mean steady-state (second-phase) C-peptide and insulin concentra-tions were calculated at 100, 110, and120min (13). ACPRmax was calculated asthe mean incremental response aboveconcentrations before the arginine bolus(14) and ACPRg as the mean incremental

1718 b-Cell Function in Youth With Dysglycemia Diabetes Care Volume 41, August 2018

response above baseline for the first 10min after glucose bolus (15).The 15-month (3 months after discon-

tinuation of metformin) steady-state C-peptide and ACPRmax, each paired with15-month insulin sensitivity, represent thecoprimary outcomes. These were pairedto express the magnitude of the b-cell re-sponse as a function of the prevailing insulinsensitivity (15). Secondary outcomes fromthe hyperglycemic clamp included the 15-month ACPRg paired with the 15-monthM/I and the 12-month C-peptide responsemeasures (steady-stateC-peptide,ACPRmax,ACPRg) each paired with M/I at the sametime point.A 3-h 75-g OGTT also was performed

on a separate day and used to determinefasting and 2-h glucose at baseline, 12months, and15months. If theparticipantwas on metformin, the last dose ofmetformin was taken the evening beforethe OGTT. Additional HbA1c testing wasdone at baseline and 3, 6, 9, 12, and 15months for robust phenotyping of par-ticipants over time.

AssaysGlucose was measured using the glu-cose hexokinase method on a Rochec501 autoanalyzer. C-peptide and insulinwere measured by a two-site immunoen-zymometric assay performed on the Tosoh2000 autoanalyzer (Tosoh Bioscience, Inc.,South San Francisco, CA). Interassay co-efficients of variation on quality controlsamples with low, medium, medium-high,and high concentrations were #2.0% forglucose,#4.3% for C-peptide, and#3.5%for insulin.

StatisticsData were accumulated centrally, andanalyses were performed according to aprespecified analysis plan. For this proof-of-principle trial, we chose two mea-sures of b-cell function as coprimaryoutcomes: clamp-derived steady-stateC-peptide and ACPRmax at 15 months(both evaluated jointly with M/I). Thesetwo measures evaluate different compo-nents ofb-cell function andwere selectedbecause there were no data in youth todeterminewhetheroneor theother waslikely to change after the interventions.No preliminary data were available in

youth on the expected differences ineither primary outcome after a periodof intervention withdrawal (washout).Several unpublished analyses provideda range of possible differences between

individuals taking medication and thosetaking placebo as well as expected corre-lations between baseline and follow-upmeasures. Therefore, sample size wasbased on aminimum effect size equal tothe difference between the two groupsdivided by the SD in favor of the glarginefollowed by metformin group at a powerof 80%. Because we targeted improve-ments in b-cell function, a one-sidedsignificance level of 0.05 to detect a ben-efit of glargine followed by metforminwas used for each comparison, with aconservative correlation of 0.5 betweenbaseline and follow-up measures. Thirty-nine participants per arm (78 total) at theend of the washout provided$80% powerto detect a minimum effect size of 0.57in favor of glargine followed by metfor-min for either primary outcomemeasure-ment using baseline-adjusted ANCOVA.Ninety participants were sought, allowingfor an ;10% loss to follow-up.

The primary analysis was the compar-ison of b-cell responses paired with M/Ibetween treatment groups at 15 months,adjusted for baseline b-cell response andM/I. Jointmodels forb-cell response andM/I were fit simultaneously using seem-ingly unrelated regression techniques(16–18), which provide a two degreesof freedom x2 test of the treatment armdifference in the joint values of b-cellresponse and M/I between treatmentgroups. The same methods were used tocompare treatment groups at 12 months.

To evaluate changeswithin each treat-ment arm over time, Hotelling T2 distri-bution was used to test changes in b-cellresponses and M/I simultaneously (19).Additional secondary analyses comparedmeans across treatment arms at specifictime points by using ANCOVA adjustedfor the baseline value of the measure andby repeated-measures ANCOVA across theactive 12-month intervention period ad-justed for thebaseline value of themeasure.

