Calorie Restriction and Matched Weight Loss From Exercise: Independent and Additive Effects on Glucoregulation and the Incretin System in Overweight Women and Men

Calorie Restriction and MatchedWeight Loss From Exercise:Independent and AdditiveEffects on Glucoregulationand the Incretin System inOverweight Women and MenDOI: 10.2337/dc14-2913

OBJECTIVE

It is not knownwhether calorie restriction (CR) has additive benefits to those fromexercise (EX)-induced weight loss. We hypothesized that weight loss from CR andEX (CREX) improves insulin sensitivity more than matched weight loss induced byEX or CR alone, and that the incretin systemmay be involved in adaptations to CR.

RESEARCH DESIGN AND METHODS

Sedentary, overweight men and women (n = 52, 45–65 years of age) were ran-domized to undergo 6–8% weight loss by using CR, EX, or CREX. Glucose, insulin,C-peptide, insulin sensitivity, and incretin hormones (glucagon-like peptide1 [GLP-1] and glucose-dependent insulinotropic polypeptide [GIP]) were mea-sured during frequently sampled oral glucose tolerance tests (FSOGTTs). Incretineffects on insulin secretion were measured by comparing insulin secretion ratesfrom the FSOGTTs to those from a glycemia-matched glucose infusion.

RESULTS

Despite similar weight losses in all groups, insulin sensitivity index values in-creased twofold more in the CREX group (2.09 6 0.35 mmol/L/kg/pmol/L 3

100) than in the CR (0.89 6 0.39 mmol/L/kg/pmol/L 3 100) and EX (1.04 6 0.39mmol/L/kg/pmol/L 3 100) groups. Postprandial GLP-1 concentrations decreasedonly in the CR group (P = 0.04); GIP concentrations decreased in all groups. Incretineffects on insulin secretion were unchanged.

CONCLUSION

CR and EX have additive beneficial effects on glucoregulation. Furthermore, theadaptations to CR may involve reductions in postprandial GLP-1 concentrations.These findings underscore the importance of promoting both CR and EX for opti-mal health. However, because data from participants who withdrew from thestudy and from those who did not adhere to the intervention were excluded,the results may be limited to individuals who are capable of adhering to a healthylifestyle intervention.

1Department of Nutrition and Dietetics, SaintLouis University, St. Louis, MO2Division of Geriatrics and Nutritional Science,Washington University School of Medicine,St. Louis, MO3Division of Endocrinology, School of Medicine,Saint Louis University, St. Louis, MO4Department of Biomedical Laboratory Science,Saint Louis University, St. Louis, MO5Division of Endocrinology, Diabetes, andMetabolism, Baylor College of Medicine,Houston, TX

Correspondingauthor: EdwardP.Weiss, [email protected].

Received 8 December 2014 and accepted 25March 2015.

Clinical trial reg. no. NCT00777621, clinicaltrials.gov.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc14-2913/-/DC1.

© 2015 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.

Edward P. Weiss,1,2 Stewart G. Albert,3

Dominic N. Reeds,2 Kathleen S. Kress,1

Uthayashanker R. Ezekiel,4

Jennifer L. McDaniel,1

Bruce W. Patterson,2 Samuel Klein,2

and Dennis T. Villareal 2,5

Diabetes Care 1

CLIN

CARE/ED

UCATIO

N/N

UTR

ITION/PSYC

HOSO

CIAL

Diabetes Care Publish Ahead of Print, published online April 15, 2015

Calorie restriction (CR) and exercise (EX)can lead to weight loss, and are effectivefor improving glucose tolerance and insu-lin action, and reducing type 2 diabetesrisk (1–3). EX can improve glucoregulationby causing weight loss and by weightloss–independent effects, including in-creases in skeletal muscle GLUT4 trans-port protein levels (4) and greaterinsulin-mediated glucose disposal (5). Incontrast, the beneficial effect of CR onglucoregulation is often attributed toweight loss alone. In this context, EX-induced weight loss would be expectedto improve glucoregulation more thanmatched weight loss induced by CR.However, we (1) and others (6) haveshown that EX-inducedweight loss (with-out CR) does not provide greater improve-ments in glucoregulation than CR alone.A plausible explanation for this unexpectedfinding is that, in addition to providingbenefits through weight loss, CR mayalso improve glucoregulation throughother mechanisms. If this is true, thenthe combination of CR and EX (CREX)would be expected to improve glucore-gulation more so than similar weightloss from EX alone; however, no studieshave evaluated this possibility.The purpose of the current study was

to evaluate the hypothesis that CR andEX have additive effects, even in the ab-sence of greater weight loss. We pro-posed that a 7% reduction in bodymass induced by CREX results in greaterimprovements in glucose tolerance andinsulin action than those resulting fromsimilar weight loss induced by EX or CRalone. Another objective was to gain in-sights about unique mechanisms bywhich CR might alter glucoregulation(i.e., independent of weight loss in-duced by EX). Because the incretinsystem is a food-sensing system, we hy-pothesized that long-term restriction offood intake (i.e., CR), but not EX-inducedweight loss, reduces postprandial incre-tin hormone levels while their actions topromote insulin secretion (i.e., incretineffects (7)) are maintained, suggestingenhanced pancreatic sensitivity to incre-tin hormones; this might be especiallyimportant for preventing the relative in-sulin deficiency that accompanies progres-sion to type 2 diabetes. Furthermore,because one of the incretin hormones(glucagon-like peptide 1 [GLP-1]) pro-motes glucose uptake in muscle andadipose tissue (8,9), a reduction in

GLP-1 with concomitant improve-ments in glycemic control would besuggestive of enhanced GLP-1 actionson glucose uptake.

RESEARCH DESIGN AND METHODS

Study Design and RandomizationSubjects were randomized, with stratifi-cation for sex, to CR, EX, or CREX, all ofwhich were designed to induce a 6–8%weight loss. The initial allocation ratio of1:1:1 was later revised to 2:2:1, withgreater enrollment in the CR and EXgroups to account for more with-drawals from these groups. Outcomemeasures were performed at baselineand after weight loss. Participants pro-vided informed written consent toparticipate in the study, which was ap-proved by the Institutional ReviewBoards at Saint Louis University andWashington University. The trial wasregistered at clinicaltrials.gov (Clinicaltrial reg. no. NCT00777621).

