Long-form but not short-form Mini-Nutritional Assessment is
appropriate for grading nutritional risk of patients on hemodialysis—A cross-sectional study
Alan C. Tsai
a,b,*, Mei-Zu Chang
caDepartmentofHealthcareAdministration,AsiaUniversity,Taichung41354,Taiwan
bDepartmentofHealthServicesManagement,SchoolofPublicHealth,ChinaMedicalUniversity,Taichung404,Taiwan
cHemodialysisUnit,ChungShanMedicalUniversityHospitalChung-HsingDivision,Taichung402,Taiwan
Whatisalreadyknownaboutthetopic?
Patientsonhemodialysisareathighriskofmalnutrition.
Routinemonitoringoftheirnutritionalstatusisimpor-
tantinpreventingprotein-caloriemalnutritioninthese patients.
SubjectiveGlobalAssessmentistherecommendedtool for monitoring their nutritional status. But extensive training and experience is needed to use this tool properly.Thus,asimple,reliable,non-invasiveandtime- savingtoolisgreatlydesired.
Long-form Mini-Nutritional Assessment (MNA-LF), a relativelysimple andnon-invasive tool,is appropriate ARTICLE INFO
Articlehistory:
Received3December2010 Receivedinrevisedform29April2011 Accepted5May2011
Keywords:
Nutritionalriskscreening Hemodialysis
MiniNutritionalAssessment Protein-energymalnutrition SubjectiveGlobalAssessment
ABSTRACT
Background:Routine screening/assessment of protein-energy status is essential for preventinguremic malnutritioninpatientsonhemodialysis(HD).Asimple,lowcost, reliableandnon-invasivetoolisgreatlydesired.
Objective:Thisstudyaimedtoevaluatetheappropriatenessofusingthelong-form(LF) andtheshort-form(SF)MiniNutritionalAssessment(MNA)forgradingtheriskofprotein- energymalnutritioninpatientsonHD.
Designandsampling: Across-sectionalstudywithpurposivesampling.
Setting:Ahospital-managedhemodialysiscenter.
Participants: 152adultambulatorypatientsonhemodialysis.
Methods:ThenutritionalstatusofeachpatientwasgradedwithMNA-LFandMNA-SF,each intwoversions—anormalized-original(content-equivalent)version(byadoptingpopula- tion-specificanthropometriccut-offpoints)andanalternativeversionthatreplacedcalf circumferenceforBMIinthescale.TheSGA,serumalbuminandserumcreatinineservedas references.Cross-tabulationtestwasusedtoevaluatetheconsistencyoftheversions.
Results:MNA-SFversionsratedfewerHDsubjectsmalnourishedoratriskofmalnutrition (32.2%and24.3%forT1andT2,respectively)comparedtoMNA-LFversions(40.8%and 36.2%)ortheSGA(47.4%).MNA-SFversions(kappa=0.450and0.446)alsodidnotperform aswellasMNA-LFversions(kappa=0.734and0.666)inpredictingtheriskofmalnutrition inHDpatientsusingtheSGAasthereference.MNA-SFalsodidnotperformaswellasthe MNA-LFusingserumalbuminorserumcreatinineasthereference.
Conclusions: TheMNA-LFisappropriateforpredictingprotein-energymalnutritioninHD patientsbutMNA-SFmayunder-ratethesepatients.Effortshouldbemadetoimprovethe MNA-SFforHDpatientssincetheshort-formismoretime-efficientandthus,greatly desiredinclinicalpractice.
ß2011ElsevierLtd.Allrightsreserved.
* Correspondingauthorat:DepartmentofHealthcareAdministration, AsiaUniversity,500LiufengRoad,Wufeng,Taichung41354,Taiwan.
Tel.:+886423323456x1943;fax:+886423321206.
E-mailaddress:atsai@umich.edu(A.C.Tsai).
ContentslistsavailableatScienceDirect
International Journal of Nursing Studies
journalhomepage:www.elsevier.com/ijns
0020-7489/$–seefrontmatterß2011ElsevierLtd.Allrightsreserved.
doi:10.1016/j.ijnurstu.2011.05.004
for grading therisk of protein-calorie malnutrition in patients on hemodialysis(HD). Butthevalidityof the short-form(SF)hasnotbeenexamined.
Whatthispaperadds
TheMNA-LFandMNA-SFdonotagreewelltoeachother ingradingnutritionalriskofpatientsonHD.
TheMNA-SFratesasmallerproportionofpatientsonHD malnourishedoratriskofmalnutritioncomparedtothat ratedwiththeMNA-LF.
Until it is further revised and validated, the MNA-SF should be used with caution for grading the risk of malnutritioninpatientsonHD.