Two participants had an HbA1c eleva-tion requiring study withdrawal beforetheir 15-month primary outcome visit. Theclamp-derived outcome data for theseparticipants were imputed using the mid-point between zero and the lowest valuein their respective treatment arm.

RESULTS

Demographic, Physical, and MetabolicCharacteristicsWe screened 236 youth and randomized91 (Supplementary Fig. 2). Table 1 lists

baseline demographic and metaboliccharacteristics of participants by treat-ment arm. The metformin alone groupwas slightly younger but not differentin Tanner stage. All other baseline phys-ical, metabolic, b-cell response, and in-sulin sensitivity measures were similarbetween treatment groups. Of the 91participants, 21 with type 2 diabeteswere receiving metformin at random-ization or were previously exposed to it,with a similar distribution of metforminexposure and IGT versus type 2 diabetesstatus between treatment groups.

Adherence to InterventionsDuring weeks 11 and 12, the mean6 SDglargine dose in the glargine followedbymetformin group was 0.76 0.4 units/kg/day, and fasting blood glucose haddecreased from a mean of 6.0 6 0.8mmol/L at the baseline OGTT to a fastingSMBG mean of 5.2 6 0.7 mmol/L, with50% achieving the 4.4–5.0 mmol/L goal(Fig. 1). As assessed by residual volume,68% of participants were.80% compliantwith the insulin. Metformin adherenceby pill count was 88 6 18% over theentire study, not differing by treatmentgroup.

Treatment Effects on TemporalChanges in b-Cell FunctionFigure 2 illustrates glucose and C-peptideconcentrations during the hyperglyce-mic clamps performed at baseline, 12months, and 15 months. Of note, iden-tical target glucose concentrations of11.1 and .25 mmol/L were achievedand not different between treatmentgroups at baseline, 12 months, and 15months. Supplementary Table 1 showsb-cell responses and M/I by treatmentgroup.

No significant differences were foundbetween treatment groups at 12 or 15months in either primary measure ofb-cell function (e.g., response [steady-state C-peptide, ACPRmax] paired withM/I) or in the secondary measure ofb-cell function (ACPRg paired with M/I).Figure 3 shows a vector plot of changesfrom baseline to 12 and 15 months foreach of the two coprimary clamp-derivedb-cell responses (steady-state C-peptide,ACPRmax) paired with M/I, and Supple-mentary Fig. 3 shows the vector plot forthe secondary b-cell response (ACPRgpaired with M/I). Explanations of thevector plots appear in the respective

care.diabetesjournals.org The RISE Consortium 1719

figure legends. Two participants in themetformin alone arm met HbA1c criteriafor study withdrawal at 9 months; thusthey had a 12-month visit but did notcomplete the 15-month visit. The inclu-sion of imputed values for their b-cellresponses and M/I at 15 months did notalter the results.Neither treatment improved nor main-

tained b-cell function at the end ofor after active treatment (Fig. 3 andSupplementary Fig. 3). Rather, b-cellfunction declined significantly frombaseline while on active treatment

through 12months within each treatmentgroup for steady-state C-peptide withM/I (glargine followed by metformin P =0.019, metformin alone P = 0.025) andACPRmax with M/I (glargine followed bymetformin P, 0.001, metformin aloneP = 0.001). b-Cell function remainedsignificantly worse at 15 months com-pared with baseline within both treat-ments for steady-state C-peptide withM/I (glargine followed by metforminP , 0.001, metformin alone P = 0.031),ACPRmax with M/I (glargine followed bymetformin P = 0.001, metformin alone

P , 0.001), and ACPRg with M/I (glarginefollowed by metformin P , 0.001, met-formin alone P = 0.014).

Approximately 60% of participants hadIGT at baseline; by study design, none hadreceived metformin before randomiza-tion. In this IGT subgroup, we observedpatterns similar to those seen in the fullcohort, with no evidence of a beneficialeffect on b-cell function of glargine fol-lowed by metformin versus metfor-min alone, and a worsening of b-cellfunction through 12 and 15 months inboth treatment groups (Supplemen-tary Fig. 4).

