Table 7.
Logistic regression analysis to predict presence of low appendicular lean tissue using the optimized five‐site protocol
| Model development | Log odds: low lean tissue | Odds ratio [95% CI] | Validation fold | c‐index [95% CI] | ||
|---|---|---|---|---|---|---|
| Five site | Sex (women = 1) | Age | ||||
| Folds 1 + 2 | 13.209 − (1.157*X2) − (3.415*X3) − (0.0217*X4) | 0.31 [0.15, 0.67] | 0.03 [0.00, 0.86] | 0.98 [0.94, 1.02] | 3 | 0.96 [0.89, 1.00] |
| Folds 1 + 3 | 17.914 − (1.554*X2) − (4.380*X3) − (0.0355*X4) | 0.21 [0.08, 0.57] | 0.01 [0.00, 0.50] | 0.97 [0.92, 1.02] | 2 | 0.85 [0.71, 0.99] |
| Folds 2 + 3 | 16.431 − (1.431*X2) − (4.198*X3) − (0.0275*X4) | 0.24 [0.10, 0.60] | 0.02 [0.00, 0.65] | 0.97 [0.93, 1.02] | 1 | 0.87 [0.74, 1.00] |
| Average | 15.851 − (1.381*X2) − (3.998*X3) − (0.0282*X4) | 0.26 [0.11, 0.61] | 0.02 [0.00, 0.67] | 0.97 [0.93, 1.02] | — | 0.89 [0.78, 1.00] |
X2 = five‐site muscle thickness (average four‐site muscle thickness + anterior upper arm muscle thickness) × height (cm × m), X3 = sex (male = 0, female = 1), X4 = age (years). c‐index, concordance index; CI, confidence interval.