Systematic review of reviews of intervention components ...

Evidence on long-term effectiveness suggested the need for greater consideration of behaviour maintenance strategies. Conclusions. This ... Skiptomaincontent Advertisement SearchallBMCarticles Search DownloadPDF Researcharticle OpenAccess Published:18February2011 Systematicreviewofreviewsofinterventioncomponentsassociatedwithincreasedeffectivenessindietaryandphysicalactivityinterventions ColinJGreaves1,KateESheppard1,CharlesAbraham2,WendyHardeman3,MichaelRoden4,PhilipHEvans5,PeterSchwarz6&TheIMAGEStudyGroup BMCPublicHealth volume 11,Article number: 119(2011) Citethisarticle 57kAccesses 623Citations 96Altmetric Metricsdetails AbstractBackgroundTodevelopmoreefficientprogrammesforpromotingdietaryand/orphysicalactivitychange(inordertopreventtype2diabetes)itiscriticaltoensurethattheinterventioncomponentsandcharacteristicsmoststronglyassociatedwitheffectivenessareincluded.Theaimofthissystematicreviewofreviewswastoidentifyinterventioncomponentsthatareassociatedwithincreasedchangeindietand/orphysicalactivityinindividualsatriskoftype2diabetes.MethodsMEDLINE,EMBASE,CINAHL,PsycInfo,andtheCochraneLibraryweresearchedforsystematicreviewsofinterventionstargetingdietand/orphysicalactivityinadultsatriskofdevelopingtype2diabetesfrom1998to2008.Tworeviewersindependentlyselectedreviewsandratedmethodologicalquality.Individualanalysesfromreviewsrelatingeffectivenesstointerventioncomponentswereextracted,gradedforevidencequalityandsummarised.ResultsOf3856identifiedarticles,30mettheinclusioncriteriaand129analysesrelatedinterventioncomponentstoeffectiveness.Theseincludedcausalanalyses(basedonrandomisationofparticipantstodifferentinterventionconditions)andassociativeanalyses(e.g.meta-regression).Overall,interventionsproducedclinicallymeaningfulweightloss(3-5kgat12months;2-3kgat36months)andincreasedphysicalactivity(30-60mins/weekofmoderateactivityat12-18months).Basedoncausalanalyses,interventioneffectivenesswasincreasedbyengagingsocialsupport,targetingbothdietandphysicalactivity,andusingwell-defined/establishedbehaviourchangetechniques.Increasedeffectivenesswasalsoassociatedwithincreasedcontactfrequencyandusingaspecificclusterof"self-regulatory"behaviourchangetechniques(e.g.goal-setting,self-monitoring).Noclearrelationshipswerefoundbetweeneffectivenessandinterventionsetting,deliverymode,studypopulationordeliveryprovider.Evidenceonlong-termeffectivenesssuggestedtheneedforgreaterconsiderationofbehaviourmaintenancestrategies.ConclusionsThiscomprehensivereviewofreviewsidentifiesspecificcomponentswhichareassociatedwithincreasedeffectivenessininterventionstopromotechangeindietand/orphysicalactivity.Tomaximisetheefficiencyofprogrammesfordiabetesprevention,practitionersandcommissioningorganisationsshouldconsiderincludingthesecomponents. PeerReviewreports BackgroundThedevelopmentoftype2diabetesisstronglyassociatedwithbeingoverweight,obeseorphysicallyinactive[1,2].Largerandomisedcontrolledtrials(RCTs)haveshownthatrelativelymodestchangesinlifestyle(increasingfibre(≥15g/1000kcal),reducingtotalfat(<30%ofenergyconsumed)andsaturatedfat(<10%ofenergyconsumed),engaginginmoderatephysicalactivity(≥30mins/day),weightreduction(5%))canreducetheriskofprogressiontotype2diabetesinadultswithimpairedglucoseregulation(alsoknownaspre-diabetes)byaround50%[3–7].Inonestudy,achievingfourormoreoftheabovetargetsledtozeroincidenceoftype2diabetesuptosevenyearslater[8].Consequently,promotingchangesinphysicalactivityanddietaryintakeisnowrecommendedinnationalandinternationalguidelinesasafirstlinetherapyforpreventingtype2diabetes[9–12].Anumberofdiabetespreventionprogrammeshavebeendevelopedinternationally(e.g.inFinland,[13]Germany,[14,15]theUS,[16,17]Australia[18]andChina[19]).However,nationaldiabetespreventionstrategiesarestilllackinginmanycountries.Thecost-effectivenessoflifestyleinterventionapproachesfordiabetespreventionisalreadywellestablishedandisfavourableincomparisontopharmacologicalapproaches[20–22].However,mostinterventionsusedtodateinaresearchsettingareconsideredtobetoointensiveforwidespreadimplementationinhealthservices[23].Forexample,theUSDiabetesPreventionProgramme[4]involved16individualcounsellingsessionsplusindividualcoachingandamaintenanceprogrammewithfurtherindividualandgroupsessions.Amajorchallengeforhealthcareprovidersthereforeishowtoachievethelifestylechangesneededtopreventtype2diabetes(anditsassociatedcardiovascularrisk)withoutoverstretchingexistingbudgetsandavailableresources[24,25].Intranslatingtheresearchevidenceintopracticalprogrammesitiscriticaltoensurethattheinterventioncomponents(i.e.behaviourchangetechniquesandstrategies)andcharacteristics(e.g.setting,deliverymode,interventionprovider)moststronglyassociatedwitheffectivenessareincluded.Wethereforeaimedtosystematicallyreviewexistingsystematicreviewstosummarisetheevidencerelatingthecontentofinterventionsforpromotingdietaryand/orphysicalactivitychangetotheireffectivenessinproducingweightandbehaviourchange.Thereviewfocusedonevidencerelatingtoindividualsatriskoftype2diabetesduetolifestyle(e.g.inactivity)orclinicalriskfactors(e.g.overweight,elevatedbloodpressure).MethodsDataSourcesandSearchStrategyOneauthor(KS)searchedMEDLINE,EMBASE,CINAHL,PsycInfo,andtheCochraneLibraryforsystematicreviewsintheEnglishlanguage,publishedbetweenJanuary1998andMay2008(thesearchtermswerereviewedbyseveralauthors(CG,CA,WH)andareprovidedinAdditionalfile1TableS1).Referencelistsofselectedreviewsandrelevantclinicalguidelineswerealsosearchedandexpertsintheareawerecontactedinordertoidentifyunpublishedreviews.ReviewselectionTworeviewers(KS,CG)independentlyexaminedtitlesandabstracts.Relevantreviewarticleswereobtainedinfull,andassessedagainsttheinclusionandstudyqualitycriteriadescribedbelow.Inter-revieweragreementoninclusionwasassessedusingkappastatisticsandanydisagreementswereresolvedthroughdiscussion.Inclusioncriteria1)Typeofstudy:Systematicreviewsandmeta-analysesincludingRCTs,observationalstudies,case-controlledorotherquasi-experimentalstudies.Comparisongroupscouldincludeusualcare,nointerventionorotherinterventions.2)Typeofintervention:Interventionspromotingphysicalactivityand/ordietarychangeattheindividual-level(i.e.interventionsdeliveredtoindividualseithersinglyoringroupsessions,butnotwhole-communityorwhole-populationlevelinterventionssuchasmediacampaignsorchangesinthelocalenvironment).3)Studypopulations:Adults(18yearsandover)atriskofdevelopingtype2diabetes,selectedbecausetheywereobese,overweight,sedentary,hadhypertension,impairedfastingglucose,impairedglucosetolerance,hyperlipidaemia,metabolicsyndrome,polycysticovariansyndrome,gestationaldiabetes,afamilyhistoryoftype2diabetesorcardiovasculardisease,orhadbeenidentifiedashavingahighcardiovasculardiseaseriskscore(e.g.usingavalidatedriskscoresuchasQ-RISKorFramingham).Exclusioncriteria1)Reviewsnotmeetingpre-definedcriteriaformethodologicalquality(Additionalfile1TableS2).2)Reviewswhichfocusedonpeoplewithexistingdiabetes,cardiovasculardisease,orsolelyonhealthyadults,orwhichwereconfinedtogroupswithsignificantco-morbidities(e.g.arthritis,mentalhealth).OutcomesWeselectedreviewswheretheprimaryoutcomemeasurewasweight,weightloss(kgorBodyMassIndex(BMI),proportionsofpeopleachievingatargetweightloss),changesinphysicalactivity(e.g.frequency,met-hrsperweek)ordietarybehaviour.Behaviourscouldbemeasuredobjectively(e.g.withaccelerometers)orbyself-report(e.g.dietaryintakequestionnaires).Cardio-respiratoryfitnesswasconsideredasaproxyforchangeinphysicalactivity.Asself-reportincreasestheriskofmeasurementbias,[26,27]wehavehighlightedfindingsbasedonself-reportinthedatatables(Additionalfile2TablesS7-S14).Wealsoexaminedpapersforotheroutcomeswhichmightbeofinterestinrelationtochangeinweight,diet,orphysicalactivitybehaviourorinrelationtotheprogressiontotype2diabetes.StudyqualityassessmentReviewqualitywasratedindependentlybytwoauthors(KS,CG)forasub-sample(35outof107)ofthearticlesidentifiedaspotentiallyrelevant,usingtheOverviewQualityAssessmentQuestionnaire(OQAQ;[28]Additionalfile1TableS2).Thereafter,reviewqualitywasratedbyoneresearcher(KS)andverifiedbyanother(CG).ReviewswereincludediftheirOQAQscorewas14ormore(possiblerange0-18)andiftheyscoredatleastonepointforeitherofthetwoOQAQcriteriaaboutassessingquality/takingqualityintoaccountinanalyses(thiswasintendedtomaximisethelikelyqualityofevidenceunderlyingthereview-levelanalyses).Apercentagescorewascalculatedforinter-rateragreement(definedas≤1pointofvariationonOQAQscores)andanydisagreementswereresolvedbydiscussion.DataextractionWeextracteddataontheeffectivenessofinterventionsandontherelationshipofeffectivenesstosevenpre-definedinterventioncomponents.Thesewere:Theoreticalbasis(i.e.weextractedanalysesrelatingeffectivenesstotheuseofanystatedtheoryofbehaviourorbehaviourchange);Behaviourchangetechniquesused(e.g.theuseofspecifictechniquessuchasgoal-setting,problem-solvingortheplanneduseofsomeclearlydefinedsetofbehaviourchangetechniques:SeeTable1forexamples);Modeofdelivery(e.g.group-based,individual,self-delivery,mixed-mode);Interventionprovider(e.g.generalpractitioner,counsellor);Intensity(e.g.numberofsessions,totalcontacttime);Characteristicsofthetargetpopulation(e.g.age,ethnicity,riskstate);andSetting(e.g.primarycare,workplace).Datawereextractedagainstadataextractiontemplatebyoneauthor(KS)andcheckedbyanother(CG)withreferencetothefulltextofthearticle.Extracteddataalsoincludedinclusionandexclusioncriteria,reportedanalysesandanalysistype.Table1Definitionsof'establishedbehaviourchangetechniques'Fullsizetable GradingofevidenceAnevidencegradewasgiventoeachreportedanalysis,basedontheScottishIntercollegiateGuidelinesNetwork(SIGN)evidencegradingsystem[29].Thissystemgradestheriskofbiasassociatedwithaparticularpieceofevidenceonahierarchyfrommeta-analysisandRCTevidence(grade1)downtoexpertopinion(grade4),withadditionalindicators(++,+or-)toindicatemethodologicalquality.TheSIGNsystemwasmodified,asourreviewaimedtoidentifytherelativeeffectivenessofinterventioncomponents,ratherthaneffectivenessperse(seeAdditionalfile1TableS3forfulldetails).AlthoughtheSIGNevidencegradingusesanalpha-numericsystem(1++,1+,1-,2++,2+,2-),foreaseofreadingwehaveconvertedthistoatext-basedformat.Foreachanalysisthequalityoftheevidence(thedegreeofconfidencethattheriskofbiasislow)isdescribedaseither"high(++),medium(+)orlow(-)".Eachanalysisisalsocategorisedasbeingeither"causal"evidence(SIGNgrade1;evidencefrommeta-analysesorsummariesofRCTswherethecomponentorcharacteristicofinterestwasexperimentallymanipulated)or"associative"evidence(SIGNgrade2;evidencefromcorrelationalorobservationalanalyses).Wealsoappliedacategoryof"verylowquality"foranalyseswithverylowapparentpower(totalN<100).Thereportingthatfollowsexcludesthisverylowqualityevidence,althoughitisincludedinthesupplementarydatatablesforcompleteness.AnalysisNostatisticalanalysesormeta-analyseswereconducted.Instead,theexistinganalysesreportedinthearticlesreviewedwereextractedandreportedinasystematicformat(Additionalfile2TablesS7toS14).EachanalysiswasgradedusingtheadaptedSIGNcriteriaasdescribedaboveandanarrativesynthesisispresentedbelow,indicatingboththequalityoftheevidence(low,medium,high)andwhetheritiscausalorassociativeinnature.Inaccordancewithreportingguidelinesforsystematicreviews,aPRISMA(PreferredReportingItemsforSystematicReviewsandMeta-Analyses)checklistisavailableforthisreview(Additionalfile3).ResultsSearchesidentified3856potentiallyrelevantarticles.Followingreviewoftitlesandabstracts,96articleswereretrievedandquality-assessed.Anadditional11articleswereidentifiedthroughreferencelistsandgreyliterature.Ofthese107articles,30metboththeselectionandqualitycriteria(Figure1)andtheseareidentifiedbyanasteriskinthereferencelist[30–59].Theinter-raterreliability(Kappa)forapplyingreviewselectioncriteriawas0.71(95%CI:0.61to0.80),andtheproportionforinter-revieweragreementonreviewqualitywas0.70(95%CI:0.55to0.85).Figure1 Flowdiagramofstudyselection.Fullsizeimage ReviewcharacteristicsThecharacteristicsoftheincludedandexcludedreviewsaresummarisedinAdditionalfile1TablesS4andS5.Tenreviewsexaminedphysicalactivityinterventions,threeexamineddietaryinterventionsandseventeenexaminedboth.Reviewsincludeddatafromarangeofpopulations(e.g.sedentary,overweight,obese,impairedglucosetolerance)anddeliverysettings(e.g.homebased,leisurecentrebased,primarycare,workplace)andusedavarietyofdescriptive,meta-analyticandmeta-regressionanalysestoinvestigatetheassociationofinterventioncomponentswitheffectiveness.Weidentified129analysesofrelationshipsbetweeninterventioncomponentsandeffectiveness,and55analysesofinterventioneffectiveness(Additionalfile2TablesS7toS14).Thedatesofpublishedstudiesincludedinthereviewsexaminedrangedfrom1966to2008.StudyqualityThemethodologicalqualityofincludedreviews(Additionalfile1TablesS4,S6)wasgenerallygood(medianOQAQscore=15.6).Themostcommonmethodologicalweaknesseswerethelackofuseofstudyqualitydatatoinformanalyses(e.g.bysensitivityanalysis,orbyconstructingseparateanalyseswhichexcludedlowqualitytrials)andpotentialbiasintheselectionofarticles(e.g.notusingindependentassessors).EvidencesynthesisTheextractedanalysesandevidencegradesforeachanalysisarepresentedinAdditionalfile2TablesS7toS14.Thefindingscanbesummarisedasfollows:-Overalleffectiveness(Additionalfile2,TableS7)WeightLossHighqualitycausalevidence(grade1++)fromeightmeta-analysesofRCTsfromfourreviewsshowedthatinterventionstopromotechangesindiet(orbothdietandphysicalactivity)producedmoderateandclinicallymeaningfuleffectsonweightloss(typically3-5kgat12months,2-3kgat36months)[37,38,42,50].Theeffectivenessofsuchinterventions(aswellasphysicalactivityonlyinterventions)inproducingweightlosswasfurthersupportedbymediumandlowqualitycausalevidence(grade1+and1-)from14meta-analysesandsummariesofRCTsfromsixreviews(eightmedium,sixlowqualityanalyses)[31,39,49,54,57,59].PhysicalActivityHighqualitycausalevidencewasfoundfromfourmeta-analysesofRCTsintworeviewsthatphysicalactivityinterventionscanproducemoderatechangesinself-reportedphysicalactivity(standardisedmeandifferencearound0.3;OddsRatioforachievinghealthyactivitytargetsaround1.2to1.3)andcardio-respiratoryfitness(standardisedmeandifferencearound0.5)ataminimum6monthsoffollowup[41,59].Thiswassupportedbylowerqualitycausalevidencefromsixmeta