Temporal Patterns of BMI andGlycemic Measures Over 15 MonthsSeveral transient changes were seenin body size and glycemic parameters.BMI did not differ between treatmentgroups at 12 or 15 months. However,BMI declined from baseline within themetformin alone group over the first9 months, resulting in a significantly lowerBMI across the 12-month interventionperiod in the metformin alone versusglargine followed by metformin group(P = 0.008) and specifically at 3 (P, 0.001),6 (P , 0.02), and 9 (P , 0.04) months.

HbA1c did not significantly differ be-tween treatment groups across the12-month intervention period or at anyindividual time point (Fig. 4). However,compared with baseline, HbA1c de-clined significantly within the glarginefollowed by metformin group at 3 monthswhile on glargine treatment (P , 0.03)and within both groups at 6 months(glargine followed by metformin P =0.001, metformin alone P = 0.01) beforerebounding to baseline levels in bothgroups at months 9 and 12, despitecontinued active metformin treatment.By 15 months, HbA1c had increased sig-nificantly from baseline in the glarginefollowed by metformin group (P = 0.03),ending at ;6% (42 mmol/mol) in bothtreatment groups.

No differences were found betweentreatment groups in fasting or OGTT 2-hglucose concentrations at 12 or 15 months.Within the glargine followed by metformingroup, fasting glucose was not different at12 months compared with baseline butrose higher than baseline by 15months(off treatment) (P = 0.02). Within theglargine followed by metformin group,2-h glucose was significantly lower thanbaseline at 12 months (P = 0.02) but

Table 1—Baselinedemographic andphysical characteristics, insulin sensitivity,andb-cell responses fromthehyperglycemic clampandOGTTby treatmentgroup

Glargine followed bymetformin (n = 44)

Metformin alone(n = 47)

Female, n (%) 27 (61.4) 38 (80.9)

Age (years) 14.9 6 2.0 13.9 6 2.1*

Tanner stage V, n (%) 32 (72.7) 25 (53.2)

Race/ethnicity, n (%)White 13 (29.5) 12 (25.5)Black 14 (31.8) 9 (19.1)Hispanic 14 (31.8) 20 (42.6)Other 3 (6.8) 6 (12.8)

Weight (kg) 102.0 6 25.7 97.7 6 23.3

BMI (kg/m2) 36.5 6 6.4 36.9 6 6.4

BMI percentile 98.4 6 2.5 98.8 6 1.3

BMI z score 2.3 6 0.4 2.4 6 0.3

Waist circumference (cm) 109.8 6 15.8 109.6 6 12.4

HbA1c% 5.7 6 0.6 5.7 6 0.6mmol/mol 39.2 6 6.6 38.6 6 6.3

IGT, n (%) 26 (59) 28 (60)

Type 2 diabetes, n (%) 18 (41) 19 (40)

Metformin use at baseline, n (%) 10 (22.7) 11 (23.4)

Systolic blood pressure (mmHg) 120.7 6 7.8 119.5 6 8.7

Diastolic blood pressure (mmHg) 67.6 6 7.7 70.1 6 7.9

Hypertension, n (%) 9 (20.5) 15 (31.9)

Triglycerides (mmol/L) 1.11 (0.39, 3.14) 1.17 (0.45, 3.06)

