Frailty Phenotype Prevalence in Community-Dwelling Older ...

The low physical activity criterion of the frailty phenotype was determined by using five different questionnaires and an accelerometer, and ... Journals WhyPublishWithUs? EditorialPolicies AuthorGuidelines PeerReviewGuidelines OpenOutlook COVID-19 Podcasts Blog Reprints SubmitNewManuscript Menu About Contact Sustainability Awards PressCenter Testimonials Blog FavoredAuthorProgram Permissions Pre-Submission Reprints Login openaccesstoscientificandmedicalresearch Advancedsearch HomeJournalsWhypublishwithus?EditorialPoliciesAuthorInformationPeerReviewGuidelinesOpenOutlookCOVID-19Podcasts Usage 36334 Monthlydownloads/views AverageArticleStatistics 20Days * Fromsubmissiontofirsteditorialdecision. 12Days * Fromeditorialacceptancetopublication. *Businessdays(Mon-Fri)RejectionRate 79% Theabovepercentageofmanuscriptshavebeenrejectedinthelast12months. 71935 PapersPublished SubmitNewManuscript Logintoviewexistingmanuscriptstatus SignupforJournalalerts AboutDovePress Openaccesspeer-reviewedscientificandmedicaljournals.Learnmore OpenAccess DoveMedicalPressisamemberoftheOAI.Learnmore Reprints Bulkreprintsforthepharmaceuticalindustry.Learnmore FavoredAuthors Weofferrealbenefitstoourauthors,includingfast-trackprocessingofpapers.Learnmore PromotionalArticleMonitoring RegisteryourspecificdetailsandspecificdrugsofinterestandwewillmatchtheinformationyouprovidetoarticlesfromourextensivedatabaseandemailPDFcopiestoyoupromptly.Learnmore SocialMedia       BacktoJournals»ClinicalInterventionsinAging»Volume15 OriginalResearch FrailtyPhenotypePrevalenceinCommunity-DwellingOlderAdultsAccordingtoPhysicalActivityAssessmentMethod            FulltextMetricsGetPermissionCitethisarticle AuthorsZillerC ,BraunT ,ThielC  Received12November2019 Acceptedforpublication11January2020 Published9March2020 Volume2020:15Pages343—355 DOIhttps://doi.org/10.2147/CIA.S238204 CheckedforplagiarismYes ReviewbySingleanonymouspeerreview Peerreviewercomments2 Editorwhoapprovedpublication: DrRichardWalker DownloadArticle[PDF]  CarinaZiller,1TobiasBraun,1ChristianThiel1,21DepartmentofAppliedHealthSciences,DivisionofPhysiotherapy,HochschulefürGesundheit(UniversityofAppliedSciences),Bochum,Germany;2FacultyofSportsScience,TrainingandExerciseScience,RuhrUniversityBochum,Bochum,GermanyCorrespondence:CarinaZillerDepartmentofAppliedHealthSciences,DivisionofPhysiotherapy,HochschulefürGesundheit(UniversityofAppliedSciences),Gesundheitscampus6-8,Bochum44801,GermanyTel+4923477727628Fax+4923477727828Email[email protected]Introduction:StudieshavedescribedvaryingprevalencesoffrailtyasdeterminedbyFried’sfrailtyphenotype.Comparabilitymaybelimitedduetofrailtyphenotypemodifications,especiallythelowphysicalactivitycriterion.Purpose:Thisstudyaimedtodeterminethevariabilityoffrailtyphenotypeprevalenceaccordingtothephysicalactivityassessmentmethod.PatientsandMethods:Inacross-sectionalanalysis,frailtyphenotypeprevalencewasassessedincommunity-dwellingolderadults.Thelowphysicalactivitycriterionofthefrailtyphenotypewasdeterminedbyusingfivedifferentquestionnairesandanaccelerometer,andthreedifferentcut-pointmodels.Results:In47participants,frailtyphenotypeprevalencevariedbetween14.9%and31.9%,dependingonthemodelusedtoassessphysicalactivity.Conclusion:Themethodofphysicalactivityassessmentandthechoiceofcut-pointsforlowphysicalactivityconsiderablyimpactfrailtyphenotypeprevalence.Moreeffortstostandardizeandadheretothelowphysicalactivitycriterionseemwarranted.Thecalculationofcorrectionfactorsbetweencommonlyusedsetsoflowphysicalactivitycriteriamightallowbettercomparisonsofpublishedprevalencerates.Keywords:physicalfrailty,physicalactivityassessments,olderadults Introduction Frailtyhasbecomeanimportantgeriatricconcept.Itdescribesanolderadult’sincreasedriskfornegativehealthoutcomessuchasfalls,physicaldisability,hospitalizationandmortality.1 Frailtyprevalenceincommunity-dwellingolderadultsvariesbetween4%and59%.2Thishighvariabilitymayreflecttruedifferencesinvariouspopulations,butitmayalsobeduetothemethodofdiagnosisoffrailty.3Abroadrangeoffrailtyinstrumentsforscreeninganddiagnosisexists,includingself-reportedquestionnairesand/orperformancetestswithimplicationsforprevalencevalues,classificationandpredictiveabilities.4,5Researchersandcliniciansarechallengedtodecidewhichinstrumenttouse.5 Twofrailtyinstrumentsareespeciallycommon:TheFrailtyIndex(FI),whichcountsaccumulatedhealthdeficitsandcoversabroadsetoffrailtydomainsasacontinuum,6andthefrailtyphenotype(FP),whichclassifiespeopleintothreecategories(fit,pre-frail,andfrail)basedonfivecriteria(unintendedweightloss,self-reportedexhaustion,lowphysicalactivity(LPA),slowwalkingspeed,andweakgripstrength).7 InareviewincludingstudiesthathaveinvestigatedfrailtyprevalencebasedonFried’smodel,prevalencevariedbetween4.9%and27.3%,avariationof22.4percentagepoints.8AccordingtoFPdatareportedbyCollardetal,2prevalencevaluesoftheincludedstudiesvariedby13.0percentagepoints(4.0%to17.0%).Thisvariationinfrailtyprevalencemightbeexplainedbytruevariation,country-specificdifferences,aswellasmodificationsoftheFPdescribedintheliterature.2,3 