did about half of women without lithium relapse within a year after a postpartum psychosis?
“the women without lithium, we saw unfortunately that half of them relapsed within a year after delivery.”
a guest, a psychiatrist and director of a women's mental health center said on Psychiatry & Psychotherapy Podcast, 2026-09-18, at 25:34. hear it in context → the speaker cited: the guest's own cohort study.
the episode: Postpartum Psychosis: Diagnosis, Treatment, Lithium, and Relapse Prevention with Dr. Veerle Bergink (at 25:34). we link the show's own page; we do not re-host audio.
first, the part that holds
the speaker's group has published this kind of finding. in their 2012 cohort, 44.4% of women with a history of postpartum-only psychosis who declined postpartum prophylaxis relapsed after delivery, while none of the 20 who took it did. that is close to "about half."
what we found when we went and looked
the closest published match is a 2012 American Journal of Psychiatry cohort of 70 high-risk pregnant women from the speaker's clinic. women were advised to start lithium immediately after delivery. "none of the 20 women with a history of postpartum psychosis only who used postpartum prophylaxis relapsed, compared to 44.4% of patients with postpartum psychosis only who declined prophylaxis." that 44.4% comes from a very small group: 29 women had postpartum-only psychosis and 20 took prophylaxis, so at most 9 declined. 44.4% of 9 is 4 women (our arithmetic; the abstract does not give the count). the abstract we read does not say the relapse window was one year, and we could not tell which of the group's cohorts the speaker meant. so the "within a year" clause, and the exact figure attached to it, could not be checked.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| in the speaker's cohort, about half of women not given lithium after delivery relapsed | confirmed | the 2012 cohort abstract reports 44.4% postpartum relapse in women with postpartum-only psychosis who declined prophylaxis (the protocol advised lithium), against 0 of 20 who used it. 44% is reasonably called about half. the group is tiny, at most 9 women, so 4 relapses (our arithmetic). the uncertainty around that figure is very wide [1] |
| the relapses happened within a year after delivery | unverifiable | the 2012 abstract says 'postpartum' and does not state a one-year window. the 2015 treatment cohort followed women to 9 months, not 12. we could not read a one-year relapse figure for women without lithium in any of the group's papers we fetched [1][2] |
| in the group's treatment cohort, women kept on lithium relapsed less than women kept on an antipsychotic alone | confirmed | the 2015 abstract on 64 women with first-onset postpartum psychosis reports sustained remission in 79.7% at 9 months postpartum. it says 'patients treated with lithium had a significantly lower rate of relapse compared with those treated with antipsychotic monotherapy.' the abstract does not give the per-group relapse percentages [2] |
| across the wider literature, the postpartum relapse risk without medication is in the range the speaker describes | confirmed | a 2016 meta-analysis (37 studies, 5,700 deliveries) found an overall postpartum relapse risk of 35%. women with a history of postpartum psychosis had a severe-episode rate of 29%. in bipolar disorder, relapse was 66% when women were medication free in pregnancy versus 23% with prophylaxis. these figures mix populations and outcome definitions and are not a direct match for 'no lithium after postpartum psychosis' [3] |
both directions
for the speaker: the speaker's own published cohort has 44.4% postpartum relapse among women with postpartum-only psychosis who declined prophylaxis, versus 0 of 20 who took it. the protocol advised lithium. the same group's 2015 cohort reports significantly less relapse on lithium maintenance than on antipsychotic monotherapy through 9 months. a 2016 meta-analysis found high postpartum relapse in unprotected high-risk women: 66% in bipolar disorder without medication in pregnancy. the 2012 contrast of 44% versus 0% is in the direction the speaker describes and is large.
against: "half" rests on a very small group. at most 9 women declined prophylaxis, so the 44.4% is about 4 women, and a single case either way moves it by about 11 points. the comparison is not randomized: women who declined prophylaxis may differ from those who accepted it. the abstract does not define the relapse window as one year. the pooled figure for women with a history of postpartum psychosis in the meta-analysis is 29%, lower than half, though that pool mixes treated and untreated women. the group's 2021 long-term cohort of 106 women found that over two thirds had no major episodes outside the postpartum period during a four-year follow-up. it reports a 32% recurrence rate outside the postpartum period, not a first-year no-lithium figure.
what this cannot say
we read abstracts and summaries, not the full-text tables of the 2012 and 2015 papers. so we could not confirm the exact denominators, the relapse window, or the per-group relapse rates in the 2015 cohort. the speaker may have meant a newer or unpublished analysis of the group's long-running cohort, which has enrolled patients since 2005. if so, we have no file to check against. observational cohorts like these cannot show that lithium caused the lower relapse rate. women chose whether to take prophylaxis. this is a description of what a study found, not advice about any person's medication.
sources, each one fetched and read
- Prevention of postpartum psychosis and mania in women at high risk. American Journal of Psychiatry, 2012. PMID 22407083 (PubMed record) What we read there: 70 pregnant high-risk women, 2003-2010; medication-free women advised to start lithium prophylaxis immediately postpartum. 'none of the 20 women with a history of postpartum psychosis only who used postpartum prophylaxis relapsed, compared to 44.4% of patients with postpartum psychosis only who declined prophylaxis.' 29 women had psychosis limited to the postpartum period. Fetched 2026-10-22 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/22407083/
- Treatment of psychosis and mania in the postpartum period. American Journal of Psychiatry, 2015;172(2):115-23. PMID 25640930 (PubMed record) What we read there: 'nearly all patients (98.4%) achieved complete remission within the first three steps... At 9 months postpartum, sustained remission was observed in 79.7%. Patients treated with lithium had a significantly lower rate of relapse compared with those treated with antipsychotic monotherapy.' Fetched 2026-10-22 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/25640930/
- Risk of postpartum relapse in bipolar disorder and postpartum psychosis: a systematic review and meta-analysis. American Journal of Psychiatry, 2016. PMID 26514657 (PubMed record) What we read there: 37 articles, 5,700 deliveries in 4,023 patients; overall postpartum relapse 35% (95% CI 29-41). Women with a history of postpartum psychosis: 29% (20-41). Bipolar disorder: 66% (57-75) if medication free in pregnancy vs 23% (14-37) with prophylaxis. Fetched 2026-10-22 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/26514657/
- Long-term outcome of postpartum psychosis: a prospective clinical cohort study in 106 women. International Journal of Bipolar Disorders, 2021 (PMC8554899) What we read there: 'Over two thirds of the women included in this study did not have major psychiatric episodes outside of the postpartum period during follow-up. The overall recurrence rate of mood/psychotic episodes outside the postpartum period was ~32%.' 76.4% received lithium during admission. 'out of the 52 women who stopped lithium, five relapsed within six months of discontinuation.' Fetched 2026-10-22 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC8554899/
- Letter: Duration of episode in postpartum psychosis treated with antipsychotic medication and lithium. American Journal of Psychiatry, 2025;182(12):1094 (PMC13489038) What we read there: data from a prospective Dutch cohort at Erasmus MC since 2005: '102 postpartum psychosis patients treated with antipsychotics and lithium and 22 patients with antipsychotics monotherapy'. Mean time from onset to full remission 7.4 weeks with combined treatment. No one-year relapse comparison is reported. Fetched 2026-10-22 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC13489038/
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