HDL cholesterol (mmol/L) 1.0 6 0.3 1.0 6 0.2

LDL cholesterol (mmol/L) 2.3 6 0.8 2.1 6 0.6

Hyperglycemic clamp variablesFasting glucose (mmol/L) 6.04 6 0.85 6.11 6 1.10Fasting C-peptide (nmol/L) 1.63 6 0.55 1.82 6 0.58Fasting insulin (pmol/L) 211.1 (54.8, 813.5) 247.4 (74.1, 825.4)Steady-state C-peptide (nmol/L) 5.22 (2.44, 11.15) 5.09 (2.32, 11.19)ACPRmax (nmol/L) 7.29 (3.29, 16.16) 8.12 (3.42, 19.26)ACPRg (nmol/L) 1.06 (0.10, 11.66) 1.14 (0.14, 9.02)Glucose infusion rate (M) (mmol/kg/min) 0.025 6 0.013 0.023 6 0.010Steady-state insulin (I) (pmol/L) 1,391 (281, 6,874) 1,371 (286, 6,572)M/I (3 10e25 mmol/kg/min per pmol/L) 1.60 (0.35, 7.38) 1.54 (0.35, 6.76)

OGTT variablesFasting glucose (mmol/L) 6.0 6 0.8 6.1 6 1.12-h glucose (mmol/L) 10.2 6 2.5 10.2 6 2.8

Data are mean 6 SD or geometric mean (95% CI) for nonnormally distributed variables. *P ,0.03; all other P values were not significant.

1720 b-Cell Function in Youth With Dysglycemia Diabetes Care Volume 41, August 2018

no longer different from baseline at15 months. Within the metformin alonegroup, no significant changes were foundin fastingor2-hglucoseat12monthswhileon active treatment or at 15 months.

Safety Outcomes

Seven serious adverse events occurred(one hospitalization each for suicidalideation/anxiety, newly diagnosed Ewingsarcoma, otitis externa, appendicitis/appendectomy, and pneumonia andtwo hospitalizations for tonsillectomy/adenoidectomy), all deemed unrelatedto the study interventions. Other non-serious adverse events are summarized inSupplementary Table 2. No participantshad severe hypoglycemia or acute meta-bolic decompensation as defined in Sup-plementary Study Methods. Accordingly,

no participants required rescue therapywith insulin.

Over the 15 months, three participantsin the glargine followed by metformingroup with IGT at baseline and none inthe metformin alone group with IGT atbaseline developed diabetes as definedby OGTT criteria and HbA1c $6.5%(48 mmol/mol) (20). In each treatmentgroup, three participants met hypergly-cemic criteria for study withdrawal. Thelast outcome visits for these partici-pants were three at 15 months in theglargine followed by metformin groupand two at 12 and one at 15 months inthe metformin alone group, after whichthe participants were withdrawn fromthe study and returned to their primarydiabetes provider for additional diabe-tes treatment.

CONCLUSIONS

In youth along the glycemic continuumof IGT to mild, recently diagnosed type2 diabetes, we tested whether initial short-term treatment with insulin glargine for3 months followed by metformin for 9months would preserve or improve b-cellfunction comparedwithmetformin alonefor 12months and a sustained effect afterwithdrawal of therapy. No significant dif-ferences were found between groups inb-cell function at 12 or 15months, and allb-cell measures at 12 months (on treat-ment) and 15 months (off treatment) wereworse than at baseline. Thus, the only twoagents approved for treating type 2 di-abetes in youth both failed to preserve orimprove b-cell function during or afterwithdrawal of treatment in youth with IGTor recently diagnosed type 2 diabetes.

Trials of insulin therapy in adults withIGT or early type 2 diabetes support theconcept that early use of insulin mayhave a beneficial long-term effect ondysglycemia. In the Outcome ReductionWith Initial Glargine Intervention (ORIGINTrial), adults with high cardiovascular dis-ease risk and prediabetes were treatedwith insulin glargine to a similar targetglucose (#5.3 mmol/L) as we achieved(21). A median of 3.3 months after in-sulin discontinuation, a similar durationof durability as we evaluated, those withprediabetes randomized to glargine treat-ment had a 20% reduction in the odds ofdeveloping diabetes on the basis of OGTTcriteria comparedwith standard care (21).In adults with recently diagnosed type2 diabetes, Weng et al. (7) found that upto 2 weeks of intensive insulin therapyas either multiple daily injections or con-tinuous subcutaneous infusion, an evenshorter treatment duration than wetested, resulted in remission and noneed for diabetes medications 1 yearlater in 45% and 51% of participants,respectively. Furthermore, after the2-week insulin treatment, the b-cell re-sponse increased, and this improvementwas sustained in the participants who re-mained in remission while off medications.Although the study by Weng et al. sug-gested benefit from,2 weeks of insulintreatment, we saw no effect from a3-month duration of treatment, whichis longer than that usedbyWenget al. yetbrief enough to realistically consider inyouth. Moreover, we chose a targetglucose lower than that used in the

