Non-alcoholic fatty liver disease and neurological defects

(1)

by Elsevier. This article is free for everyone to access, download and

read.

Any restrictions on use, including any restrictions on further

reproduction and distribution, selling or licensing copies, or posting

to personal, institutional or third party websites are defined by the

user license specified on the article.

For more information regarding Elsevier's open access licenses

please visit:

(2)

AnnalsofHepatology18(2019)563–570

ContentslistsavailableatScienceDirect

Annals

of

Hepatology

j o ur na l h o me p a g e : w w w . e l s e v i e r . e s / a n n a l s o f h e p a t o l o g y

Original

article

Non-alcoholic

fatty

liver

disease

and

neurological

defects

夽

Rita

Moretti

a,∗

_,

_Paola

_Caruso

a

_,

_Silvia

_Gazzin

b

a_Neurology_Clinic,_Department_of_Medical,_Surgical,_and_Health_Sciences,_University_of_Trieste,_Trieste,_Italy b_Italian_Liver_Foundation,_Centro_Studi_Fegato,_Trieste,_Italy

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received1February2019 Accepted11April2019 Availableonline6May2019

Keywords: Liversteatosis Vascularriskfactors Neurologicaldisorders Behaviordiseases

a

b

s

t

r

a

c

t

IntroductionandObjectives:Nonalcoholicfattyliverdisease(NAFLD)canbeconsideredoneofthemost commoncausesofliverdiseaseinourdaysandisregardedasoneofthenewestvascularriskfactorsfor cerebrovascularandotherneurologicaldiseases.

Materialsandmethods:Westudiedagroupofneurologicaloutpatients,dividedintotwohomogenous groupsbasedonthepresenceorabsenceofNAFLD.

Resultsandconclusions:WetestiﬁedanindependentrelationshipbetweenNAFLDandcommonvascular riskfactors(age,sex,educationallevel,BMI,cholesterolandlipidassessment,Hb1ac).Atthesametime, weascertainedanindependentrelationshipbetweenNAFLDandmorerecentlyrecognizedvascularrisk factors,suchaslackoffolate,vitaminB12andvitaminD-OH25,andincreasedlevelsofhomocysteine. Finally,wehavedocumentedthatNAFLDshowedworseexecutiveandfrontalfunctions,andbehavioral changes,suchasdepressivemoodandanxiety,andapathy.

1. Introduction

Non-alcoholicfattyliverdisease(NAFLD)wasdeﬁnedas“mild, moderateorseveresteatosisasdeterminedbyultrasound,inthe absenceofelevatedalcoholconsumption”(generallyestimatedas morethan20g/dayforwomenandmorethan30g/dayformen), andintheabsenceofapositivetestforviralhepatitisBorC[1,2]. NAFLDisconsideredoneofthemostcommoncausesofchronic liverdisease[3–5].Ithasbeenreportedthatitisgradually becom-ingoneofthemostprevalentliverdiseases,identiﬁedonimaging in20–33%ofadults[1,6,7].

NAFLDhasbeenrecentlyassociatedwithcardiovasculardisease andislinkedwithvascularriskfactors,suchasobesity, hyperten-sion,type2diabetes,dyslipidemia[8–10].Somestudies[4,11]also proposedthatNAFLDcouldbeconsideredanindependentvascular riskfactor.WonSeoetal.[3]hypothesizedfortheﬁrsttimethat NAFLDcouldbeindependentlyassociatedwithcognitive impair-ment.

Ourstudywasaimedtoﬁndapossibleindependent relation-shipwithmorerecentlyrecognizedvascularriskfactors,suchas lackoffolate,vitaminB12andvitaminD-OH25,andincreased

lev-夽 TheauthorsthankMr.AndrewHighgatefortherevisionofthemanuscript. ∗ Correspondingauthorat:NeurologyClinic,DepartmentofMedical,Surgical,and HealthSciences,UniversityofTrieste,Trieste,Italy.

E-mailaddress:moretti@units.it(R.Moretti).

elsofhomocysteine[12–16].Moreover,wewanttodeterminea possibleindependentrelationshipwithcognitiveandbehavioral changes,suchasfrontalexecutivefunctionsanddividedattention, depression,anxiety,andapathy,not,tothebestofourknowledge, togetherpreviouslystudied.

1.1. Subjectscharacteristics

Thestudyincluded1358,consecutivecasesofadultpatients referredtotheNeurologicalUnitoftheUniversityofTrieste,from June1st,2008toJune1st,2016.TheirreferraltoNeurologicalUnit wasdeterminedby differentdiagnosis:743suffering from cer-vicogenicheadaches,307sufferingfromamigraine,and308due toasingletransientischemicattack.Studysubjectsunderwenta standardizedbaseline assessment,thatincluded: adetailed his-tory;aphysicalexamination;laboratorytests;neuropsychological evaluations;anabdominalechography;acarotid Doppler ultra-sonography.Thephysicalexaminationincludedcardiacandblood pressureevaluation,peripheralpulses,retinalvesselexploration, electrocardiographicevaluation. Allthepatientswereevaluated withacompleteneurologicalexaminationandcompletedthetest withaseriesofneuropsychologicalexamination.

Alltherecruitedparticipantsunderwentagallbladderandliver examinationbyultrasoundwithaTOSHIBASSA-907machinewith a 3.75and5.0MHz transducerdataconcerningthepresenceof kidneytolivercontrast,echogenicwallswithintheintrahepatic vessels anddeepbeamattenuation,associated withthe

bright-https://doi.org/10.1016/j.aohep.2019.04.007

(3)

564 R.Morettietal./AnnalsofHepatology18(2019)563–570

nessoftheliverparenchymaandgallbladderwallswasobtained. Weappliedastandardizedalgorithm,todeﬁnetheextensionof thesteatosisasa4levelclassiﬁcation(1:none;2:mild;3: mod-erate;4:severesteatosis).Patientsweredividedintotwogroups accordingtothepresenceorabsenceofNAFLD.671patientswere affectedbyNAFLD:amongthem,359sufferedfromcervicogenic headaches(GroupA),161sufferingfromamigraine(GroupB),and 151duetoasingletransientischemicattack(GroupC).Amongthe 687patientswithoutNAFLD(GroupD)384sufferedfroma cervico-genicheadache,146fromamigraineand157hadasingletransient ischemicattack.GroupDwasconsideredthecontrolpopulation.

Standardquestionnaireswereconductedwithallparticipants andincludedalcoholconsumption,smoking,druguse,andphysical activity.

