Osteoporosis International

, Volume 24, Issue 2, pp 595–603

Discriminating sarcopenia in community-dwelling older women with high frequency of overweight/obesity: the São Paulo Ageing & Health Study (SPAH)

Authors

  • D. S. Domiciano
    • Bone Metabolism Laboratory, Rheumatology DivisionFaculdade de Medicina da Universidade de São Paulo
  • C. P. Figueiredo
    • Bone Metabolism Laboratory, Rheumatology DivisionFaculdade de Medicina da Universidade de São Paulo
  • J. B. Lopes
    • Bone Metabolism Laboratory, Rheumatology DivisionFaculdade de Medicina da Universidade de São Paulo
  • V. F. Caparbo
    • Bone Metabolism Laboratory, Rheumatology DivisionFaculdade de Medicina da Universidade de São Paulo
  • L. Takayama
    • Bone Metabolism Laboratory, Rheumatology DivisionFaculdade de Medicina da Universidade de São Paulo
  • P. R. Menezes
    • Department of Preventive MedicineFaculdade de Medicina da Universidade de São Paulo
  • E. Bonfa
    • Bone Metabolism Laboratory, Rheumatology DivisionFaculdade de Medicina da Universidade de São Paulo
    • Bone Metabolism Laboratory, Rheumatology DivisionFaculdade de Medicina da Universidade de São Paulo
Original Article

DOI: 10.1007/s00198-012-2002-1

Cite this article as:
Domiciano, D.S., Figueiredo, C.P., Lopes, J.B. et al. Osteoporos Int (2013) 24: 595. doi:10.1007/s00198-012-2002-1

Abstract

Summary

The criteria most used for the definition of sarcopenia, those based on the ratio between the appendicular skeletal muscle mass (ASM) and the square of the height (h2) underestimate prevalence in overweight/obese people whereas another criteria consider ASM adjusted for total fat mass. We have shown that ASM adjusted for fat seems to be more appropriate for sarcopenia diagnosis.

Introduction

Since the prevalence of overweight and obesity is a growing public health issue, the aim of this study was to evaluate the prevalence and risk factors associated with sarcopenia, based on these two criteria, among older women.

Methods

Six hundred eleven community-dwelling women were evaluated by specific questionnaire including clinical data. Body composition and bone mineral density were evaluated by dual X-ray absorptiometry. Logistic regression models were used to identify factors independently related to sarcopenia by ASM/h2 and ASM adjusted for total fat mass criteria.

Results

The prevalence of overweight/obesity was high (74.3 %). The frequency of sarcopenia was lower using the criteria of ASM/h2 (3.7 %) than ASM adjusted for fat (19.9 %) (P < 0.0001). We also note that less than 5 %(1/23) of sarcopenic women, according to ASM/h2, had overweight/obesity, whereas 60 % (74/122) of sarcopenic women by ASM adjusted for fat had this complication. Using ASM/h2, the associated factors observed in regression models were femoral neck T-score (OR = 1.90; 95 % CI 1.06–3.41; P = 0.03) and current alcohol intake (OR = 4.13, 95 % CI 1.18–14.45, P = 0.03). In contrast, we have identified that creatinine (OR = 0.21; 95 % CI 0.07–0.63; P = 0.005) and the White race (OR = 1.81; 95 % CI 1.15–2.84; P = 0.01) showed a significant association with sarcopenia using ASM adjusted for fat.

Conclusions

In women with overweight/obesity, ASM adjusted for fat seems to be the more appropriate criteria for sarcopenia diagnosis. This finding has relevant public health implications, considering the high prevalence of overweight/obesity in older women.

Keywords

Appendicular muscle massFat massObesityPrevalenceSarcopeniaWomen

Introduction

Sarcopenia is a syndrome characterised by a progressive loss of skeletal muscle mass, resulting in decreased muscle strength, impairment of physical and functional capacity and an increased risk of death. It is primarily linked to the ageing process, but may also be secondary to several causes (malnutrition, neurodegenerative diseases, muscle atrophy related to disuse and endocrine diseases). In most cases, the aetiology is multifactorial, and it is difficult to characterise one specific underlying cause for sarcopenia [1].

