umxACEv - twin variance component

As a comment, note that variance that seems specific to twins because DZ twins correlate more than siblings could be due to age and cohort effects that decay the sibling correlation relative to the DZ twins. Even prenatally, maternal diet and stress likely differ more across pregnancies than it does across placentas in a twin pregnancy. That too could easily generate lower sib than DZ correlations. These processes seem inherently at least as likely as, e.g., people going "Ooooh they're twins, let's treat them really really similarly..." - this is especially difficult to do if you don't know a person is a twin :).