ORIGINAL ARTICLE

Sarcopenia, and its association with cardiometabolic andfunctional characteristics in Taiwan: Results from I-LanLongitudinal Aging Study

Li-Kuo Liu,1,3 Wei-Ju Lee,1,4 Liang-Yu Chen,1,2,3 An-Chun Hwang,1,2,3 Ming-Hsien Lin,1,3

Li-Ning Peng1,2,3 and Liang-Kung Chen1,2

1Aging and Health Research Center, 2Institute of Public Health, National Yang Ming University, 3Center for Geriatrics and Gerontology,Taipei Veterans General Hospital, Taipei and 4Department of Family Medicine, Taipei Veterans General Hospital Yuanshan Branch,Yi-Land, Taiwan

Aim: Sarcopenia is a well-recognized geriatric syndrome, which is associated with a variety of adverse outcomes.The present study aimed to evaluate the prevalence of sarcopenia and its associative clinical characteristics in Taiwan.

Methods: Data of the I-Lan Longitudinal Aging Study (ILAS) were retrieved for this study. Sarcopenia was definedaccording to the European Working Group on Sarcopenia in Older People criteria, and comparisons of demographiccharacteristics, physical performance, body composition, cardiometabolic profiles and functionality indicators werecarried out.

Results: Overall, data of 1008 participants (mean age 65.2 ± 9.3 years, male 50.6%) were retrieved for analysis. Thecut-off value of relative appendicular skeletal muscle was 7.0 kg/m2 for men and 5.9 kg/m2 for women. Sarcopenia wassignificantly related to low body mass index, smaller waist circumference, poor nutrition and poor cognition. Themean carotid intima-media thickness and cardiometabolic parameters showed no statistically significant findings.

Conclusions: The present paper showed the epidemiology of sarcopenia, and the strong connection to function-ality indicators. However, sarcopenia was not associated with cardiometabolic risk or carotid intima media thicknessin the present study. Geriatr Gerontol Int 2014; 14 (Suppl. 1): 36–45.

Keywords: cardiovascular disease, community-dwelling elderly, frailty, sarcopenia.

Introduction

In recent years, sarcopenia has been well recognized asa geriatric syndrome,1 and has been defined as an age-related decline in skeletal muscle mass plus low musclestrength and/or low physical performance.2,3 The asso-ciation of sarcopenia with various cardiometabolicdisorders and diseases has been reported,4,5 such asdiabetes,6 chronic pulmonary diseases,7 impaired heartfunctions8,9 and atherosclerosis.10,11 In addition, sarco-penia is also related to impaired functional status, frailtyand poorer quality of life of older adults.12 Sarcopenia isalso associated with adverse clinical consequences, such

as infectious and non-infectious complications of hos-pital inpatients, and all-cause mortality.13,14

The European Working Group on Sarcopenia inOlder People (EWGSOP) recommended a diagno-stic algorithm and standardized measurements ofsarcopenia,3 which has been supported by a great varietyof studies internationally. However, based on differentassessment instruments and cut-off values in differentstudy populations, the prevalence of sarcopenia variedfrom 3% to 30% from study to study.15–17 As the mea-surements of muscle mass and muscle strength wereclosely related to body size and ethnic backgrounds,providing an optimal diagnosis for sarcopenia in Asiancountries is challenging. In particular, the populationdemography forecast clearly showed the escalation ofthe elderly population in Asia, which might conse-quently increase the impact of sarcopenia in Asia.Although Taiwan is known to be one of the fastest agingcountries in the world,18,19 little is known regarding theimpact of sarcopenia in Taiwan.20,21 Therefore, the aim

Accepted for publication 5 November 2013.

Correspondence: Dr Liang-Kung Chen MD PhD, Center forGeriatrics and Gerontology, Taipei Veterans General Hospital,No. 201, Sec 2, Shihpai Road, Taipei, Taiwan. Email:lkchen2@vghtpe.gov.tw

bs_bs_banner

Geriatr Gerontol Int 2014; 14 (Suppl. 1): 36–45

36 | © 2014 Japan Geriatrics Societydoi: 10.1111/ggi.12208

of the present study was to evaluate the prevalence ofsarcopenia among the community-dwelling middle-aged and elderly population in Taiwan, and to providecomprehensive demographic and clinical characteris-tics, physical and mental function, health behaviorsand nutritional status and cardiometabolic profiles forfurther studies of sarcopenia.

Methods

Study design

The I-Lan Longitudinal Aging Study (ILAS) is apopulation-based aging cohort study in I-Lan Countyof Taiwan. ILAS aimed to evaluate the complex inter-relationship between aging, frailty, sarcopenia and cog-nitive decline. Community-dwelling adults aged 50years and older were randomly sampled through thehousehold registrations of the county government inYuanshan Township of I-Lan County in Taiwan.Selected residents were invited to participate by mail ortelephone invitations from the research team, and wereenrolled when they had fully consented and agreed forparticipation. The inclusion criteria were: (i) inhabitantswho then lived in I-Lan County without a plan to movein the near future; and (ii) inhabitants aged 50 years orolder. Any respondents that met any one of the follow-ing conditions were excluded from the study: (i) therespondent was unable to communicate with the inter-viewer and grant an interview; (ii) the respondent had apoor function status, which could lead to a fail in evalu-ation, such as unable to complete a 6-m timed walkwithin a reasonable period of time; (iii) the respondenthad a limited life expectancy (in general, <6 months)because of major illnesses; (iv) the respondent had animplant that was contraindicated for magnetic reso-nance imaging; and (v) currently institutionalizedpeople. The whole study had been approved by theinstitutional review board of the National Yang MingUniversity.

Demographic and physical examinations

A questionnaire seeking information about demograph-ics, smoking habit, habitual alcohol use status, educa-tional years, medical history and comorbidities burdenby Charlson Comorbidity Index22 was administered bytrained interviewers. Extensive functional measure-ments were collected through questionnaires by trainedinterviewers as well, including cognitive status carriedout by the Mini-Mental State Examination (MMSE),23

functional assessment by the Functional AutonomyMeasurement System (SMAF; a tool of a 29-items scaleranging from 0 to 87 points, measured activities of dailyliving, instrumental activities of daily living, mentalfunction, mobility and communications),24 mood and

depression status described by the Center for Epidemio-logic Studies Depression Scale (CES-D),25 and thenutritional status using Mini-Nutrition Assessment(MNA).26

All participants underwent anthropometric measure-ments by research nurses, including height, bodyweightand waist circumferences. Body mass index (BMI) wascalculated as weight (kg) divided by height (m) squared.Systolic blood pressure and diastolic blood pressurewere measured at rest and recorded.

