PMID: 41207299 IMPORTANCE Prevalenceofmetabolicsyndrome(MetS)definedas3ormore cardiometabolic risk factors has grown to 43%amongUSadults.Lifestyleinterventions producebenefits, but with limited sustainability. OBJECTIVE Todeterminewhethera6-monthhabit-basedlifestyleprogramaddsbenefitto education andactivity monitoring for sustained MetS remission at 24 months. DESIGN,SETTING,ANDPARTICIPANTS TheEnhancingLifestylesinMetabolicSyndrome(ELM) study wasasingle-blind, individually randomized clinical trial of participants with MetS recruited from July 2019 toJanuary2022at5locationsintheUSandfollowedupfor24 months.Eligible participants were motivated for lifestyle change and were contacted using electronic medical records and social or mass media; those whodidnotmeetmedicalor logistical criteria were excluded. In all, 14 817 adults were screened during a 2.5 year period. DatawereanalyzedfromMarch2024toMay2025. INTERVENTIONS Allparticipants received education and an activity monitor. Those in the intervention group also received 19 small group in-person meetings during 6 months,which sought toestablish the following habits as part of their daily routine: vegetables at meals, brisk walks, sensory awareness, and emotionregulation. Repetition of these habits was encouragedbyattention toparticipants’ experience of immediate benefits and peer support. Thecomparatorgroupreceived24evidence-basededucationalmailingsmonthlyduringthe 24months. MAINOUTCOMEANDMEASURESMetSremissionat24monthsdeterminedbyblinded laboratory evaluation and clinical examination. Secondary outcomes were cardiometabolic andlifestyle risk factors at 6 and 24 months. RESULTS Theanalysisincluded 618participants (mean [SD] age, 55.5 [11.0] years; 468 female [74.7%] and150male[24.3%]individuals), of whom306(49.5%)wererandomizedtothe intervention and 312 (50.5%) tothecomparator.The24-monthfollow-upperiodwas completedby517participants (83.7%). Obesity, measured bybodymassindex,waspresent in 513 participants (83%). At 6 months,theproportionachieving MetSremissionwas24.8% in intervention group and 17.9%incomparator(adjusted oddsratio, 1.64; 95%CI, 1.07-2.53; P =.03),supportedbyimprovementsinwaistcircumference,triglycerides, fasting glucose, bodymassindex,weight,hemoglobinA1c,MetSseverity,perceived stress, vegetable intake, moderate-intensity physical activity, daily steps, sensory awareness, and the habits of daily brisk walks and vegetables at meals. At24months,theproportionachievingsustainedMetS remission was 27.8%(85of306participants)inintervention groupand21.2%(66of312)in the comparator(adjusted oddsratio, 1.46; 95%CI, 1.01-2.14; P < .05), supported by sustained improvementinfasting glucose, vegetable intake, daily steps, sensory awareness, and the habit of daily brisk walks. CONCLUSIONSANDRELEVANCE ThisrandomizedclinicaltrialfoundthatsustainedMetS remission after treatment may bepossible bypromotingsimplehabitsthrougha behavior-based lifestyle program focused on immediate benefits. TRIALREGISTRATION ClinicalTrials.gov Identifier: NCT04036006 JAMAInternMed.doi:10.1001/jamainternmed.2025.5900 Published online November9,2025. VisualAbstract Multimedia Supplementalcontent AuthorAffiliations: Author affiliations are listed at the end of this article. GroupInformation:Themembersof the ELMTrialResearchGroupappear listed in Supplement 3. CorrespondingAuthor:LyndaH. Powell, PhD, DepartmentofFamily andPreventive Medicine, RushUniversity Medical Center, 1700WVanBurenSt, Ste 470,Chicago, IL 60612 (lpowell@rush.edu). (Reprinted) E1 Downloaded from jamanetwork.com by guest on 11/10/2025Research OriginalInvestigation Metabolicsyndrome(MetS) is diagnosed by the cooccurrence of 3 or more of 5 cardiometabolic risk factors: abdominal obesity, hypertension, elevated fasting glucose, elevated triglycerides, and low high-density lipoprotein(HDL)cholesterol.1IntheUS,theprevalenceofMetS in adults was 25.3% from 1988 to 1994 and increased 43.3% from2017to2018.2-4Additionally,MetShasgrownamongsubgroupssuchasyoungadults,AsianAmericanindividuals,and the college educated.3,4 MetS is a multisystem risk factor for chronic and infectious diseases5-8 and is associated with elevatedhealthcarecosts.7,9,10Poordietqualityandphysicalinactivity are strong modifiable risk factors.7,11-13 Three highqualitytrialswithlarge,diversepopulationsofparticipantswith MetS evaluated the impact of diet, with or without physical activity, on MetS remission for 24 months or longer.14-16 They achieved,onaverage,22%MetSremissionusinglifestyletreatments that continued at various levels of intensity over the duration of the follow-up. No trial has demonstrated sustained remission of MetS after treatment has ended. The Enhanced Lifestyles for Metabolic Syndrome (ELM) multisiteefficacytrialtestedthehypothesisthatmotivatedparticipants withMetSreceivinga6-monthhabit-basedlifestyle intervention in addition to education and activity