ParticipantsOverweight men and postmenopausalwomen (45–65 years of age, BMI 25.0–29.9 kg/m2) were recruited from theSt. Louis metropolitan area. Potentialparticipants underwent screening,including a medical evaluation, andwere excluded from the study if theyhad experienced a significant (.3%)weight change within 6 months and ifthey performed regular vigorous endur-ance EX (moderate to hard effort EX,$20 min/session, and three or moretimes per week). Other exclusion criteriawere the presence of major chronic dis-eases, conditions that would interferewith EX or in which EX is contraindi-cated, or conditions that would inter-fere with interpretation of the results.Examples include diabetes (self-reportedor fasting blood glucose level of $126mg/dL), blood pressure of$160 mmHgsystolic or $100 mmHg diastolic, mus-culoskeletal problems, and smoking.Use of glucoregulatory medicationswas also exclusionary. For other medi-cations, participants were required tohave been on stable dosages for $6months prior to baseline testing andwere advised to maintain dosages dur-ing the study.

InterventionsThe interventions were designed to de-crease body mass by 6–8% over 12–14weeks. However, the intervention

duration was adjusted as needed for par-ticipants to reach the weight loss goal.CR and EX prescriptions were based onestimates of baseline total energy ex-penditure (TEE) and energy intake, asfollows: 1) dietary reference intakesequations for estimated energy require-ments (10); 2) 3-day food diaries withnutrient analysis (described below); 3)accelerometry (described below); and4) 7-day physical activity recalls (de-scribed below). Because energy intakeand TEE are equal duringweight stability,and because the participants wereweight stable at baseline, the averageof all four measures was used to reflectthe TEE and energy intake. During theinterventions, the prescriptions were ad-justed as needed, with the goal ofachieving weight loss at a rate of;0.5% per week. To eliminate the po-tentially confounding effects of negativeenergy balance on the results, bodyweight was stabilized by altering the CRand/or EX prescriptions for 2 weeks be-fore follow-up testing, with the goal be-ing to avoid weight changes of .0.5 kg,based on a 3-day rolling average weight.The participants recorded daily fastedmorning body weight at home, and vis-ited our clinic weekly to beweighed, turnin home weight logs, and undergo otherintervention-specific requirements (de-scribed below).

CR

The CR intervention was designed to de-crease energy intake by ;20% withoutchanging physical activity. During theinitial 3 weeks and periodically thereaf-ter, the participants completed 3-dayfood diaries that were used by the studydietitians for personalized dietary rec-ommendations. The strategies for de-creasing energy intake included foodportion control and replacing energy-dense foods with foods containinglower energy density. As needed to pro-mote compliance, participants under-went weeklong periods of full foodprovision on a 20% hypocaloric diet. Di-etary advice also included recommenda-tions for macronutrient intake to bewithin the recommended ranges (per-centages of total energy: carbohydrate45–65%; fat 20–35%; and protein 10–35%) (10).

EX Intervention

The EX intervention was designed to in-crease TEE by;20% by using EX without

2 CR, Exercise, and Glucoregulation Diabetes Care

changing energy intake. Weekly EX en-ergy expenditure prescriptions were cal-culated after accounting for differencesbetween gross and net EX energy expen-diture, as described previously (1). Thesubjects monitored their progress to-ward the energy expenditure goalswith heart rate (HR) monitors (PolarElectro Oy, Kempele, Finland), which es-timate EX energy expenditure based onEX HR and subject-specific characteris-tics (e.g., weight, maximal oxygen up-take [VO2max]). The monitors storeddata for EX energy expenditure, HR, EXduration, and EX frequency, all of whichwere transferred to the study databaseeach week. Specific goals for EX fre-quency and intensity were not provided;however, tomaximize weekly energy ex-penditure, the participants were en-couraged to perform daily EX and tomostly perform activities that required“moderate” and “hard” physical effort.The participants were advised to per-form cardiovascular EX and to increasefunctional physical activities (e.g., activetransportation). They were also advisedto refrain from strength/resistancetraining, as it may alter glucoregulationthrough uniquemechanisms. During theinitial three to six EX sessions, and asneeded to promote compliance thereaf-ter, the participants exercised under thesupervision of study personnel. Other-wise, the subjects were encouraged toEX on their own (i.e., fitness facility,home, or outdoors).

Caloric Restriction Plus EX Intervention

The CREX intervention was designed toinduce weight loss through a combina-tion of CR and EX, with each componentcontributing approximately half to thetotal energy deficit. The participantswere given weekly EX energy expendi-ture prescriptions equal to 10% of TEE.The remainder of the energy deficit(10%) was induced by caloric restriction.

Body Weight and CompositionOn 2 separate days at each study timepoint, fasted morning body weight wasmeasured in duplicate while the partic-ipant was wearing a hospital gown. Fatmass and fat-free mass were measuredwith DXA (Lunar iDXA, software version13.31; GE Healthcare, Madison, WI).

Energy IntakeEnergy intake was quantified by using 3-day food diaries with computerized

nutrient analysis (Food Processor SQLsoftware; ESHA Research, Salem, OR).

Energy ExpenditureTEE was calculated as the average of es-timates from physical activity recall(PAR) interviews and accelerometry.The PAR interview was a modified ver-sion of the Stanford 7-day PAR inter-view, as described elsewhere (11).Accelerometry was performed with tri-axial accelerometers (RT3; StayHealthy,Monrovia, CA).

Aerobic CapacityVO2max was measured with indirect calo-rimetry (MedGraphics CardiO2; MedicalGraphics Corporation, St. Paul,MN) duringan incremental treadmill EX test to exhaus-tion (modified Balke treadmill protocol).

Glucoregulatory FunctionGlucose tolerance and insulin actionwere assessed by using a 2-h frequentlysampled oral glucose tolerance test(FSOGTT) (12) after an overnight fastand after 3 days of consuming $150 g/day carbohydrates. The oral test wasused because, unlike infusion-basedmeasures, it involves the intestine,which we proposed to be involved inthe adaptations to CR. For follow-up as-sessments on subjects in the EX andCREX groups, tests were performed12–24 h after EX. Venous blood sampleswere obtained before and at 10, 20, 30,60, 90, and 120 min after administrationof a 75-g oral glucose load for the anal-ysis of plasma glucose (glucose oxidasemethod; YSI STAT Plus; YSI Life Sciences,Yellow Springs, OH), and insulin and C-peptide (IMMULITE ChemiluminescenceKit; Diagnostics Products Corporation,Los Angeles, CA).

Total areas under the curve (AUCs) forall analytes were calculated based onthe trapezoidal rule. Insulin sensitivityindex (ISI) was calculated according toStumvoll et al. (13), which is reproduc-ible (14) and valid (15), and according toMatsuda and DeFronzo (16). Total insu-lin secretion rate (ISR) and b-cell re-sponse were estimated by using theC-peptide minimal model (17,18) andSAAM II software (version 1.2; Univer-sity of Washington Digital Ventures). In-sulin clearance was estimated as the ISRAUC-to-insulin AUC ratio (19).

