Effect of a Lifestyle InterventionProgram With Energy-RestrictedMediterranean Diet and Exerciseon Weight Loss andCardiovascular Risk Factors:One-Year Results of thePREDIMED-Plus TrialDiabetes Care 2019;42:777–788 | https://doi.org/10.2337/dc18-0836

OBJECTIVE

The long-term impact of intentional weight loss on cardiovascular events remainsunknown. We describe 12-month changes in body weight and cardiovascular riskfactors in PREvencion con DIeta MEDiterranea (PREDIMED)-Plus, a trial designedto evaluate the long-term effectiveness of an intensive weight loss lifestyle in-tervention on primary cardiovascular prevention.

RESEARCH DESIGN AND METHODS

Overweight/obeseadultswithmetabolic syndromeaged55–75years (n=626)wererandomized to an intensive weight loss lifestyle intervention based on an energy-restrictedMediterranean diet, physical activity promotion, and behavioral support(IG) or a control group (CG). The primary and secondary outcomes were changes inweight and cardiovascular risk markers, respectively.

RESULTS

Diet and physical activity changes were in the expected direction, with significantimprovements in IG versus CG. After 12 months, IG participants lost an average of3.2 kg vs. 0.7 kg in the CG (P < 0.001), a mean difference of 22.5 kg (95% CI 23.1to21.9).Weight loss‡5%occurred in33.7%of IGparticipants comparedwith11.9%in the CG (P < 0.001). Compared with the CG, cardiovascular risk factors, includingwaist circumference, fasting glucose, triglycerides, and HDL cholesterol, signifi-cantly improved in IG participants (P < 0.002). Reductions in insulin resistance,HbA1c, and circulating levels of leptin, interleukin-18, andMCP-1 were greater in IGthan CG participants (P < 0.05). IG participants with prediabetes/diabetes signif-icantly improved glycemic control and insulin sensitivity, along with triglyceridesand HDL cholesterol levels compared with their CG counterparts.

CONCLUSIONS

PREDIMED-Plus intensive lifestyle intervention for 12 months was effective indecreasing adiposity and improving cardiovascular risk factors in overweight/obeseolder adults with metabolic syndrome, as well as in individuals with or at risk fordiabetes.

1Department of Biochemistry andBiotechnology,Human Nutrition Unit, Pere Virgili Institute forHealth Research (IISPV), University Hospital ofSant JoandeReus,Rovira iVirgiliUniversity, Reus,Spain2Centro de Investigacion Biomedica en Red deFisiopatologıa de la Obesidad y la Nutricion(CIBEROBN), Instituto de Salud Carlos III, Madrid,Spain3Department of Preventive Medicine and PublicHealth, University of Navarra, Navarra Institutefor Health Research (IdiSNA), Pamplona, Spain4Primary Care Division, Catalan Institute ofHealth, Institut d’Investigacio i Recerca en AtencioPrimaria (IDIAP)-Jordi Gol, Barcelona, Spain5Cardiovascular Risk and Nutrition ResearchGroup, Hospital del Mar Medical Research In-stitute (IMIM), Barcelona, Spain6Department of Preventive Medicine, Universityof Valencia, Valencia, Spain7Research Institute of Biomedical and HealthSciences, University of Las Palmas de GranCanaria, Las Palmas de Gran Canaria, Spain8Department of Nursing, School of Health Sci-ences, University of Malaga-Institute of Biomed-ical Research in Malaga (IBIMA), Malaga, Spain9Instituto de Investigacion Sanitaria Illes Balears(IdISBa), University Hospital of Son Espases, Palmade Mallorca, Spain10Department of Internal Medicine, HospitalClınic, Institut d’Investigacions BiomediquesAugust Pi Sunyer (IDIBAPS), University of Barce-lona, Barcelona, Spain11Endocrinology and Nutrition Department,Hospital Clinic Universitari, Barcelona, Spain12Institut d’Investigacions Biomediques AugustPi Sunyer (IDIBAPS), Barcelona, Spain

Jordi Salas-Salvado,1,2

Andres Dıaz-Lopez,1,2

Miguel Ruiz-Canela,2,3 Josep Basora,1,2,4

Montse Fito,2,5 Dolores Corella,2,6

Luıs Serra-Majem,2,7 Julia Warnberg,2,8

Dora Romaguera,2,9 Ramon Estruch,2,10

Josep Vidal,11,12,13

J. Alfredo Martınez,2,14,15

Fernando Aros,2,16 Clotilde Vazquez,2,17

Emilio Ros,2,18 Jesus Vioque,19,20

Jose Lopez-Miranda,2,21

Aurora Bueno-Cavanillas,20,22

Josep A. Tur,2,23 Francisco J. Tinahones,2,24

Vicente Martın,20,25,26 Jose Lapetra,2,27

Xavier Pinto,2,28 Lidia Daimiel,29

Miguel Delgado-Rodrıguez,20,30

Pilar Matıa,31 Enrique Gomez-Gracia,2,32

Javier Dıez-Espino,2,3,33 Nancy Babio,1,2

Olga Casta~ner,2,5 Jose V. Sorlı,2,6

Miquel Fiol,2,9 Marıa Angeles Zulet,2,14

Monica Bullo,1,2 Albert Goday,2,5,34 and

Miguel A. Martınez-Gonzalez,2,3,35 for thePREDIMED-Plus investigators*

Diabetes Care Volume 42, May 2019 777

EVOLU

TIONOFNUTR

ITIONALTH

ERAPY

The current obesity pandemic entails amajor impact on global morbidity andpremature mortality linked to chronicdiseases while severely impairing thequality of life of affected individualsand posing a significant burden to thehealth system (1,2). Thus, effective strat-egies to reduce the burden of obesityand its adverse health consequencesare urgently needed.Prior research has shown that mod-

erate weight loss (5–10% of initial bodyweight) achieved through lifestyle changesis associated with improvement of car-diometabolic abnormalities character-istic of overweight/obesity (3,4) andreduction in the risk of type 2 diabetes(5). Expectedly, greater weight losseslead to greater cardiometabolic benefit(6,7). However, there are inconsistenciesin the association between overweight/obesity and total or cardiovascular mor-tality, particularly after a controversialsystematic review (8). Additionally, thelack of benefit on cardiovascular eventsor mortality of a recent large interven-tion trial focused on weight loss amongindividuals with type 2 diabetes castsdoubts on the long-term cardiovascularimpact of weight loss (9).The current recommendation for pa-

tientswithoverweightorobesity, especially

if they harbor the metabolic syndrome(MetS), is to establish a plan for weightloss through lifestyle changes. Althoughsuccessful weight loss is expected toreduce cardiovascular risk, no large ran-domized controlled trial (RCT) has everdemonstrated that long-term intentionalweight loss reduces the incidence ofcardiovascular disease (CVD) events. Inthe only available trial (Look AHEAD)conducted to assess the long-term ef-fects on CVD of weight loss and physicalactivity (PA), the results were null (9).Look AHEAD was conducted exclusivelyamong individuals with diabetes, and thechoice of a low-fat diet as active inter-vention has been argued as one expla-nation for its lack of cardiovascularbenefit (10). Although not focused onweight loss, the Women’s Health Initia-tiveDietaryModification Trial also used alow-fat diet and failed to show anybenefit on cardiovascular events (11).Another nutritional strategy used tolose weight is carbohydrate restriction.However, low-carbohydrate diets areusually rich in saturated fat and poorin fiber and mineral content, are associ-ated with increased LDL cholesterol, andtend to lose their weight-reducing effi-cacy after 12 months (12). Another di-etary paradigm that may be an effective

alternative to low-fat or low-carbohy-drate diets in terms of weight loss isthe Mediterranean diet (MedDiet), inwhich the quality of fat and carbohy-drates is more important than theamounts of these macronutrients (13).A meta-analysis of RCTs suggests thatthe MedDiet is a useful tool to reducebody weight and obesity-related meta-bolic alterations, particularly when totalenergy intake is restricted (14). The Di-etary Intervention Randomized Con-trolled Trial (DIRECT) trial and a recentmeta-analysis (15,16) also provided fur-ther evidence supporting the strongerbeneficial effects of the MedDiet onweight loss and long-term maintenanceof modest weight loss as compared witha low-fat diet.

