Randomization to a low-carbohydrate diet advice improves health related quality of life compared with a low-fat diet at similar weight-loss in Type 2 diabetes mellitus

Randomization to a low-carbohydrate diet

advice improves health related quality of life

compared with a low-fat diet at similar

weight-loss in Type 2 diabetes mellitus

Hans Guldbrand, Torbjörn Lindström, B. Dizdar, B. Bunjaku, Carl Johan Östgren, Fredrik H.

Nyström and Margareta Bachrach-Lindström

Linköping University Post Print

N.B.: When citing this work, cite the original article.

Original Publication:

Hans Guldbrand, Torbjörn Lindström, B. Dizdar, B. Bunjaku, Carl Johan Östgren, Fredrik H.

Nyström and Margareta Bachrach-Lindström, Randomization to a low-carbohydrate diet advice

improves health related quality of life compared with a low-fat diet at similar weight-loss in

Type 2 diabetes mellitus, 2014, Diabetes Research and Clinical Practice, (106), 2, 221-227.

http://dx.doi.org/10.1016/j.diabres.2014.08.032

Copyright: Elsevier

http://www.elsevier.com/

Postprint available at: Linköping University Electronic Press

Randomization

to

a

low-carbohydrate

diet

advice

improves

health

related

quality

of

life

compared

with

a

low-fat

diet

at

similar

weight-loss

in

Type

2

diabetes

mellitus

§

,§§

H.

Guldbrand

,

T.

Lindstro¨m

,

B.

Dizdar

,

B.

Bunjaku

,

C.J.

O¨stgren

,

F.H.

Nystrom

,

M.

Bachrach-Lindstro¨m

*

CountyCouncilofO¨stergo¨tland,Linko¨ping,Sweden

diabetes research andclinical practice 106(2014) 221–227

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received28January2014 Receivedinrevisedform 29July2014

Accepted30August2014

Availableonline21September2014 Keywords:

Type2diabetesmellitus Dietaryintervention Low-carbohydratediet SF-36

a

b

s

t

r

a

c

t

Aims:Tocomparetheeffectsonhealth-relatedqualityoflife(HRQoL)ofa2-year interven-tionwithalow-fatdiet(LFD)oralow-carbohydratediet(LCD)basedonfourgroup-meetings toachieve compliance.Todescribedifferentaspectsof takingpartinthe intervention followingtheLFDorLCD.

Methods:Prospective,randomizedtrialof61adultswithType2diabetesmellitus.TheSF-36 questionnairewasusedatbaseline,6,12and24months.PatientsonLFDaimedfor55–60 energypercent(E%)andthoseonLCDfor20E%fromcarbohydrates.Thepatientswere interviewedabouttheirexperiencesoftheintervention.

Results: Meanbody-mass-indexwas32.75.4kg/m2_{atbaseline.Weight-lossdidnotdiffer}

betweengroupsandwasmaximalat6months,LFD: 3.994.1kg,LCD: 4.313.6kg (p<0.001withingroups).TherewasanincreaseinthephysicalcomponentscoreofSF-36 from44.1(10.0)to46.7(10.5)at12monthsintheLCDgroup(p<0.009)whilenochange occurredintheLFDgroup(p<0.03betweengroups).At12monthsthephysicalfunction, bodilypainandgeneralhealthscoresimprovedwithintheLCDgroup(pvalues0.042–0.009) whiletherewasnochangewithintheLFDgroup.

Conclusions: Weight-changesdidnotdifferbetweenthedietgroupswhileimprovementsin HRQoL onlyoccurredafteroneyearduring treatmentwithLCD.Nochanges ofHRQoL occurredintheLFDgroupinspiteofasimilarreductioninbodyweight.

#_{2014TheAuthors.PublishedbyElsevierIrelandLtd.Thisisanopenaccessarticleunder} theCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/3.0/).

§

Trialregistrynumber:NCT01005498atClinicalTrials.gov.

§§

ThisstudyhasbeenpresentedasanabstractatEASD24September2013,Barcelona,Spain.

*Correspondingauthorat:DepartmentofMedicalandHealthSciences,Linko¨pingUniversity,SE58185Linko¨ping,Sweden. Tel.:+4613285827;fax:+4613145004.

E-mailaddress:margareta.bachrach-lindstrom@liu.se(M.Bachrach-Lindstro¨m).

Abbreviations: SF-36,ShortForm36;LCD,low-carbohydratediet;LFD,low-fatdiet;HRQoL,health-relatedqualityoflife;PCS,physical componentscore;MCS,mentalcomponentscore.

ContentsavailableatScienceDirect

Diabetes

Research

and

Clinical

Practice

j o u r n a lh o m e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / d i a b r e s

http://dx.doi.org/10.1016/j.diabres.2014.08.032

0168-8227/#2014TheAuthors.PublishedbyElsevierIrelandLtd.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http:// creativecommons.org/licenses/by-nc-nd/3.0/).

1.

