Table 1.
LMM shows the effect of age on rTL. Age effect estimates are on a natural-log-transformed scale. The β-estimates (β), standard errors (s.e.), degrees of freedom (d.f.), t-statistics (t) and associated p-values (p) for fixed effects and variances (σ2) attributable to random effects, are shown for each model parameter. Italicized shows parameters for which p < 0.05. When this log age effect is decomposed into within- and between-individual age effects on rTL, effect slopes are statistically similar for βW and βB of log age (electronic supplementary material, table S3).
| effect | β | s.e. | d.f. | t | p | σ2 |
|---|---|---|---|---|---|---|
| intercept | 3.50 × 10−1 | 9.60 × 10−2 | 194.2 | 3.644 | <0.001 | — |
| log age | −2.30 × 10−1 | 4.22 × 10−2 | 622.7 | −5.460 | <0.001 | — |
| storage time | −1.90 × 10−4 | 5.71 × 10−5 | 706.7 | −3.322 | <0.001 | — |
| qPCR run | — | — | — | — | — | 0.082 |
| individual ID | — | — | — | — | — | 0.290 |
| (residual) | — | — | — | — | — | 0.575 |