The Japanese version of the mother-to-infant bonding scale (MIBS-J), a self-report of postpartum bonding disorder, is used in clinical settings for screening postpartum bonding disorder at various time points. However, its psychometric properties, particularly measurement invariance, have rarely been reported, and the validity of comparing scores across time points and sex is unclear. We aimed to select and validate the MIBS-J items suitable for parents at three time points. Postpartum mothers (n = 543) and fathers (n = 350) were surveyed at 5 days, 1 month, and 4 months postpartum. All participants were randomly divided into two subgroups, one for exploratory factor analyses (EFAs) and another for confirmatory factor analyses. Measurement invariance of the best model was tested using the entire sample, between fathers and mothers, and across the three observation periods. A three-item structure (items 1, 6, and 8) extracted through EFAs showed acceptable configural invariance. This model was accepted for scalar invariance between fathers and mothers and for metric invariance across the three time points. Our findings suggest that the three-item MIBS-J is sufficient for diagnosis of postpartum parental bonding disorder through continuous observation for at least 4 postpartum months, in order to detect the priority of parents who need support.
Keywords: bonding disorder; factor analysis; longitudinal study; mother-to-infant bonding scale; parents; postpartum period; psychometrics; questionnaire.