Postpartum psychosis: an important clue to the etiology of mental illness

Howard and Khalifeh ¹ masterfully review the epidemiology of perinatal mental health conditions and the evidence base for their management. Here I address a further issue and exciting opportunity: the role that the study of severe perinatal mental illness can play in advancing our understanding of the etiology of mental health con‐ditions.

The close relationship of severe episodes of mental illness to childbirth, episodes labelled postpartum psychosis, has been observed for hundreds, if not thousands, of years, and more recently this link has received support from clinical and epidemiological studies ² . Despite this long history, we have failed to take advantage of this important clue to the pathophysiology of mental illness.One reason may be the confusion that remains around classification, with both DSM and ICD not dealing adequately with severe postpartum mental illness. As with many mental health conditions, there may be fuzziness around the boundaries, but there is clarity at the core of the concept of postpartum psychosis, and this concept remains useful and in widespread use by clinicians and women themselves. For ex‐ample, the main third sector organization supporting women and their families in the UK is called Action on Postpartum Psychosis (app‐network.org). Despite this nosological confusion, however, there is no doubt that “we know it when we see it”.

What, then, is postpartum psychosis and why is this condition potentially so important in our understanding of the etiology of mental disorders? Postpartum psychosis is a severe episode of mental illness that impacts around 1 in 1,000 women following childbirth ² . Onset is in the immediate postpartum, most often the first or second postpartum week. The symptoms are most commonly of an affective psychosis, with perplexity common, and often a rapidly and constantly changing (“kaleidoscopic”) presentation.

Postpartum psychosis is a true psychiatric emergency, with admission to hospital usually required, but, despite the initial severity and rapidity of presentation, prognosis is good, with most episodes responding well to treatment, predominantly medication in the acute stage. Following the initial psychotic phase, however, women may ex‐perience longer episodes of depression, and many of them report that full recovery takes many months. Psychological interventions, including peer support, in the longer term can be very helpful in the recovery process.

Although around 50% of women with postpartum psychosis have not experienced a previous episode of mental illness, there is a clear link to bipolar disorder, especially bipolar I disorder. Women with a previous diagnosis of bipolar disorder are at high risk (around one in five deliveries) ³ . In addition, women who experience postpartum psychosis as a first episode, even if not clearly bipolar at initial presentation, are at high risk of subsequent bipolar illness ⁴ .

The evidence is clear, therefore, that child‐birth is a potent trigger of episodes of severe mental illness, and that this risk is not spread evenly across all mental illness, but shows a specific link to bipolar disorder. What are the mechanisms behind this as‐sociation? Although psychological and so‐cial factors clearly play an important role in perinatal mental health conditions in general, and postnatal depression in particular, when it comes to postpartum psychosis biological factors are likely to be primary, with hormonal, immunological, circadian rhythm, and genetic factors all suggested to play a role ² .

There is a dramatic rise in levels of reproductive hormones (oestrogen and progesterone) in pregnancy and a precipitous fall in the immediate postpartum, corresponding to the exact time that sees the peak onset for postpartum psychosis. Periods of hormonal fluctuation, in the menstrual cycle for example, are known to be associated with mood symptoms, and this had led to hormonal factors being considered in the etiology of postpartum psychosis. The evidence base for this assertion remains, however, mostly circumstantial. There have been no consistently demonstrated abnormalities in hormonal levels in women experiencing perinatal mental illness, but it remains possible that women with postpartum episodes are differentially sensitive to the normal hormonal fluctuations associated with pregnancy and childbirth ⁵ .

In recent years, the role that immunological mechanisms and inflammation play in psychiatric disorders has received considerable attention. This, combined with the fact that pregnancy is a major immunological challenge, has led some to hypothesize that immune and neuroinflammatory mechanisms play a role in the etiology of postpartum psychosis. Further support comes from the evidence of increased risk in first pregnancies, a finding shared with other pregnancy‐related disorders, such as pre‐eclampsia, which are thought to be driven by immunological mechanisms. Studies have found some evidence pointing to the role of immune biomarkers. For example, women with postpartum psychosis in one study did not display the expected T cell elevation following childbirth, but rather presented a monocytosis ⁶ . In addition, small numbers of women with postpartum psychosis (around 2%) were reported to have anti‐neuronal autoantibodies in one study ⁷ .

A further clue to etiology comes from the known link between circadian rhythm disturbance and the triggering of mood disorder, particularly mania, combined with the almost universal disturbance of sleep patterns that having a baby involves. Although it has not been studied extensively, there is some evidence in support of this hypothesis. For example, one study found that women with bipolar disorder who reported that sleep loss triggered episodes of mania were more than twice as likely to have experienced postpartum psychosis ⁸ .

A further hypothesis receiving attention is the potential involvement of genetic factors. Family and linkage studies suggest a genetic etiology, and a number of linkage and candidate gene studies have been reported, but are yet to yield replicated results ² . Sample sizes have been limited up to now, but large‐scale collaborative efforts are underway to significantly increase the numbers available.

In summary, childbirth is a potent trigger for severe mood disorder, and this link gives us unrivalled opportunities for research into etiology. In no other scenario can we identify individuals, currently well, who are at such a high risk of experiencing a severe episode of mental illness in a defined two‐week period. In addition to understanding more about etiology, we also have a significant opportunity for prevention, through the development of predictive models identifying which women are at very high risk ⁹ .

We need, therefore, to take advantage of the vital clue that postpartum psychosis represents. First, we need this condition to be better dealt with by the ICD and DSM classification systems, which currently are of little help in ensuring that these episodes are recorded. Second, we need to build large cohorts of women who have experienced this condition for international collaborations to look, for example, at its genetic underpinnings. Finally, we need prospective studies of selected populations, for example women with previous episodes of bipolar disorder, applying a range of paradigms, from imaging to other biomarkers, allowing us to better identify subjects at high risk.