Muscle strength and physical performance

Handgrip strength of the dominant hand was measuredusing digital dynamometers (Smedlay’s Dynamo Meter;TTM, Tokyo, Japan) while the participant was in anupright standing position, with arms down by theirsides. Participants needed to hold the dynamometer inthe dominant hand without squeezing the arm againstthe body, and they were urged to exhibit the greatestpossible force. The best result of three tests was used forstatistical analysis.27

A timed 6-m walk was carried out to evaluate physicalperformance.28,29 Participants walked at their usualwalking speed with a static start without decelerationthroughout a 6-m straight line, while the time con-sumed was taken by a fixed study nurse with a stopwatch (HS-70W, Casio computer co. LTC, Tokyo,Japan).

Body composition

A whole body dual-energy X-ray absorptiometry (DXA)scan was carried out for each participant to measuretotal body fat mass and percentage, and fat-free leanbody mass (LBM) was measured using a Lunar Prodigyinstrument (GE Healthcare, Madison, WI, USA).Appendicular skeletal muscle mass (ASM) was definedby the sum of the lean soft tissue mass of four limbs.30

According to our previous study, height-adjustedmuscle mass was a more suitable skeletal muscle indexfor Asian populations,31 as proposed by Baumgartneret al.32 Therefore, relative appendicular skeletal muscle(RASM), calculated by appendicular skeletal musclemass divided by height (m) squared (ASM / height2;kg/m2) was used in the present study. Bone density wasalso obtained, and then lumbar and hip T-score werecalculated.

Diagnosis of sarcopenia

In the present study, sarcopenia was defined by theEWGSOP criteria,3 as low muscle mass plus low musclestrength (measured by handgrip strength) or low physi-cal performance (measured by usual walking speed).To define the cut-off value of low muscle mass, the

Sarcopenia in Taiwan

© 2014 Japan Geriatrics Society | 37

sex-specific lowest 20% of the study population wasused, as proposed by Newman et al.33 The cut-off valuesof handgrip strength and walking speed were managedin the same approach. Figure 1 presents the diagnosticalgorithm for sarcopenia by the EWGSOP recommen-dations, and cut-off values of muscle mass, musclestrength and walking speed of the present study.

Carotid ultrasonography

In the present study, bilateral carotid arteries in longitu-dinal projections were investigated using an ultrasoundinstrument (GE LOGIQ 400 PRO; GE, Cleveland, OH,USA) equipped with a high-resolution broadband widthlinear array transducer. An experienced techniciancarried out the ultrasound examinations for all partici-pants. The participants were examined in the supineposition. Images were obtained bilaterally of the proximalcommon carotid artery (CCA) to distal CCA, includingbifurcation, internal carotid artery and external carotidartery. Each participant had intima-media thickness(IMT) measured on the far wall of the CCA by longitu-dinal view.34 Mean IMT, the average of the left and rightIMT, was measured for further analysis.

Laboratory examinations

All blood samples were drawn with the participant in theseated position after a 10-h overnight fast. Serum con-centrations of glucose, total cholesterol, low-densitylipoprotein cholesterol (LDL-C) and high-density lipo-protein cholesterol (HDL-C) were determined using an

automatic analyzer (ADVIA 1800, Siemens, Malvern,PA, USA). Whole-blood glycated hemoglobin A1c(HbA1c) was measured by an enzymatic method usingthe Tosoh G8 HPLC Analyzer (Tosoh Bioscience, Inc.,San Francisco, CA, USA).

Serum levels of total testosterone were measuredusing a chemiluminescence immunoassay analyzer(Siemens ADVIA Centaur; Siemens). The intra-assaycoefficient of variation (CV) and inter-assay CV were5.8% and 4.7%, respectively. Dehydroepiandrosteronesulfate (DHEA-S) and sex-hormone binding globulin(SHBG) were measured by using electrochemilu-minometry (Roche Elecsys e411; Roche, Indianapolis,MI, USA). Analytical sensitivity and intra- and inter-assay CV were as follows: DHEA-S: 0.1 µg/dL, 4.02%,2.4%; and SHBG: 0.35 nmol/L; 2.57%, 2.7%, respec-tively. Free T and bio T were calculated from total tes-tosterone, SHBG and albumin by the Vermeulenmethod.35 Then the free androgen index (FAI) wasdefined as total testosterone (nM/L) / SHBG (nM/L).The serum level of growth hormone, homocysteineand insulin-like growth factor-1 (IGF-1) were alsomeasured.

Statistical analysis

All descriptive continuous variables in the text andtables of the present study were reported as mean ±standard deviation, and categorical variables wereexpressed by number (percentage). Comparisonsbetween continuous variables were carried out by Stu-dent’s t-test, and comparisons of categorical data were

Figure 1 Algorithm for sarcopeniadefinition and cut-off values using theI-Lan Longitudinal Aging Studydatabase. The figure was modified fromthe European Working Group onSarcopenia in Older People and AsianWorking Group for Sarcopeniaconsensus. RASM, relativeappendicular skeletal muscle index.

L-K Liu et al.

38 | © 2014 Japan Geriatrics Society

carried out by χ2-test when indicated. The statisticallysignificant factors in the single variable analysis werelisted as independent variables, whereas sarcopenia wasregarded as the dependent variable. Binary logisticregressions with backward Wald method were used todetermine independent risk factors for GS. All statisticalanalysis was carried out by commercial statistical soft-ware (SPSS 16.0; SPSS, Chicago, IL, USA). A P-valueof <0.05 (two-tailed) was considered statisticallysignificant.

