INTRODUCTION
Thailand has entered a completed aging society, with older adults accounting for 20.83% of the population in 2024.1) As the number of older adults continues to increase, age-related conditions are becoming more common. Sarcopenia is one of the most prevalent conditions among older adults, affecting approximately 20.7% of the Thai older population.2)
According to the Asian Working Group for Sarcopenia (AWGS) 2025 consensus, sarcopenia is defined as the concurrent presence of low muscle mass and low muscle strength.3) Sarcopenia has been increasingly recognized as a major health concern in older adults because of its association with a wide range of adverse outcomes. Previous studies have shown that older adults with sarcopenia are at increased risk of geriatric syndromes, including falls4) and cognitive impairment.5) In addition, sarcopenia contributes to functional decline,6) loss of independence, and poorer quality of life.7) Given these consequences, early identification of sarcopenia is important for supporting healthy aging and reducing the burden of age-related disability in older populations.
In Thailand, the diagnosis of sarcopenia has largely been based on the AWGS 2019 criteria in both clinical practice and research settings.8,9) Recently, the AWGS 2025 consensus introduced several updates, including revised cutoff values for individuals aged 50–64 years and the incorporation of appendicular skeletal muscle mass adjusted for body mass index (ASM/BMI) as an alternative measure of low muscle mass. However, data on ASM/BMI among Thai adults remain limited, and the applicability of these new criteria to the Thai population has not been well established.
It is not yet clear how the updated AWGS 2025 criteria will affect the diagnosis of sarcopenia in Thai older adults. The revised cutoff values and the introduction of ASM/BMI may change the number of individuals identified as having sarcopenia. Since information on these new parameters in the Thai population remains limited, evaluating their impact is important. Therefore, the primary objective was to compare the prevalence and classification of low muscle mass and sarcopenia under the AWGS 2019 and 2025 criteria. Secondary objectives were to characterize discordant ASM/height²- and ASM/BMI-based phenotypes and compare their associations with handgrip strength and gait speed. Exploratory analyses examined the age- and sex-specific distributions of the two muscle mass indices.
MATERIALS AND METHODS
Study Design
This cross-sectional study used data from an ongoing geriatric cohort of community-dwelling older adults aged 60 years and older who attended a geriatric outpatient clinic at a university-affiliated tertiary care hospital between March 2020 and July 2025. Eligible patients were consecutively screened and recruited during their in-person outpatient visits. During 2020–2021, the hospital reduced non-urgent outpatient services and in-person visits in accordance with COVID-19 infection-control and social-distancing measures. Consequently, annual recruitment decreased from approximately 2,200–2,800 participants during regular service periods to approximately 150–200 participants per year during 2020–2021. The cohort was prospectively registered with the Thai Clinical Trials Registry (TCTR20200511003). The study protocol was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University (IRB No. 718/62; approved on 26 March 2020), and written informed consent was obtained from all participants before enrollment. The study was conducted and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.10)
Eligible participants were community-dwelling adults aged 60 years and older who remained independent in basic activities of daily living. Exclusion criteria included conditions that could interfere with bioelectrical impedance analysis (BIA) measurements, such as metallic implants or implanted cardiac devices. Participants were also excluded if they had active malignancy, recent cancer treatment within 1 year, hospitalization for major cardiac, pulmonary, or cerebrovascular disease within the past three months, renal replacement therapy–dependent chronic kidney disease, known cirrhosis or psychiatric disorders requiring medication adjustment during the preceding 3 months.
Data Collection
Demographic characteristics, including age and sex, were recorded at the study visit. Fall history was assessed using a single question from the adapted Thai version of the STEADI fall-risk screening algorithm: “Have you fallen in the past 12 months?” An affirmative response indicated a history of at least one fall during this period.11,12) Frailty was assessed using the Fried Frailty Phenotype, which consists of five components: unintentional weight loss, exhaustion, low physical activity, slowness, and weakness.13) Unintentional weight loss was defined as a self-reported loss of at least 4.5 kg during the previous year. Exhaustion was assessed using two items from the Center for Epidemiologic Studies Depression Scale (CES-D),14) while physical activity was evaluated using the Thai version of the Global Physical Activity Questionnaire (Thai-GPAQ).15) Slowness was determined by gait speed, and weakness was defined using handgrip strength cutoff values based on the AWGS criteria. For comparisons between the AWGS 20198) and AWGS 20253) sarcopenia definitions, the corresponding handgrip strength thresholds from each guideline were applied. Participants were categorized as robust (0 criteria), pre-frail (1–2 criteria), or frail (≥3 criteria) according to the Fried Frailty Phenotype.
