Please citemass inde

ARTICLE IN PRESSG ModelYDLD-2798; No. of Pages 6Digestive and Liver Disease xxx (2015) xxxxxx

Contents lists available at ScienceDirect

Digestive and Liver Disease

jou rna l h om epage: www.elsev ier .com/ locate /d ld

Liver, Pancreas and Biliary Tract

Sarcopenia is a risk factor for elevated aminotranindepe itsactivity

Ki Deok Y e, OByung ChDepartment of

a r t i c l

Article history:Received 19 August 2014Accepted 24 December 2014Available online xxx

Keywords:Alanine aminotransferaseAspartate amiMuscle massSarcopeniaSkeletal muscl

y is awith obesity and metabolic syndrome. However, elevated aminotransferase activity is not uncommon innon-obese and non-alcoholic patients in clinical practice.Aim: To examine the relationship between sarcopenia and aminotransferase activity in a large population-based cohort.Methods: Data from the Korean National Health and Nutrition Examinations were used. A total of 13,431subjects were included. A whole-body dual X-ray absorptiometry scan was performed on each patient to

1. Introdu

Alanine (AST) are twhepatitis, hare commoels [1]. Hoelevated livmetabolic dhepatitis, owho presenand metabo

CorresponCollege of MedKorea. Tel.: +8

E-mail add

http://dx.doi.o1590-8658/ this article in press as: Yoo KD, et al. Sarcopenia is a risk factor for elevated aminotransferase in men independently of bodyx, dietary habits, and physical activity. Dig Liver Dis (2015), http://dx.doi.org/10.1016/j.dld.2014.12.014

notransferase

e index

measure total and regional muscle mass. Appendicular skeletal muscle mass indices were also obtained.Results: The prevalence of sarcopenia was signicantly higher in the group with elevated aminotrans-ferase levels than in the normal liver enzyme group (males: 26.5% vs. 16.9%; females: 38.3% vs. 22.1%,p < 0.05). The skeletal muscle index was negatively correlated with most cardiometabolic risk factors,including fasting glucose and cholesterol levels. The frequency of elevated aminotransferase increasedin male patients with sarcopenia after adjusting for potential confounding factors including age, bodymass index, fasting glucose level, dietary, and exercise habits. However, the correlation was no longerobserved in women after adjusting for body mass index.Conclusion: Sarcopenia is a risk factor for elevated aminotransferase in men, independently of body massindex, dietary habits, and physical activity.

2015 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.

ction

aminotransferase (ALT) and aspartate aminotransferaseo well-known serologic markers of liver injury. Viral

eavy alcohol use, hepatotoxic drug use, and obesityn risk factors for abnormal aminotransferase lev-wever, it is not uncommon for patients who shower enzymes to lack the traditional risk factors forisease, including obesity, herbal medication use, viralr signicant alcohol consumption. Non-obese patientst metabolic diseases, including diabetes, hypertension,lic syndrome, are also common. Despite using strict,

ding author at: Department of Internal Medicine, Hanyang University,icine, 17 Haengdang-dong, Sungdong-gu, Seoul 133-792, Republic of2 2 2290 8338; fax: +82 2 972 0068.ress: noshin@hanyang.ac.kr (D.W. Jun).

ethnicity-specic criteria, many Asian-Pacic subjects havingdiabetes or hypertension are also non-obese [1,2]. Genetic back-ground, fat distribution, unhealthy dietary habits, and lifestyleshave all been suggested as risk factors for the development ofmetabolic disease in non-obese subjects [35].