1. Introduction
Thetotal numberof patients beingtreated for end- stagerenaldisease(ESRD)hasbeenontherisegloballyin recent years (Schober-Halstenberg, 2009) and protein- energymalnutritioniscommoninthesepatients(Qureshi etal.,1998;NationalKidneyFoundation,2000).Persons onhemodialysis(HD)areathighriskofprotein-energy malnutritionbecausethesepatientsoftenhavedepressed appetite,restricteddiet,drug–nutrientinteractionsand dialysis-relatedloss of nutrients. They also often have non-protein-energy-malnutrition-associated metabolic abnormalitiessuchashyperphosphatemia,hyperkalamia andacidosis.Toavoidseverenutritionaldeficiencyitis imperative to monitor/screen their nutritional status routinely.
The Subjective Global Assessment (SGA) has been recommended by the U.S. National Kidney Foundation KidneyDisease/Dialysis OutcomesandQuality Initiative (K/DOQI)clinicalpracticeguidelinesasaregulartoolfor assessingprotein-energymalnutritioninlargepopulations ofESRDpatients(NationalKidneyFoundation,2000).Itis alsothetool recommendedforassessingthenutritional status of HD patients in Taiwan (Taiwan Society of Nephrology,2010). Several studieshaveshownthat the SGAisreliableforassessingprotein-energymalnutritionin HDpatients(Steiberetal.,2007;Yangetal.,2007).Arecent study has also shown that the SGA is effective in identifying malnutrition and is appropriate for cross- sectional assessment of nutritionalstatus in predialysis patients(Campbelletal.,2007).Longitudinalstudieshave shownthatbaselinenutritionalstatusassessedwiththe SGAisalsoassociatedwithfuturemortalityriskinchronic dialysis patients (Dwyeret al., 2005; de Mutsertet al., 2009).However,therearealsostudiesthatquestionedthe validityandreliabilityofSGAinidentifyingESRDpatients atrisk ofmalnutrition(Cooperet al.,2002;Jones etal., 2004;Steiberetal.,2004).
TheMiniNutritionalAssessment(MNA) isanappro- priatetoolforgrading/assessingtheriskofprotein-energy malnutrition in persons withvarious health conditions (Guigozetal.,2002;Vellasetal.,1999).TheMNAiseasy- to-use,lowcost,reliableand non-invasiveanddoesnot requireextensive professionalbackgroundortraining to use. So, the tool, especially the short-form, is highly desired.However,fewstudieshaveattemptedtoexamine
the suitability of this tool for grading the risk of malnutritioninpatientsonHD.Afsaretal.(2006)found thatthefull-MNAoverestimatedtheriskofmalnutritionof patients on HD compared totheSGA (65.7%vs. 32.8%).
Nevertheless, one recent study found the full-MNA appropriatefor ratingnutritionalriskof patientson HD (Tsaietal.,2009).TheMNAalsohasashort-form(SF)for screeningtheriskofundernutrition.Becauseofitstime- saving feature, the SF is much preferred by clinical practitioners.Resultsratedwiththetwoformsgenerally agreewitheachotherwellinpatientsfrommostsettings (Tsaietal.,2010a),buttheiragreementinpatientsonHD patientshasnotbeenevaluated.Hence,thepresentstudy wasaimedtoexaminewhethertheMNA-SFagreedwell withtheMNA-LFingradingpatientsonHD.
2. Methods
2.1. Nutritionalassessment/screeningtools
Thisstudyinvolvedthreenutritionassessment/screen- ingscales:theSGA,MNA-LFandMNA-SF.TheSGAgrades nutritional risk by subjectively assessing a patient’s historical dataon weightchange,altered dietaryintake, andgastrointestinalsymptomsinfluencingnutrientintake orabsorption,andeffectsofunder-nutrition.Theprocess also involves a physical examination to detect clinical characteristicsofunder-nutrition(Detskyetal.,1987;Enia et al., 1993). However, professional knowledge, clinical experience andspecialtrainingareessentialinapplying the tool properly.The SGA is based on a 7-pointscale.
Thosewhoscored1–2pointsareconsideredathighriskof malnutrition;3–5pointsatmoderaterisk;and6–7points atlowrisk(NationalKidneyFoundation,2000).
TheMNAwasdevelopedbasedmainlyonclinicaldata ofolderadultsinEuropeandUS.Thetoolhastwoforms:
MNA-LF consists of 18 items covering four dimensions (dietary,anthropometric,globalandself-evaluatedpara- meters)forassessingnutritionalriskstatusandtheMNA- SFconsistsof6items(thefirst6itemsoftheMNA-LF),for screeningtheriskofmalnutrition(Rubensteinetal.,2001).
