Table III.
Non-parametric Bootstrap Analysis and Parameter Estimates (Relative Standard Errors) of the Testosterone Population Kinetic Model
| Model parameters | Original dataset | Non-parametric Bootstrap (N = 1000 replicates) | |
|---|---|---|---|
| Estimate (RSE%) | Mean | 95% confidence interval | |
| Base (ng/dL) | 239 (1.1) | 239 | 232, 245 |
| Amplitude (ng/dL) | 32.1 (3.5) | 32.2 | 30.2, 34.1 |
| t max (clock time) | 9:22 (0.4) | 9:21 | 8:13, 10:30 |
| t min (clock time) | 14:02 (0.4) | 14:00 | 12:49, 15:33 |
| Season on Base (%) | 8.09 (18.8) | 7.86 | 4.66, 11.0 |
| Age on Base (%/10 years) | −2.40 (24.0) | −2.08 | −0.07, −3.55 |
| Box–Cox on Base | −1.93 (2.4) | −1.98 | −1.69, −2.30 |
| Interindividual variability (CV%) | |||
| η Base | 25.1 (3.6) | 25.1 | 23.1, 27 |
| η Amplitude | 50.3 (6.6) | 50.2 | 43.5, 56.0 |
| 10.6 (19.1) | 10.4 | 7.7, 14.4 | |
| 19.0 (11.5) | 19.4 | 16.0, 22.1 | |
| Residual variability (CV%) | |||
| σ proportional | 13.8 (2.4) | 13.8 | 13.1, 14.4 |
Shrinkage values for t max, t min, Amplitude, Base parameters, and residual variability were 60, 46, 27, 4, and 2%, respectively
Base baseline value, t max time at which the peak of testosterone is reached, t min time at which the nadir of testosterone occurs