Anthropometric Measurements
Anthropometric measurements were performed by trained nurses using standardized procedures. Participants were assessed while wearing light clothing and no shoes. Body weight was measured to the nearest 0.1 kg using a calibrated digital scale, and standing height was measured to the nearest 0.1 cm using a stadiometer. Waist circumference was measured at the midpoint between the lower margin of the last palpable rib and the top of the iliac crest at the end of a normal expiration, in accordance with World Health Organization recommendations.16)
Body Composition Assessment
Body composition was evaluated using a multifrequency bioelectrical impedance analyzer (InBody 770; InBody Co., Seoul, South Korea). ASM, percent body fat (PBF), and visceral fat area (VFA) were obtained following a standardized protocol based on the European Society for Clinical Nutrition and Metabolism (ESPEN) recommendations. Participants were instructed to avoid food, alcohol, caffeine, smoking, and moderate-to-vigorous physical activity for at least 8 hours before testing. Assessments were performed after bladder emptying, with participants wearing light clothing and without metallic accessories. Measurements were deferred in the presence of fever or dehydration.17)
Physical Performance
Usual gait speed was evaluated using a 6-meter walk test following the AWGS 2019 protocol. Participants walked at their normal pace from a moving start, and timing was recorded over the 6-meter distance without a deceleration phase. The mean value from two trials was used to calculate gait speed and reported in meters per second (m/s).8)
Handgrip Strength
Handgrip strength was assessed using a digital dynamometer (T.K.K.5401; Takei Scientific Instruments Co., Ltd., Niigata, Japan). Participants performed the test in a standing position with the arm fully extended alongside the body. Two measurements were obtained from the dominant hand, and the higher value was used for analysis.8,18)
Sarcopenia Diagnosis and Category
After the assessment of handgrip strength, gait speed, and muscle mass using BIA, participants were classified according to the AWGS 2019 and 2025 criteria. Under the AWGS 2019 definition, sarcopenia was identified by the presence of low muscle mass together with either low muscle strength or poor physical performance.8) In contrast, the AWGS 2025 criteria define sarcopenia as the coexistence of low muscle mass and low muscle strength.3)
Statistical Analysis
All statistical analyses were performed using IBM SPSS Statistics for Windows, version 29.0.1.0 (IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean±standard deviation, while categorical variables are expressed as frequencies and percentages. Missing data were not imputed. All analyses were performed using a complete-case approach. Participants with incomplete BIA-derived body-composition data or missing handgrip-strength or gait-speed data were excluded from the analytical cohort. The analyses were organized according to the hierarchy of the study objectives.
Primary analyses summarized the prevalence and classification of low muscle mass and sarcopenia under the AWGS 2019 and 2025 criteria as frequencies and percentages. Participant characteristics were compared using independent-samples t-tests for continuous variables and chi-square tests for categorical variables. Secondary analyses classified participants into four mutually exclusive phenotypes: normal muscle mass, isolated low ASM/BMI, isolated low ASM/height², and low values for both indices. Differences across these groups were examined using one-way analysis of variance or chi-square tests, as appropriate.
Exploratory receiver operating characteristic (ROC) analyses assessed the classification correspondence of PBF, VFA, and waist-to-height ratio (WHtR) with the isolated low ASM/BMI phenotype. Thresholds were selected using the Youden index, with the area under the curve (AUC), sensitivity, specificity, positive likelihood ratio (LR+), and negative likelihood ratio (LR−) reported with 95% confidence intervals (CIs). Separate age- and sex-adjusted linear regression models examined the associations of each muscle mass index with handgrip strength and gait speed; standardized coefficients and adjusted R² values were reported. Effect sizes included Cohen’s d, Cramér’s V, and eta-squared (η²), as appropriate. Post hoc analyses calculated age- and sex-specific P5, P20, P50, P80, and P95 values for both indices and handgrip strength. All tests were two-sided, with p<0.05 indicating statistical significance.
RESULTS
Participant selection is summarized in Fig. 1. Among 9,973 individuals screened, 821 were excluded based on the eligibility criteria, leaving 9,152 eligible participants. Of these, 335 declined participation and 8,817 provided consent and were enrolled. A further 744 participants were excluded because of incomplete body composition data (n=590) or missing handgrip strength or gait speed (n=154), resulting in 8,073 participants included in the final cross-sectional analysis. Participants excluded because of missing measurements were slightly older than those included in the analysis (70.62±5.75 vs. 69.27±5.53 years), whereas the proportion of women was similar (78.9% vs. 76.3%) (Supplementary Table S1).
