The Version You Have Seen Is the Rarest One
Every few years a case reaches the news, and it is always the same shape: a mother, a dead child, a courtroom. For most people that story is postpartum psychosis. It is the only version they have encountered.
Those cases are real. They are also the least representative thing that could have been chosen to represent the illness.
What It Usually Looks Like
In a Dutch cohort of 130 women admitted for postpartum psychosis, researchers found three distinct symptom profiles:
- Depressive — 41%. Depressive and anxiety symptoms dominate.
- Manic — 34%. Elevated mood, agitation.
- Atypical — 25%. Disturbed consciousness, disorientation.
The most common presentation is the one that looks least like the news. And it is the one that gets missed: in that same cohort, women with the depressive profile started treatment about two weeks later than the others.
The authors also found that disorientation and depersonalization were less common than the earlier literature had suggested. The image of a woman visibly, floridly out of her mind is not the typical case.
The Number Behind the Headlines Is Disputed
You will see it written that around 4% of postpartum psychosis cases end in infanticide. That figure is contested in the peer-reviewed literature.
A 2017 review traced it back to a misreading of a single Scottish follow-up study of 82 patients, most of whom had depression rather than psychosis. Compiling roughly 4,029 documented cases of childbearing psychosis from the historical literature, the same author found filicide in about 1.2% overall — 4.5% among the depressive presentations, and under 1% among episodes without overt depression. His conclusion, in his own words: the risk “has also been exaggerated.”
That accounting comes with its own caveat. It is a compilation of case reports spanning more than two centuries, not a modern population study, and it should be read as the best available historical accounting rather than a clean rate.
Thoughts Are Not Acts, and the Difference Is Most of the Story
In that cohort of 130 hospitalized women, 19% had suicidal thoughts and 8% had thoughts of harming the infant. Which means 92% did not report that thought at all — and having the thought is categorically different from acting on it.
This distinction is the one headlines collapse, and collapsing it does real harm. It is the reason women do not speak.
How Common It Actually Is
Roughly 1 to 2 in every 1,000 births — but that figure deserves more honesty than it usually gets.
A systematic review found reported incidence ranging from 0.89 to 2.6 per 1,000, and its authors explicitly declined to pool those numbers because the underlying studies defined and identified cases too differently. So “1 to 2 per 1,000” is a reasonable midpoint of a spread, not a precise pooled estimate.
Picture 1,000 women giving birth. One or two will develop this. The other 998 will not.
It Often Arrives With No Warning at All
When researchers interviewed nine women who had recovered from postpartum psychosis about a British soap opera storyline, their most common criticism was revealing. The televised character had a known history of bipolar disorder before her episode — and the women objected that this understated how often the illness arrives with no psychiatric history whatsoever. One put it directly: it failed to convey “the alarming fact that for quite a lot of women, postpartum psychosis is totally out of the blue.”
Most women who develop postpartum psychosis have no prior mental health diagnosis. The absence of one is not protection.
Medication Changes Belong in This Conversation
This is the part that rarely makes the news, and it is among the better-evidenced pieces of the picture.
For women with bipolar disorder, a meta-analysis covering 5,700 deliveries found postpartum relapse in about 66% of those who went through pregnancy without medication, against about 23% of those who continued prophylactic medication.
For women whose only psychiatric history is a previous postpartum psychosis, a Dutch cohort found something more striking still: of 20 who began prophylaxis immediately after delivery, none relapsed. Of 9 who declined it, 4 did.
And discontinuation is common in practice. A US cohort of 214 pregnancies with a documented bipolar diagnosis found only 14% on a mood stabilizer at the first prenatal visit and 22.4% at delivery — with, as the authors noted, the reasons for stopping usually undocumented.
Four honest limits on all of that:
- These are associations from observational studies, not randomized trials. Women who stopped medication may have differed from those who continued in ways the studies could not separate.
- Every one of these findings comes from women already diagnosed with bipolar disorder or with a prior postpartum episode. They do not describe the general population.
- Medication is not a guarantee. Roughly 23 in 100 relapsed despite continuing it.
- On antidepressants specifically — starting, stopping or switching them around delivery — the evidence amounts to a three-patient case series. That is a signal clinicians have noticed, not a finding. It should not be extrapolated from the lithium and mood-stabilizer research above.
None of this is a reason to change anything on your own. It is the agenda for an appointment with a perinatal psychiatrist, ideally before delivery.
Losing a Night of Sleep Around the Birth Matters
In a UK study following 76 pregnant women with bipolar or schizoaffective disorder, losing at least one complete night of sleep across labor and delivery was associated with roughly five times the odds of developing postpartum psychosis — and that held after accounting for whether the women were on protective medication.