The diagnosis depends on the measurement of the individual skeletal muscle mass, particularly the appendicular skeletal muscle mass (ASM) evaluated by dual X-ray absorptiometry (DXA), which is the method officially recommended in clinical practice. There are several criteria for sarcopenia based on skeletal muscle mass measurement [25]. The most widely used is that proposed by Baumgartner et al., which defines sarcopenia as being present when the ratio between the ASM (in kilograms) and height squared (square metres)—ASM/h2—is at least two standard deviations below the normal mean of the reference group measured using DXA [2]. This classification criterion was associated with physical disability in a New Mexico population, but its extrapolation to other populations has resulted in low prevalence rates for sarcopenia, ranging from 3 to 16 % [68]. Obesity may account for these unexpected results, as it has been demonstrated that this criteria have the disadvantage to underestimate the prevalence in overweight/obese people [9]. In 2003, Newman et al. proposed another criterion based on ASM measurement adjusted for total fat mass, which seems to identify more individuals with sarcopenia, particularly in obese subjects [5].

This is a relevant issue because the prevalence of overweight and obesity is increasing among older age groups, not only in developed countries [1013], but also in other parts of the world including the Brazilian population [14]. Therefore, the aim of this study was to evaluate the prevalence and risk factors associated with sarcopenia, according to these two criteria in community-dwelling older women.

Materials and methods

This study is an extension of the epidemiological project “Prevalence and risk factors of radiographic vertebral fracture in community-dwelling Brazilian elderly”, a population-based survey (São Paulo Ageing and Health Study—SPAH) [15]. This study was conducted from June 2005 to July 2009, involving individuals aged 65 years or over, in the district of Butantã, in the West Zone of São Paulo, Brazil. All of the individuals were clinically healthy and showed no evidence of malabsorption, chronic diarrhoea, hepatic disease, severe chronic diseases, or cancer. None of the subjects used anabolic steroids or protein supplements.

Clinical and radiographic evaluation

Out of the 959 individuals evaluated for the first study [15], 611 women were included in this protocol. All the individuals answered a standardised questionnaire regarding lifestyle and health, including family history of hip fracture, fragility and previous fractures, track record of falls during the last year, physical activity, current alcohol intake, current tobacco use, personal history of hypertension, diabetes mellitus, cardiovascular events (myocardial infarction and cerebrovascular disease), dietary calcium intake and age at menopause.

Previous fragility fractures were also analysed separately: non-vertebral and vertebral fractures. The personal history of non-vertebral fracture was determined in individuals who had fallen from standing height or less after 50 years of age with a fracture occurring at the most common sites of bone fragility (for example, rib, forearm, humerus and femur) [16]. Vertebral fractures were diagnosed from standard lateral thoracic and lumbar spine radiographs taken from all study participants using a 40-in tube-to-film distance centred on T7 and L2. Images of the T4–L4 vertebrae were evaluated by an expert skeletal radiologist using a Genant semiquantitative approach [17]. Only grade 2 and grade 3 fractures were included in the analysis.

Individuals who had two or more falls in the last 12 months were defined as chronic fallers [18]. An alcohol user was defined as anyone with a current intake of three or more units daily [19]. Physical activity was classified as: low—not even housework is performed; moderate—performs regular housework, occasionally walks, gardening; and high—performs regular physical activity apart from their daily routine at least twice a week for 30 min [20]. Calcium intake was evaluated by dairy product consumption like milk or yoghurt (millilitres per day) and cheese (grams per day) during the 7 days prior to the interview [21]. Race was defined based on the self-reported race of second generation ancestors, an approach previously used for the Brazilian population [22]. Individuals with four grandparents reported as White were classified as White, while those individuals with Black and White ancestors (mixed race) were classified as non-White. Individuals with Asian ancestors were also classified as non-White. When racial information regarding the grandparents was not available, an individual’s race was determined by the race of his or her parents. The height (without shoes) of each participant was measured to the nearest 0.1 cm with a wall-mounted stadiometer. The weight of each participant (without shoes, wearing only light clothing) was measured to the nearest 0.25 kg using a double-beam balance scale.

