Prevalence and associated factors of sarcopenia and severe sarcopenia in older Taiwanese living in rural community: The Tianliao Old People study 04




The aim of the present study was to show the prevalence and associated factors of sarcopenia and severe sarcopenia in rural community-dwelling older Taiwanese.


Using the whole community sampling method, a total of 285 men and 264 women aged over 65 years were randomly sampled (response rate = 50%) from Tianliao District, southern Taiwan, in 2012. Participants were interviewed by trained investigators to complete a validated structural questionnaire. Body composition was measured by bioelectrical impedance analysis, and skeletal muscle mass was estimated by Janssen's equation. The Mini-Nutritional Assessment (MNA) score, Short Portable Mental Status Questionnaire, grip strength, gait speed and short physical performance battery (SPPB) were obtained by the standard procedures. Sarcopenia and severe sarcopenia were defined according to the 2010 consensus of the Report of the European Working Group on Sarcopenia in Older People.


Of the 549 study participants, 39 (7.1%) were classified as having sarcopenia and 31 (5.6%) participants were classified as having severe sarcopenia. Using multiple logistic regression models, the age, sex, working status, waist circumference, body mass index, hypertensive history, MNA and SPPB score were independently associated with different stages of sarcopenia.


Approximately one-fifth of community-dwelling older adults were facing the threat of sarcopenia in southern Taiwan. The older age, female sex, lower body mass index, higher waist circumference, a history of hypertension, lower MNA or SPPB score and not working regularly were associated factors for either sarcopenia or severe sarcopenia. Geriatr Gerontol Int 2014; 14 (Suppl. 1): 69–75.


Sarcopenia is named to reflect the status of decreased skeletal muscle mass (SMM) in aging people.[1] As the SMM is one of the major human body components responsible for daily activities, sarcopenia has an influence on self-independence in activities of daily living, such as eating, taking a shower and walking,[2-5] and results in adverse outcomes, including falls, hip fractures, comorbidities[6-8] and mortality.[9, 10] Therefore, the importance of sarcopenia is increasingly emphasized in the aging society worldwide.[6, 11, 12]

It is believed that sarcopenia is an economic burden for both caregivers and the healthcare system in an aging society.[12] However, there are still limited studies or inconsistent prevalence of sarcopenia in Chinese populations.[3, 5, 9, 13-18] In contrast, most of the reports were collected from subjects either living in nursing homes,[15] hospital-based[3, 18] or as volunteers.[13, 14] Few studies focused on the systemic sampling of older adults living in metropolitan areas.[9, 16, 17]

Bioelectrical impedance analysis (BIA) is non-radiating, relatively cheap and portable, making it possible to measure body composition in rural communities. Many studies have shown that BIA-derived skeletal muscle mass index (SMI) can be used to define sarcopenia,[3, 13, 19, 20] it is plausible to use BIA rather than dual energy X-ray absorptiometry (DXA) to measure the SMM in a community survey.

In 2010, the consensus of the Report of the European Working Group on Sarcopenia in Older People (EWGSOP) redefined the criterion for sarcopenia in which not only the muscle mass is taken into consideration, but muscle strength and muscle performance are also considered for the definition of sarcopenia.[21] Therefore, combined with muscle function and physical performance, the level of decreased SMM was newly categorized as presarcopenia, sarcopenia and severe sarcopenia.[21] The aging population is endemic, especially in rural townships. However, the healthcare facilities are relatively inadequate when compared with that in metropolitan areas. Therefore, to obtain the status of sarcopenia in rural communities is urgent and important for public health policy. To the best of our knowledge, nearly none of the systemic sampling surveys of sarcopenia have focused on the rural communities in China or Taiwan. Furthermore, none of the studies have reported the different stages of sarcopenia and the associated risk factors concomitantly.

The aims of the present study were to determine the prevalence of sarcopenia and severe sarcopenia of older adults living in rural communities in southern Taiwan. The associated factors of different stages of sarcopenia were also evaluated accordingly. Through the comprehensive evaluation of sarcopenia in rural communities, intervention and preventive strategy can be applied appropriately.


