Skip to main content

Improving prediction of the triggering of bipolar disorder episodes by childbirth

Final Report Summary - POSTPARTUM BID (Improving prediction of the triggering of bipolar disorder episodes by childbirth)

The primary goal of my fellowship was to develop clinically useful predictive models for 1) postpartum mania/psychotic depression and 2) postpartum bipolar non-psychotic depression. To individualise the risk of perinatal recurrences, information was gathered retrospectively by semi-structured interview, questionnaires and case-note review from in 887 multiparae with bipolar disorder. Clinical predictors were selected using backwards step-wise logistic regression, conditional permutation random forests and reinforcement learning trees. The latter is a powerful statistical method developed by collaborators I met during the Outgoing phase of the fellowship at the University of North Carolina.
My research found that the first lifetime episode of bipolar disorder occurs in relation to childbirth for over 1 in 4 women with bipolar disorder). of women. For women with a history of postpartum affective psychosis this proportion was even higher: 57% of them did not have any psychiatric history prior to the postpartum episode, making prediction of first psychosis difficult.
I also found a high risk of a recurrence with further pregnancies. This risk can be stratified according to previous perinatal history. Although the rates were significantly higher in women with previous perinatal psychiatric history, women without such episodes were still at risk of developing perinatal illness. As anticipated, the clinical presentation (affective psychosis versus non-psychotic depression and time of onset in relation to delivery) of the first perinatal episode was associated with the clinical presentation of subsequent perinatal episodes. Although women with postpartum affective psychosis were at the highest risk of developing a further severe postpartum episode, those with a history of perinatal non-psychotic depression had the highest rates of any mood episode recurrence.
Impact: Our study will be of benefit in individualising the risk of perinatal episodes in parous women with bipolar disorder. The research paper (currently under review) included a risk assessment flow chart that can be used for preconception and pregnancy counselling. This will help women and their clinicians make the very difficult decisions they face regarding pregnancy.

Work on postpartum depression: Models for bipolar perinatal recurrence are not replicable in postpartum depression. Results from my research on the PACT consortium dataset, the largest, international dataset available on postpartum depression showed that there might be different types and severity of perinatal depression with varying time of onset throughout pregnancy and post-partum. I was also able to show that the reporting of depression varies across cultural backgrounds and countries, but it is similar across European centres who took part to the research. This research is particularly important for clinical practice as universal screening for postpartum depression is recommended in many countries. Knowledge of whether the disclosure of depressive symptoms in the postpartum period differs across cultures could improve detection and provide new insights into the pathogenesis. Moreover, it is a necessary step to evaluate the universal use of screening instruments in research and clinical practice.
Impact: Investigators and clinicians should be aware of the potential differences in expression of phenotype of postpartum depression that women of different educational backgrounds may manifest. The increasing cultural heterogeneity of societies together with the tendency towards globalization requires a culturally sensitive approach to patients, research and policies, that considers, beyond rhetoric, the context of a person's experiences and the context in which the research is conducted.

Work on premenstrual dysphoric disorder (PMDD) and hormonal manipulation protocols: During the Outgoing phase of my Fellowship at the University of North Carolina in the USA I attended the first NIH-funded postdoctoral training program in reproductive mood disorders and developed mastery in reproductive hormonal physiology and signalling and methods for manipulating the reproductive system.
Data from a case-control study on 78 women with prospectively confirmed PMDD and 63 healthy controls did not support the view that women with PMDD are “difficult women”. Women with PMDD instead display subclinical personality traits that are present in both the luteal and follicular phases of the menstrual cycle. Such traits may contribute to the hormone sensitivity and therefore be the target of preventive strategies.
Impact: Although further research is needed to elucidate the relationship between personality characteristics and the biological substrates of the trait vulnerability to PMDD, our results suggest new venues to the study of preventative and therapeutic approaches. Currently, the first-line treatment options for premenstrual syndromes are cognitive-behavioral therapy and drugs targeting the hypothalamus-pituitary-ovary axis or brain serotonergic synapses. The overall quality of the evidence for all treatment options is however low and effect sizes are small. The sub-clinical abnormalities that we found in women with PMDD may be the target for the development of new specific interventions aimed to increase resilience towards the changes in reproductive hormones during the menstrual cycle. Despite its recent inclusion in the main text of the DSM-5, PMDD is still a highly stigmatized and controversial disorder. The view that the diagnosis is given to “‘difficult women’”, who are “victims of a process of pathologisation” is still widespread. Our results failed to support this position and will contribute to the de-stimatization of the disorder.

Related documents