Introduction

Severalstudieshaveshownthatbothindividualswithobesity

[1–3]and Type2diabetes[4,5]have loweredhealth-related quality of life (HRQoL) in comparison with normal-weight individuals. There is also an inverse relationship between degreeofobesitymeasuredwithBMIandHRQoL[6].Boththe physicalcomponentscore(PCS)andmentalcomponentscore (MCS) of the Short Form 36 (SF-36) show lower values in obesity,which indicateslowerHRQoL,butthe differenceis moreprominent inthe PCS[7]. Inthe Australian Diabetes, ObesityandLifestylestudy[8]obesityatbaselinewasrelated toadecreasedHRQoL,whenfollowedduring5years,andlow HRQoLwasalsoapredictorofweightgainduringthe follow-upperiod.Whilethescoresarelowerforpatientswhohave developeddiabeticcomplications[9],aloweredscorecanalso befoundinpatientswithnewlydiagnoseddiabeteswithout anydiabetes-relatedcomplications[10]whichshowsthatitis notonlyaresultofsuchcomplications.Therelationshipis dualaslowscoresontheSF-36havebeenfoundtopredictthe riskoflaterdevelopmentofType2diabetes,cardiovascular diseaseandmortality[11].InthestudybyWilliamsetal.,low PCS on the SF-36 was related to further increase in cardiovascular mortality in patients with diabetes and a similarassociation was foundforthe MCS.Therearethus interactions betweenHRQoLand bothType2diabetesand obesity.

Intentional weight-loss in both obesity and in Type 2 diabetesleadstoimprovementsincardiovascularriskfactors, and hasalso been foundto improveHRQoL[12,13]. Inthe SHIELDstudyimprovementofqualityoflifewasfoundinthe participantswithType2diabeteswhohadlostweightduring the last 12 months compared with respondents reporting weight gain [14]. Also weight reduction by gastric bypass, Roux-en-Ygastricbypasssurgery,hasbeenshowntoimprove HRQoLwhenfollowedupafter2yearspostoperatively[15,16]. Somestudieshave prospectivelycomparedthe effectof different diet regimens in patients with Type 2 diabetes mellitus,withthemainfocusbeingchangeofbodyweightand ofcardiovascularriskfactors.Thereareveryfewreportson HRQoLinsuchcomparative studies. InastudyBrinkworth et al.comparedthe effectsonProfile ofMoodStates,Beck DepressionInventory,andSpielbergerStateAnxiety Invento-ryscoreinoverweightorobese participantsrandomizedto low-fatorlow-carbohydratediet,andfoundgreater improve-mentofpsychologicalmoodinthelowfatgroup[17].However, diabeteswasanexclusioncriterioninthestudybyBrinkworth etal.

We performeda randomizedstudyconfinedto patients withType2diabetesmellitustocompareglycaemiccontrol andalsoofweight-lossandcardiovascularriskfactorsofa low-carbohydratedietwiththatofatraditionallow-fatdiet. The results on body weight, glycaemic control and other cardiovascularriskfactorshavebeenpreviouslyreported[18]. Incontrasttomostpreviousstudies,thepatientsrandomized to the low-carbohydrate diet were not advised to avoid saturatedfat. The interventionswere based onfour group meetingswithdurationof60mineachforthefirstyearandno furthergroupmeetingsduringtheremaining12monthswere

given. Both reduction ofenergy intakeand changes ofthe macronutrientcompositionmight,bynotwell-known mech-anisms, affectHRQoL[19]. Inthesefurtheranalysesofour study, the aim was to compare effects on health-related qualityoflifeduringalow-carbohydratedietcomparedwitha traditional low-fat diet in patients with Type 2 diabetes mellitus.Asecondaimwastodescribedifferentaspectsfrom takingpartoftheinterventionfollowingthelow-fator low-carbohydratediet.

2.

Materials

and

methods

Themethodshavebeendescribedbefore[18].Inshortpatients withType2diabeteswereincludedinthestudy,whichwas conducted at twoprimary healthcare centresinsoutheast Sweden. The patients were randomized either to a low-carbohydratedietortoatraditionallow-fatdiet,bothwitha caloriccontentof1600kcalforwomenor1800kcalformen. Randomizationwasnotstratified,andwasbasedondrawing blinded ballots. The low-carbohydrate diet had an energy contentwhere50E%wasfat,20E%carbohydratesand30E% protein.Thelow-fatdiethadanutrientcompositionthatwas similartowhatistraditionallyrecommendedfortreatmentof Type 2diabetesinSweden with30E%fat (lessthan 10E% saturatedfat),55–60E%carbohydratesand10–15E%protein. No information was given to change the level ofphysical activityoftheparticipants.

Investigations of anthropometrics and laboratory tests wereperformedatbaselineandat6,12and24months,and patientswerealsoaskedtofill-outquestionnaireson health-relatedqualityoflife(SF-36)atthesetime-points.Dietrecords were also performed at these 4 visits with one additional recordingat3months.