In July 2012, a cross-sectional survey following the Tianliao Old People (TOP) study[22, 23] was carried out in Tianliao District, Kaohsiung City in southern Taiwan. In the 2012 census report, in the total population of 7800, 1966 subjects (25.2% of the total population) were aged 65 years and over. After excluding empty houses (n = 489), death (n = 40), non-ambulatory subjects (subjects with significant disabilities, such as handicapped cerebral vascular accident, cancerous cachexia, unstable chronic diseases or psychiatric disorders, severe arthritis or inflammatory disease, uncomfortable anorexia by any medications, n = 138) and non-reachable subjects (n = 201), a total of 285 men and 264 women, aged 65–102 years, were enrolled (response rate: 549/1098 = 50%) by the whole district sampling method.[24] The sex distribution of the 549 respondents was not statistically different from the 1417 non-responders (P = 0.132). However, the mean age was relatively younger for respondents (76.0 ± 6.2 vs 76.8 ± 7.4 years, P = 0.001). For convenience and accessibility, all the participants received the survey in five locally designated examination centers evenly distributed in the Tianliao District. This study was approved by the institute review board of National Cheng Kung University Hospital (IRB number: B-ER-101-119). Each participant signed an informed consent before examination.

Participants were interviewed to complete a validated structural questionnaire,[22, 23] which included basic characteristics, smoking and drinking habits, working status, and medical history. We defined participants to have a smoking habit if they had smoked more than 100 cigarettes and still smoked one pack (20 cigarettes) at least per month for more than 6 months, and alcohol drinking was defined by if participants still drank one time per week for more than 6 months.[23] A history of hypertension or diabetes was assessed by referring to the self-reported physician's diagnosis. Participants who still worked at a farm or had regular work were defined as working regularly. The long form Mini-Nutritional Assessment (MNA) was used as a screening tool to evaluate nutritional status.[25, 26] Cognitive function was evaluated by the wrong answer of the 10-item Short Portable Mental Status Questionnaire (SPMSQ).[27]

Anthropometry and body composition

Bodyweight and standing height were measured by DETECTO (Detecto, Webb City, MO, USA), with participants dressed in light clothing and barefoot. Body mass index (BMI) was calculated by bodyweight divided by square of height (kg/m2). The waist circumference (WC) was measured (Gulick II; Gays Mills, WI, USA; to the nearest mm) midway between the lateral lower rib margin and the superior anterior iliac crest at the end of a gentle expiration phase.[22] A single frequency 8-electrode bioelectrical impedance analysis (BIA) device (BC-418; Tanita, Tokyo, Japan) was used to measure body composition, which had been validated to measure the SMM.[28] SMM (kg) was estimated using the Janssen's equation:[20] SMM = ([Ht2 / R × 0.401] + [sex × 3.825) + (age × −0.071]) + 5.102, where height is in cm; resistance is in ohms; for sex, men = 1 and women = 0; and age is in years. SMM is divided by square of height to obtain SMI (kg/m2) for defining low SMM and sarcopenia.[2, 3, 21]

Physical performance

Functional limitations were assessed by using the short physical performance battery (SPPB).[29] A higher summary performance score represents a better performance, and vice versa.[30] In the present study, the SPPB score was dichotomized as <9 or ≧9 for statistical analysis.[29, 30] Two tests (30 s separately) of grip strength (Grip-D; TKK 5401, Tokyo, Japan) of the bilateral hands were obtained for each participant. The maximal value of grip strength was used for the diagnosis of low muscle function according to the corresponding cut-offs by BMI and sex.[21] In short, the corresponding cut-offs of low grip strength for men with BMI ≤24 kg/m2, 24.1∼28 kg/m2, and >28 kg/m2 were ≤29 kg, ≤ 30 kg, and ≤32 kg, respectively; for women with BMI ≤23 kg/m2, 23.1∼26 kg/m2, 26.1∼29 kg/m2, and >29 kg/m2 were ≤17 kg, ≤17.3 kg, ≤18 kg, and ≤21 kg, respectively.[21] The gait speed was assessed by a 15-ft walking test. Low physical performance was determined when the gait speed was ≤0.8 m/s.[3]

Definition of non-sarcopenia, sarcopenia and severe sarcopenia

In the present study, we took the same young reference values that have been validated by the National Health Research Institute in Taiwan.[3] In short, 498 healthy males and 500 females aged 20–40 years were recruited for body composition assessment using the same BIA (BC-418) with segmental measures for developing the reference SMI.[3] With the definition of low muscle mass set at two standard deviations below the mean value of SMI in the young reference groups, the cut-off points for men and women were 7.70 and 5.67 kg/m2, respectively.[3] Subjects were defined according to the 2010 consensus of Report of the European Working Group on Sarcopenia in Older People.[21] Those who had SMI higher than the cut-offs[3] were classified as normal. Low SMM (also named as presarcopenia) was defined as participants with SMI under two standard deviations of the young reference of the same ancestry from 18 to 40 years-of-age.[21] Sarcopenia was defined as participants with low SMI and either low muscle function (reflected by grip strength) or low physical performance (reflected by walking speed). Severe sarcopenia was defined when the aforementioned three conditions were present concomitantly.[21] Normal and presarcopenia were reclassified as non-sarcopenia for statistical analysis.