The PREvencion con DIeta MEDi-terranea (PREDIMED) intervention trialcontributed to a large body of availableevidence supporting the effectivenessof the MedDiet for cardiovascular pre-vention (17). A cardiovascular benefit ofthe MedDiet is plausible due to its ben-eficial impact on classical and emergentcardiovascular risk factors, such as hy-pertension, dyslipidemia, increased ad-iposity, MetS, type 2 diabetes, insulinresistance, and markers of oxidativestress, inflammation, and endothelial

13Centro de Investigacion Biomedica en Red deDiabetes y EnfermedadesMetabolicas Asociadas(CIBERDEM), Institutode Salud Carlos III,Madrid,Spain14Department of Nutrition, Food Sciences, andPhysiology, Center for Nutrition Research, Uni-versity of Navarra, Pamplona, Spain15Madrid Institute for Advanced Studies (IMDEA)Food Institute, Madrid, Spain16Department of Cardiology, University HospitalAraba, Vitoria, Spain17Department of Endocrinology andNutrition, Univer-sity Hospital Fundacion Jimenez Dıaz, Madrid, Spain18Lipid Clinic, Endocrinology and Nutrition Ser-vice, Institut d’Investigacions BiomediquesAugust Pi Sunyer (IDIBAPS), Hospital Clınic, Uni-versity of Barcelona, Barcelona, Spain19Nutritional Epidemiology Unit, University of Mi-guel Hernandez, Isabial-Fisabio, Alicante, Spain20Centro de Investigacion Biomedica en Red deEpidemiologıa y Salud Publica (CIBERESP), In-stituto de Salud Carlos III, Madrid, Spain21Lipids and Atherosclerosis Unit, MaimonidesBiomedical Research Institute of Cordoba(IMIBIC), Reina Sofia University Hospital, Univer-sity of Cordoba, Cordoba, Spain22Department of Preventive Medicine and PublicHealth, University of Granada, Granada, Spain23Research Group on Community Nutrition andOxidative Stress, University of the Balearic Is-lands, Palma de Mallorca, Spain

24Department of Endocrinology and Nutrition,Virgen de la Victoria Hospital, University ofMalaga-Institute of Biomedical Research inMalaga (IBIMA), Malaga, Spain25Research Group on Gene-Environment Interac-tions and Health, University of Leon, Leon, Spain26Biomedicine Institute (IBIOMED), University ofLeon, Leon, Spain27Department of FamilyMedicine, ResearchUnit,Distrito Sanitario Atencion Primaria Sevilla, Sev-illa, Spain28Lipid Unit, Department of Internal Medicine,BellvitgeBiomedicalResearch Institute (IDIBELL)-Hospital Universitari de Bellvitge, L’Hospitaletde Llobregat, Barcelona, Spain29Nutritional Genomics and Epigenomics Group,Madrid Institute for Advanced Studies (IMDEA)Food Institute, Campus of International Excel-lence UAM+CSIC, Madrid, Spain30Department of Health Sciences, University ofJaen, Jaen, Spain31Endocrinology and Nutrition Department, Fac-ultad de Medicina, Hospital Clınico San Carlos-Instituto de Investigacion Sanitaria del HospitalClınico San Carlos (IdISSC), Universidad Complu-tense de Madrid, Madrid, Spain32Department of Preventive Medicine andPublic Health, University of Malaga, Malaga,Spain33Atencion Primaria, Servicio Navarro de Salud–Osasunbidea, Pamplona, Spain

34Department of Medicine, University of Barce-lona, Barcelona, Spain35Department of Nutrition, Harvard T.H. ChanSchool of Public Health, Boston, Massachusetts

Corresponding author: Jordi Salas-Salvado,jordi.salas@urv.cat, or Miguel A. Martınez-Gonzalez, mamartinez@unav.es

Received 17 April 2018 and accepted 29 Sep-tember 2018

Clinical trial reg. no. ISRCTN89898870, www.isrctn.org

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc18-0836/-/DC1.

*Acomplete listof thePREDIMED-Plus investigatorscan be found in the Supplementary Data online.

This article is part of a special article collectionavailable at http://care.diabetesjournals.org/evolution-nutritional-therapy.

This article is featured in a podcast available athttp://www.diabetesjournals.org/content/diabetes-core-update-podcasts.

© 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.

778 Lifestyle Intervention Program and Weight Loss Diabetes Care Volume 42, May 2019

dysfunction, respectively (13). The PRE-DIMED-Plus randomized trial, startedin 2013, provides a unique opportunityto assess the long-term cardiovasculareffects of an intensive weight loss inter-vention based on an energy-restrictedMedDiet (erMedDiet), PA promotion,and behavioral support in comparisonwith a control group (CG). In this study,we report on a studywithin the PREDIMED-Plus trial on the 12-month effects of theintervention on weight loss, adiposity,and intermediate markers of cardiovas-cular risk in overweight/obese adultswith MetS at high cardiovascular risk.

RESEARCH DESIGN AND METHODS

Study DesignThe PREDIMED-Plus trial is a new 6-yearparallel-group, multicenter RCT involving6,874 participants recruited in 23 Spanishrecruiting centers. The trial’s main ob-jective is to evaluate the effect of anintensive weight loss intervention basedon an erMedDiet, PA promotion, andbehavioral support (IG) on hard cardio-vascular events in comparison with aCG receiving usual care, including therecommendation to follow an energy-unrestricted MedDiet without any adviceto increase PA. The primary end point is acomposite of CVD events (cardiovasculardeath, nonfatal myocardial infarction,and nonfatal stroke). The recruitmentperiod lasted from 5 September 2013 to30 November 2016. The intervention isscheduled to last for an average time of6 years, with a further 2 years of ex-tended follow-up for collection of clinicalevents. The PREDIMED-Plus protocol isavailable at http://predimedplus.com/.The institutional review boards of the23 participating centers approved thestudy protocol, and all participants pro-vided written informed consent.The current report is the first longitu-

dinal assessment of the PREDIMED-PlusRCT aimed to examine the 6- and 12-month effects of the intensive lifestyleintervention on body weight, adiposityparameters, and intermediate markers ofcardiovascular risk (systolic and diastolicblood pressure, glucose metabolism–

related variables, lipid profile, and pe-ripheral levels of leptin, C-peptide, andsome inflammatory markers) in compar-ison with usual care. This analysis wasperformed in 626 participants random-ized into the trial belonging to the first1,013 candidates assessed for eligibility.

These participants were recruited from13 out of the 23 PREDIMED-Plus re-cruiting centers because these centerswere the first that started the recruit-ment of participants. The data wereanalyzed using the available completePREDIMED-Plus database, dated 12 May2017.

Participant Selection and RecruitmentFrom September 2013 until September2014, medical doctors from primaryhealth care centers associated withhospital or university recruiting centersassessed potential participants for eligi-bility. Eligible participants were men(aged 55–75 years) and women (aged60–75 years), without documented his-tory of CVD (except heart failure NewYork Heart Association class I and II orvalvular heart disease) at enrollment,who were overweight/obese (BMI $27and ,40 kg/m2) and disclosed at leastthree components of the MetS accord-ing to the harmonized definition of thejoint statement from the InternationalDiabetes Federation/National Heart,Lung, and Blood Institute/American HeartAssociation (2009), the most updatedand largely recognized set of MetScriteria (18). Detailed inclusion and ex-clusion criteria are displayed in theSupplementary Data.

Potentially eligible candidates werecontacted by telephone or interviewedin person in a clinical visit. A screeninginterview with a PREDIMED-Plus inves-tigator was scheduled to inform in detailwilling candidates about the study andobtain signed informed consent. Priorto randomization, potential participantsentered a 4-week run-in period compris-ing three screening visits aimed at de-termining adherence to study procedures.Candidates were evaluated according toinclusion and exclusion criteria and re-ceived questionnaires assessing differentlifestyle and sociodemographic variablesto be returned at the last screening visit.In a second telephone-based visit, inves-tigators ensured completion of the ad-ministered questionnaires. Importantly,in January 2014, following advice of theData Safety and Monitoring Board, theSteering Committee decided to amendthe protocol and omit the prerandom-ization requirement to sustain at least a1.5-kg weight loss. Consequently, fromthis date onwards, recruits were ad-vised to maintain their usual eating

habits, PA level, and body weight. Suchprotocol change only affected the first70 participants who were eligible andrandomized in two vanguard centers(Supplementary Fig. 1). In the third screen-ing face-to-face visit, candidates meet-ing eligibility criteria who had attendedall screening visits and correctly filled inthe administered questionnaires and re-cords were randomized to either IG or CG.

Randomization and InterventionEach recruiting center randomly allo-cated candidates in a 1:1 ratio to eitherthe IG or the CG, using a centrally con-trolled, computer-generated random-number internet-based system withstratification by center, sex, and age(,65, 65–70, and .70 years). The ran-domization procedure was internet-based and blinded to all staff and tothe principal investigators of each re-cruitment center. Couples sharing thesame household were randomized to-gether, using the couple as unit of ran-domization. The name, sex, age, center,and individual/couple randomizationstatus of participants willing to partici-pate and fulfilling inclusion criteria weresubmitted by the field team of each siteto the internet-based system. This sys-tem applied the randomization algo-rithm and then returned to the site theautomatically generated group assignedto the participant. With this system, nochanges were possible regarding groupassignments after the submission ofthe demographic data of the participant.In the specific cases of couples in whichthe first spouse was previously recruitedat a different time, the last spouse en-tering the study was directly assigned(not randomized) to the same study armas his/her partner (n = 73).

