A Cholesky decomposition was applied with Mx (M. C. Neale et al., 1999) using raw ordinal data that allowed for missing data within a twin and when no cotwin information was available. This approach decomposes the variance of the individual phenotypes, i.e., AN, MDD, and SA, and the covariance among these phenotypes to estimate: 1) the contribution of additive genetic, common environmental, and unique environmental factors; and 2) the genetic, common environmental, and unique environmental correlations (ra, rc, and re, respectively). These correlations indicate the extent to which the traits share the respective factors. When no cotwin data are available, the twin’s information is still useful because it contributes to the variance, but not correlations, of the phenotypes.