Body mass index (BMI) was calculated by dividing the participants’ weight (in kilograms) by their height squared (square metres). Individuals were classified by their BMI using both the classification used by the World Health Organisation (WHO) [23] and that recommended by Lipschitz [24]. The WHO cutoff points are: underweight (BMI <18.5 kg/m2), normal weight (BMI >18.5 and <25 kg/m2), overweight or pre-obese (BMI >25 and <30 kg/m2), subject with class I obesity (BMI >30 and <35 kg/m2), subject with class II obesity (BMI >35 and <40 kg/m2), and subject with class III obesity (BMI ≥40 kg/m2). In contrast, the classification proposed by Lipschitz considers the body composition changes from ageing, namely underweight (BMI <22 kg/m²), normal weight (BMI ≥22 and <27 kg/m²) and overweight/obese (BMI >27 kg/m²). This study was approved by the Local Ethics in Research Committee of the School of Medicine at the University of São Paulo (FM-USP), and all participants have given written informed consent.

Laboratory evaluation

The concentrations of serum calcium (adjusted for the albumin concentration), phosphorus, alkaline phosphatase, creatinine, total cholesterol, HDL, triglycerides, albumin and glucose were determined using standard automated laboratory methods. The estimated glomerular filtration rate (eGFR) was calculated using the Cockroft–Gault equation [25]. The concentration of serum 25-hydroxyvitamin D (25OHD) was measured using the radioimmunoassay technique (DiaSorin, Stillwater, MN, USA) with a lower detection limit of 5 ng/mL. Vitamin D was expressed by the mean (standard deviation) and also by vitamin D categories according to the definitions of 2011 Endocrine Society Clinical Practice Guideline [vitamin D deficiency as a 25OHD below 20 ng/mL (50 nmol/L), vitamin D insufficiency as a 25OHD of 21–29 ng/mL (52.5–72.5 nmol/L) and vitamin D sufficiency as a 25OHD ≥30 ng/mL (75 nmol/L)] [26].

The intra- and inter-assay variation coefficients in our laboratory were 10.5 and 17.8 %, respectively. Intact parathyroid hormone serum concentrations were measured by immunoradiometric assay (ELSA-PTH, CIS bio international, France), with reference variations of 11–65 pg/mL.

Body composition and bone mineral density

Body composition and bone mineral density (BMD) were determined by DXA, using Hologic QDR 4500A densitometry equipment (Hologic Inc. Bedford, MA, USA, Discovery model), at the following regions: lumbar spine, femoral neck, total femur and total body. All DXA measurements were performed by the same experienced technologist.

Appendicular lean mass was calculated as the sum of arms and legs lean soft tissue masses, assuming that all non-fat and non-bone tissue is skeletal muscle. The total body fat was expressed in grams and as a percentage of body weight.

The precision error for DXA measurements was determined based on standard ISCD protocols [27]. We calculated the least significant change with 95 % confidence to be 3.3 % for lumbar spine, 4.7 % for femoral neck, 3.9 % for total femur and 1 % for the whole body.

Definitions of sarcopenia

According to Baumgartner’s criteria, sarcopenia is defined when the relative skeletal muscle mass index (RSMI = ASM/height2) is less than 5.45 kg/m2 (value for female) [2].

The Newman’s criteria is also based on ASM measurement adjusted for height (similar to Baumgartner’s), but also for total fat mass [5]. Linear regression was used to model the association between ASM on height (in metre) and fat mass (in kilograms). The residuals from linear regression models were used to identify those individuals whose amount of ASM (obtained from DXA) was lower than expected (obtained by an equation resultant from the model) for a given amount of fat mass. A positive residual would indicate a relatively muscular individual, whereas negative values would indicate relatively sarcopenic individual. Our equation resultant from the model was:
$$ {\text{ASM}}\left( {\text{in\ kilograms}} \right) = - {14}.{51} + {17}.{27} \times {\text{height}}\left( {\text{in metres}} \right) + 0.{2}0 \times {\text{fat\ mass}}\left( {\text{in\ kilograms}} \right) $$

The 20th percentile was defined as the cutoff point for sarcopenia in accordance with the study by Newman et al. [5]. This cutoff corresponded to a residual of −1.45 in the population studied herein.

Statistical analysis

Results were expressed as the mean (standard deviation) or percentages. Differences between the two groups (sarcopenia group and no sarcopenia) according to the two criteria were evaluated using the McNemar test, Student’s t test, Mann–Whitney U test or the chi-square test.

Logistic regression models were used to analyse which factors were independently associated with sarcopenia regarding the two criteria. Only those variables significantly associated (P < 0.05) with sarcopenia in the univariate analysis were included in the logistic regression models. Lean mass was not included as an independent variable for both criteria and neither was the fat mass for ASM adjusted for fat criteria. Because collinearity exists between BMD sites, sites were sequentially added to the logistic regression, and the best model was then selected. These findings are presented as adjusted prevalence ratios with corresponding 95 % confidence intervals (95 % CI). Significance was set at P < 0.05. All analyses were performed using Stata 9.0 software.

Results

The mean age of the 611 women studied was 73.22 ± 5.21 years, 389 (63.7 %) White, 206 (33.7 %) Black or mulatto and 16 (2.6 %) Asian. The mean weight and height were, respectively, 64.75 ± 12.94 kg and 1.50 ± 0.06 m. The mean BMI was 28.67 ± 5.20 kg/m2, with 213 subjects (34.9 %) classified as overweight (BMI ≥25 and <30 kg/m2) and 235 (38.5 %) as obese (BMI >30 kg/m2), according to the WHO classification [23]. Using the BMI classification adjusted for ageing [24], 363 women (59.4 %) were classified as overweight or obese (BMI >27 kg/m2).

Prevalences of sarcopenia and overweight/obesity

The frequency of sarcopenia was significantly lower (3.7 %) using ASM/h2 than ASM adjusted for fat (19.9 %) (P < 0.0001). Twenty-two individuals were considered sarcopenic by both methods, 100 women were considered sarcopenic by ASM adjusted for fat criteria and not by ASM/h2, while only one woman was classified as sarcopenic by the ASM/h2 definition and not by ASM adjusted for fat criteria. The distribution of subjects according to the presence or absence of sarcopenia, using each method, is shown in Fig. 1.
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-012-2002-1/MediaObjects/198_2012_2002_Fig1_HTML.gif
Fig. 1

Distribution of individuals according to the two sarcopenia criteria

The distribution of individuals according to the BMI categories is presented in Figs. 2 and 3. We must also note that, according to the WHO classification of BMI [23], less than 5 % (1/23) of those classified as sarcopenic by ASM/h2 were overweight or obese (Fig. 2a), whereas 60 % (74/122) of those individuals considered sarcopenic based on the ASM adjusted for fat criteria had this complication (Fig. 2b). Using the cutoff points proposed by Lipschitz [24], a similar distribution was observed: less than 5 % (1/23) of women considered sarcopenic by ASM/h2 (Fig. 3a) and more than 45 % (55/122) of sarcopenic women classified as sarcopenic by ASM adjusted for fat criteria (Fig. 3b) were classed as overweight or obese.
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-012-2002-1/MediaObjects/198_2012_2002_Fig2_HTML.gif
Fig. 2

a Classification of individuals according to BMI categories (WHO criteria), comparing sarcopenic and non-sarcopenic women in each criteria, using ASM/h2. b Classification of individuals according to BMI categories (WHO criteria), comparing sarcopenic and non-sarcopenic women in each criteria, using ASM adjusted for fat

https://static-content.springer.com/image/art%3A10.1007%2Fs00198-012-2002-1/MediaObjects/198_2012_2002_Fig3_HTML.gif
Fig. 3

a Classification of individuals according to BMI categories (Lipschitz’s criteria), comparing sarcopenic and non-sarcopenic women in each criteria, using ASM/h2. b Classification of individuals according to BMI categories (Lipschitz’s criteria), comparing sarcopenic and non-sarcopenic women in each criteria, using ASM adjusted for fat

Risk factors according to ASM/h2 criteria

Women with sarcopenia had lower weight, lower BMI and higher frequency of alcohol intake compared to women without sarcopenia (Table 1). In relation to the laboratory variables, only the eGFR was significantly lower in cases of sarcopenia. No association between 25OHD categories and mean vitamin D levels and sarcopenia was observed (Table 1). Considering the blood samples collected among the seasons, similar distribution of sarcopenic subjects was observed among the seasons (spring and summer, 47.8 % compared with autumn and winter, 52.2 %, P = 0.63).
Table 1

Clinical, laboratory and DXA parameters in women with and without sarcopenia according to ASM/h2 criteria

 

Sarcopenia (n = 23)

No sarcopenia (n = 588)

P

Age, years

74.34 ± 4.42

73.20 ± 5.23

0.11

Race (White), %

60.9

63.8

0.32

Height, m

1.49 ± 0.06

1.50 ± 0.06

0.49

Weight, kg

47.70 ± 7.90

65.41 ± 12.64

<0.001

BMI, kg/m²

21.4 ± 2.83

29.0 ± 5.06

<0.001

Alcohol (≥3 U/day), %

17.4

5.3

0.03

Current smoking, %

17.4

9.9

0.27

Low physical activity, %

17.4

8.3

0.17

Falls (>1/year), %

26.1

38.1

0.24

Personal history of fracture, %

17.4

19.9

0.99

Radiographic vertebral fracture, %

26.1

30.1

0.68

Diabetes mellitus, %

13.0

23.3

0.25

Hypertension, %

56.5

73.0

0.08

Cardiovascular event, %

17.4

13.1

0.53

Calcium, mg/dl

9.47 ± 0.47

9.43 ± 0.45

0.76

Phosphorus, mg/dl

3.57 ± 0.49

3.52 ± 0.46

0.48

Alkaline phosphatase, U/L

199.17 ± 137.64

187.82 ± 54.90

0.33

iPTH, pg/dl

52.43 ± 54.95

40.90 ± 18.95

0.77

25OHD, ng/ml

19.27 ± 11.09

18.44 ± 8.97

0.76

25OHD ≥30 ng/ml, %

4.4

11.2

0.36

25OHD 21–29 ng/ml, %

34.7

24.0

 

25OHD ≤20 ng/ml, %

60.9

64.7

 

Creatinine, mg/dl

0.91 ± 0.26

0.94 ± 0.20

0.23

Creatinine clearance, ml/min

51.30 ± 16.04

67.65 ± 21.03

<0.001

Albumin, g/dl

4.41 ± 0.29

4.38 ± 0.27

0.62

Spine T-score

−2.67 ± 1.49

−2.25 ± 1.62

0.36

Femoral neck T-score

−2.60 ± 0.72

−1.83 ± 1.06

<0.001

Total femur T-score

−2.20 ± 0.74

−1.35 ± 1.04

<0.001

Lean mass, g

30,192.6 ± 3,285.9

39,185.4 ± 5,675.7

<0.001

Fat mass, g

15,858.9 ± 5,414.0

24,430.5 ± 7,780.2

<0.001

Fat mass, %

32.70 ± 6.17

36.71 ± 5.76

<0.001

All DXA bone parameters, except the L1–L4 T-score, were significantly lower in the sarcopenia group (Table 1). Regarding the DXA body composition, total lean mass, total fat mass and percentage of fat mass were lower in the individuals with sarcopenia.

The logistic regression model, after due adjustments for age, has shown that the femoral neck T-score (OR, 1.90; 95 % CI 1.06–3.41; P = 0.03) and current alcohol intake (three or more units daily) (OR, 4.13; 95 % CI 1.18–14.45, P = 0.03) remained significantly associated with sarcopenia.

Risk factors according to ASM adjusted for fat criteria

The sarcopenia group had a higher frequency of White race, lower weight and lower BMI compared to non-sarcopenic subjects. Concerning laboratory parameters, only creatinine was significantly lower in the sarcopenia group compared to subjects without this condition. No association between 25OHD categories and mean vitamin D levels and sarcopenia was observed (Table 2). Considering the blood samples collected among the seasons, similar distribution of sarcopenic subjects was observed among the seasons (spring and summer, 52.5 % vs. autumn and winter, 47.5 %, P = 0.95).
Table 2

Clinical, laboratory and DXA parameters in women with and without sarcopenia according to ASM adjusted for fat criteria

 

Sarcopenia (n = 122)

No sarcopenia (n = 489)