A 17-item questionnaire aimed at as-sessing adherence to the erMedDiet wasdelivered at baseline, 6, and 12 months inboth study groups. The validated 14-itemPREDIMED questionnaire was also usedto assess adherence to the traditionalMedDiet only in the CG participants (19)(Supplementary Data). A validated 143-item food frequency questionnaire wasalso completed, together with the vali-dated Regicor Short Physical ActivityQuestionnaire (20) and the validatedSpanish version of the Nurses’ HealthStudy questionnaire to assess sedentarybehaviors (21). Physical fitness was eval-uated using the validated 30-s chair-stand

care.diabetesjournals.org Salas-Salvado and Associates 779

test (22). Additional information relatedto sociodemographic and lifestyle as-pects, education level, individual andfamily medical history, and current med-ication use was collected. Anthropomet-ric and blood pressure measurementswere obtained, and samples of fastingblood and urine were collected.

Intervention

Participants allocated to the IG followedan erMedDiet plus PA promotion andbehavioral support, with the purposeof accomplishing specific weight lossobjectives. The objectives in terms ofweight loss and the interventions aredetailed in the Supplementary Data.Participants in the CG received edu-

cational sessions on an ad libitum Med-Diet with the same content as those usedin the PREDIMED study (17). No specificadvice for increasing PA or losing weightwas provided to them (SupplementaryData).In addition to the individual sessions

(Supplementary Data), participants inboth groups received periodical groupsessions and telephone calls once permonth in the IG and every 6 monthsin the CG. Briefly, group sessions forboth study groups were conducted bythe dietitians and consisted of informa-tive talks addressing lifestyle-relatedtopics, in which free extra virgin oliveoil (1 L/month) and raw nuts (125 g/month)were provided in order to reinforce ad-herence to the protocol in both arms ofthe trial.

Outcomes and AssessmentsThe primary end point of the currentstudy was between-group differences inweight loss at 6 and 12 months of in-tervention, expressed in absolute values(kilograms) and percentage as well aschanges in BMI from baseline. Weight-related secondary end points at 6 and12 months were the between-groupdifferences in the proportions of partic-ipants who had either a stable weight orweight below baseline values, those wholost at least 5% or 10% of their initialweight, and those reversing obesity(changing BMI from$30 to,30 kg/m2).Other secondary outcomes were 6- and

12-month changes in waist circumference,body composition, systolic and diastolicblood pressure, fasting glucose, HbA1c,insulin sensitivity, and lipid levels. Meth-ods for anthropometry, body composi-tion, and blood pressure measurements

are described in the SupplementaryData.

After an overnight fast, blood sam-ples were collected at baseline, 6, and12 months. Tubes of serum and plasmawere collected, and aliquots were codedand stored at 280°C in a central labo-ratory until analyses. Serum glucose,triglycerides, total cholesterol, and HDLcholesterol levels were measured usingstandard enzymatic methods, and LDLcholesterol concentrations were calcu-lated with the Friedewald formula.

Other outcomes included 12-monthchanges in circulating levels of fastingserum insulin, leptin, C-peptide, hs-CRP,interleukin-6 (IL-6), IL-8, IL-18, tumornecrosis factor-a, MCP-1, and regulatedon activation, normal T-cell expressedand secreted cytokines. The methodsused for these determinations are shownin the Supplementary Data. Laboratorypersonnel performing all assays wereblinded to group allocation.

Statistical AnalysesAll analyses were performed using STATAsoftware, version 15.0 (StataCorp LP,College Station, TX). Based on previousstudies (6,15,23), and assuming that ourintervention has only a small effect onweight change, achieving a weight lossof 3 kg in the IG and 1 kg in the CG, with aSD of 8 and a correlation of 0.7 betweenfirst and second measurements, 80%power, and a = 0.05, we would needa sample size of 151 in each group.Because the number of participants in-cluded in the present analysis is muchhigher than this figure, the study had100% power to detect at least a 2-kgdifference between groups.

We used descriptive statistics withmean (6 SD) or percentages (numbers)for participants’ baseline characteris-tics. Data were analyzed by using theintention-to-treat principle and thecompleters-only framework. The smallnumber of missing outcome data (3%for weight) were handled via multivar-iate imputation with chained equations(STATA “mi” command), generating 20imputations for each missing measure-ment from regression equations to pre-dict these outcomes. The imputationmodels included as predictors all varia-bles in Table 1 and group allocation.Analyses of completers include onlyparticipants who had all measurements,without the inclusion of imputed data.

Continuous outcomes were assessed fornormality with the Shapiro-Wilk test,visual inspection of histograms, and scat-ter plots before each analysis.

Intervention effects on weight loss andBMI changes from baseline were evalu-ated using linear regression modelsbased on between-group differencesin mean changes. We used robust var-iance estimators to account for intra-cluster correlations in all regressionmodels, considering as clusters themembers of the same household. Pre-specified subgroup analyses of the pri-mary results were conducted within strataof sex, age, BMI, diabetes status, insulinuse, statin treatment, and educationallevel. The proportions of participantsin each group achieving the differentweight loss categorical outcomes at 6and 12 months were compared with thex2 test. The between-group differencesin changes in cardiovascular risk mark-ers, dietary variables, PA, sedentarybehaviors, and medication use werecompared using linear regression, x2

test, independent-samples t test, or me-dian regression analyses if data wereskewed to examine differences in me-dians (reporting median and interquar-tile range), as appropriate. Values areshown as mean and 95% CI, if not in-dicated otherwise. Significance for allstatistical tests was set at P , 0.05for bilateral contrast.

RESULTS

Between September 2013 and Septem-ber 2014, 1,013 candidates were ass-essed for eligibility. Of these, 143 refusedto participate, and 172 did not meetinclusion or randomization criteria. Thus,698 participants were randomly allo-cated into two intervention groupsof similar size, of whom 626 were in-cluded in the final analysis (n = 327,IG; and n = 299, CG). Approximately98% of participants in each groupcompleted 12months of follow-up, with-out significant differences in attritionbetween groups (Supplementary Fig. 1).

Participants’ Baseline CharacteristicsRandomized participants were compa-rable to those not randomized due toineligibility regarding bodyweight (mean85 kg), waist circumference (107 cm),and proportion of men (44%). However,compared with nonrandomized partici-pants, those randomized were slightly

780 Lifestyle Intervention Program and Weight Loss Diabetes Care Volume 42, May 2019

younger (by 1 year; P , 0.001). The IGand CG participants were well matchedat randomization and showed similarbaseline characteristics (Table 1).

Compliance With the Dietary andLifestyle InterventionsAt baseline, participants in the two groupsreported similar adherence to the Med-Diet and similar intakes of food groups,energy, andnutrients. At 6 and12months,participants in the IG reported a signif-icantly greater achievement in 10 of the17 items of the questionnaire of adher-ence to the erMedDiet, with a net in-crease of 2 points versus CG at 12months (P , 0.001) (SupplementaryTables 1 and 2). Although the generalfood pattern improved in both IG and

CG, the consumption of some key Med-iterranean foods improved significantlymore in IG than CG participants (Supple-mentary Table 3). At 12 months, reductionsin daily energy intake were more pro-nounced in the IG than in CG (P = 0.05),and participants in the IG reporteda lower intake of carbohydrates andhigher intakes of protein, total fat, andmonounsaturated fat (SupplementaryTable 4).

Total leisure time PA increased signif-icantly from baseline in the IG whiledecreasing in the CG (P = 0.001). A sig-nificant reduction from baseline of totalsedentary time was observed in the IG,without significant differences versusCG (Supplementary Table 2). Mean timespent in television viewing was reduced

significantly more among participants inthe IG than in the CG. At baseline, theproportions of participants meeting theWorld Health Organization’s 2010 rec-ommendations of at least 150 min ofmoderate-vigorous PA/week were 52.6%in the IG and 51.5% in the CG. At 12months, this proportion increased by11.5% in the IG, whereas it decreasedby 0.7% in the CG (P = 0.001). Changesin the 30-s chair-stand test were minorand did not differ between groups. Ofnote, during the 12 months of the study,participants in the IG attended 75% and67% of the individual and group sessions,respectively, whereas respective ratesfor those in the CG were 95% and 78%.

Weight Loss and MaintenanceBecause intention-to-treat (using multi-ple imputation methods) and com-pleters-only analyses showed similarresults, only intention-to-treat resultsare reported. Weight losses betweenthe two groups differed significantlyover time (Table 2): at 6 months, themean weight losses from baselineamong participants assigned to inter-vention and control were 22.4 (22.7%)and 20.4 kg (20.5%), respectively,whereas respective values at 12 monthswere23.2 (23.7%)and20.7 kg (20.8%).The maximum between-group differ-ence in weight loss was reached at12 months, with a mean difference inweight changes between the IG and CGof22.5 kg (23.1 to21.9). Results wereconsistent among subgroups of sex,age, BMI, diabetes, insulin use, statintreatment, and educational level (Sup-plementary Table 5). Reductions inBMI from baseline to 6 and 12 monthswere greater in the IG (Table 2). Resultswere also qualitatively similar afteradjusting by baseline values.