monitoring would achieve a higher proportion of sustained 24month MetS remission than those receiving education and activitymonitoringalone.Secondaryobjectivescomparedcardiometabolicandlifestyleriskfactorsbetweengroupsat6and 24months. Methods AcentralinstitutionalreviewboardatRushMedicalCenterapproved the trial protocol (available in Supplement 1) and all amendments. Details of design, eligibility, recruitment, interventions, assessments, and baseline characteristics have been publishedelsewhere.17Trialorganizationincludedanindependentdatacoordinatingcenterandadataandsafetymonitoring board.Writteninformedconsentwasobtained(eMethods1and 2inSupplement 2), and the study followed the Consolidated StandardsofReportingTrials(CONSORT)reportingguideline18 and the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) reporting guideline.19 DesignandParticipants This was a single-blind, individually randomized behavioral efficacy clinical trial of participants withMetSrecruitedfrom 5diversegeographicregions:Rochester,NewYork;centraland northeast Pennsylvania; Chicago, Illinois; Kansas City, Missouri;andDenver,Colorado.Recruitmentwasconductedfrom July2019throughJanuary2022,withthetrialendinginApril 2024 as planned. To minimize bias from a single-blind design, blinding was extended to all operations using neutral names for trial groups, blinding to hypotheses, and training in equipoise.20 Eligibleparticipantswereage18yearsorolderandmetcriteria for MetS defined by the Joint Interim Statement.1 Exclusions were for medical safety (eg, cardiovascular risk, food E2 JAMAInternalMedicine PublishedonlineNovember 9,2025 (Reprinted) Lifestyle Intervention for Sustained Remission of Metabolic Syndrome KeyPoints Question Doesa6-monthhabit-basedgrouplifestyle intervention added to routine education and activity monitoring improvetheproportionofpatients with metabolic syndrome (MetS) whosustainremissionat24months? Findings Thismultisite randomized clinical trial of 618 participants with MetS found asignificant difference in sustained MetSremissionat24monthsinthelifestyleintervention group (28%)comparedwiththecomparatorgroup(21%). Meaning Additionofa6-monthhabit-forminglifestyle program to routine education andactivity monitoring shows that sustained MetSremissionafter treatment maybepossiblebypromoting simple habits and focusing onimmediatebenefits. allergies); logistical barriers (eg, inability to attend in-person meetings,noprimarycarephysician);motivation(eg,unwillingorunabletochangelifestyle);medicalconfounding(eg,diabetes,weightlossmedications,oralcorticosteroids);andmental health (eg, major depression, substance abuse, cognitive impairment).17 Potentialparticipants,recruitedin4wavesof150participantsperwave,wereidentifiedthroughelectronicmedicalrecords (334 [54%]), self-referred through social or mass media advertisements (247 [40%]), or from other miscellaneous sources (37 [6%]). Screening required attendance at an information session that described the time needed for lifestyle changeandprovidedtimetoconsiderpersonalprosandcons ofparticipation(eMethods1inSupplement2).21Thebaseline examination included clinic visits for medical eligibility and arun-inforexposuretotrialassessmentdemands.Inall,14817 adults were screened during a 2.5 year period. Eligible participants with confirmed MetS and continued interest were randomized by an unblinded statistician in a 1:1 ratio, stratified by site, and followed up for 24 months. COVID-19 restrictions required a pause in recruitment, remote assessments, and/or safety precautions for in-person examinations. Retention was maximized by obtaining alternative contacts, developing relationships with participants, minimizing transportation burden, reimbursing for assessment time, and conducting partial assessments with potential dropouts. Interventions CoreElementsinBothStudyGroups Both groups received education plus a Fitbit (Google Inc) activity monitor. The hypothesis was that education and activity monitoring were necessary but not sufficient to produce sustained MetS remission. ComparatorGroup Evidence-basededucationonMetSincludingnutrition,physical activity, and stress management12,22,23 was sent in 24 monthly tip sheets. A trained local coordinator made telephonecontactevery3monthstoprovidetrialupdatesandto answerqueries.Participantswerereferredtotipsheetsforlifestylequestions,andtotheirprimarycarepractitionerformediDownloaded from jamanetwork.com by guest on 11/10/2025 jamainternalmedicine.comLifestyle Intervention for Sustained Remission of Metabolic Syndrome cal questions.Driftfromthetelephonecontactsprotocolwas remediated in monthly supervisor meetings. Intervention Group Alifestyle programwasdevelopedover15yearsinwhichhabit formation became the key strategy and components; length, modeofadministration,anddosewerealsodetermined.24The rationaleforamindfulnesscomponentwastoenhancemotivationduringearlyhabitformationbyfocusingontheimmediate