The Optimal Cut-off Value of Upper Arm Circumference and Calf Circumference for Assessing Sarcopenia Among Chinese Community-Dwelling Older Adults

Introduction

Sarcopenia is becoming increasingly common in older adults due to rapid aging. Globally, more than 120 million older adults are diagnosed with sarcopenia, and the number is expected to double by 2050.¹ Sarcopenia is characterized by declined muscle mass, muscle strength and physical function,² which has been classified by WHO as an independent disease with the code ICD-10-CM (M62.84).³ Sarcopenia significantly increases the risks of health adverse events, including fracture,⁴ hospitalization,⁵ and all-cause mortality,⁶ greatly increasing the healthcare costs and the need for medical resources.⁷

However, it is almost impossible for sarcopenic older adults to realize their conditions due to no obvious symptoms in the early stages.⁸ The common situation is that when they realize their conditions, they are already physically and functionally dependent.⁸ Thus, it is imperative to diagnose sarcopenia in the early stages to provide timely interventions. However, certain limitations exist in the equipment diagnosed for sarcopenia (ie computed tomography, magnetic resonance imaging), such as radiation, high costs, and requirement for highly trained personnel.⁹ More importantly, the above equipment is not available in many Chinese primary and community hospitals, which are supposed to provide early screening and initial treatment for older adults. Consequently, it is highly important to find a simple, rapid and accurate screening approach for early identification of patients with sarcopenia, especially for countries like China with large population and limited medical resources.

Calf circumference (CC), an inexpensive, non-invasive and easy to be measured indicator,¹⁰ has been used in sarcopenia screening.^10–12 Although CC is widely used in screening for sarcopenia, limited studies are conducted in Chinese community-dwelling older adults, according to a recent scoping review.¹³ The screening ability and cut-off values of CC on sarcopenia screening interfere with geographic variations and ethnic differences.^14,15 Thus, accurate and Chinese tailor-made cut-off values of CC on the screening for sarcopenia are essential to identify potential patients with sarcopenia. Although the Asian Working Group for Sarcopenia 2019 (AWGS 2019) has recommended the cut-off values of CC on sarcopenia screening (males: 34 cm; females: 33 cm),¹⁶ some limitations exist in the recommendations¹⁶ and above studies.^10–12 First, many studies, including those referred in the AWGS 2019, did not have the “gold standard” for the diagnosis of sarcopenia, only focusing on low muscle mass,^17,18 which is different from sarcopenia. Second, these studies included middle-aged adults (age 40 years or older), not targeting older adults,^17,18 which leads to the recommended cut-off values that may not be suitable for older adults. Third, these studies were from Canada,¹⁹ Japan,¹⁷ and Korea.¹⁴ Only one study explored the evaluation effect of calf circumference on reduced muscle mass in community-dwelling elderly adults in China.²⁰ The study found that calf circumference has a good evaluation effect on reduced skeletal muscle index (SMI) in elderly adults in Shanghai. However, the study did not explore its role in screening and diagnosing sarcopenia. In addition, it also pointed out that the evaluation effect of calf circumference on reduced SMI can be influenced by factors such as geography. Overall, the cut-off values of CC screening for sarcopenia in Chinese community-dwelling older adults are still unclear.

In addition to CC, upper arm circumference (AC) is another anthropometric index associated with muscle mass¹³ and has good screening performance on sarcopenia in Brazilian older adults²¹ and Japanese patients with chronic liver disease.²² Similar to CC, the screening ability and cut-off values of AC on sarcopenia are influenced by age, gender, region and race.¹⁴ To our best known, there is no study exploring the screening ability of AC on sarcopenia in Chinese community-dwelling older adults.

AC and CC have been used as indicators for nutritional assessment,^23,24 and nutrition plays a crucial role in the development and progression of sarcopenia,² thereby affecting the health status of older adults, such as physical function. Therefore, in addition to the early detection of patients with sarcopenia, it is also pivotal to explore the evaluation effects of CC and AC on health assessments among older adults with sarcopenia. This can provide easy and rapid approaches for health evaluations and expand the use of AC and CC in this field.

