A bivariate model was fitted to raw categorical data from twins and siblings in the statistical software package Mx (Neale, 2004), which uses full information maximum likelihood estimation. A series of sub-models were tested to assess statistical significance of the A, C, T, and E influences and whether these influences could be equated across AA and EA subsamples. The difference between the -2 log likelihood fit of the full model and the nested sub-model, which is distributed as chi-square for the given degrees of freedom, was used to determine relative fit. All twin models were adjusted for age and study design variables (MOFAM low risk, high risk or very high risk).