ThepresentstudywasconductedfollowingtheDeclarationof Helsinki,andundertheEthicsGuidelines(Point4oftheCEUR Dec-laration)oftheCommitteeoftheUniversity–HospitalofTrieste, andwritteninformedconsentwasobtainedfromallthe partici-pants.

2. Methods

Diabetesmellituswasdefinedasvenousplasmaglucose concen-trationsof>120mg/dlafteranovernightfast.Glycatedhemoglobin (Hb1Ac)results have beenaligned to theassay appliedby the DiabetesControland Complication Trial(DCCT),expressedas a percentage; non-diabetic normal range has been considered a level 4–5.6%; 5.5–6.5%indicates a highrisk of developing dia-betes;avaluesuperiorto6.6%indicatesdiabetes[17,18].Clinical measurementincluded:totalserumcholesterol,triglycerides,and high-densitylipoprotein(HDL),low-densitylipoprotein(LDL)was calculatedbyFriedwald’sformula[19].Serumleveloffolate, vita-minB12andhomocysteinewerealsotested,asspecificmeasures ofvascularriskfactors.Liverenzymes,alanineaminotransferase (ALT)andaspartateaminotransferase(AST),aswellasasecond generationenzymeImmuno-assay(AbbottLaboratories,Chicago, IL)wasemployedtodetectthepresenceofhepatitisCantibodies, whileasolidphasecompetitiveimmunoassay(Abbott Laborato-ries)wasusedtodetectantibodiestothehepatitisBcoreantigen. Serumlevelsof25(OH)Dvitaminweremeasuredusingenzyme immune-assaykits(DIASourceImmunoAssay, SABelgium)and theclassificationofvitaminDfollowingtheNationalOsteoporosis Society[20].

Standardquestionnaireswereconductedwithallparticipants andincludedalcoholconsumption(AUDIT),smoking,druguse,and physicalactivity.

Weobtainedtheaveragedailyestimateforalcoholconsumption asanamountlessthan20g/dayforwomenandlessthan30g/day formen[21–23].

Thebodymassindex (BMI)wascalculated.Vascular impair-ment,tumor,orCNSprimaryinﬂammatoryordegenerativedisease wereexcludedbyaheadCT/MRIscan.

AnUltrasonographyCarotidexaminationhasbeendonetoall thepatients.

Allthepatientsunderwentasimplesequenceof neuropsychol-ogicaltests:

1.Frontal Assessment Battery (FAB; score: 0–18; 18=normal)

[24,25].

2.Beck’s Depression Inventory (version for Italian population) (score: 0–39; total score 10–19 indicated mild depression; totalscore20–29indicatedmoderatedepression;>30indicated severedepression)[26,27].

3.HamiltonAnxietyRatingScale (HAM-A)(score: 0–56;a total scorecomprised0–17,estimatedmildanxiety;18–24:mildto moderateanxiety;25–30:severeanxiety)[28].

4.ApathyEvaluationScore(AES-C)(clinicalexamination; score: 18–72;higherscoresreﬂectmoreapathy)[29].

5.StroopTest(reading,colornaming,andinterferencemistakes)

[30,31].

2.1. Statisticalanalysis

StatisticalanalysiswasperformedwithSPSSstatistics17.0(SPSS version17.0).Thedifferenceinbaselinecharacteristicsbetween theNAFLDandnoNAFLDpatientswasassessedbyANOVAtest,for categoricalvariables;incasetheANOVAresultswerefound signif-icant,themultiplecomparisonanalysiswasalsodonebytheTukey Test,toexaminethosegroups,whichweresigniﬁcantlydifferent fromeachother.Themultinomiallogisticregressionmethodwas appliedtoanalyzetherelationshipbetweendiseasestatus(NAFLD and control)consideringthem asthedependentvariable (non-metric),andage(metric),sex(non-metric),Labsvalues(allmetrics) asindependentvariables.

ToevaluatetherelationshipbetweenNAFLDandcognitiveand behavioralresults,weperformedamultivariatelinearregression analysis.

Inmodel1,weadjustedforsex,age,race,andeducationallevels; inmodel2wefurtheradjustedforvitaminD-OH25,folate,vitamin B12,andhomocysteinelevels.

3. Results

We recruited 1358 consecutive patients. 671patients were affectedbyNAFLD.687didnotsufferfromNAFLD(GroupD)and wereconsideredashealthycontrols.

Amongthe671NAFLDpatientsenrolled,ﬁve patients aban-donedthestudy,13wereexcluded forlackof compliance,653 patients completed the study. Group A included 341 patients; GroupBenrolled155patients,andGroupCcomprised151patients, foratotalof647patients.

In Group D, composed initially by 687 healthy controls, 12 patientsabandonedthestudyand22wereexcludedforthelackof compliance.Therefore,GroupDfinallycomprisedof653patients. Bydefinition,NAFLDdidnothaveHBVorHCVvirus Antibod-ies;IngroupA(341patients),83(24.3%)showedgrade2steatosis and258(75.6%)grade3steatosis.In GroupB(155patients),64 patients(41.2%)showedgrade2steatosis,89(57.4%)grade3of steatosis,and2 (1.3%)patientsgrade4 ofsteatosis.In GroupC (151patients),74patients(49%)showedgrade2steatosisand77 patients(50.9%)grade3steatosis.Thetwopatientswithgrade4 ofsteatosishavebeenfurtherstudiedandweredefinedas nonal-coholicsteatohepatitis(NASH):theyrepresent0.3%oftheNAFLD patients.

Nobodyinthecontrolgroupshowedsteatosis(Table1). AUDIT (WHO) measures indicated an average score of 6.4 (SD=2.1)in568NAFLDpatientsand7.8 (SD=1.3)in79NAFLD patients.GroupDAUDITscorewasof6.6(SD=2.5)in579patients and7.2(SD=3.1)in74patients.

(4)

R.Morettietal./AnnalsofHepatology18(2019)563–570 565

Table1

SteatosisasdeﬁnedbyThirdNationalHealthandNutritionExa.Survey,2010(numberofpatientsand%).

Steatosis GroupA(341pts) GroupB(155pts) GroupC(151pts) GroupD(653pts)

Grade1=none 0 0 0 653(100%

Grade2=mild 83(24.3%) 64(41.2%) 74(49%) 0

Grade3=moderate 258(75.6%) 89(57.4%) 77(50.9%) 0

Grade4=severe 0 2(1.3%) 0 0

Table2

Comparisonofmeanvaluesofvariousbiochemicalparametersinthethreegroupsandcontrols(MeansandSD).