Thetwoformscanidentifypersonswithprotein-energy malnutritioninatwo-stepscreeningprocess.Personswho areidentifiedas‘‘atrisk’’withtheSFarefurtherevaluated withtherestoftheLFtoconfirmthediagnosisandplan interventions.Itnormallytakesabout3–5mintocomplete theSFandabout10–20mintocompletetheLF.TheMNA- LFhasamaximumscoreof30.Ascoreof16.5orlesswas consideredmalnourished;17–23.5,atriskofmalnutrition;
and24ormore,normal(Vellasetal.,1999).TheMNA-SF hasamaximumscoreof14.Ascoreof0–7wasconsidered athighriskofmalnutrition;8–11,moderateriskand12–
14,lowrisk(Kaiseretal.,2009).
2.2. Subjects
Inthisstudywepurposivelyrecruitedoutpatientswho wereundergoingmaintenanceHDtreatmentataregional hospitalin CentralTaiwan.Of159HDpatients,152(78 menand74 women)agreedtoparticipateinthestudy.
Patientswereeligibletoparticipateiftheywere18years
ofage,non-pregnant,withoutacutediseaseorinfection, patientsofthecenterfor1monthandwillingtosignan informedconsent.Thestudywasconductedaccordingto theguidelineslaiddownintheDeclarationofHelsinkiand all procedures involving human subjects/patients were approvedbytheethicsreviewboardofthehospital.
2.3. Procedure
Aresearcher(clinicalnurse)whohadpriorexperiencein usingtheSGAwastrainedbyaclinicaldietitiantoratethe riskofmalnutritionofHDpatientswithboththeSGAand theMNA.Aftergoodconsistencywasestablishedbetween the dietician and the researcher (kappa>0.80), the researcherratedthepatientsindependently.Theresearcher interviewed each qualifiedparticipant with a structured questionnaire toelicit personal data and torate his/her nutritionalstatuswiththeMNAduringaroutinedialysis session. The same clinical nurse also evaluated each patient’snutritional statuswith the SGA duringanother dialysissession.Thetwo interviewswere approximately two weeks apart (roughly before and after the blood samplingdate)andtookplaceinrandomorder.Toavoid bias, theresearcherdidnotaccesstheratedresultsuntil both ratings were completed. Biochemical data were obtainedfrompatients’routinelaboratorytestsperformed bytheclinicallaboratoryofthehospital.Datacollectiontook placeduringAugustthroughDecember2009.
TheriskofmalnutritionintheseHDpatientswasrated withtheLFandSFoftwoTaiwanese-specificMNAversions (MNA-T1 and MNA-T2). MNA-T1was the same as the original MNA, except it adopted the Taiwanese-specific anthropometriccut-offpointstomaintaincontent-equiv- alency of the scale. MNA-T2 further substituted calf- circumference(CC,itemR)forBMI(itemF)anddoubled theweightingofitemQ(MAC)(Tsaietal.,2007,2008).The validityoftheTaiwanese-specificlong-formversionshas been shown previously (Tsai et al., 2008). Adoption of Taiwanese-specific anthropometric (BMI, MAC and CC) cut-offpoints(MNA-T1)improvedthepredictiveabilityof the scale whereas replacement of CC for BMI with an increaseofMACto2points(MNA-T2) furtherimproved the predictive abilityof the MNA. Similar resultswere observedintheshort-formversions.Adoptionofpopula- tion-specificanthropometricBMIcut-offpointsimproved thepredictiveabilityoftheMNA-T1-SF,whereasreplace- mentofCCforBMIfurtherimprovedthepredictiveability ofthescale(kappavalueswere0.596,0.742,and0.843for community-living;0.560,0.683,and0.839forcarecenter- living;and0.346,0.454,and0.522fornursinghome-living elderly for the original, MNA-T1-SF, and MNA-T2-SF, respectively, usingMNA-T1-LFas reference) (Tsaiet al., 2010a).MNA-T2had17insteadof18items(withoutBMI) but maintained the same maximum score (30 points).
MNA-T2wasshowntofunctionatleastaswellasMNA-T1 in Taiwanese(Tsaietal.,2010a,b). Weight,height,mid- armcircumferenceandcalfcircumferenceweremeasured accordingtostandardprocedures(LeeandNieman,2003) andasdescribedpreviously(Tsaietal.,2010a).
It is generally recognized that in ESRD patients, no singlemeasurecanprovideacomprehensiveindicationof
protein-energynutritionalstatusandmalnutritionmaybe identifiedwithgreatersensitivityandspecificityusinga combinationoffactors(K/DOQI,2000).Thus,inthepresent studynutritionalscores/statusgradedwiththeMNAand theSGAwerecomparedtothepredialysisserumalbumin andcreatinine,inadditiontotheSGA.Thesebiochemical indicators are considered valid and clinically useful measures of protein-energy nutritional status in main- tenance dialysis patients and are future indicators of mortality risk (National Kidney Foundation, 2000; Pifer et al., 2002; Honda et al., 2006; Phelan et al., 2008;
Herselmanetal.,2010).Wearbitrarilychose3.8g/dLasthe cut-offpointforserumalbuminand10mg/dLforserum creatinine.Thesevaluesarethesameas(orcloseto)the outcome guidelinevalues recommended bythe K/DOQI (National Kidney Foundation, 2000). It should also be mentioned that although both serum markers are con- sidered valid and clinicallyuseful measures of protein- energynutritionalstatusinmaintenancedialysispatients, thepresenceofacuteorchronicinflammationcanlimitthe specificityofserumalbuminwhereasdiminishedskeletal musclemasscanalsoimpactserumcreatinineconcentra- tion(NationalKidneyFoundation,2000).
