Abstract
BackgroundHistory of psychiatric disorders, particularly depression, is the strongest risk factor for perinatal depression (PND). Yet many women without such history experience their first depression episode perinatally, whilst other women with depression history do not experience any episodes during the perinatal period. PND may itself be heterogenous, according to differences in psychiatric history. However, a comprehensive investigation of characteristics of women with PND, with and without a prior psychiatric history, has not been attempted.MethodsIn a large sample of parous women with depression, we sought to identify risk factors associated with PND after previous depression episodes, or as first-onset depression. Using data from the Australian Genetics of Depression Study, we identified two subgroups of PND cases (Edinburgh Postnatal Depression Scale score >= 13) with and without prior depression history. For both subgroups, we investigated lifetime prevalence, length and severity of PND. Logistic regression compared a range of characteristics of cases to those of a comparison group with major depression without any perinatal episodes. ResultsCriteria for PND was met by 5,058 (70%) of 7,182 parous women who met criteria for major depression. Of women reporting depression onset before first pregnancy, 2,261 (77%) PND cases were compared to 672 (23%) without PND. Among women reporting their first depression episode during or after their first pregnancy, 878 women for whom this first episode was PND were compared to 2,124 parous women who had experienced depression but never perinatally. Of women who experienced depression prior to first pregnancy, PND cases were significantly more likely to report more episodes of depression (OR=1.1 per additional depression episode, CI=[1.1-1.1], P=1.3E-13), non-European ancestry (OR=1.8, CI=[1.3-2.5], P=1.2E-03), severe nausea during pregnancy (OR=1.3, CI=[1.1-1.6], P=6.6E-03) and emotional abuse (OR=1.4, CI=[1.1-1.7], P=2.0E-03). Women without any depression before their first perinatal episode were significantly more likely to report emotional abuse (OR=1.3, CI=[1.1-1.6], P=1.1E-02) than women with depression without PND.ConclusionsThe majority of parous women in this study experienced PND, associated with more complex, severe depression. Results highlight the importance of perinatal assessments of depressive symptoms, particularly for women with a history of depression or childhood adverse experiences.
Are the lifetime prevalence estimates in the ECA study accurate?
