checkedThe Carlat Psychiatry Podcast2026-11-09last verified 2026-11-09

did a study find nearly half of women with postpartum psychosis said symptoms waxed and waned?

verdict: confirmed
the primary source says it, at that magnitude
“In a new study of postpartum psychosis, nearly half the women said that their symptoms waxed and waned.”

a host, a psychiatrist said on The Carlat Psychiatry Podcast, 2026-11-09, at 10:01. hear it in context → the speaker cited: a new study of postpartum psychosis.

the episode: Kraepelin on Clancy Part 2 (at 10:01). we link the show's own page; we do not re-host audio.

first, the part that holds

the study is real and the number holds. a large postpartum psychosis cohort from the massachusetts general hospital postpartum psychosis project, published in molecular psychiatry, found that 45.7% reported their symptoms waxed and waned and 51.9% reported more steady symptoms. "nearly half" is a fair way to say 45.7%.

what we found when we went and looked

the paper behind this is most likely the 2024 molecular psychiatry report from the massachusetts general hospital postpartum psychosis project. it covers 248 people with at least one postpartum psychosis episode, assessed by a structured retrospective interview (the MINI-PDS plus the project's own questionnaire). the study team's own summary of the paper says 45.7% (n=118) reported symptoms that waxed and waned and 51.9% (n=134) reported more consistent symptoms. the full text is paywalled, so we read that figure in the investigators' summary and could not read the passage in the journal itself. the arithmetic suggests the percentage was calculated over the 258 episodes reported, not the 248 people (118/258 = 45.7%, 118/248 = 47.6%), but either way it is "nearly half".

assertion by assertion

assertionverdictwhat the record says
a study of postpartum psychosis asked women about their symptoms, and they reported them themselvesconfirmedthe molecular psychiatry report describes 248 participants with a postpartum psychosis episode within 6 months of delivery. it was assessed retrospectively by structured interview (MINI-PDS and the MGHP3 questionnaire), which captured onset, duration, symptoms and treatment. so the pattern of symptoms is what participants recalled and told the interviewers. [1][2]
nearly half reported that their symptoms waxed and wanedconfirmed45.7% (n=118) reported waxing and waning symptoms and 51.9% (n=134) reported more consistent ones. we read this figure in the study team's own published summary of the paper. the journal full text is paywalled, and the abstract we could read does not include the figure. that summary page is not on our list of citable sources, so it is described here rather than linked. [1]
the study is newconfirmedit was published online 6 december 2024, about 23 months before the episode aired. that is recent for a research paper, though calling it 'new' is loose. if the speaker meant a different, later study, we did not find one on postpartum psychosis that reports this figure. [1]
the 'half' refers to women (people), not episodesconfirmed118 out of 248 people would be 47.6%, not 45.7%. but 118 out of the 258 episodes reported comes to exactly 45.7%, and 134/258 comes to exactly 51.9%. so the denominator looks like episodes (a few women had more than one). the difference does not change 'nearly half'. [1]
popular framing: about half of all postpartum psychosis cases follow a waxing-and-waning courseoverstatedthe cohort was recruited through a website, social media, advocacy and support networks, physician referral and one hospital department. episodes were recalled up to 10 years later, with some protocol deviations to include older episodes. this is a self-selected, retrospective sample, not a population estimate of how the illness unfolds. [1][2]

both directions

for the speaker: the number matches the study team's report: 45.7% waxing and waning versus 51.9% consistent. the study is the largest cohort with validated postpartum psychosis described to date (248 participants), so it is a reasonable study to quote. "women said" correctly describes interview-based self-report. the finding also fits long-standing clinical descriptions of postpartum psychosis as a fast-changing presentation, which is one reason it can be missed when a clinician sees someone during a calmer stretch.

against: the study is about two years old, not new. the 45.7% seems to be calculated over episodes rather than individual women. the data are retrospective recall, sometimes of episodes a decade or more earlier, from a sample recruited largely through online and advocacy channels. people whose illness was more dramatic or memorable, or who are active in support communities, may be overrepresented. "waxed and waned" was a participant-reported description, not a pattern measured by repeated clinical ratings.

what this cannot say

we could not open the paywalled full text of the journal article. the 45.7% / 51.9% split comes from the investigators' own summary of the paper, and the abstract we read does not contain it. the paper is not in pubmed central, and the pubmed page would not load for us. the episode-versus-person denominator is our inference from the arithmetic, not a statement we read. we assumed the speaker meant this study. the episode does not name the paper, and a different study could be meant. none of this tells anyone how a particular person's illness will unfold.

sources, each one fetched and read

  1. massachusetts general hospital postpartum psychosis project investigators, 'the phenomenology of postpartum psychosis: preliminary findings from the massachusetts general hospital postpartum psychosis project,' molecular psychiatry, published online 6 december 2024, doi:10.1038/s41380-024-02856-3 What we read there: 248 participants with postpartum psychosis within 6 months of delivery, assessed retrospectively with the MINI-PDS and the MGHP3 questionnaire (onset, duration, symptoms, treatment); delusions 87.6%, persecutory delusions 75.2%, visual hallucinations 52.3%, auditory hallucinations 48.1%; protocol deviations allowed episodes more than 10 years before interview. the waxing-and-waning figure was not visible in the portion we could access. Fetched 2026-11-09 (HTTP 200). https://www.nature.com/articles/s41380-024-02856-3?error=cookies_not_supported&code=3813b66e-abce-4e5f-8b4a-046e4148dba2
  2. massachusetts general hospital postpartum psychosis project investigators, 'establishment of the MGH postpartum psychosis project: MGHP3,' plos one, 2023;18(2):e0281133, doi:10.1371/journal.pone.0281133 What we read there: recruitment via 'web presence (i.e., womensmentalhealth.org), social media, physician referral, advocacy-based digital networks, and within Massachusetts General Hospital's (MGH) Psychiatry Department'; 311 enrolled from 44 states, DC and 7 countries as of july 2022; roughly 90-minute structured telephone interview; symptom timing relies on patient self-report, clarified with medical records when available. Fetched 2026-11-09 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC9910633/

speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam