Background
Methods
Study population
Assessment of muscle function
Measurement of muscle mass by DXA
Measurement of muscle mass by BIA
Method specific cut-offs for reduced muscle mass
Comparison of the DXA- and BIA-based approaches to detect reduced muscle mass and diagnose sarcopenia
Statistics
Results
Study population
Total, n (%) | Female, n (%) | Male, n (%) |
p valuec
| |
---|---|---|---|---|
Number of participants | 60 (100) | 42 (70) | 18 (30) | |
Age (y, mean ± SD) | 81.6 ± 5.28 | 80.4 ± 5.29 | 84.5 ± 4.08 |
p = 0.005 |
Community dwelling | 55 (92) | 37 (88) | 18 (100) | n.s. |
Malnutrition (MNA <17) | 26 (43) | 21 (50) | 5 (28) | n.s. |
Obesity (BMI > 30 kg/m2) | 14 (23) | 10 (24) | 4 (22) | n.s. |
Morbidity: | ||||
Coronary heart disease | 18 (30) | 9 (21) | 9 (50) |
p = 0.035 |
Chronic heart failure | 16 (27) | 9 (21) | 7 (39) | n.s. |
Cerebrovascular diseasea
| 20 (33) | 11 (26) | 9 (50) | n.s. |
Art. Hypertension | 49 (82) | 35 (83) | 14 (78) | n.s. |
COPD | 6 (10) | 3 (7) | 3 (17) | n.s. |
Diabetes | 16 (27) | 11 (26) | 5 (28) | n.s. |
CKD (≥ stage 3) | 22 (37) | 12 (29) | 10 (56) | n.s. |
Cancerb
| 10 (17) | 6 (14) | 4 (22) | n.s. |
Mild or moderate dementia | 7 (12) | 3 (7) | 4 (22) | n.s. |
Comorbidity (≥2 of listed diseases) | 46 (77) | 30 (71) | 16 (89) | n.s. |
Comorbidity (≥3 of listed diseases) | 35 (58) | 20 (48) | 15 (83) |
p = 0.012 |
Polypharmacy (≥5 drugs taken) | 48 (80) | 33 (78) | 15 (83) | n.s. |
Dependency in ADL (Barthel <70) | 27 (45) | 19 (45) | 8 (44) | n.s. |
Gait speed [m/s] (mean ± SD) | 0.86 ± 0.37 | 0.77 ± 0.34 | 1.05 ± 0.36 |
p = 0.005 |
Grip strength [kg] (mean ± SD) | 24.8 ± 9.90 | 20.9 ± 7.36 | 33.6 ± 9.45 |
p < 0.001 |
Correlation of muscle mass measurements between DXA and BIA
Agreement between the DXA- and BIA-based approaches to detect reduced muscle mass
a | Aagreement % (CI) | Cohens’ κ-coeff. (CI) |
DXA vs. BIA (threshold Janssen et al.) | 73 % (61.7–83.3) | 0.437 (0.218–0.653) |
DXA vs. BIA (threshold Chien et al.) | 80 % (68.3–90.0) | 0.492 (0.219–0.734) |
b | Agreement % (CI) | Cohens’ κ-coeff. (CI) |
DXA vs. BIA (threshold Janssen et al.) | 79 % (65.8–92.1) | 0.579 (0.309–0.829) |
DXA vs. BIA (threshold Chien et al.) | 84 % (71.1–94.7) | 0.636 (0.321–0.883) |
a | Patients with reduced muscle mass (n = 60, before functional prescreening) | ||||||||
True pos. (n) | False pos. (n) | false neg. (n) | True neg. (n) | PPV | NPV | Sens. | Spec. | Acc. | |
DXA | 18 | - | - | 42 | |||||
BIA (threshold Janssen et al.) | 14 | 12 | 4 | 30 | 54 % | 88 % | 77 % | 71 % | 73 % |
BIA (threshold Chien et al.) | 10 | 4 | 8 | 38 | 71 % | 83 % | 55 % | 90 % | 80 % |
b | Patients with sarcopenia (n = 38, after functional prescreening) | ||||||||
True pos. (n) | False pos. (n) | False neg. (n) | True neg (n) | PPV | NPV | Sens. | Spec. | Acc. | |
DXA | 13 | - | - | 25 | |||||
BIA (threshold Janssen et al.) | 12 | 7 | 1 | 18 | 63 % | 94 % | 92 % | 72 % | 79 % |
BIA (threshold Chien et al.) | 9 | 2 | 4 | 23 | 82 % | 85 % | 69 % | 92 % | 84 % |