TheFPhasoftenbeenestablishedbasedondifferentmeasurementmethods,complicatingcomparativeanalysis.9Accordingtoasystematicreview,unintendedweightlossandLPAwerethecriteriathathadbeenalteredthemostfrequently.3Originally,inordertodefineaLPAlevel,Friedandcolleagues7includeditemsoftheMinnesotaLeisureTimePhysicalActivityQuestionnaire(MLTPAQ)10andcut-pointvalueswerederivedfromthelowest20thpercentileoftheCardiovascularHealthStudy(CHS)population.7TheappropriatenessoftheMLTPAQforthegeneralgeriatricpopulationhasbeenquestioned,asitfocusesonmoderatetovigorousphysicalactivities(MVPA)notrelatedtoolderadultsandshowsflooreffects.11,12Later,variousmethods,includingdifferentquestionnaires,singlequestionsorobjectivemeasuressuchasanactivitymonitor(accelerometer)havebeenusedtoassesstheLPAcriterion.3 InpublishedstudiesusingtheFP,mostcut-pointvalueswereeitherbasedontheonesproposedbyFriedetal7(population-independent)orderivedinrelationtothevariabilitywithinthestudysample,egbyusingthelowestquintileoftherespectivestudypopulation.3Eventhoughitseemsreasonabletoassumethatthediversityofexistingphysicalactivity(PA)measurementsandcut-pointvaluesmayinfluencefrailtyprevalence,themagnitudeofitspotentialinfluencehasneverbeenquantified. Thus,theaimofthisstudywastoanalyzethevarianceinprevalenceofphysicalfrailtyincommunity-dwellingolderadultsaccordingtotheFriedmodelbyusingdifferentmethodsandmodelsformeasuringPAfortheLPAcriterion.WeassumedthatmodificationsoftheLPAcriterionwouldshifttheprevalenceofphysicalfrailtybymorethanhalfofthe17.7percentagepoints,themeanofthetwomeanvaluesreportedinthementionedreviews.2,8 MaterialsandMethods StudyDesign ThecurrentanalysisisanextensionoftwodifferentstudiesconductedbetweenOctober2015andFebruary2018,inwhichadditionaldataonPAhadbeencollected.Onestudy(studyA)hadestablishedtheprevalenceoffrailtyinasampleofolderpeopleattendingoutpatientphysiotherapyservices,13andthesecondstudy(studyB),apilotrandomizedcontrolledtrial,hadassessedtheeffectofaphysiotherapy-ledhome-basedinterventioninolderadultswithfrailty.14ThepresentstudywasapprovedasanamendmenttothetwooriginalethicalapprovalsbytheEthicalReviewBoardoftheGermanConfederationforPhysiotherapy.ThisstudywasconductedinaccordancewiththeDeclarationofHelsinkiinitsmostrecentversion(64thWMAGeneralAssembly,Fortaleza,Brazil,October2013).Allparticipantsprovidedwritteninformedconsentforstudyparticipation.ReportingfollowstheStrengtheningtheReportingofObservationalStudiesinEpidemiologystatement.15 RecruitmentStrategy Participantswererecruitedinthecontextofthetwomentionedstudies(Figure1).13,14Recruitmentproceduresofthosetwostudieshavebeenreportedindetailelsewhere.13,14Shortly,instudyA(theprevalencestudy),participantswererecruitedinoutpatientphysiotherapyclinicsinBochum,Germany.AssessmentsinstudyAattheseoutpatientphysiotherapyclinicsincludeddifferentmeasurementinstrumentsoffrailty,afterwhichparticipantswereinstantlyinvitedtotakepartinthepresentstudy.InstudyB(theRCT),olderadultswererecruitedinsocialfacilitiessuchasseniorcitizens’offices,bynewspaperadvertisements,aswellasusingcareservicesandtheuniversitynetwork.RespondentswerescreenedforpredefinedinclusionandexclusioncriteriaofstudyB.Iftheydidnotfulfillthosecriteria,and/oriftheylackedthetimeorwillingnesstoparticipateinstudyB,theywerelaterinvitedbytelephonetoparticipateinthepresentstudy. Figure1Flow-chartofrecruitmentprocess. Setting DatacollectionwasconductedinthecityofBochumintheWesternpartofGermany.Measurementsoftheparticipantsrecruitedfromtheprevalencestudy(studyA)wereeitherperformedintheoutpatientphysiotherapyclinics,orathome.MeasurementsoftherespondentsnotparticipatingintheRCT(studyB)wereperformedattheirhomes. Participants Tobeeligible,olderadultshadtobecommunity-dwellingandatleast65yearsold.Exclusioncriteriawere:1)theinabilitytobothunderstandthestudyinformationandgiveinformedconsent(egduetoseverecognitiveorvisualimpairment);2)acuteconditionswhichmightaffectphysicalactivitybehavior(egacuteillness,infection,injury);3)theinabilitytowalkatleast10m(withawalkingaid,ifrequired);4)insufficientGermanlanguageskills;and5)lackoftimetoparticipateinthestudy.Chronicmorbiditieswerenotanexclusioncriterion,sincetheyareatypicalfeatureoffrailty,andthemeasurementofphysicalactivityisnotassociatedwithanyrisksforhealth. Procedures Atthefirstappointment,anaccelerometer(ActiGraphwGT3x-BT)wasgivenandexplainedtoeachparticipantasdescribedbelow.Atasecondappointment,approximatelyoneweeklater,theaccelerometerwascollected,frailtyassessmentswereconducted,andPAquestionnaireswerecollectedinrandomorderintheformofaninterview. Variables PhysicalfrailtywasmeasuredbyusingseveralversionsoftheFP.7OverallfrailtywasmeasuredusingaFI.6 DataMeasurementFrailty