Results

In the present study, data of 1008 participants aged 50years and older of ILAS were retrieved, and 25 of them

were excluded as a result of incomplete data. Theirmean age was 65.2 ± 9.3 years, and men accounted for50.6% of participants. Among them, 497 were men(mean age 66.2 ± 9.8 years) and 486 were women (meanage 64.2 ± 8.7 years). Table 1 summarizes the demo-graphic characteristics of the participants. The analysisshowed that BMI was similar between men and women.However, women had higher total body fat percentageand higher total body fat mass than men. On the con-trary, men had more lean body mass, skeletal musclemass and higher skeletal muscle index (RASM). Inphysical performance, men walked faster (1.6 ± 0.5 m/svs 1.4 ± 0.4 m/s, P < 0.001) and had significantly stron-ger handgrip strength than women (34.9 ± 8.3 kg vs21.8 ± 5.3 kg, P < 0.001). In the present study, themultimorbidity of the study participants was low (CCI

Table 1 Demographic characteristics of participants of the I-LanLongitudinal Aging Study

Men(n = 497)

Women(n = 486)

P-value

Age (years) 66.2 ± 9.8 64.2 ± 8.7 <0.001Anthropometric measurements

Height (cm) 163.7 ± 6.1 152.8 ± 5.5 <0.001Weight (kg) 66.9 ± 10.4 58.2 ± 9.9 <0.001Body mass index (kg/m2) 24.9 ± 3.4 24.9 ± 3.9 0.859

Dual-energy X-ray absorptiometryLean body mass (kg) 48.1 ± 5.4 35.6 ± 4.2 <0.001ASM (kg) 21.2 ± 2.9 14.8 ± 2.1 <0.001RASM (kg/m2) 7.9 ± 0.8 6.3 ± 0.7 <0.001Total fat mass (kg) 17.0 ± 6.4 21.6 ± 6.7 <0.001Total body fat percentage (%) 25.4 ± 6.7 37.0 ± 6.4 <0.001Lumbar T-score −0.35 ± 1.50 −1.28 ± 1.39 <0.001Hip T-score −0.77 ± 1.09 −1.27 ± 1.10 <0.001

Physical performanceWalking speed (m/s) 1.6 ± 0.5 1.4 ± 0.4 <0.001Handgrip strength (kg) 34.9 ± 8.3 21.8 ± 5.3 <0.001

Cigarette Smoking (%)Never 35.1 94.5 <0.001Past 27.2 0.8Present 37.7 4.6

Alcohol drinker (%) 51.2 20.8 <0.001Functional status

SMAF −0.3 ± 2.2 −0.2 ± 1.8 0.398CES-D 2.6 ± 4.2 3.9 ± 6.5 <0.001Mini-nutrition assessment 27.3 ± 1.8 26.8 ± 2.1 <0.001MMSE 25.9 ± 3.7 24.7 ± 4.3 <0.001Education, years 6.3 ± 4.9 4.4 ± 4.3 <0.001

Charlson Comorbidity Index 0.9 ± 1.1 0.8 ± 1.2 0.149Mean carotid IMT (mm) 0.7 ± 0.2 0.7 ± 0.1 <0.001

ASM, appendicular skeletal muscle mass; CES-D, the Center for EpidemiologicStudies Depression Scale; IMT, Intima-media thickness; MMSE, Mini-Mental StateExamination; RASM, relative appendicular skeletal muscle index (appendicularskeletal muscle mass / height2); SMAF, the Functional Autonomy MeasurementSystem.

Sarcopenia in Taiwan

© 2014 Japan Geriatrics Society | 39

0.9 ± 1.1 in men and 0.8 ± 1.2 in women) with highlypreserved functional status, low score of depressivesymptoms and well-nourished status. Nevertheless, themean IMT of men were thicker than women.

The participants were divided into two groups foranalysis. For those who were younger than 65 years, thecomparison was carried out between the low musclemass group (defined as the age- and sex-specific RASMmass within the lower 20% of this population) andnormal muscle mass participants. For those aged 65years and older, the diagnostic criteria of sarcopenia wereapplied, and the comparisons were made between par-ticipants with and without sarcopenia. Figure 1 showsthe diagnosis of sarcopenia algorithm modified from theEWGSOP algorithm,3 and 25 men (9.4%) and 21 women(9.8%) were categorized as having sarcopenia.

Table 2 summarized the comparisons between par-ticipants aged younger than 65 years with a normalamount or low muscle mass in both sexes. In bothsexes, participants with low muscle mass were signifi-cantly thinner (both BMI and waist circumference) thanthose with normal muscle mass. However, the age wasnot statistically different between the two groups in bothsexes. Furthermore, participants with low muscle masshad less lean body mass and less appendicular skeletalmuscle mass when compared with their compatriots.The total body fat percentage showed no significantdifferences between groups in both sexes. Other asso-ciated factors are listed in Table 2.

Table 3 summarizes the comparisons between par-ticipants with and without sarcopenia in both sexes. Inboth sexes, participants with sarcopenia were signifi-cantly older and thinner (both BMI and waist circum-ference [in men]) than those without sarcopenia.Furthermore, sarcopenic participants also had less leanbody mass and less appendicular skeletal muscle masswhen compared with non-sarcopenic participants. Thetotal body fat percentage showed no significant differ-ences between groups in both sexes.

Participants with sarcopenia had a higher cigarettesmoking rate in women, but no differences in habitualalcohol consumption in both sexes. The sarcopeniagroup, compared with those without sarcopenia, had ahigher Charlson Comorbidity Index, poorer cognitiveperformance based on MMSE score, fewer educationalyears and poorer nutritional status by Mini-NutritionAssessment. The participants with sarcopenia alsoshowed a trend of poorer functional state by SMAF andmore depressive symptoms by CES-D, though theywere not statistically different. In terms of mean carotidIMT and cardiometabolic parameters, no significantdifferences could be identified between groups with orwithout sarcopenia in both sexes.

Considering hormonal change, there were no statis-tical differences in growth hormone, homocysteine orIGF-1. The serum free androgen index was lower in

the sarcopenic group compared with the other groupin both sexes, and the differences in dehydroepian-drosterone sulfate (DHEA-S) were borderline.