2.4. Statisticalanalysis
Results were analyzed with SPSS 15.0 (Statistical Package for the Social Science, SPSS Inc., Chicago, IL).
Simplestatistics(n,%)wereusedtoshowthedistribution of subjects’basiccharacteristics.Biochemical datawere showninmeansandstandarddeviations(SD).Wilcoxon’s Signed-Ranktestwasusedtoevaluatethesignificanceof differencesamongthedistributionsof nutritionalstatus graded with various scales. Cross-tabulation test was performed to determine the agreement (sensitivity, specificityandkappa)ofnutritionalstatuspredictedwith the MNA with that predicted with serum albumin, creatinineortheSGA.Statisticalsignificancewasaccepted atalpha=0.05.
3. Results
Table1showsthesociodemographic,anthropometric andbiochemicalcharacteristicsofsubjects.Mostpatients (83% of male and 90% of female) had at least one comorbidity.Averageserumalbuminconcentrationswere 3.9 and 3.7g/dL;and creatinine, 10.9and 8.9mg/dLfor men and women, respectively. Table 2 shows the distributions of nutritional status (or risk levels) rated withtheSGAandtheMNA.ResultsratedwiththeSGAand MNA-T1-LFandMNA-T2-LFwerenotsignificantlydiffer- entfromeachother.Thetwoshort-formversionsdidnot agree well with the SGA or the respective long-form versions.Table3showsresultsofcross-tabulationtests.
Using theSGA as a reference, theconsistency withthe MNA-T1-LFandMNA-T2-LFwasmoderate(kappa=0.734 and 0.666, respectively); the consistency with the two short-formversionswaslower(kappa=0.450and0.446, respectively).The consistency with serum albuminand creatininewasgenerallyrelativelylow.Subjects’MNAand
SGAscoringpatternsareshown inAppendices1 and2, respectively.
4. Discussion
4.1. PredictiveabilityoftheMNA
ResultsofthepresentstudysuggestthattwoMNA-LF versionsareappropriateforgradingthenutritionalriskof malnutritioninpatientsonHD.ThetwoLFversionshave moderately high consistency with the SGA, the tool recommendedby K/DOQIformonitoring thenutritional status of patients on HD (National Kidney Foundation, 2000). ThetwoMNA-LFversionsarealsocomparableto theSGAingradingthenutritionalriskofpatientsonHD usingbiochemicalparameters(albuminandcreatinine)as references.However,thetwoMNA-SFversionsperformed
lessfavorablycomparedtothetwoMNA-LFversionsorthe SGA. Both MNA-T1-SF and MNA-T2-SF rated fewer subjects ‘‘malnourished+at risk of malnutrition’’ than their respective long-form versions or the SGA. These resultssuggest thatMNA-SFmay under-ratetherisk of malnutrition and cause a delay in the revelation of emergingprotein-energymalnutritioninpatientsonHD.
InadditiontotheSGA,twoscalesdevelopedfromthe SGA (DialysisMalnutrition Score,DMSand Malnutrition Inflammation Score, MIS), have also been used for assessingnutritionalstatusofpatients onHD(Kalantar- Zadeh et al., 1999). DMS was developed using the componentsoftheconventionalSGA butwithaquanti- tativescoringsystem.Eachcomponenthasascoreranging from1(normal)to5(severelyabnormal).Thesumofall seven items ranges from 7 (normal) to 35 (severely malnourished). MISwasdeveloped tomakethe scoring Table1
Sociodemographic,anthropometricandbiochemicalcharacteristicsofsubjects.
