What causes postpartum psychosis?

Roughly 1-2 in 1000 births trigger a psychotic episode, known as ‘postpartum psychosis’. Symptoms appear acutely, usually within the first two weeks of delivery, and present as a mixture of psychotic (delusions, hallucinations) and mood (mania, affective lability) symptoms. Where delusions are present, they often centre on the newborn baby.

Postpartum psychosis is called a ‘psychiatric emergency’. This reflects the fast changing symptoms, as well as risk to the infant and mother. Tragically, suicide is the leading cause of maternal mortality in the first year after birth.

Periodically, patient accounts of postpartum psychosis and recovery are published in outlets like the Guardian: see examples from Catherine Cho, Laura Dockrill, and Jessica Pidsley. These personal stories are both deeply moving and insightful into the experience of postpartum psychosis.

Most women experiencing postpartum psychosis do not have a history of mental illness but over half go on to have further episodes that are not associated with childbirth. If we knew what made certain individuals vulnerable to the trigger of childbirth, we might be able to prevent a lifetime of relapsing mental disorder.

What we already know

Some facts about postpartum psychosis are well established. Firstly, the greatest risk is in a woman’s first pregnancy. Second births have a much smaller risk and third births have no elevated risk above baseline. Partly, this could be that women who have experienced postpartum illness are reluctant to become pregnant again but it doesn’t fully explain the effect. Other obstetric disorders like pre-eclampsia show a similar pattern.

Secondly, risk of relapse following childbirth is highest for bipolar disorder. Women with postpartum psychosis who go on to experience further episodes most commonly receive a diagnosis of bipolar disorder. This has led to the conceptualisation of postpartum psychosis as a bipolar-spectrum disorder.

Thirdly, if a woman has had a previous episode of postpartum psychosis, the chance of it happening in a subsequent pregnancy is about 50%. Given the baseline rate is ~0.15%, this shows that there is a small group of women with a massively increased vulnerability to the trigger of childbirth.

Possible causes

Adverse life events

Some of the biggest and most replicated risk factors for depression following childbirth are adverse life events. This encompasses a broad range of negative exposures including trauma, intimate partner violence, and chronic stressors. Adverse life events are also associated with the development of psychosis outwith the postpartum period. So it seems logical to investigate whether adverse events may also contribute to postpartum psychosis.

Contrary to expectation, various large studies have failed to find an association with adverse events. These have included childhood trauma, complications in pregnancy or birth, and death of a close relative. In line with this, women who develop postpartum psychosis do not seem different in terms of personality traits, cognitive styles, or temperament.

A lack of association between life events and postpartum psychosis suggests that other, more biological causes should be considered.

Hormonal shifts

During pregnancy, oestrogen and progesterone rise to high levels, then drop precipitously following birth. As I have previously written, there is convincing evidence to support the oestrogen hypothesis of psychosis.

Could the fall in oestrogen trigger postpartum psychosis? Relatively few studies have directly investigated. One small (n=29) open label trial by Channi Kumar and colleagues failed to find an effect of transdermal oestrogen on relapse rates.

One of the few novel psychiatric drugs to gain FDA approval is for postpartum depression. Brexanolone, an analogue of the progesterone metabolite allopregnanolone, acts as a GABA-A receptor modulator, and produces a quick and sustained antidepressant effect in early postpartum.

To my knowledge, no-one is studying whether allopregnanolone analogues might also be effective for postpartum psychosis but, given GABA-A receptors are implicated in psychosis more generally, it is worth considering.

Sleep disturbance

Most new parents are well acquainted with sleep deprivation. It is well known that sleep loss can trigger episodes of mania in bipolar disorder (particularly in women). More generally, sleep disturbances are common to both bipolar and schizophrenia.

Women with bipolar disorder who experience episodes triggered by sleep loss are twice as likely to have experienced an episode of postpartum psychosis compared with women without this trigger.

One doubt I have about the role of sleep disturbance is the timing. Postpartum psychosis emerges in the first days to weeks following the arrival of a newborn, whereas the sleep disturbance in new parents will continue for at least a few more months. If sleep disturbance is causative, would we also expect an increased risk of psychosis in new fathers? From one large study, they have no increased risk of mental disorder.

Immune dysregulation

Pregnancy and postpartum are associated with interesting changes in the mother’s immune system, which has to tolerate the fetus while also providing protection from pathogens. The immune system is widely implicated in psychosis, with a particular focus in recent years on whether there is a subset of patients with ‘autoimmune psychosis’, caused by a person’s own immune system targeting antigens in the brain.

Autoimmune thyroid disease is a common problem affecting ~5% of new mothers, thought to result from a ‘rebound’ of the immune system following childbirth.

In a review article, Katie Hazelgrove argues that dysregulation of the immune system could cause postpartum psychosis. She presents evidence for specific inflammatory markers being raised. While the evidence to date is preliminary, it seems like a promising area for future research.

Genetics

Postpartum psychosis, like most psychiatric disorders, clearly has a genetic component: a positive family history increases the risk 3-fold. The nature of this genetic liability is not well characterised but research is underway. Arianna Di Florio, a professor at Cardiff University, led the first genetic study of postpartum psychosis last year. She found that women with first onset postpartum psychosis had increased genetic risk for bipolar disorder and schizophrenia but not for depression. Specific genetic markers for postpartum psychosis are yet to be discovered and how this genetic vulnerability is specifically expressed following childbirth is not known.

Why don’t we already know what causes postpartum psychosis?

The risk of psychosis following childbirth is estimated to be increased 23-fold compared with any other times of a woman’s life. Given the high rate of recurrence (50%), the short-time window of highest risk (two weeks), and clear trigger - why haven’t we already discovered what causes postpartum psychosis?

I’m sure part of the explanation is the relative low-priority that has historically held back female-specific neuroscience. However, another important factor is that while the relative risk of postpartum psychosis is high compared with baseline, the absolute risk is actually very small. In other words, postpartum psychosis is a rare disorder.

Consider that there were roughly 111,000 births in London last year. We should expect ~170 cases of postpartum psychosis. Given there are 28 different maternity hospital sites in London, each hospital might see only a handful of cases per year. Like other rare diseases, this forces us to use large record-linkage study designs or alternatively to form networks in order to recruit enough patients to take part in studies. Improving our understanding of this disorder is crucial, though, if we are to prevent new mothers developing a severe mental illness at such a crucial time of life.