More participants in the IG than inthe CG lost weight below their initialweight at 6 months (81.3% vs. 58.4%;P , 0.001) and 12 months (84.1% vs.57.9%; P , 0.001) (Table 3). In allweight loss thresholds (losses of $5%,or $10%, and change of baselineBMI $30 to ,30 kg/m2) at 6 and12 months, the proportions of partic-ipants attaining these targets weresignificantly higher in the IG than inthe CG. At 12 months, 33.7 and 6.9% ofparticipants in the IG achieved a weightloss of $5% and of at least 10%, re-spectively.

Table 1—Characteristics of participants at randomization

Characteristic All (n = 626)Intervention group

(n = 327)Control group(n = 299)

Age (years) 65 6 5 66 6 5 65 6 5

Male 46 (289) 45 (148) 47 (141)

Same household (couples) 12 (73) 12 (38) 12 (35)

Weight (kg) 86.3 6 12.9 85.8 6 13.1 86.9 6 12.7

BMI (kg/m2) 32.5 6 3.5 32.3 6 3.4 32.6 6 3.6

Waist circumference (cm) 106.8 6 9.3 106.3 6 8.9 107.3 6 9.6

Obese (BMI $30 kg/m2) 73 (459) 73 (240) 73 (219)

Prediabetes* 40 (248) 39 (129) 40 (119)

Type 2 diabetes† 45 (281) 44 (144) 46 (137)

Hypertension 88 (551) 87 (285) 89 (266)

Dyslipidemia 74 (461) 72 (234) 76 (227)

Number of MetS components#3 components 57 (354) 57 (185) 57 (169)4 components 28 (175) 27 (90) 28 (85)5 components 15 (97) 16 (52) 15 (45)

Current smokers 11 (68) 11 (36) 11 (32)

Former smokers 41 (260) 39 (126) 45 (134)

MedicationsLipid-lowering drugs 55 (342) 52 (171) 57 (171)Statin use 48 (298) 46 (150) 49 (148)

Antihypertensive therapy 78 (488) 76 (250) 80 (238)Thiazide drugs‡ 26 (163) 26 (86) 26 (77)ACEi/ARB use 61 (381) 60 (195) 62 (186)

Oral antidiabetic medications§ 36 (227) 36 (117) 37 (110)Insulin treatment 7 (44) 6 (21) 8 (23)

Educational levelPrimary school 52 (324) 50 (165) 53 (159)First-degree high school 28 (173) 27 (89) 28 (84)High school or university 21 (129) 22 (73) 19 (56)

Data aremean6SDorpercentage (number). ACEi, ACE inhibitor;ARB, angiotensin type2 receptorblocker. *Prediabetes was defined as fasting plasma glucose of 100–125 mg/dL (5.6–6.9 mmol/L)or glycated hemoglobin (HbA1c) of 5.7–6.4% (39–47 mmol/mol). †Diabetes was defined asprevious diagnosis of diabetes or HbA1c $6.5% (48 mmol/mol), use of antidiabetic medicationor having fasting glucose .126 mg/dL (7.0 mmol/L) in the screening visit plus fastingglucose.126mg/dL (7.0mmol/L) at baseline visit. ‡Thiazide drugs include thiazides and thiazide-like diuretics. §Oral antidiabetic medications include treatment with meglitinides, sulfonylureas,biguanides, thiazolidinediones, dipeptidyl peptidase IV inhibitors, and a-glucosidase inhibitors,each used as a single oral agent or combined with one or more oral antidiabetic drugs.

care.diabetesjournals.org Salas-Salvado and Associates 781

Compared with the CG, the IG expe-rienced greater decreases in total bodyfat. Significant reductions in total leanmass were observed only in the IG,whereas no changes occurred in theCG. Compared with the CG, the IGshowed greater improvement in totallean mass/total body fat ratio (Supple-mentary Table 6).

Waist Circumference, Blood Pressure,Glucose Metabolism–RelatedParameters, and Lipid ProfileFor several risk factors, between-groupdifferences were most apparent at 12months, when a maximum difference inweight loss was reached (Table 4). Waistcircumference showed a greater reduc-tion in the IG than in the CG both at 6and 12 months. Significant reductionsin all glucose metabolism–related pa-rameters (fasting glucose, HbA1c, insulin,and HOMA of insulin resistance [HOMA-IR] index) were observed in the IG after12 months, whereas no changes oc-curred in the CG.

The IG experienced greater improve-ments in HDL cholesterol and triglycer-ide levels. Systolic and diastolic bloodpressure and total and LDL cholesteroldecreased in the two groups, withoutbetween-group differences. For all anal-yses, results were also qualitatively sim-ilar after adjusting by baseline values. Wehave further conducted subgroup anal-yses stratified according to diabetes sta-tus (normal glycemia, prediabetes, anddiabetes) to assess the changes in glu-cose metabolism–related parametersand lipid profile (Supplementary Tables7–9). Among participants with diabetes(n = 281) and prediabetes (n = 181), theintervention obtained improvements inglycemic control and insulin sensitivity,along with amelioration in triglycerideand HDL cholesterol levels significantlyhigher than the CG (P , 0.05) (Sup-plementary Tables 7 and 8).

On-trial changes in medications tocontrol blood glucose, dyslipidemia, orhypertension were similarly distributedbetween the two groups (SupplementaryTable 9).

Leptin, Pancreatic Insulin Secretion,and Inflammatory MarkersNovel cardiovascular risk biomarkers weremeasured only at baseline and 12 months.Circulating leptin had a greater reduc-tion in the IG than in the CG. Levels of IL-18

Table

2—Meanweightloss

at6and12

monthsofinterven

tion,exp

ressedasabso

lute

bodyweightloss

(kilograms),percen

treductionrelative

tobase

line,andabso

lute

changein

BMIbytreatm

entgro

up:intention-to-treat(m

ultiple

imputation)andco

mpleters-o

nly

Variable

Interven

tiongroup

Controlgroup

Interven

tionvs.control

Between-groupdifference

Intention-to-treat

(MI)(n

=32

7)Completers-only

(n=30

2)Intention-to-treat

(MI)(n

=29

9)Completers-only

(n=28

2)Intention-to-treat

(MI)

Pvalue

Completers-only

Pvalue

Change

inbodyweight(kg)

Atmonth

622.4(2

2.8to

22.0)

22.4(2

2.8to

22.0)

20.4(2

0.8to

20.1)

20.4(2

0.8to

20.1)

21.9(2

2.4to

21.4)

,0.001

22.0(2

2.5to

21.4)

,0.001

Atmonth

1223.2(2

3.7to

22.8)

23.3(2

3.8to

22.8)

20.7(2

1.1to

20.3)

20.7(2

1.2to

20.3)

22.5(2

3.2to

21.9)

,0.001

22.6(2

3.2to

21.9)

,0.001

Change

inbodyweight(%

)Atmonth

622.7(2

3.2to

22.3)

22.7(2

3.2to

22.3)

20.5(2

0.9to

20.1)

20.5(2

0.9to

20.1)

22.2(2

2.8to

21.6)

,0.001

22.2(2

2.8to

21.6)

,0.001

Atmonth

1223.7(2

4.3to

23.2)

23.8(2

4.3to

23.2)

20.8(2

1.3to

20.3)

20.8(2

1.3to

20.3)

23.0(2

3.6to

22.3)

,0.001

23.0(2

3.7to

22.3)

,0.001

Change

inBMI(kg/m

2)

Atmonth

620.9(2

1.0to

20.8)

20.9(2

1.0to

20.8)

20.2(2

0.3to

20.1)

20.2(2

0.3to

20.1)

20.7(2

0.9to

20.5)

,0.001

20.7(2

0.9to

20.5)

,0.001

Atmonth

1221.2(2

1.4to

21.0)

21.2(2

1.4to

21.1)

20.3(2

0.4to

20.1)

20.3(2

0.4to

20.1)

21.0(2

1.2to

20.7)

,0.001

21.0(2

1.2to

20.7)

,0.001

Dataaremean(95%

CI).Pvalues

forbetween-groupdifferenceswerecalculatedusinglinearregressionmodelswithrobust

SEto

accountforintracluster

correlations.MI,multiple

imputation.

782 Lifestyle Intervention Program and Weight Loss Diabetes Care Volume 42, May 2019

and MCP-1 also decreased significantlyversus control. Circulating C-peptide de-creased significantly only in the IG, butwithout significant differences versuscontrol. Changes in other inflammatorymarkers, such as hs-CRP, IL-6, IL-8, tu-mor necrosis factor-a, and regulated onactivation, normal T-cell expressed andsecreted, were minor and not statisti-cally different between groups (Sup-plementary Table 10).

CONCLUSIONS

In this study conducted in the first par-ticipants recruited into the PREDIMED-Plus trial who had a 12-month follow-up,we show that the intensive lifestyle in-tervention was effective in producing aclinically meaningful weight loss amongoverweight/obese adults with MetS. Asexpected, participants allocated to theIG lost more weight and also showedgreater improvements in some cardio-vascular risk factors at 6 and 12 monthsthan CG participants. Weight loss wasmaximal after 12months,when improve-ments in risk factors were more evident,suggesting a strong relationship betweenthe magnitude of weight loss achievedand changes in cardiovascular risk.