Recently, several studies have addressed metabolic diseasesin non-obese patients [2,3,6]. Some researchers have reportedthe impact of muscle mass on metabolic disease incidence [69].Sarcopenia is a syndrome characterized by progressive and gen-eralized loss of skeletal muscle mass and strength [10]. Severalresearch groups have found that sarcopenia is associated withinsulin resistance, type 2 diabetes, dyslipidemia, and hypertension[8,9,11]. Sarcopenia often occurs in elderly people with normalbody mass index (BMI) [12]. However, the relationship betweensarcopenia and liver enzymes or fatty liver disease has been rarelystudied. Hong et al. showed that the skeletal muscle index (SMI) isnegatively associated with intrahepatic fat accumulation [13]. Thisprovided a novel insight into the mechanism linking sarcopenia

rg/10.1016/j.dld.2014.12.0142015 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.ndently of body mass index, dietary hab

oo, Dae Won Jun , Kang Nyeong Lee, Hang Lak Leul Yoon, Ho Soon Choi

Internal Medicine, Hanyang University, College of Medicine, Seoul, Republic of Korea

e i n f o a b s t r a c t

Background: Aminotransferase activitsferase in men, and physical

h Young Lee,

surrogate marker of liver injury showing strong correlations

Please cit r elevmass inde /dx.d

ARTICLE IN PRESSG ModelYDLD-2798; No. of Pages 62 K.D. Yoo et al. / Digestive and Liver Disease xxx (2015) xxxxxx

and non-alcoholic fatty liver disease (NAFLD) [13]. However, itis still unclear whether an elevated aminotransferase activity isassociated with sarcopenia in non-obese subjects or in the generalpopulation with few or no risk factors for liver disease.

The effeprovide clienzymes inevaluation iapproachesto identify aactivity, indand amoun

2. Method

2.1. Subject

This stuNational He1) (200920the KNHANPopulation type, the ntive districtmethod to with the powere extracregion. Gropling unit, was samples

The KNHsurvey, a ninterview sviewer in ohypertensioother diseaand, in thishol intake wuse by the measured nmethod, astrained inte

2.2. Selectio

A total underwentabsorptiom472 individmales and fa week, resliver cirrhotest or for were also ethe study (S

2.3. Termin

Appendiwhole-bodyUSA) to meeral contenbody fat petargeted thsubtractingfollows: sta

and bone mineral contents were subtracted, supposing that the tis-sue without bone and fat is skeletal muscle. The SMI was calculatedby dividing the ASM by weight according to the method used inprevious studies [SMI (%) = total skeletal muscle mass (kg)/weight

100] ne stts (19

than

ood c

od teast, wterolDL), iatelthe sal chteroler (Sichemated

atisti

Stuaractmineactorwas f sar

and.05 wicag

ults

ver es

freqn ma(mal1). d lowglucoactivl dels. In

oldeto then snciesevels

amo abnoed (

rrelan bot

en wiver en coLT lee this article in press as: Yoo KD, et al. Sarcopenia is a risk factor fox, dietary habits, and physical activity. Dig Liver Dis (2015), http:/

cts of muscle mass on aminotransferase activity maynicians a better understanding of the elevated liver

non-obese patients who show no risk factors uponn outpatient clinics, and could also suggest therapeutic

for lean NAFLD patients. The purpose of this study was correlation between sarcopenia and aminotransferaseependently of BMI, body fat percentage, dietary habits,t of exercise.

s

s

dy used source data from the Fourth and Fifth Koreaalth and Nutrition Examination Surveys (KNHANES V-10, Korea Centers for Disease Control and Prevention);ES includes every household and individual in its latestand Housing Census. With regard to region and housingation was divided into 29 strata, and 200 administra-s were extracted by applying a proportional allocationmatch the sample component ratios for each stratumpulation stratication variable. Enumeration districtsted to reect the house characteristics in the selectedups of 2030 houses, which made up the third sam-ere extracted from the enumeration districts selected

in systematic sampling.ANES survey was subdivided into a health interview

utrition survey, and a health examination. The healthurvey was carried out face-to-face by a trained inter-rder to identify individuals diagnosed with diabetes,n, hepatitis C, liver cirrhosis, liver cancer, or variousses, as well as to identify the subjects taking medicines

latter case, the type of medicines being taken. Alco-as calculated by multiplying the frequency of alcohol

amount of alcohol consumption. The nutrition surveyutrient intake levels using the one-day 24-hour recall

determined in the one-on-one conversations with arviewer.