Figure 1—Insulin doses and corresponding fasting glucose values over time while on glarginetreatment. A: Mean fastingmorning SMBG values every 2 weeks over 12 weeks. B: Mean glarginedose corresponding to each SMBG value. C: Corresponding percentage of participants whoachieved the goal fasting SMBG of 4.4–5.0 mmol/L every 2 weeks. Data are mean (95% CI).

care.diabetesjournals.org The RISE Consortium 1721

ORIGIN Trial to maximize the effect of theinsulin treatment. To further increasethe likelihood of insulin having an effect,we added the oral glucose-loweringagent metformin shown in the DPP tobe effective in preventing the develop-ment of diabetes in adults with IGT (6). Incontrast to the adult studies, we foundno durable effect of insulin glargine tar-geted to achieve a plasma glucose of4.4–5.0 mmol/L followed by metforminon measures of b-cell function or glycemiain youth. Furthermore, the lack of a ben-eficial effect of metformin alone on b-cellfunction, even when youth were still onthe medication at 12 months, contrastswithobservations after 1 year of therapyin adults with IGT (5) or type 2 diabetes(22). Although a longer duration of in-sulin treatment may have had more ofan effect onb-cell function, longer-termtreatment also may have led to additional

weight gain, with an uncertain net bal-anceof benefit andharm.Moreover, theinjections and glucose monitoring re-quired for longer-term insulin treatmentmight be met with resistance by youth,particularly those who are not yet di-agnosed with diabetes.

The RISE Consortium (9) reportedthat youth with IGT or recently diag-nosed, drug-naıve type 2 diabetes areprofoundly more insulin resistant, havegreater insulin responses for any de-gree of insulin sensitivity, and havelower hepatic insulin clearance thanadults. This finding, combined with thepoor response of youth to the RISEinterventions, suggests that althoughb-cell dysfunction is the foundationfor dysglycemia in both age-groups,greater b-cell demand may place youthat a higher risk for more rapid b-celldamage. Additional support for this

interpretation comes from the appar-ently more rapid progression of type2 diabetes observed in youth in theTODAY study than in a similar studyin adults treated with the sameglucose-lowering medications (2,3). Fur-thermore, the current observation ofa lack of a beneficial effect of metfor-min on b-cell function in youth after12 months of therapy contrasts theimprovement in b-cell function ob-served in adults with IGT (5) or type 2diabetes (22). The RISE Adult Medica-tion Study, which includes these sametwo intervention arms as part of a four-arm study, should provide additionalinsight into these differences (8).

The current study has several strengths,including the randomized design anda robust, multicenter approach toquantification of insulin sensitivity andb-cell responses to both glucose and

Figure 2—Glucose and C-peptide values during the hyperglycemic clamp at baseline, after 12 months of treatment, and 3 months after discontinuingthe intervention (15 months). The glargine followed by metformin group (n = 44 [green]) and the metformin alone group (n = 47 [brown]) are shown.The steady-state glucose targets were 11.1 mmol/L between 90 and 120 min and .25 mmol/L at 150 min. Data are mean 6 SEM.