The statistically significant factors in the single vari-able analysis, including age, BMI, cigarette smoking,SMAF, MNA, MMSE, CCI, serum growth hormoneand DHEA-S level were listed as independent univariatevariables, whereas sarcopenia was regarded as thedependent variable. By using binary logistic regression(backward Wald method), we found that older age (oddsratio [OR] 1.112, 95% confidence interval [CI] 1.041–1.188; P = 0.002) and a lower BMI (OR 0.774, 95% CI0.679–0.883; P < 0.001) were both independent riskfactors for the presence of sarcopenia. (Table 4)

Discussion

The present study showed the prevalence of sarcopeniaamong community-dwelling middle-aged and elderlypeople in Taiwan. The prevalence of sarcopenia amongthe ILAS men and women (mean age 65 years) was9.4% and 9.8%, respectively. Female sex was regardedas a risk factor of sarcopenia and a higher prevalencehas been reported before,1,36 but many Asian andother international studies showed the prevalence ofsarcopenia to be lower in women than in men.37,38 Thecut-off values of relative skeletal muscle index in thepresent study (men: 7.0 kg/m2, women: 5.9 kg/m2) weresimilar to the Rosetta study,32 Health ABC study39 andsome Asian reports.40 However, unlike our previousstudy, the cut-off of skeletal muscle mass index was notdetermined by the sex-specific young reference group.21

A similar approach has been reported by Newman, et al.,which might be more suitable for Asian populations,because many Asian sarcopenia researchers encoun-tered the same challenge that the prevalence ofsarcopenia was extremely low in elderly women.21,40 Toovercome this difficulty, some researchers prefer usingweight-adjusted skeletal muscle mass instead of height-adjusted skeletal muscle mass. However, it has beenclearly shown that the “low muscle mass” defined byheight-adjusted and weight-adjusted skeletal muscleindex eventually differ from each other greatly.41 To bemore compatible with functional status, we prefer usingheight-adjustment rather than weight-adjustment.However, using the 20th percentile of height-adjustedskeletal muscle index of the study sample as the cut-offvalue of low muscle mass might be a better approach forsarcopenia.

The prevalence of sarcopenia among men and womenaged 65 years and older in ILAS was 9.4% and 9.8%,and would increase to 18.6% and 15.0% in men andwomen aged 80 years and older. The increasing preva-lence along with aging is similar to that in theiISIRENTE study,15 and many others. In the presentstudy, sarcopenic participants were significantly thinner

L-K Liu et al.

40 | © 2014 Japan Geriatrics Society

Table 2 Comparisons between middle-aged participants with normal and low muscle mass by sex

Men P-value Women P-valueNormal Low muscle

massNormal Low muscle

mass

n 188 42 223 49Age (years) 57.2 ± 3.9 57.8 ± 4.1 0.380 57.5 ± 3.7 57.7 ± 3.9 0.812Body mass index (kg/m2) 26.1 ± 3.4 22.8 ± 2.7 <0.001 25.4 ± 3.9 21.5 ± 2.1 <0.001Waist circumference (cm) 88.8 ± 8.4 83.9 ± 7.6 0.001 82.9 ± 12.8 75.7 ± 6.6 <0.001

Dual-energy X-ray absorptiometryLean body mass (kg) 51.5 ± 4.9 45.3 ± 3.1 <0.001 36.6 ± 4.1 31.3 ± 2.6 <0.001ASM (kg) 23.1 ± 2.5 19.5 ± 1.4 <0.001 15.5 ± 1.9 12.7 ± 1.3 <0.001RASM (ASM/ht2) 8.4 ± 0.6 7.2 ± 0.3 <0.001 6.5 ± 0.7 5.3 ± 0.3 <0.001Total fat mass (kg) 18.5 ± 6.9 15.0 ± 5.5 0.001 22.6 ± 6.8 19.0 ± 4.6 <0.001Total body fat percentage (%) 25.7 ± 6.9 24.2 ± 6.1 0.205 37.5 ± 6.1 37.3 ± 6.1 0.821Lumbar T-score −0.29 ± 1.21 −0.47 ± 1.60 0.509 −0.77 ± 1.32 −1.13 ± 1.06 0.078Hip T-score −0.40 ± 0.96 −0.75 ± 1.13 0.041 −0.83 ± 1.00 −1.39 ± 0.94 <0.001

Physical performanceWalking speed (m/s) 1.7 ± 00.5 1.7 ± 0.6 0.950 1.5 ± 0.4 1.6 ± 0.4 0.001Handgrip strength (kg) 40.2 ± 6.6 37.4 ± 6.9 0.018 23.9 ± 4.8 22.3 ± 4.9 <0.001

Cigarette Smoking (%)Never 43.7 35.7 0.345 95.0 95.7 0.820Past 20.8 16.7 0 0Present 35.5 47.6 5.0 4.3

Alcohol drinker (%) 67.0 64.3 0.734 28.3 23.4 0.494Functional status

SMAF −0.40 ± 2.54 −1.94 ± 5.56 0.193 −0.17 ± 0.65 −2.79 ± 8.13 0.156CES-D 2.7 ± 4.4 3.4 ± 4.7 0.495 3.7 ± 6.2 4.2 ± 6.7 0.681Mini-nutrition assessment 27.1 ± 1.9 25.9 ± 2.4 0.005 26.9 ± 1.8 24.9 ± 3.6 0.022MMSE 24.7 ± 3.7 21.1 ± 4.8 0.001 22.5 ± 4.1 20.5 ± 5.4 0.030Education (years) 4.4 ± 4.5 1.6 ± 2.2 <0.001 1.6 ± 2.6 0.9 ± 1.8 0.209

Charlson Comorbidity Index 1.2 ± 1.2 1.9 ± 1.3 0.025 1.2 ± 1.3 1.7 ± 1.4 0.106Mean carotid IMT (mm) 0.8 ± 0.2 0.8 ± 0.1 0.418 0.7 ± 0.1 0.7 ± 0.1 0.367Cardiometabolic parameters

SBP (mmHg) 136.4 ± 17.7 134.8 ± 14.0 0.667 137.1 ± 19.9 130.7 ± 12.9 0.150DBP (mmHg) 81.6 ± 12.6 80.5 ± 9.1 0.675 82.3 ± 13.7 78.8 ± 12.4 0.265Fasting blood glucose (mg/dL) 100.1 ± 18.6 98.9 ± 35.6 0.792 102.8 ± 24.2 102.7 ± 19.6 0.992HbA1c (%) 6.0 ± 0.8 6.1 ± 1.2 0.853 6.3 ± 0.9 6.3 ± 1.2 0.818Total cholesterol (mg/dL) 184.7 ± 33.3 185.3 ± 34.4 0.927 197.1 ± 35.2 190.6 ± 36.6 0.407LDL-cholesterol (mg/dL) 115.1 ± 30.8 117.1 ± 32.4 0.760 118.6 ± 33.0 110.7 ± 29.9 0.271HDL-cholesterol (mg/dL) 50.4 ± 12.2 46.4 ± 10.2 0.118 56.5 ± 13.9 60.9 ± 13.9 0.152