In the primary analysis (Table 1), the prevalence of low muscle mass increased from 43.2% (n=3,490) under AWGS 2019 to 44.6% (n=3,604) after applying the revised age-specific AWGS 2025 ASM/height² cutoffs and to 51.4% (n=4,151) after including ASM/BMI, which identified 547 additional participants. Sarcopenia prevalence similarly increased from 15.1% (n=1,215) under AWGS 2019 to 17.5% (n=1,414) using the revised ASM/height² cutoffs and to 19.7% (n=1,587) under the complete AWGS 2025 criteria. The prevalence of possible sarcopenia based on low handgrip strength increased from 25.0% (n=2,016) to 29.5% (n=2,381), representing 365 additional participants. Under both definitions, participants with sarcopenia were older, were more frequently prefrail or frail, and had slower gait speed than those without sarcopenia. Fall history did not differ significantly under AWGS 2025 (14.4% vs. 13.0%, p=0.132), whereas the difference under AWGS 2019 was statistically significant but negligible in magnitude (15.1% vs. 13.0%, p=0.047; Cramér’s V=0.02).
To characterize the participants additionally identified after incorporating ASM/BMI, clinical characteristics were compared across the four muscle mass phenotypes (Table 2). In both sexes, participants with isolated low ASM/BMI had higher PBF, VFA, and WHtR than those with isolated low ASM/height². Participants with low values for both indices also exhibited greater adiposity than those with isolated low ASM/height². The between-phenotype effects were large for PBF and VFA (η²=0.17–0.27) but small for WHtR (η²=0.04).
Exploratory ROC analyses were subsequently performed to examine the classification correspondence between obesity-related indicators and isolated low ASM/BMI (Table 3). PBF showed the highest classification correspondence in both women and men. The cohort-derived thresholds maximizing the Youden index were PBF ≥40.85% in women and ≥30.3% in men, VFA ≥129.5 cm² in women and ≥92.5 cm² in men, and WHtR ≥0.56 in both sexes.
In the age- and sex-adjusted regression analyses (Table 4), ASM/height² was positively associated with handgrip strength (B=2.019; 95% CI, 1.828–2.211; standardized β=0.262; p<0.001), whereas ASM/BMI was not associated with handgrip strength (B=0.149; 95% CI, −0.797–1.096; standardized β=0.004; p=0.757). In contrast, ASM/BMI was positively associated with gait speed (B=0.252; 95% CI, 0.196–0.308; standardized β=0.122; p<0.001), while ASM/height² showed a statistically significant but weak inverse association with gait speed (B=−0.013; 95% CI, −0.025 to −0.001; standardized β=−0.035; p=0.027). The ASM/height² and ASM/BMI models explained 39.3% and 36.1% of the variance in handgrip strength and 4.2% and 5.1% of the variance in gait speed, respectively.
Table 5 summarizes the age- and sex-specific distributions of ASM/height², ASM/BMI, and handgrip strength in this cohort of 8,073 Thai community-dwelling older adults. Mean ASM/height² and handgrip strength were lower in older age groups in both sexes. Among women aged ≥80 years, mean ASM/height² was 5.34±0.68 kg/m², compared with 5.80±0.67 kg/m² among those aged 60–69 years; the corresponding handgrip strength values were 16.85±5.17 and 20.53±5.04 kg. Among men, mean ASM/height² was 6.74±0.74 versus 7.52±0.69 kg/m², and mean handgrip strength was 23.58±9.00 versus 31.71±7.36 kg in the respective age groups. In contrast, ASM/BMI differed little across age groups in either sex. Across older age groups, raw ASM, body weight, and BMI were lower in both sexes, whereas PBF and VFA were higher (Supplementary Table S2).
DISCUSSION
This study provides an early evaluation of how implementation of the AWGS 2025 criteria may alter the classification of low muscle mass and sarcopenia in older adults attending a tertiary geriatric outpatient service. The observed changes did not arise from a single modification. Rather, they reflected the combined effects of the revised age-specific criteria, the proportion of participants meeting the handgrip-strength criterion, and the inclusion of ASM/BMI as an alternative muscle mass index. A larger numerical increase in low muscle mass classification was observed after incorporating ASM/BMI, whereas the increase in sarcopenia classification was observed after applying changes to both the muscle-mass and muscle-strength criteria.3,8)
The prevalence of sarcopenia under AWGS 2025 was broadly consistent with previous Thai estimates of 17.3%–18.1%.2,9) A recent nationwide Thai study similarly found that AWGS 2025 identified more individuals, particularly those who were younger and had a higher BMI, and reported higher fall risk among individuals newly classified under AWGS 2025.19) In our cohort, however, fall history was self-reported, was the only clinical outcome examined for this purpose, and showed little or no meaningful difference. Thus, the present study demonstrates a change in classification but cannot determine whether the additional cases are at greater risk of clinically important outcomes. Longitudinal data are needed to establish the clinical value of identifying these additional cases. As the present cohort was recruited from a tertiary geriatric clinic, these findings support the feasibility of applying AWGS 2025 in this setting but should not be considered nationally representative.