P

Age, years

73.68 ± 4.97

73.11 ± 5.26

0.15

Race (White), %

74.6

60.9

0.001

Height, m

1.51 ± 0.06

1.50 ± 0.06

0.11

Weight, kg

62.02 ± 13.06

65.43 ± 12.84

0.006

BMI, kg/m²

27.11 ± 5.13

29.06 ± 5.15

<0.001

Alcohol (≥3 U/day), %

7.4

5.3

0.38

Current smoking, %

9.8

10.2

0.90

Low physical activity, %

11.5

8.0

0.41

Falls (>1/year), %

43.4

36.2

0.14

Personal history of fracture, %

23.8

18.8

0.22

Radiographic vertebral fracture, %

28.7

30.3

0.73

Diabetes mellitus, %

19.7

23.7

0.34

Hypertension, %

73

72.2

0.87

Cardiovascular event, %

11.5

13.7

0.52

Calcium, mg/dl

9.45 ± 0.49

9.42 ± 0.44

0.66

Phosphorus, mg/dl

3.55 ± 0.47

3.52 ± 0.46

0.62

Alkaline phosphatase, U/L

188.97 ± 77.94

188.08 ± 54.63

0.67

iPTH, pg/dl

41.61 ± 27.30

41.26 ± 19.75

0.71

25OHD, ng/ml

19.22 ± 10.47

18.29 ± 8.66

0.70

25OHD ≥30 ng/ml, %

13.9

10.3

0.40

25OHD 21–29 ng/ml, %

21.3

25.2

 

25OHD ≤20 ng/ml, %

64.8

64.6

 

Creatinine, mg/dl

0.89 ± 0.21

0.95 ± 0.20

0.001

Creatinine clearance, ml/min

67.76 ± 22.75

66.86 ± 20.67

0.81

Albumin, g/dl

4.41 ± 0.27

4.37 ± 0.28

0.09

L1-L4 T-score

−2.12 ± 1.54

−2.30 ± 1.64

0.21

Femoral neck T-score

−1.88 ± 1.09

−1.86 ± 1.05

0.60

Total femur T-score

−1.49 ± 0.99

−1.35 ± 1.05

0.18

Lean mass, g

35,471.7 ± 5,341.9

39,688.9 ± 5,680.7

<0.001

Fat mass, g

24,570 ± 8,090

23,990 ± 7,820

0.72

Fat mass %

38.97 ± 5.38

35.97 ± 5.78

<0.001

Out of the DXA parameters, the total lean mass was lower in individuals with sarcopenia, while the percentage of fat mass was higher in this group (Table 2). The logistic regression model, after due adjustments for age, showed that White race (OR, 1.81; 95 % CI 1.15–2.84; P = 0.01) and creatinine (OR, 0.21; 95 % CI 0.07–0.63; P = 0.005) remained significantly associated to sarcopenia.

Comparison between sarcopenic women according to ASM/h2 and ASM adjusted for fat criteria

Women with sarcopenia according to ASM/h2 criteria presented lower weight, BMI, femoral neck T-score, total femur T-score, lean mass, fat mass and percentage of fat mass compared to sarcopenic women according to the criteria of ASM adjusted for fat (Table 3).
Table 3

Clinical, laboratory and DXA parameters in sarcopenic women comparing ASM/h2 and ASM adjusted for fat criteria

 

ASM/h2 (n = 23)

ASM adjusted for fat (n = 122)