Statistical analysis

Statistical analysis was carried out by using the Statistical Package of Social Science for Windows software Version 17 (SPSSWIN, version 17.0; Chicago, IL, USA). Continuous and categorical variables were expressed as means ± SD and percentages, respectively. The comparisons between groups in regard to categorical variables were analyzed using the χ2-test, and continuous variables were analyzed using one-way ANOVA. Multiple logistic regression models were used to evaluate the independently associated factors between different stages of sarcopenia. Statistical significance was defined as P < 0.05 for two-tailed analysis.


All the basic characteristics are shown in Table 1. Of the 549 subjects, including 285 males and 264 females, 101 (18.4%) subjects with 35 (12.3%) men and 66 (25.0%) women had SMI less than 7.70 and 5.67 kg/m2, respectively.[3] The prevalence of sarcopenia and severe sarcopenia were 7.1% (n = 39) and 5.6% (n = 31), respectively. Women had a higher prevalence of sarcopenia and severe sarcopenia than men (Table 2). Non-sarcopenic participants were younger, more likely to be working regularly, had higher BMI, WC, SMI, MNA and SPPB score, but lower gait speed and SPMSQ score than sarcopenic and severe sarcopenic participants of both sexes. No statistical difference of demographic characteristics could be found between sarcopenia and severe sarcopenia.

Table 1. Demographic characteristics of 549 older Taiwanese adults living in a rural community
 TotalNormalPresarcopeniaNon-sarcopeniaSarcopeniaSevere sarcopeniaP-value*
  1. *Comparison between non-sarcopenia, sarcopenia and severe sarcopenia. Continuous variables: one-way ANOVA. Categorical variables: χ2-test. Comparison between non-sarcopenia and sarcopenia: P < 0.001. Comparison between non-sarcopenia and severe sarcopenia: P < 0.001. Comparison between sarcopenia and severe sarcopenia: non-significance. Non-sarcopenia: combined with normal and presarcopenia. SPMSQ, short portable mental status questionnaire; SPPB, short physical performance battery.
Age (years)76.0 ± 6.275.2 ± 5.976.4 ± 5.975.3 ± 5.980.4 ± 5.981.8 ± 5.7<0.001
Sex (male/female)285/264250/19812/19262/21711/2812/190.002
Height (cm)154.6 ± 8.4155.2 ± 8.2156.1 ± 7.6155.3 ± 8.2150.8 ± 8.4149.4 ± 8.8<0.001
Weight (kg)58.8 ± 10.961.1 ± 10.252.7 ± 7.760.5 ± 10.348.7 ± 7.545.5 ± 7.2<0.001
Body mass index (kg/m2)24.6 ± 4.125.4 ± 3.921.6 ± 2.325.1 ± 4.021.5 ± 3.020.4 ± 2.4<0.001
Waist circumference (cm)87.0 ± 10.388.8 ± 9.980.8 ± 7.188.3 ± 9.978.9 ± 9.278.3 ± 8.5<0.001
Skeletal muscle mass (kg)18.1 ± 4.919.0 ± 4.715.1 ± 3.918.8 ± 4.713.3 ± 3.113.5 ± 3.8<0.001
Skeletal muscle mass index (kg/m2)7.4 ± 1.57.8 ± 1.46.1 ± 1.27.7 ± 1.45.8 ± 0.95.9 ± 1.1<0.001
Working regularly (%)44.048.338.747.718.419.4<0.001
Habitual smoking (%)26.228.319.427.710.522.60.062
Alcohol drinking (%)16.517.912.917.613.23.20.095
History of hypertension (%)49.551.322.649.555.341.90.546
History of diabetes (%)17.217.916.117.813.212.90.618
Mini-Nutritional Assessment score25.9 ± 2.626.2 ± 2.325.6 ± 2.726.2 ± 2.323.4 ± 3.424.2 ± 2.6<0.001
SPMSQ score1.9 ± 1.91.6 ± 1.52.9 ± 2.21.7 ± 1.83.0 ± 2.53.0 ± 2.4<0.001
Gait speed (m/s)0.92 ± 0.290.94 ± 0.291.10 ± 0.180.95 ± 0.280.82 ± 0.220.59 ± 0.14<0.001
SPPB score ≧9 (%)72.276.493.577.556.89.7<0.001
Table 2. Prevalence of different stages of sarcopenia in 549 older Taiwanese adults living in a rural community
  1. *Comparison between sex, χ2-test.
Case no.549285264
Average age (years)76.0 ± 6.276.2 ± 6.575.9 ± 5.80.633
Non-sarcopenia479 (87.3%)262 (91.9%)217 (82.2%)<0.001
Normal448 (81.7%)250 (87.7%)198 (75.0%)<0.001
Presarcopenia31 (5.6%)12 (4.2%)19 (7.2%)0.002
Sarcopenia39 (7.1%)11 (3.9%)28 (10.6%)<0.001
Severe sarcopenia31 (5.6%)12 (4.2%)19 (7.2%)0.002