The discordance between the two indices appeared to be largely related to body composition. Isolated low ASM/BMI represented a phenotype with greater overall and central adiposity than isolated low ASM/height², suggesting that BMI-adjusted muscle mass may provide complementary information in individuals with obesity. This interpretation is consistent with the 2023 Korean Working Group on Sarcopenia consensus and findings from the FNIH Sarcopenia Project supporting BMI-adjusted muscle mass indices in this population.20-22) The thresholds identified in the exploratory ROC analyses should nevertheless be interpreted cautiously. Although PBF, VFA, and WHtR are not direct components of ASM/BMI, their correlations with BMI may partly explain their classification correspondence with isolated low ASM/BMI. Moreover, all thresholds were derived from the same cohort. These values should not be considered validated cutoffs for sarcopenic obesity without external validation.23)
The associations of the two muscle mass indices were outcome-specific. After adjustment for age and sex, ASM/height² was positively associated with handgrip strength, whereas ASM/BMI was positively associated with gait speed. ASM/height² represents muscle quantity normalized for stature, while handgrip strength is an absolute force measure influenced by body size.24) In contrast, ASM/BMI may reflect muscle reserve relative to the body load involved in walking; preferred gait speed has been shown to decrease with increasing BMI, accompanied by altered gait mechanics.25) Similarly, studies in Japanese and Turkish older outpatients found ASM/BMI more closely associated with low physical function or gait speed than ASM/height², while considering both indices improved the identification of fall risk.26-28) In our cohort, the limited variation in ASM/BMI across age groups may have attenuated its association with handgrip strength. However, because the models adjusted only for age and sex, the gait-speed association may partly reflect the BMI denominator and residual confounding by adiposity or other mobility determinants. Moreover, handgrip strength is part of the sarcopenia diagnostic framework,3) and the gait-speed models explained limited outcome variance. Therefore, these cross-sectional associations do not provide evidence for choosing one muscle-mass index over the other. Longitudinal studies using independent outcomes, including chair-stand performance, Short Physical Performance Battery (SPPB), disability, hospitalization, and mortality, are needed to clarify their relative clinical relevance.
Although ASM/height² and handgrip strength were lower in older age groups, ASM/BMI differed little across age groups. This pattern differs somewhat from the lower ASM/BMI values observed in older age groups in a recent multinational Asian study29) and from the corresponding age-group differences in ALM/BMI among nondisabled community-dwelling older Japanese adults.30) In the present cohort, raw ASM, body weight, and BMI were lower in older age groups in both sexes, whereas PBF and VFA were higher (Supplementary Table S2). Because BMI is the denominator of ASM/BMI, the lower BMI observed in older age groups may partly offset their lower ASM, thereby attenuating differences in the ratio. This pattern is consistent with a previous Thai study reporting lower mean BMI in older age groups after age 60,31) while Chittawatanarat et al.32) reported higher percentage fat mass but lower fat-free mass in older age groups. Differences from broader Asian data may also reflect the predominance of participants aged 60–79 years, the relatively small proportion aged ≥80 years, and the selected nature of the cohort, as suggested by its relatively low frailty prevalence. These characteristics should be considered when interpreting the reported muscle-mass distributions.
Based on the observed clinical profiles of adults aged ≥60 years, we proposed a conceptual framework illustrating when ASM/BMI may provide complementary information to ASM/height² (Table 6).
The principal strengths of this study include its large sample size, standardized body-composition and physical-performance assessments, direct application of both AWGS definitions within the same cohort, and stepwise analyses that separated the contributions of the revised ASM/height² criteria and the addition of ASM/BMI.
Several limitations should be considered. The cross-sectional design precludes causal and prognostic inference. Five-time chair-stand performance and the SPPB were not collected; therefore, participants with impairment detectable by these tests may have been missed, potentially underestimating the prevalence of sarcopenia under AWGS 2019 and affecting comparisons between the two criteria. Participants were recruited from a single urban tertiary geriatric outpatient service rather than through population-based sampling. Eligibility required independence in basic activities of daily living, and several serious conditions were excluded. Thus, the cohort was relatively healthy and functionally selected and may underrepresent frail or disabled older adults. These selection factors, together with the predominance of women, limit the generalizability of the reported muscle-mass distributions and percentile values to the overall Thai older population. Recruitment was also markedly reduced during the COVID-19 restrictions in 2020–2021, potentially introducing temporal selection bias. In addition, the complete-case analysis excluded participants with incomplete body-composition, handgrip-strength, or other required diagnostic data; selection bias may have occurred if missingness was related to health status.
In conclusion, implementation of AWGS 2025 broadened the classification of low muscle mass and sarcopenia in this tertiary geriatric outpatient cohort through both revised age-specific cutoffs and the inclusion of ASM/BMI. ASM/BMI provided complementary information to ASM/height² by identifying a distinct phenotype characterized by greater adiposity. However, the associations of the two indices with handgrip strength and gait speed were outcome-specific, and neither index was consistently superior.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
agmr-26-0128-Supplementary-Table-S1.pdf
