Abstract
Objectives
Sarcopenia, an age-related skeletal muscle disorder, is assessed by handgrip strength, gait speed, and muscle mass, yet population-specific reference data and standardized scores for Asian adults remain limited. We aimed to establish sex- and age-specific normative reference distributions, smoothed centile curves, and model-based standardized T scores for these indicators in Asian adults.
Methods
We retrospectively analyzed adults aged ≥ 40 years attending a tertiary-center Adult Preventive Health Program in 2023–2024. To assess construct validity, we compared sarcopenia indicators across age groups using analysis of variance (ANOVA). Sex- and age-specific reference distributions were estimated using Generalized Additive Models for Location, Scale, and Shape (GAMLSS), from which smooth model-based age-specific centiles (5th–95th) were derived. Individual values were standardized to model-based Z scores and transformed to T scores on an approximate 100-point scale. Asian Working Group for Sarcopenia (AWGS) cut-offs were used in receiver operating characteristic (ROC) analyses to determine corresponding T-score thresholds.
Results
Among 8095 participants, model-based mean handgrip strength, gait speed, and skeletal muscle mass were 20.03±5 kg, 1.12±0.37 m/s, and 6.93±1.43 kg/m² in women and 30.69±7.46 kg, 1.12±0.37 m/s, and 8.79±1.95 kg/m² in men. GAMLSS-derived centile curves demonstrated age-related declines in strength and gait speed and milder changes in muscle mass. AWGS cut-offs corresponded to T scores in the mid-40s for strength and gait and around 40–41 for muscle mass.
Conclusions
We established sex- and age-specific normative values and GAMLSS-based reference distributions, and derived model-based Z and T scores that provide standardized sarcopenia metrics in Asian adults, align with AWGS thresholds, and support detection of muscle decline and risk stratification.
Keywords: Muscle, Muscle strength, Reference standards, Sarcopenia, Walking speed
1. Introduction
Sarcopenia, a major geriatric syndrome, is characterized by a progressive decline in skeletal muscle mass and strength associated with aging [1]. The term was first introduced by Rosenberg in 1989 to describe age-related changes in body composition and associated functional decline [2]. Over the past decades, various definitions and diagnostic criteria have emerged based on different ethnic populations and methodological approaches. In 2010, the European Working Group on Sarcopenia in Older People (EWGSOP) proposed a widely accepted operational definition based on the assessment of muscle mass, muscle strength, and physical performance, which was subsequently updated in 2018 [3,4]. Given the marked differences in body size and composition between Asian and Western populations, the Asian Working Group for Sarcopenia (AWGS) was established in 2014 to develop region-specific diagnostic criteria and cut-off values for sarcopenia [5].
In 2019, the AWGS updated its consensus and recommended using questionnaire-based tools for sarcopenia risk screening in primary care and community settings. The SARC-F questionnaire comprises five items: strength, assistance with walking, rising from a chair, climbing stairs, and falls. A modified version, SARC-CalF, incorporates calf circumference to enhance screening sensitivity. Individuals with positive sarcopenia screening results should undergo confirmatory evaluation. The AWGS recommends diagnostic evaluation based on handgrip strength for muscle strength, gait speed, five-times chair stand test, or Short Physical Performance Battery (SPPB) for physical performance, and muscle mass assessed by bioelectrical impedance analysis (BIA) or dual-energy X-ray absorptiometry (DXA) [6].
Among the assessment methods for sarcopenia, handgrip strength, gait speed, and appendicular skeletal muscle mass (ASM) measured by BIA are frequently used in healthcare settings due to their validity, feasibility, and non-invasiveness [4,6,7]. However, the direct measurement values require standardization to enhance clinical interpretability and enable comparisons across individuals and populations. Standardized metrics such as Z scores, T scores, and percentile ranks are essential for contextualizing individual results within reference distributions [8]. Although several international studies have reported normative values based on means and standard deviations [[9], [10], [11], [12], [13], [14], [15]], few have converted them into standardized scoring systems to support clinical and public health applications. As muscle mass, strength, and gait speed progressively decline with age [16], establishing age- and sex-specific normative references based on representative samples, with construct validity supported by age-related trends, is critical for early detection, risk stratification, and clinical decision-making [17].
This study aimed to establish normative reference values for sarcopenia-related indicators, including handgrip strength, gait speed, and appendicular skeletal muscle mass, as recommended by the AWGS. These measures reflect the three core domains of sarcopenia, muscle strength, physical performance, and muscle mass, thereby ensuring construct validity of the reference framework. Using data from nearly 10,000 adults aged 40 years and older who underwent standardized assessments at a tertiary medical center in northern Taiwan over a two-year period, normative values were stratified by sex and 5-year age groups. Analyses included mean differences, percentile-based reference ranges, and standardized Z and T scores. T scores were linearly transformed from Z scores to approximate a 100-point scale, enhancing interpretability across variables with different units. These normative data provide a robust reference to support early risk identification and guide clinical and preventive strategies for sarcopenia management in Asian populations.
2. Methods
2.1. Study population and data collection
This retrospective study utilized data from the Healthcare Information System of Far Eastern Memorial Hospital, a tertiary medical center located in New Taipei City, Taiwan. The hospital primarily serves an urban–suburban catchment area in northern Taiwan, encompassing New Taipei City and Taipei City. Consequently, the study cohort mainly reflects community-dwelling, relatively health-conscious middle-aged and older adults residing in these urban and suburban areas who voluntarily attended preventive health check-ups. Data were collected from adults aged 40 years and older who participated in the National Adult Preventive Health Program and completed sarcopenia-related assessments, including handgrip strength, gait speed, and skeletal muscle mass measurements, between January 2023 and December 2024. Individuals were excluded if they were hospitalized, institutionalized, functionally impaired in daily living, or unable to follow testing instructions.
Among the 27,048 adult health checkup records collected during the two-year period, 9919 individuals completed sarcopenia-related assessments. Records that lacked data on handgrip strength, gait speed, or skeletal muscle mass, or contained implausible values (including extreme outliers or data entry errors), were excluded from analysis. To ensure consistency, only the first assessment record was retained for participants who received repeated examinations across years. A total of 8095 participants were included in the final analytic sample for normative reference construction and standardized scoring. The participant selection flowchart is illustrated in Fig. 1.
Fig. 1.
Flowchart of data processing and participant selection.
2.2. Sarcopenia measurements and cutoff points
Sarcopenia measurements were assessed according to the 2019 diagnostic criteria proposed by the AWGS [6], covering three domains: muscle strength, physical performance, and skeletal muscle mass. All measurements were conducted by trained medical technicians and recorded through the physician order entry system.
Handgrip strength was measured using a spring-type digital hand dynamometer (Camry EH101, Taichung, Taiwan) with participants standing and the dominant elbow fully extended; three trials were performed, and the highest value was used for analysis. Low muscle strength was defined as < 18 kg in women and < 28 kg in men [6,18]. Gait speed was measured over a 6-m walkway with 2-m acceleration and deceleration zones, and calculated as distance divided by time (m/s); values < 1.0 m/s were classified as low physical performance [6,7].
Skeletal muscle mass was estimated using BIA (Tanita BC-418, Tokyo, Japan) with participants in light clothing and bare feet. ASM (kg) was divided by height squared (m2) to derive the ASM index (kg/m2), with low skeletal muscle mass defined as < 5.7 kg/m2 in women and < 7.0 kg/m2 in men [6].
2.3. Statistical analysis
Means and standard deviations of handgrip strength, gait speed, and skeletal muscle mass were calculated by sex across 5-year age intervals, and age-related differences were examined using analysis of variance (ANOVA). To assess construct validity, we examined whether these measures demonstrated progressive age-related declines consistent with international sarcopenia definitions [1], [19].
Reference distributions for these measures were then estimated using generalized additive models for location, scale, and shape (GAMLSS) [20], [21], fitted separately for sex. Age was modelled as a smooth function to accommodate non-linear trends and age-dependent variability. Smoothed centile curves and centiles from the 5th to the 95th percentile were derived from the models.
To facilitate comparison across indicators measured on different scales, individual values were standardized to model-based Z scores using formula Z = (X − μ)/σ, where X is the observed value, and μ and σ are the corresponding age- and sex-specific means and standard deviations derived from the GAMLSS reference distributions. Z scores were linearly transformed to T scores using T = 10Z + 50, resulting in a mean of 50; values from 20 to 80 correspond to ±3 standard deviations, encompassing approximately 99.7% of the reference population. This transformation approximates a 100-point scale and provides an intuitive framework for interpreting performance across sarcopenia domains [8].
For additional clinical interpretability, we conducted a secondary analysis to derive optimal T-score cut-off points for low handgrip strength, slow gait speed, and low skeletal muscle mass. Receiver operating characteristic (ROC) curves were constructed using the AWGS diagnostic cut-off points as the reference standard. Youden's index was used to identify the optimal T-score thresholds for these indicators in women and men, and was calculated as sensitivity + specificity −1, ranging from 0 to 1, with higher values indicating better combined sensitivity and specificity [22].