-analysesofRCTsandsummariesofRCTsandotherstudies(threemediumandthreelowqualityanalyses)fromthreesystematicreviewsthatinterventionstoincreasephysicalactivityincreasedself-reportedphysicalactivity(typicallyequivalentto30-60minutesofwalkingperweek)atamedianof6weeksto19monthsoffollowup[38,40,51].However,itisworthnotingthattherewerefewexamplesoftrialswithsuccessfuloutcomesatmorethan12months.DietaryIntakeMediumandlowerqualitycausalevidencefrommeta-analysesanddescriptivesummariesofRCTs(nineanalysesfromthreeseparatereviews:sixmedium,threelow)thatfoundpositivechangesinself-reporteddiet(calorie,fat,fibre,fruitandvegetableintake)at6to19monthsoffollowupfordietaryinterventions[38,34,44].OtherOutcomesHighqualitycausalevidence(grade1++)fromonemeta-analysisofRCTs[43]showedthatinterventionstopromotechangesindietorphysicalactivity(orboth)producedmoderateandclinicallymeaningfuleffectsontheriskofprogressiontotype2diabetes(relativeriskreductionof49%at3.4years)inpeoplewithimpairedglucoseregulation.Onereviewwhichexaminedvariationsineffectivenessovertime[37]showedthatweightlosstendedtoreverseonceinterventionsceasedormovedfromanactivetoamaintenancephase(netweightlossduringactivephase0.08BMIunitspermonth;netweightgainduringmaintenancephase0.03BMIunitspermonth).Theoreticalbasis(Additionalfile2,TableS8)Onemeta-regressionanalysisprovidedmediumqualityassociativeevidence(grade2+)suggestingthatinterventionswithanexplicitlystatedtheoreticalbasis(e.g.SocialCognitiveTheory,[60]TheoryofPlannedBehaviour[61])werenomoreeffectiveinproducingchangesineitherweightorincombineddietaryandphysicalactivityoutcomesthaninterventionswithnostatedtheoreticalbasis[38].However,fourmeta-regressionanalyses(allmediumqualityassociativeanalyses)intworeviews[38,48]didfindanassociationbetweentheuseofatheoreticallyspecifiedclusterof'self-regulatory'interventiontechniques(specificgoal-setting,promptingself-monitoring,providingfeedbackonperformance,goalreview)andincreasedeffectivenessintermsofa)weightloss,b)changeindietaryoutcomes,c)changeinphysicalactivityandd)combined(standardisedmeandifferenceforeitherdietarychangeorphysicalactivity)outcomes.Behaviourchangetechniques(Additionalfile2,TableS9)Categorisationofinterventionsvariedgreatlybetweenreviews,withcategoriesoftenconceptuallyoverlappingandvaguelydefined(e.g.dietvs.exercisevs.behaviouralintervention).Despitethis,wehavesummarisedevidenceontheuseofwhatwehavecalled"established,welldefinedbehaviourchangetechniques",basedonthosereviewswhereclearandspecificdefinitionswereprovided(seeTable1fordefinitions).FurtherdefinitionofthespecificbehaviourchangetechniquescitedinTable1andthosementionedinthetextbelowcanbefoundinarecenttaxonomyofbehaviourchangetechniques[62].Causalevidencefromonemediumqualitymeta-analysisindicatedthatchangeinweightwasgreaterwhenestablished,welldefinedbehaviourchangetechniqueswereaddedtointerventions(e.g.whendietaryadviceplusawell-definedbehaviouralinterventionusingestablishedbehaviourchangetechniqueswascomparedwithdietaryadvicealone).Theweightlossachievedbyaddingestablishedbehaviourchangetechniquestointerventionswas4.5kgatamedian6monthsoffollowup[54].Thiswassupportedbytwoassociativeanalyses(onemediumandonelowquality)whichcomparedtheresultsofdifferentgroupsofstudiesinwhichtheinterventionseitherdidordidnotuseestablished,well-definedbehaviourchangetechniques.Usingestablishedbehaviourchangetechniqueswasassociatedwithincreasedweightloss(2.5to5.5kg)comparedwithnon-behaviouralinterventions(0.1to0.9kg)[46,47].Evidencefromfivelowtomediumqualityassociativeanalysesintworeviewsattemptedtorelatethenumberofbehaviourchangetechniquesusedtoeffectivenessintermsofweightlossorchangesindietorphysicalactivity.Theevidencewasequivocalwiththepatternofdatasuggestingapossibleassociation,butonlyoneanalysisapproachedsignificance[38,48].UseofspecificbehaviourchangetechniquesHighqualitycausalevidencewasfoundthataddingsocialsupporttointerventions(usuallyfromfamilymembers)providedanadditionalweightlossof3.0kgatupto12months(comparedwiththesameinterventionwithnosocialsupportelement)[31].Mediumtolowqualityassociativeevidence(fromthreemeta-regressionanalysesandtwoassociativeanalysesinthreereviews)suggestedthateffectivenessforinitialbehaviourchange(i.e.changeinweight,dietorphysicalactivitywasassociatedwithusingthefollowingtechniques(NB:definitionsofthesecanbefoundinarecenttaxonomyofbehaviourchangetechniques[62]):1)Fordietarychange:providinginstruction,establishingself-monitoringofbehaviour,useofrelapsepreventiontechniques[38,48].2)Forphysicalactivitychange:promptingpractice,establishingself-monitoringofbehaviour,individualtailoring(e.g.ofinformationorcounsellingcontent)[38,40,48].Onereviewalsoprovidedmediumqualitycausalevidence(adescriptivesummaryofindividualRCTfindings)thatbriefadvice,whichusuallyincludedgoal-setting,ledtoanincreaseinwalkingactivity(27mins/weekwalkingat12monthsoffollowup)[51].Goal-settingalongsidetheuseofpedometerswasalsoassociatedwithincreasedwalking(seebelow).Furthermediumqualityassociativeevidencesuggestedthatincreasedmaintenanceofbehaviourchangewasassociatedwiththeuseoftimemanagementtechniques(forphysicalactivity)andencouragingself-talk(forbothdietarychangeandphysicalactivity)[38].Threereviewsexaminedinterventionsthatusedpedometers(i.e.self-monitoringofphysicalactivity)topromotewalking:Mediumqualitycausalevidence(twoanalysesfromtworeviews)supportedtheeffectivenessofpedometerbasedinterventionsforincreasingwalkingactivity[33,51](meanincreaseof2004stepsperdayatamedian11weeks;medianincreaseintimewalkingof+54minperweekatamedian13weeks).Itmustbenotedthatthevastmajorityoftheinterventionsincludedinthesemeta-analysesincludedeitherstep-goalsorstepdiaries(orboth)alongsidetheuseofpedometers,sotheevidencedoesnotsupporttheuseofpedometersinisolationfromtheseadditionaltechniques.Indeed,associativeanalysesfromonereview[33]suggestedthattheuseofa)astepdiary(onelowqualityanalysis)andb)goal-setting(onelowandonemediumqualityanalysis)incombinationwithuseofapedometerwasassociatedwithincreasedwalking.Mediumtohighqualityassociativeevidence(basedonmeta-analysisofonlytheinterventionarmsofstudies)fromtworeviews[33,52]suggestedthatsmallchangesinweightmightalsobeachievablewithpedometerbasedinterventions(e.g.changeinBMIof0.38kg/m2at11weeks).MotivationalinterviewingMotivationalinterviewingisadistinctcombinationofbehaviourchangetechniques(includingdecisionalbalanceandrelapsepreventiontechniques)deliveredinaspecificstyle(usingpatientcentredempathybuildingtechniques,suchasrollingwithresistance;affirmationandreflectivelistening)[63].Highqualitycausalevidencefromonemeta-analysisofRCTs[53]foundthatmotivationalinterviewingwassignificantlymoreeffectivethantraditionaladvice-givingforinitiatingchangesinweight(producinganetdifferenceof0.72BMIunitscomparedwithtraditionaladvice-giving)at3to24monthsoffollowup(mostlyunder6months).Afurthermeta-analysisofRCTs[35]providedmediumqualitycausalevidenceoftheeffectivenessofmotivationalinterviewingforacombinedphysicalactivityanddietaryoutcome,atupto4monthsoffollowup(StandardisedMeanDifference0.53).TargetingmultiplebehavioursCausalevidencefromnineanalysesinfourreviews(onehigh