AtranslatedGermanversion16oftheFPwiththefollowingfivecriteriawasusedtoassessparticipantsforphysicalfrailty:1)unintentionalweightloss(≥4.5kginthelastyear);2)exhaustion(twoself-reportedquestionsaboutthepastweekfromtheCentreforEpidemiologicalStudies-Depressionscale);3)slowgaitspeed(walkingtimeoveradistanceof4.57m[15ft]);4)weakness(gripstrengthofthedominanthandusingaJAMARdynamometer(PattersonMedical,Model5030J1)followingmeasurementproceduresrecommendedbyRobertsetal17);and5)lowphysicalactivity(usingdifferentassessmentsandcut-points;seesection“PAassessments”andTable1).Morethanthreepositivecriteriacharacterizeapersonasfrail,oneortwopositiveitemsaspre-frailandnoneasnon-frailorfit.7Fried’soriginalphenotypeversionwasdefinedastheprimaryphenotypeinthisstudy.Additionally,eighteenmodifiedversionswerecreatedbyincludingdifferentPAassessmentsaswellasusingdifferentcut-pointmodels(Table1). Table1DetailedDescriptionoftheModelsUsedtoDetermineLowPhysicalActivity TheFIwasderivedfromadeficitaccumulationassessmentbasedontheprocedureproposedbySearleetal.31TheslightlymodifiedFIusedinthisstudyconsistedof41items,including39questionscoveringphysical,psychological,socialandcognitivedomains,comorbidities,andtwophysicalperformancemeasurements(gripstrengthandgaitspeed,asdescribedabove).13Cut-pointsforpre-frailtyandfrailtywere0.08and0.25,respectively.32 PAAssessments TheinstrumentstoassessPA(accelerometer,questionnaires)andthecut-pointvaluemodelswereselectedbasedonthemodificationsoftheLPAcriterionlistedinasystematicreviewbyTheouetal,3andfurtherbasedontheresultsofanon-systematicliteraturesearch.Tobeincluded,questionnaireshadtobeaccessibleinGermanlanguage. Intheprimaryphenotype,PAwasassessedbythe18-itemshortversionoftheMLTPAQ.7,10Alternatively,theGermanPhysicalActivityQuestionnaire50+(GPAQ50+),33thelongGermanversionoftheInternationalPhysicalActivityQuestionnaire(IPAQ),34theStanfordSevenDayPhysicalActivityRecall(7D-PAR),35recommendedforfrailpeople,3andaGermanversion36ofthePhysicalActivityScalefortheElderly(PASE)37wereused(Table1). Physicalactivityenergyexpenditure(PAEE)wascalculatedfromtotalenergyexpenditure(TEE)followingCalabroetal(PAEE=TEE∗0.9–RMR).38Toestimaterestingmetabolicrate(RMR),theequationsofMifflinetalwereused.39 ForthedirectmeasurementofPA,participantswereinstructedtowearanaccelerometer(ActiGraphwGT3X-BT)ontherighthipforsevenconsecutivedaysduringwakinghours.Publishedrecommendationsforaccelerometerdatacollection(100Hz,10-secondepochs),dataprocessing(ActiLifeSoftware6,ActiGraph,LLC)andreportingwerefollowed.40Non-wearingperiods,definedas90minofcontinuouszerocountstimes,wereremovedbeforeanalysis.41Fourtosevendayswitheachatleast6 hoursofweartimewereincluded.EnergyexpenditurewascalculatedbytheFreedsonCombinationAlgorithmforuniaxialcountsperminute(cpm).27Twodifferentcut-pointswereusedtodeterminetheamountofMVPA,ageneralmodelbyFreedsonetal(MVPA≥1952cpm)27andaspecificmodelforolderadultsbyCopelandandEsliger(MVPA≥1041cpm).28AnyactivityabovethethresholdswascountedasMVPAirrespectiveoftheboutlength.Sedentarytimewasdefinedbyasignal<100cpm.42 Cut-PointModelsforLPA TocategorizepersonsfulfillingtheLPAcriterion,threecut-pointswereconsidered:A)usingthecut-pointsdescribedbyFried,<270kcal/weekforwomenand<383kcal/weekformen;7B)usingspecificvaluesoftherespectivestudy/questionnaire;andC)notreachingpublishedhealthrecommendationsforMVPAproposedbytheWorldHealthOrganization(WHO).18Ifnotspecifiedotherwiseintherespectivequestionnaire,MVPAwasassumedforallactivityitemsequalorgreaterthan3METs.43 Bias Relevantsourcesofbiasweretherecruitmentandselectionprocess,andthechoiceofquestionnaires.Wetriedtoaddressthislatterissuebyselectingwell-establishedmethodswhichareappropriateforcommunity-dwellingolderadultslivinginGermany. SampleSize Wetriedtoincludeasmanyparticipantsaspossiblebutaimedataminimumsamplesizeof30participants,asstatedinourapplicationapprovedbytheethicalreviewboard.Thisminimumsamplesizeisbasedonthe“adequate”samplesizeforcriterionvalidity(“30–50patientsinthesmallestgroup”)proposedbytheCOSMINgroup.44ThenumberofparticipantswaslimitedtothenumberofpersonsinterestedinparticipatinginthetwomotherstudiesAandB. QuantitativeVariables PAdatawasexportedandfurtherprocessedwithExcel2013(MicrosoftCorporation,Redmond,USA)tocalculateenergyexpenditure(kcal/week),timeatMVPA(min/week),dailysteps(steps/day)andsedentarytime(hours/day). Foreachofthe19phenotypes,relativeandabsoluteprevalenceoffrailty,pre-frailty,andnon-frailtyarepresented. Togradethemagnitudeandimpactoftheresultingvariationofprevalencevalues,theresultsoftworeviewsweretakenintoaccount.2,8Adeviationof17.7percentagepoints,whichisthemeanofthevariationinprevalencesreportedinthetworeviews,wouldberegardedasdenotingaveryhighrelevance,morethanhalfofit(>8.85percentagepoints)aconsiderablerelevance,and8.85orlessasmalltomoderaterelevance. StatisticalMethods