Stratum sizes of about 200 participants in each sex-specific 5-year age band are generally considered sufficient for reasonably precise percentile estimation; in this study, the overall sample of 8095 adults yielded at least 200 participants in each such band, supporting stable estimation of key centiles across age groups. Descriptive statistics, ANOVA, empirical summaries, and initial standardized score calculations were performed using SPSS version 22.0 (IBM Corp., Armonk, NY, USA), whereas GAMLSS models, model-based centiles, and Z and T scores were obtained in R version 4.5.2 (R Foundation for Statistical Computing, Vienna, Austria) using the GAMLSS package. Two-sided P-values < 0.05 were considered statistically significant.
2.4. Ethics statement
This study was approved by the Institutional Review Board of Far Eastern Memorial Hospital, New Taipei City, Taiwan (114127-E).
3. Results
3.1. Participant characteristics
A total of 8095 individuals were included in the study, consisting of 4477 women (55%) and 3618 men (45%). The mean age was 64.2 ± 11 years. When stratified into 5-year age intervals, the age distribution was as follows: 40–44 years (5.8%), 45–49 (7.7%), 50–54 (8.3%), 55–59 (8.6%), 60–64 (10.9%), 65–69 (24.8%), 70–74 (19.0%), 75–79 (8.3%), and ≥ 80 years (6.5%). Notably, the largest proportion of participants (43.8%) were aged between 65 and 74 years.
3.2. Normative references and construct validity for sarcopenia measurements
Sex- and age-specific means and standard deviations were calculated for handgrip strength, gait speed, and skeletal muscle mass, providing normative references (Table 1). In women, mean handgrip strength was 20.03 ± 5 kg. A relatively stable pattern was observed below age 50, followed by a decline ranging from 0.3 to 2.1 kg across increasing age groups. In men, mean handgrip strength was 30.69 ± 7.46 kg and declined by 1.2–3.2 kg across increasing age groups after age 50. ANOVA indicated significant age-group differences in both sexes. The mean gait speed was 1.12 ± 0.37 m/sec in both sexes, with a significant reduction observed with advancing age. For skeletal muscle mass measured by BIA, the mean among women was 6.93 ± 1.43 kg/m2, showing minimal change across age groups with no statistically significant differences. In contrast, the mean for men was 8.79 ± 1.95 kg/m2, with significantly lower values noted in individuals aged 70 and older compared to those younger than 55 years. Together, these age-associated trends support the construct validity of the normative references.
Table 1.
Sex- and age-stratified normative values for handgrip strength, gait speed, and skeletal muscle mass in middle-aged and older adults.
| Total |
Middle age |
Elderly |
40–44 yr |
45–49 yr |
50–54 yr |
55–59 yr |
60–64 yr |
65–69 yr |
70–74 yr |
75–79 yr |
≥ 80 yr |
||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | Mean ± SD (Median) | P | |
| Total number | 8095 | 3343 | 4752 | 470 | 620 | 671 | 699 | 883 | 2008 | 1542 | 674 | 528 | |
| Female | 4477 | 1972 | 2505 | 263 | 378 | 416 | 415 | 500 | 1122 | 801 | 324 | 258 | |
| Male | 3618 | 1371 | 2247 | 207 | 242 | 255 | 284 | 383 | 886 | 741 | 350 | 270 | |
| Handgrip strength (kg) | |||||||||||||
| Female | 20.03 ± 5 (19.8) | 22 ± 4.8 (21.9) | 18.49 ± 4.59 (18.6) | 23.35 ± 4.96 (23.2) | 23.79 ± 4.6 (23.7) | 22.38 ± 4.59 (22.2) | 21.32 ± 4.88 (21.4) | 20.18 ± 4.21 (20) | 19.85 ± 4.32 (19.7) | 18.31 ± 4.16 (18.5) | 17.01 ± 4.3 (17.3) | 14.96 ± 4.8 (14.85) | <0.001 |
| Male | 30.69 ± 7.46 (30.5) | 34.34 ± 7.45 (34.34) | 28.46 ± 6.53 (28.5) | 36.78 ± 8.04 (36) | 36.44 ± 7.22 (35.85) | 35.07 ± 7.55 (35) | 33.73 ± 6.82 (33.9) | 31.67 ± 6.72 (32) | 30.46 ± 6.25 (30.6) | 28.76 ± 6.03 (28.6) | 26.65 ± 6.12 (26.5) | 23.43 ± 5.99 (23.1) | <0.001 |
| Gait speed (m/sec) | |||||||||||||
| Female | 1.12 ± 0.37 (1.09) | 1.21 ± 0.36 (1.14) | 1.05 ± 0.37 (1) | 1.22 ± 0.31 (1.18) | 1.24 ± 0.37 (1.2) | 1.23 ± 0.36 (1.19) | 1.2 ± 0.31 (1.14) | 1.18 ± 0.39 (1.1) | 1.13 ± 0.35 (1.09) | 1.06 ± 0.36 (1) | 0.94 ± 0.37 (0.92) | 0.78 ± 0.3 (0.75) | <0.001 |
| Male | 1.12 ± 0.37 (1.08) | 1.23 ± 0.37 (1.17) | 1.06 ± 0.35 (1) | 1.26 ± 0.36 (1.2) | 1.29 ± 0.38 (1.2) | 1.24 ± 0.38 (1.2) | 1.22 ± 0.36 (1.17) | 1.17 ± 0.38 (1.1) | 1.12 ± 0.35 (1.07) | 1.07 ± 0.33 (1.01) | 1 ± 0.35 (1) | 0.85 ± 0.35 (0.83) | <0.001 |
| Skeletal muscle mass (kg/m2) | |||||||||||||
| Female | 6.93 ± 1.43 (6.69) | 6.93 ± 1.76 (6.67) | 6.94 ± 1.11 (6.71) | 7 ± 1.18 (6.76) | 6.95 ± 1.05 (6.73) | 7.09 ± 3.28 (6.65) | 6.76 ± 0.9 (6.6) | 6.87 ± 1 (6.71) | 6.91 ± 1.02 (6.67) | 6.96 ± 1.3 (6.71) | 6.94 ± 1.02 (6.71) | 6.97 ± 0.92 (6.86) | 0.1 |
| Male | 8.79 ± 1.95 (8.62) | 9.1 ± 1.89 (8.89) | 8.61 ± 1.96 (8.42) | 9.25 ± 1.29 (9.16) | 9.31 ± 2.52 (9.03) | 9.3 ± 2.6 (9) | 9.05 ± 1.2 (8.85) | 8.8 ± 1.46 (8.64) | 8.77 ± 2.39 (8.54) | 8.5 ± 1.33 (8.37) | 8.55 ± 2.22 (8.23) | 8.41 ± 1.38 (8.17) | <0.001 |
Values are presented as mean ± standard deviation (SD) with median in parentheses. Handgrip strength is expressed in kilograms (kg), gait speed in meters per second (m/s), and skeletal muscle mass as skeletal muscle mass index (kg/m2). P values were obtained from ANOVA across five-year age groups.
SD, standard deviation.
3.3. Model-based centile distributions for sarcopenia indicators
Under the sex- and age-specific GAMLSS reference distributions, the overall predicted mean handgrip strength was 20.03 ± 5 kg in women and 30.69 ± 7.46 kg in men; mean gait speed was 1.12 ± 0.37 m/s in both sexes; mean skeletal muscle mass was 6.93 ± 1.43 kg/m² in women and 8.79 ± 1.95 kg/m² in men. Across strata, model-based mean values for the sarcopenia indicators were located close to the 50th percentile. Using the AWGS diagnostic thresholds, low handgrip strength corresponded approximately to the 5th–20th percentiles in middle-aged adults and shifted toward the 50th percentile in those aged 80 years and older. For gait speed, the percentile position of the diagnostic cutoff similarly increased from about the 5th–20th percentiles in younger adults to around the 50th percentile in the oldest group. In contrast, among men, skeletal muscle mass at the 5th percentile remained above the diagnostic cutoff throughout middle age, whereas in the oldest age group, individuals at the 5th percentile met the criteria for low muscle mass. Detailed percentile distributions for all indicators are presented in Table 2. Smoothed age-specific centile curves derived from the GAMLSS models for handgrip strength, gait speed, and skeletal muscle mass in women and men are shown in Fig. 2. These curves demonstrate the expected age-related decline in handgrip strength and gait speed, with more modest changes in skeletal muscle mass, indicating heterogeneous aging trajectories across sarcopenia indicators.
Table 2.