ThegenericShortForm-36(SF-36)questionnairedesigned tomeasureindividualsHRQoLinclinicalpractice,research, healthpolicyevaluationsandgeneralpopulationsurveyswas used. The 36 item questionnaire comprise eight health domains;physicalfunctioning(PF,10items),rolelimitations due to physical problems (RP, 4 items), bodily pain (BP, 2 items),generalhealth(GH,5items),vitality(VT,4items),social function (SF, 2 items), role limitations due to emotional problems(RE,3items),andmentalhealth(MH,5items)and onesingleitemratinghealthstatusoveroneyear[20].Each domain is separately scored and transformed in values between 0 and 100 where a higher score indicates higher HRQoL. ThecombinedPhysical componentscore(PCS)and Mentalcomponentscore(MCS)werecalculated.Both reliabil-ityandvalidityhasbeenextensivelyevaluatedunderSwedish conditions[21].Noimputationofdatawasdoneinthecaseofa missingquestionnaire.

The participants were interviewed following a semi-structured interview guide with eight questions regarding differentaspectsoftakingpartoftheinterventionfollowing the low-fat or low-carbohydrate diet. The answers were written down by the interviewer duringthe interview and the text was analyzedusing conventional contentanalysis followingHsiehandShannon[22].At12months,butnotat other time points, they also answered 3 VAS-scales about appetiteandsatiety.

diabetes researchandclinicalpractice 106(2014) 221–227

2.1. Statistics

Statisticalcalculationsweredone withPASW20.0software (SPSS Inc., Chicago, IL, USA). Linear correlations were calculated, as stated in the text. Comparisons within and between groups were done with Student’s paired and unpaired2-tailed t-testor inthe caseofquestionnaires by WilcoxonandMann–Whitneytests.Mean(SD)isgivenunless otherwisestated.Statisticalsignificancewasconsideredtobe present at the 5% level (p0.05). ANOVA with repeated measureswasusedforcalculationsofthechangesduringthe totalstudyduration.

2.2. Ethics

ThestudywasapprovedbytheRegionalEthicsCommitteeof Linko¨pingandperformedinaccordancewiththeDeclaration ofHelsinki.Writteninformedconsentwasobtainedfromall participating subjects. The study was registered with trial numberNCT01005498atClinicalTrials.gov.

3.

Results

3.1. Totalstudypopulation

Thestudypopulationandtheflow chartofthe studyhave beenpreviouslydescribedindetail[18]. Intotal61patients enteredthestudy,andat baselinetheSF-36questionnaires werecompletedby30patientsinthelow-fatgroupandby30 patientsin the low-carbohydrate group. Thequestionnaire wasansweredby22patientsat6months,28patientsat12 monthsand29patientsat24monthsintheLFDgroup,while thecorrespondingfiguresfortheLCDgroupwere23,27and25, respectively.

Themeanageinthelow-fatgroupwas62.711yearsand therewere13menand18womenhavingadiabetes-duration of8.86.2years.Correspondingfiguresforthe low-carbohy-drategroupwere61.29.5years,14menand16womenanda known diabetes-duration of 9.85.5 years. Age, gender compositionandknowndurationofdiabetesdidnotdiffer betweenthegroups(allp>0.05).

AmongthepatientswhoansweredtheSF-36questionnaire bodyweightintheLFDgroupwas98.821.6kgatbaseline, 93.719.3kgat 6months,92.918.7kgat 12monthsand 95.021.5kgat24months.CorrespondingfiguresfortheLCD group were 91.418.9kg, 84.818.6kg, 88.419.3kg and 85.216.1kg.Therewerenodifferencesinweightreduction betweenthesegroups(at6months:LFDgroup: 4.54.5kg, LCDgroup: 4.53.6kg,p=0.99,at 12months:LFDgroup: 3.44.3kg,LCDgroup: 3.14.1kg,p=0.84,at24months LFD group: 3.44.5kg,LCD group: 3.24.3kg, p=0.85). Weightreductionwithineithergroupwassignificantbetween baselineandthesametimepoints(p<0.001).

TheresultsofHRQoLaredescribedinTable1.Therewasno differenceinbaselinevaluesinanyofthedomainsbetween thelow-fatandlowcarbohydrategroups.The low-carbohy-drateshowedimprovementsinthePhysicalfunction,Bodily Pain,GeneralHealthandVitalitydomainsoftheSF-36at12 months compared to baseline (p values 0.042–0.009). In

contrasttothis,nochangewasobservedinthelow-fatgroup atanytimeduringthestudy.Thechangefrombaselineto12 months showed improvements inBodily Pain and General Healthinthelow-carbohydrategroupincomparisonwiththe low-fatgroup(p=0.017and0.022,respectively).At12months there was also an improvement inthe combined Physical Component Score (PCS) in the low-carbohydrate group compared tobaseline(p=0.009) and alsocompared tothe low-fat group (p=0.028) (Fig. 1). We also recalculated the results basedon the patients whowerejudged tobe fully compliantwiththeprescribeddietbythecriteriaofintakeof fat>40E%togetherwithcarbohydrates<40E%fortheLCD and fat<40E%togetherwithcarbohydrates>40E%forthe LFD, and this did not change the main results (data not shown).