n of the subject group

of 15,119 subjects older than 19 years of age who a blood test, dietary survey, and dual-energy X-rayetry (DXA) were initially selected. Of these, we excludeduals who tested positive for hepatitis B antigen, 1197emales who drank more than 210 and 140 g of alcoholpectively, and 19 subjects with a history of hepatitis C,sis, or liver cancer. Patients who did not receive a bloodwhom there was insufcient data on nutrient intakexcluded. Finally, 14,628 individuals were included inupplementary Figure S1).

ology

cular skeletal muscle mass (ASM) was measured by DXA using a fan beam densitometer (Hologic Inc., MA,asure total and regional lean mass (kg), total bone min-ts (kg), BMD (kg/m2), total body fat mass (kg), and totalrcentage. Two X-ray beams with different energy levelse subjects bones. The BMD could then be calculated by

the soft tissue absorption. The ASM was calculated asrting from the total lean mass, the total body fat mass

(kg) least osubjechigher[14].

2.4. Bl

Blohour fcholestein (Himmedlab on for totcholesanalyselectroautom

2.5. St

Theical chto exaother fmass mine ihabits,A p < 0Inc., Ch

3. Res

3.1. Lisubject

Theboth igroup p < 0.00showeblood ferase mineracontroobese,pared

Whfrequeeride las thecies ofincreas

3.2. Colevels i

Whmass, lrelatiowith Aated aminotransferase in men independently of bodyoi.org/10.1016/j.dld.2014.12.014

[9,10,13]. Sarcopenia was dened if the result was as atandard deviation (SD) below the muscle mass of young39 years) [9,10,13]. Abnormal ALT was dened as ALT

30 U/L for males, and higher than 19 U/L for females

hemistry

sts, which were performed on all subjects after a 12-ere conducted for hepatitis B antigen, AST, ALT, total

, low density lipoprotein (LDL), high density lipopro-and triglycerides. Blood samples were centrifuged, theny refrigerated and sent in iceboxes to a single centralame day of blood collection. Routine biochemical testsolesterol, triglycerides, glucose, HDL cholesterol, LDL, ALT, and AST were performed with an ADIVIA 1650emens, Deereld, IL, USA). HBsAg was measured with aniluminescence immunoassay method using an E-170

analyser (Roche, Penzberg, Germany).

cal analysis

dents t-test was used to compare the patients clin-eristics, and partial Pearsons correlations were used

the relationships between skeletal muscle mass ands associated with metabolic syndrome. Skeletal muscledivided into four levels for the analysis. To deter-copenia was a risk factor independent of BMI, dietary

exercise, a logistic regression analysis was performed.as considered statistically signicant. SPSS 17.0 (SPSSo, IL) for Windows was used for all statistical analyses.

nzymes and metabolic parameters in sarcopenic

uency of elevated aminotransferase activity was higherles and females with sarcopenia than in the controles: 26.5% vs. 16.3%, p < 0.001; females: 38.3% vs. 16.3%,Although male and female subjects with sarcopeniaer total calorie intake, they showed higher fastingse, insulin, triglyceride, cholesterol, and aminotrans-ity levels (Table 1). Weekly exercise hours and bonensity were lower in the sarcopenic group than normal

the sarcopenic group we found more subjects that werer, more sedentary, and had lower calorie intake com-e control group, and this was true for both genders.keletal muscle mass was classied into quartiles, the

of abnormal ALT, fasting blood glucose, and triglyc- were negatively correlated with muscle mass. Namely,unt of skeletal muscle mass decreased, the frequen-rmal LFT, fasting blood glucose, and triglyceride levelsFig. 1).

tions between skeletal muscle mass and liver enzymeh genders after adjusting for body mass index and age

e examined the correlations between skeletal musclenzymes, and metabolic parameters using Pearsons cor-efcients, ASM was found to be negatively correlatedvels for both genders (r = 0.222 for males and r = 3.17

Please citemass inde

ARTICLE IN PRESSG ModelYDLD-2798; No. of Pages 6K.D. Yoo et al. / Digestive and Liver Disease xxx (2015) xxxxxx 3

Table 1Comparison of clinical and metabolic parameters according to skeletal muscle index.