Matched Glucose InfusionOn a separate day after the FSOGTT, avariable rate glucose infusion was

performed with the goal of matchingthe glycemic response from the FSOGTT.As for the FSOGTT, follow-up assess-ments on subjects in the EX and CREXgroups were performed 12–24 h afterEX. Intravenous dextrose (20%) was in-fused into an antecubital vein, and bloodsamples were drawn from the contralat-eral arm. Blood samples (1 mL) weredrawn every 5 min for quantification ofplasma glucose concentrations (YSISTAT Plus) to inform decisions aboutthe glucose infusion rates. At the sametime points as described for the FSOGTT,larger blood samples were obtained forquantification of insulin and C-peptide.

Combined results from the matchedglucose infusion (MGI) and FSOGTTwere used to calculate incretin effects,as follows:

Relative incretin effectð%Þ¼ 1003ðAUCFSOGTT2AUCMGIÞ=AUCFSOGTT

Absolute incretin effectðAUC unitsÞ¼ AUCFSOGTT 2AUCMGI

Additionally, the incretin effect on insulinclearance was evaluated by comparinginsulin clearance from the FSOGTT to thatfrom the MGI.

Incretin Hormones and DipeptidylPeptidase-IVA portion of the blood samples fromthe FSOGTT were collected directlyinto tubes that contained a dipeptidylpeptidase-IV (DPP-IV) inhibitor (DPP4–010; Millipore, Billerica MA) to preventthe degradation of active forms of GLP-1and glucose-dependent insulinotropicpolypeptide (GIP). Plasma was analyzedfor concentrations of active humanGLP-1 (7–36 and 7–37 amides) and ac-tive human GIP (1–42 amide) with ELISA(IBL International, Toronto, ON, Canada).DPP-IV was measured in fasting plasmaby using ELISA (R&D Systems, Minneap-olis, MN).

Statistical AnalysesAs planned a priori, the primary analyseswere performed on a per-protocol basis,and excluded subjects who did not com-plete the study and those who did notlose weight. Baseline characteristicsamong groups were compared withFisher exact tests and ANOVAs. Out-comeswere compared by using ANCOVAs,in which the study group was theindependent variable; change in the

care.diabetesjournals.org Weiss and Associates 3

outcome (i.e., final value minus baselinevalue) was the dependent variable, andthe baseline value was a covariate.Between-group post hoc comparisonswere performed using the protectedF test principle and least significant dif-ference tests. Baseline-adjusted leastsquares means were used to evaluatethe significance of within-group changes.Associations were evaluated with Pear-son correlations. Intention-to-treat (ITT)analyses (including data from all 69subjects who underwent baseline test-ing and were randomized) were alsoperformed. Missing data were handledby using the last observation carriedforward. The analysis of outcome dataincluded the magnitude of weight lossas a covariate. Because the ITT approachresulted in differences among groupsfor weight loss, which is problematicwhen evaluating the effects of matchedweight losses, it was considered a sup-plementary analysis. All statisticaltests were two tailed, and significancewas accepted at P # 0.05. Data arepresented as the arithmetic mean 6SE, unless otherwise noted. Analyseswere performed using SAS for Win-dows (version 9.3; SAS Institute Inc.,Cary, NC).

RESULTS

ParticipantsAmong 525 individuals who inquiredabout study participation, 63 were not in-terested after learning more about thestudy, and 393 were screened out, withthe single most common reason beingBMI$30 kg/m2. The remaining 69 partic-ipants were enrolled, underwent baselinetesting, and were randomized (Supple-mentary Fig. 1, consort diagram). Twelveparticipants discontinued participationbefore providing follow-up data, and 5participants were noncompliant (weightloss ,1%; range 20.5% to 1.3%). There-fore, analyses for the present report werebased on 52 participants. By design, theparticipants were at increased risk of di-abetes by virtue of being middle to olderaged (mean age 57 6 1 years), over-weight (mean BMI 27.7 6 0.2 kg/m2),and physically inactive (VO2max values inthe 10th to 15th percentile for age andsex) (20) (Table 1). Accordingly, based onfasting and 2-h plasma glucose and hemo-globin A1c levels (21), 54% of the partici-pants (n = 28) hadprediabetes at baseline.Furthermore, although subjects with

diagnosed diabetes and those with fast-ing blood glucose levels of $126 mg/dLwere excluded during screening, threeparticipants (6%) had 2-h glucose toler-ance test glucose values that met thecriteria for diabetes; according to clinicalstandards, these cases were considered“provisional diabetes” because a formaldiagnosis of diabetes would haverequired a second test to confirm theinitial results (21) (Table 1).

Body Weight and CompositionBody mass decreased by ;7% in allthree groups (Fig. 1), as intended bydesign. The time required to reachthe weight loss goal was shorter in theCREX group (13 6 2 weeks) than in theCR group (196 2weeks, P = 0.02) and EXgroup (20 6 2 weeks, P = 0.007). Therewere nonsignificant tendencies forgreater fat mass reductions and better

preservation of fat-free mass in the EXand CREX groups (Fig. 1). During the 2-week weight stability period prior tofollow-up testing, body mass did notchange (CR group 20.1 6 0.2 kg; CREXgroup 20.2 6 0.2 kg; EX group 20.2 60.2 kg; all P $ 0.28).

The ITT analyses revealed smaller re-ductions in body mass in the CR group(24.8 6 0.7%) and EX group (24.6 60.7%) compared with the CREX group(27.2 6 0.8%). Likewise, the reductionsin fat mass were significantly less in theCR and EX groups (Supplementary Fig. 3).

EX Training Volume and ModeEX energy expenditure in the CREXgroup was 217 6 23 kcal/day, which isequivalent to 10% of baseline TEE. TheEX energy expenditure in the EX groupwas 412 6 26 kcal/day (equivalent to22% of baseline TEE) and greater than

Table 1—Baseline characteristics of study participants

CR group CREX group EX group Among-group P*

Participants, n 17 19 16

Female sex 13 (76) 15 (79) 11 (69) 0.85

Age, years 57 6 1 57 6 1 56 6 1 0.86

Race 0.26Caucasian 16 (94) 14 (74) 12 (75)African American 0 (0) 4 (21) 3 (19)Other or not specified 1 (6) 1 (5) 1 (6)

Body weight, kgWomen 73.2 6 1.8 77.2 6 1.7 74.2 6 1.8 0.26Men 92.4 6 5.6 98.7 6 5.6 86.3 6 5.0 0.30

BMI, kg/m2 27.7 6 0.4 28.3 6 0.4 27.0 6 0.4 0.08

Waist circumference, cmWomen 88.6 6 2.6 88.7 6 1.8 90.4 6 2.1 0.81Men 101.9 6 2.0 108.0 6 4.8 96.6 6 3.1 0.11