benefits of vegetables at meals and daily brisk walks. Dietitians conducted individual 60-minute dietary consultations at baseline and 6 months. The 6-month intervention wasdeliveredin19in-persongroupmeetingsof90minutedurationwith15participantspergroup.Themeetingswere co-led by a psychologist and a dietitian, and supported by a lay healthcoachwhoprovidedlogisticsassistance,including weighingparticipants.Therationaleforgrouptreatmentwas to encourage peer support and facilitate vicarious learning through watching peers make changes. Content progressed fromeating(ie,nutrition,whattoeat,howtocook,andemotional/opportunistic eating) to sustaining change with physicalactivityandongoingcontactswithhealth-consciouspeers. Thegoalwastomake4simplehabitsanautomaticpartofdaily routine: (1) vegetables at meals25; (2) daily brisk walks26,27; (3) sensory awareness of smells, colors, and tastes28-30; and (4) emotionregulationbypausingbeforereactingtostressor opportunistic eating.28-30 Correlations between each simple habit and the complex risk factor they were intended to represent ranged from 0.18 to 0.44.31 The strategy for new habit formation was to repeat a new habit in response to repeated daily cues until the habit became automatic.32,33 This was accomplished in group meetings with an ecologically valid format that mirrored daily activities. Participants walked together, practiced sensory awareness, prepared a vegetable dish,andatetogetherwhileincreasingsensitivitytoimmediatebenefitsofnewhabits.24,34Athome,dailyrepetitionofall 4habitswasfosteredbypeer-to-peeraccountabilityusingthe Fitbit Friends/Community apps. COVID-19 restrictions resultedintemporaryshiftstoremotemeetingsandimplementation of safety protocolsforin-persongroups.Driftfromthe protocol was remediated in monthly supervisor meetings. In the 18 months after the intervention, participants were offered monthly support meetings but no new content. These meetingscombinedallparticipantsrandomizedtotheintervention at a site, and were given inperson, hybrid, or remotely. Fidelity Monitoring Intervention fidelity assessed delivery and receipt.35 In the comparator group, delivery was assessed by rate of successful tip sheet mailings and 3-month telephone calls; the extenttowhichmaterialswerereadwasnottracked.Receiptwas assessed by frequency of problemswithtipsheetsortheFitbit reported during routine telephone calls. In the interventiongroup,deliverywasassessedbyobserverchecklistsfordelivery of requiredcomponentsateachgroupmeeting,andby contentandprocessvariablesratedfrom20%ofrandomlyselectedvideorecordingsfromeachsiteandrecruitmentwave. Receipt wasassessedbygroupattendanceandacomprehenjamainternalmedicine.com Original Investigation Research sionquestionnairecompletedbyparticipantsaftereachgroup meeting. Outcomes Self-reported sociodemographic information was collected usingdefinedoptionsthatincludedan“other”categorywith aspacetowrite-inaresponse.Raceandethnicityallowedfor assessment of balance between groups, subgroup differences, and generalizability of results. TheprimaryoutcomewasremissionofMetSat24months, definedaslessthan3cardiometabolicriskfactorsmeetingMetS diagnosticcriteria1:elevatedbloodpressure(systolic≥130mm Hgordiastolic≥85mmHgorantihypertensivedrugtherapy); abdominalobesity(waistcircumference≥102cmformaleand ≥88 cm for female individuals); elevated triglycerides (≥150 mg/dLordrugtreatment; to convert mg/dL to mmol/L, multiplyby0.0113);elevatedfastingglucose(≥100mg/dLordrug treatment;toconvertmg/dLtommol/L,multiplyby0.0555); andlowHDLcholesterol(<40mg/dLformaleand<50mg/dL forfemaleindividuals,ordrugtreatment;toconvertmg/dLto mmol/L, multiply by 0.0259). Blood was drawn after a 12hour fast and sent to anindependent andblinded laboratory for assessment of HDL cholesterol, triglycerides, and glucose. Blood pressure was the average of 3 measurements after a5-minuterest.36Waistcircumferencewastheaverageof 2 assessments on skin at the iliac crest.37 At baseline and 6 months,areportwassenttoparticipantsand,ifrequested,their primary care practitioner. Secondary clinical outcomes included MetS components; glycated hemoglobin (HbA1c); MetS severity38; weight; and body mass index (BMI; calculated as weight in kilograms divided by height in meters squared). Targets of the intervention included self-reported vegetable intake (US National Cancer Institute Vegetable Screener39); moderateintensity physical activity minutes per week and daily steps (accelerometer40,41); self-reported sensory awareness and emotion regulation (Observe and Non-React Facets of 5-Facet Mindfulness Questionnaire42); and self-reported automatic daily habits (Self-Report Habit Index43). Intervention satisfaction at 24 months was assessed as the percentage of respondents who would “definitely recommend this program to family and friends.”17 Outcome measures with scoring and