Table 2.
Parameter estimates of the statistical model.
| Effect | Estimate (nMU) | Prediction (nMU, reference subject) |
Degrees of freedom | f | p |
|---|---|---|---|---|---|
| Intercept | 1.10 | 1.10 | – | – | – |
| Age | −0.013/year | 1.10 | 1/34 | 0.00 | 0.99 |
| Recovery | −0.65 | 0.45 | 1/480 | 263 | <0.0001 |
| Recovery:age at onset | −0.015/year | 0.45 | 1/431 | 12.8 | 0.0004 |
| Recovery:age | −0.19/year | 0.45 | 1/518 | 5.18 | 0.023 |
| Recovery:age2 | 0.032/year2 | 0.45 | 1/504 | 0.740 | 0.39 |
| Recovery:treatment (IFN-A) | 0.08 | 0.53 | 3/463 | 6.46 | 0.0003 |
| Recovery:treatment (IFN-R) | 0.0067 | 0.45 | |||
| Recovery:treatment (GA) | −0.23 | 0.22 | |||
| Random effects | χ2 = 223; p < 0.00001 | ||||
| Lesion | Variance = 0.056 | ||||
| Timepoint | Variance = 0.000 | ||||
| Subject | Variance = 0.040 | ||||
| Residual | Variance = 0.066 | ||||
p > 0.05 using a standard general linear mixed model f-test with denominator degrees of freedom estimated by the Satterthwaite approximation, as discussed in the methods.