A longitudinal study of postnatal dysphoria
Background: Postnatal depression (PND) follows 10-15% of deliveries. Postnatal blues occur in the first postpartum week and are thought to have little significance. Studies report links between them, PND and premenstrual symptomatology but are methodologically flawed. Hypotheses: Women with severe blues are more likely to become depressed in the 6 months after delivery and more likely to experience premenstrual symptomatology when menstruation resumes. Subjects: First-time mothers who were literate, English speaking, had a singleton pregnancy, no current severe mental illness and whose fetus was healthy. Method: Written informed consent was obtained in late pregnancy. Baseline data include the Eysenck Personality Questionnaire, Edinburgh Postnatal Depression Scale (EPDS), sociodemographic and obstetric details. The Blues Questionnaire was completed on postpartum days 3 & 5. Obstetric data were recorded. Subjects (scores >/75th centile on the Blues Questionnaire) and controls (\<25th centile) were matched for age, marital status and social class. All participants completed monthly postal EPDS. When menstruation returned, daily Menstrual Distress Questionnaires and visual analogue scales for premenstrual symptoms were completed for 2 cycles. At 6 months all women with EPDS scores >/9 at any time postpartum were interviewed using the Schedule for Affective Disorder and Schizophrenia (Lifetime version). Research Diagnostic Criteria diagnoses were made for current or past psychiatric disorder. 1 in 5 women with EPDS scores <9 were interviewed to exclude false negatives. Results: Women with severe blues were 3.8 times more likely to become depressed and to have a major rather than minor illness. Their depressions began sooner after delivery and lasted longer than those with no blues. They were more likely to experience premenstrual symptoms. Discussion: Results support the idea that these conditions are variants of affective disorder, severe blues acting as a marker of affective vulnerability. Clinical applications of the results and areas for further research are explored.