Model-based centiles for handgrip strength, gait speed, and skeletal muscle mass by sex and age group.
| Female |
Male |
|||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 5th | 15th | 20th | 50th | 75th | 85th | 95th | 5th | 15th | 20th | 50th | 75th | 85th | 95th | |
| Handgrip strength (kg) | ||||||||||||||
| Total | 12 | 15.2 | 16.1 | 19.8 | 23.02 | 24.9 | 28.5 | 18.7 | 23.4 | 24.9 | 30.5 | 35.3 | 38 | 43.4 |
| Middle age | 14.5 | 17.4 | 18.2 | 21.9 | 25 | 26.6 | 30.1 | 23.1 | 27.2 | 28.4 | 34 | 39 | 41.9 | 46.7 |
| Elderly | 10.8 | 13.9 | 14.96 | 18.6 | 21.2 | 22.9 | 25.78 | 17.6 | 21.8 | 23.2 | 28.5 | 32.8 | 35 | 38.9 |
| 40–44 yr | 16 | 18.68 | 19.56 | 23.51 | 26.84 | 28.67 | 31.84 | 25.16 | 29.04 | 30.36 | 36.49 | 41.92 | 45.04 | 50.61 |
| 45–49 yr | 15.71 | 18.36 | 19.24 | 23.16 | 26.44 | 28.26 | 31.39 | 24.38 | 28.45 | 29.8 | 35.88 | 41.03 | 43.89 | 48.86 |
| 50–54 yr | 14.97 | 17.56 | 18.41 | 22.22 | 25.41 | 27.18 | 30.22 | 23.32 | 27.58 | 28.95 | 34.95 | 39.81 | 42.44 | 46.89 |
| 55–59 yr | 14.08 | 16.57 | 17.4 | 21.07 | 24.15 | 25.85 | 28.79 | 22.06 | 26.43 | 27.8 | 33.63 | 38.19 | 40.6 | 44.6 |
| 60–64 yr | 13.44 | 15.87 | 16.67 | 20.25 | 23.25 | 24.91 | 27.76 | 20.78 | 25.11 | 26.45 | 32.04 | 36.33 | 38.58 | 42.27 |
| 65–69 yr | 12.77 | 15.17 | 15.97 | 19.51 | 22.47 | 24.11 | 26.92 | 19.6 | 23.75 | 25.03 | 30.39 | 34.5 | 36.64 | 40.18 |
| 70–74 yr | 11.47 | 13.84 | 14.63 | 18.12 | 21.06 | 22.68 | 25.47 | 18.32 | 22.18 | 23.4 | 28.53 | 32.54 | 34.65 | 38.16 |
| 75–79 yr | 9.91 | 12.28 | 13.07 | 16.59 | 19.55 | 21.19 | 24.01 | 16.69 | 20.32 | 21.47 | 26.44 | 30.39 | 32.5 | 36.04 |
| ≥ 80 yr | 8.35 | 10.73 | 11.52 | 15.08 | 18.08 | 19.74 | 22.62 | 14.71 | 18.22 | 19.34 | 24.2 | 28.08 | 30.16 | 33.66 |
| Gait speed (m/sec) | ||||||||||||||
| Total | 0.58 | 0.8 | 0.85 | 1.09 | 1.27 | 1.5 | 1.82 | 0.6 | 0.79 | 0.85 | 1.08 | 1.24 | 1.5 | 1.84 |
| Middle age | 0.75 | 0.92 | 1 | 1.13 | 1.38 | 1.51 | 1.9 | 0.75 | 0.93 | 1 | 1.17 | 1.4 | 1.54 | 1.96 |
| Elderly | 0.5 | 0.72 | 0.76 | 1 | 1.2 | 1.41 | 1.71 | 0.55 | 0.75 | 0.8 | 1 | 1.2 | 1.36 | 1.67 |
| 40–44 yr | 0.77 | 0.91 | 0.95 | 1.18 | 1.4 | 1.53 | 1.78 | 0.8 | 0.93 | 0.98 | 1.22 | 1.47 | 1.63 | 1.95 |
| 45–49 yr | 0.76 | 0.9 | 0.95 | 1.2 | 1.43 | 1.57 | 1.83 | 0.78 | 0.92 | 0.97 | 1.22 | 1.47 | 1.63 | 1.95 |
| 50–54 yr | 0.73 | 0.88 | 0.94 | 1.19 | 1.44 | 1.58 | 1.85 | 0.76 | 0.89 | 0.94 | 1.19 | 1.45 | 1.61 | 1.92 |
| 55–59 yr | 0.7 | 0.86 | 0.91 | 1.18 | 1.43 | 1.58 | 1.85 | 0.72 | 0.86 | 0.91 | 1.16 | 1.41 | 1.56 | 1.86 |
| 60–64 yr | 0.65 | 0.81 | 0.87 | 1.14 | 1.4 | 1.55 | 1.83 | 0.68 | 0.82 | 0.87 | 1.12 | 1.36 | 1.51 | 1.8 |
| 65–69 yr | 0.6 | 0.76 | 0.82 | 1.09 | 1.35 | 1.5 | 1.77 | 0.64 | 0.78 | 0.83 | 1.08 | 1.32 | 1.46 | 1.74 |
| 70–74 yr | 0.52 | 0.69 | 0.74 | 1.01 | 1.26 | 1.4 | 1.67 | 0.58 | 0.73 | 0.78 | 1.03 | 1.26 | 1.4 | 1.67 |
| 75–79 yr | 0.44 | 0.6 | 0.65 | 0.91 | 1.14 | 1.27 | 1.51 | 0.5 | 0.64 | 0.69 | 0.94 | 1.18 | 1.32 | 1.58 |
| ≥ 80 yr | 0.36 | 0.5 | 0.55 | 0.79 | 1 | 1.12 | 1.34 | 0.4 | 0.55 | 0.6 | 0.85 | 1.08 | 1.22 | 1.47 |
| Skeletal muscle mass (kg/m2) | ||||||||||||||
| Total | 5.66 | 6 | 6.11 | 6.69 | 7.4 | 7.97 | 8.8 | 6.8 | 7.4 | 7.6 | 8.61 | 9.58 | 10.06 | 11.09 |
| Middle age | 5.65 | 5.97 | 6.07 | 6.66 | 7.33 | 7.85 | 8.73 | 7.26 | 7.77 | 7.92 | 8.89 | 9.81 | 10.31 | 11.39 |
| Elderly | 5.66 | 6.01 | 6.15 | 6.71 | 7.45 | 8.05 | 8.84 | 6.64 | 7.21 | 7.42 | 8.41 | 9.43 | 9.91 | 10.92 |
| 40–44 yr | 5.63 | 5.97 | 6.09 | 6.74 | 7.44 | 7.92 | 9 | 7.36 | 7.98 | 8.19 | 9.17 | 10.05 | 10.56 | 11.47 |
| 45–49 yr | 5.62 | 5.96 | 6.08 | 6.71 | 7.37 | 7.81 | 8.77 | 7.38 | 7.93 | 8.12 | 9.05 | 9.94 | 10.47 | 11.5 |
| 50–54 yr | 5.61 | 5.94 | 6.06 | 6.67 | 7.31 | 7.73 | 8.62 | 7.34 | 7.85 | 8.03 | 8.93 | 9.81 | 10.37 | 11.47 |
| 55–59 yr | 5.6 | 5.94 | 6.06 | 6.67 | 7.3 | 7.71 | 8.57 | 7.21 | 7.73 | 7.91 | 8.81 | 9.68 | 10.23 | 11.3 |
| 60–64 yr | 5.59 | 5.95 | 6.07 | 6.7 | 7.34 | 7.76 | 8.62 | 7 | 7.56 | 7.75 | 8.69 | 9.56 | 10.09 | 11.08 |
| 65–69 yr | 5.6 | 5.96 | 6.09 | 6.74 | 7.4 | 7.83 | 8.7 | 6.74 | 7.36 | 7.58 | 8.56 | 9.45 | 9.96 | 10.89 |
| 70–74 yr | 5.61 | 5.98 | 6.12 | 6.78 | 7.46 | 7.89 | 8.77 | 6.54 | 7.19 | 7.41 | 8.44 | 9.36 | 9.88 | 10.82 |
| 75–79 yr | 5.62 | 6 | 6.14 | 6.82 | 7.5 | 7.93 | 8.8 | 6.43 | 7.07 | 7.29 | 8.32 | 9.27 | 9.83 | 10.85 |
| ≥ 80 yr | 5.63 | 6.02 | 6.16 | 6.85 | 7.52 | 7.95 | 8.79 | 6.36 | 6.97 | 7.18 | 8.2 | 9.17 | 9.76 | 10.88 |
Values are presented as model-based centile cutoffs (5th, 15th, 20th, 50th, 75th, 85th, and 95th) derived from sex- and age-specific GAMLSS reference distributions. Handgrip strength is expressed in kilograms (kg), gait speed in meters per second (m/s), and skeletal muscle mass as skeletal muscle mass index (kg/m2). Values meeting the AWGS 2019 diagnostic cutoff criteria are highlighted with a bold italic characters.
Fig. 2.
Age-specific centile curves for handgrip strength, gait speed, and appendicular skeletal muscle mass
Smoothed age-specific centile curves for (A, D) handgrip strength, (B, E) gait speed, and (C, F) appendicular skeletal muscle mass index (ASM/height2) in women (A–C) and men (D–F), derived using Generalized Additive Models for Location, Scale, and Shape (GAMLSS). Lines represent selected centiles (5th, 15th, 20th, 50th, 75th, 85th, and 95th) across the age range from 40 years onward.
