A test of the tripartite model of anxiety and depression in postpartum psychiatric inpatients

The aims of the present thesis were to (a) examine the psychometric integrity in a postpartum inpatient population of two measures, the Edinburgh Postnatal Depression Scale (EPDS; Cox, Holden, & Sagovsky, 1987) and the Depression Anxiety Stress Scales (DASS; S. H. Lovibond & P. F. Lovibond, 1995); (b) examine whether the tripartite model – a conceptual model of anxiety and depression previously supported in non-postpartum populations – provided an optimal fit to self-reported anxiety and depression data in a postpartum clinical sample; and (c) examine the relationship of the factors underlying postpartum anxiety and depression symptoms to diagnoses of unipolar mood disorders, anxiety disorders, and disorders characterised by mixed anxiety and depressive symptoms. These aims were tested using self-report anxiety and depression data from questionnaires completed by inpatients with postpartum psychiatric disorders.

An initial aim was to determine whether two self-report questionnaires were suitable for measuring depression and anxiety in the postpartum inpatient sample. The EPDS was considered given evidence that it measures both anxiety and depression. However, its factor structure has varied across previous studies, and has not previously been examined in a postpartum inpatient sample. Thus, Chapter 2 aimed to examine the factor structure of the EPDS in the postpartum inpatient sample. The factor structure of the EPDS was not consistent across admission and discharge, which was in line with previous literature indicating that the EPDS lacks factorial invariance. However, it was unclear whether the non-invariant factor structure was specific to the EPDS, or whether it was characteristic of the postpartum inpatient sample more generally.

Therefore, Chapter 3 aimed to examine the psychometric properties and factor structure of the DASS, a self-report measure previously found to have a robust, replicable factor structure. In this study, the same three factor structure identified previously in other populations provided the best fit in the postpartum sample, both at hospital admission and discharge. These findings suggested that the DASS was reliable and valid in the inpatient sample, and suggested that the variability of the EPDS factor structure was specific to the EPDS, rather than due to the nature of distress in the postpartum sample. Furthermore, the results provided evidence that the latent structure of anxiety and depression symptoms in postpartum inpatients is similar to that observed in non-postpartum populations.

Chapter 4 turned to the second objective of the thesis: to test whether the tripartite model provided an adequate fit to anxiety and depression data (collected at hospital admission) in a sample of postpartum psychiatric inpatients. Confirmatory factor analyses revealed that a three factor model consistent with the tripartite model (comprising factors reflecting negative affect, positive affect, and autonomic arousal) provided an adequate fit to the data, and was superior to alternative, simpler models. Consistent with the tripartite model, the factors were differentially related to depression, anxiety, and mixed anxiety/depressive diagnoses.

In conclusion, the present thesis provided evidence for validity of the DASS in a postpartum inpatient sample, and extended the utility of the tripartite model of anxiety and depression to a postpartum inpatient sample. In addition to demonstrating a tripartite symptom structure consistent with that observed in non-postpartum populations, the present study also found that anxiety, depressive, and mixed diagnoses were differentially associated with the tripartite model symptom dimensions. The present thesis demonstrated that the structure of anxiety and depressive symptoms in postpartum psychiatric inpatients is similar to that at other life stages, consistent with the view that postpartum anxiety and depression are the same clinical entities as non-postpartum anxiety and depression. The results from the three studies contained in the thesis are summarised in the General Discussion (Chapter 5) and are interpreted in the context of the wider literature on postpartum (and non-postpartum) anxiety and depression. Implications and future directions for the measurement and treatment of postpartum emotional disorders are also discussed.