Using multiple logistic regression models, the associated factors for different stages of sarcopenia are shown in Table 3. Age was a positive independent factor for all stages of sarcopenia. BMI was a negative independent factor for all stages of sarcopenia. Sex was an independent factor in models I, II and III. WC was shown to be an associated factor for all stages of sarcopenia, but was only significant in model I. A lack of regular work was an independent factor in model I. A History of hypertension and low MNA score were also the independent factors in model III. Low SPPB score was also an independent factor in models II and IV. Other variables including WC, habitual smoking, drinking alcohol, history of diabetes and SPMSQ score showed no statistic significance in all different stages of sarcopenia.

Table 3. Multiple logistic regression models of the risk factors for sarcopenia and severe sarcopenia in 549 older Taiwanese adults living in rural community
VariablesModel IModel IIModel IIIModel IV
(n = 549)(n = 518)(n = 518)(n = 510)
Odds ratio (95% CI)Odds ratio (95% CI)Odds ratio (95% CI)Odds ratio (95% CI)
  1. Model I: normal (n = 448) versus participants with low skeletal muscle mass (n = 101). Model II: non-sarcopenia (n = 479) versus sarcopenia and severe sarcopenia (n = 70). Model III: non-sarcopenia (n = 479) versus sarcopenia (n = 39). Model IV: non-sarcopenia (n = 479) versus severe sarcopenia (n = 31). *P < 0.05, **P < 0.01, ***P < 0.001. SPMSQ, short portable mental status questionnaire; SPPB, short physical performance battery.
Nagelkerke R2 value0.4750.4900.3870.563
Age (years)1.104 (1.045∼1.166)***1.118 (1.049∼1.192)***1.113 (1.030∼1.202)**1.170 (1.054∼1.300)**
Waist circumference (cm)1.087 (1.018∼1.161)*1.027 (0.953∼1.107)1.000 (0.991∼1.098)1.042 (0.929∼1.169)
Sex (male = 1, female = 0)0.187 (0.083∼0.425)***0.305 (0.118∼0.791)*0.255 (0.079∼0.828)*0.348 (0.082∼1.485)
Body mass index (kg/m2)0.498 (0.399∼0.620)***0.595 (0.469∼0.755)***0.706 (0.531∼0.939)*0.467 (0.319∼0.684)***
Working regularly (yes = 1, no = 0)0.494 (0.263∼0.928)*0.509 (0.231∼1.120)0.467 (0.174∼1.251)0.662 (0.201∼2.186)
Habitual smoking (yes = 1, no = 0)0.867 (0.339∼2.219)0.856 (0.277∼2.645)0.386 (0.075∼1.994)1.740 (0.346∼8.746)
Alcohol drinking (yes = 1, no = 0)1.059 (0.393∼2.855)0.961 (0.288∼3.205)2.918 (0.641∼13.284)0.195 (0.018∼2.175)
Mini-Nutritional Assessment score0.990 (0.884∼1.107)0.937 (0.825∼1.063)0.857 (0.744∼0.988)*1.125 (0.900∼1.407)
History of hypertension (yes = 1, no = 0)0.733 (0.401∼1.329)1.464 (0.720∼2.976)2.454 (1.005∼5.995)*1.016 (0.328∼3.147)
History of diabetes (yes = 1, no = 0)0.680 (0.310∼1.492)0.661 (0.260∼1.681)0.537 (0.160∼1.808)1.284 (0.324∼1.413)
SPMSQ Score0.996 (0.848∼1.170)1.019 (0.852∼1.219)0.998 (0.794∼1.255)1.089 (0.839∼1.413)
SPPB score (≧9)0.996 (0.868∼1.143)0.842 (0.721∼0.983)*0.996 (0.826∼1.202)0.707 (0.561∼0.892)**