Item Men(N=78) Women(N=74)
N(%) MeanSD N(%) MeanSD
Age(years) 58.813.0 64.912.2
18–39 6(7.7) 3(4.1)
40–64 49(62.8) 31(41.8)
65–85 23(29.5) 40(54.1)
Formaleducation(years)
6 30(38.4) 49(66.2)
7–9 10(12.8) 12(16.2)
10–12 19(24.4) 11(14.9)
>12 19(24.4) 2(2.7)
Smokingstatus
No 39(50) 74(100)
Yes 18(23.1) 0(0.0)
Pastsmokers 21(26.9) 0(0.0)
Alcoholdrinkinga
No 74(94.9) 73(98.6)
Yes 4(5.1) 1(1.4)
Routineexerciseb
No 37(47.4) 47(63.5)
Yes 41(52.6) 27(36.5)
No.ofco-morbidityc
0 13(16.7) 7(9.5)
1 24(30.8) 16(21.6)
2 23(29.5) 27(36.5)
3 18(23.1) 24(30.5)
Functionalstatus
Totallyindependent 49(62.8) 28(37.8)
Needhelpbyothers 27(34.6) 45(60.8)
Totallydependentonothers 2(2.6) 1(1.4)
BMI(kg/m2) 22.73.1 22.53.8
Mid-armcircumference(cm) 27.63.1 27.23.7
Calfcircumference(cm) 32.93.4 29.73.3
Biomarkersd
Hemoglobin(g/dL) 11.21.4 10.31.2
Serumalbumin(g/dL) 3.90.3 3.70.2
SerumTG(mg/dL) 140.098.5 166.0104
Serumcholesterol(mg/dL) 144.034.2 166.037.6
Fastingbloodglucose(mg/dL) 105.046.3 118.054.6
Serumcreatinine(mg/dL) 10.92.4 8.91.3
Serumuricacid(mg/dL) 7.11.2 7.11.4
Ureaclearancerate(Kt/v) 1.320.1 1.40.1
Serumcalcium(mg/dL) 9.50.7 9.70.7
SerumP(mg/dL) 4.61.3 4.61.1
aDrinkatleastonetime/week.
b Exercised3daysormore/week.
cIncludinghypertension,diabetes,heartdisease,psychiatricdisease,respiratorydisease,strokeandcancer.
d Allbiochemicalvaluesarepredialysis.
systemmorecomprehensiveandquantitative.Itincluded thesevenDMSitems(withrevisions)andthreenewitems:
BMI,serumalbuminlevelandtotaliron-bindingcapacity (TIBC).Eachitemcanscorefrom0(normal)to3(severely abnormal)andthetotalscorerangesfrom0(normal)to30 (severelymalnourished).
Allthreescales(SGA,DMS,andMIS)andMNAevaluate items covering weightloss, appetite, GI function, func- tionalcapacity(ormobility),comorbidity(yearsofdialysis or comorbidity of dialysis for SGA, DMS and MIS; and psychologicalstress,acutedisease,depressionanddemen- tiaforMNA),lossofsubcutaneousfatandmusclewasting (MAC and CC for MNA). MIS and MNA also rate BMI.
However,onlyMISincludesbiochemicalindicators(serum andTIBC)andonlytheMNAincludesself-viewofhealth and nutritional status and intake of water, protein-rich
foodsandfruitsandvegetables.TheMNAhasaweighted scoringsystemwhiletheotherthreescaleshaveanequal- scoresystemforallitems.MNA-SFisthefirstsectionofthe MNA-LF and does not cover dietary intakes or self- evaluateditems.Itis aprescreeningtoolrather thanan assessmenttool.TheMIShasbeensuggestedtobesuperior to the conventional SGA and DMS because it includes biochemical indicators (Kalantar-Zadeh et al., 2001).
However,italsomakestheprocessinvasive,morecostly andmoretime-consuming.Accordingtotheresultsofthe presentstudy,theMNA-LFis atleastcomparable tothe SGAinassessingtheriskofmalnutritioninpatientsonHD.
ItshouldalsobementionedthatcomparedtotheSGA, thetwolong-formMNAversionsratedslightlylower(not statisticallysignificant)proportions(47.4%vs.40.8%and 36.2%) of subjectsat risk of malnutrition.On theother Table2
DistributionofnutritionalstatusratedwiththeMNA-LF,MNA-SForSGA(N=152).
Scaleandversion Distributionofratedstatus(N,%)
Malnourished Atrisk Normal Test
SGA(scorerange) (1–2) (3–5) (6–7)
Distribution 1(0.7) 71(46.7) 80(52.6) aa
Long-form(scorerange) (0–16.5) (17–23.5) (24–30)
MNA-T1 10(6.6) 52(34.2) 90(59.2) a
MNA-T2 10(6.6) 45(29.6) 97(63.8) a,b
Short-form(scorerange) (0–7) (8–11) (12–14)
MNA-T1 4(2.6) 45(29.6) 103(67.8) b
MNA-T2 4(2.6) 33(21.7) 115(75.7) c
SGA=SubjectiveGlobalAssessment;MNA=MiniNutritionalAssessment;LF=long-form;SF=short-form;T1=Taiwanversion-1;T2=Taiwanversion-2.
a a–cindicatewhetherthedistributionsratedwiththescalesaredifferentfromeachother.Setsofdistributionnotindicatedbythesamealphabetsare significantlydifferentfromeachotheronthebasisofWilcoxon’sSigned-Ranktest(p<0.05).
Table3
Binaryclassificationtestsofnutritionalassessment/screeningscaleswithreferencestandards(N=152).