,fourmediumandfourlowquality)showedthatinterventionswhichtargetedbothphysicalactivityanddietratherthanonlyoneofthesebehavioursproducedhigherweightchange(additionalweightlossaround2-3kgatupto12months)[31,36,37,54].Modeofdelivery(Additionalfile2,TableS10)Theevidencefromfivereviewsofdietaryand/orphysicalactivityinterventionwasmixed.Fiveassociativeanalyses(threemediumandtwolowquality)fromfourreviewsfailedtofindaclearassociationbetweeneffectivenessandmodeofinterventiondeliveryforweightloss,dietarychangeorphysicalactivitychange[38,46,48,51].Onereviewfoundmediumqualityassociativeevidencethat'mixedmode'(individualandgroup)deliverywassignificantlyrelatedtogreatereffectiveness,comparedwithindividualdelivery,forinitialweightloss(upto6months),butnotforweightlossmaintenance(atamean19months)[38].However,itisworthnotingthatthereisevidencefromindividualhighqualityRCTs(basedondataintheevidencetablesoftheincludedreviews)thatindividual,group,andmixedmodeinterventionscanallbeeffectiveinchangingdietand/orphysicalactivity[31,38,51].Interventionprovider(Additionalfile2,TableS11)Therewasalackofhighqualityevidenceinthisareaforcomparisonsbetweenspecifictypesofinterventionprovider.Fourassociativeanalyses(twomedium,twolow)fromfourreviewsprovidednoconsistentorsignificantrelationshipbetweeninterventionproviderandweight,physicalactivityordietaryoutcomesatupto12monthsoffollowup[38,40,48,51].However,strongevidencefromindividualRCTs(basedondataintheevidencetablesoftheincludedreviews)showedthatawiderangeofproviders(withappropriatetraining)includingdoctors,nurses,dieticians/nutritionists,exercisespecialistsandlaypeople,candelivereffectiveinterventionsforchangingdietand/orphysicalactivity[38,40,43,48,51,52].Interventionintensity(Additionalfile2,TableS12)Definitionsofinterventionintensityreportedinthereviewsvariedconsiderably,incorporatingfrequencyandtotalnumberofcontacts,totalcontacttime,durationoftheinterventionandthenumberofbehaviourchangetechniquesused.Thefrequencyanddurationofclinicalcontactvariedwidely,rangingfrom1toaround80sessions,delivereddailytomonthlyandlastinganythingfrom15to150minutes,overperiodsrangingfrom1dayto2years.Forinstance,onereviewof17weightlossinterventionsthatcompareddifferentinterventionintensities,reportedthatthemediancontactfrequencywasweekly,themediansessionduration60minutes,andthemediandeliveryperiod10weeks[54].Physicalactivityinterventionsareoftenmuchmoreintensiveduetoafocusonpractisingthetargetbehaviour(e.g.Shawetal.[55]reportinterventionslasting3to12monthswith3to5sessionsperweeklastingamedian45minuteseach).WeightLossOverall,7outof9analysesofinterventionintensityfavouredhigherintensityinterventions.Onemeta-analysisoftensmallRCTs(N=306)comparingdifferentinterventionintensities[54]foundmediumqualitycausalevidencethatmoreintensiveinterventions(thoseincludingmorebehaviourchangetechniques,morecontacttimeoralongerdurationofintervention)generatedsignificantlymoreweightlossthanlessintensiveinterventions(anadditional2.3kgatamediansevenmonthsfollowup).Thiswassupportedbyamediumqualityassociativeanalysisfromthesamereview.However,itwasnotpossibletodeducefromtheavailabledatawhichcomponentofintensitydrivesthisrelationship.Mediumtolowqualityevidencefromthreeanalysesinthreereviews(onemediumquality,twolowquality)showedapositiveassociationbetweenthetotalnumberofcontactsandweightlossat12to38months[46,50,57].Associativeevidencefromtwoanalysesintworeviews(onehighquality,onelowquality)foundarelationshipbetweenincreasedfrequencyofcontactsandweightlossat6to15monthsoffollowup[37,47].However,twoassociativeanalyses(onehighandonemediumquality)intworeviews[37,38]foundnosuchrelationshipat6to60months.Twomediumqualityassociativeanalysesfoundmixedevidence(onepositiveonenegative)ontheassociationbetweeninterventiondurationandweightloss.DietaryChangeTwolowqualityassociativeanalyseswithinthesamereviewfoundapositiverelationshipbetweennumberofcontactsandself-reporteddietarychangeat12monthsoffollowup[34].PhysicalActivityTherewasalackofevidenceontherelationshipbetweeninterventionintensityandphysicalactivityoutcomes.Twolowqualityassociativeanalysesintworeviews[33,40]foundnoclearrelationshipbetweeninterventionintensity(duration)andphysicalactivityoutcomes.Characteristicsofthetargetpopulation(Additionalfile2,TableS13)GenderEightassociativeanalyses(threemediumquality,fivelowquality)fromsixreviewsfoundnoconsistentassociationbetweengenderandchangesinweightorphysicalactivityat10weeksto16monthsoffollowup[33,38,41,48,55,58].EthnicityAlthoughthereisevidence(withinsomeofthecomponenttrialsinthereviewsexamined)thatinterventionscanbeeffectiveforanumberofethnicgroups[4]therewasverylittlereview-levelevidenceontherelationshipbetweenethnicityandinterventioneffectiveness.Oneassociativeanalysis(lowquality)suggestedthatinterventionstudieswithahigherpercentageofwhiteCaucasianparticipantsachievedlargerdecreasesinBMIatamedianof12weeksoffollowup[33].Another(lowquality)associativeanalysisinthesamereviewreportednoassociationbetweenethnicityandincreasedwalking.AgeAssociativeanalyses(onemediumquality,onelowquality)fromtworeviews[33,55]suggestedthatolderpeoplelostmoreweightthanyoungerpeopleat10.5to16weeksoffollowup[33].Twofurther(lowquality)analysesfromtworeviewsfoundnorelationshipbetweenageandphysicalactivityat3and6monthsoffollowup[33,41].AtriskpopulationsArangeofevidence,includingstrongcausalevidencefromtwometa-analysesofsub-groupsofstudiesandassociativeevidencefrommeta-regressionanalysesfromseveralfurtherreviewsfoundthatchangesinweightand(atleastshort-term)physicalactivityarepossibleinhighriskaswellaslowerriskpopulations,includinghighandlowweight,highcardiovascularriskgroupsandsedentaryandnon-sedentarygroups,atbetween3and36monthsoffollowup[33,37,38,41–43,48,51].Fiveanalysesfromfourreviewsprovidedmixedevidenceastowhethertargetingofinterventionsatpeoplewhoaremoresedentarywasassociatedwithlargerincreasesintheamountofphysicalactivity(twomediumanalyses(onepositive,onenegative),threelowqualityanalyses(twonegative,onetrend)[33,41,48,51].DiabetesIntwoassociativeanalyses(onehighquality,onemediumquality),effectivenessforweightloss(at3to60months)wasfoundtobeconsiderablylowerforpeoplewithtype2diabetesthanforpeoplewithouttype2diabetes[37,38].WeightFouranalysesinfourreviews[33,41,42,48]providedmixedassociativeevidence(twomedium(onepositive,onenegative),twolowqualityanalyses(onepositive,onenegative))astowhethertargetingmoreoverweightpeoplewasassociatedwithlargerincreasesintheamountofweightlossachieved.However,onehighqualityassociativeanalysisshowedthatpeoplewithahigherstartingweightachievebetterhealthimprovementsat2to4.6years,intermsofareducedincidenceoftype2diabetes[43].Setting(Additionalfile2,TableS14)Exampleswerefound(basedondataintheevidencetablesofincludedreviews)ofeffectiveinterventionsdeliveredinawiderangeofsettings,includinghealthcaresettings,theworkplace,thehome,andinthecommunity[30,34].Fewreviewsformallyexaminedtheimpactofinterventionsettingoneffectiveness.However,onemediumqualityassociativeanalysisrevealednosignificantdifferencesinoutcomes(eitherdietaryorphysicalactivitychange)atsixmonthsbetweeninterventionsinprimarycare,communityandworkplacesettings[48].DiscussionThisreviewhas,forthefirsttime,systematicallyidentified,synthesisedandgradedawiderangeofevidenceabouttherelationshipofinterventioncontenttoeffectivenessinindividual-levelinterventionsforpromotingchangesindietand/orphysicalactivityinadultsatriskoftype2diabetes.Interventionsproducedsignificantandclinicallymeaningfulchangesinphysicalactivity(typicallyequivalentto30-60minutesofwalkingperweek,forupto18months)andinweight(typically3-5kgat12months,2-3kgat36months).Greatereffectivenessofinterventionswascausallylinked(inmeta-analysesandrandomisedtrialswhichexperimentallymanipulatedtheuseoftheseelements)withtargetingbothdietandphysicalactivity,mobilisingsocialsupportandtheuseofwell-described/establishedbehaviourchangetechniques.Greatereffectivenesswasalsoassociated(incorrelationalanalysesandnon-randomisedcomparisons)withusingaclusterofself-regulatorytechniques(goal-setting,promptingself-monitoring,providingfeedbackonperformance,goalreview[62,64]),andprovidingahighercontacttimeorfrequencyofcontacts.However,withregardtointensity,theamountofclinicalcontactininterventionsvariedwidely(seerangesreportedabove)andtheevidencedidnotsupporttherecommendationofanyparticularminimumthreshold.Theevidenceonpatternsofeffectivenessovertime[37]alsosuggestedthatthereisaneedforanincreasedfocusontheuseoftechniquestosupportbehaviourmaintenance.Therewerenoclearassociationsbetweenprovider,setting,deliverymode,ethnicityandageofthetargetgroupandeffectiveness.This(andevidencefromarangeofindividualRCTscitedinthereviewsexamined)suggeststhatinterventionscanbedeliveredsuccessfullybyawiderangeofprovidersinawiderangeofsettings,ingrouporindividualorcombinedmodes,andcanbeeffectiveforawiderangeofethnicandagegroups.Whiletheuseof"established,well-definedbehaviourchangetechniques"wasassociatedwithincreasedeffectiveness,itisworthemphasisingthatindividualtechniquesarerarelyappliedinisolationandshouldformpartofacoherentinterventionmodel.Therefore,aplannedapproachtointerventiondesignisrecommended,suchas"interventionmapping",[65]orothersystematicinterventiondevelopmentprocesses[66]whichselectinterventiontechniquestoaddresstargetedbehaviourchangeprocesses(andthataretailoredforthetargetpopulationandsetting).Takentogether,thefindingssuggestanumberofrecommendationsforoptimisingpracticeinthedevelopmentanddeliveryofinterventionstopromotechangesindietand/orphysicalactivityandtheseareoutlinedinTable2.Itishopedthatapplyingthesefindingswillhelptomeetthegrowingneedforlesscostly,butnonethelesseffective,type2diabetespreventionprogrammes.Table2RecommendationsforpracticeFullsizetable Althoughprovidingagreaterdegreeofdepthwithregardtointerventioncomponents,thesefindingsareconsistentwithUKguidanceforthepreventionandtreatmentofobesity(whichrecommendsengagingsocial(especiallyfamilybased)support,andtargetingbothdietandexercise)[67].ThefindingsarealsoconsistentwithrecentguidancefromtheAmericanHeartAssociation[68]onthepreventionofheartdiseaseinadultsagedover18,whichrecommendtheuseofmotivationalinterviewingaswellasgoal-setting,self-monitoringandahighcontactfrequency.Recentevidence-basedguidancefromtheUSAssociationofDiabetesEducatorsalsorecommendsgoal-setting,problem-solving(relapseprevention)andself-monitoringofplans(self-regulation)forsupportinghealthyeatingandincreasedphysicalactivityinpeoplewithtype2diabetes[69].Ourfindingsmayalsobemorewidelygeneralisabletoadultswithdiagnosedchronicdisease(e.g.type2diabetes,heartdisease)ortoapparentlyhealthyadults.StrengthsandlimitationsOurreviewfocusedonlyonhigherqualitysystematicreviews.WeidentifiedasubstantialnumberofreviewswhichsynthesiseddatafromalargenumberofRCTsandotherstudies,inawiderangeofagegroups,clinical/riskgroupsandsettings.Drawingtogetherthesefindingsinoneplacehasgeneratedacomprehensive,evidence-basedoverviewofwhichinterventioncomponentsaremostlikelytofacilitateeffectiveness.However,severalchallengesaffectingthesynthesisandinterpretationoftheavailableevidencewereencountered.Oneofthelimitationsmostcommonlycitedbyreviewauthorswasaninadequatedescriptionofbehaviouralinterventionsintheindividualstudyreports.Thiscausesdifficultiesforthereviewerincategorisinginterventioncontentandconductingsubsequentanalysestorelatecontenttoeffectiveness.Wethereforesuggestthatfutureinterventionstudyreports(andreviewsofindividualstudies)useanappropriatetaxonomytodescribe(andcategorise)behaviourchangetechniques[62].Amajorlimitationinassessingtheutilityofspecifictheoriesandtechniquesunderpinninginterventionsisthattechniquesmaynotbeimplementedrigorouslyormaynotfaithfullyrepresentthespecifiedtheories[62,70].Notably,noneofthe30reviewsthatweexaminedtookinterventionfidelityintoaccount.Hence,thelackofanassociationbetweentheuseofastatedtheoryandeffectivenessmayreflectalackofgoodtheoriesoritmayreflectpoorimplementationoftheories.Otherpotentiallyimportantsourcesofbiasincludemeasurementissues(especiallyinrelationtotheuseofself-reportdata);self-selectionofinterventionparticipants;andafailuretoconsiderpotentialbiasesduetostudyqualityinsomereviews.Furthermore,itisworthnotingthatwithassociativeevidence,othercovariatesthanthoseanalysedmayaccountforthestatedrelationships(e.g.theassociationbetweenintensityandeffectivenessmightbeexplainedtosomeextentbylowerqualityofinterventionbeingassociatedwithlowerintensity).Afurtherpotentialsourceofbiaswhichnoreviewaccountedforwasthelowsamplesizecontributingtosomeoftheanalysesexamined.Inparticular,itisworthnotingthat,whilstourrecommendation(Table2)ontheusefulnessofsocialsupporttechnicallymeritsagradeA(asitisbasedonlevel1+evidencefromameta-analysisofrandomisedcontrolledtrials),thetotalnumberofparticipantscontributingtothemeta-analysiswasonly127.Ifthegradingsystemhadtakensamplesizeintoaccount,wemayhavegiventhisrecommendationalowergrade.Ininterpretingtheaboveinformation,itshouldbenotedthattheanalysesconsideredwereinmanycasesbasedonoverlappingsetsoftrials(andotherstudies).Itshouldalsobenoted,asthisisareviewofreviewswewerenotabletosynthesiseormeta-analysedatafromindividualstudies,whichmayhaveyieldedvaluableevidence.Itisalsoworthnotingthatatthetimeoftheliteraturesearchtherewerenohighqualityreviewsontheuseofinternet-basedinterventions,sonoevidenceispresentedinthisarea.ImplicationsforpracticeandpolicyOurreviewhasgeneratedclearrecommendationsonhowinterventionsforpromotinglifestylechangewithindiabetespreventionprogrammescouldbedevelopedorrefinedtomaximiseeffectiveness(Table2).Ourrecommendationsgoconsiderablybeyondthedataonbasiceffectivenesspresentedintrialsandsystematicreviewsofdiabetespreventionprogrammestodate[3–8].Theycanbeuseful,forexample,inguidingthetranslationofeffective,high-intensity/highresource-useinterventionsinresearchcontextsintolower-cost(yetstilleffective)interventionsforimplementationinclinicalpractice.DirectionsforfutureresearchMorerigorousevaluationsoftheeffectivenessandcost-effectivenessofspecificinterventioncomponentsandclustersoftechniquesforpromotingandmaintainingchangeindietandphysicalactivityareneeded.Thiswillrequireexperimentalandtheoreticallydrivenmanipulationofinterventioncomponentsinwell-poweredandhigh-qualitytrials.Interventionstudiesneedtoprovidecarefuldescriptionsofthehypothesisedcausalprocessesforachievingbehaviourchangeandthespecifictechniquesusedtomodifytheseprocesses.Trialsshouldincludeprocessanalysestoestablishthevalidityorotherwiseofthecausalmodelsproposed.Researchisurgentlyneededtocomparethecost-effectivenessofinterventionswithdifferentproviders,interventionmodesandintensities(usingclearandconsistentconceptualisationsofintensityandattemptingtodisentanglethedifferentelementsofintensitysuchascontacttime,numberofcontactsandcontactfrequency).Thisshouldincludetheevaluationofremotelydeliveredand/orself-delivered(e.g.internet-based)approachesandotherapproachesthatmightprovidehigheffectivenessforlowercost.Researchisalsoneededtoestablishtheimpactoftheinterventionsettingoneffectiveness;tooptimiseinterventionproceduresfordifferentethnic,ageandgendergroups;toestablisheffectivetechniquesforimprovingrecruitmenttointerventions(andtoaddressgenderimbalances);andtoassessthepossibleadverseaffectsofdietaryandphysicalactivityinterventions.ConclusionsInterventionstopromotechangesindietand/orphysicalactivityinadultswithincreasedriskofdiabetesorcardiovasculardiseasearemorelikelytobeeffectiveiftheya)targetbothdietandphysicalactivity,b)involvetheplanneduseofestablishedbehaviourchangetechniques,c)mobilisesocialsupport,andd)haveaclearplanforsupportingmaintenanceofbehaviourchange.Theymayalsobenefitfromprovidingahigherfrequencyortotalnumberofcontacts.Tomaximisetheeffectivenessofinterventionprogrammestopromotechangesindietand/orphysicalactivityfordiabetesprevention,practitionersandcommissioningorganisationsshouldcarefullyconsidertheinclusionoftheabovecomponents. References1.AlbertiKGMM,ZimmetP,ShawJ:InternationalDiabetesFederation:aconsensusonType2diabetesprevention.DiabetMed.2007,24:451-463.10.1111/j.1464-5491.2007.02157.x.CAS  Article  PubMed  GoogleScholar  2.WHO:10factsaboutdiabetes.AccessedJuly2008,[http://www.who.int/features/factfiles/diabetes/01_en.html]3.ErikssonK,LindgärdeF:Preventionoftype2(non-insulindependent)diabetesmellitusbydietandphysicalexercise.The6-yearMalmofeasibilitystudy.Diabetologia.1991,34:891-898.10.1007/BF00400196.CAS  Article  PubMed  GoogleScholar  4.KnowlerWC,Barrett-ConnorE,FowlerSE,HammanRF,LachinJM,WalkerEA,NathanDM:Reductionintheincidenceoftype2diabeteswithlifestyleinterventionormetformin.NEnglJMed.2002,346:393-403.10.1056/NEJMoa012512.CAS  Article  PubMed  GoogleScholar  5.KosakaK,NodaM,KuzuyaT:Preventionoftype2diabetesbylifestyleintervention:aJapanesetrialinIGTmales.DiabetesResClinPract.2005,67:152-162.10.1016/j.diabres.2004.06.010.Article  PubMed  GoogleScholar  6.RamachandranA,SnehalathaC,MaryS,MukeshB,BhaskarAD,VijayV,IndianDiabetesPreventionProgramme:TheIndianDiabetesPreventionProgrammeshowsthatlifestylemodificationandmetforminpreventtype2diabetesinAsianIndiansubjectswithimpairedglucosetolerance(IDPP-1).Diabetologia.2006,49:289-297.10.1007/s00125-005-0097-z.CAS  Article  PubMed  GoogleScholar  7.TuomilehtoJ,LindströmJ,ErikssonJG,ValleTT,HamalainenH,Ilanne-ParikkaP,Keinanen-KiukaanniemiS,LaaksoM,LouherantaA,RastasM,SalminenV,UusitupaM,AunolaS,CepaitisZ,MoltchanovV,HakumakiM,MannelinM,MartikkalaV,SundvallJ:Preventionoftype2diabetesmellitusbychangesinlifestyleamongsubjectswithimpairedglucosetolerance.NEnglJMed.2001,344:1343-1350.10.1056/NEJM200105033441801.CAS  Article  PubMed  GoogleScholar  8.LindstromJ,Ilanne-ParikkaP,PeltonenM,AunolaS,ErikssonJG,HemioK,HamalainenH,HarkonenP,Keinanen-KiukaanniemiS,LaaksoM,LouherantaA,MannelinM,PaturiM,SundvallJ,ValleTT,UusitupaM,TuomilehtoJ:Sustainedreductionintheincidenceoftype2diabetesbylifestyleintervention:follow-upoftheFinnishDiabetesPreventionStudy.Lancet.2006,368:1673-1679.10.1016/S0140-6736(06)69701-8.Article  PubMed  GoogleScholar  9.AmericanDiabetesAssociation:SummaryofRevisionsforthe2009ClinicalPracticeRecommendations.DiabetesCare.2009,32(Suppl1):S3-S5. GoogleScholar  10.CanadianDiabetesAssociationClinicalPracticeGuidelinesExpertCommittee:CanadianDiabetesAssociation2008clinicalpracticeguidelinesforthepreventionandmanagementofdiabetesinCanada.CanadianJournalofDiabetes.2008,32:S1-S201. GoogleScholar  11.HerpertzS,PetrakF,AlbusC,HirschA,KruseJ,KulzerB:Evidence-basedGuidelinesoftheGermanDiabetesAssociation:PsychosocialFactorsandDiabetesMellitus.GermanDiabetesAssociation&TheGermanCollegeforPsychosomaticMedicine.Berlin.Editedby:ScherbaumWA,LandgrafR.2003,1-84. GoogleScholar  12.PaulweberB,ValensiP,LindströmJ,LalicNM,GreavesCJ,McKeeM,Kissimova-SkarbekK,LiatisS,CossonE,SzendroediJ,SheppardKE,CharlesworthK,FeltonAM,HallM,RissannenA,TuomilehtoJ,SchwarzPE,RodenM,fortheWritingGroup,onbehalfoftheIMAGEStudyGroup:AEuropeanEvidence-BasedGuidelineforthePreventionofType2Diabetes.HormMetabRes.2010,42(Suppl1):S3-S36.10.1055/s-0029-1240928.CAS  Article  PubMed  GoogleScholar  13.SaaristoT,PeltonenM,Keinanen-KiukaanniemiS,VanhalaM,SaltevoJ,OksaH,Korpi-HyövältiH,TuomilehtoJ:NationalType2DiabetesProgrammeinFinland:FIN-D2D.IntJCircumpolarHealth.2007,66:101-112.Article  PubMed  GoogleScholar  14.SchwarzP,GruhlU,BornsteinS,LandgrafR,HallM,TuomilehtoJ:TheEuropeanPerspectiveonDiabetesPrevention:developmentandImplementationofAEuropeanGuidelineandtrainingstandardsfordiabetesprevention(IMAGE).DiabVascDisRes.2007,4:353-357.10.3132/dvdr.2007.064.Article  PubMed  GoogleScholar  15.SchwarzP,SchwarzJ,SchuppeniesA,BornsteinS,SchulzeJ:Developmentofadiabetespreventionmanagementprogramforclinicalpractice.PublicHealthRep.2007,122:258-263.PubMed  PubMedCentral  GoogleScholar  16.AckermannRT,MarreroDG:AdaptingtheDiabetesPreventionProgramlifestyleinterventionfordeliveryinthecommunity:TheYMCAModel.DiabetesEduc.2007,33:69-78.10.1177/0145721706297743.Article  PubMed  GoogleScholar  17.DelahantyLM,NathanDM:ImplicationsoftheDiabetesPreventionProgramandtheLookAHEADclinicaltrialsforlifestyleintervention.AmericanDieteticAssociation.2008,108(Suppl1):S66-S72.10.1016/j.jada.2008.01.026.Article  GoogleScholar  18.LaatikainenT,DunbarJ,ChapmanA,KilkkinenA,VartiainenE,HeistaroS,PhilpotB,AbsetzP,BunkerS,O'NeilA,ReddyP,BestJ,JanusE:PreventionofType2diabetesbylifestyleinterventioninanAustralianprimaryhealthcaresetting:GreaterGreenTriangle(GGT)DiabetesPreventionProject.BMCPublicHealth.2007,7:249-10.1186/1471-2458-7-249.Article  PubMed  PubMedCentral  GoogleScholar  19.QiaoQ,PangZ,GaoW,WangS,DongY,ZhangL,NanH,RenJ:Alarge-scalediabetespreventionprograminreal-lifesettingsinQingdaoofChina(2006-2012).PrimCareDiabetes.2010,4:99-103.10.1016/j.pcd.2010.04.003.Article  PubMed  GoogleScholar  20.GilliesCL,LambertPC,AbramsKR,SuttonAJ,CooperNJ,HsuRT,DaviesMJ,KhuntiK:Differentstrategiesforscreeningandpreventionoftype2diabetesinadults:costeffectivenessanalysis.BMJ.2008,336:1180-1185.10.1136/bmj.39545.585289.25.Article  PubMed  PubMedCentral  GoogleScholar  21.HermanWH,HoergerTJ,BrandleM,HicksK,SorensenS,ZhangP,HammanRF,AckermannRT,EngelgauMM,RatnerRE,DiabetesPreventionProgram:Thecost-effectivenessoflifestylemodificationormetformininpreventingtype2diabetesinadultswithimpairedglucosetolerance.AnnInternMed.2005,142:323-332.Article  PubMed  PubMedCentral  GoogleScholar  