The objectives of this study were 1) to explore the screening ability and cut-off values of CC and AC on sarcopenia in Chinese community-dwelling older adults and 2) to explore the evaluation effects of AC and CC on health indicators.

Materials and Methods

Overall Trial Design

This is a cross-sectional project, including sub-study 1 being diagnostic research and sub-study 2 correlational research. Firstly, we included a larger population to explore the screening abilities and cut-off values of AC and CC on sarcopenia diagnosis (sarcopenia-AC and sarcopenia-CC) in Chinese community-dwelling older adults (Sub-study 1). Thereafter, a portion of participants were invited to have additional health assessments to explore the differential effects of sarcopenia-AC and sarcopenia-CC on health indicators (Sub-study 2). Sarcopenia-CC or sarcopenia-AC was defined if the CC or AC of participant was lower than the cut-off value of CC or AC on sarcopenia obtained from Sub-study 1.

Sub-Study 1

Participants and Groups

This cross-sectional sub-study was conducted from November 2019 to October 2020. A total of 1537 participants were included. The participants were aged 60 years or older, able to stand and willing to participate in this sub-study. The exclusion criteria were 1) unable to have BIA measurement due to any reason, including but not limited to the presence of cardiac stents, pacemakers, steel plates, steel nails within the body; 2) unable to communicate; 3) having clinically visible edema.

The participants were randomly divided into the experimental group (EG) or the validation group (VG) at the ratio of 7:3. To avoid data contamination, the characteristics and correlation between muscle indicators and CC or AC were analyzed from the data in the EG.

Sample Sizes

The sensitivity (S_N) of CC screening for sarcopenia in the study of Kawakami et al¹⁷ was referred to calculate the sample size of participants in the EG. The S_N was 88% in males and 76% in females. The permissible Error (d) was 10%, and statistical significance (α) was 0.05. According to the following equation,²⁵ the required sample size for the EG was 239 males and 413 females.

Based on the ratio of 7:3 between the EG and the VG, the required sample size for the VG was 103 males and 177 females. Considering an attribution rate of 20% (eg dropping out during the tests), the total required sample size was 428 males and 738 females.

Measurements

In addition to the demographic data, body composition was measured by bioelectrical impedance analysis (BIA) (Tanita, MC-180, Japan).²⁶ The included indicators were body mass index (BMI), fat mass, body fat percentage (BFP), fat-free mass (FFM), appendix skeletal muscle mass (ASM) and SMI, which was calculated by ASM (kg)/Height² (m²).

Handgrip strength was assessed by a hydraulic dynamometer (Jamar, 563213, America).²⁷ After resting for 5 minutes, the subjects sat in front of the instrument with their torso upright, shoulder joint in a neutral position, and elbow joint at 90 degrees, with the forearm and hip joint in a neutral position. Each hand was tested three times with a 30-second interval between tests and a 1-minute rest between sets. If the coefficient of variation (CV%) across the three sets exceeded 10%, the test was repeated after a 20-minute rest. In this study, the maximum handgrip strength of the dominant hand was adopted.

The 6-m gait was measured by instructing participants to walk a straight line of 10 meters at their usual gait speed, and the time to walk the middle 6 meters was recorded to minimize the bias. CC and AC were both measured by inelastic but flexible measuring tapes without compressing the skin. CC measurement was performed at the largest part of the calf with participants in the seated position and the calf at 90° to the thigh. AC measurement was performed at the midpoint of the acromion and olecranon. CC and AC were both measured twice, and the average results were used.

Diagnosis of Sarcopenia

The AWGS 2019 criteria¹⁶ was the gold standard for sarcopenia as our participants were Chinese older adults. Specifically, sarcopenia was defined as the presence of low SMI (male: <7.0 kg/m², female: <5.7 kg/m²) and low handgrip strength (male: <28 kg, female: <18 kg) or/and low gait speed (<1.0/s).