Scale Serumalbumin Creatinine SGA
Atrisk Normal Atrisk Normal Atrisk Normal
SGA
Atrisk 43(59.3)a 15(18.8)b 56(77.8) 25(31.2) – –
Normal 29(40.7)c 65(81.2)d 6(22.2) 55(68.8)
Kappa 0.414*** 0.462***
MNA-T1-LF
Atrisk 40(64.5) 18(20.0) 48(77.4) 33(36.7) 57(91.9) 15(16.7)
Normal 22(35.5) 72(80.0) 14(23.6) 57(63.3) 5(8.1) 75(83.3)
Kappa 0.450*** 0.389*** 0.734***
MNA-T2-LF
Atrisk 38(69.1) 20(20.6) 44(80.0) 37(38.1) 51(92.7) 21(21.6)
Normal 17(30.9) 77(79.4) 11(20.0) 60(61.9) 4(7.3) 76(78.4)
Kappa 0.479*** 0.380*** 0.666***
MNA-T1-SF
Atrisk 28(57.1) 30(29.1) 38(77.6) 43(41.7) 40(81.2) 32(31.1)
Normal 21(42.9) 73(70.9) 11(22.4) 60(58.3) 9(18.8) 71(68.9)
Kappa 0.267*** 0.306*** 0.450***
MNA-T2-SF
Atrisk 26(17.1) 32(21.1) 30(19.7) 51(33.6) 34(91.9) 38(26.7)
Normal 11(7.2) 83(54.6) 7(4.6) 64(42.1) 3(8.1) 77(73.3)
Kappa 0.356*** 0.262*** 0.446***
Atriskofmalnutritionwasdefinedas<3.8g/dLofserumalbuminor<10mg/dLofcreatinine.
a n(sensitivity).
b n(1 specificity).
cn(1 sensitivity).
d n(specificity).
*** p<0.001.
hand,Afsar et al.(2006) suggested thatthe MNAover- ratedtheproportionofHDpatientsatriskofmalnutrition comparedtotheSGA.Thereason(s)forthesediscrepancies is notapparent at thepresent time, butcould beuser- related.TheSGAhasbeenobservedtobeassociatedwith relativelylargeinter-ratervariations(Eketal.,1996;Vellas etal.,2001)presumablyduetothesubjectiveandnon- quantitativenatureofthetool.Furtherstudiesareneeded toclarifythediscrepancy.
4.2. Comparisonamongscales
TheseresultssuggestthattheMNA-LFversionsareat leastasgoodastheSGAingradingthenutritionalriskof HD patients. This result is in line with theobservation madeina prospectivesingle-blind studywhichshowed that theMNA wasbetterin identifyingmalnutrition in hospitalized (non-ESRD) patients (Barone et al., 2003).
However,resultsofthepresentstudyshowthattheMNA- SFversionsarenotasgoodastheSGAortheMNA-LFin predictingunder-nutritioninHDpatients.Thereasonsfor theMNA-SFto underpredictrisk of malnutritionin HD patientscanessentiallybeexplainedbythecompositionof thescales.TheMNA-LFversionscontain18items(Vellas etal.,1999)(17itemsforMNA-T2)whereastheMNA-SF versionscontainonlythefirst6ofthe18itemsintheMNA (Kaiseretal.,2009).Althoughthe6itemsarethekeyitems, someaspectsimportanttoHDpatientsarenotadequately representedinthe6items.HDpatientshaveratherunique appetite,medicationandfluidintakeproblemsandthey arelesssatisfiedwiththeirhealthandnutritionalstatus.
TheseitemsarenotadequatelyreflectedintheMNA-SF versions.
4.3. AlternativeMNAversions
ResultsofthepresentstudyalsoconfirmagainthatCC isanacceptablealternativetoBMIintheMNA.Overall,the T2versionperformedrelativelywellcomparedtotheT1 versions.ReplacingCCforBMIinthesescaleshascertain advantages. It enables easy completion of these scales withouttheneedformeasuringweightandheightwhich canbequitecumbersomeinfrailelderlyorpersonswith disability.
4.4. Studylimitations
The study has some limitations. (a) Subjects were purposivelyrecruitedfromonlyonehospitalsothesample may not totally represent the entire HD population in Taiwan.(b)Althoughwehavedoneourbesttominimizeit, there is still a potential of bias given that the same investigatorevaluatedparticipantswiththeMNAandthe SGA. (c) Although we have attempted touse the most appropriatereferencescalesand biochemical indicators, no-oneiscertainthattheyareindeedthemostidealones.
(d) Some serum parameters such as C-reactive protein (CRP),atissueinflammationbiomarker,couldhavebeen determined and its impact on appetite and nutritional status beenevaluated. However, CRP is not a routinely measureditem.
