Table 2.
Diagnostic accuracy of the diagnostic algorithm for sarcopenia with ASM estimated by the anthropometric equation and using DXA as the reference standard.
| Variables | Men (N=131) | Women (N=299) | Total (N=430) | |||
| ASMa (kg), Lin’s concordanceb, mean (SD, 95% CI) | ||||||
|
|
DXAc | 21.4 (3.3, 20.8-21.9) | 14.8 (2.4, 14.5-15.1) | 16.8 (4.1, 16.4-17.2) | ||
|
|
Equation | 21.5 (2.6, 21.0-21.9) | 14.7 (2.2, 14.5-15.0) | 16.8 (3.9, 16.4-17.1) | ||
| SMId (kg/m2), Lin’s concordance, mean (SD, 95% CI) | ||||||
|
|
DXA | 7.92 (1.1, 7.7-8.1) | 6.50 (0.9, 6.4-6.6) | 6.9 (1.2, 6.8-7.0) | ||
|
|
Equation | 7.97 (0.8, 7.8-8.1) | 6.47 (0.8, 6.4-6.6) | 6.9 (1.1, 6.8-7.0) | ||
| Sarcopenia, % (95% CI) | ||||||
|
|
DXA | 17.6 (11.5-25.2) | 13.7 (10.0-18.1) | 14.9 (11.7-18.6) | ||
|
|
Equation | 18.3 (12.1-26.0) | 15.4 (11.5-20.0) | 16.3 (12.9-20.1) | ||
| Summary statistics for diagnostic tests (equation) compared to true disease status (DXA) | ||||||
|
|
Sensitivity, % (95% CI) | 95 (75.1-99.9) | 75 (57.8-87.9) | 82.1 (69.6-91.1) | ||
|
|
Specificity, % (95% CI) | 97.3 (92.3-99.4) | 93.9 (90.3-96.5) | 94.9 (92.2-96.9) | ||
|
|
PPVe, % (95% CI) | 86.4 (65.1-97.1) | 62.8 (46.7-77) | 70.8 (58.2-81.4) | ||
|
|
NPVf, % (95% CI) | 99.1 (95-100) | 96.5 (93.4-98.4) | 97.3 (95.0-98.7) | ||
|
|
Positive likelihood ratio, % (95% CI) | 35.1 (11.5-98) | 12.3 (7.4-20.5) | 16.2 (10.3-25.5) | ||
|
|
Negative likelihood ratio, % (95% CI) | 0.05 (0.01-0.35) | 0.27 (0.15-0.47) | 0.19 (0.11-0.33) | ||
aASM: appendicular skeletal muscle mass.
bLin’s concordance: Lin’s concordance correlation coefficient was 0.96.
cDXA: Dual-energy x-ray absorptiometry.
dSMI: appendicular skeletal muscle mass index.
ePPV: positive predictive value.
fNPV: negative predictive value.