1722 b-Cell Function in Youth With Dysglycemia Diabetes Care Volume 41, August 2018

nonglucose secretagogue arginine per-formed to provide thorough, longitu-dinal phenotyping of participants. Byquantifying both insulin sensitivity andb-cell responses simultaneously, wewereable to account for the well-recognizedrelationship between insulin sensitivityand b-cell responsivity (15) and thusgain mechanistic insight into the effectof the interventions on b-cell functionover time. Of key importance for this

multicenter study, we reached matchinglevels of hyperglycemia during each ofthree repeated hyperglycemic clamptests, showing a high degree of bothreproducibility and reliability for the out-come measures across time points andacross study sites. Study retention andmetformin adherence alsowere excellentdespite adolescents presenting a challeng-ing study population. Limitations includethe absence of a placebo arm because of

the inclusion of youth with type 2 diabeteswho could not remain untreated. Althougha decline in b-cell function might havebeen even greater in a placebo group, themarked decline in b-cell function evidentduring andafter stopping the interventionssuggests that themedications largely weremasking disease progression. In addition,despite the high reported insulin doses,50% of the participants in the insulinglargine followed by metformin arm didnot fully meet the fasting goal of 4.4–5.0 mmol/L, regardless of IGT or type 2diabetes status. Of note, adults with pre-diabetes or type 2 diabetes in the ORIGINTrial were treated with insulin glargine toa target fasting glucose of #5.3 mmol/L,a slightly less-aggressive target than wechose, yet we still observed no significantbeneficial effect onb-cell function despitereaching a mean fasting glucose of 5.2mmol/L (21).Moreover, after 1 year in theORIGIN Trial, only 50% of the adult par-ticipants achieved the targeted glucose,similar to the proportion of youth in thecurrent study, yet positive effects stillwere seen on diabetes development inthese adults (21). Finally, the sample sizewas too small to obtain reliable estimatesof the effects of the interventions in thesubgroups of IGT versus type 2 diabetes.However, in a sensitivity analysis of thesubgroup with IGT who had not receivedmetformin before randomization and com-prised the majority of participants in eacharm, results were similar to the full cohort(Supplementary Fig. 4). Specifically, in theIGT group, treatment with glargine fol-lowed by metformin did not have a ben-eficial effect on b-cell function outcomescompared with metformin alone, andb-cell function declined within both treat-ment groups.

In conclusion, neither 3 months of in-sulin glargine followed by 9 months ofmetformin nor 12 months of metforminalone improved or preserved b-cell func-tion when administered to youth with IGTor recently diagnosed type 2 diabetes. b-Cell function deteriorated with both treat-ments, the only treatments approved foruse in type 2 diabetes in youth, highlight-ing an urgent need for alternate approachesto preserve b-cell function in youth andprevent the progression of dysglycemia.

Appendix

Writing Group: Kristen J. Nadeau (chair),Tamara S. Hannon, Sharon L. Edelstein,Silva A. Arslanian, Sonia Caprio, EllenW.

Figure 3—Relationship of the two coprimary outcomes: hyperglycemic clamp-derived b-cellresponses (steady-state C-peptide and ACPRmax) paired with M/I. Changes are shown frombaseline to 12 and 15months for the clamp-derived b-cell responses of steady-state C-peptide(A) and ACPRmax (B) pairedwithM/I. The black line depicts the joint relationship betweenb-cellresponse and M/I at baseline for the full cohort, with the mean value at baseline for the fullcohort indicated by the black box with a 0. The dotted lines to boxes at months 12 and 15 showthe trajectory of values from baseline to 12 months of intervention and then to 3 months afterdiscontinuation of the intervention (15months) for glargine followedbymetformin (green) andmetformin alone (brown). Values above the black line represent improved b-cell function, andvalues below the line represent poorer b-cell function. The ellipses depict the 95% confidencebands around the points at months 12 and 15. No significant differences were observed at anytime point between treatment arms; however, significant within-group declines were seenfrom baseline while on active treatment through 12 months for steady-state C-peptide withM/I (glargine followed by metformin P = 0.019, metformin alone P = 0.025) and ACPRmax withM/I (glargine followed by metformin P , 0.001, metformin alone P = 0.001) and through15 months for steady-state C-peptide with M/I (glargine followed by metformin P , 0.001,metformin alone P = 0.031) and ACPRmax withM/I (glargine followed by metformin P = 0.001,metformin alone P , 0.001).

care.diabetesjournals.org The RISE Consortium 1723

Leschek, Philip S. Zeitler, Thomas A.Buchanan, David A. Ehrmann, KierenJ. Mather, and Steven E. Kahn.