FPandFIprevalenceand95%confidenceintervalwerecalculated.45PrevalencesbasedonmodifiedFPversionsweredescriptivelycomparedtoeachother,andtotheFIbasedprevalence.Interval-basedvariablesweretestedfornormality(Shapiro–Wilktest).Fortheprimaryphenotype,baselinecharacteristicsaswellasPAdataofthefrail,pre-frailandnon-frailgroupswerecomparedusingANOVAfornormallydistributedandKruskal–Wallis-Testfornotnormallydistributedinterval-baseddataandchi-squaretestsforproportions. Ifdatawasmissingfortheprimaryoutcome(FP)orthemainoutcomeofPAmeasures,participantswereexcludedfromfurtheranalysis. AnalyseswereperformedusingSPSSforWindows(Version20;SPSSStatistics,Armonk,NY,IBMCorp.)andMATLAB(R2017b,TheMathWorks,Inc.,Massachusetts,USA).ThelevelofstatisticalsignificancewassettoP<0.05. Results Participants Of426personspotentiallyeligible,52participatedinthestudy(Figure1).Reasonsfornon-participationwerenotsystematicallyevaluated.Fiveparticipantshadtobeexcludedposthocbecauseofincompleteaccelerometerdata.Theseparticipants(4women,1man)wereofcomparableage(70.4years)andwereallnon-frailexceptonepersonwhowaspre-frail. AspresentedinTable2,the47olderadultsfinallyincludedintheanalysis(31women(66%),74±6yearsold)wereabletowalkindependentlywithout(77%)orwithawalkingaid(23%).Participantswithphysicalfrailtyweresignificantlymoresedentary,walkedlesssteps,expendedlessenergy,andaccumulatedlessMVPA(seeTable3). Table2CharacteristicsfortheTotalStudySampleandforSubgroupsofthePrimaryPhenotype Table3PhysicalActivityDatafortheTotalStudySampleandforSubgroupsofthePrimaryPhenotype FrailtyPhenotypePrevalenceAccordingtotheAssessmentandDefinitionoftheLPACriterion DependingontheassessmentanddefinitionofLPA,frailtyprevalencevariedbetween14.9%(n=7,95%CI:7.4%to27.7%)and31.9%(n=15,95%CI:20.4%to46.2%)(Figure2).AsseeninTable4,between27.7%(n=13,95%CI:16.9%to41.8%)and66.0%(n=31,95%CI:51.7%to77.8%)ofparticipantswereclassifiedaspre-frail.Overall,usingdifferentLPAcriteriaresultedinavariationof17.0percentagepoints.Usingonlypublishedphenotypemodels,thevariationwasstill12.8percentagepoints,whichcorrespondstomorethanhalfofthepre-definedvalueof17.7percentagepoints. Table4FrailtyPhenotypePrevalenceAccordingtotheAssessmentandDefinitionofLowPhysicalActivity Figure2Frailtyprevalenceaccordingtotheassessmentanddefinitionofthelowphysicalactivitycriterion. Notes:Cut-pointmodels:(A):Fried’svalues.(B):specificvalues.(C):MVPArecommendation. Abbreviations:Accel,Accelerometer;CI,confidenceinterval;GPAQ50+,GermanPhysicalActivityQuestionnaire50+;IPAQ,InternationalPhysicalActivityQuestionnaire;MLTPAQ,MinnesotaLeisureTimePhysicalActivityQuestionnaire;MVPA,moderatetovigorousphysicalactivity;PASE,PhysicalActivityScalefortheElderly;7D-PAR,SevenDayPhysicalActivityRecall. ThehighestfrailtyprevalenceoccurredwhentheLPAcriterionwasbasedonaccelerometermeasurementscombinedwiththeMVPArecommendationsforsteps.UsingtheGPAQ50+withFriedcut-points,7the7D-PARwithFriedcut-pointsoranaccelerometerwiththeCopelandandEsliger28algorithmtodefineMVPAresultedinthelowestprevalence,witheverysingleparticipantbeingsufficientlyphysicallyactive. BesidesLPA,themostfrequentpositivecriterionoftheFPwasslowgaitspeed(n=16,34%),followedbyexhaustion(n=15,32%)andweakness(n=13,28%).Unintendedweightlosswasrarelyobserved(n=3,6%).AsstatedinTable5,approximately50%oftheparticipantsshowednosignofphysicalfrailty. Table5AnalysisofPositiveItemsoftheFrailtyPhenotype Discussion TheaimofthisstudywastoanalyzetheinfluenceoftheassessmentanddefinitionofLPAontheprevalenceoftheFPincommunity-dwellingolderadults. ModifyingtheLPAcriterionoftheFPinthesamesampleresultedinavariationofFPprevalencerangingfrom14.9%to31.9%.ForthosephenotypeversionswithexactlythesameLPAcriterionsusedinpreviouspublications,suchasthePASEphenotypeversions,22–24theprevalencestillvariedby12.8percentagepoints.Thisvariationcorrespondstomorethanhalfofthe17.7percentagepointsreportedinrecentlypublishedreviews.2,8 Accordingtotheseresults,thechoiceofPAassessmentandcut-pointforlowactivitydoesindeedhaveaconsiderableimpactontheoccurrenceofphysicalfrailty.Thus,differencesinpublishedprevalencevaluesmayonlypartlyreflecttruedifferenceswithinorbetweenpopulations.Thisshouldbecarefullyconsideredwhencomparingprevalencedata,drawingconclusionsfrompublishedstudies,andplanningfuturestudies. OurresultssupportthefindingsofTheouetal,3whohaveconcludedthatingeneral,modificationsoftheFPresultinsubstantialdifferencesinphysicalfrailtyprevalence,predictivemortalityandinternalconsistencyoftheassessment.Inaretrospectiveanalysis,theycreated262differentphenotypesandreportedFPprevalenceestimationsrangingfrom12.7%to28.2%.3Thecorrespondingrangeof15.5percentagepointsiscomparabletothe17.0percentagepointsfoundinourstudy. Afterdatacollectionforourstudy,ameta-analysiswaspublished,whichfoundanevenhighervarianceofprevalence(rangefrom8.6%to50.9%)basedontheFrieddefinitioninMexicanolderadults.46 FrailtyPrevalence