Background

Maternal psychological distress (depression, anxiety, stress, “blues”) and difficulties in mother–infant bonding have each been linked to adverse child outcomes. Although many studies report co-occurrence of distress and bonding problems, a comprehensive quantitative synthesis had been lacking. This meta-analysis estimates the strength of these associations and examines whether effects vary by distress domain, timing across the first postpartum year, and prior (pre-birth) distress history.

Methodology

The authors followed established international guidelines (PRISMA and MOOSE) for systematic reviews and meta-analyses. They began by searching six major research databases — including MEDLINE, PsycINFO, CINAHL, and Embase — along with doctoral dissertations and grey literature, to find any studies that explored links between a mother’s psychological distress and how she bonded with her infant. Only studies published in English and reporting quantitative data were included.

To be eligible, a study had to measure both maternal distress (such as depression, anxiety, stress, or “baby blues”) and mother–infant bonding within the first year after birth. The distress and bonding measures had to come from self-report questionnaires rather than clinician ratings, so that effect sizes could be compared across studies. The researchers excluded case reports, qualitative studies, and literature reviews.

Once they collected the eligible studies, the team extracted the results into a single dataset and converted every association into a common statistic (the correlation coefficient r), which allows results from different scales and samples to be compared. Because most studies did not statistically adjust for other variables, the authors used these unadjusted correlations to capture the overall relationship between distress and bonding across a wide range of contexts.

Altogether, the review covered 133 studies representing more than 119,000 mothers worldwide. Nearly all studies were published since 2011, and most came from high-income countries. The most common measures were the Edinburgh Postnatal Depression Scale (EPDS) for depressive symptoms and the Postpartum Bonding Questionnaire (PBQ) or Maternal Postnatal Attachment Scale (MPAS) for bonding.

The authors used a random-effects meta-analysis, a statistical approach that assumes differences among studies (for example, in culture, timing, or measurement). They also looked at whether certain factors influenced the results — such as the type of distress (depression vs. anxiety), when bonding was measured (early or late postpartum), and which questionnaire was used. Finally, they checked for publication bias using visual inspection (funnel plots) and statistical tests to make sure results weren’t skewed toward studies showing stronger effects.

Findings

• Overall association. Maternal psychological distress showed a strong aggregate association with impaired mother–infant bonding (overall r ≈ .38).
• By domain (postpartum, concurrent). Associations ranged from moderate to very strong, with the largest for depression (typical r ≈ .41–.47 after the first postpartum week); anxiety and stress also showed moderate–strong links.
• Timing across the first year. For depression, the distress–bonding association was smaller in the first postpartum week than at later time points; timing did not significantly moderate anxiety or stress effects.
• Measurement differences. Effect sizes depended on the bonding instrument: associations were smaller with the MIBS and larger with MPAS/PBQ, suggesting instruments capture somewhat different facets of bonding.
• Antenatal distress → postpartum bonding. Antenatal depression, anxiety, and (to a lesser extent) stress showed small-to-moderate prospective associations with later bonding problems, often with wide confidence intervals (especially for stress).
• Pre-conception distress history. In two longitudinal cohorts, persistent depressive/anxiety symptoms across adolescence and young adulthood predicted poorer subsequent bonding (d = −0.34; r ≈ −.17). Effects attenuated, but persisted, after adjusting for concurrent distress.
• Publication bias & robustness. No evidence of publication bias; effects largely robust across sample type, parity, country income, and publication type. An informal 2023 search update yielded consistent direction and magnitude.

Authors’ Stated Limitations

• Unadjusted effects only were synthesized; causal inferences and mechanistic pathways cannot be established from these correlations.
• Directionality was not examined (i.e., whether impaired bonding might increase maternal distress).
• Limited studies on anxiety, stress, blues, and pre-conception exposures reduce precision and generalizability for those domains.
• Sampling/attrition risks in primary studies (low recruitment and higher attrition in longitudinal work).
• English-language restriction may bias estimates and limit generalizability.

Additional Considerations

• Potential misclassification/construct heterogeneity. “Bonding” was operationalised with multiple instruments tapping overlapping but not identical constructs; differences in valence, cognitive vs affective content, and recall window may inflate heterogeneity and complicate cross-study comparability (e.g., MPAS/PBQ vs MIBS).
• Shared method variance. Many associations rely on self-report scales completed at similar time points, which can inflate correlations via common-method bias. (This is consistent with the authors’ unadjusted-effects approach but is not directly quantified.)
• Clinical overlap/confounding. Early postpartum “blues” symptomatology overlaps with depression screening content, potentially diluting or distorting week-1 estimates; similarly, infant medical status (e.g., NICU) and social context (partner support, IPV, SES) are probable unmeasured confounders in many primary studies.
• Cultural/contextual limits. ~90% of studies came from high-income countries; applicability to low-income contexts and to racially/ethnically marginalised groups requires caution.

Practice Implications

• Screen both, not just mood. Routine perinatal screening typically targets depression/anxiety; these findings support adding a brief, validated bonding instrument (e.g., PBQ/MPAS)—particularly after the first postpartum week—when mood symptoms are present.
• Distress ≠ bonding difficulty (and vice versa). Co-occurrence is common but not guaranteed; assessment and care plans should differentiate targets (mood vs relational bond) and consider combined interventions when both are present.
• History matters. A persistent pre-conception distress history flags elevated bonding-risk; preconception/pregnancy care pathways may warrant proactive mental-health support.

Citation

O’Dea, G. A., Youssef, G. J., Hagg, L. J., Francis, L. M., Spry, E. A., Rossen, L., Smith, I., Teague, S. J., Mansour, K., Booth, A., Davies, S., Hutchinson, D., & Macdonald, J. A. (2023). Associations between maternal psychological distress and mother–infant bonding: A systematic review and meta-analysis. Archives of Women’s Mental Health, 26(4), 441–452. https://doi.org/10.1007/s00737-023-01332-1