cut points were previously published.17 Blinded telephone calls every 3 monthsascertained vital status, adverse events, health care utilization, and nontrial treatments.A15-monthassessmentofMetSwasconductedfor retention purposes only. Safety was evaluated monthly by a safetycommitteeandmonitoredbyanindependentdataand safety monitoringboard.Datawerecollectedandtransferred to the central secure database using 4 electronic systems, ie, SnapSurveyonlinedatacaptureforms;CentrePointforaccelerometry data; Quanum Portal for laboratory results; and Fitabase for Fitbit data. Statistical Analysis SampleSize Fortheprimaryoutcome(MetSinremissionat24months),we used6assumptions:(1)MetSremissionininterventiongroup, E3 Downloaded from jamanetwork.com by guest on 11/10/2025 (Reprinted) JAMAInternalMedicine PublishedonlineNovember 9,2025Research OriginalInvestigation Lifestyle Intervention for Sustained Remission of Metabolic Syndrome Figure 1. ConsortDiagramofStudyParticipantsinStudyofSustainedRemissionofMetabolicSyndrome(MetS) 14 817 Adults with potential MetS screened 12 021 Excluded 2796 Eligible for baseline examination 10 243 Not interested 797 Medical reasons 441 Logistical reasons 331 Behavioral reasons 209 Did not have MetS 2178 Excluded 1773 Not interested 297 Did not have MetS 297 Medical reasons 40 Behavioral reasons 2 Logistical reasons 618 Randomized 312 Randomized to the comparator group 262 Received allocated interventiona 306 Randomized to the treatment group 237 Received allocated interventionb 6 mo 265 Completed examination 40 No examination 1 Died 6 Withdrew 6 mo 262 Completed examination 34 No examination 0 Died 10 Withdrew 24 mo 262 Completed examination 3 No examination 15 Lost to follow-up 3 Diedc 29 Withdrewd 312 Included in analysis 24 mo 255 Completed examination 7 No examination 14 Lost to follow-up 0 Died 30 Withdrewe 306 Included in analysis 40%or more; MetS remission in comparator group, 20% or less; (2) minimum power of 90%; (3) significance level of P < .05;(4)mixedlogisticmodelwithclusteringin1group44;(5) intraclasscorrelationof0.15estimatedfrompriorreviews45,46; and (6) 20% loss to follow-up. The 40% remission in the intervention was clinically significant based on comparability, with50%adherencetodrugtherapy47,48anda35%orgreater successrateforUSFoodandDrugAdministrationweightloss studies,49 and accounted for erosion from the 54% MetS remissionobservedintheproofofconcept.24The20%MetSremission inthecomparatorwasderivedfromtheaverage15% MetS remission in education controls in lifestyle interventions for MetS with follow-up periods of 1 year or more.50-55 These assumptions produced a sample size of 572, rounded upwardtoasampleof600,andthenincreasedto618toavoid turning away those in the process of enrolling when the target of 600wasreached. aReceived 100%ofmailingsinthe first 6 months. bAttended atleast 50%ofsessionsin the first 6 months. cCause ofdeath: stage IVlungcancer, stage IV liver cancer, COVID-19. dUnknown,19;relocation, 5; logistics, 4; andhealth, 1. eUnknown,11;health, 10; logistics, 7; relocation, 1; and administrative, 1. remission at 24 months using mixed-effects logistic regressiondrawnfromthelogisticversionofmodelsbyCandlish,56 included a random effect for partial clustering, and adjustmentforprespecifiedbaselinefixed-effectcovariatesage,sex, race, ethnicity, education, income, geographic site, BMI, and numberofcomorbidities.44,57 Secondary analyses compared groups onoutcomesat6and24monthsinasimilarmanner. Analyses were performed using SAS, version 9.4 (SAS Institute) from March 2024toMay2025. Statistical Analysis Plan Thestatisticalanalysisplanisavailableinthetrialprotocolin Supplement1.TheprimaryanalysiscomparedgroupsonMetS E4 JAMAInternalMedicine PublishedonlineNovember 9,2025 (Reprinted) Theprespecifiedapproachtomissingdata(trialprotocolin Supplement 1) was to identify the pattern of missing data and link it to an appropriate imputation plan. Missing data were judgedtobemissing not at random because missingness at 24 monthscouldnotbepredictedbyanybaselineor6-monthmeasures. Theprimaryimputationmodelwaslastobservationcarriedforward,justifiedbyafailuretosupporttheconcernthatit is an optimistic estimate of treatment benefit.58 Missingness could not be explained by status on treatment targets such as MetS,weight,orBMIassessedposttreatmentat6months.Missing data were replaced with the most recent assessment from the 15-month, 6-month, or baseline examination. Sensitivity jamainternalmedicine.com Downloaded from jamanetwork.com by guest on 11/10/2025Table1.BaselineParticipantCharacteristics,inTotalCohortandbyStudyGroup Characteristic Total(N=618) Interventiongroup (n=306) Comparatorgroup (n=312) Demographic,No.