Energy intake, kcal/day 2,2436 145 2,3106 137 1,9096 149 0.12

TEE, kcal/day 2,068 6 88 2,092 6 86 2,143 6 94 0.84

Activity energyexpenditure, kcal/day 425 6 43 439 6 43 423 6 46 0.96

VO2max, mL/kg/minWomen 25 6 1 22 6 1 24 6 1 0.22Men 32 6 3 28 6 3 28 6 3 0.60

Fasting glucose, mmol/L 5.4 6 0.1 5.4 6 0.1 5.2 6 0.1 0.48

2-h glucose, mmol/L 7.3 6 0.5 6.3 6 0.5 7.0 6 0.5 0.37

Hemoglobin A1c% 5.6 6 0.1 5.7 6 0.1 5.7 6 0.1 0.69mmol/mol 38 6 1 39 6 1 39 6 1 0.69

Prediabetes prevalence 9 (53) 10 (53) 9 (56) 1.00

Provisional diabetes prevalence 2 (12) 1 (5) 0 (0) 0.51

Total cholesterol, mg/dL 202 6 7 206 6 9 178 6 6 0.15

Systolic BP, mmHg 127 6 3 127 6 3 125 6 3 0.79

Diastolic BP, mmHg 82 6 2 78 6 2 82 6 2 0.33

Values are reported as counts (%) for categorical data and the mean6 SE for quantitative data,unless otherwise indicated. Prediabetes and provisional diabetes were determined based onclinical criteria from the American Diabetes Association (21). *Between-group P values forquantitative data are from ANOVAs, and those for categorical data are from Fisher exact tests.

4 CR, Exercise, and Glucoregulation Diabetes Care

that in the CREX group (P, 0.0001). EXduration was 4.4 6 0.5 h/week in theCREX group and 7.4 6 0.5 h/week inthe EX group (P = 0.0002). EX frequencywas 6 6 1 sessions/week in the CREXgroup and 8 6 1 sessions/week (i.e.,.1 session/day) in the EX group (P = 0.08vs. CREX). EX HR did not differ between theCREX and EX groupswith respective EXHRsof 74 6 1% and 77 6 1% of measured

maximal HR (P = 0.17). Brisk walking wasthe most common mode of EX; however,other commonly used modes included cy-cling, elliptical machine EX, stair climbing,and running.

Energy IntakeEnergy intake in the CR and CREX groupsdecreased significantly, while the en-ergy intake in the EX group was not

different from baseline (Fig. 1). The re-ductions in energy intake in the CR andCREX groups were attributed to respec-tive reductions in all macronutrient in-takes including fat (239 6 7% and227 6 6%, both P # 0.0001), carbohy-drate (229 6 5% and 229 6 5%, bothP # 0.0001), and protein (222 6 5%and 217 6 5%, both P # 0.001). Basedon the ITT analysis, which included datafrom participants who dropped out andthose who were noncompliant, themagnitude of the decrease in energy in-take in the CR group was less than thatfrom the per protocol analysis, as ex-pected (Supplementary Fig. 3).

Energy ExpenditureTEE increased from baseline in the EXand CREX groups and remained un-changed in the CR group (Fig. 1). As ex-pected, due to reductions in body massduring the intervention, the increases inTEE were less than the increases in EXenergy expenditure. Based on physicalactivity recall data, there were nochanges in sedentary and light physicalactivity within any study group (all P .0.19), and there were no differencesamong groups (P = 0.96). Comparedwith the per protocol analyses, the ITTanalyses revealed smaller increases inenergy expenditure in the CREX and EXgroups (Supplementary Fig. 3).

Aerobic CapacityChanges in VO2max during the interven-tion correspondedwith the EX dose (Fig.1), with the CR group having no change(0 6 1 mL/kg/min, P = 0.79), the CREXgroup having a modest increase (3 6 1mL/kg/min, P = 0.0009), and the EXgroup having the largest increase (5 61 mL/kg/min, P , 0.0001). When ana-lyzed using the ITT approach (includingdropouts and noncompliant subjects),the improvements in VO2max in theCREX and EX groups were smaller thanthose based on the per protocol analysis(Supplementary Fig. 3).

Glucoregulatory Responses Duringthe FSOGTTISI increased in all groups; however, theincrease in the CREX group was twofoldgreater than those in the CR and EXgroups (Table 2). The glucose AUC de-creased in the CREX group but not inthe CR or EX groups; however, thesechanges did not differ among thegroups (P = 0.12) (Table 2). Among the

Figure 1—Body weight and composition changes (top panel) and measures of interventioncompliance (bottom panel). Values are reported as the least-squares mean6 SE from ANCOVAsin which the change in the outcome was the dependent variable, study group was the indepen-dent variable, and baseline values were the covariate. P values reflect the significance of theoverall ANCOVA. When overall ANCOVAs were significant at P# 0.05, post hoc paired compar-isons were performed according to the principle of protected F tests and least significantdifference tests. *P # 0.05 vs. zero (significance of within-group change). †P # 0.05 vs. CRgroup. ‡P # 0.05 vs. CREX group.

care.diabetesjournals.org Weiss and Associates 5

Table 2—FSOGTT responses

CR CREX EX Among-group P

ISI (Stumvoll et al. [13]), mU/kg/pmol/LBaseline 0.072 6 0.006 0.076 6 0.009 0.085 6 0.005 0.43Final 0.083 6 0.006 0.098 6 0.005 0.093 6 0.007Adjusted change 0.009 6 0.004 0.021 6 0.004* 0.010 6 0.004 0.04Within-group P 0.02 ,0.0001 0.01

ISI (Matsuda and DeFronzo [16] index)Baseline 4.4 6 0.7 5.5 6 0.9 5.5 6 0.6 0.52Final 5.3 6 0.6 8.2 6 1.2 6.3 6 0.8Adjusted change 0.8 6 0.6 2.9 6 0.6* 0.7 6 0.7 0.03Within-group P 0.23 ,0.0001 0.27

Glucose AUC, mmol/L/minBaseline 986 6 46 910 6 55 897 6 44 0.40Final 981 6 49 846 6 36 921 6 53Adjusted change 14 6 34 271 6 32 12 6 35 0.12Within-group P 0.68 0.03 0.74

Insulin AUC, 3103 pmol/L/minBaseline 56.8 6 6.7 56.8 6 9.5 44.1 6 5.8 0.43Final 43.5 6 3.5 43.3 6 8.3 36.4 6 4.8Adjusted change 212.0 6 3.4 212.2 6 3.2 210.6 6 3.6 0.94Within-group P 0.001 0.0004 0.005

C-peptide AUC, nmol/L/minBaseline 1,184 6 69 1,113 6 101 1,052 6 80 0.58Final 1,041 6 64 957 6 103 958 6 80Adjusted change 2128 6 54 2156 6 51 2108 6 56 0.82Within-group P 0.02 0.004 0.06