Twenty-nineparticipants(58%)wererecruitedfromastudysamplewheretheFPprevalencewas17.8%(95%CI:13.2%to22.5%),13whichisonlyslightlylowerthaninthepresentstudy.Therefore,weassumethatthesamplewassufficientlyrepresentativeofthepopulationincludedintheprimarystudyA. InthepopulationoftheCHS,6.9%oftheolderadultswereclassifiedasfrail,7whichislessthaninourstudy.Olderadultswithaprescriptionforphysiotherapyusuallyhavemedicalconditionsandmayshowphysicalimpairmentsanddisabilitymorefrequently,explainingthehigherfrailtyprevalenceinourstudy.Arelationshipbetweenactivitiesofdailyliving,disabilityandfrailtyhasbeenreported.22 ComparedtothephysicalFPversions,theFIbasedprevalenceestimation(34%frail,51%pre-frail)washigherthangenerallyreportedintheliterature.2ItisimportanttonotethattheFPandtheFIarenotalternativesbutcomplementaryandhavedifferentpurposesintheclinicalsetting.47TheFIwasnotoriginallydesignedtobecategorized,asitdescribesafitness-frailty-continuum.32 Cut-PointModels Accordingtoourresults,thechoiceofcut-pointvaluesseemstohaveamajorinfluenceonthevariabilityoffrailtyprevalence.Thecut-pointmodelA–thevaluesproposedbyFriedandcolleagues7–mayproduceaverylowprevalenceofpositiveLPAcriteria.ApplyingmodelAontheGPAQ50+orthe7D-PAR,noparticipantshowedapositiveLPAcriterion.Thesedifferences,possiblyduetovariationsinthescopeandnumberofactivitiesincludedinthesequestionnaires,shouldbecarefullyconsidered. Otherstudieshaveusedpopulation-dependentapproachesfordefiningcut-pointvaluesfortheLPAcriterion,mostlyreferringtoaspecificpopulation.Relatedtooursecondcut-pointapproach(modelB:cut-pointvaluesrecommendedforaspecificpopulation),thedifferencebetweenoursamplepopulation(85.5%positiveLPAcriterion)andparticipantswhoshowedapositiveLPAcriterioninastudybyChenetal26(19.5%)isparticularlystriking.Thus,werecommenddefiningalargepopulationinwhichthesamecut-pointcanbeused,insteadofseveralspecificcut-pointsforanumberofsmallpopulations.Thismayalsobemorefeasibleinclinicalcareandmayeasecomparisonsbetweenstudies. WealsoincludedpublichealthrecommendationsforMVPAtodefinetheLPAcriterion(modelC),18eventhoughtheseareunderdebateandmaybetooambitiousforolderadults.48Here,participantsshowedcomparablelevelsofLPA(between500and700min/week)irrespectiveofthePAquestionnaireused,butvaryinglevelsofLPAwhenaccelerometerswereused.ThisisinlinewithareviewonolderadultsmeetingPArecommendations,whichshowedarangefrom2%to83%.49 AssessingPA UsingquestionnairestoassessPAinolderadultsremainschallengingwithrespecttopsychometricproperties,contextspecificity,recallorobservationtime,suitabilityforcognitivelyimpairedindividuals,andthedomainstobeincluded.50Inallmodels,exceptfortheMLTPAQ,wefoundthatindividualsclassifiedasfrailshowedsignificantlylessPAcomparedtothoseclassifiedasnon-frail.ThismightindicatethattheMLTPAQmightnotbeverywellapplicableforolderadults. WealsoassessedPAbyaccelerometry.Thereissomeevidencethataccelerometerdatahavethestrongestassociationwithfrailty.12AlthoughaccelerometersshowedfarbetterresultsinvalidationstudiescomparedtoPAquestionnairesintermsofcorrelationtothedoublylabeledwatermethod(egColbertetal51),outcomesrelyalotontheinvestigator’schoiceofcut-points.52Inoursample,150minMVPAperweek(asrecommendedbytheWHO)werehardlysurpassedwhenusingtheFreedson27model,whereaseverysingleparticipantfulfilledtheWHOrecommendationifthemodelproposedbyCopelandandEsliger28wasused.ApossibleexplanationforthismightbethedifferentdefinitionsofMVPAduringwalking(Freedson:274.0km/h,CopelandandEsliger:283.2km/h). Toinvestigatedailystepsofolderadultsinclinicalpractice,asinglepedometerorsmartphonemightbeapracticalchoice.However,theproblemofdefiningappropriatecut-pointsforastepremains.Olderadultsmaypresentgaitpatternsandbiomechanicsdifferentfromyoungeradults. Sedentarytimewasfoundtobecorrelatedwithphysicalfrailtyandmightbeanindependentriskfactor.53Otherstudiesstresstheimportanceofsedentarytimeintermsoffrailtylevels.54Participantsinthepresentstudyweresedentaryfor9.9±2.9 hoursperday,whichcorrespondsto73%ofweartime.Themoretimeparticipantsspentinsedentarylevelsrelativetotheirweartime,thefrailertheywere.ThevaluesareinaccordancewithdatafromahugecohortoftheOsteoarthritisInitiative,whereolderadultsspent9.9 hoursperday(or66%ofwakingtime)beingsedentary.53 Limitations Toourknowledge,thisisthefirststudytoanalyzetheimpactofmodifyingtheLPAcriterionintheFP.However,thereareseverallimitations. Thetotalnumberofparticipantswithvaliddata(n=47)wasrelativelysmallandnoformalsamplesizecalculationwasperformed.However,thisisamethodologicalratherthanaprevalencestudy,andtheconsiderablemagnitudeoftheresultingconfidenceintervalstakesthesmallsamplesizeintoaccount.44 Therecruitmentprocessitselfmighthaveinducedbiasaswestruggledtofindenoughparticipantsforthestudy.Thepopulationofthepresentstudydoesnotrepresentarandomandrepresentativesampleofoldercommunity-dwellingadultslivinginanurbanregioninGermany. Usinganestablishedassessmentofcognitivestatusinsteadofsubjectiveratingscouldhaveresultedinaslightlydifferentcompositionofoursample.Forsomeparticipants,itmayhavebeendifficulttoconcentratethroughoutallthestudyassessments.Weattemptedtominimizethequestion-orderbiasbyselectingarandomorderinaskingPAquestionnaires. TheselectionofPAquestionnairesandthecorrespondingcut-pointsemergedfromanon-systematicliteratureresearchonstudiesusingtheFPupto2015andunderlyingcriteriasuchastheavailabilityofaGermanlanguageversion.Asystematicreviewwouldhaveproducedamorerepresentativesampleofmeasurementinstruments.Therefore,ourfindingscanbegeneralizedtoalimitedextentonly. Perspectives