3.4. Model-based standardized scores (Z and T scores)
Standardized scores for handgrip strength, gait speed, and skeletal muscle mass were derived directly from the GAMLSS reference distributions. Model-based Z scores for handgrip strength, gait speed, and skeletal muscle mass (Z handgrip strength, Z gait speed, Z muscle mass) represent the number of standard deviations from the corresponding age- and sex-specific mean. T scores (T handgrip strength, T gait speed, T muscle mass) were obtained by linear transformation. For handgrip strength, values below the 50th percentile were consistently lower than the age-specific mean, resulting in negative Z scores and T scores < 50. Across age and sex strata, the model-based 5th–95th percentile range corresponded to Z handgrip strength of approximately −1.76 to 1.80 and T handgrip strength of about 32–68 (Table 3). For gait speed, values below the mean extended up to the 50th percentile, with Z gait speed ranging from roughly −1.60 to 2.08 and T gait speed from about 34 to 71 (Table 4). Similarly, for skeletal muscle mass, values below the age- and sex-specific mean were observed up to the 50th percentile, with Z muscle mass ranging from approximately −1.55 to 2.02 and T muscle mass from about 35 to 70 (Table 5). These model-derived standardized scores provide a unified quantitative framework for interpreting sarcopenia parameters across age groups and between indicators.
Table 3.
Model-based standardized Z and T scores and centiles for handgrip strength by sex and age group.
| Female |
Male |
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | 5th | 15th | 20th | 50th | 75th | 85th | 95th | Mean ± SD | 5th | 15th | 20th | 50th | 75th | 85th | 95th | |
| Total | 20.03 ± 5 | 12 | 15.2 | 16.1 | 19.8 | 23.02 | 24.9 | 28.5 | 30.69 ± 7.46 | 18.7 | 23.4 | 24.9 | 30.5 | 35.3 | 38 | 43.4 |
| Z score | −1.61 | −0.97 | −0.79 | −0.05 | 0.6 | 0.97 | 1.69 | −1.61 | −0.98 | −0.78 | −0.02 | 0.62 | 0.98 | 1.7 | ||
| T score | 34 | 40 | 42 | 50 | 56 | 60 | 67 | 34 | 40 | 42 | 50 | 56 | 60 | 67 | ||
| Middle age | 22.00 ± 4.81 | 14.5 | 17.4 | 18.2 | 21.9 | 25 | 26.6 | 30.1 | 34.34 ± 7.45 | 23.1 | 27.2 | 28.4 | 34 | 39 | 41.9 | 46.7 |
| Z score | −1.56 | −0.96 | −0.79 | −0.02 | 0.62 | 0.96 | 1.69 | −1.51 | −0.96 | −0.8 | −0.05 | 0.63 | 1.01 | 1.66 | ||
| T score | 34 | 40 | 42 | 50 | 56 | 60 | 67 | 35 | 40 | 42 | 50 | 56 | 60 | 67 | ||
| Elderly | 18.49 ± 4.59 | 10.8 | 13.9 | 14.96 | 18.6 | 21.2 | 22.9 | 25.78 | 28.46 ± 6.53 | 17.6 | 21.8 | 23.2 | 28.5 | 32.8 | 35 | 38.9 |
| Z score | −1.68 | −1 | −0.77 | 0.02 | 0.59 | 0.96 | 1.59 | −1.66 | −1.02 | −0.8 | 0.01 | 0.66 | 1 | 1.6 | ||
| T score | 33 | 40 | 42 | 50 | 56 | 60 | 66 | 33 | 40 | 42 | 50 | 57 | 60 | 66 | ||
| 40–44 yr | 23.35 ± 4.96 | 16 | 18.68 | 19.56 | 23.51 | 26.84 | 28.67 | 31.84 | 36.77 ± 8.04 | 25.16 | 29.04 | 30.36 | 36.49 | 41.92 | 45.04 | 50.61 |
| Z score | −1.48 | −0.94 | −0.76 | 0.03 | 0.7 | 1.07 | 1.71 | −1.44 | −0.96 | −0.8 | −0.04 | 0.64 | 1.03 | 1.72 | ||
| T score | 35 | 41 | 42 | 50 | 57 | 61 | 67 | 36 | 40 | 42 | 50 | 56 | 60 | 67 | ||
| 45–49 yr | 23.79 ± 4.6 | 15.71 | 18.36 | 19.24 | 23.16 | 26.44 | 28.26 | 31.39 | 36.42 ± 7.23 | 24.38 | 28.45 | 29.8 | 35.88 | 41.03 | 43.89 | 48.86 |
| Z score | −1.76 | −1.18 | −0.99 | −0.14 | 0.58 | 0.97 | 1.65 | −1.67 | −1.1 | −0.92 | −0.07 | 0.64 | 1.03 | 1.72 | ||
| T score | 32 | 38 | 40 | 49 | 56 | 60 | 67 | 33 | 39 | 41 | 49 | 56 | 60 | 67 | ||
| 50–54 yr | 22.38 ± 4.61 | 14.97 | 17.56 | 18.41 | 22.22 | 25.41 | 27.18 | 30.22 | 35.08 ± 7.56 | 23.32 | 27.58 | 28.95 | 34.95 | 39.81 | 42.44 | 46.89 |
| Z score | −1.61 | −1.05 | −0.86 | −0.04 | 0.66 | 1.04 | 1.7 | −1.56 | −0.99 | −0.81 | −0.02 | 0.63 | 0.97 | 1.56 | ||
| T score | 34 | 40 | 41 | 50 | 57 | 60 | 67 | 34 | 40 | 42 | 50 | 56 | 60 | 66 | ||
| 55–59 yr | 21.32 ± 4.88 | 14.08 | 16.57 | 17.4 | 21.07 | 24.15 | 25.85 | 28.79 | 33.73 ± 6.82 | 22.06 | 26.43 | 27.8 | 33.63 | 38.19 | 40.6 | 44.6 |
| Z score | −1.48 | −0.97 | −0.8 | −0.05 | 0.58 | 0.93 | 1.53 | −1.71 | −1.07 | −0.87 | −0.01 | 0.65 | 1.01 | 1.59 | ||
| T score | 35 | 40 | 42 | 50 | 56 | 59 | 65 | 33 | 39 | 41 | 50 | 57 | 60 | 66 | ||
| 60–64 yr | 20.18 ± 4.21 | 13.44 | 15.87 | 16.67 | 20.25 | 23.25 | 24.91 | 27.76 | 31.68 ± 6.72 | 20.78 | 25.11 | 26.45 | 32.04 | 36.33 | 38.58 | 42.27 |
| Z score | −1.6 | −1.02 | −0.83 | 0.02 | 0.73 | 1.12 | 1.8 | −1.62 | −0.98 | −0.78 | 0.05 | 0.69 | 1.03 | 1.57 | ||
| T score | 34 | 40 | 42 | 50 | 57 | 61 | 68 | 34 | 40 | 42 | 51 | 57 | 60 | 66 | ||
| 65–69 yr | 19.85 ± 4.32 | 12.77 | 15.17 | 15.97 | 19.51 | 22.47 | 24.11 | 26.92 | 30.46 ± 6.25 | 19.6 | 23.75 | 25.03 | 30.39 | 34.5 | 36.64 | 40.18 |
| Z score | −1.64 | −1.08 | −0.9 | −0.08 | 0.61 | 0.99 | 1.64 | −1.74 | −1.07 | −0.87 | −0.01 | 0.65 | 0.99 | 1.55 | ||
| T score | 34 | 39 | 41 | 49 | 56 | 60 | 66 | 33 | 39 | 41 | 50 | 56 | 60 | 66 | ||
| 70–74 yr | 18.31 ± 4.16 | 11.47 | 13.84 | 14.63 | 18.12 | 21.06 | 22.68 | 25.47 | 28.76 ± 6.03 | 18.32 | 22.18 | 23.4 | 28.53 | 32.54 | 34.65 | 38.16 |
| Z score | −1.64 | −1.07 | −0.89 | −0.04 | 0.66 | 1.05 | 1.72 | −1.73 | −1.09 | −0.89 | −0.04 | 0.63 | 0.98 | 1.56 | ||
| T score | 34 | 39 | 41 | 50 | 57 | 60 | 67 | 33 | 39 | 41 | 50 | 56 | 60 | 66 | ||
| 75–79 yr | 17.01 ± 4.3 | 9.91 | 12.28 | 13.07 | 16.59 | 19.55 | 21.19 | 24.01 | 26.64 ± 6.12 | 16.69 | 20.32 | 21.47 | 26.44 | 30.39 | 32.5 | 36.04 |
| Z score | −1.65 | −1.1 | −0.92 | −0.1 | 0.59 | 0.97 | 1.63 | −1.62 | −1.03 | −0.84 | −0.03 | 0.61 | 0.96 | 1.53 | ||
| T score | 33 | 39 | 41 | 49 | 56 | 60 | 66 | 34 | 40 | 42 | 50 | 56 | 60 | 65 | ||
| ≥ 80 yr | 14.96 ± 4.8 | 8.35 | 10.73 | 11.52 | 15.08 | 18.08 | 19.74 | 22.62 | 23.43 ± 5.99 | 14.71 | 18.22 | 19.34 | 24.2 | 28.08 | 30.16 | 33.66 |
| Z score | −1.38 | −0.88 | −0.72 | 0.02 | 0.65 | 1 | 1.59 | −1.46 | −0.87 | −0.68 | 0.13 | 0.78 | 1.12 | 1.71 | ||
| T score | 36 | 41 | 43 | 50 | 56 | 60 | 66 | 35 | 41 | 43 | 51 | 58 | 61 | 67 | ||
Values are presented as model-based mean ± standard deviation (SD) and percentile cutoffs (5th, 15th, 20th, 50th, 75th, 85th, and 95th) for handgrip strength. Handgrip strength is expressed in kilograms (kg). Percentiles and standardized scores were derived from sex- and age-specific GAMLSS models; Z scores were calculated as Z = (X − μ)/σ and converted to T scores using T = 50 + 10 × Z. Data are stratified by sex and age group.