Acknowledgments. The RISE Consortium thanksthe RISE Data and Safety Monitoring Board andBarbara Linder, the National Institute of Diabetesand Digestive and Kidney Diseases Program Offi-cial for RISE (Rockville, MD), for support andguidance. The Consortium also thanks the par-ticipants who, by volunteering, are furthering theability to reduce the burden of diabetes.Funding and Duality of Interest. RISE is sup-ported by grants from the National Institutesof Health (U01-DK-094406, U01-DK-094430, U01-DK-094431, U01-DK-094438, U01-DK-094467,P30-DK-017047, P30-DK-020595, P30-DK-045735,P30-DK-097512, UL1-TR-000430, UL1-TR-001082,UL1-TR-001108, UL1-TR-001855, UL1-TR-001857,UL1-TR-001858, UL1-TR-001863), the Depart-ment of Veterans Affairs, and Kaiser PermanenteSouthern California. Additional financial and ma-terial support from the American Diabetes Asso-ciation, Allergan, Apollo Endosurgery, AbbottLaboratories, and Novo Nordisk was received.S.A.A. and S.E.K. serve as paid consultants onadvisory boards for Novo Nordisk. S.A.A. isa participant in a Novo Nordisk–sponsoredclinical trial. T.A.B. has received research supportfrom Allergan and Apollo Endosurgery. K.J.M.holds an investigator-initiated research grantfrom Novo Nordisk. No other potential conflictsof interest relevant to this article were reported.Author Contributions. Members of the RISEConsortium recruited participants and collectedstudy data. K.J.N. wrote the first draft. K.J.N.,

S.L.E., S.A.A., S.C., E.W.L., T.A.B., D.A.E., K.J.M.,and S.E.K. (chair) comprise the steering commit-tee (principal investigator at each site, the datacoordinating center, and the National Instituteof Diabetes and Digestive and Kidney Diseasesproject scientist) and designed and imple-mented the study, contributed to the discus-sion, and edited the manuscript. K.J.N., T.S.H.,S.L.E., S.A.A., S.C., E.W.L., P.S.Z., T.A.B., D.A.E., K.J.M.,and S.E.K. researched data. S.L.E. performedall data analyses. K.J.N., S.L.E., and S.E.K. are theguarantors of this work and, as such, had fullaccess to all the data in the study and takeresponsibility for the integrity of the data and theaccuracy of the data analysis.Prior Presentation. Parts of this study werepresented in oral form at the 78th ScientificSessions of the American Diabetes Association,Orlando, FL, 22–26 June 2018.

References1. Mayer-Davis EJ, Lawrence JM, Dabelea D,et al.; SEARCH forDiabetes inYouthStudy. Incidencetrends of type 1 and type 2 diabetes among youths,2002-2012. N Engl J Med 2017;376:1419–14292. TODAY Study Group. A clinical trial to main-tain glycemic control in youth with type 2 di-abetes. N Engl J Med 2012;366:2247–22563. Kahn SE, Haffner SM, HeiseMA, et al.; ADOPTStudy Group. Glycemic durability of rosiglitazone,metformin, or glyburide monotherapy. N Engl JMed 2006;355:2427–24434. TODAY Study Group. Effects of metformin,metformin plus rosiglitazone, and metformin pluslifestyle on insulin sensitivity and b-cell functionin TODAY. Diabetes Care 2013;36:1749–17575. Diabetes Prevention Program ResearchGroup. Role of insulin secretion and sensitivity