ThecalculationofcorrectionfactorsbetweencommonlyusedLPAcriteriasetsmightallowabettercomparisonofpublishedprevalencerates.However,thiswouldrequiresettingupastudyinwhichvariousmethodsofPAassessmentanddifferentcut-pointsareappliedinaverylargeanddiversesample. Despitethediagnosticlimitations,itisveryimportanttodetectfrailtyinolderadults,asitisassociatedwithvariousnegativehealthoutcomes.1Infuturestudiesweshouldendeavortomeasurephysicalfrailtywithstandardized,reliableandvalidmethods. ConsumerwearablesmayincreasinglybecomeagoodoptiontoassessthePAlevelinanaccuratewayforuseintheFP.Combiningmeasurementsofaccelerometry,heartrateandlocation(GPS),ashasbeendoneinTheouetal,12promisestoimprovethevalidityofPAmeasurementsasonecriterionforphysicalfrailty. Conclusion OurresultsindicatethatthemethodofPAassessmentandthechoiceofcut-pointvaluesforLPAimpactstheprevalenceofphysicalfrailtyquiteconsiderably.Thus,theinterpretationandcomparisonofpublishedphysicalfrailtyprevalencedataneedstobedonewithmorecaution,andmoreefforttostandardizeandadheretotheLPAcriterionseemswarranted.Developingcorrectionformulasandstandardizedassessmentmethods(includingcut-pointvalues)forclearlydefinedpopulations(egagecategories)andcontextscouldhelptoensureandimproveinternalvalidityoftheassessmentandenablebetterinterstudycomparison.ThiswouldbeanimportantprerequisiteforamorefrequentuseoftheFPinbothresearchandclinicalsettings. Abbreviations CHS,CardiovascularHealthStudy;cpm,countsperminute;FI,FrailtyIndex;FP,frailtyphenotype;GPAQ50+,GermanPhysicalActivityQuestionnaire50+;IPAQ,InternationalPhysicalActivityQuestionnaire;LPA,lowphysicalactivity;MLTPAQ,MinnesotaLeisureTimePhysicalActivity510Questionnaire;MVPA,moderatetovigorousphysicalactivity;PA,physicalactivity;PAEE,physicalactivityenergyexpenditure;PASE,PhysicalActivityScalefortheElderly;RMR,restingmetabolicrate;TEE,totalenergyexpenditure;WHO,WorldHealthOrganization;7D-PAR,SevenDayPhysicalActivityRecall. EthicsApprovalandInformedConsent ApprovalwasobtainedfromtheEthicalReviewBoardoftheGermanConfederationforPhysiotherapy(registrationnumber:2015-07).ThisstudywasconductedinaccordancewiththeDeclarationofHelsinkiinitsrecentversion(64thWMAGeneralAssembly,Fortaleza,Brazil,October2013).Allparticipantsprovidedwritteninformedconsentforstudyparticipation. DataSharingStatement Datacanbeobtainedfromthecorrespondingauthoruponreasonablerequest. Acknowledgments Wewouldliketothanktheresearchteaminvolvedinthedatacollectionprocess:CarolinBahns,TheresaRetzmann,LisaHappe,JuliaRascheandSvenjaHansen. AuthorContributions Allauthorscontributedtodataanalysis,draftingorrevisingthearticle,gavefinalapprovaloftheversiontobepublished,andagreetobeaccountableforallaspectsofthework. Funding ThisresearchwasfundedbyanintramuralgrantfromtheHochschulefürGesundheitBochum(UniversityofAppliedSciences).Wedidnotreceiveanyspecificgrantfromfundingagenciesinthepublic,commercial,ornot-for-profitsectors.Thefundingbodyhadnoroleinstudydesign,datacollectionandanalysis,decisiontopublish,orpreparationofthemanuscript. Disclosure Theauthorsdeclarethattherearenoconflictsofinterestinthiswork. References 1.VermeirenS,Vella-AzzopardiR,BeckweeD,etal.Frailtyandthepredictionofnegativehealthoutcomes:ameta-analysis.JAmMedDirAssoc.2016;17(12):1163.e1161–1163.e1117.doi:10.1016/j.jamda.\2016.09.010 2.CollardRM,BoterH,SchoeversRA,OudeVoshaarRC.Prevalenceoffrailtyincommunity-dwellingolderpersons:asystematicreview.JAmGeriatrSoc.2012;60(8):1487–1492.doi:10.1111/j.1532-5415.2012.04054.x 3.TheouO,CannL,BlodgettJ,WallaceLM,BrothersTD,RockwoodK.Modificationstothefrailtyphenotypecriteria:systematicreviewofthecurrentliteratureandinvestigationof262frailtyphenotypesintheSurveyofHealth,Ageing,andRetirementinEurope.AgeingResRev.2015;21:78–94.doi:10.1016/j.arr.2015.04.001 4.ApostoloJ,CookeR,Bobrowicz-CamposE,etal.Predictingriskandoutcomesforfrailolderadults:anumbrellareviewoffrailtyscreeningtools.JBIDatabaseSystemRevImplementRep.2017;15(4):1154–1208.doi:10.11124/JBISRIR-2016-003018 5.deVriesNM,StaalJB,vanRavensbergCD,HobbelenJS,OldeRikkertMG,Nijhuis-vanderSandenMW.Outcomeinstrumentstomeasurefrailty:asystematicreview.AgeingResRev.2011;10(1):104–114.doi:10.1016/j.arr.2010.09.001 6.MitnitskiAB,MogilnerAJ,RockwoodK.Accumulationofdeficitsasaproxymeasureofaging.SciWorldJ.2001;1:323–336.doi:10.1100/tsw.2001.58 7.FriedLP,TangenCM,WalstonJ,etal.Frailtyinolderadults:evidenceforaphenotype.JGerontolaBiolSciMedSci.2001;56(3):M146–M156.doi:10.1093/gerona/56.3.M146 8.ChoiJ,AhnA,KimS,WonCW.Globalprevalenceofphysicalfrailtybyfried’scriteriaincommunity-dwellingelderlywithnationalpopulation-basedsurveys.JAmMedDirAssoc.2015;16(7):548–550.doi:10.1016/j.jamda.2015.02.004 9.ChangSF,LinPL.Frailphenotypeandmortalityprediction:asystematicreviewandmeta-analysisofprospectivecohortstudies.IntJPsychiatrNursRes.2015;52(8):1362–1374. 