SD, standard deviation.
Table 4.
Model-based standardized Z and T scores and centiles for gait speed by sex and age group.
| Female |
Male |
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | 5th | 15th | 20th | 50th | 75th | 85th | 95th | Mean ± SD | 5th | 15th | 20th | 50th | 75th | 85th | 95th | |
| Total | 1.12 ± 0.37 | 0.58 | 0.8 | 0.85 | 1.09 | 1.27 | 1.5 | 1.82 | 1.12 ± 0.37 | 0.6 | 0.79 | 0.85 | 1.08 | 1.24 | 1.5 | 1.84 |
| Z score | −1.45 | −0.87 | −0.73 | −0.09 | 0.4 | 1.02 | 1.88 | −1.4 | −0.89 | −0.73 | −0.11 | 0.32 | 1.02 | 1.94 | ||
| T score | 35 | 41 | 43 | 49 | 54 | 60 | 69 | 36 | 41 | 43 | 49 | 53 | 60 | 69 | ||
| Middle age | 1.21 ± 0.36 | 0.75 | 0.92 | 1 | 1.13 | 1.38 | 1.51 | 1.9 | 1.23 ± 0.37 | 0.75 | 0.93 | 1 | 1.17 | 1.4 | 1.54 | 1.96 |
| Z score | −1.3 | −0.82 | −0.59 | −0.21 | 0.48 | 0.85 | 1.94 | −1.28 | −0.79 | −0.61 | −0.14 | 0.46 | 0.83 | 1.94 | ||
| T score | 37 | 42 | 44 | 48 | 55 | 59 | 69 | 37 | 42 | 44 | 49 | 55 | 58 | 69 | ||
| Elderly | 1.05 ± 0.37 | 0.5 | 0.72 | 0.76 | 1 | 1.2 | 1.41 | 1.71 | 1.05 ± 0.35 | 0.55 | 0.75 | 0.8 | 1 | 1.2 | 1.36 | 1.67 |
| Z score | −1.49 | −0.91 | −0.78 | −0.13 | 0.41 | 0.98 | 1.79 | −1.43 | −0.86 | −0.72 | −0.15 | 0.41 | 0.86 | 1.74 | ||
| T score | 35 | 41 | 42 | 49 | 54 | 60 | 68 | 36 | 41 | 43 | 48 | 54 | 59 | 67 | ||
| 40–44 yr | 1.22 ± 0.31 | 0.77 | 0.91 | 0.95 | 1.18 | 1.4 | 1.53 | 1.78 | 1.26 ± 0.36 | 0.8 | 0.93 | 0.98 | 1.22 | 1.47 | 1.63 | 1.95 |
| Z score | −1.43 | −1.01 | −0.86 | −0.12 | 0.59 | 1.02 | 1.82 | −1.29 | −0.93 | −0.8 | −0.13 | 0.57 | 1.01 | 1.9 | ||
| T score | 36 | 40 | 41 | 49 | 56 | 60 | 68 | 37 | 41 | 42 | 49 | 56 | 60 | 69 | ||
| 45–49 yr | 1.24 ± 0.37 | 0.76 | 0.9 | 0.95 | 1.2 | 1.43 | 1.57 | 1.83 | 1.29 ± 0.38 | 0.78 | 0.92 | 0.97 | 1.22 | 1.47 | 1.63 | 1.95 |
| Z score | −1.3 | −0.91 | −0.78 | −0.11 | 0.52 | 0.91 | 1.62 | −1.32 | −0.97 | −0.84 | −0.19 | 0.48 | 0.9 | 1.74 | ||
| T score | 37 | 41 | 42 | 49 | 55 | 59 | 66 | 37 | 40 | 42 | 48 | 55 | 59 | 67 | ||
| 50–54 yr | 1.23 ± 0.36 | 0.73 | 0.88 | 0.94 | 1.19 | 1.44 | 1.58 | 1.85 | 1.24 ± 0.38 | 0.76 | 0.89 | 0.94 | 1.19 | 1.45 | 1.61 | 1.92 |
| Z score | −1.38 | −0.96 | −0.82 | −0.11 | 0.56 | 0.96 | 1.71 | −1.26 | −0.9 | −0.77 | −0.11 | 0.55 | 0.97 | 1.79 | ||
| T score | 36 | 40 | 42 | 49 | 56 | 60 | 67 | 37 | 41 | 42 | 49 | 56 | 60 | 68 | ||
| 55–59 yr | 1.2 ± 0.31 | 0.7 | 0.86 | 0.91 | 1.18 | 1.43 | 1.58 | 1.85 | 1.22 ± 0.36 | 0.72 | 0.86 | 0.91 | 1.16 | 1.41 | 1.56 | 1.86 |
| Z score | −1.6 | −1.11 | −0.93 | −0.08 | 0.72 | 1.2 | 2.08 | −1.38 | −0.99 | −0.86 | −0.17 | 0.52 | 0.94 | 1.76 | ||
| T score | 34 | 39 | 41 | 49 | 57 | 62 | 71 | 36 | 40 | 41 | 48 | 55 | 59 | 68 | ||
| 60–64 yr | 1.18 ± 0.39 | 0.65 | 0.81 | 0.87 | 1.14 | 1.4 | 1.55 | 1.83 | 1.17 ± 0.38 | 0.68 | 0.82 | 0.87 | 1.12 | 1.36 | 1.51 | 1.8 |
| Z score | −1.33 | −0.92 | −0.78 | −0.09 | 0.56 | 0.94 | 1.64 | −1.29 | −0.91 | −0.78 | −0.12 | 0.53 | 0.92 | 1.68 | ||
| T score | 37 | 41 | 42 | 49 | 56 | 59 | 66 | 37 | 41 | 42 | 49 | 55 | 59 | 67 | ||
| 65–69 yr | 1.13 ± 0.35 | 0.6 | 0.76 | 0.82 | 1.09 | 1.35 | 1.5 | 1.77 | 1.12 ± 0.35 | 0.64 | 0.78 | 0.83 | 1.08 | 1.32 | 1.46 | 1.74 |
| Z score | −1.52 | −1.06 | −0.89 | −0.12 | 0.6 | 1.02 | 1.79 | −1.38 | −0.98 | −0.83 | −0.12 | 0.57 | 0.99 | 1.77 | ||
| T score | 35 | 39 | 41 | 49 | 56 | 60 | 68 | 36 | 40 | 42 | 49 | 56 | 60 | 68 | ||
| 70–74 yr | 1.06 ± 0.36 | 0.52 | 0.69 | 0.74 | 1.01 | 1.26 | 1.4 | 1.67 | 1.07 ± 0.33 | 0.58 | 0.73 | 0.78 | 1.03 | 1.26 | 1.4 | 1.67 |
| Z score | −1.5 | −1.05 | −0.89 | −0.13 | 0.56 | 0.97 | 1.7 | −1.47 | −1.04 | −0.89 | −0.15 | 0.57 | 1 | 1.79 | ||
| T score | 35 | 39 | 41 | 49 | 56 | 60 | 67 | 35 | 40 | 41 | 49 | 56 | 60 | 68 | ||
| 75–79 yr | 0.94 ± 0.37 | 0.44 | 0.6 | 0.65 | 0.91 | 1.14 | 1.27 | 1.51 | 1 ± 0.35 | 0.5 | 0.64 | 0.69 | 0.94 | 1.18 | 1.32 | 1.58 |
| Z score | −1.37 | −0.95 | −0.8 | −0.1 | 0.53 | 0.89 | 1.55 | −1.43 | −1.02 | −0.88 | −0.17 | 0.5 | 0.9 | 1.64 | ||
| T score | 36 | 41 | 42 | 49 | 55 | 59 | 65 | 36 | 40 | 41 | 48 | 55 | 59 | 66 | ||
| ≥ 80 yr | 0.78 ± 0.3 | 0.36 | 0.5 | 0.55 | 0.79 | 1 | 1.12 | 1.34 | 0.85 ± 0.35 | 0.4 | 0.55 | 0.6 | 0.85 | 1.08 | 1.22 | 1.47 |
| Z score | −1.41 | −0.93 | −0.77 | 0.01 | 0.71 | 1.11 | 1.82 | −1.26 | −0.85 | −0.71 | 0 | 0.67 | 1.06 | 1.78 | ||
| T score | 36 | 41 | 42 | 50 | 57 | 61 | 68 | 37 | 41 | 43 | 50 | 57 | 61 | 68 | ||
Values are presented as model-based mean ± standard deviation (SD) and percentile cutoffs (5th, 15th, 20th, 50th, 75th, 85th, and 95th) for gait speed. Gait speed is expressed in meters per second (m/s). Percentiles and standardized scores were derived from sex- and age-specific GAMLSS models; Z scores were calculated as Z = (X − μ)/σ and converted to T scores using T = 50 + 10 × Z. Data are stratified by sex and age group.