(%) Age,mean(SD),y 55.5(11.0) 55.5(11.0) 55.5(11.0) Female 468(75.7) 239(79.1) 229(73.4) Male 150(24.3) 67(21.9) 83(26.6) Race(self-reported) AfricanAmerican/Black 106(17.1) 53(17.3) 53(17.0) Asian 9(1.5) 3(1.0) 6(1.9) White 456(73.8) 225(73.5) 231(74.0) Othera 47(7.6) 25(8.2) 22(7.1) Hispanic/Latineethnicity 60(9.7) 33(10.8) 27(8.7) Lessthancollegedegree 250(40.5) 127(41.5) 123(39.4) Marriedorcohabitating 381(61.7) 191(62.4) 190(60.9) Familyincome<$90000 288(45.6) 145(47.4) 143(45.8) Clinical,mean(SD)b Waistcircumference,cm 115.1(14.0) 114.7(14.1) 115.5(14.0) Systolicbloodpressure,mmHg 128.1(15.4) 127.6(15.6) 128.6(15.1) Diastolicbloodpressure,mmHg 83.1(10.0) 82.4(10.1) 83.8(9.9) Triglycerides,mg/dL 175.3(96.1) 179.6(88.4) 171.1(103.1) Fastingglucose,mg/dL 100.0(12.6) 100.0(12.2) 100.1(13.0) HDLcholesterol,mg/dL 47.0(11.1) 47.4(11.8) 46.6(10.5) Medications,No.(%) Bloodpressure 398(64.4) 191(62.4) 207(66.4) Triglycerides 24(3.9) 15(4.9) 9(2.9) Glucose(metformin) 59(9.6) 31(10.1) 28(9.0) HDLcholesterol 21(3.4) 12(3.9) 9(2.9) Bodymassindexc 36.5(7.0) 36.3(7.1) 36.7(6.9) Weight,kg 102.0(21.8) 100.5(21.6) 103.4(22.0) LDLcholesterol,mg/dL 116.5(33.9) 116.5(34.8) 116.5(33.2) HemoglobinA1c,d% 5.6(0.4) 5.6(0.4) 5.6(0.4) MetSseveritye 0.97(0.52) 0.98(0.53) 0.96(0.51) Currentsmoking,No.(%) 30(4.9) 15(4.9) 15(4.8) Qualityoflifef Perceivedhealth 63.9(17.4) 63.3(17.0) 64.6(17.8) Energy/vitality 56.0(18.4) 55.4(19.1) 56.5(17.7) Depressivesymptoms 2.3(2.3) 2.4(2.4) 2.3(2.2) Perceivedstress 19.9(6.7) 20.2(6.8) 19.6(6.6) Lifestyletargetsofinterventiong Vegetableintake,servings/d 3.0(2.2) 3.1(2.2) 2.9(2.1) Moderate-intensityphysicalactivity,min/wk 104.9(97.6) 112.7(102.4) 97.2(92.1) Dailysteps 4363(2228.0) 4583.6(2327.7) 4146.7(2107.8) Sensoryawareness 3.5(0.6) 3.5(0.6) 3.5(0.6) Emotionregulation 3.5(0.6) 3.5(0.6) 3.5(0.6) Automatichabits Briskwalks 2.8(1.0) 2.8(0.9) 2.8(1.0) Vegetablesatmeals 2.8(1.0) 2.8(0.9) 2.8(0.9) Noticesensoryexperiences 3.7(0.9) 3.7(0.9) 3.8(0.8) Pausebeforereacting 3.3(0.8) 3.3(0.8) 3.3(0.8) Abbreviations:HDL,high-density lipoprotein;LDL,low-density lipoprotein;MetS,metabolic syndrome. aIncludedAmericanIndian/Alaska Native,NativeHawaiian/other PacificIslander,multiracial,other, andnoresponse. bMetScutpoints:waist circumference, 102cm(male)and 88cm(female);systolicblood pressure, 130mmHgortaking antihypertensivemedication(for mmHgtokPa,multiplyby0.133); diastolicbloodpressure, 85mm Hgortakingantihypertensive medication(formmHgtokPa, multiplyby0.133);triglycerides, 150mg/dLortreatmentfor elevatedtriglycerides(formg/dLto mmol/Lmultiplyby0.0113); glucose, 100-125mg/dLortaking metformin(upperlimitexcludes diabetes;formg/dLtommol/L, multiplyby0.0555); HDLcholesterol,<40mg/dL(male) andor<50mg/dL(female),or takingmedicationforlowHDL(for mg/dLtommol/L,multiply by0.0259). cWeight(kg)dividedbyheightin meterssquared(overweight,25to <30;obese, 30). dNormal,<5.7%;prediabetic, 5.7%-6.4%;diabetic, 6.5%. eTotalcohortrange,−0.51to2.60 (higherscoresindicatemore severe). fAssessmenttools:perceivedhealth, energy/vitality,ShortForm-36, (scale,0-100;higherscoresindicate betterquality);depressive symptoms,PatientHealth Questionnaire-8(scale0-24;higher scoresindicatemoresymptoms); andperceivedstress,Perceived StressScale(scale,0-56;higher scoresindicatemorestress). gAssessmenttools:vegetableintake (servings/d),NationalCancer InstituteAllDayScreener,vegetable subscale(higherscoresindicate moreservings);physicalactivity, ActigraphWGT3X-BTaccelerometer (guidelinesare150min/wkof moderate-intensityphysicalactivity and7500steps/d);sensory awarenessandemotionregulation, 5-FactorMindfulnessquestionnaire; andautomatichabits,Self-Report HabitIndex(scale1-5;higherscores indicatemoreautomatichabit). LifestyleInterventionforSustainedRemissionofMetabolicSyndrome OriginalInvestigation Research jamainternalmedicine.com (Reprinted) JAMAInternalMedicine PublishedonlineNovember9,2025 E5 Downloaded from jamanetwork.com by guest on 11/10/2025Research OriginalInvestigation Figure 2. MetabolicSyndromeRemission,byStudyGroupandTime 35 Metabolic syndrome remission, % 30 25 20 15 10 5 0 a 6 a Treatment Comparator 0 24 Time, mo Remission at 6months:adjustedoddsratio,1.64(95%CI, 1.07-2.53; P = .03); andremission at24months:adjustedoddsratio, 1.46(95%CI, 1.01-2.14; P <.05).Adjustmentsforclustering and prespecified covariates age, sex, race, ethnicity, education, income, geographic site, body mass index, and number of comorbidities. aP < .05. analyses using pattern mixture models for a series of tipping points59 are presented in eMethods 3inSupplement2. Results The analysis included 618 participants (mean [SD] age, 55.5 [11.0] years; 468 female [74.7%] and 150 male [24.3%] individuals), whowererandomized,306(49.5%)totheintervention and 312 (50.5%) to the comparator (Figure 1). The 24month follow-up period was completed by 517 participants (83.7%). The loss, withdrawal, or death rate of 16.3% (101 of 618) at 24 months was similar across groups and lower than the 20% assumed in sample size calculations. Table 1 presents comparability between groups at