Sub-Study 2

Participants

Due to better screening ability of CC on sarcopenia than AC, participants with and without sarcopenia-CC were included in Sub-study 2 at the ratio of 1:1. The exclusion criteria were having severe osteoarthritis or surgeries or other conditions that cannot perform the physical function tests, besides the exclusion criteria of Sub-study 1.

Sample Size

With the aim to evaluate the differences between participants with and without sarcopenia-CC or sarcopenia-AC on health indicators, the following equation of comparison between two sample sizes was used in Sub-study 2:

Handgrip strength was the primary indicator in Sub-study 2 and the specific values were referred to Sub-study 1. In males, the handgrip strengths were 32.23 kg and 25.07 kg for older adults without and with sarcopenia, respectively. Thus, δ = 7.16, σ = 6.41 and the sample size for males was 24 per group considering 20% drop-out rate. In females, the handgrip strengths were 20.30 kg and 16.96 kg for older adults without and with sarcopenia, respectively. Thus, δ = 3.34, σ = 4.41 and the sample size for females was 48 per group considering 20% drop-out rate.

Measurements

Regarding demographic data, the following aspects were added to the demographic questionnaire of Sub-study 1: living alone, sleep quality, exercise habit, walking with assistant, clinical treatment, weight loss, history of hypertension, diabetes, hyperlipidemia, cardiovascular disease, and sedentary hours. Additionally, nutritional status was assessed using the mini nutritional assessment (MNA).²⁸

Based on the definition and diagnostic criteria of sarcopenia, fat and muscle mass, muscle strength, and physical function were assessed. Besides the Measurements in Sub-study 1, the additional indicators and measurements were as follows.

Senior fitness tests (SFTs) including the back scratch test (BST), the chair sit and reach test (CSRT), the biceps curl test (BCT), time-up and go test (TUGT), 30-second chair stand test (30 CST), 6-Minute Walk Test (6MWT) were tested based on the standard procedures recommended by Rikli and Jones.²⁹

Digital muscle strengths, including hip flexion strength (HFS), knee extension strength (KES), knee flexion strength (KFS), ankle dorsiflexion strength (ADS), elbow flexion strength (EFS) and elbow extension strength (EES), were assessed by MicroFET3 muscle strength tester (FET3, Hoggan, USA). The detailed information about limb position, MicroFET3 placement and required movements for each group of muscle testing are presented in Table S1.^30,31 Before the test, participants were instructed to practice each test with mild effort 1–2 times. During the test, participants were instructed to perform the corresponding movement for each test quickly and with maximum effort. Each test lasted for 3–5 seconds and ended when the dynamometer reading stabilizes. Each group of muscles were tested twice, and the maximum value was recorded.

Statistical Analysis

T-test or Mann-Whitney U-test was used for normal or skewedly distributed continuous data, respectively and or Fisher test was used for categorical data. The correlation between muscle indicators and CC or AC was analyzed by Pearson (for normal distributed data) or Spearman analysis (for skewedly distributed data).

Receiver operating characteristic (ROC) analysis was used to explore the screening ability and cut-off values of CC and AC on sarcopenia. The indicators included sensitivity , specificity , AUC with 95% confidence interval, and Youden index (YI). AUC values of 0.5–0.7, 0.7–0.8, 0.8–0.9 and >0.9 were considered presenting low, acceptable, excellent and outstanding accuracy of screening ability.³² The difference between AUCs was compared by test. The cut-off values of CC and AC for screening sarcopenia were determined by YI. Positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), and negative likelihood ratio (LR-) were also calculated. Statistical analyses were performed with SPSS 23.0 (IBM, Chicago) and Stata 16.0 (Stata Corp, College Station, USA). Two-sided P values <0.05 were considered statistically significant.

Results

Sub-Study 1

Characteristic of Participants

A total of 1537 participants were included with 690 males (44.89%) and 847 females (55.11%) after screening 1681 community-dwelling older people. The prevalence of sarcopenia was 11.58% (males: 8.41%; females: 14.17%, P<0.001). When compared with non-sarcopenic older adults, sarcopenic older adults were older, had less muscle mass (eg ASM, SMI) and fat mass (eg total fat mass, BFP), lower gait speed, lower handgrip strength, smaller AC and CC (P<0.05, Table 1).