in the evolution of type 2 diabetes in the Dia-betes Prevention Program: Effects of lifestyleintervention and metformin. Diabetes 2005;54:2404–24146. Diabetes Prevention Program Research Group.Reduction in the incidence of type 2 diabeteswith lifestyle intervention or metformin. N EnglJ Med 2002;346:393–4037. Weng J, Li Y, Xu W, et al. Effect of intensiveinsulin therapy on b-cell function and glycaemiccontrol in patients with newly diagnosed type 2diabetes: a multicentre randomised parallel-group trial. Lancet 2008;371:1753–17608. The RISE Consortium. Restoring Insulin Secre-tion (RISE): Design of studies of b-cell preservationin prediabetes and early type 2 diabetes acrossthe life span. Diabetes Care 2014;37:780–7889. The RISE Consortium. Metabolic contrastsbetween youth and adults with impaired glucosetolerance or recently diagnosed type 2 diabetes:I. Observations using the hyperglycemic clamp.Diabetes Care 2018;41:1696–170610. Hannon TS, Kahn SE, Utzschneider KM, et al.;The RISE Consortium. Review of methods formeasuringb-cell function:Design considerationsfrom the Restoring Insulin Secretion (RISE) Con-sortium. Diabetes Obes Metab 2018;20:14–2411. Andres R, Swerdloff R, Posefsky T, ColemanDL. Manual feedback technique for the controlof blood glucose concentration. In Automationin Analytical Chemistry. Skeegs LT Jr., Ed. NewYork, Mediad, 1966, p. 486–49112. DeFronzo RA, Tobin JD, Andres R. Glucoseclamp technique: a method for quantifying in-sulin secretion and resistance. Am J Physiol 1979;237:E214–E22313. Elahi D. In praise of the hyperglycemic clamp.Amethod for assessment ofb-cell sensitivity and

Figure 4—BMI and glycemia over time. BMI (A), HbA1c (B), fasting glucose (C), and OGTT 2-h glucose over 15 months (D). In the glargine followedby metformin group (n = 44 [green]), glargine insulin was given from baseline to 3 months followed by metformin from months 3 through 12. In themetformin alone group (n = 47 [brown]), metformin was given from baseline through 12 months. Treatment was discontinued in both groupsat 12 months. Data are mean 6 SEM. *P , 0.05 for the difference between treatment groups at the specified time point; §P , 0.05 for thedifference within treatment group between the specified time point and baseline.

1724 b-Cell Function in Youth With Dysglycemia Diabetes Care Volume 41, August 2018

insulin resistance. Diabetes Care 1996;19:278–28614. Ward WK, Bolgiano DC, McKnight B, HalterJB, Porte D Jr. Diminished B cell secretorycapacity in patients with noninsulin-dependentdiabetes mellitus. J Clin Invest 1984;74:1318–132815. Kahn SE, Prigeon RL, McCulloch DK, et al.Quantification of the relationship between in-sulin sensitivity and b-cell function in humansubjects. Evidence for a hyperbolic function. Dia-betes 1993;42:1663–1672

16. Zellner A. An efficient method of estimatingseemingly unrelated regression equations andtests for aggregation bias. J Am Stat Assoc 1962;57:348–36817. Srivastava VK, Giles DEA. Seemingly Un-related Regression Equations Models: Estimationand Inference. New York, Marcel Dekker, 198718. Henningsen A, Hamann J. Systemfit: a pack-age for estimating systems of simultaneous equa-tions in R. J Stat Softw 2007;23:1–4019. Hotelling H. The generalization of student’sratio. Ann Math Stat 1931;2:360–378

20. American Diabetes Association. 2 Classifica-tion and diagnosis of diabetes: Standards ofMedical Care in Diabetesd2018. Diabetes Care2018;41(Suppl. 1):S13–S2721. The ORIGIN Trial Investigators. Basal insu-lin and cardiovascular and other outcomes indysglycemia. N Engl J Med 2012;367:319–32822. Kahn SE, Lachin JM, Zinman B, et al.; ADOPTStudy Group. Effects of rosiglitazone, glyburide,and metformin on b-cell function and insu-lin sensitivity in ADOPT. Diabetes 2011;60:1552–1560

care.diabetesjournals.org The RISE Consortium 1725