10.TaylorHL,JacobsDRJr,SchuckerB,KnudsenJ,LeonAS,DebackerG.Aquestionnairefortheassessmentofleisuretimephysicalactivities.JChronicDis.1978;31(12):741–755.doi:10.1016/0021-9681(78)90058-9 11.Jorstad-SteinEC,HauerK,BeckerC,etal.Suitabilityofphysicalactivityquestionnairesforolderadultsinfall-preventiontrials:asystematicreview.JAgingPhysAct.2005;13(4):461–481.doi:10.1123/japa.13.4.461 12.TheouO,JakobiJM,VandervoortAA,JonesGR.Acomparisonofphysicalactivity(PA)assessmenttoolsacrosslevelsoffrailty.ArchGerontolGeriatr.2012;54(3):e307–e314.doi:10.1016/j.archger.2011.12.005 13.BraunT,ThielC,ZillerC,etal.PrevalenceoffrailtyinolderadultsinoutpatientphysiotherapyinanurbanregioninthewesternpartofGermany:across-sectionalstudy.BMJOpen.2019;9(6):e027768.doi:10.1136/bmjopen-2018-027768 14.ThielC,BraunT,GrünebergC.KörperlichesTrainingalsKernkomponentemultimodalerBehandlungältererMenschenmitFrailty–StudienprotokolleinerrandomisiertenkontrolliertenPilotstudie.ZGerontolGeriat.2018;52:45–60. 15.vonElmE,AltmanDG,EggerM,PocockSJ,GotzschePC,VandenbrouckeJP.TheStrengtheningtheReportingofObservationalStudiesinEpidemiology(STROBE)statement:guidelinesforreportingobservationalstudies.JClinEpidemiol.2008;61(4):344–349.doi:10.1016/j.jclinepi.2007.11.008 16.BraunT,ThielC,SchulzR,GrünebergC.DiagnostikundBehandlungphysischerFrailty.DtschMedWochenschr.2017;142(02):117–122.doi:10.1055/s-0042-101631 17.RobertsHC,DenisonHJ,MartinHJ,etal.Areviewofthemeasurementofgripstrengthinclinicalandepidemiologicalstudies:towardsastandardisedapproach.AgeAgeing.2011;40(4):423–429.doi:10.1093/ageing/afr051 18.GlobalRecommendationsonPhysicalActivityforHealth.Geneva:WorldHealthOrganization;2010.Availablefrom:https://www.ncbi.nlm.nih.gov/books/NBK305057/.AccessedNovember8,2019. 19.EckelSP,Bandeen-RocheK,ChavesPH,FriedLP,LouisTA.Surrogatescreeningmodelsforthelowphysicalactivitycriterionoffrailty.AgingClinExpRes.2011;23(3):209–216.doi:10.1007/BF03324962 20.IPAQResearchCommittee.Guidelinesfordataprocessingandanalysisoftheinternationalphysicalactivityquestionnaire(IPAQ)-shortandlongforms.2005.Availablefrom:https://sites.google.com/site/theipaq/scoring-protocol.AccessedNovember15,2015. 21.ReisJúniorWM,CarneiroJAO,CoqueiroR,SantosKT,FernandesMH.Pre-frailtyandfrailtyofelderlyresidentsinamunicipalitywithalowHumanDevelopmentIndex.RevLatAmEnfermagem.2014;22(4):654–661.doi:10.1590/0104-1169.3538.2464 22.AlSnihS,GrahamJE,RayLA,Samper-TernentR,MarkidesKS,OttenbacherKJ.Frailtyandincidenceofactivitiesofdailylivingdisabilityamongoldermexicanamericans.JRehabilMed.2009;41(11):892–897.doi:10.2340/16501977-0424 23.GrahamJE,SnihS,BergesIM,RayLA,MarkidesKS,OttenbacherKJ.Frailtyand10-yearmortalityincommunity-livingMexicanAmericanolderadults.Gerontology.2009;55(6):644–651.doi:10.1159/000235653 24.CawthonPM,EnsrudKE,LaughlinGA,etal.Sexhormonesandfrailtyinoldermen:theosteoporoticfracturesinmen(MrOS)study.JClinEndocrinolMetab.2009;94(10):3806–3815.doi:10.1210/jc.2009-0417 25.WuI-C,LinX-Z,LiuP-F,TsaiW-L,ShieshS-C.Lowserumtestosteroneandfrailtyinoldermenandwomen.Maturitas.2010;67(4):348–352.doi:10.1016/j.maturitas.2010.07.010 26.ChenS,HondaT,ChenT,etal.Screeningforfrailtyphenotypewithobjectively-measuredphysicalactivityinawestJapanesesuburbancommunity:evidencefromtheSasaguriGenkimonStudy.BMCGeriatr.2015;15:36.doi:10.1186/s12877-015-0037-9 27.FreedsonPS,MelansonE,SirardJ.Calibrationofthecomputerscienceandapplications,Inc.accelerometer.MedSciSportsExerc.1998;30(5):777–781.doi:10.1097/00005768-199805000-00021 28.CopelandJL,EsligerDW.Accelerometerassessmentofphysicalactivityinactive,healthyolderadults.JAgingPhysAct.2009;17(1):17–30.doi:10.1123/japa.17.1.17 29.BastiaanseLP,HilgenkampTI,EchteldMA,EvenhuisHM.Prevalenceandassociatedfactorsofsarcopeniainolderadultswithintellectualdisabilities.ResDevDisabil.2012;33(6):2004–2012.doi:10.1016/j.ridd.2012.06.002 30.Tudor-LockeC,CraigCL,AoyagiY,etal.Howmanysteps/dayareenough?Forolderadultsandspecialpopulations.IntJBehavNutrPhysAct.2011;8:80.doi:10.1186/1479-5868-8-80 31.SearleSD,MitnitskiA,GahbauerEA,GillTM,RockwoodK.Astandardprocedureforcreatingafrailtyindex.BMCGeriatr.2008;8:24.doi:10.1186/1471-2318-8-24 32.RockwoodK,AndrewM,MitnitskiA.Acomparisonoftwoapproachestomeasuringfrailtyinelderlypeople.JGerontolaBiolSciMedSci.2007;62(7):738–743.doi:10.1093/gerona/62.7.738 33.HuyC,SchneiderS.InstrumentfürdieErfassungderphysischenAktivitätbeiPersonenimmittlerenundhöherenErwachsenenalter.ZGerontolGeriat.2008;41(3):208–216.doi:10.1007/s00391-007-0474-y 34.CraigCL,MarshallAL,SjostromM,etal.Internationalphysicalactivityquestionnaire:12-countryreliabilityandvalidity.MedSciSportsExerc.2003;35(8):1381–1395.doi:10.1249/01.MSS.0000078924.61453.FB 35.SallisJF,BuonoMJ,RobyJJ,MicaleFG,NelsonJA.Seven-dayrecallandotherphysicalactivityself-reportsinchildrenandadolescents.MedSciSportsExerc.1993;25(1):99–108.doi:10.1249/00005768-199301000-00014 36.MärkiA.EntwicklungundEvaluationeinesBeratungsinstrumenteszurFörderungderkörperlichenAktivitätbeiälterenMenschenunterBrücksichtigungdesTranstheoretischenModellsderVerhaltensänderung[dissertation].Zürich:Philosophisch-HistorischeFakultät,UniversitätBasel;2004. 