Table 1 Characteristics of Participants Using the AWGS 2019 Sarcopenia Criteria

Screening Ability and Cut-off Values of AC and CC on Sarcopenia

Shown in Table 2, the AUCs of AC and CC on screening of sarcopenia were statistically significant in males and females and all greater than 0.7 (0.7–0.9, P<0.05). In males, there was no significant difference between the AUCs of AC and the AUCs of CC Figure 1A. In females, the AUC of CC was greater than that of AC Figure 1B. The cut-off values of sarcopenia-CC were 33.7 cm in males and 33.0 cm in females . The cut-off values of sarcopenia-AC were 25.9 cm in males and 26.5 cm in females Table 2. The PPVs of AC and CC on screening of sarcopenia ranged from 0.283 to 0.343, while the NPVs were all greater than 0.9. The LR+s of AC and CC were greater than 2.0 and the LR- were less than 0.5. The conformity rate of sarcopenia-AC and sarcopenia-CC were all greater than 60% and the Kappa values ranged from 0.241 to 0.416 (P<0.001, Table S2).

Table 2 Screening Ability and Cut-off Values of AC and CC on Sarcopenia in the Experimental Group

Figure 1 ROC Curves of AC and CC on the Screening of Sarcopenia in the Older People (A) Male; (B)Female. The areas under the green and blue curves presented the screening abilities of AC and CC screening for sarcopenia in older males, respectively. The yellow line was the reference line and the area under the yellow line was 0.5. The green and blue line were above the yellow line, which mean that areas under the curves of AC and CC in older males and females were greater than 0.5. Abbreviations: AC, arm circumference; CC, calf circumference; ROC, receiver operating characteristic.

The Results of the VG

A total of 461 participants aged 65–90 years (72.30 ± 5.44) with 44.90% (207 males) were included in the VG. The prevalence of sarcopenia-AC and sarcopenia-CC were 17.87% and 31.88% in males, respectively, with the conformity rate with the AWGS 2019 all greater than 60%. In females, the prevalence of sarcopenia-AC and sarcopenia-CC were 35.43% and 42.52%, respectively, with the conformity rate with the AWGS 2019 all greater than 60%. The kappa values in male and female participants were statistically significant (P<0.05, Table S3).

Sub-Study 2

Demographics

A total of 269 participants were included in Sub-study 2. In males, participants with sarcopenia-CC were older, had higher rate of living alone, higher prevalence of hypertension and lower score of MNA, compared to participants with non-sarcopenia-CC (P<0.05). In females, participants with sarcopenia-CC had a higher prevalence rate of hypertension and a lower score of MNA (P<0.05, Table 3).

Table 3 Demographic Data of Participant in Sub-Study 2

The Evaluation Effects of Sarcopenia-AC and Sarcopenia-CC on Muscle and Fat Mass

As shown in Table 4, participants with sarcopenia-AC or sarcopenia-CC had lower fat and muscle mass indicated by all of 11 indicators included in Sub-study 2 (eg BMI, total muscle mass), compared to participants without sarcopenia-AC or sarcopenia-CC, respectively (P<0.05). The differentiation effects of sarcopenia-AC and sarcopenia-CC on fat and muscle mass were all consistent with the assessments of sarcopenia-AWGS.

Table 4 The Evaluation Effects of Sarcopenia-AC and Sarcopenia-CC on Fat and Muscle Masses

The Evaluation Effects of Sarcopenia-AC and Sarcopenia-CC on Muscle Strength

There were 9 indicators of muscle strength (eg handgrip strength, BCT, HFS), Table 5. Based on the sarcopenia criteria of AC and CC, the muscle strength of sarcopenic older people was significantly lower than their counterparts without sarcopenia indicated by handgrip strength, BCT, EFS, EES, ADS, HFS, KES, and KFS (P<0.05), in both sexes. However, 30 CST showed no difference in either males (P=0.123 and 0.805, respectively) or females (P=0.913 and 0.681, respectively), details in Table 5. The evaluation effects of sarcopenia-AC and sarcopenia-CC on muscle strength were consistent with the assessments of sarcopenia-AWGS in both males and females, except for the evaluation of 30 CST in females. Specifically, female participants with sarcopenia-AC or CC had similar Results on 30 CST to participants without sarcopenia-AC or CC (P=0.913 and 0.681, respectively), which was different from the results that female participants with sarcopenia-AWGS had significantly less 30 CST than those without sarcopenia-AWGS (P < 0.05), details in Table 5.