SD, standard deviation.
Table 5.
Model-based standardized Z and T scores and centiles for skeletal muscle mass by sex and age group.
| Female |
Male |
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | 5th | 15th | 20th | 50th | 75th | 85th | 95th | Mean ± SD | 5th | 15th | 20th | 50th | 75th | 85th | 95th | |
| Total | 6.9 ± 1 | 5.66 | 6 | 6.11 | 6.69 | 7.4 | 7.97 | 8.8 | 8.74 ± 1.33 | 6.8 | 7.4 | 7.6 | 8.61 | 9.58 | 10.06 | 11.09 |
| Z score | −1.24 | −0.9 | −0.79 | −0.21 | 0.5 | 1.07 | 1.9 | −1.46 | −1.01 | −0.86 | −0.1 | 0.63 | 0.99 | 1.77 | ||
| T score | 38 | 41 | 42 | 48 | 55 | 61 | 69 | 35 | 40 | 41 | 49 | 56 | 60 | 68 | ||
| Middle age | 6.86 ± 1 | 5.65 | 5.97 | 6.07 | 6.66 | 7.33 | 7.85 | 8.73 | 9.04 ± 1.29 | 7.26 | 7.77 | 7.92 | 8.89 | 9.81 | 10.31 | 11.39 |
| Z score | −1.21 | −0.89 | −0.79 | −0.2 | 0.47 | 0.99 | 1.87 | −1.38 | −0.99 | −0.87 | −0.12 | 0.59 | 0.98 | 1.81 | ||
| T score | 38 | 41 | 42 | 48 | 55 | 60 | 69 | 36 | 40 | 41 | 49 | 56 | 60 | 68 | ||
| Elderly | 6.93 ± 1 | 5.66 | 6.01 | 6.15 | 6.71 | 7.45 | 8.05 | 8.84 | 8.55 ± 1.32 | 6.64 | 7.21 | 7.42 | 8.41 | 9.43 | 9.91 | 10.92 |
| Z score | −1.27 | −0.92 | −0.78 | −0.22 | 0.52 | 1.12 | 1.91 | −1.45 | −1.02 | −0.86 | −0.11 | 0.67 | 1.03 | 1.8 | ||
| T score | 37 | 41 | 42 | 48 | 55 | 61 | 69 | 35 | 40 | 41 | 49 | 57 | 60 | 68 | ||
| 40–44 yr | 7 ± 1.18 | 5.63 | 5.97 | 6.09 | 6.74 | 7.44 | 7.92 | 9 | 9.25 ± 1.29 | 7.36 | 7.98 | 8.19 | 9.17 | 10.05 | 10.56 | 11.47 |
| Z score | −1.17 | −0.88 | −0.77 | −0.22 | 0.37 | 0.78 | 1.7 | −1.47 | −0.99 | −0.82 | −0.06 | 0.62 | 1.01 | 1.72 | ||
| T score | 38 | 41 | 42 | 48 | 54 | 58 | 67 | 35 | 40 | 42 | 49 | 56 | 60 | 67 | ||
| 45–49 yr | 6.95 ± 1.05 | 5.62 | 5.96 | 6.08 | 6.71 | 7.37 | 7.81 | 8.77 | 9.17 ± 1.23 | 7.38 | 7.93 | 8.12 | 9.05 | 9.94 | 10.47 | 11.5 |
| Z score | −1.26 | −0.94 | −0.82 | −0.23 | 0.39 | 0.81 | 1.73 | −1.45 | −1.01 | −0.85 | −0.1 | 0.62 | 1.06 | 1.89 | ||
| T score | 37 | 41 | 42 | 48 | 54 | 58 | 67 | 35 | 40 | 41 | 49 | 56 | 61 | 69 | ||
| 50–54 yr | 6.79 ± 0.92 | 5.61 | 5.94 | 6.06 | 6.67 | 7.31 | 7.73 | 8.62 | 9.16 ± 1.39 | 7.34 | 7.85 | 8.03 | 8.93 | 9.81 | 10.37 | 11.47 |
| Z score | −1.29 | −0.92 | −0.79 | −0.12 | 0.57 | 1.03 | 2 | −1.31 | −0.94 | −0.81 | −0.17 | 0.47 | 0.87 | 1.66 | ||
| T score | 37 | 41 | 42 | 49 | 56 | 60 | 70 | 37 | 41 | 42 | 48 | 55 | 59 | 67 | ||
| 55–59 yr | 6.76 ± 0.9 | 5.6 | 5.94 | 6.06 | 6.67 | 7.3 | 7.71 | 8.57 | 9.05 ± 1.2 | 7.21 | 7.73 | 7.91 | 8.81 | 9.68 | 10.23 | 11.3 |
| Z score | −1.3 | −0.92 | −0.78 | −0.1 | 0.6 | 1.06 | 2.02 | −1.52 | −1.09 | −0.94 | −0.2 | 0.53 | 0.98 | 1.87 | ||
| T score | 37 | 41 | 42 | 49 | 56 | 61 | 70 | 35 | 39 | 41 | 48 | 55 | 60 | 69 | ||
| 60–64 yr | 6.87 ± 1 | 5.59 | 5.95 | 6.07 | 6.7 | 7.34 | 7.76 | 8.62 | 8.77 ± 1.29 | 7 | 7.56 | 7.75 | 8.69 | 9.56 | 10.09 | 11.08 |
| Z score | −1.28 | −0.93 | −0.8 | −0.17 | 0.47 | 0.89 | 1.76 | −1.37 | −0.94 | −0.79 | −0.06 | 0.62 | 1.03 | 1.8 | ||
| T score | 37 | 41 | 42 | 48 | 55 | 59 | 68 | 36 | 41 | 42 | 49 | 56 | 60 | 68 | ||
| 65–69 yr | 6.91 ± 1.02 | 5.6 | 5.96 | 6.09 | 6.74 | 7.4 | 7.83 | 8.7 | 8.68 ± 1.25 | 6.74 | 7.36 | 7.58 | 8.56 | 9.45 | 9.96 | 10.89 |
| Z score | −1.28 | −0.93 | −0.8 | −0.17 | 0.48 | 0.89 | 1.75 | −1.55 | −1.05 | −0.88 | −0.09 | 0.62 | 1.03 | 1.76 | ||
| T score | 37 | 41 | 42 | 48 | 55 | 59 | 68 | 35 | 40 | 41 | 49 | 56 | 60 | 68 | ||
| 70–74 yr | 6.93 ± 0.98 | 5.61 | 5.98 | 6.12 | 6.78 | 7.46 | 7.89 | 8.77 | 8.5 ± 1.33 | 6.54 | 7.19 | 7.41 | 8.44 | 9.36 | 9.88 | 10.82 |
| Z score | −1.35 | −0.97 | −0.83 | −0.15 | 0.54 | 0.98 | 1.87 | −1.48 | −0.99 | −0.82 | −0.05 | 0.64 | 1.04 | 1.75 | ||
| T score | 37 | 40 | 42 | 48 | 55 | 60 | 69 | 35 | 40 | 42 | 50 | 56 | 60 | 67 | ||
| 75–79 yr | 6.94 ± 1.02 | 5.62 | 6 | 6.14 | 6.82 | 7.5 | 7.93 | 8.8 | 8.46 ± 1.37 | 6.43 | 7.07 | 7.29 | 8.32 | 9.27 | 9.83 | 10.85 |
| Z score | −1.3 | −0.92 | −0.78 | −0.12 | 0.55 | 0.98 | 1.83 | −1.48 | −1.01 | −0.86 | −0.1 | 0.59 | 1 | 1.74 | ||
| T score | 37 | 41 | 42 | 49 | 56 | 60 | 68 | 35 | 40 | 41 | 49 | 56 | 60 | 67 | ||
| ≥ 80 yr | 6.97 ± 0.92 | 5.63 | 6.02 | 6.16 | 6.85 | 7.52 | 7.95 | 8.79 | 8.41 ± 1.38 | 6.36 | 6.97 | 7.18 | 8.2 | 9.17 | 9.76 | 10.88 |
| Z score | −1.46 | −1.04 | −0.88 | −0.14 | 0.6 | 1.06 | 1.98 | −1.48 | −1.04 | −0.89 | −0.15 | 0.55 | 0.98 | 1.79 | ||
| T score | 35 | 40 | 41 | 49 | 56 | 61 | 70 | 35 | 40 | 41 | 48 | 56 | 60 | 68 | ||
Values are presented as model-based mean ± standard deviation (SD) and percentile cutoffs (5th, 15th, 20th, 50th, 75th, 85th, and 95th) for skeletal muscle mass. Skeletal muscle mass is expressed as skeletal muscle mass index (kg/m2). Percentiles and standardized scores were derived from sex- and age-specific GAMLSS models; Z scores were calculated as Z = (X − μ)/σ and converted to T scores using T = 50 + 10 × Z. Data are stratified by sex and age group.