In multiple linear regression analyses usingchange in PCSasdependentvariableandchangesinBMIandHbA1cas independentvariables,thechangeinPCSremained associ-atedwith thechangeinBMIintheLFDgroupat6and12 months (b= 0.872, p=0.001, respectively b= 0.679, p<0.0005)andthechangeinHbA1cintheLFDgroupat6 months(b=0˙ 496,p=0.028).Therewerenoassociationsin changesintheLCDgroupinthesameanalyses.Therewere noassociationsbetweenregainofbodyweightfrom6to12 months and changes of SF-36 variables during the same period (data not shown). Changes in Mental Component Score(MCS)showednoassociationstochangesinBMIand HbA1cinlinearregressionanalysisat6and12months,for bothgroups.

TheVASscale‘‘Iwillsucceed/notsucceedwiththisdiet’’ showedsimilarresults6.5(2.6)inthelowfatgroupand7.4(2.7) inthelowcarbohydrategroupat12months(p=0.95).TheVAS scale‘‘Iamoftenhungry/satisfied’’showedsimilarresults6.6 (2.7)inthelowfatgroupand7.2(2.7)inthelowcarbohydrate group(p=0.81).TheVASscale ‘‘Ihavecravingforfood/not cravingforfood’’alsoshowedsimilarresults6.4(3.1)inthe low fat group and 6.2 (3.4) in the low carbohydrate group (p=0.38).

3.2. Interview

Eachtranscriptwasreadfrombeginningtoendseveraltimes andkeywordsorphraseswerehighlightedandcoded.Four categoriesemergedduringthisprocessthatdescribes differ-entaspectsoftakingpartintheintervention.

3.3. Bigandsmallchanges

Participants inbothdiet-groups did notfindthe necessary changes very burdensome. In the low-fat group it was commontohavebeeneatingquitesimilarbeforetakingpart ofthestudy.Wheneatingoutitcouldbehardtofollowthediet strictly.

Theparticipantsthatfoundthechangesdifficultwereonly foundinthelowcarbohydrate-dietgroup,andpotatoeswere mentionedashardtorestrainfromaswellaschangefrom low-fat tohigh fatproducts.Participantsinthisgroupalso founditdifficulttosubstitutepastaandpotatoesaswellas cookiesandsnackswithinthedietprovided.Itwashardtoeat somuchfattyproducts.

Table 1 – SF-36 at baseline, 6, 12 and 24 months in 30 patients with Type 2 diabetes during treatment with low fat diet and in 30 patients with Type 2 diabetes during treatment with low carbohydrate diet (means (SD); Wilcoxon signed rank test and Mann–Whitney U test). The questionnaire was answered by 22 patients at 6 months, 28 patients at 12 months and 29 patients at 24 months in the low fat group while the corresponding figures for the low carbohydrate group were 23, 27 and 25, respectively. SF-36 domains Diet baseline 6 months P value 0–6 12 months P value 0–12 24 months P value 0–24 P-value between groups 0 and 6 P-value between groups 0 and 12 P value between groups 0 and 24 Friedman’s 2-way analysis of variance Physical function Low fat 79.8 (20.0) 84.5 (12.1) 0.65 83.8 (15.7) 0.10 81.6 (17.7) 0.62 0.48 0.27 0.38 0.49 Low carb 77.7 (19.3) 79.4 (15.6) 0.15 83.6 (18.2) 0.009 78.7 (19.7) 0.12 0.004 Role Physical Low fat 78.1 (35.9) 84.5 (26.8) 0.55 87.5 (27.6) 0.34 67.8 (42.7) 0.13 0.68 0.89 0.87 0.042 Low carb 79.5 (29.7) 70.5 (39.1) 0.78 78.7 (33.0) 0.64 66.3 (43.7) 0.13 0.60

Bodily Pain Low fat 69.3 (27.1) 66.2 (22.3) 0.49 65.7 (26.5) 0.59 61.6 (28.34) 0.11 0.85 0.017 0.12 0.12

Low carb 61.0 (23.0) 61.0 (25.0) 0.95 71.4 (22.1) 0.021 60.6 (25.6) 0.72 0.031

General Health

Low fat 62.5 (21.1) 67.7 (18.2) 0.41 63.3 (18.4) 0.46 66.1 (23.4) 0.51 0.85 0.022 0.88 0.40

Low carb 63.2 (22.3) 63.5 (25.6) 0.29 70.7 (22.7) 0.031 63.8 (26.7) 0.55 0.031

Vitality Low fat 65.0 (20.3) 71.0 (17.4) 0.66 66.9 (22.9) 0.83 67.8 (23.9) 0.49 0.39 0.12 0.75 0.87