Hormones and EndocrinesGrowth hormone (ng/mL) 0.47 ± 0.78 1.12 ± 2.44 0.197 0.63 ± 1.05 0.58 ± 0.64 0.818Free Androgen Index (%) 31.4 ± 10.2 27.6 ± 8.5 0.078 2.1 ± 1.9 0.9 ± 0.6 <0.001DHEA-S (ug/dL) 99.9 ± 60.4 82.7 ± 38.2 0.051 66.5 ± 48.3 46.0 ± 33.4 0.050Homocysteine (umol/L) 16.5 ± 7.4 15.6 ± 5.8 0.550 13.7 ± 5.0 15.2 ± 8.4 0.383IGF-1 (ng/mL) 120.4 ± 52.6 115.0 ± 56.0 0.629 111.6 ± 51.8 93.9 ± 40.0 0.116

ASM, appendicular skeletal muscle mass; CES-D, the Center for Epidemiologic Studies Depression Scale; DBP, diastolic bloodpressure; DHEA-S, dehydroepiandrosterone sulfate; HbA1c, glycated hemoglobin; IGF-1, insulin-like growth factor-1; IMT,intima-media thickness; MMSE, Mini-Mental State Examination; RASM, relative appendicular skeletal muscle index(appendicular skeletal muscle mass / height2); SBP, systolic blood pressure; SMAF, the Functional Autonomy MeasurementSystem.

Sarcopenia in Taiwan

© 2014 Japan Geriatrics Society | 41

Table 3 Sex-specific comparisons of demographics and serum biomarkers in older adults with and withoutsarcopenia

Men P-value Women P-valueNormal Sarcopenia Normal Sarcopenia

n 242 25 193 21Age (years) 73.5 ± 6.1 78.5 ± 4.1 <0.001 72.2 ± 4.9 76.4 ± 5.2 <0.001Body mass index (kg/m2) 24.7 ± 3.1 21.7 ± 2.7 <0.001 25.4 ± 3.6 22.6 ± 4.2 0.001Waist circumference (cm) 88.9 ± 9.0 83.5 ± 11.4 0.006 86.0 ± 10.3 83.4 ± 11.1 0.277

Dual-energy X-ray absorptiometryLean body mass (kg) 47.0 ± 4.7 41.1 ± 3.4 <0.001 36.0 ± 3.8 31.7 ± 3.1 <0.001ASM (kg) 20.5 ± 2.4 17.1 ± 1.6 <0.001 14.8 ± 1.9 12.5 ± 1.3 <0.001RASM (ASM/ht2) 7.7 ± 0.7 6.6 ± 0.5 <0.001 6.5 ± 0.6 5.5 ± 0.4 <0.001Total fat mass (kg) 16.7 ± 5.9 13.4 ± 6.4 0.009 21.3 ± 6.4 19.2 ± 8.7 0.165Total body fat percentage (%) 25.6 ± 6.6 23.6 ± 7.3 0.140 36.5 ± 6.3 35.9 ± 9.9 0.781Lumbar T-score −0.33 ± 1.65 −0.62 ± 1.60 0.412 −1.78 ± 1.30 −2.36 ± 1.42 0.064Hip T-score −0.95 ± 1.04 −1.63 ± 1.41 0.004 −1.61 ± 1.01 −2.28 ± 1.37 0.007

Physical performanceWalking speed (m/s) 1.5 ± 0.5 1.1 ± 0.3 <0.001 1.2 ± 0.4 0.9 ± 0.2 0.001Handgrip strength (kg) 31.7 ± 6.9 23.0 ± 5.5 <0.001 20.3 ± 4.5 13.5 ± 4.3 <0.001

Cigarette smoking (%)Never 29.5 20.8 0.569 94.7 85.7 <0.001Past 32.5 41.7 0.5 14.3Present 38.0 37.5 4.8 0

Alcohol drinker (%) 37.6 41.7 0.696 12.2 14.3 0.780Functional status

SMAF −0.40 ± 2.54 −1.94 ± 5.56 0.193 −0.17 ± 0.65 −2.79 ± 8.13 0.156CES-D 2.7 ± 4.4 3.4 ± 4.7 0.495 3.6 ± 6.2 4.6 ± 6.9 0.484Mini-Nutrition Assessment 27.1 ± 1.9 25.9 ± 2.4 0.005 26.9 ± 1.8 24.9 ± 3.6 0.022MMSE 24.7 ± 3.7 21.1 ± 4.8 0.001 22.6 ± 4.1 20.2 ± 5.4 0.015Education (years) 4.4 ± 4.5 1.6 ± 2.2 <0.001 1.6 ± 2.6 0.9 ± 1.9 0.126

Charlson Comorbidity Index 1.2 ± 1.2 1.9 ± 1.3 0.025 1.2 ± 1.3 1.9 ± 1.4 0.030Mean carotid IMT (mm) 0.8 ± 0.2 0.8 ± 0.1 0.418 0.7 ± 0.1 0.7 ± 0.1 0.283Cardiometabolic parameters

SBP (mmHg) 136.4 ± 17.7 134.8 ± 14.0 0.667 137.1 ± 19.9 130.7 ± 12.9 0.150DBP (mmHg) 81.6 ± 12.6 80.5 ± 9.1 0.675 82.3 ± 13.7 78.8 ± 12.4 0.265Fasting blood glucose (mg/dL) 100.1 ± 18.6 98.9 ± 35.6 0.792 102.8 ± 24.1 102.9 ± 20.6 0.980HbA1c (%) 6.0 ± 0.8 6.1 ± 1.2 0.853 6.3 ± 0.9 6.3 ± 1.2 0.677Total cholesterol (mg/dL) 184.7 ± 33.3 185.3 ± 34.4 0.927 197.3 ± 35.2 187.7 ± 35.4 0.234LDL-cholesterol (mg/dL) 115.1 ± 30.8 117.1 ± 32.4 0.760 118.7 ± 33.0 109.1 ± 29.4 0.202HDL-cholesterol (mg/dL) 50.4 ± 12.2 46.4 ± 10.2 0.118 56.7 ± 13.9 59.9 ± 14.1 0.320