Sleep quality earlier in pregnancy showed no such association, and neither did postpartum depression. It was the acute loss around the birth itself.
Small study, specific population. But it is a concrete, actionable thing to plan for.
It Responds to Treatment, and That Is the Part Nobody Reports
In a specialist unit treating 64 women with first-onset postpartum psychosis under a structured four-step protocol, 98.4% reached complete remission, none required ECT, and 79.7% were still in remission at nine months.
That is a single specialist center using a specific protocol on first-onset cases, so it is a best case rather than an average. But it is the right counterweight to a news archive made entirely of the worst outcomes.
If the Dark Thoughts Are Yours, Say So
The depressive presentation is the most common and the most likely to be missed. It is also the one most likely to be endured quietly, because it does not feel dramatic enough to report.
Some things worth knowing before you decide to stay silent:
Telling a clinician about frightening thoughts is not the same event as a child-protection referral. Clinicians who work with new parents expect to hear about intrusive thoughts, and the distress those thoughts cause you is precisely what tells them what they are looking at.
If you are able to sit and carefully examine whether something might be wrong with your mind, that capacity is itself meaningful — psychosis characteristically impairs it. This is clinical reasoning rather than a research finding, but it is worth holding onto.
And the illness is treatable, fast, with a very high remission rate when it is caught.
Call or text 988, or call, text or chat the National Maternal Mental Health Hotline at 1-833-852-6262, free and confidential, 24 hours a day. Call 911 or go to an emergency room if you are seeing or hearing things others do not, believing things others find strange, feeling confused or disconnected from reality, or have gone days without sleep.
More detail on the condition itself is on our page about postpartum psychosis, and every number and what happens when you call is on crisis help and hotlines.
Sources
- Brockington I. “Suicide and filicide in postpartum psychosis.” Arch Womens Ment Health. 2017;20(1):63-69. doi:10.1007/s00737-016-0675-8. PMID 27778148.
- Kamperman AM, Veldman-Hoek MJ, Wesseloo R, et al. “Phenotypical characteristics of postpartum psychosis: A clinical cohort study.” Bipolar Disord. 2017;19(6):450-457. doi:10.1111/bdi.12523. PMID 28699248.
- VanderKruik R, Barreix M, Chou D, et al. “The global prevalence of postpartum psychosis: a systematic review.” BMC Psychiatry. 2017;17(1):272. doi:10.1186/s12888-017-1427-7. PMID 28754094.
- Roberts L, Berrisford G, Heron J, et al. “Qualitative exploration of the effect of a television soap opera storyline on women with experience of postpartum psychosis.” BJPsych Open. 2018;4(2):75-82. doi:10.1192/bjo.2018.9. PMID 29971150.
- Wesseloo R, Kamperman AM, Munk-Olsen T, Pop VJ, Kushner SA, Bergink V. “Risk of Postpartum Relapse in Bipolar Disorder and Postpartum Psychosis: A Systematic Review and Meta-Analysis.” Am J Psychiatry. 2016;173(2):117-127. doi:10.1176/appi.ajp.2015.15010124. PMID 26514657.
- Bergink V, Bouvy PF, Vervoort JS, Koorengevel KM, Steegers EA, Kushner SA. “Prevention of Postpartum Psychosis and Mania in Women at High Risk.” Am J Psychiatry. 2012;169(6):609-615. doi:10.1176/appi.ajp.2012.11071047. PMID 22407083.
- Kernizan NJ, et al. “Prescribing Patterns for the Treatment of Bipolar Disorder in Pregnancy: A Retrospective Cohort Study.” J Clin Med. 2025;14(5):1638. doi:10.3390/jcm14051638. PMID 40095612.
- Perry A, Gordon-Smith K, Lewis KJS, Di Florio A, Craddock N, Jones L, Jones I. “Perinatal sleep disruption and postpartum psychosis in bipolar disorder: Findings from the UK BDRN Pregnancy Study.” J Affect Disord. 2024. doi:10.1016/j.jad.2023.11.005. PMID 37940050.
- Bergink V, Burgerhout KM, Koorengevel KM, et al. “Treatment of psychosis and mania in the postpartum period.” Am J Psychiatry. 2015;172(2):115-123. doi:10.1176/appi.ajp.2014.13121652. PMID 25640930.
- Sharma V. “A cautionary note on the use of antidepressants in postpartum depression.” Bipolar Disord. 2006;8(4):411-414. doi:10.1111/j.1399-5618.2006.00336.x. PMID 16879142.
- Bergink V, Rasgon N, Wisner KL. “Postpartum Psychosis: Madness, Mania, and Melancholia in Motherhood.” Am J Psychiatry. 2016;173(12):1179-1188. doi:10.1176/appi.ajp.2016.16040454. PMID 27609245.