Males Females

Normal SMI Sarcopenia *p Normal SMI Sarcopenia *pN = 4064 N = 960 N = 5315 N = 1901

Age (years) 43.3 19.0 48.1 22.4

Please cite this article in press as: Yoo KD, et al. Sarcopenia is a risk factor for elevated aminotransferase in men independently of bodymass index, dietary habits, and physical activity. Dig Liver Dis (2015), http://dx.doi.org/10.1016/j.dld.2014.12.014

ARTICLE IN PRESSG ModelYDLD-2798; No. of Pages 64 K.D. Yoo et al. / Digestive and Liver Disease xxx (2015) xxxxxx

Table 2Partial correlations of skeletal muscle index and clinical parameters.

Partial correlation (I) Age Partial correlation (II) Age, BMI

Male Female Male Female

r *p r *p r *p r *p

ALT (U/L) 0.284

Please cite elevmass inde /dx.d

ARTICLE IN PRESSG ModelYDLD-2798; No. of Pages 6K.D. Yoo et al. / Digestive and Liver Disease xxx (2015) xxxxxx 5

3.3. Multivariate analysis of the effect of skeletal muscle mass onthe frequencies of elevated liver enzymes and fatty liver disease

Multiple logistic regression analysis with abnormal liverenzyme as an increasecopenic maglucose lev(Supplemenobserved in

4. Discussi

BMI, diewith aminomuscle mascontrolling in males. Sshowed grepresenting

Several metabolic ssimilar to thgated the efHong and MDXA, and bcle mass waet al. showfat accumulanism linkisubjects wegeneralizatwhich inclu

An emeris a risk facbody shapecopenia onadjust for fsarcopenia ple ratio beaffected byadjust for Bmass, as wecle mass onimpact of saeffects of mfor women tal muscle dto play an e

To bettemen and wand 2893 psis showed disease did(Supplemendependent liver is not dfact that woimpact of b

Other mdietary haba relationshpendently omuch lowewith the co

subjects were higher than those of the controls (Table 1). Anotherinteresting nding is that low protein intake, and not high fatintake, was correlated with sarcopenia and elevated liver enzymes.Sarcopenic subjects were also characterised by lower serum vita-

levelenic thaned pd as tion.

cohtion te thcidecise.ur st

masndic

squaf sathe easeht; height0% inropris wa

c lowhis stscleopen

for senia,

prelow tlinicaal L

histdiolo

somgic eplaina coulaineare is stu

habethoietarere gatoucat

to pts (0ce atwees are, althl userformc livee theNHA, and. Thi

hav this article in press as: Yoo KD, et al. Sarcopenia is a risk factor forx, dietary habits, and physical activity. Dig Liver Dis (2015), http:/

the dependent variable was performed and showedd frequency of elevated aminotransferase levels in sar-le patients after controlling for age, BMI, smoking status,el, insulin level, as well as dietary and exercise habitstary Table S1). However, this correlation was no longer

women after controlling for BMI.

on

tary habits, and physical exercise have been correlatedtransferase activity. Our results show that the skeletals was correlated with aminotransferase activity afterfor BMI, dietary habits, and amount of exercise onlyarcopenic subjects were older, more sedentary, andater central obesity compared with controls, despitea lower total calorie intake.studies have investigated the impact of sarcopenia onyndrome [9,11,15]. The pathophysiology of NAFLD isat of metabolic syndrome, but few studies have investi-fect of muscle mass on the incidence of NAFLD. Recently,oon used abdominal computed tomography (CT) scan,ody impedance assay (BIA) to show that skeletal mus-s associated with the incidence of NAFLD [13,16]. Honged that SMI is negatively associated with intrahepatication [13]. This provided a novel insight into the mech-ng sarcopenia and NAFLD. Although more than 450re enrolled in their study, a selection bias may limit theion of their results. We used data from the KNHANES,des the most relevant and representative data.ging issue regarding sarcopenia is whether muscle masstor for NAFLD and metabolic disease independently of