In most long-term weight loss studiesusing lifestyle (diet or/and exercise) in-terventions or drugs, maximum weightloss was typically achieved at 6 monthsof intervention, and thereafter a plateauor, more frequently, weight regain oc-curred (5,9,24,25). Notwithstanding, inour study, maximum weight loss wasachieved at 12 monthsdfor the entirestudy population and by subgroups ofsex, age, BMI, diabetes status, insulin andstatin treatment, and educational leveldwithout any evidence of weight regain,thereby highlighting the sustainedefficacy of our intensive lifestyle inter-vention based on an erMedDiet, PA, andbehavioral enforcement. A maximumincrease in adherence to the erMedDietscore was evident after 12 months inboth groups, with a net increase in favorof the IG, as expected. In addition, theincrease in the percentage of individualsachieving the goals for each of the 17score items was apparent in 11 of the17 items at 6 months, but also at 12months, reflecting the large potential forlong-term sustainability of the interven-tion in PREDIMED-Plus. These changescan be explained by a higher and

Table

3—Pro

portio

nofparticip

ants

(%)whometdiffe

rentweightloss

criteria

at6and12

monthsofinterve

ntio

nbytre

atm

entgro

up:intentio

n-to

-treat(m

ultip

leim

putatio

n)

andco

mpleters-o

nly

Criteria

Proportio

nofparticip

ants

(%)

Interven

tiongro

up

Contro

lgro

up

Interven

tionvs.

contro

l

Betw

een-grou

pdifferen

ce

Inten

tion-to

-treat(M

I)(n

=327)

Completers-o

nly

(n=302)

Inten

tion-to

-treat(M

I)(n

=299)

Completers-

only

(n=282)

Inten

tion-to

-treat(M

I)Pvalu

eCompleters-o

nly

Pvalu

e

Atorbelo

wbaselin

eweigh

tAtmonth

681.3

(76.9–85.6)

81.4(76.6

–85.5)58.4

(52.6–64.2)

58.5(52.6

–64.1)22.9

(15.5–30.1)

,0.001

22.7(15.3

–30.0),0.001

Atmonth

1284.1

(80.0–88.2)

84.4(80.3

–88.5)57.9

(52.2–63.7)

58.2(52.4

–63.9)26.3

(19.1–33.5)

,0.001

26.0(18.7

–33.1),0.001

Atleast

5%belo

wbaselin

eweigh

tAtmonth

620.7

(16.2–25.3)

20.8(16.3

–25.4)6.5

(3.6–9.3)

6.4(3.5

–9.2)14.2

(8.6–19.7)

,0.001

14.5(8.9

–20.0),0.001

Atmonth

1233.7

(28.4–39.1)

34.1(28.7

–39.5)11.9

(8.1–15.7)

11.7(7.9

–15.5)21.8

(15.2–28.3)

,0.001

22.4(15.7

–29.0),0.001

Atleast

10%belo

wbaselin

eweigh

tAtmonth

64.2

(1.8–6.5)

3.6(1.5

–5.8)0.9

(20.3

to2.2)

0.7(2

0.2to

1.6)3.3

(0.5–5.9)

0.022.9

(0.5–5.3)

0.02Atmonth

126.9

(4.1–9.8)

6.6(3.8

–9.4)2.2

(0.4–3.9)

2.1(0.4

–3.8)4.7

(1.3–8.1)

0.0074.5

(1.2–7.8)

0.008

Change

from

baselin

eBMI$30

toBMI,30

kg/m2

Atmonth

611.3

(7.8–14.8)

10.9(7.3

–14.5)6.3

(3.4–9.1)

6.0(3.2

–8.8)5.0

(0.4–9.6)

0.034.9

(0.3–9.4)

0.03Atmonth

1215.7

(11.6–19.7)

15.6(11.4

–19.7)6.7

(3.7–9.6)

6.4(3.5

–9.2)9.0

(3.9–14.1)

0.0019.2

(4.1–14.2)

,0.001

Data

arepercen

tage(95%

CI).

Pvalu

esforbetw

een-gro

upsdifferen

ceswere

calculated

usin

glinear

regressionmodels

with

robust

SEto

accountforintraclu

stercorrelatio

ns.MI,multip

leim

putation

.

care.diabetesjournals.org Salas-Salvado and Associates 783

Table

4—Base

lineand6-and12

-month

changesin

adiposity,bloodpressure,andca

rdiova

scularrisk

factors

bytreatm

entgro

up:intention-to-treat(m

ultiple

imputation)

andco

mpleters-o

nly

Variable

Interven

tiongroup

Controlgroup

Interven

tionvs.control

Between-groupdifference

Intention-to-treat

(MI)

Completers-only

Intention-to-treat

(MI)

Completers-only

Intention-to-treat

(MI)

Pvalue

Completers-only

Pvalue

Waist

circumference

(cm)

n=32

7n=29

2n=29

9n=27

5Baseline

106.3(105

.3–10

7.2)

106.3(105

.2–10

7.3)

107.3(106.2–10

8.4)

107.3(106

.1–10

8.4)

6-month

change

23.0(2

3.6to

22.5)

23.1(2

3.7to

22.5)

20.9(2

1.5to

20.3)

20.9(2

1.4to

20.3)

22.1(2

2.9to

21.3)

,0.001

22.2(2

3.1to

21.4)

,0.001

12-m

onth

change

23.1(2

3.8to

22.5)

23.3(2

3.8to

22.8)

20.7(2

1.3to

0.03

)20.7(2

1.4to

20.2)

22.5(2

3.4to

21.5)

,0.001

22.5(2

3.5to

21.6)

,0.001

Systolic

BP(m

mHg)

n=32

7n=29

1n=29

9n=27

0Baseline

139.0(137

.0–14

0.7)

139.1(137.2–14

0)13

8.1(136.2–14

0.0)

138.1(136

.1–14

0.1)

6-month

change

20.8(2

2.6to

1.0)

20.7(2

2.5to

1.1)

0.9(2

1.0to

2.8)

0.8(2

1.0to

2.7)

21.7(2

4.3to

0.9)

0.19

21.6(2

4.2to

0.9)

0.21

12-m

onth

change

23.5(2

5.4to

21.6)

23.5(2

5.4to

21.5)

22.0(2

4.1to

0.1)

21.9(2

4.0to

0.2)

21.5(2

4.4to

1.4)

0.24

21.6(2

4.4to

1.3)

0.28

Diastolic

BP(m

mHg)

n=32

7n=29

1n=29

9n=27

0Baseline

79.9

(78.8–81

.1)

79.8

(78.6–81

)79

.0(77.7–80

.0)

79.0

(77.8–80

.2)

6-month

change

21.8(2.8

to20.8)

21.7(2.7

to20.7)

20.4(2

1.4to

0.6)

20.3(2

1.3to

0.7)

21.4(2

2.8to

20.04

)0.05

21.4(2

2.8to

0.05)

0.06

12-m

onth

change

22.1(2

3.0to

21.2)

22.2(2

3.1to

21.3)

21.4(2

2.4to

20.3)

21.3(2

2.4to

20.2)

20.7(2

2.1to

0.7)

0.34

20.9(2

2.3to

0.5)

0.22

Glucose

(mmol/L)

n=32

7n=29

4n=29

9n=26

1Baseline

6.53

(6.34–6.72)

6.52

(6.331–6.72)

6.50

(6.30–6.70

)6.45

(6.25–6.65)

6-month

change

20.13

(20.24

to20.01

)20.13

(20.24

to20.01

)0.08

(20.05

to0.20)

0.07

(20.05

to0.21)

20.21

(20.37

to20.03)

0.02

20.20

(20.38

to20.03

)0.02

12-m

onth

change

20.23

(20.36

to20.09

)20.22

(20.35

to20.08

)0.12

(20.04

to0.29)

0.16

(20.01

to0.33)

20.35

(20.56

to20.13)

0.00

220.39

(20.59

to20.16

)0.00

1

HbA1c(%

)*n=32

7n=19

3n=29

9n=17

4Baseline

6.0(5.6–6.5)

6.0(5.7–6.5)

6.0(5.6–6.6)

6.0(5.7–6.6)

6-month

change

20.10

(20.33

to0.10)

20.10

(20.30

to0.10)

0.0(2

0.21

to0.26)

0.0(2

0.20

to0.10)

20.10

(20.17

to20.03)

0.00

620.10

(20.18

to20.02

)0.01

12-m

onth

change

20.12

(20.42

to0.14)

20.10

(20.40

to0.10)

0.0(2

0.23

to0.22)

0.0(2

0.20

to0.20)

20.12

(20.21

to20.02)

0.01

20.10

(20.17

to20.02

)0.01

HbA1c(m

mol/mol)*

n=32

7n=19

3n=29

9n=17

4Baseline

42.07(37.7–48

.26)

42.07(38.79–47

.54)

42.07(38.0–48

.70)

42.07(38.79–48

.73)

6-month

change

21.08

(23.46

to1.18)

21.09

(23.27

to1.09)

0.04

(22.19

to3.12)

0.0(2

2.18

to1.09)

21.11

(21.91

to20.32)

