Karin Gidén: Beyond the early weeks: Remission and persistence of postpartum depressive symptoms

Abstract

Peripartum depression is a common complication in the perinatal period, affecting approximately 17% of women worldwide. Many women (30–50%) continue experiencing long-term symptoms. The aim of this thesis was to investigate the course of postpartum depressive symptoms measured with the Edinburgh Postnatal Depression Scale (EPDS), the treatment provided following positive screening, and factors associated with remission or persistence of symptoms at six months postpartum.

Paper I (n = 593) was a descriptive nested cohort study examining identification, interventions received, and remission status at six months postpartum among women scoring ≥12 on the EPDS. Approximately 59% of women with depressive symptoms were not identified by the healthcare system. Among those identified, around 90% received some form of intervention, and symptom reduction was greatest in this group. Remission rates at six months postpartum were 30–50% depending on identification.

Paper II (n = 654) was a prediction model study including women with early postpartum depressive symptoms (EPDS ≥12) from a large longitudinal cohort. Cohort data and medical record information were analyzed to develop a model identifying women at risk of persistent depressive symptoms at six months postpartum. The model demonstrated high specificity (92%) and moderate sensitivity (48%) at an EPDS cut-off of 11/12 at six months postpartum. The area under the curve was 0.73, indicating acceptable discriminative ability.

Paper III (n = 206) was an external validation study of the prediction model in an independent longitudinal cohort. Predictive performance was slightly lower but remained acceptable, maintaining high specificity and strong reliability in positive classifications.

Paper IV (n = 18) was a qualitative descriptive interview study exploring perceived barriers and facilitators linked to remission. Five themes were identified: (1) Others take responsibility; (2) Practical support; (3) Emotional validation; (4) Thresholds; and (5) Struggling to prioritize oneself.

In conclusion, many women with postpartum depressive symptoms remain undetected. However, when identified, most receive treatment in Sweden today. This thesis presents a clinically applicable prediction model for identifying women at risk of persistent symptoms and demonstrates that recovery depends not only on effective treatment, but also on social, relational, and structural factors influencing access to support and remission.