Table 5 The Evaluation Effects of Sarcopenia-AC and Sarcopenia-CC on Muscle Strength

The Evaluation Effects of Sarcopenia-AC and Sarcopenia-CC on Physical Function

A total of five indicators of physical function were included in Sub-study 2, Table 6. In males, older adults with sarcopenia-AC had a longer time cost of TUGT and less meters of 6MWT than those without sarcopenia-AC (P < 0.05). In addition to the similar results of TUGT and 6MWT, older males with sarcopenia-CC had greater gait speed than those without sarcopenia-CC (P<0.05). The evaluation effects of sarcopenia-CC on physical functions were consistent with the assessments of sarcopenia-AWGS in male participants, while not the evaluation of gait speed based on sarcopenia-AC. In females, none of the five indicators of physical function had statistically significant (P>0.05) based on either sarcopenia-AC or sarcopenia-CC, which was different from the evaluation on CSRT and 6MWT based on sarcopenia-AWGS. The female participants with sarcopenia-AWGS had longer CSRT and less distance of 6MWT than those without sarcopenia-AWGS (P<0.05).

Table 6 The Evaluation Effects of Sarcopenia-AC and Sarcopenia-CC on Physical Function

Discussion

This study provided the accurate and Chinese population targeted cut-off values of AC and CC on screening for sarcopenia (25.9 cm and 33.7 cm in males; 26.5 cm and 33.0 cm in females, respectively). In addition, we found that AC and CC were able to distinguish the differences in muscle mass, fat mass, and physical function between males with and without sarcopenia-AC or sarcopenia-CC, but not the physical function of females. Our study provides efficient and population-tailored evidence on the use of AC and CC in the screening and health evaluation of sarcopenia in Chinese community-dwelling older people.

This study has demonstrated that AC and CC both have excellent screening abilities on sarcopenia (AUC≥0.7) in Chinese community-dwelling older people, which has been further confirmed by the data in the VG with the conformity rates between sarcopenia-AC/CC and sarcopenia-AWGS all greater than 60%. AC and CC are noninvasive, inexpensive and easily assessed,³³ which are feasible to use in facilities and community health service centers, especially for countries with large population and limited medical resource in primary or community hospitals, like China. Similarly, Esteves et al²¹ investigating Brazilian community of older people and Nishikawa et al²² including Japanese patients with chronic liver disease, have found that AC has a good screening ability on sarcopenia with AUCs greater than 0.7 in both sexes. However, the cut-off values of AC screening for sarcopenia are different from ours. These studies have indicated that the cut-off values were 27 cm both for males and females,^21,22 which is higher than our study (25.9 cm for males, 26.5 cm for females). Currently, there are few studies related to AC in sarcopenia screening, and to our best knowledge, no studies have been found in the Chinese community-dwelling older adults. However, one study has found that the cut-off value of AC for the diagnosis of malnutrition is 24 cm in Chinese inpatients.²⁴ Malnutrition is closely associated with the development of sarcopenia,² which indirectly proves the value of the application of AC in the screening of sarcopenia. However, more studies are needed to find the optimal cut-off values of AC on sarcopenia screening, especially in different regions.