5. Conclusion
Both versions of the long-form MNA are effective in gradingtheriskofmalnutritioninpatientsonHD.However, theMNA-SFversionsarenotasgoodastheMNA-LForthe SGA in screening under-nutrition in these patients. Itis advisedthattheMNA-SFnotbeusedforratingtheriskof malnutrition in patientson HD underthe presentform.
EffortsshouldbemadetoadjustorrevisetheMNA-SFforHD patientsbecausetheshort-formversionismuchmoretime- savingandgreatlydesiredinclinicalpractice.
Acknowledgments
The authors wish to thankthe administrators of the dialysis center for permitting the interviews and the participantsfortheircooperationduringthecourseofthis study.
Conflictofinterest.Nonedeclared.
Funding.Thepresentstudyreceivednospecificgrantfrom any fundingagency in thepublic, commercialornot-for- profitsectors.
Ethical approval. Chung Shan Medical University Hospital InstitutionalReviewBoard(#CS09-031).
AppendixA. Supplementarydata
Supplementarydataassociated withthisarticlecanbe found, in the online version, at doi:10.1016/j.ijnurstu.
2011.05.004.
References
Afsar,B.,Sezer,S.,Arat,Z., etal.,2006.ReliabilityofMiniNutritional AssessmentinhemodialysiscomparedwithSubjectiveGlobalAssess- ment.JournalofRenalNutrition16(3),277–282.
Barone,L.,Milosavljevic,M.,Gazibarich,B.,2003.Assessingtheolder person:istheMNAamoreappropriatenutritionalassessmenttool thantheSGA? JournalofNutrition,Health&Aging7(1),13–17.
Campbell,K.L.,Ash,S.,Bauer,J.D.,Davies,P.S.W.,2007.Evaluationof nutritionassessmenttoolscomparedwithbodycellmassforthe assessmentofmalnutritioninchronickidneydisease.JournalofRenal Nutrition17(3),189–195.
Cooper,B.A.,Bartlett,L.H.,Aslani,A., etal.,2002.ValidityofSubjective GlobalAssessmentasanutritionalmarkerinend-stagerenaldisease.
AmericanJournalofKidneyDiseases40(1),126–132.
deMutsert,R.,Grootendorst,D.C.,Boeschoten,E.W., etal.,2009.Sub- jectiveglobalassessmentofnutritionalstatusisstronglyassociated with mortality in chronic dialysis patients. AmericanJournal of ClinicalNutrition89(3),787–793.
Detsky,A.S.,McLaughlin,J.R.,Baker,J.P., etal.,1987.Whatissubjective globalassessmentofnutritionalstatus? JournalofParenteraland EnteralNutrition11(1),8–13.
Dwyer,J.T., Larive,B.,Leung,J., etal., 2005. Arenutritional status indicatorsassociatedwithmortalityintheHemodialysis(HEMO) Study? KidneyInternational68(4),1766–1776.
Ek,A.C.,Unosson,M.,Larsson,J., etal.,1996.Interratervariabilityin subjectivenutritionalassessmentofolderpersonpatients.Scanndi- navianJournalofCaringSciences10(3),163–168.
Enia,G.,Sicuso,C.,Alati,G.,etal.,1993.SubjectiveGlobalAssessmentof nutritionindialysispatients.NephrologyDialysisTransplantation8 (10),1094–2008.
Guigoz,Y.,Langue,S.,Vellas,B.J.,2002.Identifyingtheelderlyatriskfor malnutrition—theMiniNutritionalAssessment.ClinicsinGeriatric Medicine18(4),737–757.
Herselman,M.,Esau,N.,Kruger,J.M., etal.,2010.Relationshipbetween serumproteinandmortalityinadultsonlong-termhemodialysis:
exhaustivereviewandmeta-analysis.Nutrition26(1),10–32.
Honda,H.,Qureshi,A.R.,Heimburger,O.,etal.,2006.Serumalbumin,C- reactiveprotein,interleukin6,andfetuinAasfactorsofmalnutrition, cardiovasculardisease,andmortalityinpatientswithESRD.Amer- icanJournalofKidneyDiseases47(1),139–148.
Jones,C.H.,Wolfenden,R.C.,Wells,L.M.,2004.IsSubjectiveGlobalAssess- ment a reliable measure of nutritional status in hemodialysis?
JournalofRenalNutrition14(1),26–30.
Kaiser,M.J.,Bauer,J.M.,Ramsch,C., etal.,2009.ValidationoftheMini NutritionalAssessmentShort-form(MNA-SF): apracticaltoolfor identification ofnutritional status.JournalofNutrition,Health&
Aging13,782–788.
Kalantar-Zadeh, K.,Kleiner, M.,Dunne, E., etal., 1999.Amodified quantitativeSubjectiveGlobalAssessmentofnutritionfordialysis patients.NephrologyDialysisTransplantation14(7),1732–1738.