Hormones and endocrinesGrowth hormone (ng/mL) 0.47 ± 0.78 1.12 ± 2.44 0.197 0.63 ± 1.05 0.58 ± 0.64 0.818Free Androgen Index (%) 31.4 ± 10.2 27.6 ± 8.5 0.078 2.1 ± 1.9 0.9 ± 0.6 <0.001DHEA-S (µg/dL) 99.9 ± 60.4 82.7 ± 38.2 0.051 66.5 ± 48.3 46.0 ± 33.4 0.050Homocysteine (umol/L) 16.5 ± 7.4 15.6 ± 5.8 0.550 13.7 ± 5.0 15.2 ± 8.4 0.383IGF-1 (ng/mL) 120.4 ± 52.6 115.0 ± 56.0 0.629 111.6 ± 51.8 93.9 ± 40.0 0.116

ASM, appendicular skeletal muscle mass; CES-D, the Center for Epidemiologic Studies Depression Scale; DBP, diastolic bloodpressure; DHEA-S, dehydroepiandrosterone sulfate; HbA1c, glycated hemoglobin; IGF-1, insulin-like growth factor-1; IMT,intima-media thickness; MMSE, Mini-Mental State Examination; RASM, relative appendicular skeletal muscle index(appendicular skeletal muscle mass / height2); SBP, systolic blood pressure; SMAF, the Functional Autonomy MeasurementSystem.

L-K Liu et al.

42 | © 2014 Japan Geriatrics Society

than non-sarcopenic participants of both sexes.Sarcopenic participants also had lower lean body massand lower appendicular skeletal muscle mass comparedwith non-sarcopenic participants. Domiciano et al. alsoshowed that more than 95% of sarcopenic participantsdefined by RASM were lean.42 In theory, sarcopenia ismore prone to co-occur among frail older people withlow BMI, and is more likely to be associated withadverse consequences. In terms of obesity in the elderly,it is clear that mortality and morbidity only increasewhen a BMI is higher than 30 kg/m2,43 which is far moreobese than the present study population. Furtheroutcome-based cohort research is required to clarify theclinical impact of sarcopenia defined by height-adjustedor weight-adjusted skeletal muscle index.

In the present study, sarcopenic participants showedpoorer nutritional status and poorer cognitive functionwith fewer educational years compared with thosewithout sarcopenia. However, the physical function anddepressive symptoms were not significantly differentbetween participants with and without sarcopenia,which was consistent with many previous studies.44–46

The higher multimorbidities in sarcopenic participantsin the present study was compatible with many studies,because the prevalence of metabolic syndrome and car-diovascular disease were more common in sarcopenicparticipants.47,48 However, the cardiometabolic profile ofsarcopenic participants was similar to non-sarcopenicparticipants in the present study. Besides, a close rela-tionship between sarcopenia and bone mass has beenreported,49,50 which was compatible with the presentstudy, especially in women.

In the present study, the mean carotid IMT and allcardiometabolic parameters were similar between par-ticipants with and without sarcopenia. Peters, et al.showed that the carotid intima-media was moreecholucent in older adults than in middle-aged people,51

which could underestimate the IMT of the studyparticipants. A strong relationship between reducedthigh muscle mass area and carotid IMT has beenreported,11,52 and physical activity was favorably associ-ated with IMT in adolescents.53 In theory, based on theaforementioned discoveries, sarcopenia should be posi-tively associated with carotid IMT in the elderly, butthis association was not identified in the presentstudy. Similar to the present results, Abe, et al. alsoshowed that the cardiovascular risk parameters weresimilar among participants with or without sarcopenia.54

However, some studies clearly showed the associationbetween metabolic syndrome and sarcopenic obesity orfrailty,48,55,56 which deserve further investigation in thefuture.

Despite the extensive effort that went into the presentstudy, there were still some limitations. First, the studyhad a cross-sectional study design, so it was not possibleto obtain the causal relationship. However, as ILAS isa longitudinal cohort study per se, we believe moreoutcome-based study results will be available in the nearfuture. Second, the determination of cut-offs for lowmuscle mass, low muscle strength and low physicalperformance was based on the study sample. The ILASparticipants were relatively healthier than community-dwelling older adults because we basically excluded par-ticipants with any disability. Therefore, the study results

Table 4 Binary logistic regression on sarcopenia in I-Lan Longitudinal Aging Study older population

Dependent Independent OR (95% CI)Univariate Multivariate*

Sarcopenia Age 1.135 (1.079–1.194) 1.112 (1.041–1.188)Sex

WomenMen

Reference0.858 (0.472–1.559)

BMI 0.748 (0.670–0.835) 0.774 (0.679–0.883)Cigarette smoking

Never ReferencePast 0.545 (0.222–1.342)Present 0.554 (0.268–1.146)

SMAF 0.881 (0.809–0.961)MNA 0.747 (0.657–0.850)MMSE 0.863 (0.808–0.922)CCI 1.382 (1.088–1.757)Growth hormone 1.221 (0.992–1.053)DHEA-S 0.992 (0.986–0.999)

*Backward Wald method was used in multivariate binary logistic regression with all variables entered at the beginning. Oddsratios and confidence intervals were presented only for significant variables with P < 0.05. BMI, body mass index; CCI, CharlsonComorbidity Index; CI, confidence interval; DHEA-S, dehydroepiandrosterone sulfate; MMSE, Mini-Mental State Examination;MNA, Mini-Nutrition Assessment; OR, odds ratio; SMAF, the Functional Autonomy Measurement System.

Sarcopenia in Taiwan

© 2014 Japan Geriatrics Society | 43

might underestimate the prevalence of sarcopenia andits associated health impact. However, the recruitmentstrategy was of great help in providing the functionaltrajectory of healthy older people living in communities.Third, study participants were recruited from a subur-ban area, and they were leading a more physically activelifestyle, which might be different from subjects enrolledfrom urban areas.

