Psychometric properties of the Maternal Postnatal Attachment Scale and the Postpartum Bonding Questionnaire in three German samples

Abstract

Background

Forming an emotional bond towards the infant is an important developmental aspect of the mother-child relationship. Two questionnaires frequently used for the assessment of mother-infant bonding, namely the Maternal Postnatal Attachment Scale (MPAS) and the Postpartum Bonding Questionnaire (PBQ), have shown inconclusive psychometric properties. To ensure comparability of results across studies, it is crucial to examine the replicability of psychometric properties and previously proposed factor structures of measurements when adapted to other languages.

Aim

The study aim was to investigate the psychometric properties of the German versions of both MPAS and PBQ, across three different German-speaking study samples.

Methods

Maternal data from three longitudinal studies from Hamburg, Germany (PAULINE-PRINCE study, N = 229), and Dresden, Germany (MARI study, N = 286; DREAM study, N = 1,968), were used to investigate the psychometric properties (descriptive statistics, item difficulty, inter-item correlations) and the factorial structure (confirmatory factor analysis, CFA; principal axis factoring, PAF) of both MPAS and PBQ. Correlations with maternal-fetal bonding, adult romantic attachment style, attachment style to one’s own mother, postpartum depressive symptoms, and education level were investigated.

Results

Across the three samples, both MPAS and PBQ showed convincing results regarding the psychometric properties for their total scores, with satisfying to excellent internal consistencies. A strong correlation between the MPAS and PBQ total scores was observed (r=-.71, p < .001). In PAF, for both questionnaires, factor structures on subscale level differed across samples and assessment points. For MPAS and PBQ total scores, significant small to medium-sized associations in the expected directions with maternal-fetal bonding and depressive symptoms, as well as for MPAS with adult romantic attachment style, and for PBQ with attachment towards one’s own mother were found. In two samples, higher educated participants reported less optimal MIB.

Conclusion

The results across the three included samples provide evidence for the validity of the construct assessed with the German adaptations of both MPAS and PBQ. However, the factor analytical results on subscale level highlight the need to further investigate the concept of mother-infant bonding in the first year after birth as well as to develop instruments applicable for use in clinical and community samples with satisfying psychometric properties.