0.00

621.09

(21.92

to20.26

)0.01

12-m

onth

change

21.25

(24.61

to1.54)

21.09

(24.37

to1.09)

0.02

(22.39

to2.41)

0.0(2

2.18

to2.19)

21.27

(22.29

to20.25)

0.01

21.09

(21.92

to20.26

)0.01

Insulin

(pmol/L)†

n=29

2n=23

7n=26

7n=21

2Baseline

132.8(124

.6–14

1.1)

132.9(123

.5–14

2.4)

133.2(123.1–14

3.2)

132.4(122

.8–14

2.1)

6-month

change

dd

dd

dd

dd

12-m

onth

change

226

.2(2

34.3

to218

.1)

227

.0(2

34.7

to219

.2)

27.4(2

15.8

to1.1)

26.9(2

15.3

to1.5)

218

.8(2

30.7

to26.9)

0.00

2220

.1(2

31.5

to28.7)

0.00

1

HOMA-IRindex‡

n=29

2n=23

7n=26

7n=21

2Baseline

5.20

(4.8–5.58)

5.19

(4.83–5.57)

5.12

(4.73–5.51

)5.09

(4.72–5.47)

6-month

change

dd

dd

dd

dd

12-m

onth

change

21.16

(21.51

to20.81

)21.12

(21.45

to20.79

)20.07

(20.48

to0.33

)20.11

(20.47

to0.26)

21.09

(21.60

to20.58)

,0.001

21.02

(21.52

to20.52

),0.001

HOMA-IRindex§

n=18

3n=15

0n=16

2n=12

6Baseline

4.57

(4.14–4.99)

4.50

(4.13–4.88)

4.79

(4.33–5.25

)4.85

(4.39–5.31)

6-month

change

dd

dd

dd

dd

12-m

onth

change

21.03

(21.42

to20.64

)20.93

(21.27

to20.58

)20.18

(20.65

to0.29

)20.29

(20.67

to0.10)

20.85

(21.45

to20.26)

0.00

520.64

(21.16

to20.13

)0.01

Con

tinu

edon

p.78

5

784 Lifestyle Intervention Program and Weight Loss Diabetes Care Volume 42, May 2019

Table

4—Continued

Variable

Interven

tiongroup

Controlgroup

Interven

tionvs.control

Between-groupdifference

Intention-to-treat

(MI)

Completers-only

Intention-to-treat

(MI)

Completers-only

Intention-to-treat

(MI)

Pvalue

Completers-only

Pvalue

Totalcholesterol

(mmol/L)

n=32

7n=29

2n=29

9n=25

7Baseline

5.14

(5.03–5.25)

5.09

(4.98–5.20)

5.12

(5.01–5.24

)5.19

(5.07–5.31)

6-month

change

20.05

(20.14

to0.04)

20.03

(20.12

to0.06)

20.07

(20.15

to0.01

)20.07

(20.16

to20.01

)0.02

(20.10

to0.13)

0.76

0.04

(20.08

to0.16)

0.49

12-m

onth

change

20.13

(20.22

to20.04

)20.12

(20.21

to20.03

)20.15

(20.24

to20.06)

20.16

(20.26

to20.07

)0.02

(20.11

to0.15)

0.72

0.04

(20.08

to0.17)

0.48

HDLcholesterol

(mmol/L)

n=32

7n=29

0n=29

9n=25

4Baseline

1.26

(1.22–1.29)

1.26

(1.23–1.30)

1.28

(1.24–1.31

)1.27

(1.23–1.30)

6-month

change

0.07

(0.05–0.09)

0.07

(0.05–0.09)

0.04

(0.02–0.06

)0.04

(0.02–0.06)

0.03

(0.01–0.06)

0.02

0.03

(0.01–0.06)

0.02

12-m

onth

change

0.06

(0.04–0.08)

0.06

(0.04–0.08)

0.0(2

0.02

to0.02)

0.01

(20.01

to0.02)

0.06

(0.03–0.09)

,0.001

0.06

(0.03–0.09)

,0.001

LDLcholesterol

(mmol/L)

n=32

7n=28

9n=29

9n=25

3Baseline

3.13

(3.04–3.23)

3.08

(2.98–3.18)

3.08

(2.99–3.18

)3.14

(3.04–3.24)

6-month

change

20.06

(20.14

to0.02)

20.04

(20.12

to0.04)

20.08

(20.15

to0.0)

20.08

(20.16

to20.01

)0.02

(20.09

to0.12)

0.76

0.04

(20.07

to0.14)

0.48

12-m

onth

change

20.12

(20.19

to20.04

)20.11

(20.19

to20.02

)200.16

(20.24

to20.07

)20.15

(20.24

to20.07

)0.04

(20.08

to0.15)

0.48

0.05

(20.07

to0.16)

0.42

Totalcholesterol/HDL

cholesterolratio

n=32

7n=29

0n=29

9n=25

3Baseline

4.27

(4.15–4.39)

4.21

(4.08–4.33)

4.17

(4.05–4.28

)4.23

(4.11–4.35)

6-month

change

20.26

(20.34

to20.18

)20.25

(20.33

to20.16

)0.19

(20.26

to20.11

)0.18

(20.26

to20.11)

20.07

(20.18

to0.04)

0.19

20.06

(20.17

to0.05)

0.28

12month

change

20.30

(20.38

to20.21

)20.29

(20.37

to20.20

)20.14

(20.23

to20.05)

20.15

(20.23

to20.06

)20.16

(20.28

to20.03)

0.01

20.14

(20.26

to20.02

)0.02

Triglycerides

(mmol/L)

n=32

7n=29

1n=29

9n=25

9Baseline

1.64

(1.57–1.72)

1.62

(1.54–1.69)

1.67

(1.59–1.74

)1.69

(1.60–1.77)

6-month

change

20.13

(20.19

to20.07

)20.14

(20.20

to20.07

)20.06

(20.13

to0.01

)20.05

(20.13

to0.02)

20.07

(20.16

to0.02)

0.13

20.08

(20.18

to0.01)

0.09

12-m

onth

change

20.17

(20.23

to20.10

)20.16

(20.22

to20.10

)0.0(2

0.08

to0.09)

0.0(2

0.08

to0.09)

20.17

(20.27

to20.06)

0.00

220.16

(20.27

to20.06

)0.00

3

Triglyceride/HDL

cholesterolratio

n=32

7n=28

9n=29

9n=25

4Baseline

1.43

(1.34–1.53)

1.41

(1.32–1.50)

1.45

(1.35–1.56

)1.45

(1.35–1.56)

6-month

change

20.16

(20.23

to20.10

)20.17

(20.24

to20.10

)20.10

(20.18

to20.03)

20.10

(20.17

to20.02

)20.06

(20.16

to0.04)

0.22

20.07

(20.18

to0.03)

0.19

12-m

onth

change

20.20

(20.27

to20.12

)20.19

(20.26

to20.12

)0.0(2

0.09

to0.09)

20.02

(20.11

to0.07)

20.19

(20.31

to20.08)

0.00

120.17

(20.29

to20.05

)0.00

4

Dataaremean

(95%

CI)unless

otherw

iseindicated.BP,bloodpressure;MI,multiple

imputation.*Fo

rHbA1c,dataaremedian(interquartile

range

),an

dthebetw

een-groupdifference

ismedian(95%

CI).†,‡Determ

inedonly

inparticipan

tswithoutinsulintreatment.§D

eterm

inedonly

inparticipan

tswithoutdiabetes.

Pvaluesforbetw

een-groupsdifferenceswere

calculated

usinglinear

regressionwithrobust

SEto

acco

untforintraclusterco

rrelationsormedianregressionan

alysesifdatawere

skewed.

care.diabetesjournals.org Salas-Salvado and Associates 785

sustained increase in the consumption ofnuts, whole-grain cereals, and fish alongwith a higher reduction in the consump-tion of refined cereals, red meat, pastries,cakes, and sweets in participants assignedto the IG compared with those in the CG.These data demonstrate the effect of theintervention increasing the consumptionof foods typical of the MedDiet and de-creasing the consumption of those char-acteristic of an unhealthy dietary pattern(e.g., red and processed meat, soft drinks,and refined foods).Our study also showed a sustained

effect of the PA intervention. Comparedwith control, a significant and higher de-crease in time spent in television view-ing and a higher increase in leisure-timePA was observed in the IG, and thesechanges were most apparent at 12months. In fact, the proportion of par-ticipants meeting the World HealthOrganization’s 2010 recommendationsof at least 150 min of moderate-vigorousPA/week was maximal at 12 months andhigher in the IG than CG.Although we observed a weight loss

effect maintained over time, only a me-dian 3.7% decrease in body weight wasachieved in the IG compared with the 8%decrease aimed at. In addition, only a 3%between-group difference in weight losswas observed compared with the objec-tive of 5%. The modest effect of thePREDIMED-Plus intervention on bodyweight and waist circumference in re-lation to our a priori objectives can beexplained in part because our populationwas aged, had a low educational level(;50% only have primary-school level),and 45% of participants had diabetes atbaseline. All of these factors have beenrecognized as predictors of suboptimaladherence to and efficacy of the interven-tion in weight loss trials (26). A consider-able strength of our study is that therewere very few dropouts in comparisonwith previous weight loss trials. In addi-tion, attendance to the scheduled individ-ual and group sessions during follow-upwas also higher than in previous weightloss trials. High retention into the trialprovides sound evidence of the long-termpalatability andsustainabilityof anerMed-Diet. In our study, the IG achieved weightloss especially at expenses of total bodyfat. Participants in the IG showed improve-ments in the leanbodymass/total body fatratio, which is typically considered as amore favorable body composition, hereby

suggesting that lifestyle interventions forweight loss should include the use of leanbody mass–preserving strategies (e.g.,PA) in order to prevent or delay sarco-penia.

