Prenatal bonding linked to late-pregnancy EEG coherence

By Olivia Hart

A resting-state scalp electroencephalography (EEG) study of 35 pregnant women found that stronger maternal-fetal attachment, particularly anticipation and interaction, tracked with greater alpha-band coherence between specific maternal cortical regions. Mean gestational age was 36.8 weeks, and the association with left frontoparietal and intra-frontal synchrony persisted after adjustment for gestational age and parity.

Kathy Ayala and Helena Rutherford of Yale University, with colleagues, conducted the study. It appeared as “Bonding before birth: Maternal–fetal attachment and maternal cortical connectivity in late pregnancy” in the Journal of Neuroendocrinology in 2026. Its DOI is 10.1111/jne.70251. The recordings came from the mothers; the report did not include fetal EEG.

Pregnancy entails psychological and neurobiological adaptation during the transition to motherhood and development of caregiving behavior. Maternal-fetal attachment (MFA) describes the emotional bond between a pregnant woman and her unborn baby, expressed through thoughts, feelings and behavior toward the fetus.

Previous work treats MFA as multidimensional, spanning attention to bodily and fetal movement, interaction with the fetus, imagining the baby, anticipating life together and motivation to promote the baby's health and well-being. Earlier studies linked higher MFA during pregnancy to stronger postnatal bonding, more sensitive maternal behavior and more responsive mother-infant interactions after birth.

Despite decades of research, the neurobiological basis of MFA remained poorly understood. The study did not evaluate an FDA-cleared or investigational device, define a screening protocol or test a treatment; it examined an observational association in humans.

Each participant underwent resting-state scalp EEG, which recorded maternal cortical activity without an active task. The researchers calculated alpha-band coherence within frontal regions, within parietal regions and between frontoparietal regions.

Coherence, a measure of coordinated activity between recorded cortical regions, estimates functional synchrony rather than directly imaging anatomical connections. The measure had been associated with cognitive, attentional and emotional processes relevant to parenting. Researchers therefore tested whether regional coherence covaried with questionnaire-reported attachment.

Attachment was assessed with the Prenatal Attachment Inventory-Revised (PAI-R), which covers three dimensions:

• Anticipation
• Interaction
• Differentiation

Across the broader construct, relevant behaviors included imagining the baby, anticipating life together and engaging with the fetus. The reported analysis focused on these subscales rather than treating attachment as one undifferentiated score.

Higher left frontoparietal and intra-frontal alpha coherence was associated with higher PAI-R anticipation and interaction scores. Women who imagined their baby more often, anticipated life together and engaged more with the fetus also displayed greater coordination of activity in those regions.

Those relationships remained after researchers accounted for gestational age and parity, denoting whether a woman had previously given birth. The reported associations concerned anticipation and interaction; no corresponding differentiation or intra-parietal association was identified.

Rutherford said: “Our study suggests that individual differences in maternal brain connectivity are associated with specific aspects of the emotional bond with the fetus.” She described the work as offering new insight into the neurobiological basis of the early mother-child relationship. Because the study reported a concurrent association in a small late-pregnancy sample, it cannot establish whether connectivity causes attachment or attachment changes connectivity, nor predict later postnatal behavior.

For EEG laboratories, the study demonstrates that maternal resting-state scalp activity can be quantified as regional alpha synchrony during late pregnancy. The reported methods did not specify a scalp montage, acquisition settings or artifact-rejection protocol, limiting independent assessment of technical reproducibility. No diagnostic threshold or clinically usable classifier emerged.

What does this change for prenatal care and EEG practice?

At present, it does not change routine prenatal care. The work supplies no validated neural measure for maternal-fetal attachment and no evidence for EEG screening during pregnancy. A clinical EEG reading therefore should not infer a bonding disorder from alpha-band coherence values in a pregnant patient.

For researchers, the result supports treating prenatal attachment as a multidimensional outcome, with alpha coherence focused on anticipatory and interactional engagement rather than a single maternal-fetal bond score. Larger studies could test whether connectivity patterns precede postpartum behavior and whether they persist after accounting for gestational age, parity and additional physiological and psychological factors.

The study provides a candidate neurobiological correlate for longitudinal work on prenatal attachment, not a clinical test. Its clearest contribution is the observed covariation between specific prenatal engagement measures and left frontoparietal and intra-frontal alpha synchrony in late pregnancy.