baseline. Demographic,clinical,andlifestylecharacteristicsdidnotdiffersignificantly between groups. Figure2andTable2presentMetSremissionbygroupand time. At 6 months, the proportion of MetS remissions was higherintheintervention(76of306[24.8%])thaninthecomparator(56of312[17.9%]);adjustedoddsratio[OR]1.64;95% CI, 1.07-2.53; P = .03). At 24 months, the proportion of MetS remissionswashigherintheintervention(85of306[27.8%]) thaninthecomparator(66of312[21.2%])group,withanadjusted OR of 1.46 (95% CI, 1.01-2.14; P < .05). Approximately 15 patients needtoreceivetheinterventiontoachieve1MetS remission.Forcompleteness,eFigure1inSupplement2shows MetSremissionwith15-monthdataadded,butitspurposewas for retention only, and it had a high rate of missing data. Secondary end points are presented in Table 2 as groupspecific means or percentages with adjusted P values for differences, and also in the eMethods 5 in Supplement 2 as adjusted effect size estimates of the difference between groups (aES),with95%CIs.Amongtheclinicalendpointsat6months, theinterventionwassuperiortothecomparatoronwaistcircumference,triglycerides,fastingglucose,BMI,weight,HbA1c, MetS severity, and quality of life. At 24 months, statistically E6 JAMAInternalMedicine PublishedonlineNovember 9,2025 (Reprinted) Lifestyle Intervention for Sustained Remission of Metabolic Syndrome significant differences between the intervention and comparator groups were sustained only for fasting glucose (97.7 mg/dL vs 100.7 mg/dL; aES, −2.74; 95% CI, −5.31 to −0.16; P =.04).Theinterventionwasassociatedwithareductionin blood pressure medication over the comparator (192 of 306 [62.7%] vs 223 of 312 [75.5%], respectively; aES, −0.06; 95% CI, −0.103 to −0.009; P = .02). Table 2 and Figure 3 present lifestyle intervention targets, by groupandtime.At6months,theinterventiongroup was significantly different from the comparator on all lifestylefactorsexceptemotionregulation.At24months,significantdifferencesweresustainedforvegetableintake(4.0vs3.4; aES, 0.52; 95%CI,0.04-0.99; P = .04),daily steps (4822.6 vs 4157.8; aES, 432.5; 95% CI, 66.4-798.6; P = .02), and sensory awareness(3.8vs 3.6;aES,0.17;95%CI,0.08-0.26;P < .001). Amongdailyhabits,briskwalksandvegetablesatmealswere significant at 6 months and sustained at 24 months for brisk walksonly.Interventionsatisfaction(ie,percentagewhowould definitely recommend this intervention) was higher in intervention than in the comparator group (172 of 218 [78.9%] vs 97 of205[47.3%]; P < .001). Subgroup analyses evaluated odds of MetS remission for intervention vs comparator groups within prespecified subgroupsat6and24months(eFigure2inSupplement2).There wasnoobservedheterogeneityoftreatmenteffectacrosssubgroups. Interventionfidelitywasexcellent.Inthecomparatorgroup, 84%oftheparticipantsreceived100%ofthemailingsinthefirst 6 months, 92% of the 3-month telephone calls were completed,and292participants(94%)reportednoproblems.Inthe interventiongroup,98%(299of306participants)and70%(182 of 262) completed individual dietary consultations at baseline and6months,respectively.Duringthe6-monthintensivephase, deliveryofrequiredcontentwasaccomplishedin94%ofmeetings with an average process score of 28.6 of 31.0 points. Each participant attended an average of 67% of group meetings and comprehensionofcontentaveraged4.4on5-pointscale.Averageattendanceatthefollow-upsupportmeetingsbetween6and 24monthswas37%. Nosafetyconcernswerereported.Therewere437blinded adjudicationsthatwereclassifiedas115seriousadverseevents, 290 adverse events, and 32 nonevents. No events were determinedtoberelatedtotrial participation. Discussion MetS is growing and is fundamentally affected by lifestyle. Lifestyle interventions have achieved MetS remission but sustaining it after treatmentendsisachallenge.60Theevolvingscienceofsustainedbehaviorchangesuggeststhatunique strategiesareneededtoachievesustainability,61oneofwhich is new habit formation.32,33 ELM was designed to determine whether a habit-formation program could augment educationandactivitymonitoringtoachievesustainedMetSremission over a long-term follow-up period after treatment. The primary result was that 28% of participants in the intervention vs 21% in the comparator group were in MetS Downloaded from jamanetwork.com by guest on 11/10/2025 jamainternalmedicine.comremissionat24months.Theinterventionproduced25%MetS remissionafteritsintensive6-monthphaseandsustainedit withouttreatmentforthenext18months.Thisresultismeaningfulwhencomparedtotheaverage22%MetSremissionobservedinlifestyletrialsofsimilarparticipantswhoreceived sometypeofongoingtreatmentoverthecourseoftheir follow-up.14-16SustainedMetSremissionwassupportedbysustainedimprovementinfastingglucose.AlthoughlifestyleinterventionsforMetShavebeenshowntoimproveglucose,60 ELMisthefirsttoshowsustainedimprovementafteractive treatmentended.Differencesbetweengroupsinweightloss anditscorrelatesat6monthswerenolongersignificantat24 months,whichsuggeststhattheyweremoreimportantforinitiationthansustainingMetSremission—anobservationwe madepreviously.24 Themostconvincingsupportforlifestylechangewasin dailysteps.At24months,participantsintheintervention groupwalkedanaverageof4823dailysteps,thatis,665steps Table2.SecondaryOutcomes,byStudyGroupandTime Outcome Follow-uptime 6mo 24mo Interventiongroup Comparatorgroup Pvaluea Interventiongroup Comparatorgroup Pvaluea Participants,No. 306 312 NA 306 312 NA Participantswithcompletedata, No.