or BMI. Therefore, when analysing the impact of sar- fatty liver and metabolic diseases, it is important toat mass and BMI. In Moons study, the denition offollowed the Janssens method [17]. However, the sim-tween body weight and muscle mass may be greatly

the amount of total fat. Likewise, Hong et al did notMI [13]. In the present study, we adjusted for body fatll as total body fat, when analysing the effects of mus-

metabolic disease and fatty liver. We found that thercopenia on liver enzymes was gender-dependent: theuscle mass on liver enzymes were no longer signicantafter controlling for BMI. The absolute volume of skele-iffers between genders, and sex hormones are knownssential role in muscle growth and development [18].r understand the reasons for these differences betweenomen, we compared 4020 premenopausal women

ostmenopausal women. However, multivariate analy-that the effects of ASM on liver enzymes and fatty liver

not differ between pre- or post-menopausal womentary Table S2). Thus, it seems likely that the gender-effect of sarcopenia on both ALT and incidence of fattyue to female sex hormones, but rather results from themen have less muscles than men, resulting in a greaterody fat mass rather than muscle mass.ajor risk factors of abnormal liver enzyme levels areits, physical activity, and age. This study clearly showedip between liver enzymes and peripheral ASM, inde-f BMI, dietary habits, and physical activity. Despite ther total calorie intake of the sarcopenic group comparedntrol group, BMI and body weight of the sarcopenic

min D sarcoprather increasmenderestrica largepopulaevalualevel inof exer

In omuscleOther iby thecator obelow be incrof heigoverwbeing be appand thuspeci

In tthe muon SarcinitionsarcopsideredSDs be

In cabnormcarefuland raever, inradiolocan excopeniunexpis not r

Thidietarythis mterm ddata winvesticic edabilitypatienin Franing bereportSecondalcohonot pechroniDespitfrom Kbiopsysurveytend toated aminotransferase in men independently of bodyoi.org/10.1016/j.dld.2014.12.014

s and less physical activity. The clinical characteristics ofobesity were sedentary lifestyle and low protein intake

high calorie intake or high fat diet. This suggests thathysical activity and protein intake should be recom-lifestyle modications rather than calorie or fat intakeAnother strong point of this study is that it includedort of 14,628 subjects representative of the Koreanand based on the KNHANES data. This allowed us toe impact of sarcopenia on abnormal aminotransferasence after correcting for BMI, dietary habits, and amount

udy, we used the SMI, which is dened as total skeletals (kg)/weight (kg) 100, as an indicator of sarcopenia.ators of sarcopenia were also present. The ASM dividedred height (ASM/height2) is also often used as an indi-rcopenia. Sarcopenia can also be dened as two SDsmean ASM/height2. As the height increases, ASM cand. The indicator ASM/height2 could adjust the inuenceowever, this indicator could be inuenced by obesity. In

subjects the prevalence of sarcopenia was 8.9%, while obese people. Therefore, this indicator seems not toate for obese people in the diagnosis of sarcopenia [19],s not used in our study. Another indicator is the gender-est 20th percentile of ASM.udy, we dened sarcopenia as a value at least 1 SD below

mass of young subjects. The European Working Groupia in Older People (EWGSOP) developed a practical def-arcopenia [10]: in their report, they used SDs to dene

measured in terms of SMI. Class-I sarcopenia was con-sent in subjects whose SMI was between one and twohe mean values for young adults [9,10,13,17,20].l practice, there are several aetiologies that could causeFT. In almost all cases of patients with abnormal LFT, aory and physical examination with a simple blood testgic test can reveal the aetiology of abnormal LFT. How-e cases, the aetiology is unclear after both serologic andvaluation. Our current study suggests that sarcopenia

several abnormal LFT of unclear aetiology. Also, sar-ld be an aetiologic clue for non-obese patients with