ISR AUC, 3103 pmolBaseline 29.2 6 1.8 26.9 6 2.5 25.9 6 2.0 0.57Final 25.9 6 1.7 22.7 6 2.4 23.3 6 2.2Adjusted change 22.8 6 1.4 24.3 6 1.4 22.9 6 1.5 0.69Within-group P 0.06 0.002 0.05

F, 109 min21

Baseline 34.5 6 2.1 32.9 6 2.8 34.5 6 2.8 0.89Final 31.3 6 2.2 29.0 6 3.1 30.3 6 2.2Adjusted change 22.9 6 2.2 24.4 6 2.1 23.9 6 2.3 0.89Within-group P 0.20 0.04 0.09

Insulin clearance index(ISR:insulin AUC ratio)

Baseline 0.58 6 0.04 0.57 6 0.04 0.66 6 0.05 0.26Final 0.62 6 0.03 0.64 6 0.05 0.70 6 0.05Adjusted change 0.04 6 0.03 0.07 6 0.03 0.05 6 0.03 0.66Within-group P 0.20 0.01 0.09

GLP-1 AUC, pmol/L/minBaseline 1,357 6 124 1,582 6 139 1,273 6 104 0.20Final 1,169 6 100 1,636 6 157 1,407 6 143Adjusted change 2199 6 95† 86 6 91 106 6 99 0.05Within-group P 0.04 0.35 0.29

GIP AUC, pmol/L/minBaseline 4,299 6 350 4,050 6 398 3,445 6 433 0.32Final 3,898 6 295 3,654 6 371 3,460 6 353Adjusted change 2292 6 216 2364 6 203 2139 6 224 0.76Within-group P 0.18 0.08 0.54

DPP-IV (fasting), ng/mLBaseline 402 6 15 371 6 24 368 6 24 0.48Final 373 6 14 375 6 27 348 6 25Adjusted change 222.9 6 17.3 2.0 6 16.2 223.2 6 17.7 0.47Within-group P 0.19 0.90 0.20

Values are reported as arithmetic means6 SE, except for adjusted change values, which are least squares means6 SE that have been adjusted fordifferences in baseline values among groups. F, pancreatic b-cell sensitivity to glucose according to the C-peptide minimal model analysis (17,18).Insulin clearance index was calculated as the ratio of total ISR AUC to insulin AUC (19). Among-group P, significance of the among-group differencesin change values after adjustment for baseline values using ANCOVA. *P # 0.05 vs. CR and EX groups; †P # 0.05 vs. CREX and EX groups.

6 CR, Exercise, and Glucoregulation Diabetes Care

participants with prediabetes at baseline(Table 1), 33% (n = 3) in the CR group,30% in the CREX group, and 11% in theEX group became normoglycemic at thefollow-up assessment; these conversionrates did not differ among groups (P =0.46). FSOGTT insulin and C-peptideAUCs decreased to a similar extent inall study groups (Table 2). These changeswere accompanied by reductions in ISRAUC (212%, P = 0.0002). Changes in in-sulin clearance did not differ amonggroups (Table 2); however, with allgroups, the combined insulin clearanceincreased by 9% (P = 0.002).Based on the ITT analyses, including

data from dropouts and noncompliantparticipants, the improvements ininsulin sensitivity (the method of Stum-voll et al. [13]) were smaller thanthose from the per protocol analysis(0.008 6 0.003, 0.018 6 0.004, and0.005 mU/kg/mmol/L in the CR, CREX,and EX groups, respectively), and thetest for differences among groups be-came marginally nonsignificant (P =0.054). Results from the ISI of Matsudaand DeFronzo (16) were similarly af-fected, with the test for differencesamong groups becomingmarginally non-significant (P = 0.08). The statistical sig-nificance of results for glucose, insulin,and C-peptide AUCs and ISR were notdifferent between the ITT and per pro-tocol analyses. However, the decreasesin insulin AUC and ISR were smaller inthe ITT analysis (insulin AUC 3103

pmol/L/min: CR 210.4 6 3.3, CREX212.2 6 3.7, EX 25.2 6 3.2; ISR AUC3103 pmol: CR 22.5 6 1.2, CREX23.5 6 1.3, EX 21.7 6 1.1).

Incretin Hormone Responses to theFSOGTTPlasma GLP-1 response (AUC) during theFSOGTT decreased by 15% in the CRgroup and remained unchanged in theCREX and EX groups (Table 2). Changesin GIP AUC did not differ among groups(P = 0.76); however, there was a margin-ally significant 7% decrease (P = 0.058)with all groups combined. DPP-IV con-centrations did not change in any of thegroups (Table 2). The statistical signifi-cance of the incretin hormone resultsfrom the ITT analyses did not differfrom the results of the per protocolanalyses described above. However,the magnitude of the decrease in GLP-1 AUC in response to CR was 23%

smaller (2199 6 95 vs. 2153 6 73pmol/L/min, both P = 0.04).

Glucoregulatory Responses to theMatched Glucose InfusionPlasma glucose concentrations from theFSOGTT and MGI were well matched(Supplementary Fig. 2). There were nodifferences among groups in terms ofchanges in MGI insulin, C-peptide, or in-sulin secretions rates (SupplementaryTable 1). However, with all groups com-bined, significant decreases wereobserved for MGI insulin AUC (225%,P = 0.002) and C-peptide AUC (217%,P = 0.0003). These changes were accom-panied by a 17% reduction in ISR AUC(P = 0.002). Insulin clearance during theMGI exhibited a 14% increase (P = 0.001)for all groups combined, with no differ-ences among groups.

Incretin EffectsAs expected, ISRs and postprandial in-sulin and C-peptide concentrationswere greater after oral glucose ingestionthan during the glycemia-matched glu-cose infusion, indicating the presence ofincretin effects (Table 3 and Supplemen-tary Fig. 2). None of the interventionsaltered the absolute or relative incretineffects on insulin secretion (Table 3).Likewise, the relative incretin effectson insulin AUC and C-peptide AUC didnot change. Although no among-groupdifferences were observed, absolute in-cretin effects on insulin and C-peptidedecreased with weight loss when allgroups were combined (26.6 6 2.4 3103 pmol/L/min [P = 0.009] and2919642 pmol/L/min [P = 0.002], respec-tively). Insulin clearance was lower afteroral glucose infusion than during thematched glucose infusion, indicating anincretin effect to suppress insulin clear-ance; the magnitude of this incretin ef-fect did not change during any of theinterventions (Table 3). The use of ITTanalyses did not alter the significanceof the incretin effect results (i.e., neitherthe ITT or per protocol analysis yieldedany significant differences amonggroups).

