Past wounds, present bonds, future mothers: item-level analysis of perinatal risk profiles

Perinatal depression remains one of the most prevalent yet complex mental health complications associated with childbearing, affecting a significant percentage of women during pregnancy and the postpartum period. While traditional clinical assessments have historically relied on isolated risk indicators and variable-centered statistical models, a pioneering prospective cohort study conducted by researchers in Southern Italy reveals that a person-centered, item-level approach can more accurately capture the multifaceted nature of maternal vulnerability. Published in Frontiers in Psychology, the study offers a granular examination of how early life adversity, interpersonal dynamics, social support, and prenatal attachment intersect to dictate maternal mental health outcomes across the perinatal timeline.

Uncovering Hidden Heterogeneity Through Advanced Methodology

Conducted between January 2024 and March 2025, the observational study was structured around a clinical screening initiative at the ARNAS "Civico–Di Cristina–Benfratelli" Hospital in Palermo, home to one of the largest high-volume obstetric units in Southern Italy. The core objective was to move beyond generalized screening paradigms that aggregate broad scores, opting instead for an item-level $k$-means cluster analysis across 44 discrete psychometric variables.

The investigation initially enrolled a baseline cohort of 774 pregnant women during their second or third trimesters. These participants completed a comprehensive digital protocol utilizing validated metrics including the Edinburgh Postnatal Depression Scale (EPDS), the Childhood Trauma Questionnaire-Short Form (CTQ-SF), the Dyadic Adjustment Scale-7 (DAS-7), the Maternity Social Support Scale (MSSS), and the Maternal Antenatal Attachment Scale (MAAS). To evaluate the persistence of these psychological trajectories, a subset of 200 women completed a 6-month postpartum follow-up assessment.

By examining responses at the individual item level rather than relying solely on summarized subscale aggregates, the analytical team bypassed the statistical blind spots common in standard psychiatric epidemiology. This methodological shift allowed researchers to identify specific behavioral configurations and qualitative distinctions among participants that would otherwise be obscured by aggregate scoring.

Chronology and Dataset Demographics

The institutional approval process for the study began with clearances secured from the Bioethics Committee at the University of Palermo and the ethics board at the Civico-Di Cristina-Benfratelli Hospital in late 2022 and late 2023. Consecutive recruitment of eligible pregnant women commenced in January 2024, running continuously for 15 months to capture seasonal variations.

The baseline demographic profile of the 774 participants reflected a broad cross-section of the local population. The mean age of the cohort was 32.65 years, with the vast majority (89.7%) being married or cohabiting. Approximately half of the participants (50.1%) reported stable employment, while 35.0% held a high-school diploma, 26.1% a bachelor’s degree, and 21.3% postgraduate qualifications. Clinical indicators showed that 95.2% of participants were assessed during their third trimester, 72.7% reported planned pregnancies, and 26.4% were classified as experiencing a high-risk pregnancy.

Following the initial antenatal screening (T0), participants who reached 6 months postpartum (T1) were contacted via telephone for follow-up evaluation. Ultimately, 200 valid matched questionnaires were retained for longitudinal analysis, providing a retention rate of 25.8%. Rigorous attrition analyses confirmed that non-response was primarily driven by general service engagement factors—such as marital status and attendance at antenatal classes—rather than baseline severity of depressive symptoms or cluster assignment.

Three Distinct Psychosocial Profiles

The item-level cluster analysis successfully categorized the pregnant cohort into three distinct psychosocial profiles, each demonstrating unique configurations of risk and resilience:

1. Low-Risk Profile (53.7% of the cohort, $n = 416$): Characterized by minimal exposure to childhood trauma, optimal perceived social support, high dyadic adjustment, and strong positive prenatal attachment. Women in this group consistently recorded low depressive symptom scores, remaining well below clinical thresholds.
2. Moderately Adjusted Profile (31.7% of the cohort, $n = 245$): Exhibited intermediate levels of early adversity paired with the least favorable current relational resources at the scale level. This group reported the lowest dyadic adjustment scores and the weakest total antenatal attachment, alongside suppressed perceived social support.
3. High-Adversity Profile (14.6% of the cohort, $n = 113$): Marked by profound exposure to childhood trauma—particularly emotional and physical neglect and abuse—combined with low perceived social support. Interestingly, while their developmental trauma burden was severe, their current couple functioning sat at an intermediate level compared to the other groups.

Cross-sectional analysis at baseline revealed significant differences in depressive symptoms across the three profiles ($F(2, 771) = 62.52, p < 0.001$). These differences remained robust even after adjusting for sociodemographic and obstetric covariates. Furthermore, the proportion of women scoring above the standard Italian clinical cut-off for perinatal depression (EPDS $ge 9$) escalated sharply from 11.1% in the Low-Risk group to 35.9% in the Moderately Adjusted group, and reached 44.2% among those in the High-Adversity profile.

Longitudinal Persistence and Clinical Implications

To evaluate whether antenatal vulnerability profiles forecasted postpartum outcomes, the researchers utilized a mixed repeated-measures ANOVA on the longitudinal subsample. The analysis revealed a prominent main effect of cluster membership ($F(2, 197) = 22.74, p < 0.001$), demonstrating that between-group differences in depressive symptom severity persisted stably across the perinatal period. Neither the main effect of time nor the cluster-by-time interaction achieved statistical significance, indicating parallel symptom trajectories where at-risk profiles maintained elevated distress levels from pregnancy through six months postpartum.

These empirical findings carry substantial implications for modern obstetric care and mental health screening protocols. Lead researchers and affiliated clinical stakeholders emphasize that standard psychological screening during pregnancy must evolve beyond simple symptom checklists. Because aggregate scores frequently mask underlying structural deficits—such as a lack of shared life goals within a partnership or defensive, non-affective cognitive personification of the fetus—clinicians must adopt trauma-informed, multidimensional assessment frameworks.

The identification of these distinct profiles paves the way for stratified preventative interventions. For patients exhibiting a Moderately Adjusted profile, short-term couple-focused therapies aimed at improving relational communication and shared alignment may yield substantial clinical benefits. Conversely, individuals presenting with a High-Adversity profile require specialized, trauma-informed psychological support designed to process early developmental neglect, mend relational security, and foster genuine affective attunement between mother and child.

Ultimately, the study underscores that early developmental trauma and current psychosocial environments act as joint architects of maternal mental health. By recognizing these risk configurations prior to delivery, healthcare systems can deploy targeted resources to buffer vulnerable families, mitigating long-term psychiatric risks for both mothers and their developing children.