Mother-to-Infant Bonding in Women with Postpartum Psychosis

Mother-to-Infant Bonding in Women with Postpartum Psychosis and Severe Postpartum Depression: A Clinical Cohort Study

Mother-to-infant bonding can impact a child’s long-term emotional, cognitive, and behavioral development. Affective and protective feelings toward a child are established during the first trimester and grow throughout pregnancy, especially in response to fetal movements. After delivery, bonding progresses gradually after birth and lays the foundation for the child’s later attachment. Impaired bonding, which can occur when a mother suffers from a psychiatric disorder, can create high levels of parenting stress for the mother and put her child at risk for later maltreatment. It has also been associated with child behavioral problems and an increased likelihood of the child developing psychopathology as an adult.

The postpartum period is a time of heightened risk for the development of psychiatric disorders; postpartum depression (PD) affects approximately 10% of mothers. Postpartum psychosis (PP) occurs in fewer than 1 per 1000 women after giving birth, although it has a 17% prevalence among women with a prior diagnosis of bipolar disorder. Severe cases of both PD and PP can be treated with psychiatric hospitalization, and joint mother-baby admission to a Mother and Baby Unit (MBU) is preferrable, given evidence of reduced time to recovery when the mother is admitted with her baby.

Gilden, Molenaar, and colleagues conducted a study examining bonding in women admitted to an MBU with postpartum onset of psychosis, mania, or severe depression. Bonding was measured using a self-report instrument called the Postpartum Bonding Questionnaire (PBQ), which assesses four different categories, namely 1) impaired bonding, 2) rejection and anger, 3) anxiety about care, and 4) risk of abuse. On the PBQ, a score between 26 and 39 identifies mothers with medium impaired bonding, while a score of 40 or above signals severe impaired bonding. Depressive symptoms were measured with the Edinburgh Postnatal Depression Scale (EPDS), on which a score above 12 indicates a depressive disorder. Mania was measured using the Young Mania Rating Scale (YMRS), an 11-item, interviewer-rated assessment designed to measure the severity of symptoms in bipolar patients. On the YMRS, a score of 12 or below indicates that the patient is euthymic.

Throughout hospitalization, these three assessments were completed weekly. Patients with PP were treated with antipsychotic medications and/or lithium, and those with PD were given either tricyclic antidepressants or a selective serotonin reuptake inhibitor (SSRI). The women were also treated with non-pharmacological interventions, including a support group, direction from nursing staff, and video interaction guidance, strategies given to optimize mother-baby interaction. Women with PD and PP were exposed to these interventions equally.

Of the participants in the study, 91 were diagnosed with PP, while 64 were diagnosed with PD. Depressive symptoms, based on the results of the EPDS, were higher among women with PD. Between the two groups, there were not any substantial differences in psychiatric history, maternal age, length of admission, relationship stats, education level, or infant gender.


The percentage of patients with medium to severe impaired bonding, indicated by a PBQ score above 26, was 17.6% in women with PP and 57.1% in women with PD upon admission (p-value < 0.001). At the time of discharge, only 5.9% of women with PP and 18.2% of women with PD still had a PBQ score above 26 (p-value = 0.02). At admission, 5.9% of patients with PP and 37.5% of patients with PD had a PBQ score above 40, which is associated with severe impaired bonding (p-value < 0.001). These percentages had decreased to 1.2% in women with PP and 5.5% in women with PD by discharge (p-value = 0.30). For both PP and PD patients, PBQ scores declined gradually during the first five weeks of MBU admission, but did not continue to go down further through eight weeks of admission. Formal statistical testing showed that the course of bonding problems while at the MBU differed significantly between those with PP and those with PD.

Upon admission, the mean EPDS score was 19.0 in the PD group and 14.9 in the PP group. These scores gradually decreased to 10.3 and 4.1 respectively over the course of treatment. Among the women with PP, manic symptoms also declined gradually, from a mean YMRS score of 17.7 at admission to one of 1.1 at discharge. There was a substantial association between reduction in depressive symptoms and improved bonding in patients with PD over the admission period, while a slightly weaker association was found between reduction in depressive symptoms and improved bonding in patients with PP. There was also a correlation between a decrease in manic symptoms and an increase in bonding in the PP group, though impaired bonding was found to be more closely associated with depressive symptoms than manic ones.


Impaired mother-to-infant bonding was found to be a major problem in inpatient mothers with severe PD (affecting 57.1% of participants in this group), while only 17.6% of participants with PP reported impaired bonding. Additionally, there was found to be a higher proportion of women with severe impaired bonding in the PD group than in the PP group. Previous research has reported that depressed mothers perceive bonding with their baby more negatively than psychotic mothers do, and that this discrepancy is an expression of the natures of the two disorders; mothers suffering from PD tend to have negative cognitions with feelings of inadequacy and self-doubt, while most mothers with PP do not have these patterns.

Studies on impaired bonding have shown prevalence rates in the general population of 12.2% at 48 hours postpartum, 7.1% at 2 weeks postpartum, and 8.9% at 12 weeks postpartum. In this study, the percentage of participants with PP who had impaired bonding had been reduced to 5.9%, and could be considered equivalent to that of the general population. The results of this study were therefore encouraging, as the combined pharmacological and non-pharmacological treatment significantly improved postpartum symptoms and mother-to-infant bonding. However, a limitation was that the researchers could not establish differences in effectiveness between the types of interventions given.

Many mothers with mild to moderate symptoms of depression do not have the opportunity for treatment in an intensive eight-week MBU, and their postpartum disorders may go completely undetected or unaddressed. This could lead to inadequate mother-to-infant bonding, putting children at risk for long- and short-term adverse outcomes. The findings of this study thus maintain the importance of establishing MBUs and other accessible programs tailored to the postpartum experience. Since bonding begins in the first trimester and prenatal depression is also very prevalent, the researchers emphasize that diagnoses and interventions should be given as early as possible within the perinatal period, which is unfortunately not yet standard clinical practice everywhere.

Written by Margaux Kanamori

Read the full article here.

Gilden, J.; Molenaar, N.M.; Smit, A.K.; Hoogendijk, W.J.G.; Rommel, A.-S.; Kamperman, A.M.; Bergink, V. Mother-to-Infant Bonding in Women with Postpartum Psychosis and Severe Postpartum Depression: A Clinical Cohort Study. J. Clin. Med. 2020, 9, 2291.


185 Cambridge St
Boston, MA, USA 02114

  • Facebook Social Icon

©2018 by Massachusetts General Hospital Postpartum Psychosis Project (MGHP3).