Abstract/Summary

Postnatal mental health is often assessed using self-assessment questionnaires in epidemiological research. Differences in response style, influenced by language, culture and experience, may mean that the same response may not have the same meaning in different settings. These differences need to be identified and accounted for in cross-cultural comparisons. We describe the development and application of anchoring vignettes to investigate the cross-cultural functioning of the Edinburgh Postnatal Depression Scale (EPDS) in urban community samples in India (n = 828) and the UK (n = 549), alongside a UK calibration sample (n = 226). Participants completed the EPDS and anchoring vignettes when their children were 12-24 months old. In an unadjusted item-response theory model, UK mothers reported higher depressive symptoms than Indian mothers (d = .48, 95% CI: .358, .599). Following adjustment for differences in response style, these positions were reversed (d = -.25, 95% CI: -.391, -.103). Response styles vary between India and the UK, indicating a need to take these differences into account when making cross-cultural comparisons. Anchoring vignettes offer a valid and feasible method for global data harmonisation.