CONCLUSIONS

Weight loss involving CR and EX im-proves glucoregulation and reduces di-abetes risk in overweight and obesesubjects (2,3); however, the indepen-dent contributions of CR and EX arepoorly understood. Results from the

current study demonstrate that CR andEX together improve insulin sensitivitytwofold more than does the sameamount of weight loss induced by CRalone or EX alone. Furthermore, to en-sure similar weight losses in the threegroups, the CREX group underwent lessCR and less EX training than those usedin the CR and EX groups, respectively,and, despite this, they had much largerimprovements. The finding of additiveeffects indicates that some of the adap-tations to CR are mechanistically dis-tinct from those caused by EX-inducedweight loss. Thus, although a combina-tion of CR and EX is often recommendedformaximizing weight loss because bothcontribute to a negative energy balance,findings from the current study providean additional rationale for encouragingboth CR and EX, because together theyprovide a greater improvement in insu-lin sensitivity than either alone.

A secondary objective of the currentstudy was to gain insights about distinctmechanisms by which CR improves glu-coregulation. Postprandial GLP-1 con-centrations decreased in response toCR, but not in response to matchedweight loss from EX. Others have alsoreported that CR decreases postprandialGLP-1 levels (22,23) (as long as the post-prandial response at baseline is notblunted [24]), and that EX trainingdoes not affect GLP-1 concentrations(25,26). Because GLP-1 influences glycemiccontrol through actions on the pan-creas, skeletal muscle, and adipose tis-sue (7–9), the finding of CR-inducedreductions in GLP-1 in the current studyraises the possibility that some of thebeneficial effects of CR on glucoregula-tion may involve GLP-1. In contrast tothe findings for GLP-1, changes in post-prandial GIP concentrations did not dif-fer among groups. However, with allgroups combined, a marginal decreasein GIP occurred (P = 0.058), suggestingthat weight loss, per se, mediates thiseffect. Others have reported similar GIPresponses to weight loss (24,27,28) andthat EX training without weight loss doesnot alter GIP (25,28).

It is plausible that the observed re-ductions in GLP-1 and GIP levels mighthave resulted from reduced intestinalsecretion. Because we measured thebioactive forms of these hormones, analternate explanation would be en-hanced degradation to the inactive

care.diabetesjournals.org Weiss and Associates 7

forms by DPP-IV, which occurs withinminutes after secretion (29,30). How-ever, because DPP-IV concentrationsdid not change, this scenario does notseem likely.Despite the reductions in postpran-

dial incretin hormone levels, none ofthe weight loss interventions affectedthe incretin effects on insulin secretion,clearance, or concentrations, suggestingthat the net effect of the gut on insulinmetabolism was not altered. This disso-ciation of incretin hormone concentra-tions from incretin effects suggests thatthe b-cells of the pancreas became

more sensitive to GLP-1 (CR grouponly) and GIP (all groups). Other studieson the effect of weight change on incre-tin effects are mixed, with one showingthat CR had no effect (31), anothershowing that short-term weight gain re-duced the incretin effect from 72% to43% (32), and others reporting twofoldto fivefold increases from weight lossafter gastric bypass (31,33,34).

Data from the current study indicatethat CR and EX have additive effects oninsulin sensitivity when weight loss ismatched. However, this finding mightunderstate the benefits of combined

CR and EX because together, they wouldlikely lead to greater weight loss suc-cess. Evidence from the current studysupports this notion. First, the ITT anal-ysis of weight changes (including all sub-jects, regardless of withdrawals andpoor compliance) revealed;50% greaterweight loss in the CREX group than in theother groups. Furthermore, despite thefact that the interventions were designedfor similar rates of weight loss, the CREXgroup attained the weight loss goal inone-third less time than the other groups.Additionally, 95%of the subjects random-ized to CREX eventually achieved the

Table 3—Absolute and relative incretin effects

CR CREX EX Among-group P

Incretin effect on ISR total AUC, %Baseline 53.2 6 3.0 56.6 6 4.1 43.2 6 8.2Final 53.2 6 4.2 59.5 6 2.9 41.4 6 7.5 0.27Adjusted change 0.5 6 3.6 5.6 6 3.5 2.7 6 3.6Within-group P 0.89 0.12 0.47 0.60

Incretin effect on ISR AUC, 3103 pmolBaseline 16.2 6 1.6 14.1 6 1.5 12.2 6 1.3 0.19Final 14.0 6 1.4 12.9 6 1.4 12.1 6 1.4Adjusted change 21.0 6 1.3 21.2 6 1.2 21.1 6 1.3 1.00Within-group P 0.44 0.34 0.39

Incretin effect on insulin AUC, %Baseline 63.2 6 3.6 68.2 6 3.0 55.1 6 5.5 0.09Final 63.9 6 4.4 69.1 6 3.5 63.6 6 4.5Adjusted change 1.4 6 4.1 5.4 6 4.1 3.0 6 4.2 0.79Within-group P 0.73 0.19 0.48

Incretin effect on insulin AUC, 3103 pmol/L/minBaseline 38.4 6 5.7 36.8 6 7.5 25.5 6 4.2 0.27Final 28.4 6 2.9 30.3 6 7.3 22.3 6 2.8Adjusted change 28.1 6 3.5 25.2 6 3.4 26.5 6 3.5 0.84Within-group P 0.03 0.13 0.07

Incretin effect on C-peptide AUC, %Baseline 82.7 6 1.0 84.0 6 1.3 80.4 6 1.5 0.14Final 82.7 6 1.2 84.5 6 0.9 83.2 6 1.4Adjusted change 0.2 6 1.1 1.6 6 1.1 1.4 6 1.1 0.64Within-group P 0.86 0.16 0.22

Incretin effect on C-peptide total AUC, nmol/L/minBaseline 1,001 6 65 912 6 84 847 6 65 0.33Final 882 6 55 790 6 95 795 6 65Adjusted change 2100 6 52 2124 6 50 269 6 52 0.75Within-group P 0.06 0.02 0.19

Incretin effect on insulin clearance, %Baseline 236 6 7 240 6 6 226 6 7 0.32Final 236 6 6 238 6 7 238 6 7Adjusted change 22 6 7 24 6 7 25 6 7 0.96Within-group P 0.77 0.59 0.49

Incretin effect on insulin clearance, ISR:insulin AUC ratioBaseline 20.19 6 0.04 20.21 6 0.03 20.15 6 0.04 0.45Final 20.22 6 0.04 20.24 6 0.05 20.26 6 0.06Adjusted change 20.04 6 0.05 20.05 6 0.05 20.08 6 0.05 0.80Within-group P 0.47 0.32 0.11

Values are reported as the arithmetic mean6 SE, except for adjusted change values, which are reported as the least squares mean6 SE that havebeen adjusted for differences in baseline values among groups. Among-group P, significance of the among-group differences in change values afteradjustment for baseline values using ANCOVA. Incretin effect data are missing for one subject in the CR group and two subjects in the CREX groupbecause of missing matched glucose infusion data (refused test n = 2, technical problems during test n = 1).

