For many years, genetic data have been employed to support a separation of unipolar from bipolar affective illnesses: relatives of those with bipolar are more likely to develop bipolar, and conversely relatives of unipolar probands more likely to develop unipolar illness (MD, in other words) (Perris, 1966a). With few exceptions, subsequent studies have confirmed this observation: bipolar illness aggregates in the families of bipolar probands much more than in families of unipolar probands (Weissman et al., 1984). However, it is also true that in comparison to the general population, relatives of both bipolar and unipolar probands have increased risks of both forms of affective disorder (Gershon et al., 1982, Lieb et al., 2002, Weissman et al., 1984). The risk for bipolar disorder in relatives of MD probands is only modestly increased, approximately 2-fold across studies (on a relative risk scale) (Tsuang and Faraone, 1990). Conversely, there is about a 3-fold increase in risk of developing unipolar depression for a first-degree relative with bipolar disorder. Note that the base rates of unipolar and bipolar illnesses are very different: about 1% for