P

Age, years

74.34 ± 4.42

73.68 ± 4.97

0.53

Race (White), %

60.9

74.6

0.20

Height, m

1.49 ± 0.06

1.51 ± 0.06

0.15

Weight, kg

47.70 ± 7.90

62.02 ± 13.06

0.006

BMI, kg/m²

21.4 ± 2.83

27.11 ± 5.13

<0.001

Alcohol (≥3 U/day), %

17.4

7.40

0.13

Current smoking, %

17.4

9.80

0.28

Low physical activity, %

17.4

11.5

0.48

Falls (>1/year), %

26.1

43.40

0.16

Non-vertebral fracture, %

17.40

23.80

0.59

Radiographic vertebral fracture, %

26.1

28.7

0.99

Diabetes mellitus, %

13.04

19.7

0.56

Hypertension, %

56.52

73

0.13

Cardiovascular event, %

17.40

11.5

0.48

Calcium, mg/dl

9.47 ± 0.47

9.45 ± 0.49

0.88

Phosphorus, mg/dl

3.57 ± 0.49

3.55 ± 0.47

0.84

Alkaline phosphatase, U/L

199.17 ± 137.64

188.97 ± 77.94

0.60

iPTH, pg/dl

52.43 ± 54.95

41.61 ± 27.30

0.15

25OHD, ng/ml

19.27 ± 11.09

19.22 ± 10.47

0.97

Creatinine, mg/dl

0.91 ± 0.26

0.89 ± 0.21

0.76

Creatinine clearance, ml/min

51.30 ± 16.04

67.76 ± 22.75

0.001

Albumin, g/dl

4.41 ± 0.29

4.41 ± 0.27

0.93

Spine T-score

−2.67 ± 1.49

−2.12 ± 1.54

0.12

Femoral neck T-score

−2.60 ± 0.72

−1.88 ± 1.09

0.002

Total femur T-score

−2.20 ± 0.74

−1.49 ± 0.99

0.001

Lean mass, g

30,192.6 ± 3,285.9

35,471.7 ± 5,341.9

<0.001

Fat mass, g

15,858.9 ± 5,414.0

24,570 ± 8,090

<0.001

Fat mass %

32.70 ± 6.17

38.97 ± 5.38

<0.001

Discussion

Our study has shown that the criteria based on ASM adjusted for fat seems to be more appropriate for identifying sarcopenia in community-dwelling older women with high prevalence of overweight and obesity and distinct associated factors. The great advantage of the design of the present study is the evaluation of a large population of individuals over 65 years old, as the highest prevalence of sarcopenia and frailty syndrome occurs after this age [58, 28]. In fact, the inclusion of younger patients may hamper the interpretation of risk factors [7, 8, 29], given that age-related factors for sarcopenia are expected to be distinct according to the age bracket. In addition, the use of the community rather than subjects in hospitals or institutions reduces other confounding variables relevant for sarcopenia [30, 31]. In this regard, an evaluation of general health has shown that community-dwelling older Brazilian individuals took exercise more regularly than institutionalised ones [32]. Another important feature of our study is gender restriction given that risk factors for sarcopenia, osteoporosis and fractures are different in men and women, with dissimilar low skeletal muscle mass (ASM) cutoffs [2, 5, 8, 28, 30].

The prevalence of sarcopenia using ASM/h2 criteria is lower when compared to other countries including Asian and Hispanic populations [6, 3335]. The most likely explanation for this is methodological, as the ASM/h2 definition cannot adequately identify overweight or obese people as sarcopenic. Indeed, most of the individuals in this study were overweight or obese. Only one subject in this category was identified as sarcopenic by ASM/h2 contrasting with more than half of those overweight/obese women who were diagnosed as sarcopenic based on the criteria of ASM adjusted for fat. In fact, another study also showed that older subjects with a BMI higher than 21 kg/m2 have a lower risk of being sarcopenic using ASM/h2 criteria [30].

The WHO recommends the same cutoffs of BMI for adults aged 18 and over [23]. On the other hand, Lipschitz proposed another classification considering the body composition changes resulting from ageing, since older people have decreased height, less body water and a higher percentage of fat when compared with adults. Thus, this recommendation modifies the cutoff point for the definition of underweight (18 to 22 kg/m2) and overweight (25 to 27 kg/m2) [24]. The higher prevalence of sarcopenic women with overweight/obesity using the criteria of ASM adjusted for fat, compared with the criteria of ASM/h2, was observed regardless of the cutoff used for the definition of overweight/obesity (25 or 27 kg/m2).