8 CR, Exercise, and Glucoregulation Diabetes Care

target weight, while only two-thirds ofthe CR and EX group subjects reachedthe weight loss goal. These findings re-lated to weight loss efficacy, in addi-tion to the weight loss–independentbenefits of CR and EX, underscore theimportance of using both a low-caloriediet and EX for minimizing diabetesrisk.The current study has limitations.

First, the study was not powered to de-tect improvements in clinical status (i.e.,prediabetes vs. normoglycemia), espe-cially after eliminating participantswho had normoglycemia at baseline. Fu-ture studies with more participants andthat target individuals with prediabetesarewarranted. Another limitation is thatwe enrolled only overweight individualsbecause obese men and women wouldbe at an increased risk of orthopedicproblems during vigorous EX. Therefore,it is not clear whether these findings canbe generalized to obese individuals.Another limitation is that the primary

statistical analyses were performedusing a “per protocol” analysis, whichexcluded data from participants whowithdrew from the study and thosewho did not adhere to the interventions.ITT analyses including data from all ran-domized participants are the standardfor clinical trials and are especially im-portant for evaluating the effectivenessof clinical treatments in a real-world set-ting. However, the current study wasnot an effectiveness trial; rather, it wasdesigned to compare the efficacy of theinterventions when fully adhered to.Furthermore, ITT analyses were notideal for the current study because theCR and EX groups had lower adher-ence and more withdrawals, and conse-quently less weight loss when using anITT analysis (Supplementary Fig. 3); thisdifference in weight loss among groupsprecludes the evaluation of the weightloss–independent effects of CR, CREX,and EX, and might lead to the invalidconclusion that the differences in glu-coregulation among groups are attri-butable to differences in weight loss.Nonetheless, ITT analysis was perfor-med as a secondary analysis, and theresults were adjusted for the differencesin weight loss among groups (this ad-justs the results to reflect an averageweight loss of 25.4% based on all sub-jects, regardless of compliance). Whilethe results from this analysis largely

support those from the per protocolanalysis, most of the improvementsin outcomes were attenuated, and thedifferences among groups for insulinsensitivity became marginally nonsignif-icant (P = 0.054). A limitation in usingper protocol analyses is that it mightcause a selection bias. That is, some ofthe benefits of random allocation maybe lost, and the results may be attribut-able to unknown characteristics thatdiffered among groups. Therefore, theresults from the current study should beinterpreted with this possibility in mind.

In conclusion, data from the currentstudy indicate that a combination of CRand EX results in greater improvements inglucoregulation than matched weightloss induced by CR or EX alone. This find-ing of additive benefits suggests thatboth CR-induced and EX-induced weightloss provide mechanistically distinctadaptations that may not be directly at-tributable to weight loss. Although pre-liminary, the CR-specific adaptation mayinvolve alterations in postprandial GLP-1secretion and actions. From a clinical per-spective, these findings underscore theimportance of recommending both ahealthy low-calorie diet and EX for mini-mizing the risk of type 2 diabetes.

Acknowledgments. The authors thank thestudy participants for their cooperation, andthe staff of the National Institutes of Health/Institute of Clinical and Translational SciencesClinical Research Unit for their skilled assis-tance. The authors also thank the followinggraduate students from the Department ofNutrition and Dietetics at Saint Louis Universityfor their contributions to the conduct of thisstudy: Katie Niekamp, Sophia Liu, RichardJordan, Kory Grench, Laura Kahle, Susan Caciano,Christy Kelly, Meredith Young, Alyson Heller,Emily Freeman, Cameron Sisler, Kayli Rice,Ashley Byrd, and Kelly Trom.Funding. This work was supported by NationalInstitutes of Health grants K01-DK-080886 andDK-56341 (Nutrition and Obesity Research Cen-ter), and UL1-RR-024992 (Clinical TranslationalScience Award); and by a grant from the SaintLouis University President’s Research Fund.Duality of Interest. No potential conflicts ofinterest relevant to this article were reported.Author Contributions. E.P.W. developed thestudy design, performed and supervised thestudy intervention, performed and supervisedthe data collection, performed the data analy-ses and interpretation, and wrote the manu-script. S.G.A. and D.N.R. performed andsupervised the study intervention, performedthe data analyses and interpretation, and wrotethe manuscript. K.S.K. performed and super-vised the study intervention. U.R.E. performed

and supervised the data collection. J.L.M. per-formed and supervised the study interventionand the data collection. B.W.P. performed thedata analyses and interpretation. S.K. per-formed the data analyses and interpretation,and wrote themanuscript. D.T.V. developed thestudy design, performed and supervised thedata collection, performed the data analysesand interpretation, and wrote the manuscript.E.P.W. is the guarantor of this work and, as such,had full access to all the data in the study andtakes responsibility for the integrity of the dataand the accuracy of the data analysis.Prior Presentation. Parts of the data in thisreport were presented previously at theAmerican College of Sports Medicine’s 60th(Indianapolis, IN, 28 May–1 June 2013) and62nd (San Diego, CA, 26–30 May 2015) NationalMeetings, and at the 2014 Food & NutritionConference & Expo of the Academy of Nutritionand Dietetics, Atlanta, GA, 18–21 October 2014.

