Long-Term Outcomes of Postpartum Psychosis: What is the Risk of Recurrence?

Postpartum psychosis is the most severe form of postpartum psychiatric illness. It is a rare event that occurs in fewer than one out of every 1000 women after childbirth. Many women who have suffered from postpartum psychosis (PP) wonder about their long term prognosis. Will they have another psychotic episode if they have another child? Will they have another episode if they taper off their medication? Many of these questions are still unanswered, although we do know that, in women who have experienced PP, recurrent episodes can be effectively prevented. A recent systematic review and meta-analysis published in the Journal of Clinical Psychiatry reviews the probability of recurrent psychiatric illness following an episode of PP.  


Within the medical literature, there is considerable debate about whether postpartum psychosis lies on the bipolar spectrum or should be classified as its own disorder. The DSM-V does not currently classify PP as a distinct disorder, and since its presentation often includes mania or mixed affective symptoms, women with PP are often given a diagnosis of bipolar disorder.   


How the disorder is labelled at the time of onset determines future treatment and prognosis. On one hand, bipolar disorder is a lifelong psychiatric illness that requires long term management, typically with mood stabilizing medications. On the other hand, postpartum psychosis may represent a distinct, more episodic disorder which only occurs during the postpartum period.  


Whether an episode of postpartum psychosis is viewed as the onset of an affective disorder such as bipolar disorder versus an isolated episode of psychosis has important implications for acute treatment and planning for later pregnancies. 


Gilden and colleagues conducted a literature search and meta-analysis of longitudinal studies of postpartum psychosis and aimed to calculate the rate of recurrence, as well as elucidate whether recurrent episodes occurred during a subsequent postpartum period or were unrelated to pregnancy. They included six studies of 645 patients in their meta-analysis, and follow up periods ranged from 11-26 years.


Researchers divided recurrence into 3 categories:

  • (a) at least 1 subsequent postpartum episode but no episodes outside the postpartum period

  • (b) at least 1 subsequent episode outside the postpartum period, and

  • (c) no subsequent episode of mania, psychosis, or severe depression (sustained remission)

In estimating risk of recurrence, they observed the following:

  • 43.5% (279/645) of women had “isolated postpartum psychosis,” meaning either they experienced PP once and had no recurrence (36.1% of total sample), or experienced psychosis more than once but exclusively during postpartum periods (7.1% of total sample). 

  • For 56.7% of women (366/645) their first postpartum psychosis episode was the onset of a lifelong severe psychiatric disorder, with episodes of mania, psychosis, or major depression with psychosis occurring outside of the postpartum period. 

  • Of the overall sample, 64% (412/645) women with first-onset PP experienced a recurrence at some point during the follow up period (mean of 16 years).  

  • 36% had no recurrence, (95% CI, 32.3% to 39.7%) and remained in remission for duration of follow up which was a mean of 16 years


Limitations of the Study

The review and meta-analysis have several significant limitations. Other, less severe clinical outcomes, such as minor episodes, roughening, or treatment response couldn’t be included.   Although the mean follow up period was 16 years, the study cannot draw conclusions regarding the likelihood of recurrence during menopause, a high-risk period for mood episodes.  All included studies were conducted in Western Europe during the 1970s and 1980s, limiting generalizability. The authors point out that more recent studies have shown a lower rate of recurrence, possibly due to pharmacotherapy and prophylactic management. Finally, the authors note, “it is not possible to study a true recurrence risk in a naturalistic setting, since many of the women included in these follow-up studies might have received targeted treatment to prevent recurrence.”


Clinical Implications and Recommendations

The DSM-V is widely used to classify psychiatric disorders, and it does not identify postpartum psychosis as a distinct disorder. Since many women with PP present with features of mania or a mixed episode, many are diagnosed with bipolar disorder at the time they present with PP. The findings of this study, however, raise some doubts about this protocol. Given that 43.5% of women had no manic or psychotic recurrence outside of the postpartum period, and 36% had no recurrence at all, a distinct diagnostic category of postpartum psychosis may be more appropriate than a label of bipolar disorder. 


However, recognizing the significant probability of recurrence is essential to successful treatment and prevention. This review indicates that for the majority (56.5%) of women who experienced first-onset postpartum psychosis, they later experienced a severe non-postpartum recurrence.This suggests that an episode of PP should not be “written off” as a one-time occurrence, and treatment and medication management should be carefully monitored.


Unfortunately this analysis does not have specific predictive value; there is still no way to identify whose postpartum psychosis will recur, a priori. However, it does help providers manage treatment and expectations about long-term recovery. Understanding the likelihood of PP recurrence is important to the risk-benefit analysis of maintenance pharmacotherapy as well as preparing for subsequent pregnancies. As the article points out, an overestimation of recurrence risk could lead to “unfounded concerns for health care providers, patients, and their families, resulting in excessive medication use, unnecessary prevention strategies, or altered family planning.” On the other hand, underestimation of recurrence risk could lead to insufficient medical treatment and increased risk of hospitalization or suicide.

Miranda Arakelian, BA  

Gilden J, Kamperman AM, Munk-Olsen T, Hoogendijk WJG, Kushner SA, Bergink V.  Long-Term Outcomes of Postpartum Psychosis: A Systematic Review and Meta-Analysis.  J Clin Psychiatry. 2020 Mar 10;81(2). 


This post was originally shared on the blog of MGH's Center for Women's Mental Health. See the original post here.

617-643-7205

185 Cambridge St
Boston, MA, USA 02114

  • Facebook Social Icon

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