Low carb 62.2 (19.9) 65.5 (22.4) 0.16 69.8 (19.3) 0.042 61.2 (23.9) 0.92 0.09 Social Function Low fat 91.7 (14.8) 93.2 (18.0) 0.57 93.8 (15.4) 0.54 88.3 (20.3) 0.18 0.94 0.96 0.69 0.73 Low carb 90.8 (15.7) 87.0 (23.7) 0.72 92.1 (15.2) 0.34 88.0 (19.9) 0.39 0.41 Role Emotional Low fat 83.3 (33.6) 90.5 (26.1) 0.71 91.7 (26.6) 0.10 85.1 (34.0) 0.86 0.50 0.67 0.73 0.67 Low carb 90.8 (17.6) 78.8 (35.0) 0.20 91.4 (21.9) 0.73 93.9 (19.6) 0.91 0.06 Mental Health Low fat 83.3 (15.8) 84.5(17.0) 0.80 83.7 (17.4) 0.57 82.3 (16.9) 0.78 0.40 0.93 0.59 0.14 Low carb 80.1 (13.4) 81.6 (17.) 0.28 83.2 (10.6) 0.54 81.0 (10.7) 0.78 0.35 PCS Low fat 45.3 (10.5) 45.8 (8.2) 0.88 45.9 (8.9) 0.86 43.6 (10.5) 0.17 0.23 0.028 0.44 0.35 Low carb 44.1 (10.0) 43.2 (12.4) 0.10 46.7 (10.5) 0.009 41.4 (14.0) 0.77 0.011 MCS Low fat 51.7 (9.8) 53.5 (10.1) 0.50 52.8 (9.5) 0.64 52.0 (9.4) 0.91 0.43 0.64 0.58 0.28 Low carb 51.7 (7.2) 50.0 (13.0) 0.33 52.6 (5.3) 0.23 53.1 (4.2) 0.48 0.66

Bold values denotes significance ( p < 0.05). If you find this helpful, please, include this in the legend to the figure.

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 1 0 6 ( 2 0 1 4 ) 2 2 1 – 2 2 7

224

Participantsinbothdiet-groupsfoundtherecipesprovided tobe‘‘dull’’andwantedamorevariedfoodselection.They alsofeltguiltywhennotstrictlyfollowingthediet.

3.4. Gains

Both diet-groups mentionedexpected gains inhealth with theirdiet.

In the low-fat group the diet was described as easy to follow,tastyandcheapinprice.

Thelow-carbohydrategroupdescribedthattheyfeltless hungryandwerelesspronetosweets.

3.5. Strategies

Bothdiet-groupsmentioneddifficultieswheneatinginother people’s homes or when going to restaurants and during holidays.Strategiesmentionedweretoeatverylittleofthe food not allowed. During Christmas it was easy to find alternativestoeat.

3.6. Support

Both diet-groups mentioned that it was supportive if the whole family could follow the same diet. In the low-carbohydrategroupthefamiliessometimeschoosetoprepare twodifferentmealsatatime.

4.

Discussion

Inobesity,withorwithoutcoexistingType2diabetesmellitus, intentionalreductionofbodyweightisconsideredtoimprove health-related qualityof life[1,3,4]and it was therefore of interestthatnosucheffectcouldbefoundaccordingtoSF-36 after6monthswhichwasthetimewhenthereductionofbody weightinbothdietgroupswasmaximalinthepresentstudy.

Later,after12months,improvementsofHRQoLwerefound only in the low carbohydrate group. The improvements relatedtophysicalfunctionandtovitalityandgeneralhealth, while the domains relating to mental health showed no change.IntheSF-36,adifferenceoffivepointsinanindividual domainor2–3pointsinPCSorMCSareconsideredclinically significant[23,24].ThechangesofthesesubscalesandofPCS that were found in our study are thus considered to be clinicallysignificant.Aloweredenergyintakeof1600kcalfor womenand1800kcalformenwasprescribedforbothstudy groups,butthemainfocusinthestudywasthecomparisonof adietwithalowcarbohydrateintakewithalowfatdiet,which isthetraditionaldietrecommendedfortreatmentofType2 diabetesmellitus. Inrecent yearstherehasbeen increased interestintheuseoflow-carbohydratedietsinthetreatment ofdiabetes,butwearenotawareofanypreviousstudythat has addressed theeffects on HRQoLmeasured bySF-36in patientswithType2diabetesmellitusrandomizedtoa low-carbohydratediet.

