Genetic correlation data to test this hypothesis are limited: one twin study of 67 pairs of twins with bipolar and 177 with unipolar depression yielded a genetic correlation of 0.65 between the two disorders. However, the data were not consistent with the threshold model, namely that bipolar is a more severe subform of unipolar (McGuffin et al., 2003). A larger study of 486 twin pairs with affective illness provided some support for the threshold model, but the number of bipolar probands was small, so power to discriminate models was low (Kendler et al., 1995b).