Variable(normalvalues) GroupA(341) GroupB(155) GroupC(151) Controls(653) Fchi2value DF p-Value

Age(years) 54.1(2.1) 57.3(1.4) 56.7(1.2) 59.1(1.4) 2.66 2.24 0.63

Educationallevel(yrs) 7.7(1.3) 7.9(1.1) 8.1(1.7) 8.2(0.9) 0.67 3.11 0.45

GenderM/F 187/166 91/70 82/69 352/301 0.79 2.31 079 Hb1Ac(%) 6.2(0.3) 6.3(0.1) 6.7(0.2) 5.3(0.7) 0.67 3.11 0.05 Cholesterol(mg/dl) 207.4(24.1) 215.5(21.3) 213(20.7) 167.7(10.3) 0.71 3.21 0.05 Triglycerides(mg/dl) 139.4(17.6) 144.6(21.5) 147.7(20.1) 110.1(10.7) 0.75 3.13 0.01 HDL(mg/dl) 31.7(10.2) 29.7(5.7) 28.7(12.3) 33.2(5.7) 0.67 2.45 0.63 LDL(mg/dl) 147.8(10.1) 156.8(15.3) 154.7(13.6) 114.18(7.1) 0.69 2.7 0.01 vitaminDOH25(20–100mg/dl) 12.7(1.5) 11.7(2.9) 13.4(1.2) 22.9(3.7) 2.66 2.5 0.01 Folate(3.8–26ng/ml) 2.4(1.1) 1.8(1.1) 2.1(1.1) 5.1(1.1) 0.71 2.15 0.01 VitaminB12(201–870pg/ml) 129.2(11.7) 111.7(21.3) 112.4(12.3) 237.2(12.5) 0.91 2.32 0.01 Homocysteine(3–15mcmol/l) 19.3(2.1) 17.3(2.5) 19.7(3.9) 19.7(3.4) 5.77 2.41 0.01 ALT(U/l) 54.1(7.1) 57.7(8.1) 52.3(3.1) 21.3(1.1) 3.41 2.31 0.01 AST(u/L) 38.7(6.2) 35.7(3.2) 37.9(3.1) 18.1(1.3) 2.75 2.41 0.01

BMI(num.and%)

<18.5 17(4.9%) 15(9.6%) 8(5.2%) 57(8.7%) 0.79 3.21 0.43 18.5–24.9 32(9.3%) 17(10.9%) 14(9.2%) 152(23.4%) 0.87 2.48 0.05 25–29.9 105(30.7%) 56(36.1%) 49(32%) 284(43.3%) 0.81 3.41 0.05 30–34.9 137(40.1%) 60(38.7%) 67(44.3%) 155(23.7%) 0.82 3.46 0.01 >35 50(14.7%) 7(4.6%) 13(8.6%) 5(0.7%) 0.87 3.23 0.01 Table3

Neurologicalconditionsinthetwogroupsconsidered.

NAFLD Not-NAFLD Controlgroup

Neurologicalcondition GroupA GroupB GroupC GroupD

Cervicogenicheadaches 359 0 0 384

Migraine 0 161 0 146

Transientischemicattacks 0 0 151 157

a signiﬁcantincrease ofglycated hemoglobin(p<0.05),of total cholesterol(p<0.05),triglycerides(p<0.01),andofLDL(p<0.01) thancontrols.TheyshowedadecreasedlevelofvitaminD-OH25 (p<0.01),folate(p<0.01),vitaminB12(p<0.01),elevatedlevelsof homocysteine(p<0.01),andhigherlevelsofASTandALT(p<0.01), thancontrols.

As expected, there is a higher percentage of increased BMI (30–35ormore)inNAFLD(p<0.01)thancontrols.Moreover,there isahigherpercentageofnot-hemodynamicstenosis(50–75%)in NAFLD(Table4).

PatientsofthethreegroupsdidworseintheFABtestthan con-trols(p<0.05) showingmoresevere executivealterations.They showedhigherlevelsofdepressivemood(accordingtoBeck’stest) (p<0.01)thancontrols.Atthesametime,theyshowedgreater anx-iety,accordingtoHAM-A(p<0.01)whencomparedtocontrolsand higherlevelsofapathy(AES-C)(p<0.01)(Table5).Moreover,they madehigherquantitiesofinterferencemistakesinSTROOPtest thancontrols(p<0.01).

We have considered four biochemical variables (low folate, vitamin B12, low vitamin D-OH25, and high homocysteine) and found them signiﬁcantly different in the three Groups (Table 6). In NAFLD groups, the mean level of vitamin D-OH was lower than controls (p<0.01) (Table 6), and as seen in

Table 3, Group A population was entirely in an insufﬁciency condition,whereas GroupBin adeﬁciency status(NOSCriteria

[20]).

In NAFLDpopulations,mean Hcylevelswerehigher thanin controls(p<0.01)(Table6),aswellasfolateandvitaminB12 lev-els,weresignificantlylowerthancontrols(p<0.01)(Table6).The Multiple-comparisonTukeyTestconfirmedtheseresults(Table7). As demonstrated by the Univariate Regression Analysis, as showninTable8,crudeoddsratiosaresignificantforthe asso-ciationbetweenNAFLDandvitamininsufficiencyanddeficiency, folateandvitaminB12deficiencyandhighlevelofhomocysteine. Atthesametime,thereisapositiveassociationbetweenNAFLD andlowperformancesinFABandInterferencePartofStrooptest, higherlevelsofBeckHamiltonandAES-Ctest.

Thepresenceofarelationshipbetweenthemwascheckedbased onthestatisticalsignificanceofthefinalmodelchi-square,andthe existenceofarelationshipwasestablished.Alltheindependent variables,exceptedage,sex,andeducationallevel,hadsignificant contributionstowardNAFLDgroups(Table9).

(5)

Table4

Dopplerresults.

Doppler GroupA(341) GroupB(155) GroupC(151) Controls(653) Fchi2value DF p-Value

Intimalmedialthickening 293(85.9%) 97(62.5%) 20(13.5%) 453(69.3%) 0.27 2.17 0.63

Stenosis<50% 24(7.1%) 40(25.8%) 97(64.2%) 116(17.8%) 0.37 2.21 0.01

Stenosis50–75% 23(6.9%) 17(0.9%) 31(20.5%) 84(12.6%) 1.17 2.19 0.01

Stenosis>75% 1(0.3%) 1(0.6%) 3(1.9%) 0(0%) 4.31 2.41 0.54

Occlusion 0 0 0 0 0 0 0

Table5

Comparisonofmeanvaluesofneuropsychologicaltestsinthethreegroupsandcontrols(MeansandSD).