22.LindgrenP,LindströmJ,TuomilehtoJ,UusitupaM,PeltonenM,JönssonB,DeFaireU,HelléniusML:Lifestyleinterventiontopreventdiabetesinmenandwomenwithimpairedglucosetoleranceiscost-effective.IntJTechnolAssessHealthCare.2007,23:177-183.10.1017/S0266462307070286.Article  PubMed  GoogleScholar  23.AhmadLA,CrandallJP:Type2diabetesprevention:Areview.ClinicalDiabetes.2010,28:53-59.10.2337/diaclin.28.2.53.Article  GoogleScholar  24.SchwarzP,LinJ,LindstromJ,BergmannA,GruhlU,SaaristoT,TuomilehtoJ:Howshouldtheclinicianmosteffectivelypreventtype2diabetesintheobesepersonathighrisk?.CurrentDiabetesReports.2008,7:353-362.10.1007/s11892-007-0058-7.Article  GoogleScholar  25.DepartmentofHealth:Puttingpreventionfirst-vascularchecks:riskassessmentandmanagementLondon.2009,27- GoogleScholar  26.JanzKF:Physicalactivityinepidemiology:movingfromquestionnairetoobjectivemeasurement.BrJSportsMed.2006,40:191-192.10.1136/bjsm.2005.023036.CAS  Article  PubMed  PubMedCentral  GoogleScholar  27.ShephardRJ:Limitstothemeasurementofhabitualphysicalactivitybyquestionnaires.BrJSportsMed.2003,37:197-206.10.1136/bjsm.37.3.197.CAS  Article  PubMed  PubMedCentral  GoogleScholar  28.OxmanAD,GuyattGH:Validationofanindexofthequalityofreviewarticles.JClinEpidemiol.1991,44:1271-1278.10.1016/0895-4356(91)90160-B.CAS  Article  PubMed  GoogleScholar  29.ScottishIntercollegiateGuidelinesNetwork(SIGN):ManagementofDiabetes:Anationalclinicalguideline.SIGNPublicationNumber55.2001,Edinburgh:SIGN GoogleScholar  30.AshworthNL,ChadKE,HarrisonEL,ReederBA,MarshallSC:Homeversuscenterbasedphysicalactivityprogramsinolderadults.CochraneDatabaseofSystematicReviews.2005,CD004017- GoogleScholar  31.AvenellA,BroomJ,BrownTJ,PoobalanA,AucottL,StearnsSC,SmithWCS,JungRT,CampbellMK,GrantAM:Systematicreviewofthelong-termeffectsandeconomicconsequencesoftreatmentsforobesityandimplicationsforhealthimprovement.HealthTechnolAssess.2004,8(21):1-194.Article  GoogleScholar  32.BoschC,AbbaK,PrictorM,GarnerP:Contractsbetweenpatientsandhealthcarepractitionersforimprovingpatients'adherencetotreatment,preventionandhealthpromotionactivities.CochraneDatabaseofSystematicReviews.2007,CD004808- GoogleScholar  33.BravataDM,Smith-SpanglerC,SundaramV,GiengerAL,LinN,LewisR,StaveCD,OlkinI,SirardJR:Usingpedometerstoincreasephysicalactivityandimprovehealth:Asystematicreview.JAMA.2007,298:2296-2304.10.1001/jama.298.19.2296.CAS  Article  PubMed  GoogleScholar  34.BrunnerEJ,ReesK,WardK,BurkeM,ThorogoodM:Dietaryadviceforreducingcardiovascularrisk(Review).CochraneDatabaseofSystematicReviews.2007,CD002128- GoogleScholar  35.BurkeBL,ArkowitzH,MencholaM:Theefficacyofmotivationalinterviewing:Ameta-analysisofcontrolledclinicaltrials.JConsultClinPsychol.2003,71:843-861.10.1037/0022-006X.71.5.843.Article  PubMed  GoogleScholar  36.CurioniCC,LourentoPM:Long-termweightlossafterdietandexercise:asystematicreview.IntJObes.2005,29:1168-1174.10.1038/sj.ijo.0803015.CAS  Article  GoogleScholar  37.DansingerM,TatsioniA,WongJ,ChungM,BalkE:Meta-analysis:theeffectofdietarycounselingforweightloss.AnnInternMed.2007,147:41-50.Article  PubMed  GoogleScholar  38.DombrowskiSU,SniehottaFF,AvenellA,JohnstonM,MacLennanG,Araujo-SoaresV:Effectsofspecificbehaviourchangetechniquesonbehaviourandweightchange:Asystematicreview.HealthPsychologyReview.39.DouketisJD,MacieC,ThabaneL,WilliamsonDF:Systematicreviewoflong-termweightlossstudiesinobeseadults:Clinicalsignificanceandapplicabilitytoclinicalpractice.IntJObes.2005,29:1153-1167.10.1038/sj.ijo.0802982.CAS  Article  GoogleScholar  40.EakinEG,GlasgowRE,RileyKM:Reviewofprimarycare-basedphysicalactivityinterventionstudies-Effectivenessandimplicationsforpracticeandfutureresearch.JFamPract.2000,49:158-168.CAS  PubMed  GoogleScholar  41.FosterC,HillsdonM,ThorogoodM:Interventionsforpromotingphysicalactivity.CochraneDatabaseofSystematicReviews.2005,CD003180- GoogleScholar  42.GalaniC,SchneiderH:Preventionandtreatmentofobesitywithlifestyleinterventions:reviewandmeta-analysis.IntJPublicHealth.2007,52:348-359.10.1007/s00038-007-7015-8.Article  PubMed  GoogleScholar  43.GilliesCL,AbramsKR,LambertPC,CooperNJ,SuttonAJ,HsuRT,KhuntiK:Pharmacologicalandlifestyleinterventionstopreventordelaytype2diabetesinpeoplewithimpairedglucosetolerance:systematicreviewandmeta-analysis.BMJ.2007,334:299-308.10.1136/bmj.39063.689375.55.Article  PubMed  PubMedCentral  GoogleScholar  44.HalcombE,MoujalliS,GriffithsR,DavidsonP:Effectivenessofgeneralpracticenurseinterventionsincardiacriskfactorreductionamongadults.InternationalJournalofEvidence-BasedHealthcare.2007,5:269-295.10.1111/j.1479-6988.2007.00070.x.PubMed  GoogleScholar  45.KahnEB,RamseyLT,BrownsonRC,HeathGW,HowzeEH,PowellKE,StoneEJ,RajabMW,CorsoP,BrissPA:Theeffectivenessofinterventionstoincreasephysicalactivity-Asystematicreview.AmJPrevMed.2002,22:73-108.10.1016/S0749-3797(02)00434-8.Article  PubMed  GoogleScholar  46.McTigueKM,HarrisR,HemphillB,LuxL,SuttonS,BuntonAJ,LohrK:Screeningandinterventionsforobesityinadults:SummaryoftheevidencefortheU.S.PreventiveServicesTaskForce.AnnInternMed.2003,139:933-949.Article  PubMed  GoogleScholar  47.McTigueK,HessR,ZiourasJ:Obesityinolderadults:asystematicreviewoftheevidencefordiagnosisandtreatment.Obesity.2006,14:1485-1497.10.1038/oby.2006.171.Article  PubMed  GoogleScholar  48.MichieS,AbrahamC,WhittingtonC,McAteerJ,GuptaS:Effectivetechniquesinhealthyeatingandphysicalactivityinterventions:Ameta-regression.HealthPsychol.2009,28:690-701.10.1037/a0016136.Article  PubMed  GoogleScholar  49.MurphyM,NevillA,MurtaghE,HolderR:Theeffectofwalkingonfitness,fatnessandrestingbloodpressure:ameta-analysisofrandomised,controlledtrials.PrevMed.2007,44:377-385.10.1016/j.ypmed.2006.12.008.Article  PubMed  GoogleScholar  50.NorrisSL,ZhangX,AvenellA,GreggE,SchmidtCH,LauJ:Long-termnon-pharmacologicalweightlossinterventionsforadultswithprediabetes.CochraneDatabaseofSystematicReviews.2007,CD005270- GoogleScholar  51.OgilvieD,FosterC,RothnieH,CavillN,HamiltonV,FitzsimonsC,MutrieN:Interventionstopromotewalking:Systematicreview.BMJ.2007,334:1-11.10.1136/bmj.39198.722720.BE.Article  GoogleScholar  52.RichardsonC,NewtonT,AbrahamJ,SenA,JimboM,SwartzA:Ameta-analysisofpedometer-basedwalkinginterventionsandweightloss.AnnFamMed.2008,6:69-77.10.1370/afm.761.Article  PubMed  PubMedCentral  GoogleScholar  53.RubakS,SandbaekA,LauritzenT,ChristensenB:Motivationalinterviewing:Asystematicreviewandmeta-analysis.BrJGenPract.2005,55:305-312.PubMed  PubMedCentral  GoogleScholar  54.ShawK,RourkeP,DelM,KenardyJ:Psychologicalinterventionsforoverweightorobesity.CochraneDatabaseofSystematicReviews.2005,CD003818- GoogleScholar  55.ShawK,GennatH,RourkeP,DelM:Exerciseforoverweightorobesity.CochraneDatabaseofSystematicReviews.2006,CD003817- GoogleScholar  56.ThompsonRL,SummerbellCD,HooperL,HigginsJPT,LittlePS,TalbotD,EbrahimS:Dietaryadvicegivenbyadietitianversusotherhealthprofessionalorself-helpresourcestoreducebloodcholesterol.CochraneDatabaseofSystematicReviews.2003,CD001366- GoogleScholar  57.TsaiAG,WaddenTA:Systematicreview:AnevaluationofmajorcommercialweightlossprogramsintheUnitedStates.AnnInternMed.2005,142:56-66.Article  PubMed  GoogleScholar  