In conclusion, the prevalence of sarcopenia in thepresent study was similar to most previous Asianreports, and was lower than that from Western coun-tries. Sarcopenic participants were slightly older,remarkably thinner, lower in educational levels, morelikely to smoke (in women), lower in bone mineraldensity, with poorer cognitive function and poorernutritional status. However, sarcopenia was not associ-ated with cardiometabolic risk and carotid IMT. Furtherstudy is required to evaluate the clinical impact ofsarcopenia, and the interrelationship with physical,mental function and mood status.

Disclosure statement

The authors declare no conflict of interest.

References

1 Cruz-Jentoft AJ, Landi F, Topinkova E, Michel JP. Under-standing sarcopenia as a geriatric syndrome. Curr Opin ClinNutr Metab Care 2010; 13: 1–7.

2 Kavita Karnik DJM. Review of tools and technologies toassess multi-system functional impairment and frailty. ClinMed Geriatr 2009; 3: 1–8.

3 Cruz-Jentoft AJ, Baeyens JP, Bauer JM et al. Sarcopenia:European consensus on definition and diagnosis: report ofthe European Working Group on Sarcopenia in OlderPeople. Age Ageing 2010; 39: 412–423.

4 Goodpaster BH, Park SW, Harris TB et al. The loss ofskeletal muscle strength, mass, and quality in older adults:the health, aging and body composition study. J Gerontol ABiol Sci Med Sci 2006; 61: 1059–1064.

5 Karakelides H, Nair KS. Sarcopenia of aging and its meta-bolic impact. Curr Top Dev Biol 2005; 68: 123–148.

6 Park SW, Goodpaster BH, Lee JS et al. Excessive loss ofskeletal muscle mass in older adults with type 2 diabetes.Diabetes Care 2009; 32: 1993–1997.

7 Cesari M, Pedone C, Chiurco D et al. Physical perfor-mance, sarcopenia and respiratory function in olderpatients with chronic obstructive pulmonary disease. AgeAgeing 2012; 41: 237–241.

8 Zamboni M, Rossi A, Corzato F, Bambace C, Mazzali G,Fantin F. Sarcopenia, Cachexia and Congestive HeartFailure in the Elderly. Endocr Metab Immune Disord DrugTargets 2013; 13: 58–67.

9 von Haehling S, Steinbeck L, Doehner W, Springer J,Anker SD. Muscle wasting in heart failure: an overview. IntJ Biochem Cell Biol 2013; 45: 2257–2265.

10 Lee JS, Auyeung TW, Kwok T, Lau EM, Leung PC, WooJ. Associated factors and health impact of sarcopenia in

older chinese men and women: a cross-sectional study.Gerontology 2007; 53: 404–410.

11 Ochi M, Kohara K, Tabara Y et al. Arterial stiffness isassociated with low thigh muscle mass in middle-aged toelderly men. Atherosclerosis 2010; 212: 327–332.

12 Cesari M, Pahor M, Lauretani F et al. Skeletal muscle andmortality results from the InCHIANTI Study. J Gerontol ABiol Sci Med Sci 2009; 64: 377–384.

13 Metter EJ, Talbot LA, Schrager M, Conwit R. Skeletalmuscle strength as a predictor of all-cause mortality inhealthy men. J Gerontol A Biol Sci Med Sci 2002; 57: B359–B365.

14 Cosqueric G, Sebag A, Ducolombier C, Thomas C, PietteF, Weill-Engerer S. Sarcopenia is predictive of nosocomialinfection in care of the elderly. Br J Nutr 2006; 96: 895–901.

15 Landi F, Cruz-Jentoft AJ, Liperoti R et al. Sarcopenia andmortality risk in frail older persons aged 80 years and older:results from ilSIRENTE study. Age Ageing 2013; 42: 203–209.

16 Abellan van Kan G. Epidemiology and consequences ofsarcopenia. J Nutr Health Aging 2009; 13: 708–712.

17 Tichet J, Vol S, Goxe D, Salle A, Berrut G, Ritz P. Preva-lence of sarcopenia in the French senior population. J NutrHealth Aging 2008; 12: 202–206.

18 Chen LK, Inoue H, Won CW et al. Challenges of urbanaging in Taiwan: summary of urban aging forum. J ClinGerontol Geriatr 2013; 4: 97–101.

19 Chen CY. Meeting the challenges of eldercare in Taiwan’saging society. J Clin Gerontol Geriatr 2010; 1: 2–4.

20 Lin CC, Li CI, Meng NH et al. Frailty and its associatedfactors in an elderly taiwanese metropolitan population.J Am Geriatr Soc 2013; 61: 292–294.

21 Lee WJ, Liu LK, Peng LN, Lin MH, Chen LK. Compari-sons of sarcopenia defined by IWGS and EWGSOP criteriaamong older people: results from the I-Lan longitudinalaging study. J Am Med Dir Assoc 2013; 14: 528 e1–528 e7.

22 Charlson ME, Pompei P, Ales KL, MacKenzie CR. A newmethod of classifying prognostic comorbidity in longitudi-nal studies: development and validation. J Chronic Dis 1987;40: 373–383.

23 Folstein MF, Folstein SE, McHugh PR. “Mini-mentalstate”. A practical method for grading the cognitive stateof patients for the clinician. J Psychiatr Res 1975; 12: 189–198.

24 Hebert R, Carrier R, Bilodeau A. The FunctionalAutonomy Measurement System (SMAF): description andvalidation of an instrument for the measurement of handi-caps. Age Ageing 1988; 17: 293–302.

25 Radloff LS. The CES-D Scale: a self-report depressionscale for research in the general population. Appl PsycholMeas 1977; 1: 385–401.

26 Guigoz Y. The Mini Nutritional Assessment (MNA) reviewof the literature–What does it tell us? J Nutr Health Aging2006; 10: 466–485; discussion 85-7.

27 Xue QL, Bandeen-Roche K, Varadhan R, Zhou J, FriedLP. Initial manifestations of frailty criteria and the devel-opment of frailty phenotype in the Women’s Health andAging Study II. J Gerontol A Biol Sci Med Sci 2008; 63:984–990.

28 Witham MD, Sumukadas D, McMurdo ME. ACE inhibi-tors for sarcopenia–as good as exercise training? Age Ageing2008; 37: 363–365.