Characteristics GroupA(341) GroupB(155) GroupC(151) Controls(653) Fchi2value DF p-Value

FABtest 11.3(2.5) 11.6(1.7) 10.9(2.1) 115.6(1.2) 0.87 2.43 0.01 Beck’stest 27.1(1.3) 28.3(2.7) 29.1(2.3) 12.7(1.3) 0.74 2.9 0.01 HAM-Atest 36.4(2.7) 31.3(1.3) 27.1(3.1) 16(2.1) 0.08 2.02 0.01 AES-C 24.3(1.7) 26.7(2.1) 27.1(2.7) 11(2.1) 0.37 2.42 0.01 Strooptest Reading(correct) 84.5(2.7) 86.7(2.1) 83.1(1.7) 91.1(1.7) 0.75 2.41 0.36

Colornaming(correct) 61.7(3.1) 63.1(2.7) 64.1(1.9) 67.1(1.3) 0.77 2.34 0.54

Interferencemistakes(correct) 31.9(2.1) 32.7(3.1) 31.1(1.7) 17.3(1.7) 0.89 2.48 0.01

Table6

Comparisonofmeanvaluesofspeciﬁcbiochemical(dividedbydifferentlevels)andneuropsychologicalparametersinthethreegroupsandcontrols(MeansandSD).

Variable(normalvalues) GroupA(341) GroupB(155) GroupC(151) Controls(653) Fchi2value DF p-Value

vitD-OH25 <12ng/ml 0 11.7(2.9) 0 0 0.67 2.25 0.01 vitD-OH25 12–20ng/ml 12.7(1.5) 0 13.4(1.2) 0 0.87 2.13 0.01 vitD-OH25 >25ng/ml 0 0 0 22.9(3.7) 0.71 2.6 0.01 Hcy(3–14mcmol/l) 0 0 0 11.1(3.4) 0.34 2.11 0.01 Hcy(15–30mcmol/l) 19.3(2.1) 17.8(2.5) 19.7(3.9) 0 0.71 2.34 0.01 Folate(0.5–3.7ng/ml) 2.4(1.1) 1.8(1.1) 2.1(1.1) 0 0.47 2.17 0.01 Folate(3.8–26ng/ml) 0 0 0 5.1(1.1) 0.23 2.41 0.01 VitaminB12(90–204pg/ml) 129.2(11.7) 111.7(21.3) 112.4(12.3) 0 0.77 2.35 0.01 VitaminB12(205–870pg/ml) 0 0 0 237.2(12.5) 0.87 2.6 0.01

thattheincreaseofvitaminD-OH25decreasedthelikelihoodof NAFLD,withanexponentialBvalue0.85,whichimpliesthatforan increaseofvitaminB12levels,theoddsofhavingNAFLDdecreased by7%.Theregressioncoefﬁcient(B)forthehomocysteinewas+0.57 forNAFLD;itindicatesthatthedecreaseofhomocysteinedecreased thelikelihoodofNAFLD,withanexponentialBvalue0.98,which impliesthatforadecreaseofhomocysteinelevels,theoddsof hav-ingNAFLDdecreasedby15%.

InordertoevaluatetherelationshipbetweenNAFLDand cogni-tiveandbehavioralimpairment,weperformedamultivariatelinear regressionanalysis:inmodel1,weadjustedforsex,age, educa-tionallevelandinmodel2wefurtheradjustedforvitaminD-OH25, folate,vitaminB12,andhomocysteine(Table10).

FAB,Beck,HAM-A, AES-Cand InterferenceMistakes number failedtohavealowerscoreconcerningage,sex,andeducational levels(seeallthemodel1regressioninTable9).Ontheotherhand, comparedtocontrols,NAFLDpatientshadlowerperformanceson FAB,higherscoresinBeck,HAM-A,andAES-Candhigher Inter-ferenceNumber of mistakesin Strooptest aftercontrolling for vitaminD-OH25,folate,vitaminB12andhomocysteine (respec-tively:FAB:B=0.72,95%CI:0.105–1.347,p<0.01;Beck:B=0.88, 95%CI:0.97–10.65,p<0.01;HAM-A:B=0.74,95%CI:0.32–2.31, p<0.01;AES-C:B=0.65,95%CI:0.4–2.6,p<0.01;Interference Mis-takesStroopTest:B=0.87,95%CI:0.3–2.3,p<0.05).

Theclassificationaccuracyrateofthelogisticmodelwas56.7%, whichwashigherthantheproportionalbychanceaccuracy;the criteriafortheclassificationaccuracyweresatisfied.

Finally,wedidnotfindacorrelationbetweenhigherlevelsof BeckScoresandlowerFABscores,orhigherInterferencemistakes test.WedidnotfindanyrelationbetweenhigherlevelsofHAM-A andlowerFABscoresorhigherInterferencemistakestest.Wedid findanot-significant,buteffectiverelationbetweenhigherlevels

ofAES-CandlowerFABscores(r=−0.67,p=0.064,Spearman’s cor-relationcoefficient),butnotwithhigherInterferencemistakestest. Asexpected,wedidfindapositivecorrelationbetweenBeck’sscore andHAM-A(r=0.89,p<0.01)(Spearman’scorrelationcoefficient), butanywithAES-C.

4. Discussion

This work is a cross-sectional study, covering 647 NAFLD-neurologicaloutpatients,and653control-outpatients.

WetestiﬁedthatNAFLD,followingdatapreviouslypublished, relatedtocommonvascularriskfactors,suchaslipidassessment and lowerHDL levels, Hb1ac,ASTand [9,11,32,33] and obesity

[34,35].

NAFLDhasbeenonlyrecentlyrelatedtolowerlevelsofvitamin Bcomplex[36],ithasbeenshylyrelatedtohigherlevelof homo-cysteine[37],andﬁnally,ithasbeentightenedtolowerlevelsof vitaminD(independenttobodymassandobesitystatus)[38,39].