Many RCTs have reported the bene-ficial effects of losing 5–10% body weightirrespective of intervention on cardio-vascular risk factors associated withoverweight/obesity, including abdominaladiposity, blood pressure, triglycerides,HDL cholesterol, and insulin resistance(3,4). Moderate weight loss in our studyhad the same beneficial effects in theseparameters except for blood pressure.These findings suggest that an intensiveweight loss lifestyle intervention basedon an erMedDiet and increased PA is asafe strategy for treating the MetS andameliorate some associated cardiovas-cular risk factors. Weight-losing low-carbohydrate diets are usually rich insaturated fatty acids and cholesterol andthereby have the unwanted effect ofincreasing LDL cholesterol (12,27), a po-tent and recognized risk factor for ath-erosclerotic CVD. However, as in otherstudies using low-fat (7) or MedDiets(15,28), in the current study, no delete-rious effects of losing weight on LDLcholesterol were observed. A relevantbeneficial effect of the intervention in ourstudy occurred on fasting glucose, insulinlevels, andHbA1c, reinforcing the hypoth-esis that an erMedDiet intervention haslong-term beneficial effects on insulinresistance and glucose control in over-weight and obese elderly people at highrisk for cardiovascular events. In thisregard, it is worth noting that for indi-viduals with both prediabetes and di-abetes, our intervention, althoughexerting a modest weight loss, impor-tantly resulted in improvements in gly-cemic control and insulin sensitivity,along with improvements in triglycerideand HDL cholesterol levels. Our findingssupport prior results from RCTs (5–7,9),confirming the potential of weight loss–based intensive lifestyle interventions inthe management of diabetes as well asin the delay or prevention of diabetesin those people at risk.

Yet, intervention in our study resultedin little changes of medications to controlblood glucose, dyslipidemia, or hyper-tension. Possibly greater and sustainedweight loss over time is necessary toreduce use of medications against obe-sity-associated risk factors.

Our results also demonstrate benefi-cial effects of the intensive interventionon some circulating proinflammatoryparameters related to obesity, such asleptin and IL-18. Leptin is a proinflam-matory protein produced mainly by adi-pocytes, which reflects the degree ofadiposity and insulin resistance and isimplicated in the pathogenesis of severalof its major complications (29). IL-18 isanother proinflammatory cytokine pro-duced by macrophages and other cellsthat has been shown to increase inobesity (30) and decrease after weightloss (31). Reduction of leptin and IL-18 inthe intervention arm can be explained inpart by weight loss, but we cannot dis-count effects derived from a higher ad-herence to the MedDiet and increasedPA. The intervention was also associ-ated with reduced concentrations ofMCP-1, one of the key chemokines reg-ulating inflammation via migration andinfiltration of monocytes/macrophages(32). This protein is increased in obesityand diabetes and reduced by weight loss(31). Given that high levels of MCP-1 andIL-18 are also implicated in the develop-ment of atherosclerosis (31,33), theirsustained reduction over time might beassociated with decreased cardiovascu-lar events.

Our findings are limited to adults withhigh BMI who also meet the criteria forMetS and were living in aMediterraneancountry. Therefore, they cannot be gen-eralized to other populations or to allindividuals with MetS.

In conclusion, we have shown in over-weight/obese adults with MetS that anintensive lifestyle intervention using anerMedDiet, PA promotion, and behav-ioral support resulted in clinically mean-ingful weight loss, high adherence torecommendations, and improvementsin MetS components and other interme-diate markers of cardiovascular riskafter a 6-month follow-up, with thesebeneficial effects being enhanced afterintervention for 12 months. Addition-ally, such lifestyle intervention causedmodest, yet potentially important, im-provements in glycemic control, insulinsensitivity, and dyslipidemia in individu-als with or at risk for diabetes. Based onthese results and past research on thecardiovascular effects of the MedDiet(13,17), we hypothesize that long-termweight loss maintenance in response tothe PREDIMED-Plus lifestyle program

786 Lifestyle Intervention Program and Weight Loss Diabetes Care Volume 42, May 2019

might provide the same or even greaterbenefit in terms of hard cardiovascularevents.

Acknowledgments. The authors especiallythank the PREDIMED-Plus participants for theenthusiastic collaboration, the PREDIMED-Pluspersonnel for outstanding support, and thepersonnel of all associated primary care centersfor the exceptional effort. CIBEROBN, CIBERESP,and CIBERDEM are initiatives of Instituto deSalud Carlos III (ISCIII), Madrid, Spain. The au-thors also thank the PREDIMED-Plus BiobankNetwork as a part of the National BiobankPlatform of the ISCIII for storing and managingthe PREDIMED-Plus biological samples.Funding. The PREDIMED-Plus trial was sup-ported by the official funding agency for bio-medical research of the Spanish government,ISCIII, through the Fondo de Investigacion para laSalud (FIS), which is co-funded by the EuropeanRegional Development Fund (three coordinatedFIS projects led by J.S.-S. and J.Vid., including thefollowing projects: PI13/00673, PI13/00492,PI13/00272, PI13/01123, PI13/00462, PI13/00233, PI13/02184, PI13/00728, PI13/01090,PI13/01056, PI14/01722, PI14/00636, PI14/00618, PI14/00696, PI14/01206, PI14/01919,PI14/00853, PI14/01374, PI16/00473, PI16/00662, PI16/01873, PI16/01094, PI16/00501,PI16/00533, PI16/00381, PI16/00366, PI16/01522, PI16/01120, PI17/00764, PI17/01183,PI17/00855, PI17/01347, PI17/00525, PI17/01827, PI17/00532, PI17/00215, PI17/01441,PI17/00508, PI17/01732, and PI17/00926),the Especial Action Project entitled: Im-plementacion y evaluacion de una inter-vencion intensiva sobre la actividad f ısicaCohorte PREDIMED-PLUS grant to J.S.-S., theEuropean Research Council (Advanced ResearchGrant 2013–2018, 340918) to M.A.M.-G., theRecercaixa grant to J.S.-S. (2013ACUP00194),grants from the Consejerıa de Salud de laJunta de Andalucıa (PI0458/2013, PS0358/2016,and PI0137/2018), a grant from the GeneralitatValenciana (PROMETEO/2017/017), a SEMERGENgrant, and funds from the European RegionalDevelopment Fund (CB06/03). O.C. is sup-ported by ISCIII grant JR17/00022. Food com-panies Hojiblanca (Lucena, Spain) and PatrimonioComunal Olivarero (Madrid, Spain) donatedextra virgin olive oil, and the Almond Boardof California (Modesto, CA), American PistachioGrowers (Fresno, CA), and Paramount Farms(Wonderful Company, LLC, Los Angeles, CA)donated nuts.None of the funding sources took part in the

design, collection, analysis, or interpretation ofthe data or in the decision to submit the man-uscript for publication.Duality of Interest. J.S.-S. reports serving on theboard of and receiving grant support through hisinstitution from the International Nut and DriedFruit Council, receiving consulting personal feesfrom Danone, Font Vella Lanjaron, Nuts for Life,and Eroski, and receiving grant support throughhis institution from Eroski. E.R. reports grants,nonfinancial support, and other fees from theCalifornia Walnut Commission and Alexion; per-sonal fees and nonfinancial support from Merck

Sharp & Dohme; personal fees, nonfinancialsupport, and other fees from Aegerion Pharma-ceuticals and Ferrer International; grants andpersonal fees from Sanofi Aventis; grants fromAmgen and Pfizer; and personal fees from AkceaTherapeutics, outsideof thesubmittedwork. X.P.reports serving on the board of and receivingconsulting personal fees from Sanofi, Amgen,and Abbott Laboratories and receiving personallecture fees from Esteve, Lacer, and Rubio Lab-oratories. L.D. reports grants from FundacionCerveza y Salud. M.D.-R. reports receivinggrants from the Diputacion Provincial de Jaenand the Caja Rural de Jaen. No other potentialconflicts of interest relevant to this article werereported.Author Contributions. J.S.-S., A.D.-L., M.Fit.,D.C., L.S.-M., J.W., D.R., R.E., J.Vid., J.A.M., F.A.,C.V., E.R., J.Vio., J.L.-M., A.B.-C., J.A.T., F.J.T.,V.M., J.L., X.P., L.D., M.D.-R., P.M., E.G.-G.,M.Fio., and M.A.M.-G., conducted the research.J.S.-S., M.Fit., D.C., L.S.-M., J.W., D.R., R.E., J.Vid.,J.A.M., F.A., C.V., E.R., J.Vio., J.L.-M., A.B.-C.,J.A.T., F.J.T., V.M., J.L., X.P., L.D., M.D.-R., P.M.,E.G.-G., M.Fio., and M.A.M.-G. designed theresearch. J.S.-S. and A.D.-L., analyzed the dataand wrote the article. J.S.-S., A.D.-L., M.R.-C., J.B.,M.Fit., D.C., L.S.-M., J.W., D.R., R.E., J.Vid., J.A.M.,F.A., C.V., E.R., J.Vio., J.L.-M., A.B.-C., J.A.T., F.J.T.,V.M., J.L., X.P., L.D., M.D.-R., P.M., E.G.-G.,J.D.-E., N.B., O.C., J.V.S., M.Fio., M.A.Z., M.B.,A.G., and M.A.M.-G. revised the manuscript forimportant intellectual content and read and ap-proved the final manuscript. J.S.-S. and A.D.-L.are the guarantors of this work and, as such,had full access to all of the data in the study andtake responsibility for the integrity of the dataand the accuracy of the data analysis.