(%) 262(85.6) 265(84.9) NA 255(83.3) 262(84.0) NA Clinical,mean(SD) Metabolicsyndromeremission, No./TotalNo.(%) 76/306(24.8) 56/312(17.9) .03 85/306(27.8) 66/312(21.2) .05 Waistcircumference,cm 110.2(14.7) 113.6(14.1) <.001 111.7(15.1) 113.1(14.0) .17 Systolicbloodpressure,mmHg 125.7(15.7) 125.3(15.0) .40 125.7(15.9) 125.2(15.7) .30 Diastolicbloodpressure,mmHg 80.4(10.9) 81.1(10.9) .78 80.0(11.3) 80.7(10.3) .95 Triglycerides,mg/dL 154.0(75.8) 159.9(92.1) .008 163.1(93.5) 161.8(101.2) .52 Fastingglucose,mg/dL 97.9(11.1) 100.4(13.7) .01 97.7(13.5) 100.7(15.5) .04 HDLcholesterol,mg/dL 49.0(11.7) 47.5(10.3) .13 49.5(12.4) 48.4(11.0) .40 Bodymassindex 34.9(7.2) 36.4(7.0) <.001 35.4(7.3) 36.0(7.0) .20 Weight,kg 96.6(21.9) 102.3(21.9) <.001 97.7(22.0) 101.1(21.9) .22 LDLcholesterol,mg/dL 113.2(34.7) 114.8(34.1) .39 110.0(36.3) 112.7(35.9) .25 HemoglobinA1c,% 5.5(0.4) 5.6(0.4) .02 5.6(0.4) 5.7(0.4) .16 Metabolicsyndromeseverity 0.7(0.6) 0.8(0.6) <.001 0.7(0.7) 0.8(0.6) .06 MedicationsNo./TotalNo.(%) Bloodpressure 186/306(60.8) 209/312(67.0) .07 192/306(62.7) 223/312(71.5) .02 Triglycerides 13/306(4.3) 9/312(2.9) .39 13/306(4.3) 9/312(2.9) .39 Glucose(metformin) 24/306(7.8) 26/312(8.3) .88 29/306(9.5) 35/312(11.2) .51 HDLcholesterol 10/306(3.3) 9/312(2.9) .82 10/306(3.3) 9/312(2.9) .82 Qualityoflife,mean(SD) Perceivedhealth 70.6(16.9) 67.2(17.7) <.001 68.4(17.2) 67.3(17.8) .13 Energy/vitality 61.5(19.3) 56.3(18.6) <.001 59.6(19.9) 57.8(19.5) .12 Depressivesymptoms 2.4(3.0) 2.8(3.1) .07 2.9(3.4) 2.7(3.0) .71 Perceivedstress 19.1(6.7) 20.0(7.1) .006 19.4(7.2) 19.4(7.2) .36 Lifestyletargetsofintervention,mean(SD) Vegetableintake,servings/d 4.8(3.5) 3.5(2.6) <.001 4.0(4.0) 3.4(2.4) .03 Moderate-intensityphysical activity,min/wk 127.5(122.6) 100.4(103.1) .02 122.5(164.3) 101.3(123.6) .18 Dailysteps 5356.0(2956.4) 4316.2(2417.0) <.001 4822.6(2784.6) 4157.8(2393.2) .02 Sensoryawareness 3.8(0.6) 3.6(0.7) <.001 3.8(0.7) 3.6(0.6) <.001 Emotionregulation 3.5(0.6) 3.5(0.6) .87 3.6(0.6) 3.5(0.6) .48 Automatichabits Briskwalks 3.3(0.9) 3.0(0.9) <.001 3.3(0.9) 3.0(1.0) .001 Vegetablesatmeals 3.4(0.9) 3.2(0.9) .004 3.4(1.0) 3.3(1.0) .12 Noticesensoryexperiences 3.8(0.8) 3.8(0.8) .30 3.9(0.8) 3.8(0.8) .20 Pausebeforereacting 3.4(0.8) 3.3(0.8) .12 3.4(0.8) 3.3(0.9) .94 Interventionsatisfaction, No./TotalNo.(%) NA NA NA 172/218(78.9) 97/205(47.3) <.001 Abbreviations:HDL,high-densitylipoprotein;LDL,low-densitylipoprotein; NA,notapplicable. aAdjustedforclusteringandtheprespecifiedbaselinecovariatesofage,sex, raceandethnicity,education,income,geographicsite,bodymassindex,and numberofcomorbidities. LifestyleInterventionforSustainedRemissionofMetabolicSyndrome OriginalInvestigation Research jamainternalmedicine.com (Reprinted) JAMAInternalMedicine PublishedonlineNovember9,2025 E7 Downloaded from jamanetwork.com by guest on 11/10/2025morethanthoseinthecomparator.Althoughlessthanthegoal of7500steps(atwhichhealthbenefitsplateau62)andlessthan thedefinitionofsedentarybehavior(≤5000stepsdaily),63recentmeta-analysesfoundthatwalkingmorethan4000steps dailyreducedriskofall-cause,cardiovascular,andcancer mortality,26,64anddailystepswerebetterpredictorsofmortalitythanintensityofphysicalactivity.26Significantdifferencesbetweengroupswereobservedforsensoryawareness andthehabitofdailybriskwalks.Takentogether,thesimple habitofdailybriskwalksfortheirimmediatebenefitsmaybe apromisingalternativeforsustainabilitythansettingmoreintensivegoalstoachievefuturehealthgains. Participantsinthecomparatorgroupunderwentgreater MetSremissionthanexpected.ReceivingonlyaFitbitactivitymonitorand24monthlymailededucationtipsheets, 21%achievedMetSremissionat24months,higherthanthe Figure3.LifestyleInterventionTargets,byStudyGroupandTime 5.0 4.0 4.5 3.0 3.5 2.5 Mean vegetable intake, servings/d Time, mo Comparator Treatment 24 0 6 a b Vegetable intake A 140 130 110 120 80 90 100 Mean moderate-intensity physical activity, min/wk Time, mo Comparator Treatment 24 0 6 b Moderate-intensity physical activity B 6000 5500 4000 4500 5000 Mean daily steps Time, mo Comparator Treatment 24 0 6 a b Daily steps C 3.9 3.8 3.7 3.6 3.5 3.4 Observe subscale, mean score Time, mo Comparator Treatment 24 0 6 a a Sensory awareness D 3.9 3.8 3.7 3.6 3.5 3.4 Nonreact subscale, mean score Time, mo 24 0 6 Emotion regulation E Treatment Comparator