Toourknowledge,thisistheﬁrstworkwhichreporteda posi-tivecorrelationwithNAFLDandlowlevelsofvitaminB12,folate, andvitaminDandwithhigherlevelsofhomocysteine.Moreover, wedocumentedthatforanincreaseofvitaminDlevels,theodds ofhavingNAFLDdecreasedby18%;foranincreaseoffolate lev-els,theoddsofhavingNAFLDdecreasedby9%;foranincreaseof vitaminB12levels,theoddsofhavingNAFLDdecreasedby7%,and foradecreaseofhomocysteinelevels,theoddsofhavingNAFLD decreasedby15%.

We demonstrated a higher percentage of not-hemodynamic carotidstenosis(50–75%)inNAFLDthaninthecontrolgroup,in linewithotherworks[40].

(6)

Table7

Multiplecomparisonanalysis(Tukeytest)ofvariousbiochemicalandneuropsychologicalparametersinthefourgroups.

Variables MeanDifferences SEofmeandifferences p-Value

VitaminD-OH25 AvsD −10.2 −2.1 0.05 BvsD −11.2 −1.96 0.01 CvsD −9.5 −1.7 0.01 Folate AvsD −2.7 −0.1 0.01 BvsD −3.3 −0.3 0.01 CvsD −3.0 −0.7 0.01 Homocysteine AvsD +8.1 0.6 0.01 BvsD +6.2 0.7 0.01 CvsD +8.6 1.1 0.01 VitaminB12 AvsD −101.7 −10.1 0.01 BvsD −119.7 −12.3 0.01 CvsD −118.6 −11.7 0.01 FAB AvsD −4.3 −1.1 0.01 BvsD −4.6 −0.7 0.01 CvsD −4.1 −1.2 0.01 HAM-A AvsD +20.4 +3.4 0.01 BvsD +15.7 +3.7 0.01 CvsD +21.1 +2.7 0.01 Beck AvsD +14.4 +2.3 0.01 BvsD +15.6 +3.1 0.01 CvsD +16.4 +2.7 0.01 AES-C AvsD +13.3 +2.1 0.01 BvsD +15.7 +1.9 0.01 CvsD +16.1 +1.6 0.01

InterferencemistakesStroop

AvsD +14.6 +2.5 0.01

BvsD +15.4 +2.9 0.01

CvsD +13.8 +2.1 0.01

Table8

Univariateregressionanalysis.

Dependentvariable Independentvariable Oddsratio 95%CI p

NAFLD vitD-OH25insufﬁciency 5.4 10.93–13.56 0.023

vitD-OHdeﬁciency 4.3 8.93–11.56 0.01

Folate 3.9 2.1–4.2 0.037 vit.B12 2.7 29.3–36.7 0.034 Hcy 4.5 17.3–19.7 0.036 FAB 2.7 10.9–11.8 0.045 BECK 2.3 12.7–29.1 0.047 HAmiltonAnxiety 5.73 11.2–31.1 0.011 InterferenceMistakes 5.4 12.3–35.7 0.019 Table9

Summaryofmultinomiallogisticregressionanalysis.

Variables Univariateassoc. Regressioncoefﬁcient(B) SE p-Value EXP(B) 95%CIforexp.(B)

Dependent Independent OR 95%CI p-Value

(7)

Table10

AssociationbetweenNAFLDandneuropsychologicalvariableswithamultivariatelinearregressionanalysis.

B p-Value SE 95%CI FAB Model1 0.54 0.6 3.56 0.48–1.1 Model2 0.72 0.01 3.76 0.105–1.347 Beck Model1 0.12 0.7 3.1 0.002–0.21 Model2 0.88 0.01 3.44 0.97–10.65 HAM-A Model1 0.32 0.5 3.23 0.012–0.023 Model2 0.74 0.01 3.56 0.32–2.31 AES-C Model1 0.34 0.56 3.4 0.092–0.2 Model2 0.65 0.01 4.2 0.4–2.6

InterferenceMistakesStroopTest

Model1 6.67 0.67 3.65 0.5–13.1

Model2 0.87 0.05 3.4 0.3–2.3

Model1adjustedforage(continuous),sex(m/f),educationallevels(yearsschool).Model2adjustedforvitaminD-OH25,folate,vitaminB12andhomocysteine.

ofinterferencemistakes[30],higherscoresofBeckandHamilton AnxietyTest.Finally,amultivariatelinearregressionanalysis,as showninourmodel2,NAFLDpatientshadlowerperformanceson FAB,higherscoresinBeck,HAM-A,andAES-Candhigher Interfer-enceNumberofmistakesinStrooptestaftercontrollingforvitamin D-OH25,folate,vitaminB12,andhomocysteine.

Variousstudies havedemonstrated thatobese subjectshave lowerfolateserumlevelsincomparisonwithnormal-rangeBMI subjects[41,42].Furthermore,obesityis awell-known risk fac-torforNAFLD,asshowninourstudytoo[43].Therefore,ithas beensuggestedthatfolatedeﬁciencymayhaveapathogenicrole inNAFLD[36],anditleadstoamoresevereformofthedisease, mainlyinmetabolicsyndromeanddiabetesmellitustype2patients

[43].Moreover,Sidetal.[43]stressedtheimportanceoffolatein theprogressionofNAFLD,thusraisingthepossibilitythat supple-mentalfolatemaybeatreatmentoption(conﬁrmedbyXiaetal.

[44]).

Anotherimportantﬁndingofthisstudywastheroleoflow vita-minB12levelsasanindependentpredictorofNAFLD:ithasbeen hypothesizedarelationshipbetweenvitaminB12lowscoresand NASHhistologicalseverity,andﬁbrosisgrade[36].Theseresults corroboratedpreviousstudies,which demonstrated thatNAFLD patientshavelowerserumlevelsofvitaminB12incomparisonto controls[45–47]andareincontrastwithsomeothers[48,49].In anotherstudy[50,51],higherlevelsofhomocysteinewereobserved in71NAFLDpatients,whichcorrelatednegativelywithfolateand B12concentrations.Therefore,Mahamidetal.[36]speculatedthat lowlevelsofvitaminB12mightpredictthedevelopmentofasevere formofNASH;highlevelsmayindicatethecommencementof end-stageliverdisease.