Compared with studies carried out in Taiwan[3, 13, 16, 17] and other countries,[6, 9, 14, 15, 18, 31, 32] the prevalence of sarcopenia and severe sarcopenia in Tianliao, Kaohsiung, is consistently within the range of 5–30%. Interestingly, in the present study, the prevalence of sarcopenia and severe sarcopenia for women is notably higher than that of men. Although men are supposed to have a higher prevalence of sarcopenia than women because of the higher rate of muscular atrophy than women,[11] the prevalence might be inverted while using different diagnostic criteria of sarcopenia.[16, 33] As the present study was carried out in a farming village, most of the male participants (50.2%) were still working more regularly than the female participants (37.1%). Furthermore, gait speed was higher in men than women (1.00 ± 0.28 m/s vs 0.83 ± 0.26 m/s), and grip strength was higher in men than women (33.0 ± 7.82 kg vs 20.4 ± 5.20 kg) in the present study (data not shown). Therefore, women might have a higher prevalence of sarcopenia than men,[16] but this paradox needs to be reconfirmed by further study.

Consistent with previous reports, age was the major determinant of sarcopenia[17, 34] and physical performance[35] in the present study. Using the BIA-derived SMM,[3, 13] the prevalence of sarcopenia defined by the EWGSOP definition[21] was found to be substantially increased with age in both men and women.[3] A higher BMI indicates better nutrition intake[24] and relatively higher SMM.[11] Therefore, BMI is negatively correlated with the prevalence of sarcopenia,[6-8, 15, 16, 18] which is consistent with our findings. The fact that the MNA was not independently associated with sarcopenia after it was adjusted with BMI might reflect the superiority of BMI in relation to sarcopenia. Interestingly, the WC showed a positive trend of low SMM after adjusting the BMI and major variables in multiple logistic model I. From the viewpoint of body composition, central obesity might reflect the status of relatively higher bone mass and abdominal fat accumulation, but lower fat free mass.[22] Central obesity is a well-known risk factor of cardiovascular disease and metabolic aberrations. Recently, SMI has been negatively correlated with insulin resistance or C-reactive protein, and might share the pathophysiological mechanism with non-alcoholic fatty liver disease.[36] Sarcopenia and WC could potentiate each other to induce hypertension.[37] In summary, the association between low SMM and lower BMI or higher WC is compatible with the concept of sarcopenic obesity.[33, 38] The interrelationships between sarcopenia, sarcopenic obesity, hypertension and metabolic syndrome have been emphasized,[37, 39] and warranted further evaluation.

Studies have suggested that adequate physical activity might reduce the risk of sarcopenia.[34, 40] In the present study, the median and mean weekly calorie expenditure (International Physical Activity Questionnaires) were 4587.8 kcal and 7914.2 kcal, respectively (data not shown). However, participants who worked regularly or had a higher SPPB score showed a lower risk of sarcopenia in the present study. Grip strength and gait speed are both the major surrogates of muscle function and physical performance in combination with SMI for the diagnosis of sarcopenia.[21] As the association between grip strength, gait speed and sarcopenia was inconsistent in Chinese participants,[5, 35, 41] the SPPB is not consistently associated with different stages of sarcopenia. Nevertheless, physical performance should be measured universally beyond measurements of SMM.[35]

The present study had several limitations. First, the cross-sectional study did not allow us to identify any causal relationship. Second, only the ambulatory participants were surveyed. Despite the 50% response rate and locally designated examination centers, the prevalence of presarcopenia (low SMM), sarcopenia and severe sarcopenia might still be underestimated. Third, there are still some associated factors cannot be evaluated adequately. As the Nagelkerke R2 values are 0.387–0.563, most of the major factors were evaluated as possible in the present study. Finally, the comparison of the superiority of BIA-derived SMI with DXA-derived SMI in the evaluation of sarcopenia is still inconclusive. However, as the awareness of sarcopenia in recent decade has improved, the cheap, portable and well-validated BIA will become more popular in the near future.

In conclusion, the prevalence of sarcopenia and severe sarcopenia is 7.1% and 5.6%, respectively. Older age, female sex, not working regularly, lower BMI, higher WC, a history of hypertension, and lower MNA or SPPB score were independently associated with sarcopenia and severe sarcopenia in older Taiwanese adults living in a rural community.


This work was supported by the Ministry of Education, Taiwan, R.O.C. under the NCKU Aim for the Top University Project (D101-35001). The authors thank the staff of Tianliao district Public Health Center for their generous support, as well as Ms Yu-Chen Shih for her administrative assistance.

Disclosure statement

The authors declare no conflict of interest.