Risk factor profiles for depression following childbirth or a chronic disease diagnosis: case–control study

Abstract

Background Progress towards understanding the aetiology of major depression is compromised by its clinical heterogeneity. The variety of contexts underlying the development of a major depressive episode may contribute to such heterogeneity. Aims To compare risk factor profiles for three subgroups of major depression according to episode context. Method Using self-report questionnaires and administrative records from the UK Biobank, we characterised three contextual subgroups of major depression: postpartum depression (3581 cases), depression following diagnosis of a chronic disease (409 cases) and a more typical (named heterogeneous) major depression phenotype excluding the two other contexts (34 699 cases). Controls with the same exposure were also defined. We tested each subgroup for association with the polygenic risk scores (PRS) for major depression and with other risk factors previously associated with major depression (bipolar disorder PRS, neuroticism, reported trauma in childhood and adulthood, socioeconomic status, family history of depression, education). Results Major depression PRS was associated with all subgroups, but postpartum depression cases had higher PRS than heterogeneous major depression cases (OR = 1.06, 95% CI 1.02–1.10). Relative to heterogeneous depression, postpartum depression was more weakly associated with adulthood trauma and neuroticism. Depression following diagnosis of a chronic disease had weaker association with neuroticism and reported trauma in adulthood and childhood relative to heterogeneous depression. Conclusions The observed differences in risk factor profiles according to the context of a major depressive episode help provide insight into the heterogeneity of depression. Future studies dissecting such heterogeneity could help reveal more refined aetiological insights.

Assessing the validity of a self-reported clinical diagnosis of schizophrenia

“…However, the validity of self-reported diagnoses is likely to differ between psychiatric disorders, contexts, and cultures. Large-scale genomic datasets, such as 23andMe [12][13][14][15][16][17] , UK Biobank 13,16,[18][19][20][21] and the Million Veterans Programme 16,22 , have exploited self-report methods, but the reliability and validity of self-reported diagnoses is unclear 23 .…”