A multivariate investigation of postpartum mood disturbance.

The interrelationships of 'blues' and later postpartum depression with a number of biochemical, medical, and psychosocial variables have been examined in 52 subjects. The two syndromes shared only an impressive association with a prior history of gynaecological problems. Puerperal 'blues' was characterised in addition by associations with primiparity, tearfulness during pregnancy, and reduced plasma total tryptophan in the early puerperium. Depressive symptomatology up to nine months postpartum was related to an excess of male births and to an altered pattern of decline of non-esterified fatty acids immediately postpartum. In each case, the 'risk' variables were statistically independent and combined linearly. Stepwise discriminant analysis successfully discriminated 'blues' and depression from their respective non-cases. 'Blues and postpartum depression were only weakly related and, apart from gynaecological history, each was associated with separate and independent causative factors.