Kalantar-Zadeh,K.,Kopple,J.D.,Block,G., etal.,2001.AMalnutrition–
InflammationScoreiscorrelatedwithmorbidityandmortalityin maintenancehemodialysispatients.AmericanJournalofKidneyDis- eases38(6),1251–1263.
Lee,R.D.,Nieman,D.C.,2003.Assessmentofthehospitalizedpatient.In:
NutritionalAssessment,thirded.McGrawHill,NewYork,NY,pp.
216–250(Chapter7).
NationalKidneyFoundation,2000.K/DOQIclinicalpracticeguidelinesfor nutritioninchronicrenalfailure.AmericanJournalofKidneyDiseases 35(Suppl.2),S1–S140.
Phelan,P.J.,O’Kelly,P.,Walshe,J.J.,etal.,2008.Theimportanceofserum albuminandphosphorousaspredictorsofmortalityinESRDpatients.
RenalFailure30(4),423–429.
Pifer,T.B.,McCullough,K.P.,Port,F.K., etal.,2002.Mortalityriskin hemodialysispatientsandchangesinnutritionalindicators:DOPPS.
KidneyInternational62(6),2238–2245.
Qureshi,A.R.,Alvestrand,A.,Danielsson,A.,etal.,1998.Factorspredicting malnutritioninhemodialysispatients:across-sectionalstudy.Kid- neyInternational53(3),773–782.
Rubenstein,L.Z.,Harker,J.O.,Salva,A., etal.,2001.Screeningforunder- nutritioningeriatricpractice:developingtheshort-formMini-Nutri- tionalAssessment(MNA-SF).TheJournalsofGerontology.SeriesA:
BiologicalSciencesandMedicalSciences56(6),M366–M372.
Schober-Halstenberg,H.J.,2009.End-stagerenaldiseaseinagingsocieties:
aglobalperspective.JournalofRenalNutrition19(5Suppl.),S3–S4.
Steiber,A.L.,Kalantar-Zadeh,K.,Secker,D.,etal.,2004.SubjectiveGlobal Assessmentinchronickidneydisease:areview.JournalofRenal Nutrition14(4),191–200.
Steiber,A.,Leon,J.B.,Secker,D., etal.,2007.Multicenterstudyofthe validityandreliabilityofSubjectiveGlobalAssessmentinthehemo- dialysispopulation.JournalofRenalNutrition17(5),336–342.
TaiwanSocietyofNephrology,2010.ClinicalGuidelinesforHemodialysis inTaiwan. December 30, 2004. (accessed1.6.10)http://www.tsn.
org.tw/UI/F/F005.aspx.
Tsai,A.C.,Lu,S.J.,Chang,T.L.,2009.ModifiedMini-NutritionalAssessment caneffectivelyassessthenutritionalstatusofpatientsonhemodia- lysis.JournalofRenalNutrition19(5),380–388.
Tsai,A.C.,Chang,T.L.,Wang,Y.C.,etal.,2010a.Population-specificshort- formMiniNutritionalAssessmentwithBMIorcalf-circumferencecan predictriskofmalnutritionincommunity-livingorinstitutionalized elderlyinTaiwan.JournaloftheAmericanDieteticAssociation110 (9),1328–1334.
Tsai,A.C.,Ho,C.S.,Chang,M.C.,2007.Population-specificanthropometric cut-pointsimprovethefunctionalityoftheMiniNutritionalAssess- ment(MNA)inelderlyTaiwanese.AsiaPacificJournalofClinical Nutrition16(4),656–662.
Tsai,A.C.,Ku,P.Y.,Tsai,J.D.,2008.Population-specificanthropometric cutoffstandardsimprovethefunctionalityoftheMiniNutritional AssessmentwithoutBMIininstitutionalizedelderlyinTaiwan.Jour- nalofNutrition,Health&Aging12(10),696–700.
Tsai,A.C.,Chang,T.L.,Yang,T.W., etal.,2010b.AmodifiedMiniNutri- tionalAssessmentwithoutBMIpredictsnutritionalstatusofcom- munity-livingelderlyinTaiwan.JournalofNutrition,Health&Aging 14(3),183–189.
Vellas,B.,Guigoz, Y.,Garry, P.J., etal., 1999.TheMini Nutritional Assessment(MNA)anditsuseingradingthenutritionalstateof elderlypatients.Nutrition15(2),116–122.
Vellas,B.,Laugue,S.,Andrieu,S.,etal.,2001.Nutritionalassessmentinthe elderly.CurrentOpinioninClinicalNutrition&MetabolicCare4(1), 1–4.
Yang,F.L.,Lee,R.P.,Wang,C.H., etal.,2007.AcohortstudyofSubjective GlobalAssessmentandmortalityinTaiwanesehemodialysispatients.
RenalFailure29(8),997–1001.