Inobesesubjects,lowervaluesofthescalesoftheSF-36 relating to physical function than in normal weight and overweightindividualsarefound[7],anditisthereforelikely toseeanimprovementinthesescalesduringanintervention. There aredifferent possible mechanisms for the improve-mentsoftheseduringthestudyincludingreductionofbody weight in the patients, who had a mean BMI of 32 when entering the study, improvementofglycaemic control, but alsothechangeofmacronutrientsperse.Ourfindingthatthere was nochangeinSF-36after6months,whenreduction of bodyweightwasmaximalinbothgroups,arguesagainstthis beingonlyaneffectofweightreduction.Alimitationofour studyisthatnotallpatientsfilledoutthequestionnaireatthis timewhiletheansweringfrequencywasgoodat12months. This makes the6 monthsresultsmoreuncertain. Another limitationisthatafewpatientswerenotfullycompliantwith their dietsthroughout thestudybut calculationsexcluding thesepatientsshowedsimilarSF-36results.Furthermorethe

0 10 20 30 40 50 60 70 PCS LFD PCS LCD MCS LFD MCS LCD 0 months 6 months 12 months 24 months **=p<0.01 vs baseline **

Fig.1–Physicalcomponentscore(PCS)andmentalcomponentscore(MCS)ofSF-36atbaseline,6,12and24monthsin30 patientswithType2diabetesduringtreatmentwithlowfatdiet(LFD)andin30patientswithType2diabetesduring treatmentwithlowcarbohydratediet(LCD)(means(SD);WilcoxonsignedranktestandMann–WhitneyUtest).The questionnairewasansweredby22patientsat6months,28patientsat12monthsand29patientsat24monthsinthelow fatgroupwhilethecorrespondingfiguresforthelowcarbohydrategroupwere23,27and25,respectively.

magnitudeofweightreductionwassimilarinbothgroups,but theimprovementinHRQoLoccurredonlyintheLCDgroup.In theLFDgroupassociationswerefoundbetweenchangesofPCS andbothchangesofBMIandHbA1c,whilenosuchassociations werefoundintheLCDgroupwhichmightsuggestthatother factors,possiblythedietperse,mighthavebeenofimportance fortheimprovementsfoundinPCSintheLCDgroup.

Inthisstudynoscalespecificallyaddressinganxietyand depressionwasused,butwecouldnotfindsupportforthatthe lowcarbohydratedietcauseddeteriorationofthementalscores ofSF-36.TheMCSwasunchangedinbothstudygroupsatall timesevaluated.AsmentionedBrinkworthetal.[17]studiedthe effectofaLCDandaLFDforoneyearonqualityoflife,butthey didnotincludeSF-36.Theyreportedshort-termresultsupto8 weeksshowingveryrapid,alreadyafter2weeks,andsimilar improvementsinmoodforbothgroups[25].Whenthepatients werefollowedforalongerperiod,uptooneyear,someofthe scores returned towards baseline in the LCD group, and a significantdifferenceemergedcomparedwiththeLFDgroup

[17].Theseresultsmightseemcontradictorytoourresultsbut there aremajor differences betweenthis studyand ours.A majordifferencewasthataverylowcarbohydratedietaiming atonly4%oftheenergyintakeconsistingofcarbohydrateswas used while we prescribed a more moderate carbohydrate restrictionto20E%andreached26–31E%accordingtothediet registrations.Anotherdifferencewasthatintheirstudyobese participants were included but diabetes was an exclusion criterion.Alsothereductionofbodyweightwasgreater,almost 14kgattheterminationofthestudy,whichmighthavebeena consequenceofmoreintenselifestyletreatmentregimen.From theresultsofourstudyitcanbeconcludedthataimingformore moderatereductionofcarbohydratesisfeasibleinpatientswith Type2diabetesinroutineprimarycare,andhasnoadverse effects on the mental aspects of HRQoL. This is also in agreement with a study by Davis et al. [26] who found a tendencytowardsimprovementoftheanxietyandworryitems oftheDiabetes-39questionnaireduringadietinterventionin Type2diabetes.Thereareveryfewstudiesofshortduration thathavecomparedtheeffectsofdifferentdietsonHRQoLand moodinobesity.Yancyetal.[27]foundthattreatmentwitha lowcarbohydratedietimprovedtheMCSinSF-36inoverweight volunteers.Inasmall3-weekstudyd’Ancietal.foundworse performanceonmemory-basedtasksandlessconfusionduring a low-carbohydrate diet compared with a low-fat diet. The studywasperformedinoverweightorobesewomen.Inour study we did not measure acute effects on HRQoL, but in contrastfoundnochangesinanyofthegroupsinSF-36after6 months,whichmightsuggestthatlongertreatmentis neces-sarytoobtainimprovementsofHRQoL.

IntheLookAHEADstudy,changesofdepressivesymptoms wereevaluatedatbaselineandafteroneyearwiththeBeck Depressive Inventory [12]. Thelifestyle intervention inthe patientswhohadType2diabeteswasbasedonalowfatdietin combinationwithincreasedphysicalactivity.Itwas conclud-edthatintentionalweightlosswasnotassociatedwiththe precipitationofdepressivesymptoms,butinsteadappearedto protectagainstthisoccurrence.