Evenmoredifficultistoexplainthedysexecutiveand atten-tiondisordersshowedinourNAFLDpatients, associatedwitha somewhat increased level of depression and anxiety. The cog-nitivealterations observedin ourstudywerein line, withthat describedby WonSeoet al.[3]: themagnitude ofthe associa-tionwithNAFLDwasrobustandpersistedwhen othervascular factorshavebeenincludedinmodels.AssuggestedbyWonSeo etal.[3]itremainspossiblethattheassociationbetweenNAFLD and cognitiveimpairment couldbean epiphenomenon,mainly duetothesuperimposingcardiovascularriskfactors, character-isticofthetwoconditions(NAFLDandcognitiveimpairment)or duetoaninflammatorycondition(commontoNAFLDand cogni-tiveimpairment[35,52–55]).Ourpatients,though,underwentto neuroimaging,andwedidnotfindanysignificantvascular alter-ation.Thebehavioraldisturbancesstudiedinourworkreflected

thatNAFLDwasmoredepressedandmoreanxious(inaccordance topreviousstudies,i.e.[56–59]),andasimagined,thereisapositive correlationbetweendepressionandanxietyscores.Curiously,none ofthepatientsrecruitedinthestudyauto-deﬁnedthemselvesas depressedoranxious.Toourbestknowledge,thisistheﬁrststudy whichanalyzesapathyscoreandfoundoutthatNAFLDrelatedwith ahigherscoreofapathy,whichdidnotrelatetodepressionor anx-iety[60].Thereisasinglework[61]that,employingnear-infrared spectroscopy(NIRS),foundoutthatJapanesewomen,withNAFLD, weremoredepressed,withlessconcentrationandshowedaminor valueofoxy-Hbconcentrationinthefrontallobe.

NIRStechniqueisnotentirelyanduniversallyaccepted[61,62], butitissuggestiveofsupportingourﬁndings,whicharestrongly relatedtofrontal–subcorticalsufferance.

One of the open questions could be if the observed spec-trumcouldbe a primaryeffect, concernedsomehowtoNAFLD orsecondarytoallthevascularfactors,whicharesuper-imposed to NAFLD? In accordance to the Literature, NAFLD comprises metabolicriskfactors,suchashypertension,highertriglycerides level,higherLDLandlowerHDL,diabetes,andobesity,andeven thecombinationofalltheriskfactors[63–65].Sincewehave con-sideredtwoneurologicalpopulations,indeed,onewithNAFLDand theotherwithoutit,themetabolicsyndrome,whichcharacterized NAFLD,didnotinﬂuenceneurologicalpathologies,perse.Thetwo groupsarematchedforneurologicalcomorbidities,whichdidnot seemtoberelatedtothemetabolicfactorswefoundintheNAFLD group.Ontheotherhand,theotherneuropsychologicalspectrum ofsymptoms,i.e.,apathy,depression,dysexecutivefunctions,has beenobservedinNAFLDgroup.Thequestionremainsunsolved: aretheyrelatedtoNAFLDorthemetabolicsymptomsboundto NAFLD?Ifso,behaviorandvascularriskscouldevenbemore pro-nounced inpatientswithNASH:a furtherpossibilityfor future studiescouldbethecomparisonofNAFLDandNASHpopulation; thiswillhopefullygivemoredata,withahealthygrouptobetter understandtherealroleofliveralterationinthe neuropsychologi-calproﬁle.

(8)

differenceshavethesameneurologicalconditionsthantheother groupandcanbethoroughlystudied.

Our data werethe ﬁrstto relatevitamin B12 and D, folate, andhomocysteine anddifferentneuropsychologicalvariablesto NAFLD,havingexcludedalcoholconsumption.

Morestudieswillbeneededtoimplementfurtherknowledge.

Conﬂictofinterest

Theauthorshavenoconﬂictsofinteresttodeclare.

References

[1]SanyalAJ,BruntEM,KleinerDE,KowdleyKW,ChalasaniN,LavineJE,etal. End-pointsandclinicaltrialdesignfornonalcoholicsteatohepatitits.Hepatology 2011;54:344–53.

[2]Sattar N, Forrest E, Preiss D. Non alcoholic fatty liver disease. BMJ 2014;349:g4596,http://dx.doi.org/10.1136/bmj/g.4596.

[3]WonSeoS,GottesamnRF,ClarkJM,HernaezR,ChangY,KimC,etal.non alcoholicfattyliverdiseaseisassociatedwithcognitivefunctioninadults. Neurology2016;22(86):1136–42.

[4]FargionS,PorzioM,FracanzaniAL.Nonalcoholicfattyliverdiseaseandvascular disease:stateoftheart.WorldJGastroenterol2014;20(37):13306–24.

[5]LazoM,ClarkJM.Theepidemiologyofnonalcoholicfattyliverdisease:aglobal perspective.SeminLiverDis2008:339–50.

[6]RatziuV,BellentaniS,Cortez-PintoH,DayC,MArchesiniG.Aposition state-mentonNAFLD/NASHbasedontheEASL2009specialconference.JHepatol 2010;53:372–84.

[7]KneemanJM,MisdrajiJ,CoreyKE.Secondarycausesofnonalcoholicfattyliver disease.TherapAdvGastroent2012;5(3):199–207.

[8]KhoonsariM,AzarMMH,GhavamR,HatamiK,AsobarM,GholamiA,etal. Clinicalmanifestationsanddiagnosisofnonalcoholicfattyliverdisease.IrJ Pathol2017;82(12):99–105.

[9]StepanovaM,YounoussiZM.Independentassociationbetweennonalcoholic fattyliverdiseaseandcardiovasculardiseaseintheUSpopulation.Clin Gas-troenterolHepatol2012;10:646–50.

[10]PaschosP,PaletasK.Nonalcoholicfattyliverdiseaseandmetabolicsyndrome. Hyppokrata2009;13(1):9–19.

[11]TargherG,DayCP,BonoraE.Riskofcardiovasculardiseaseinpatientswithnon alcoholicfattyliverdisease.NEnglJMed2010;363:1341–50.

[12]DeCaterinaR,ZampolliA,MadonnaR,FiorettiP,VagnuzzoD.New cardio-vascularriskfactors:homocysteineandvitaminsinvolvedinhomocysteine metabolism.ItalHeartJ2004;5(Suppl.6):19S–24S.

[13]SadeghianS,FallahiF,SalarifarM,DavoodiG,MahmoodianM,FallahN,etal. Homocysteine,vitaminB12andfolatelevelsinprematurecoronaryartery dis-ease.BMCCardiovascDisord2006;6:38, http://dx.doi.org/10.1186/1471-2261-6-38.Published2006Sep26.

[14]MohanM,KulkarniMV,GargMK,NaikSS.VitaminB12deﬁciencyand hyper-homocysteinemiaascorrelatesofcardiovascularriskfactorsinIndiansubjects withcoronaryarterydisease.JCardiol2013;61(4):289–94.