SD, standard deviation.
3.5. Optimal T-score cut-off points for sarcopenia indicators
Using the AWGS diagnostic cut-off points as the reference standard, ROC analyses were performed to identify optimal T-score thresholds for low handgrip strength, slow gait speed, and low appendicular skeletal muscle mass index. The corresponding T-score cut-offs were 45.83 in women and 46.33 in men for handgrip strength, 46.70 and 46.74 for gait speed, and 41.42 and 40.76 for appendicular skeletal muscle mass index in women and men, respectively. These thresholds yielded sensitivities of 0.997–1.000 and specificities of 0.996–1.000, with areas under the ROC curve (AUCs) ranging from 0.999 to 1.000, indicating almost perfect discrimination.
4. Discussion
This study is, to our knowledge, the first to establish sex- and age-specific normative reference values, percentile ranks, GAMLSS-based smoothed centile curves, and standardized Z and T scores for handgrip strength, gait speed, and skeletal muscle mass in a large cohort of middle-aged and older adults attending preventive health check-ups in Taiwan. By stratifying participants into 5-year age groups, we demonstrated consistent and statistically significant declines in all three sarcopenia indicators with advancing age. These differential patterns across age strata provide construct validity, confirming that the measures capture the expected age-related deterioration in muscle strength, physical performance, and muscle mass, which are central to the clinical definition of sarcopenia [1].
Comparable trends have been reported in normative studies from other Asian populations. Compared with a population-based study from Singapore [9], Taiwanese adults aged 40 years and older demonstrated lower handgrip strength during midlife, though this difference became less pronounced in advanced age. Average gait speed was comparable between cohorts. However, Taiwanese participants, regardless of sex, exhibited higher skeletal muscle mass than their Singaporean counterparts by approximately 1–2 kg/m2. In terms of age-related trends, muscle mass in Taiwanese men showed a modest decline, while it remained relatively stable in women. Conversely, the Singaporean data showed more substantial age-associated muscle loss in both sexes.
These inter-population differences may reflect underlying variations in body composition, lifestyle patterns, and health-related behaviors [23]. The relatively preserved skeletal muscle mass observed among older Taiwanese women in this cohort may be influenced by demographic or environmental factors. Additionally, methodological differences in muscle mass assessment may partially account for the observed discrepancies. The Singaporean study employed DXA to assess muscle mass, which provides high measurement precision and is less susceptible to short-term fluctuations in hydration or environmental conditions. In contrast, our study utilized BIA, a more feasible modality for large-scale or community-based screening but one that can be influenced by factors such as hydration status, ambient temperature, and metallic implants [24]. These methodological differences in body composition assessment likely contributed, at least in part, to the discrepancies observed when comparing skeletal muscle mass values across studies.
From the GAMLSS-based centile models, the percentile distribution analysis revealed that mean values for handgrip strength, gait speed, and skeletal muscle mass were consistently located near the 50th percentile across sex- and age-specific strata. With advancing age, the percentile thresholds corresponding to low handgrip strength, slow gait, and reduced muscle mass demonstrated progressive upward shifts, ranging from the 5th to the 50th percentile depending on the indicators. These findings reflect an age-associated shift in the normative distribution of muscle strength, functional performance, and muscle quantity.
Given the limited availability of percentile-based reference values for sarcopenia, we compared our results with the 20th percentile thresholds proposed by Wu et al. for Taiwanese community-dwelling adults aged 65 years and older [25]. In our cohort, the 20th percentile values for handgrip strength were consistently lower across sex and age groups. In contrast, 20th percentile gait speed values in our sample were higher than those reported by Wu et al. For skeletal muscle mass, women in our study exhibited higher 20th percentile values across all age groups, while values in men were generally comparable between datasets. These discrepancies may reflect differences in participant age distribution and assessment methodology. In addition, the life-course handgrip percentiles reported by Dodds et al. for United Kingdom community-dwelling adults [21] were generally higher than those in our Taiwanese study; across most percentiles, the UK GAMLSS curves lay above the corresponding curves for both men and women aged 40 years and older. This pattern suggests population differences in muscle strength and supports the need for region- and population-specific reference standards.
The use of standardized scores based on the mean and standard deviation is a well-established method in clinical research, as demonstrated in bone mineral density classification for osteoporosis [26]. In this study, we applied this approach to standardize sarcopenia-related measurements, including handgrip strength, gait speed, and skeletal muscle mass, by converting raw values into Z scores. At the 20th percentile, handgrip strength showed the largest negative deviation from the population mean, followed by gait speed and skeletal muscle mass (Zhandgrip strength < Zgait speed < Zmuscle mass). This pattern was consistent across sexes and among adults aged 65 and older, suggesting that reduced muscle strength is the most prominent deficit in this population. In contrast, a nationwide study of Taiwanese older adults conducted from 2017 to 2021 identified low muscle mass as the main deficit at the 20 percentile [27]. Such discrepancies likely reflect temporal variations in population composition and measurement variability, which can influence both average values and the relative distribution of individuals.
To enhance clinical interpretability, Z scores were further linearly transformed into T scores on an approximate 100-point scale. While Z scores provide statistically robust estimates of deviation from the population mean, T scores offer a more intuitive framework for clinical decision-making. In this study, T scores corresponding to the 5th, 50th, and 95th percentiles were approximately 35–40, 50, and 65–70, respectively, and this pattern remained consistent across sex and age subgroups. T scores are reference group–dependent, and their interpretation is influenced by the underlying distribution of the comparison group. To support stratified comparisons, we generated sex-specific T score reference values for the overall population, middle-aged and older adults, and for each five-year age band. For instance, a T score of 50 in an 80-year-old reflects average performance within the ≥ 80-year group, but may correspond to a lower T score when compared with the broader ≥ 65-year population, due to differences in group-specific means and standard deviations. This standardized scoring model allows for clearer interpretation of individual performance across diverse age groups.
In addition, we expressed the AWGS diagnostic cut-off points for low handgrip strength, slow gait speed, and low skeletal muscle mass in the T-score metric, allowing clinicians to relate individual T scores to widely used diagnostic thresholds. In our data, T scores in the mid-40s for handgrip strength and gait speed and around 40 for skeletal muscle mass corresponded closely to the AWGS cut-off points. The near-perfect ROC performance largely reflects this internal derivation from the same dataset rather than external diagnostic validation, and future prospective studies are needed to determine which T-score thresholds best predict incident sarcopenia, functional decline, falls, and response to interventions.
The use of T scores in sarcopenia assessment represents a novel and clinically relevant approach that enhances interpretability and facilitates communication across clinical and public health settings. T scores provide a standardized, intuitive metric to quantify deviations from population norms, supporting risk stratification, early detection, and longitudinal monitoring. Percentile ranking further situates individual performance within reference populations. Although T scores are widely used in neurological and behavioral assessments, such as Neuro-QoL and Conners 3 [28,29], their application in sarcopenia research remains limited. Further studies are needed to define diagnostic thresholds and inform clinical decision-making.
This study has several limitations. First, skeletal muscle mass was measured using BIA, which, despite its practicality and accessibility in large-scale settings, may be influenced by hydration status and environmental conditions, potentially introducing minor measurement variability. Second, participants were individuals who voluntarily attended adult preventive health check-ups at a single tertiary hospital and may therefore represent a relatively healthier and more health-conscious subgroup of the population, potentially introducing selection bias and limiting the generalizability of the findings. Third, the cohort consisted predominantly of Taiwanese adults, with limited ethnic diversity, which may restrict the generalizability of these normative references to non-Asian populations; future studies including more diverse ethnic groups would help enhance representativeness and support broader international application of standardized scores. In addition, we did not formally quantify within-session measurement error or intra- and inter-rater reliability for the sarcopenia measurements in this cohort, and some degree of measurement error cannot be excluded.
Nonetheless, the availability of normative data enables the development of clinically interpretable cut-off values, such as Z scores and T scores, to identify individuals at risk for sarcopenia. These reference thresholds may facilitate early diagnosis, guide intervention planning, and support the monitoring of muscle health in both preventive and clinical contexts.
5. Conclusions
In conclusion, this study established sex- and age-specific normative reference values for handgrip strength, gait speed, and skeletal muscle mass in Taiwanese adults aged 40 years and older. By applying standardized scoring methods and converting Z scores into T scores on a scale approximating 100 points, this approach improves the interpretability of sarcopenia-related metrics. The observed age-related declines across all indicators support construct validity, confirming that these measures reflect expected aging-related deterioration. Together, these findings offer a practical reference framework to support sarcopenia screening, early intervention, and public health strategies aimed at promoting healthy aging in Taiwan.
CRediT author statement
Hsiu-Wen Kuo: Conceptualization, Methodology, Data collection, Analysis, and Interpretation, Writing-original draft, & editing. Chih-Dao Chen: Conceptualization, Methodology, Writing-review & editing, Supervision. Ariel Chang-Yu Wu: Conceptualization, Methodology, Writing-review & editing, Supervision.