This issue is relevant since sarcopenic obesity has been associated with reduced walking speed and functional limitation [36, 37]. Moreover, some studies have demonstrated that obesity and muscle impairment act synergistically on the risk of functional incapacity, and sarcopenic obese subjects seem to be more disabled than people who are just obese or sarcopenic [38]. This may also be explained by the fact that, in obese people, lean body mass is always higher than in slim subjects because both lean mass and fat mass increase with weight gain. Normally, an increase in lean mass is expected to accompany an increase in fat mass by a ratio of approximately 1:4 [39]. Due to this dependence of lean mass on fat mass, the adequacy of an individual’s lean mass cannot be gauged in isolation. Therefore, obese subjects may not seem sarcopenic, as their muscle mass is actually inadequate for the size of their body and physical performance. Although the functional status has not been evaluated in our population, nevertheless the recommended cutoff for ASM adjusted for fat criteria used herein is probably related to disability as also has been reported in previous studies [5, 9].

In addition, sarcopenic obesity has been associated with increased cardiovascular risk [33, 35, 40] and mortality [41, 42]. The underlying mechanism is probably associated with the known proinflammatory state induced by visceral obesity [42], including muscle inflammation generated by fat infiltration, which may enhance the loss of muscle quality in older obese people [35, 43, 44].

We described an equation, based on the linear regression models, which allows the calculation of the expected ASM according to total fat mass. The result obtained from this formula should be subtracted from the ASM obtained by DXA, and this result is the residual of the ASM. Whether this residual is less than −1.45, we considered the woman is sarcopenic. Our sample was characterised by homogenous selection of the group; all subjects lacked vitamin supplementation, were from the same gender, and all were living independently. Our sample is, in fact, representative of the Brazilian older population according to the Brazilian Institute of Geography and Statistics (IBGE) [45], as reported in our previous studies with the same population (SPAH) [15, 46, 47]. Thus, this equation can be applied to Brazilian females aged over 65 years with a residual less than −1.45 for the sarcopenia cutoff point.

As for features related to sarcopenia, we found different factors according to each criterion. The probable explanation for this is the distinct group of subjects selected by each criterion. In this regard, the ASM/h2 criteria included only slim people as sarcopenic, while using the ASM adjusted for fat, only a minority of sarcopenic women were underweight or of normal weight. It is therefore expected that the two groups have different risk factors for sarcopenia. The femoral neck T-score was associated with sarcopenia as determined by the ASM/h2 method. This finding is most likely related with lower BMI, given that this parameter is a recognised risk for osteoporosis and fractures [48]. In contrast, a higher T-score is expected in sarcopenic subjects, according to the criteria of ASM adjusted for fat, in view of the fact that a higher fat mass will result in a greater BMD [49]. Furthermore, alcohol intake (three or more units daily) was also associated with sarcopenia, according to the ASM/h2. This is a relevant finding, as the prevalence of alcohol consumption is high in the general population, including older age groups [50]. In addition, chronic alcohol consumption is associated with reduced lean body mass, and lower lean body mass is related to mortality among alcoholics [51]. We analysed the alcohol intake as a dichotomous variable, considering three or more units of alcohol daily as being a significant alcohol intake, the same cutoff point used for the risk of osteoporosis and fractures [19].

The association between sarcopenia and alcohol intake was observed only for the criteria based on ASM/h2 and not for the criteria of ASM adjusted for fat. This is due to the fact that chronic alcohol users tended to be thinner than individuals with no alcohol intake, and the criteria based on ASM/h2 is able to identify only thin people as sarcopenic.

Regarding the criterion of ASM adjusted for fat, we have confirmed and extended the previous observation that the non-White ethnic group was found to be protective for sarcopenia [5]. We also demonstrated that serum creatinine was also a protective factor. Creatinine is the expression of muscle creatinine content and total muscle mass of the subject, and individuals with less muscle mass tend to have lower creatinine values [52, 53]. This result may be useful in clinical practice, as serum creatinine is a routine laboratory examination and, hence, low levels may be predictive of sarcopenia.

We concluded that, among overweight and obese women, the criterion of ASM adjusted for fat seems to be more appropriate for the diagnosis of sarcopenia. This has relevant public health implications, considering the risk factors identified herein and also the high prevalence of overweight and obesity among older women.

Acknowledgments

This work was supported by grants from the Fundação de Amparo e Pesquisa do Estado de São Paulo (FAPESP) #03/09313-0 and #04/12694-8, Conselho Nacional de Ciência e Tecnologia #300559/2009-7 (RMRP), Federico Foundation (EB, RMRP) and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (JBL, CPF).

Conflicts of interest

None.

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2012