29 Rolfson DB, Majumdar SR, Tsuyuki RT, Tahir A,Rockwood K. Validity and reliability of the Edmonton FrailScale. Age Ageing 2006; 35: 526–529.

30 Woods JL, Iuliano-Burns S, King SJ, Strauss BJ, WalkerKZ. Poor physical function in elderly women in low-level

L-K Liu et al.

44 | © 2014 Japan Geriatrics Society

aged care is related to muscle strength rather than tomeasures of sarcopenia. Clin Interv Aging 2011; 6: 67–76.

31 Liu LK, Lee WJ, Liu CL et al. Age-related skeletal musclemass loss and physical performance in Taiwan: implica-tions to diagnostic strategy of sarcopenia in Asia. GeriatrGerontol Int 2013; 13: 964–971.

32 Baumgartner RN, Koehler KM, Gallagher D et al. Epide-miology of sarcopenia among the elderly in New Mexico.Am J Epidemiol 1998; 147: 755–763.

33 Newman AB, Kupelian V, Visser M et al. Sarcopenia: alter-native definitions and associations with lower extremityfunction. J Am Geriatr Soc 2003; 51: 1602–1609.

34 Kanters SD, Algra A, van Leeuwen MS, Banga JD. Repro-ducibility of in vivo carotid intima-media thickness mea-surements: a review. Stroke 1997; 28: 665–671.

35 Vermeulen A, Verdonck L, Kaufman JM. A critical evalu-ation of simple methods for the estimation of free testos-terone in serum. J Clin Endocrinol Metab 1999; 84 (10):3666–3672.

36 Janssen I, Heymsfield SB, Low RR. relative skeletal musclemass (sarcopenia) in older persons is associated with func-tional impairment and physical disability. J Am Geriatr Soc2002; 50: 889–896.

37 Lim S, Kwon SY, Yoon JW et al. Association between bodycomposition and pulmonary function in elderly people: theKorean Longitudinal Study on Health and Aging. Obesity(Silver Spring) 2011; 19: 631–638.

38 Pongchaiyakul C, Limpawattana P, Kotruchin P,Rajatanavin R. Prevalence of sarcopenia and associatedfactors among Thai population. J Bone Miner Metab 2013;31: 346–350.

39 Delmonico MJ, Harris TB, Lee JS et al. Alternative defini-tions of sarcopenia, lower extremity performance, andfunctional impairment with aging in older men andwomen. J Am Geriatr Soc 2007; 55: 769–774.

40 Kim YS, Lee Y, Chung YS et al. Prevalence of sarcopeniaand sarcopenic obesity in the Korean population based onthe Fourth Korean National Health and NutritionalExamination Surveys. J Gerontol A Biol Sci Med Sci 2012; 67:1107–1113.

41 Bijlsma AY, Meskers CG, Ling CH et al. Definingsarcopenia: the impact of different diagnostic criteria onthe prevalence of sarcopenia in a large middle aged cohort.Age (Dordr) 2013; 35: 871–881.

42 Figueiredo CP, Domiciano DS, Lopes JB et al. Prevalenceof sarcopenia and associated risk factors by two diagnosticcriteria in community-dwelling older men: the São PauloAgeing & Health Study (SPAH). Osteoporos Int 2013; doi:10.1007/s00198-013-2455-x.

43 Mathus-Vliegen EM. Obesity and the elderly. J ClinGastroenterol 2012; 46: 533–544.

44 Kim JK, Choi SR, Choi MJ et al. Prevalence of and factorsassociated with sarcopenia in elderly patients with end-stage renal disease. Clin Nutr 2013; pii: S0261-5614(13)00102-7. doi:10.1016/j.clnu.2013.04.002

45 Bottiggi KHA. The association between change in motorfunction and cognition in older adults: a descriptive review.Phys Ther Rev 2008; 13: 91–101.

46 Bahat G, Saka B, Tufan F et al. Prevalence of sarcopeniaand its association with functional and nutritional statusamong male residents in a nursing home in Turkey. AgingMale 2010; 13: 211–214.

47 Stephen WC, Janssen I. Sarcopenic-obesity and cardiovas-cular disease risk in the elderly. J Nutr Health Aging 2009;13: 460–466.

48 Chung JY, Kang HT, Lee DC, Lee HR, Lee YJ. Bodycomposition and its association with cardiometabolic riskfactors in the elderly: a focus on sarcopenic obesity. ArchGerontol Geriatr 2013; 56: 270–278.

49 Verschueren S, Gielen E, O’Neill TW et al. Sarcopenia andits relationship with bone mineral density in middle-agedand elderly European men. Osteoporos Int 2013; 24: 87–98.

50 Capozza RF, Cointry GR, Cure-Ramirez P, Ferretti JL,Cure-Cure C. A DXA study of muscle-bone relationshipsin the whole body and limbs of 2512 normal men and pre-and post-menopausal women. Bone 2004; 35: 283–295.

51 Peters SA, Lind L, Palmer MK et al. Increased age, highbody mass index and low HDL-C levels are related to anecholucent carotid intima-media: the METEOR study.J Intern Med 2012; 272: 257–266.

52 Kato A, Ishida J, Endo Y et al. Association of abdominalvisceral adiposity and thigh sarcopenia with changes ofarteriosclerosis in haemodialysis patients. Nephrol DialTransplant 2011; 26: 1967–1976.

53 Pahkala K, Heinonen OJ, Simell O et al. Association ofphysical activity with vascular endothelial function andintima-media thickness. Circulation 2011; 124: 1956–1963.

54 Abe T, Thiebaud RS, Loenneke JP, Bemben MG, LoftinM, Fukunaga T. Influence of severe sarcopenia on cardio-vascular risk factors in nonobese men. Metab Syndr RelatDisord 2012; 10: 407–412.

55 Sanada K, Iemitsu M, Murakami H et al. Adverse effects ofcoexistence of sarcopenia and metabolic syndrome in Japa-nese women. Eur J Clin Nutr 2012; 66: 1093–1098.

56 Lee JS, Auyeung TW, Leung J, Kwok T, Leung PC, Woo J.Physical frailty in older adults is associated with metabolicand atherosclerotic risk factors and cognitive impairmentindependent of muscle mass. J Nutr Health Aging 2011; 15:857–862.

Sarcopenia in Taiwan

© 2014 Japan Geriatrics Society | 45