Statisticaltestsadjustedforclusteringandprespecifiedcovariates:age,sex, race,ethnicity,education,income,geographicsite,bodymassindex,and numberofcomorbidities.Sensoryawarenessandemotionalregulation assessedbythe5-FactorMindfulnessquestionnaireastheaverageitem responsetoa5-pointLikertscale,withhigherscoresrepresentingmore mindfulness.Rangeintotalcohort:sensoryawareness,1.6-5.0,andemotion regulation,1.4-5.0. aP<.001. bP<.05. Research OriginalInvestigation LifestyleInterventionforSustainedRemissionofMetabolicSyndrome E8 JAMAInternalMedicine PublishedonlineNovember9,2025 (Reprinted) jamainternalmedicine.com Downloaded from jamanetwork.com by guest on 11/10/2025Lifestyle Intervention for Sustained Remission of Metabolic Syndrome average 15% remission observed in previous education controls.50-55Thiscouldbeexplainedbythehighlevelofmotivation in the ELMcohort.Screeningrequiredpotential participants to be informed about the time needed to make lifestylechanges,tohavesufficienttimetoreflectontheirpersonal prosandcons,andtocompletearun-inexposingthemtoassessmentdemands.Thosewhoremainedinterestedmayhave hadsufficientmotivationtomakelifestylechangeswithminimal intervention.65 Activity monitoring, by itself, has been showntoincrease daily steps.66 Participants in the intervention underwent less MetS remission than hypothesized. It is possible that COVID-19 restrictions during the trial, which required paused, hybrid, and socially distancedmeetings,mayhavediluteditseffects.However, 28% MetS remission has clinical relevance when comparedtothe1711participantswithMetSintheDiabetesPrevention Program(DPP).16Bothtrialshadstrict eligibility to protect internal validity, with DPP enrolling only 2%ofscreened67and ELMenrolling 4% of screened patients. The trials had almost comparable participants with MetS and featured intensive 6-monthlifestyleinterventionswith18-monthfollow-ups.However,DPPfeaturedintensivefollow-upafterinitialtreatment,includingongoinggroupcourses,motivationalcampaigns,andas manyindividualcontactsasneeded,68whereasELMofferedonly voluntarysupportcontactstolargegroupsofparticipants.MetS remissionat24monthswas11%intheDPPlifestylegroupcomparedto28%intheELMinterventiongroup.Thus,ELMachieved greater long-term remission than DPP despite having stopped the intervention. Strengths andLimitations Amongthis trial’s strengths are that ELM fills the gap left by too few efficacy trials and inadequate follow-up periods as identified in systematic reviews of lifestyle interventions for MetS.60,65,69,70 ELM featured rigorous methods for singleblind behavioral efficacy trials,20 including progressive development of treatment,24 expansion of blinding to minimizesingle-blindbias,20andpriorityonhighretentionovera long follow-up. The cohort was large and diverse, recruited from5geographicareasintheUSandwithrepresentationby sex, race, ethnicity, education, and rural residence. The priOriginal Investigation Research maryresultof28%MetSremissionobserved18monthsafter theendoftreatmentcomparesfavorablytopasttrialsofsimilar patients receiving ongoing intervention, is supported by improvements in secondary outcomes, did not differ significantly in any subgroups, and incurred no safety risks. Limitationsmustbeconsidered.71Theresultwasstatisticallysignificantbutmodest.The24-monthORof1.46wassurroundedbyawide95%CIof1.01to2.14inacomparisonof2 active treatments, only one of which ended after 6 months. Nonspecific attention may have explained results for treatmentsthatdifferedinintensityofcontact.Thetrialmayhave been overpowered to detect small differences because assumptionsaboutlossesandtheintraclasscorrelationinsample size calculations weremoreconservativethanobserved.The effect size may have been diluted by the differential impact of COVID-19restrictions on the in-person meetings in the intervention group, which did not affect the remote comparator. Anintensivein-personprogrammaybedifficulttoimplementandnotcost-effective,suggestingthevalueofexploring fully remote versions. Conclusions This randomized clinical trial found that sustained MetS remissionis possible using theELMstrategyofsimplifyingtargetsandmotivatingintegrationofimprovedlifestylehabitsinto the daily routine by focusing on immediate benefits. Interventionsthatproducesustainedhealthbenefitsovertimeare needed60;however,targetinglifestylealonemaylimitthepotentialstrengthofaninterventionforMetS.Thenewglucagonlike peptide-1 obesity drugs reduce caloric intake—only during treatment72—and appear to act in synergy with lifestyle interventions.73,74Itiscompellingtoconsiderthesuccessand cost-effectivenessthatcouldresultfromcombiningadrugthat works while on treatment with a lifestyle intervention that works after treatment is discontinued. At a more basic level, changes in thesocioecological systems associated with MetS areneeded.75However,untilthesebasicchangesaremade,the prevalence of MetS continues to rise and evidence-based interventions are needed urgently.