The interview suggests that both study groups made relevantchangesoftheirpreviousdiettofollowtheprescribed diet during the study. In general, they did not find these

changes difficult butthe needtoreduceintakeofpotatoes whichisamaincarbohydrate-richingredientofSwedishdiet (Itiseateninaverage0.7timesadayaccordingtoasurvey presentedbytheNationalFoodAgency)waspointedoutas troublesome by the LCD group. Reduction ofcarbohydrate intake whenchangingto avery lowcarbohydrate diethas been describedastroublesome[28]andeven themoderate reduction of carbohydrates in our study caused some problemswhenchangingfromlow-fattohigh-fatproducts, as wellas substitute cookiesand snacks. Awish formore variedfoodrecipeswaspronouncedbybothgroups.

Bothstudygroupsmentionedhopesthattheirdietwould incurhealthbenefitsandfeltguiltywhennotstrictlyadhering toit.Benefitsforthelowfatdietwasthatitwastastyand cheapinpricewhilefeelinglesshungryandeasiertoavoid sweetswerementionedinthelowcarbohydrategroup,butit shouldbenotedthatthequestionsbasedonVAS-scales,that wereadministeredonlyat12months,showednodifference betweenthestudygroupsonsatietyandhunger.Bothstudy groupsmentioneddifficultiesfollowingthedietwhen social-izingwithothersi.e.duringholidaysoreatingout.Eatingonly smallamountsof‘‘forbidden’’fooditemswerementionedasa way to handlethese situations. When eating at their own homeitwasmoreconvenientifthewholefamilycouldeatthe same food. When implementing and sustaining dietary changesocialrelationshipswithinandwithoutthehousehold hasanimpactoncompliance.Itseemsimportanttoinvolve thefamilyinthesechanges[29].

Inconclusion,weight-changesdidnotdifferbetweenthe dietgroupswhileimprovementsinHRQoLonlyoccurredafter oneyearduringtreatmentwithLCD.NochangesofHRQoL occurredintheLFDgroupinspiteofasimilarreductionin bodyweight.

Funding

ThestudywassupportedbyUniversityHospitalofLinko¨ping ResearchFunds,Linko¨pingUniversity,theCountyCouncilof O¨ stergo¨tland,andtheDiabetesResearchCentreofLinko¨ping University.

Conflict

of

interest

The authors declare that there is no conflict of interest associatedwiththismanuscript.

Acknowledgements

We thank Anna-Karin Scho¨ld and Maja Holm, Christina Andersson,MaudArnemyrandBirgittaBo¨ttinger,allRNfor theirvaluableworkwiththisstudy.

r

e

f

e

r

e

n

c

e

s

[1] HassanMK,JoshiAV,MadhavanSS,AmonkarMM.Obesity andhealth-relatedqualityoflife:across-sectionalanalysis

diabetes researchandclinicalpractice 106(2014) 221–227

oftheUSpopulation.IntJObesRelatMetabDisord 2003;27:1227–32.

[2] SachTH,BartonGR,DohertyM,MuirKR,JenkinsonC, AveryAJ,etal.Therelationshipbetweenbodymassindex andhealth-relatedqualityoflife:comparingtheEQ-5D, EuroQolVASandSF-6D.IntJObes(Lond)2007;31:189–96.

[3] BaiardiF,PugliaMG,ValpianiG,SturaniA,BaraccaniC, GentileM,etal.Relationshipbetweenbodyweight,quality oflifeandcardiovascularriskfactors:ageneralpopulation basedItalianexperience.EatWeightDisord2005;10:19–24.

[4] EckertK.Impactofphysicalactivityandbodyweighton health-relatedqualityoflifeinpeoplewithType2diabetes. DiabetesMetabSyndrObes2012;5:303–11.

[5] WingRR,MarcusMD,BlairEH,BurtonLR.Psychological responsesofobesetypeIIdiabeticsubjectsto very-low-caloriediet.DiabetesCare1991;14:596–9.

[6] FontaineKR,BarofskyI.Obesityandhealth-relatedquality oflife.ObesRev2001;2:173–82.

[7] Ul-HaqZ,MackayDF,FenwickE,PellJP.Meta-analysisof theassociationbetweenbodymassindexand health-relatedqualityoflifeamongadults,assessedbytheSF-36. Obesity(SilverSpring)2013;21:E322–7.

[8] CameronAJ,MaglianoDJ,DunstanDW,ZimmetPZ, HeskethK,PeetersA,etal.Abi-directionalrelationship betweenobesityandhealth-relatedqualityoflife:evidence fromthelongitudinalAusDiabstudy.IntJObes(Lond) 2012;36:295–303.

[9] RejeskiWJ,LangW,NeibergRH,VanDorstenB,FosterGD, MatthewL,etal.Correlatesofhealth-relatedqualityoflife inoverweightandobeseadultswithType2diabetes. Obesity(SilverSpring)2006;14:870–83.