References1. Guh DP, Zhang W, Bansback N, Amarsi Z,Birmingham CL, Anis AH. The incidence ofco-morbidities related to obesity and over-weight: a systematic review and meta-analysis.BMC Public Health 2009;9:882. Hruby A, Hu FB. The epidemiology of obes-ity: a big picture pharmacoeconomics. 2015;33:673–689.3. Pi-Sunyer X, Blackburn G, Brancati FL, et al.;Look AHEAD Research Group. Reduction inweight and cardiovascular disease risk factorsin individuals with type 2 diabetes: one-yearresults of the Look AHEAD trial. Diabetes Care2007;30:1374–13834. Zomer E, Gurusamy K, Leach R, et al. Inter-ventions that cause weight loss and the impacton cardiovascular risk factors: a systematic re-view and meta-analysis. Obes Rev 2016;17:1001–10115. Knowler WC, Fowler SE, Hamman RF, et al.;Diabetes Prevention Program Research Group.10-year follow-up of diabetes incidence andweight loss in the Diabetes Prevention ProgramOutcomes Study. Lancet 2009;374:1677–16866. Wing RR; Look AHEAD Research Group. Long-term effects of a lifestyle intervention on weightand cardiovascular risk factors in individualswith type 2 diabetes mellitus: four-year resultsof the Look AHEAD trial. Arch Intern Med2010;170:1566–15757. Wing RR, Lang W, Wadden TA, et al.; LookAHEAD Research Group. Benefits of modest

weight loss in improving cardiovascular riskfactors in overweight and obese individualswith type 2 diabetes. Diabetes Care 2011;34:1481–14868. Ma C, Avenell A, Bolland M, et al. Effects ofweight loss interventions for adults who areobese on mortality, cardiovascular disease,and cancer: systematic review and meta-analy-sis. BMJ 2017;359:j48499. Wing RR, Bolin P, Brancati FL, et al.; LookAHEAD Research Group. Cardiovascular effectsof intensive lifestyle intervention in type 2 di-abetes. N Engl J Med 2013;369:145–15410. Martınez-Gonzalez MA, Salas-Salvado J,Estruch R. Intensive lifestyle intervention intype 2 diabetes. N Engl J Med 2013;369:235711. Howard BV, Van Horn L, Hsia J, et al. Low-fatdietary pattern and risk of cardiovascular dis-ease: the Women’s Health Initiative Random-ized Controlled Dietary Modification Trial.JAMA 2006;295:655–66612. Foreyt JP, Salas-Salvado J, Caballero B, et al.Weight-reducing diets: are there any differen-ces? Nutr Rev 2009;67(Suppl. 1):S99–S10113. Martınez-Gonzalez MA, Salas-Salvado J,Estruch R, Corella D, Fito M, Ros E; PREDIMEDInvestigators. Benefits of the mediterraneandiet: insights from the PREDIMED study. ProgCardiovasc Dis 2015;58:50–6014. Esposito K, Kastorini C-M, PanagiotakosDB, Giugliano D. Mediterranean diet andweight loss: meta-analysis of randomizedcontrolled trials. Metab Syndr Relat Disord2011;9:1–1215. Shai I, Schwarzfuchs D, Henkin Y, et al.;Dietary Intervention Randomized ControlledTrial (DIRECT) Group. Weight loss with a low-carbohydrate, Mediterranean, or low-fat diet.N Engl J Med 2008;359:229–24116. Mancini JG, Filion KB, Atallah R, EisenbergMJ. Systematic review of the Mediterraneandiet for long-term weight loss. Am J Med2016;129:407–415.e417. Estruch R, Ros E, Salas-Salvado J, et al.;PREDIMED Study Investigators. Primary preven-tion of cardiovascular disease with a Mediter-ranean diet supplemented with extra-virginolive oil or nuts. N Engl J Med 2018;378:e3418. Alberti KGMM, Eckel RH, Grundy SM, et al.;International Diabetes Federation Task Forceon Epidemiology and Prevention; NationalHeart, Lung, and Blood Institute; AmericanHeart Association; World Heart Federation;International Atherosclerosis Society; Interna-tional Association for the Study of Obesity.Harmonizing the metabolic syndrome: a jointinterim statement of the International DiabetesFederation Task Force on Epidemiology and Pre-vention; National Heart, Lung, and Blood In-stitute; American Heart Association; WorldHeart Federation; International Atherosclero-sis Society; and International Association forthe Study of Obesity. Circulation 2009;120:1640–164519. Schroder H, Fito M, Estruch R, et al. A shortscreener is valid for assessing Mediterraneandiet adherence among older Spanish men andwomen. J Nutr 2011;141:1140–114520. Molina L, Sarmiento M, Pe~nafiel J, et al.Validation of the Regicor short physical activityquestionnaire for the adult population. PLoS One2017;12:e0168148

care.diabetesjournals.org Salas-Salvado and Associates 787

21. Martınez-Gonzalez MA, Lopez-Fontana C,Varo JJ, Sanchez-Villegas A,Martinez JA. Validationof the Spanish version of the physical activityquestionnaire used in the Nurses’ Health Studyand the Health Professionals’ Follow-up Study.Public Health Nutr 2005;8:920–92722. Jones CJ, Rikli RE, Beam WCA. A 30-s chair-stand test as a measure of lower body strengthin community-residing older adults. Res Q ExercSport 1999;70:113–11923. Sacks FM, Bray GA, Carey VJ, et al. Com-parison of weight-loss diets with different com-positions of fat, protein, and carbohydrates. NEngl J Med 2009;360:859–87324. Franz MJ, VanWormer JJ, Crain AL, et al.Weight-loss outcomes: a systematic review andmeta-analysis of weight-loss clinical trials witha minimum 1-year follow-up. J Am Diet Assoc2007;107:1755–176725. Johnston BC, Kanters S, Bandayrel K, et al.Comparison of weight loss among named diet

programs in overweight and obese adults:a meta-analysis. JAMA 2014;312:923–93326. Leung AWY, Chan RSM, Sea MMM, Woo J.An overview of factors associated with adher-ence to lifestyle modification programs forweight management in adults. Int J EnvironRes Public Health 2017;1427. Mansoor N, Vinknes KJ, Veierød MB,Retterstøl K. Effects of low-carbohydrate dietsv. low-fat diets on body weight and cardio-vascular risk factors: a meta-analysis of rand-omised controlled trials. Br J Nutr 2016;115:466–47928. Estruch R,Martınez-GonzalezMA, Corella D,et al.; PREDIMED Study Investigators. Effects ofa Mediterranean-style diet on cardiovascularrisk factors: a randomized trial. Ann InternMed 2006;145:1–1129. Correia ML de G, Haynes WG. Leptin, obe-sity and cardiovascular disease. Curr Opin Neph-rol Hypertens 2004;13:215–223

30. Zilverschoon GRC, Tack CJ, Joosten LAB,Kullberg BJ, van der Meer JWM, Netea MG.Interleukin-18 resistance in patientswith obesityand type 2 diabetes mellitus. Int J Obes 2008;32:1407–141431. Schernthaner G-H, Kopp H-P, Kriwanek S,et al. Effect of massive weight loss induced bybariatric surgery on serum levels of interleukin-18 and monocyte-chemoattractant-protein-1in morbid obesity. Obes Surg 2006;16:709–71532. Kanda H, Tateya S, Tamori Y, et al. MCP-1contributes to macrophage infiltration intoadipose tissue, insulin resistance, and hepaticsteatosis in obesity. J Clin Invest 2006;116:1494–150533. Gerszten RE, Garcia-Zepeda EA, Lim YC,et al. MCP-1 and IL-8 trigger firm adhesionof monocytes to vascular endothelium un-der flow conditions. Nature 1999;398:718–723

788 Lifestyle Intervention Program and Weight Loss Diabetes Care Volume 42, May 2019