58.WhitlockE,WilliamsS:Theprimarypreventionofheartdiseaseinwomenthroughhealthbehaviorchangepromotioninprimarycare.WomensHealthIssues.2003,13:122-141.10.1016/S1049-3867(03)00036-7.Article  PubMed  GoogleScholar  59.WilliamsN,HendryM,FranceB,LewisR,WilkinsonC:Effectivenessofexercise-referralschemestopromotephysicalactivityinadults:systematicreview.BrJGenPract.2007,57:979-986.10.3399/096016407782604866.Article  PubMed  PubMedCentral  GoogleScholar  60.BanduraA:Socialfoundationsofthoughtandaction:Asocialcognitivetheory.1986,EnglewoodCliffs,NJ:Prentice-Hall GoogleScholar  61.AzjenI:Thetheoryofplannedbehavior.OrganizationalBehaviorandHumanDecisionProcesses.1991,50:179-211.10.1016/0749-5978(91)90020-T.Article  GoogleScholar  62.AbrahamC,MichieS:Ataxonomyofbehaviourchangetechniquesusedininterventions.HealthPsychol.2008,27:379-387.10.1037/0278-6133.27.3.379.Article  PubMed  GoogleScholar  63.MillerW,RollnickS:Motivationalinterviewing:Preparingpeopleforchange.2002,NewYork:GuildfordPress,2 GoogleScholar  64.DeRidderD,DeWitJ:Self-regulationinhealthbehavior.2006,Chichester,UK:WileyBook  GoogleScholar  65.BartholomewLK,ParcelGS,KokG,GottliebNH:Planninghealthpromotionprograms.Aninterventionmappingapproach.2006,SanFrancisco:JohnWiley&Sons GoogleScholar  66.GreenLW,KreuterM:Healthprogramplanning:Aneducationalandecologicalapproach.2005,NewYork:McGraw-Hill,4 GoogleScholar  67.NationalInstituteforHealthandClinicalExcellence:Obesityguidanceontheprevention,identification,assessmentandmanagementofoverweightandobesityinadultsandchildren.London.2006 GoogleScholar  68.ArtinianNT,FletcherGF,MozaffarianD,Kris-EthertonP,VanHornL,LichtensteinAH,KumanyikaS,KrausWE,FlegJL,RedekerNS,MeiningerJC,BanksJ,Stuart-ShorEM,FletcherBJ,MillerTD,HughesS,BraunLT,KopinLA,BerraK,HaymanLL,EwingLJ,AdesPA,DurstineJL,Houston-MillerN,BurkeLE,onbehalfoftheAmericanHeartAssociationPreventionCommitteeoftheCouncilonCardiovascularNursing:Interventionstopromotephysicalactivityanddietarylifestylechangesforcardiovascularriskfactorreductioninadults:AscientificstatementfromtheAmericanHeartAssociation.Circulation.2010,122(4):406-441.10.1161/CIR.0b013e3181e8edf1.Article  PubMed  GoogleScholar  69.AmericanAssociationofDiabetesEducators:AADEguidelinesforthepracticeofdiabetesself-managementeducationandtrainingChicago.2009 GoogleScholar  70.HardemanW,GriffinS,JohnstonM,KinmonthAL,WarehamNJ:Interventionstopreventweightgain:asystematicreviewofpsychologicalmodelsandbehaviourchangemethods.IntJObes.2000,24:131-143.10.1038/sj.ijo.0801100.CAS  Article  GoogleScholar  71.SmithJ,DenoS,PronkN,ReinhardtL:Behaviouraltherapyprogramsforweightlossinadults:ICSITechnologyAssessmentReport.2005,Minnesota:InstituteforClinicalSystemsImprovement GoogleScholar  Pre-publicationhistoryThepre-publicationhistoryforthispapercanbeaccessedhere:http://www.biomedcentral.com/1471-2458/11/119/prepub DownloadreferencesAcknowledgementsFunding:EuropeanCommissionCommunityActionProgrammeforPublicHealth(Project2006309);SouthWestGeneralPracticeTrust;NationalInstituteforHealthResearch.Theviewsexpressedinthispublicationarethoseoftheauthorsandnotnecessarilythoseofthefundingbodies.WewouldalsoliketothankRodTaylor,TiffanyMoxham,membersoftheSocietyforAcademicPrimaryCare(SAPC),thePsychosocialAspectsofDiabetes(PSAD)StudyGroup,DenisedeRidder,StantonNewmanandRobertMKaplan,whoprovidedusefulcommentary.ThisreviewwasconductedtoinformthebehaviourchangesectionofaEuropeanUnionguidelineonthepreventionofdiabeteshttp://www.image-project.eu.AuthorinformationAffiliationsUniversityofExeter,PeninsulaMedicalSchool,SmeallBuilding,StLuke'sCampus,MagdalenRoad,Exeter,EX12LU,UKColinJGreaves & KateESheppardUniversityofSussex,SchoolofPsychology,PevenseyBuilding,Falmer,BN19QG,UKCharlesAbrahamUniversityofCambridge,GeneralPracticeandPrimaryCareResearchUnit,16ColwynClose,Cambridge,CB43NU,UKWendyHardemanHeinrich-HeineUniversity,InstituteforClinicalDiabetology,GermanDiabetesCentreandDepartmentofMetabolicDiseases,Auf'mHennekamp65,40225,Düsseldorf,GermanyMichaelRodenUniversityofPlymouth,PeninsulaMedicalSchool,SmeallBuilding,StLuke'sCampus,MagdalenRoad,Exeter,EX12LU,UKPhilipHEvansTechnicalUniversityofDresden(CarlGustavCarusMedicalFaculty),MedizinischeKlinikIII,Fetscherstraße74,Dresden,D-01307,GermanyPeterSchwarzAuthorsColinJGreavesViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarKateESheppardViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarCharlesAbrahamViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarWendyHardemanViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarMichaelRodenViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarPhilipHEvansViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarPeterSchwarzViewauthorpublicationsYoucanalsosearchforthisauthorin PubMed GoogleScholarConsortiaTheIMAGEStudyGroupCorrespondingauthorCorrespondenceto ColinJGreaves.AdditionalinformationCompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.Authors'contributionsCGconceivedandcoordinatedthestudy.KSandCGconductedliteraturesearches,dataextraction,reviewselection,qualityratingandevidencegradinganddraftedthemanuscript.CA,WH,MR,PEandPScontributedtothedesignofthestudyandinterpretationoftheresults.Allauthorsreadandapprovedthefinalmanuscript.Electronicsupplementarymaterial 12889_2010_2863_MOESM1_ESM.DOCAdditionalfile1:TableS1:SearchStrategy.TableS2(andexplanatorytext):OQAQ:Qualityassessmenttoolforsystematicreviewsandmeta-analyses.TableS3(andexplanatorytext):EvidenceGradingSystem.TableS4:CharacteristicsofIncludedReviews.TableS5:Excludedpapers.TableS6:OQAQscores.(DOC328KB)12889_2010_2863_MOESM2_ESM.DOCAdditionalfile2:TablesS7-14:Datafromanalysesof:S7)InterventionEffectiveness;S8)Theoreticalbasis;S9)Behaviourchangetechniques;S10)Modeofdelivery;S11)Interventionprovider;S12)Interventionintensity;S13)Interventionpopulation;S14)Interventionsetting.(DOC279KB)12889_2010_2863_MOESM3_ESM.DOCAdditionalfile3:PRISMA(PreferredReportingItemsforSystematicReviewsandMeta-Analyses)2009Checklist.(DOC68KB)Authors’originalsubmittedfilesforimagesBelowarethelinkstotheauthors’originalsubmittedfilesforimages. Authors’originalfileforfigure1Rightsandpermissions ThisarticleispublishedunderlicensetoBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited. ReprintsandPermissionsAboutthisarticleCitethisarticleGreaves,C.J.,Sheppard,K.E.,Abraham,C.etal.Systematicreviewofreviewsofinterventioncomponentsassociatedwithincreasedeffectivenessindietaryandphysicalactivityinterventions. BMCPublicHealth11,119(2011).https://doi.org/10.1186/1471-2458-11-119DownloadcitationReceived:20July2010Accepted:18February2011Published:18February2011DOI:https://doi.org/10.1186/1471-2458-11-119SharethisarticleAnyoneyousharethefollowinglinkwithwillbeabletoreadthiscontent:GetshareablelinkSorry,ashareablelinkisnotcurrentlyavailableforthisarticle.Copytoclipboard ProvidedbytheSpringerNatureSharedItcontent-sharinginitiative KeywordsPhysicalActivityPhysicalActivityInterventionMediumQualityDiabetesPreventionProgrammeBehaviourChangeTechnique DownloadPDF Advertisement BMCPublicHealth ISSN:1471-2458 Contactus Submissionenquiries:[email protected] Generalenquiries:[email protected]