Identifying distinct subtypes of mother-to-infant bonding using latent profile analysis in a nationwide Japanese study

Abstract

Background

Mother-to-infant bonding (MIB) is foundational for nurturing behaviors and an infant’s development. Identifying risk factors for difficulties or problems in MIB is vital. However, traditional research often dichotomizes MIB using cutoff thresholds, overlooking its underlying complexities. This research utilizes latent profile analysis (LPA) to discern MIB subtypes in a nationwide Japanese dataset.

Methods

We conducted LPA on data from the Mother-to-Infant Bonding Scale (MIBS), collected from 3,877 postpartum women within one year of childbirth. To empirically validate the derived profiles, we examined their associated risk factors, focusing on sociodemographic, health, and perinatal variables.

Results

Four distinct MIB profiles emerged. Profile 1 indicated minimal difficulties, while Profile 4 exhibited severe multifaceted difficulties. Profiles 2 and 3 showed moderate difficulties distinguished by lack of positive affection and presence of negative affection (especially indifference), respectively. Compared to Profile 1, women in Profiles 2–4 had a higher likelihood of postpartum depression and low family support. Each profile also presented unique risk factors: medium family support in Profile 2, maternal working status in Profile 3, and pre-pregnancy underweight status in Profile 4. Notably, both Profiles 3 and 4 were also linked to increased feelings of loneliness since the onset of the COVID-19 pandemic.

Conclusion

This study represents the first application of LPA to MIB, revealing distinct subtypes and their respective risk profiles. These insights promise to enhance and personalize early interventions for difficulties in MIB, affirming the necessity of acknowledging MIB’s heterogeneity.