[10] TappRJ,DunstanDW,PhillipsP,TonkinA,ZimmetPZ, ShawJE,etal.Associationbetweenimpairedglucose metabolismandqualityoflife:resultsfromtheAustralian diabetesobesityandlifestylestudy.DiabetesResClinPract 2006;74:154–61.

[11] WilliamsED,RawalL,OldenburgBF,RenwickC,ShawJE, TappRJ,etal.Riskofcardiovascularandall-cause mortality:impactofimpairedhealth-relatedfunctioning anddiabetes:theAustralianDiabetes,ObesityandLifestyle (AusDiab)study.DiabetesCare2012;35:1067–73.

[12] FaulconbridgeLF,WaddenTA,RubinRR,WingRR,Walkup MP,FabricatoreAN,etal.One-yearchangesinsymptoms ofdepressionandweightinoverweight/obeseindividuals withType2diabetesintheLookAHEADstudy.Obesity (SilverSpring)2012;20:783–93.

[13] SamsaGP,KolotkinRL,WilliamsGR,NguyenMH,Mendel CM.Effectofmoderateweightlossonhealth-related qualityoflife:ananalysisofcombineddatafrom4 randomizedtrialsofsibutraminevsplacebo.AmJManage Care2001;7:875–83.

[14] GrandyS,FoxKM,BazataDD.Associationofself-reported weightchangeandqualityoflife,andexerciseandweight managementbehaviorsamongadultswithType2diabetes mellitus:theSHIELDstudy.CardiolResPract

2012;2012:892564.

[15] AdamsTD,PendletonRC,StrongMB,KolotkinRL,Walker JM,LitwinSE,etal.Healthoutcomesofgastricbypass patientscomparedtononsurgical,nonintervenedseverely obese.Obesity(SilverSpring)2010;18:121–30.

[16] KarlssonJ,TaftC,RydenA,SjostromL,SullivanM. Ten-yeartrendsinhealth-relatedqualityoflifeaftersurgical andconventionaltreatmentforsevereobesity:theSOS interventionstudy.IntJObes(Lond)2007;31:1248–61.

[17] BrinkworthGD,BuckleyJD,NoakesM,CliftonPM,Wilson CJ.Long-termeffectsofaverylow-carbohydratedietanda low-fatdietonmoodandcognitivefunction.ArchIntern Med2009;169:1873–80.

[18] GuldbrandH,DizdarB,BunjakuB,LindstromT, Bachrach-LindstromM,FredriksonM,etal.InType2diabetes, randomisationtoadvicetofollowalow-carbohydratediet transientlyimprovesglycaemiccontrolcomparedwith advicetofollowalow-fatdietproducingasimilarweight loss.Diabetologia2012;55:2118–27.

[19] CarsonTL,HidalgoB,ArdJD,AffusoO.Dietary

interventionsandqualityoflife:asystematicreviewofthe literature.JNutrEducBehav2014;46:90–1.

[20] WareJrJE,SherbourneCD.TheMOS36-itemshort-form healthsurvey(SF-36):I.Conceptualframeworkanditem selection.MedCare1992;30:473–83.

[21] SullivanM,KarlssonJ,WareJrJE.TheSwedishSF-36 HealthSurvey:I.Evaluationofdataquality,scaling assumptions,reliabilityandconstructvalidityacross generalpopulationsinSweden.SocSciMed1995;41: 1349–58.

[22] HsiehHF,ShannonSE.Threeapproachestoqualitative contentanalysis.QualHealthRes2005;15:1277–88.

[23] WareJE,editor.SF-36healthsurveymanualand interpretationguide.Boston:TheHealthInstituteNew EnglandMedicalCenter;1993.

[24] WareJE,editor.SF-36physicalandmentalhealthsummary scales:auser’smanual.Boston:TheHealthInstituteNew EnglandMedicalCenter;1994.

[25] HalyburtonAK,BrinkworthGD,WilsonCJ,NoakesM, BuckleyJD,KeoghJB,etal.Low-andhigh-carbohydrate weight-lossdietshavesimilareffectsonmoodbutnot cognitiveperformance.AmJClinNutr2007;86:580–7.

[26] DavisNJ,TomutaN,IsasiCR,LeungV,Wylie-RosettJ. Diabetes-specificqualityoflifeafteralow-carbohydrate andlow-fatdietaryintervention.DiabetesEduc 2012;38:250–5.

[27] YancyJrWS,AlmirallD,MaciejewskiML,KolotkinRL, McDuffieJR,WestmanEC,etal.Effectsoftwoweight-loss dietsonhealth-relatedqualityoflife.QualLifeRes 2009;18:281–9.

[28] AlhassanS,KimS,BersaminA,KingAC,GardnerCD. Dietaryadherenceandweightlosssuccessamong overweightwomen:resultsfromtheAtoZweightloss study.IntJObes2008;32:985–91.

[29] RydenPJ,SydnerYM.Implementingandsustainingdietary changeinthecontextofsocialrelationships.ScandJCaring Sci2011;25:583–90.