References1. Weiss EP, Racette SB, Villareal DT, et al.;Washington University School of MedicineCALERIE Group. Improvements in glucose toler-ance and insulin action induced by increasingenergy expenditure or decreasing energy in-take: a randomized controlled trial. Am J ClinNutr 2006;84:1033–10422. Knowler WC, Barrett-Connor E, Fowler SE,et al.; Diabetes Prevention Program ResearchGroup. Reduction in the incidence of type 2 di-abetes with lifestyle intervention or metformin.N Engl J Med 2002;346:393–4033. Tuomilehto J, Lindstrom J, Eriksson JG, et al.;Finnish Diabetes Prevention Study Group. Pre-vention of type 2 diabetes mellitus by changesin lifestyle among subjects with impairedglucose tolerance. N Engl J Med 2001;344:1343–13504. Gulve EA, Spina RJ. Effect of 7-10 days ofcycle ergometer exercise on skeletal muscleGLUT-4 protein content. J Appl Physiol (1985)1995;79:1562–15665. O’Gorman DJ, Karlsson HK, McQuaid S, et al.Exercise training increases insulin-stimulatedglucose disposal and GLUT4 (SLC2A4) proteincontent in patients with type 2 diabetes. Diabe-tologia 2006;49:2983–29926. Ross R, Dagnone D, Jones PJ, et al. Reductionin obesity and related comorbid conditions afterdiet-induced weight loss or exercise-inducedweight loss in men. A randomized, controlledtrial. Ann Intern Med 2000;133:92–1037. Drucker DJ. The biology of incretin hor-mones. Cell Metab 2006;3:153–1658. Gonzalez N, Acitores A, Sancho V, Valverde I,Villanueva-Pe~nacarrillo ML. Effect of GLP-1 onglucose transport and its cell signalling in hu-man myocytes. Regul Pept 2005;126:203–2119. Perea A, Vi~nambres C, Clemente F, Villanueva-Pe~nacarrillo ML, Valverde I. GLP-1 (7-36) amide:effects on glucose transport and metabolism inrat adipose tissue. Horm Metab Res 1997;29:417–42110. Institute of Medicine. Food and NutritionBoard: Dietary Reference Intakes for Energy, Car-bohydrate, Fiber, Fat, Fatty Acids, Cholesterol,Protein, and Amino Acids (Macronutrients).Washington, D.C., The National Academies Press,2002

care.diabetesjournals.org Weiss and Associates 9

11. Sallis JF. Seven-day physical activity recall.Med Sci Sports Exerc 1997;29(Suppl. 6):89–10312. Dalla Man C, Campioni M, Polonsky KS,et al. Two-hour seven-sample oral glucosetolerance test and meal protocol: minimalmodel assessment of b-cell responsivity andinsulin sensitivity in nondiabetic individuals.Diabetes 2005;54:3265–327313. Stumvoll M, Mitrakou A, Pimenta W, et al.Use of the oral glucose tolerance test to assessinsulin release and insulin sensitivity. DiabetesCare 2000;23:295–30114. Gordon BA, Fraser SF, Bird SR, Benson AC.Reproducibility of multiple repeated oral glu-cose tolerance tests. Diabetes Res Clin Pract2011;94:e78–e8215. Otten J, Ahren B,Olsson T. Surrogatemeasuresof insulin sensitivity vs the hyperinsulinaemic-euglycaemic clamp: a meta-analysis. Diabeto-logia 2014;57:1781–178816. Matsuda M, DeFronzo RA. Insulin sensitiv-ity indices obtained from oral glucose tolerancetesting: comparison with the euglycemic insulinclamp. Diabetes Care 1999;22:1462–147017. Breda E, Cavaghan MK, Toffolo G, PolonskyKS, Cobelli C. Oral glucose tolerance test mini-mal model indexes of b-cell function and insulinsensitivity. Diabetes 2001;50:150–15818. Van Cauter E, Mestrez F, Sturis J, PolonskyKS. Estimation of insulin secretion rates fromC-peptide levels. Comparison of individual andstandard kinetic parameters for C-peptide clear-ance. Diabetes 1992;41:368–37719. Polonsky KS, Given BD, Hirsch L, et al. Quan-titative study of insulin secretion and clearancein normal and obese subjects. J Clin Invest 1988;81:435–441

20. American College of Sports Medicine.ACSM’s Guidelines for Exercise Testing and Pre-scription. Baltimore, Williams & Wilkins, 201421. American Diabetes Association. Standardsof medical care in diabetesd2014. DiabetesCare 2014;37(Suppl. 1):S14–S8022. Adam TC, Jocken J, Westerterp-PlantengaMS. Decreased glucagon-like peptide 1 releaseafter weight loss in overweight/obese subjects.Obes Res 2005;13:710–71623. Adam TC, Lejeune MP, Westerterp-PlantengaMS. Nutrient-stimulated glucagon-like peptide1 release after body-weight loss and weightmaintenance in human subjects. Br J Nutr2006;95:160–16724. Verdich C, Toubro S, Buemann B, LysgardMadsen J, Juul Holst J, Astrup A. The role ofpostprandial releases of insulin and incretinhormones in meal-induced satietydeffect ofobesity and weight reduction. Int J Obes RelatMetab Disord 2001;25:1206–121425. Weiss EP, Royer NK, Fisher JS, Holloszy JO,Fontana L. Postprandial plasma incretin hor-mones in exercise-trained versus untrained sub-jects. Med Sci Sports Exerc 2014;46:1098–110326. Heden TD, Liu Y, Kearney ML, et al. Priorexercise and postprandial incretin responses inlean and obese individuals. Med Sci Sports Exerc2013;45:1897–190527. Solomon TP, Haus JM, Kelly KR, Rocco M,Kashyap SR, Kirwan JP. Improved pancreaticb-cell function in type 2 diabetic patients afterlifestyle-induced weight loss is related toglucose-dependent insulinotropic polypeptide.Diabetes Care 2010;33:1561–156628. Kelly KR, Brooks LM, Solomon TP, KashyapSR, O’Leary VB, Kirwan JP. The glucose-

dependent insulinotropic polypeptide and glucose-stimulated insulin response to exercise training anddiet in obesity. Am J Physiol EndocrinolMetab 2009;296:E1269–E127429. Mentlein R, Gallwitz B, Schmidt WE.Dipeptidyl-peptidase IV hydrolyses gastric in-hibitory polypeptide, glucagon-like peptide-1(7-36)amide, peptide histidine methionineand is responsible for their degradation inhuman serum. Eur J Biochem 1993;214:829–83530. Kieffer TJ, McIntosh CH, Pederson RA. Deg-radation of glucose-dependent insulinotropicpolypeptide and truncated glucagon-like pep-tide 1 in vitro and in vivo by dipeptidyl peptidaseIV. Endocrinology 1995;136:3585–359631. Laferrere B, Teixeira J, McGinty J, et al. Ef-fect of weight loss by gastric bypass surgeryversus hypocaloric diet on glucose and incretinlevels in patients with type 2 diabetes. J ClinEndocrinol Metab 2008;93:2479–248532. Hansen KB, Vilsbøll T, Bagger JI, Holst JJ,Knop FK. Reduced glucose tolerance andinsulin resistance induced by steroid treat-ment, relative physical inactivity, and high-calorie diet impairs the incretin effect inhealthy subjects. J Clin Endocrinol Metab2010;95:3309–331733. Bose M, Teixeira J, Olivan B, et al. Weightloss and incretin responsiveness improve glu-cose control independently after gastric bypasssurgery. J Diabetes 2010;2:47–5534. Laferrere B, Heshka S, Wang K, et al. Incre-tin levels and effect are markedly enhanced 1month after Roux-en-Y gastric bypass surgery inobese patients with type 2 diabetes. DiabetesCare 2007;30:1709–1716

10 CR, Exercise, and Glucoregulation Diabetes Care