d elevation of serum aminotransferase, a condition thatn the Asian population.dy had some limitations. First, the KNHANES surveyedits using the one-day 24-hour recall method. Thus,d does not allow an easy quantication of the long-y habits, since it relies on the subjects memory. Allcollected by well-trained investigators; despite thers were not physicians, they received regular and spe-ion for this large-scale survey providing them a gooderform a survey. In our study, 19 among the 15,000.13%) had liver cirrhosis. Previous studies conductednd UK reported a prevalence of liver cirrhosis rang-n 0.076% and 0.3% [21,22]. Thus, we believe that our

comparable with the results of the previous studies.ough we excluded HBsAg-positive subjects, signicantrs, and subjects with a history of liver disease, we did

further analysis to help revealing the aetiology of ther disease (e.g. HCV-RNA, PCR test, and autoantibodies).

large part of the data used in our study was obtainedNES, platelet counts, abdominal ultrasonography, liver

liver stiffness test were not performed within this samerd, studies analysing large amounts of data generallye low p-values. Although several metabolic parameters

Please cit r elevmass inde /dx.d

ARTICLE IN PRESSG ModelYDLD-2798; No. of Pages 66 K.D. Yoo et al. / Digestive and Liver Disease xxx (2015) xxxxxx

showed correlations with sarcopenia in our study, the presence ofa statistical signicance is not always indicative of clinical signi-cance.

In conclusion, sarcopenia is a risk factor for elevated amino-transferase levels in men, independently of BMI, dietary habits, andphysical activity. Sarcopenic subjects with abnormal aminotrans-ferase levels showed different clinical characteristics comparedwith non-sarcopenic subjects. We believe that not all patients withabnormal liver function test require further examination of mus-cle mass. However, subjects who have abnormal aminotransferaseactivity with uncertain aetiology, and especially those with a nor-mal BMI, need to undergo an estimation of muscle mass.

Conict of interestNone declared.

Funding

This study was supported by a grant from the Korea HealthcareTechnology R&D Project, Ministry of Health & Welfare, Republicof Korea (A121185). The funding source had no role in the studydesign or conduct; in data collection, analysis, or interpretation; orin manuscript preparation, review, or approval.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.dld.2014.12.014.

References

[1] Liu CJ. PAsian peo2012;27:

[2] Yamakawof the T2non-obesGenetics

[3] Aung K, Lvascular dhealthy o2014;99:

[4] Dressler Hto body mthe Acade

[5] Lin YC, Chang PF, Chang MH, et al. Genetic variants in GCKR and PNPLA3 confersusceptibility to nonalcoholic fatty liver disease in obese individuals. AmericanJournal of Clinical Nutrition 2014;99:86974.

[6] Levine ME, Crimmins EM. The impact of insulin resistance and inammation onthe association between sarcopenic obesity and physical functioning. Obesity(Silver Spring) 2012;20:21016.

[7] Esposito K, Chiodini P, Capuano A, et al. Fracture risk and bone mineral densityin metabolic syndrome: a meta-analysis. Journal of Clinical Endocrinology andMetabolism 2013;98:330614.

[8] Lim KI, Yang SJ, Kim TN, et al. The association between the ratio of visceral fatto thigh muscle area and metabolic syndrome: the Korean Sarcopenic ObesityStudy (KSOS). Clinical Endocrinology 2010;73:58894.

[9] Lim S, Kim JH, Yoon JW, et al. Sarcopenic obesity: prevalence and associa-tion with metabolic syndrome in the Korean Longitudinal Study on Health andAging (KLoSHA). Diabetes Care 2010;33:16524.

[10] Cruz-Jentoft AJ, Baeyens JP, Bauer JM, et al. Sarcopenia: European consensus ondenition and diagnosis: report of the European Working Group on Sarcopeniain Older People. Age and Ageing 2010;39:41223.

[11] Guillet C, Boirie Y. Insulin resistance: a contributing factor to age-related musclemass loss? Diabetes and Metabolism 2005;31, 5S2026.

[12] Kim YS, Lee Y, Chung YS, et al. Prevalence of sarcopenia and sarcopenic obe-sity in the Korean population based on the Fourth Korean National Health andNutritional Examination Surveys. Journals of Gerontology Series A, BiologicalSciences and Medical Sciences 2012;67:110713.