Similar to this study, these studies investigating community older adults in Indonesia³⁴ and Brazil,²¹ older adults with walking dysfunction in Taiwan³⁵ and middle-aged and older Japanese men and women¹⁷ have found that CC has good screening ability for sarcopenia with AUCs greater than 0.7. However, the optimal cut-off values of CC screening for sarcopenia vary in different populations, with a greater variation in cut-off values in women (29–33 cm) and a smaller variation in men (33–34 cm). The different cut-off values further demonstrated that the use of CC screening for sarcopenia interferes with populations and regions^14,15 and it is necessary to explore the accurate and Chinese population-targeted cut-off values due to limited studies in Chinese community-dwelling older people. We also explored the screening abilities of multiple factors (ie BMI, age, gait speed, handgrip strength) and found that the AUCs of multiple factors were similar to those of CC in both sexes while higher than that of AC in females, details in Table S4.

Sex differences in the screening abilities and variations in the cut-off values of AC and CC for sarcopenia may be related to higher adiposity (especially the subcutaneous fat) and greater individual variability in women compared to men.³⁶ This further underscores the need to determine optimal cut-off values for different populations, especially for females. Regarding the better screening ability of CC compared to AC in females, the reason may be that the screening ability of AC for sarcopenia is more likely influenced by subcutaneous fat than CC, as the muscle mass of the upper arm is less than that of the calf.³⁷

Based on the results of Sub-study 1, which have shown that AC and CC have excellent screening abilities on sarcopenia, we further explored the evaluation effects of sarcopenia-AC and sarcopenia-CC on health outcomes in Chinese community-dwelling older people. We have found that sarcopenia-AC and sarcopenia-CC are able to distinguish the differences in fat mass, muscle mass and physical function between males with and without sarcopenia but not physical function of females. In line with our study, studies also have found that physical function was significantly associated with sarcopenia-CC and sarcopenia-AWGS in males but not in females.^33,38 One explanation is that physical function of females may be more related to fat mass (especially the intramuscular fat mass)³⁹ and muscle quality,⁴⁰ instead of muscle mass. For example, one study³⁹ has reported that physical function of females was significantly associated with intramuscular fat, whereas in males, it was associated with muscle mass. These results suggest that we should pay special attention to obesity or fat mass in older females to prevent the further decline of their physical function. Besides, we also found that limited studies evaluate the body flexibility, which is crucial to maintain the body postures and mobility and are associated with muscle mass and muscle endurance.⁴¹ A comprehensive assessment of physical function, including flexibility, in patients with sarcopenia is important to maintain their holistic physical health.

Regarding muscle strength, we have found that sarcopenia-AC and sarcopenia-CC had good differentiation effects on muscle strength indicators except for the number of 30 CST in both males and females. The reason may be that although the 30 CST is a simple method for assessing muscle strength, study has shown that its test results can be influenced by a variety of factors, such as co-morbidities, cognitive impairment, and physical inactivity.⁴² The relationship between these factors and sarcopenia is not clear, thus leading to the poor assessment of sarcopenia-AC and sarcopenia-CC on 30 CST in this population.

Limitation

First, although the relatively large number of participants, the specific targeted population and the “gold standard” guarantee the accuracy of results of this study, we did not explore the effect of obesity on the screening abilities and cut-off values of AC and CC on sarcopenia as the prevalence of obesity (BMI ≥ 28.0kg/m², 8.72%) was relatively low in our study. Second, despite that the validation in the VG ensures the reliability and robustness of results of this study, the generalization of the cut-off values needs to be validated in other regions, populations or ethnic groups. Third, our study is a cross-sectional study and the predictive effects of sarcopenia-AC and sarcopenia-CC on health outcomes still need to be explored. Lastly, as older adults with edema were excluded from our study, the usage of AC and CC on the screening of sarcopenia in such population needs further explanation.

Conclusions

Our study has provided the accurate and population-targeted cut-off values of AC and CC screening for sarcopenia in Chinese community-dwelling older people (25.9 cm and 33.7 cm in males; 26.5 cm and 33.0 cm in females) and found that sarcopenia-AC and sarcopenia-CC have good evaluation effects on health assessments in older males. The results of this study provide clinical professions, especially these in Chinese primary or community hospitals, the precise and tailor-made tools for screening of sarcopenia and simple approaches to health assessments of sarcopenia on older adults of the community.