[15]MandarinoNR,JúniorFD,SalgadoJV,LagesJS,FilhoNS.Isvitaminddeﬁciencya newriskfactorforcardiovasculardisease?OpenCardiovascMedJ2015;9:40–9,

http://dx.doi.org/10.2174/1874192401509010040.

[16]MorettiR,CarusoP,DalBenM,ContiC,GazzinS,TiribelliC.VitaminD, homo-cysteine,andfolateinsubcorticalvasculardementiaandAlzheimerdementia. FrontAgingNeurosci2017,May,http://dx.doi.org/10.3389/fnagi.2017.00169. [17]WHO,IDF.Deﬁnitionanddiagnosisofdiabetesmellitusandintermediate

hyperglycemia.2006.Geneva.

[18]WeykampC.HbA1C:areviewofanalyticalandclinicalaspects.AnnLabMed 2013;33:393–400,http://dx.doi.org/10.3343/alm.2013.33.6.393.

[19]FriedwaldWT,LewyRI,FredricksonDS.Estimationoftheconcentrationof low-densitylipoproteincholesterolinplasma,withoutuseofthepreparative ultracentrifuge.ClinChem1972;18:499–502.

[20]AsprayTJ,BowringC,FraserW,GittoesN,JavaidMK,MacDonaldH,etal. National osteoporosis society vitamin D guidelinesummary. Age Ageing 2014;43:592–5,http://dx.doi.org/10.1093/ageing/afu093.

[21]ClarkJM,BrancatiFL,DiehlAM.Theprevalenceandetiologyofelevated amino-transferaselevelsintheUnitedStates.AmJGastroentorol2003;98:9690–967.

[22]BeckerU,DeisA,SorensenTI,GronbaekM,Borch-JohnsenK,MullerCF,Schnohr P,jensenG.Predictionofriskofliverdiseasebyalcoholintake,sexandage:a prospectivepopulationstudy.Hepatology1996;23:1025–9.

[23]Sattar N, Forrest E, Preiss D. Non alcoholic fatty liver disease. BMJ 2014;349:g4596.

[24]DuboisB,PillonB,SlachevskyA,LitvanI.FrontalAssessmentBatteryatbedside. Neurology2000;55:1621–6.

[25]ApollonioI,LeoneM,IsellaV,PiamartaF,ConsoliT,VillaML,etal.TheFrontal AssessmentBattery:normativevaluesinanItalianpopulationsample.Neurol Sci2005;26:108–16,http://dx.doi.org/10.1007/s10072-005-0443-4. [26]BeckAT,SteerRA,BrownGK.BeckDepressionInventory-SecondEdition.

Man-ual.SanAntonio,TX:ThePsychologicalCorporation;1996.

[27]GhisiM,FlebusGB,MontanaroA,SanavioE,SicaC.BeckDepressionInventory II:ManualeItaliano.Firenze:OrganizzazioniSpeciali;2006.

[28]Hamilton M. The assessment of anxiety states by rating. Br. J Med Psychol 1959;32:50–5. Italian Validation; Technology Systems Healthcare.

[29]MarinRS,BiedrzyckiRC,FirinciogiullariS.Reliabilityandvalidityoftheapathy evaluationscale.PsychRes1991;38(2):143–62.

[30]StroopJR.Studiesofinterferenceinspatialandverbalreactions.JExpPsychol 1935;18:643–62.

[31]ValgimigliS,PAdovaniR,BudriessiC,LeoneME,LugliD,NichelliPF.Testdi Stroop:datinormativeItalianidiunaversionecartaceaperl’usoclinic.Giornale ItPsicol2010;37:945–53.

[32]TargherG, bertoliniL, PoliF, etal.Nonalcoholicfatty liverdiseaseand riskoffuturecardiovasculareventsamongtype2diabeticpatients.Diabetes 2005;54:3541–6.

[33]LazoM,HernaezR,BonekampS,etal.Non-alcoholicfattyliverdiseaseand mortalityamongUSadults:prospectivecohortstudy.BMJ2011;343:d6891.

[34]dePianoA,EstadellaD,OyamaLM,RIbeiroEB,DamasoA,daPenhaOlierde NascimentoC.Nonalcoholicfattyliverdisease(NAFLD),amanifestationofthe metabolicsyndrome:newperspectivesonthenutritionaltherapy.Endocrinol MetabSynd2014;3:3,http://dx.doi.org/10.4172/2161-1017.1000135. [35]VanGaalLF,MertensIL,DeBlockCE.Mechanismslinkingobesitywith

cardio-vasculardisease.Nature2006;444:875–80.

[36]MahamidM,MahroumN,BragazziNL,ShalaataK,YAneY,AdawiM,AmitalH, WatadA.FolateandB12levelscorrelatewithhistologicalseverityinNASH patients.Nutrients 2018;10(4):440, http://dx.doi.org/10.3390/nu10040440. Published2018Apr2.

[37]BarchettaI,AngelicoF,DelBenM,BAroniMG,PozzilliP,etal.Strongassociation betweennonalcoholicfattyliverdisease(NAFLD)andlow25(OH)vitamin Dlevelsinanadultpopulationwithnormalserumliverenzymes.BMCMed 2011;9:85.

[38]BarchettaI,CarottiS,LabbadiaG,GentilucciUV,MudaAO,etal.Livervitamin Dreceptor,CYP2R1andCYP27A1expression:relationshipwithliverhistology andvitaminD3levelsinpatientswithnonalcoholicsteatohepatisorhepatitis Cvirus.Hepatology2012;56:2180–7.

[39]WendellCR,ZondermanAB,MetterEJ,NAjjarSS,WaldsteinSR.Carotid inti-malmedialthicknesspredictscognitivedeclineamongadultswithoutclinical vasculardisease.Stroke2009;40:3180–5.

[40]CasanuevaE,DrijanskiA,Fernández-GaxiolaAC,MezaC,PfefferF.Folate deﬁ-ciencyisassociatedwithobesityandanemiainMexicanurbanwomen.Nutr Res2000;20:1389–94,http://dx.doi.org/10.1016/S0271-5317(00)80020-2. [41]HirschS,PoniachickJ,AvendanoM,CsendesA,BurdilesP,SmokG,DiazJC,

delaMazaMP.Serumfolateandhomocysteinelevelsinobesefemaleswith non-alcoholicfattyliver.Nutrition2005;21:137–41,http://dx.doi.org/10.1016/ j.nut.2004.03.022.

