Factors associated with the development of postnatal depression at 6 months postpartum in young mothers: a cohort study.
Various psychological, social, and obstetric factors are known to predispose a mother to the development of postnatal depression (PND), but there is limited research into the factors that may be important in the development of PND in an adolescent population. Current literature estimates that the prevalence in mothers is between 10 and 15%; in the adolescent population the prevalence rate has been found to be as high as 37%. PND can result in a number of adverse outcomes for both mother and child, and it is important to identify the women at risk in order to try and prevent the development of the disorder.
The aim of this study was to identify which factors are associated with the development of PND. The study utilised a large data set that was collected as part of the Building Blocks trial. Data was collected from 1618 pregnant young women aged 13-19 years old during a baseline interview (prior to 24+6 weeks gestation) and a subsequent interview at 6 months postpartum, at which time they completed the Edinburgh Postnatal Depression Scale (EPDS).Multivariable logistic regression analyses were conducted for caseness cut points and a linear model for the continuous EPDS.
The prevalence of PND in this cohort was 12.0% when the standard cut-off of >13 (validated for the adult population) on the EPDS was used. When a cut-off of ≥9 (revised for use in the adolescent population) was used, prevalence was found to be 30.7%. Pregnant young women who report having been homeless (OR 1.95, p<0.001), not having good social support (OR 2.00, p = 0.049) and continuing to use alcohol after discovering they were pregnant (OR 1.69, p=0.028) were at higher risk of developing PND at 6 months postpartum, whilst those with higher self-efficacy (OR 0.94, p=0.012) were less likely to have PND utilizing the traditional cut-point on EPDS. For the lower cut-point, self-efficacy (OR 0.91, p<0.001) and alcohol use during pregnancy (OR 1.50, p=0.037) remain significant and prior use of drugs (OR 1.60, p=0.008) is also a predictor. The linear model included the same factors as the lower cut-point model, with the addition of history of homelessness.
The predictors of postnatal depression in an adolescent population are different to those reported in the adult population and as such health workers should be aware of the possible predictors of depression and in order to identify those most at risk. Further work should consider if the factors identified represent unidentified mood disorder during pregnancy or if they are true risk factors themselves that could be targeted to reduce PND.