[13] Hong HC, Hwang SY, Choi HY, et al. The relationship between sarcopenia andnon-alcoholic fatty liver disease: the Korean sarcopenic obesity study. Hepa-tology 2014;59:17728.

[14] Prati D, Taioli E, Zanella A, et al. Updated denitions of healthy ranges for serumalanine aminotransferase levels. Annals of Internal Medicine 2002;137:110.

[15] Kim TN, Park MS, Lim KI, et al. Skeletal muscle mass to visceral fat arearatio is associated with metabolic syndrome and arterial stiffness: the KoreanSarcopenic Obesity Study (KSOS). Diabetes Research and Clinical Practice2011;93:28591.

[16] Moon JS, Yoon JS, Won KC, et al. The role of skeletal muscle in develop-ment of nonalcoholic fatty liver disease. Diabetes and Metabolism Journal2013;37:27885.

ssen I) in olabilityny AMtors ostrogences man

ociatioiety 2mgar

amon8;147nard s in a troenming Kirrhody. Joue this article in press as: Yoo KD, et al. Sarcopenia is a risk factor fox, dietary habits, and physical activity. Dig Liver Dis (2015), http:/

revalence and risk factors for non-alcoholic fatty liver disease inple who are not obese. Journal of Gastroenterology and Hepatology155560.a-Kobayashi K, Natsume M, Aoki S, et al. The combined effectDM susceptibility genes is an important risk factor for T2DM ine Japanese: a population based case-control study. BMC Medical2012;13:11.orenzo C, Hinojosa MA, et al. Risk of developing diabetes and cardio-isease in metabolically unhealthy normal-weight and metabolicallybese individuals. Journal of Clinical Endocrinology and Metabolism4628., Smith C. Environmental, personal, and behavioral factors are relatedass index in a group of multi-ethnic, low-income women. Journal ofmy of Nutrition and Dietetics 2013;113:16628.

[17] Janniadis

[18] Kendicof eSci

[19] NewassSoc

[20] Baunia199

[21] PoytorGas

[22] Fleof cstuated aminotransferase in men independently of bodyoi.org/10.1016/j.dld.2014.12.014

, Heymseld SB, Ross R. Low relative skeletal muscle mass (sarcope-der persons is associated with functional impairment and physical. Journal of the American Geriatrics Society 2002;50:88996., Dawson L, Kleppinger A, et al. Prevalence of sarcopenia and pre-

f skeletal muscle mass in nonobese women who are long-term usersen-replacement therapy. Journals of Gerontology Series A, Biologicaland Medical Sciences 2003;58:M43640.

AB, Kupelian V, Visser M, et al. Sarcopenia: alternative denitions andns with lower extremity function. Journal of the American Geriatrics003;51:16029.tner RN, Koehler KM, Gallagher D, et al. Epidemiology of sarcope-g the elderly in New Mexico. American Journal of Epidemiology:75563.T, Lebray P, Ingiliz P, et al. Prevalence of liver brosis and risk fac-general population using non-invasive biomarkers (FibroTest). BMCterology 2010;10:1040.M, Aithal GP, Solaymani-Dodaran M, et al. Incidence and prevalence

sis in the United Kingdom, 19922001: a general population-basedrnal of Hepatology 2008;49:7328.

Sarcopenia is a risk factor for elevated aminotransferase in men independently of body mass index, dietary habits, and phy...1 Introduction2 Methods2.1 Subjects2.2 Selection of the subject group2.3 Terminology2.4 Blood chemistry2.5 Statistical analysis

3 Results3.1 Liver enzymes and metabolic parameters in sarcopenic subjects3.2 Correlations between skeletal muscle mass and liver enzyme levels in both genders after adjusting for body mass index ...3.3 Multivariate analysis of the effect of skeletal muscle mass on the frequencies of elevated liver enzymes and fatty liv...

4 DiscussionConflict of interestFundingAppendix A Supplementary dataReferences