Table 2.
Results of CTM and 2-level and 3-level MLM analyses for the full sample: path coefficient estimates with standard errors (SE) and standardized variance components of the 2-level and the 3-level models (with age and sex as covariates)
| Parameter | CTM | 2-Level model | 3-Level model |
|---|---|---|---|
| Intercept | |||
| Intercept (SE) | 128.4 (0.11) | 128.4 (0.11) | 128.5 (0.14) |
| Covariates | |||
| βsex (SE) | − 0.62 (0.12) | − 0.62 (0.12) | − 0.62 (0.12) |
| βage (SE) | 1.43 (0.08) | 1.42 (0.08) | 1.42 (0.08) |
| a, c, e, Region path loadings | |||
| a (SE) | 4.70 (0.08) | 4.70 (0.08) | 4.70 (0.08) |
| c (SE) | 2.90 (0.16) | 2.90 (0.16) | 2.80 (0.16) |
| e (SE) | 1.39 (0.03) | 1.39 (0.03) | 1.39 (0.03)) |
| Region (SE) | 92.33 (2.13) | ||
| Total variance (a2 + c2 + e2 (+ region2)) | 32.46 | 32.46 | 32.47 |
| A (standardized) | 0.681 | 0.681 | 0.681 |
| C (standardized) | 0.259 | 0.259 | 0.241 |
| E (standardized) | 0.060 | 0.060 | 0.060 |
| REGION (standardized) | 0.018 | ||
Key differences between the 2-Level and the 3-Level model are printed in bold
N = 7346 twins in 3724 families