Conflicts of interest
The authors declare no competing interests.
Acknowledgments
The authors thank the Preventive Health Center of Far Eastern Memorial Hospital for their assistance with the National Adult Preventive Health Program and sarcopenia assessments, from which the study data were obtained. ORCID Hsiu-Wen Kuo: 0009-0004-5447-6132. Chih-Dao Chen: 0000-0003-0338-7346. Ariel Chang-Yu Wu: 0009-0001-5124-1846.
Contributor Information
Chih-Dao Chen, Email: cdchen@mail.femh.org.tw.
Ariel Chang-Yu Wu, Email: lspsariel@gmail.com.
References
- 1.Cooper C, Dere W, Evans W, Kanis JA, Rizzoli R, Sayer AA, et al. Frailty and sarcopenia: definitions and outcome parameters. Osteoporos Int. 2012;23:1839–1848. doi: 10.1007/s00198-012-1913-1. [DOI] [PubMed] [Google Scholar]
- 2.Rosenberg IH. Sarcopenia: origins and clinical relevance. J Nutr. 1997;127(5 Suppl):990S–991S. doi: 10.1093/jn/127.5.990S. [DOI] [PubMed] [Google Scholar]
- 3.Cruz-Jentoft AJ, Baeyens JP, Bauer JM, Boirie Y, Cederholm T, Landi F, et al. European working group on Sarcopenia in older people. Sarcopenia: European consensus on definition and diagnosis: report of the european working group on Sarcopenia in older people. Age Ageing. 2010;39:412–423. doi: 10.1093/ageing/afq034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Cruz-Jentoft AJ, Bahat G, Bauer J, Boirie Y, Bruyère O, Cederholm T, et al. Writing group for the European working group on Sarcopenia in older people 2 (EWGSOP2), and the extended group for EWGSOP2. Sarcopenia: revised european consensus on definition and diagnosis. Age Ageing. 2019;48:16–31. doi: 10.1093/ageing/afy169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chen LK, Liu LK, Woo J, Assantachai P, Auyeung TW, Bahyah KS, et al. Sarcopenia in Asia: consensus report of the Asian working group for Sarcopenia. J Am Med Dir Assoc. 2014;15:95–101. doi: 10.1016/j.jamda.2013.11.025. [DOI] [PubMed] [Google Scholar]
- 6.Chen LK., Woo J, Assantachai P, Auyeung TW, Chou MY, Iijima K, et al. Asian working group for Sarcopenia: 2019 consensus update on Sarcopenia diagnosis and treatment. J Am Med Dir Assoc. 2020;21:300.e2–307.e2. doi: 10.1016/j.jamda.2019.12.012. [DOI] [PubMed] [Google Scholar]
- 7.Peel NM, Kuys SS, Klein K. Gait speed as a measure in geriatric assessment in clinical settings: a systematic review. J Gerontol A Biol Sci Med Sci. 2013;68:39–46. doi: 10.1093/gerona/gls174. [DOI] [PubMed] [Google Scholar]
- 8.De Beurs E, Boehnke JR, Fried EI. Common measures or common metrics? A plea to harmonize measurement results. Clin Psychol Psychother. 2022;29:1755–1767. doi: 10.1002/cpp.2742. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pang BWJ, Wee SL, Lau LK, Jabbar KA, Seah WT, Ng DHM, et al. Prevalence and associated factors of sarcopenia in Singaporean adults-the yishun study. J Am Med Dir Assoc. 2021;22:885.e1–885.e10. doi: 10.1016/j.jamda.2020.05.029. [DOI] [PubMed] [Google Scholar]
- 10.Tee ML, Tee CA, Montemayor EB. Determination of normative reference for the definition of sarcopenia among Filipinos. Osteoporos Sarcopenia. 2016;2:186–190. doi: 10.1016/j.afos.2016.07.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dodds RM, Syddall HE, Cooper R, Kuh D, Cooper C, Sayer AA. Global variation in grip strength: a systematic review and meta-analysis of normative data. Age Ageing. 2016;45:209–216. doi: 10.1093/ageing/afv192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Imboden MT, Swartz AM, Finch HW, Harber MP, Kaminsky LA. Reference standards for lean mass measures using GE dual energy x-ray absorptiometry in Caucasian adults. PLoS One. 2017;12 doi: 10.1371/journal.pone.0176161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Pan PJ, Lin CH, Yang NP, Chen HC, Tsao HM, Chou P, et al. Normative data and associated factors of hand grip strength among elderly individuals: the Yilan study, Taiwan. Sci Rep. 2020;10:6611. doi: 10.1038/s41598-020-63713-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Suetta C, Haddock B, Alcazar J, Noerst T, Hansen OM, Ludvig H, et al. The Copenhagen Sarcopenia study: lean mass, strength, power, and physical function in a Danish cohort aged 20-93 years. J Cachexia Sarcopenia Muscle. 2019;10:1316–1329. doi: 10.1002/jcsm.12477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Steiber N. Strong or weak handgrip? Normative reference values for the German population across the life course stratified by sex, age, and body height. PLoS One. 2016;11 doi: 10.1371/journal.pone.0163917. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Landi F, Calvani R, Tosato M, Martone AM, Fusco D, Sisto A, et al. Age-related variations of muscle mass, strength, and physical performance in community-Dwellers: results from the Milan EXPO survey. J Am Med Dir Assoc. 2017;18:88.e17–88.e24. doi: 10.1016/j.jamda.2016.10.007. [DOI] [PubMed] [Google Scholar]
- 17.O’Connor PJ. Normative data: their definition, interpretation, and importance for primary care physicians. Fam Med. 1990;22:307–311. [PubMed] [Google Scholar]
- 18.Roberts HC, Denison HJ, Martin HJ, Patel HP, Syddall H, Cooper C, et al. A review of the measurement of grip strength in clinical and epidemiological studies: towards a standardised approach. Age Ageing. 2011;40:423–429. doi: 10.1093/ageing/afr051. [DOI] [PubMed] [Google Scholar]
- 19.Strauss ME, Smith GT. Construct validity: advances in theory and methodology. Annu Rev Clin Psychol. 2009;5:1–25. doi: 10.1146/annurev.clinpsy.032408.153639. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Stasinopoulos DM, Rigby RA. Generalized additive models for location, scale and shape (GAMLSS) in R. J Stat Software. 2007;23:1–46. [Google Scholar]
- 21.Dodds RM, Syddall HE, Cooper R, Benzeval M, Deary IJ, Dennison EM, et al. Grip strength across the life course: normative data from twelve British studies. PLoS One. 2014;9 doi: 10.1371/journal.pone.0113637. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Youden WJ. Index for rating diagnostic tests. Cancer. 1950;3:32–35. doi: 10.1002/1097-0142(1950)3:1<32::aid-cncr2820030106>3.0.co;2-3. [DOI] [PubMed] [Google Scholar]
- 23.Tyrovolas S, Koyanagi A, Olaya B, Ayuso-Mateos JL, Miret M, Chatterji S, et al. Factors associated with skeletal muscle mass, sarcopenia, and sarcopenic obesity in older adults: a multi-continent study. J Cachexia Sarcopenia Muscle. 2016;7:312–321. doi: 10.1002/jcsm.12076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Maden-Wilkinson TM, Degens H, Jones DA, McPhee JS. Comparison of MRI and DXA to measure muscle size and age-related atrophy in thigh muscles. J Musculoskelet Neuronal Interact. 2013;13:320–328. [PubMed] [Google Scholar]
- 25.Wu IC, Lin CC, Hsiung CA, Wang CY, Wu CH, Chan DC, et al. Sarcopenia and translational aging research in Taiwan team. Epidemiology of sarcopenia among community-dwelling older adults in Taiwan: a pooled analysis for a broader adoption of sarcopenia assessments. Geriatr Gerontol Int. 2014;14(Suppl 1):52–60. doi: 10.1111/ggi.12193. [DOI] [PubMed] [Google Scholar]
- 26.Sheu A, Diamond T. Bone mineral density: testing for osteoporosis. Aust Prescr. 2016;39:35–39. doi: 10.18773/austprescr.2016.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kuo HW, Chen CD, Yen AM, Chen C, Fan YT. Development and validation of the sarcopenia composite index: a comprehensive approach for assessing sarcopenia in the ageing population. Ann Acad Med Singapore. 2025;54:101–112. doi: 10.47102/annals-acadmedsg.2024272. [DOI] [PubMed] [Google Scholar]
- 28.Cella D, Lai JS, Nowinski CJ, Victorson D, Peterman A, Miller D, et al. Neuro-QOL: brief measures of health-related quality of life for clinical research in neurology. Neurology. 2012;78:1860–1867. doi: 10.1212/WNL.0b013e318258f744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Conners CK, Pitkanen J, Rzepa SR, Conners . In: Encyclopedia of clinical neuropsychology. third ed. Kreutzer JS, DeLuca J, Caplan B, editors. Springer; New York: 2011. Conners 3; Conners 2008) pp. 675–678. [Google Scholar]