[42]FabbriniE,SullivanS,KleinS.Obesityandnonalcoholicfattyliverdisease: biochemical,metabolicandclinicalimplications.Hepatology2010;51:679–89,

http://dx.doi.org/10.1002/hep.23280.

[43]SidV,SiowYLOK.Roleoffolateinnonalcoholicfattyliverdisease.CanJPhysiol Pharmacol2017;95:1141–8,http://dx.doi.org/10.1139/cjpp-2016-0681. [44]XiaMF,BianH,ZhuXP,YanHM,ChangXX,ZhangLS,etal.Serumfolicacid

levelsareassociatedwiththepresenceandseverityofliversteatosisinChinese adults.ClinNutr2017,http://dx.doi.org/10.1016/j.clnu.2017.06.021. [45]KoplayM, Gulcan E, Ozkan F. Association between serum Vitamin B12

levels and the degree of steatosis in patients with nonalcoholic fatty liverdisease.JInvestigMed2011;59:1137–40,http://dx.doi.org/10.2310/JIM. 0b013e31822a29f5.

[46]MechieNC,GoralzcykAD,ReinhardtL,MihmS,AmanzadaA.Associationof serumVitaminB12levelswithstageofliverﬁbrosisandtreatmentoutcome inpatientswithchronichepatitisCvirusgenotype1infection:a retrospec-tivestudy.BMCResNotes2015;8(260), http://dx.doi.org/10.1186/s13104-015-1248-z.

[47]GoelA,RamakrishnaB,MuliyilJ,MadhuK,SajithKG,ZachariahU,etal.Use ofserumVitaminB12levelasamarkertodifferentiateidiopathic noncir-rhoticintrahepaticportalhypertensionfromcryptogeniccirrhosis.DigDisSci 2013;58:179–87,http://dx.doi.org/10.1007/s10620-012-2361-7.

[48]PolyzosSA,KountourasJ,PatsiaouraK,KatsikiE,ZafeiriadouE,ZavosC,etal. SerumVitaminB12andfolatelevelsinpatientswithnon-alcoholicfatty liverdisease.IntJFoodSciNutr2012;63:659–66,http://dx.doi.org/10.3109/ 09637486.2011.649249.

[49]BaltaciD,KutlucanA,TurkerY,YilmazA,KaracamS,DelerH,etal. Associa-tionofVitaminB12withobesity,overweight,insulinresistanceandmetabolic syndrome,andbodyfatcomposition;primarycare-basedstudy.MedGlas 2013;10:203–10.

[50]GulsenM,YesilovaZ,BagciS,UygunA,OzcanA,ErcinCN,etal.Elevatedplasma homocysteineconcentrationsasapredictorofsteatohepatitisinpatientswith non-alcoholicfattyliverdisease.JGastroenterolHepatol2005;20:1448–55,

http://dx.doi.org/10.1111/j.1440-1746.2005.03891.x.

[51]PolyzosSA,KountourasJ,PatsiaouraK,KatsikiE,ZafeiriadouE,DeretziG,etal. Serumhomocysteinelevelsinpatientswithnonalcoholicfattyliverdisease. AnnHepatol2012;11(1):68–76.

[52]KimDG,KrenzA,ToussaintLE,MaurerKJ,RobinsonSA,YanA,etal. Non-Alcoholicfatty liverdiseaseinduces signsofAlzheimer’s disease(AD)in wild-typemiceandacceleratespathologicalsignsofADinanADmodel.J Neuroinﬂamm2016;13:1,http://dx.doi.org/10.1186/S12974-15-0467-S. [53]VillanovaN,MoscatielloS,RamilliS,BugianesiE,MagalottiD,VanniE,ZoliM,

(9)

[54]RosenbergGA,BjerkeM,WallinA.Multimodalmarkersofinﬂammationinthe subcorticalischemicvasculardiseasetypeofvascularcognitiveimpairment. Storke2014;45:1531–8.

[55]TakedaS,SatoN,MorishitaR.Systemicinﬂammation,bloodbrainbarrier vulnerabilityandcognitive/noncognitivesymptomsinAlzheimerdisease: rel-evancetopathogenesisandtherapy.FrontAgingNeurosci2014;6:171–9.

[56]ElwingJE,LustmanPJ,WangHL,ClouseRE.Depression,anxiety,andnon alco-holicsteatohepatitis.PSychosomMed2006;68:563–9.

[57]WeinsteinAA,KallmanPriceJ,StepanovaM,PomsLW,FangY,MoonJ,etal. Depressioninpatientswithnonalcoholicfattyliverdiseaseandchronicviral hepatitisBandC.Psychosomatics2011;52:127–32.

[58]YoussefNA,AbdelmalekMK,BinksM,GuyCD,OmenettiA,SmithAD,etal. Asso-ciationsofdepression,anxietyandantidepressantswithhistologicalseverity ofnonalcoholicfattyliverdisease.LiverInt2013;33:1062–70.

[59]TomenoW,KawashimaK,YonedaM,SaitoS,OgawaY,HondaY,etal.Non alcoholicfattyliverdiseasecomorbidwithmajordepressivedisorder.The pathologicalfeaturesandpoortherapeuticefﬁcacy.JGastroenterolHepatol 2015;30:1009–14.

[60]MorettiR,SignoriR.Neuralcorrelatesforapathy:frontal–prefrontaland pari-etalcortical–subcorticalcircuits.FrontAgingNeurosci2016;8:289,http://dx. doi.org/10.3389/fnagi.2016.00289.

[61]TakahashiT,TakikawaY,KawagoeR,ShibuyaS,IwanoT,KitazawaS. Influ-enceofskinbloodflowonnear-infraredspectroscopysignalsmeasuredonthe foreheadduringaverbalfluencytask.Neuroimage2011;57:991–1002.

[62]KirilinaE,JelzowA,HeineA,NiessingM,WabnitzH,BruhlR,etal.The phys-iologicaloriginoftask-evokedsystemicartifactsinfunctionalnearinfrared spectroscopy.Neuroimage2012;61:70–81.

[63]NationalGuideCentreUK:NonAlcoholicfattyliverdisease.Assessmentand management.LondonNationalInstituteforHealthandCareExcellence(UK): 2016,July.NiceGuideline,no49,5,RiskfactorsforNAFLD.

[64]BruntEM,WongWW,NobiliV,DayCP,SookolanS,MaherJJ,BugianesiE,Sirlin CB,Neuschweinder-TeriBA,RinellaME.Nonalcoholicfattyliverdisease.Nat RevDiseasePrimers2015;1:15080.