37.WashburnRA,SmithKW,JetteAM,JanneyCA.ThePhysicalActivityScalefortheElderly(PASE):developmentandevaluation.JClinEpidemiol.1993;46(2):153–162.doi:10.1016/0895-4356(93)90053-4 38.CalabroMA,KimY,FrankeWD,StewartJM,WelkGJ.Objectiveandsubjectivemeasurementofenergyexpenditureinolderadults:adoublylabeledwaterstudy.EurJClinNutr.2015;69(7):850–855.doi:10.1038/ejcn.2014.241 39.MifflinMD,StJeorST,HillLA,ScottBJ,DaughertySA,KohYO.Anewpredictiveequationforrestingenergyexpenditureinhealthyindividuals.AmJClinNutr.1990;51(2):241–247.doi:10.1093/ajcn/51.2.241 40.ThielC,GabrysL,VogtL.RegistrierungkörperlicherAktivitätmittragbarenAkzelerometern.DtschZSportmed.2016;67(2):44–48.doi:10.5960/dzsm.2016.220 41.ChoiL,LiuZ,MatthewsCE,BuchowskiMS.Validationofaccelerometerwearandnonweartimeclassificationalgorithm.MedSciSportsExerc.2011;43(2):357–364.doi:10.1249/MSS.0b013e3181ed61a3 42.HagstromerM,OjaP,SjostromM.Physicalactivityandinactivityinanadultpopulationassessedbyaccelerometry.MedSciSportsExerc.2007;39(9):1502–1508.doi:10.1249/mss.0b013e3180a76de5 43.NelsonME,RejeskiWJ,BlairSN,etal.Physicalactivityandpublichealthinolderadults:recommendationfromtheAmericanCollegeofSportsMedicineandtheAmericanHeartAssociation.MedSciSportsExerc.2007;39(8):1435–1445.doi:10.1249/mss.0b013e3180616aa2 44.TerweeCB,MokkinkLB,KnolDL,OsteloRW,BouterLM,deVetHC.Ratingthemethodologicalqualityinsystematicreviewsofstudiesonmeasurementproperties:ascoringsystemfortheCOSMINchecklist.QualLifeRes.2012;21(4):651–657.doi:10.1007/s11136-011-9960-1 45.BlandJM,AltmanDG.Statisticalmethodsforassessingagreementbetweentwomethodsofclinicalmeasurement.Lancet.1986;1(8476):307–310.doi:10.1016/S0140-6736(86)90837-8 46.JocabedRocha-BalcázarL,CortésSarmientoD,CastellanosPerillaN,Núñez-AguirreS,Salinas-MartínezR,Perez-ZepedaM.Systematicreviewandmeta-analysisoffrailtyprevalenceinMexicanolderadults.JLatAmGeriatricMed.2018;4:44–49. 47.CesariM,GambassiG,vanKanGA,VellasB.Thefrailtyphenotypeandthefrailtyindex:differentinstrumentsfordifferentpurposes.AgeAgeing.2014;43(1):10–12.doi:10.1093/ageing/aft160 48.WarburtonDE,BredinSS.Reflectionsonphysicalactivityandhealth:whatshouldwerecommend?CanJCardiol.2016;32(4):495–504.doi:10.1016/j.cjca.2016.01.024 49.SunF,NormanIJ,WhileAE.Physicalactivityinolderpeople:asystematicreview.BMCPublicHealth.2013;13(1):449.doi:10.1186/1471-2458-13-449 50.vanPoppelMM,ChinapawMM,MokkinkL,vanMechelenW,TerweeC.Physicalactivityquestionnairesforadults.SportsMed.2010;40(7):565–600.doi:10.2165/11531930-000000000-00000 51.ColbertLH,MatthewsCE,HavighurstTC,KimK,SchoellerDA.Comparativevalidityofphysicalactivitymeasuresinolderadults.MedSciSportsExerc.2011;43(5):867–876.doi:10.1249/MSS.0b013e3181fc7162 52.PedisicZ,BaumanA.Accelerometer-basedmeasuresinphysicalactivitysurveillance:currentpracticesandissues.BrJSportsMed.2015;49(4):219–223.doi:10.1136/bjsports-2013-093407 53.SongJ,LindquistLA,ChangRW,etal.Sedentarybehaviorasariskfactorforphysicalfrailtyindependentofmoderateactivity:resultsfromtheosteoarthritisinitiative.AmJPublicHealth.2015;105(7):1439–1445.doi:10.2105/AJPH.2014.302540 54.BlodgettJ,TheouO,KirklandS,AndreouP,RockwoodK.FrailtyinNHANES:comparingthefrailtyindexandphenotype.ArchGerontolGeriatr.2015;60(3):464–470.doi:10.1016/j.archger.2015.01.016 ThisworkispublishedandlicensedbyDoveMedicalPressLimited.Thefulltermsofthislicenseareavailableathttps://www.dovepress.com/terms.phpandincorporatetheCreativeCommonsAttribution-NonCommercial(unported,v3.0)License. ByaccessingtheworkyouherebyaccepttheTerms.Non-commercialusesoftheworkarepermittedwithoutanyfurtherpermissionfromDoveMedicalPressLimited,providedtheworkisproperlyattributed.Forpermissionforcommercialuseofthiswork,pleaseseeparagraphs4.2and5ofourTerms. DownloadArticle[PDF]   ContactUs  • PrivacyPolicy  • Associations&Partners  • Testimonials  • Terms&Conditions  • Recommendthissite • Top ContactUs • PrivacyPolicy Accept Inordertoprovideourwebsitevisitorsandregistereduserswithaservicetailoredtotheirindividualpreferencesweusecookiestoanalysevisitortrafficandpersonalisecontent.YoucanlearnaboutouruseofcookiesbyreadingourPrivacyPolicy.Wealsoretaindatainrelationtoourvisitorsandregisteredusersforinternalpurposesandforsharinginformationwithourbusinesspartners.Youcanlearnaboutwhatdataofyoursweretain,howitisprocessed,whoitissharedwithandyourrighttohaveyourdatadeletedbyreadingourPrivacyPolicy. IfyouagreetoouruseofcookiesandthecontentsofourPrivacyPolicypleaseclick'accept'.