Zuranolone

The active substance in Zurzuvae, a capsule for postpartum depression. It is a neuroactive steroid that acts on GABA-A receptors in the brain, as brexanolone does.

Drug
In this atlas

Overview

Technical summary

A positive allosteric modulator of synaptic and extrasynaptic GABA-A receptors and a neuroactive steroid (SKYLARK abstract). NIMH says it acts on similar brain receptors to brexanolone. RxNorm concept 2669905 (ingredient); FDA UNII 7ZW49N180B (openFDA drugsfda record). The approved product is Zurzuvae. The SKYLARK trial tested 50 mg a day for 14 days in severe postpartum depression (see the claims on this page). The recipient is the adult postpartum woman. Fetal harm is a labelled concern (FDA press announcement of 2023); no fetal or newborn outcome claim is recorded here.

About this drug

Level: Substance
Mechanism: Positive allosteric modulator of synaptic and extrasynaptic GABA-A receptors (neuroactive steroid), as described in the SKYLARK abstract.

Claims and evidence

Context Women aged 18 to 45 years with severe postpartum depression, 14-day oral course (SKYLARK, 2023)

Recipient: Maternal
Population: Women with severe postpartum depression who were randomized 1:1 to zuranolone 50 mg a day or placebo for 14 days. The registry gives the age range as 18 to 45 years and the sex as female. The abstract says severe postpartum depression but prints no score cut-off or onset window.
Care setting: Not stated in the abstract. The registry lists study sites in the United States, Spain and the United Kingdom.
Geography: United States, Spain and United Kingdom (registry locations)
Life stage: Postpartum. The abstract and the registry fields read give no postnatal age limits.
Gestational age: Not reported
Postnatal age: Not reported
Follow-up: 45-day study: 14 days of treatment, primary endpoint at day 15, secondary endpoints at days 3, 28 and 45 (abstract).

Eligibility: 2 inclusion criteria; 1 exclusion criterion

Included: Women aged 18 to 45 years (registry)
Severe postpartum depression (abstract)
Excluded: Not stated in the parts of the abstract and registry fields that were read

Effect Interventional Certainty not assessed

In SKYLARK, a double-blind phase 3 trial, 196 women with severe postpartum depression were randomized 1:1 to zuranolone 50 mg a day or placebo for 14 days (98 in each group), and 170 (86.7%) completed the 45-day study. At day 15, the primary endpoint, the average change from baseline in the 17-item Hamilton Depression Rating Scale (HAM-D) total score was -15.6 with zuranolone and -11.6 with placebo, a least-squares mean difference of -4.0 (95% CI -6.3 to -1.7). The authors report significant improvement at days 3, 28 and 45 too.

Source problem. A cited source is corrected. Check the sources before relying on this claim.
Applies to Maternal
- Women with severe postpartum depression who were randomized 1:1 to zuranolone 50 mg a day or placebo for 14 days. The registry gives the age range as 18 to 45 years and the sex as female. The abstract says severe postpartum depression but prints no score cut-off or onset window.
- Not stated in the abstract. The registry lists study sites in the United States, Spain and the United Kingdom.
- United States, Spain and United Kingdom (registry locations)
- Postpartum. The abstract and the registry fields read give no postnatal age limits.
- Gestational age Not reported
- Postnatal age Not reported
Follow-up 45-day study: 14 days of treatment, primary endpoint at day 15, secondary endpoints at days 3, 28 and 45 (abstract).
Certainty method
Not assessed

Limitations
Only the PubMed abstract was read, not the full paper. PubMed lists an erratum for the paper (PMID 39659161) that has no abstract and was not read, so these numbers are as printed in the abstract and may have been corrected. The registry names an industry lead sponsor (Biogen). The trial enrolled women with severe postpartum depression, so the result does not apply to milder illness. A rating-scale change at day 15 is not remission or relapse, and follow-up ended at day 45. The registry records 200 enrolled while the abstract prints 196 randomized; both counts are kept on the study page.

Last checked
2026-10-14

Next evidence check due
Not scheduled
Sources and citations (8)
1. Supports Zuranolone for the Treatment of Postpartum Depression
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. PubMed lists a published erratum for this paper (PMID 39659161). PubMed holds no abstract for the erratum, and its text was not read, so any figure taken from this abstract is stated as printed in the abstract and may have been corrected. · Published 2023-09-01 · DOI: 10.1176/appi.ajp.20220785 · PMID: 37491938
  
  Locator
  Abstract, METHODS, sentence 1, fragment 1 of 2
  
  How it applies
  Design, randomization and dose.
  
  Retrieved
  2026-10-14T18:18:20Z
  
  Source effective date
  Not reported
  
  Source state
  Corrected
  
  Snapshot hash
  12c6ed30532a24be00a480e7f5dcb6792d5416f1ad5788ee1cf27d8f2a0c0b35
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
2. Supports Zuranolone for the Treatment of Postpartum Depression
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. PubMed lists a published erratum for this paper (PMID 39659161). PubMed holds no abstract for the erratum, and its text was not read, so any figure taken from this abstract is stated as printed in the abstract and may have been corrected. · Published 2023-09-01 · DOI: 10.1176/appi.ajp.20220785 · PMID: 37491938
  
  Locator
  Abstract, RESULTS, sentence 1
  
  How it applies
  Numbers randomized and completing the study.
  
  Retrieved
  2026-10-14T18:18:20Z
  
  Source effective date
  Not reported
  
  Source state
  Corrected
  
  Snapshot hash
  12c6ed30532a24be00a480e7f5dcb6792d5416f1ad5788ee1cf27d8f2a0c0b35
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
3. Supports Zuranolone for the Treatment of Postpartum Depression
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. PubMed lists a published erratum for this paper (PMID 39659161). PubMed holds no abstract for the erratum, and its text was not read, so any figure taken from this abstract is stated as printed in the abstract and may have been corrected. · Published 2023-09-01 · DOI: 10.1176/appi.ajp.20220785 · PMID: 37491938
  
  Locator
  Abstract, METHODS, sentence 2, fragment 1 of 3
  
  How it applies
  Primary endpoint and its timing.
  
  Retrieved
  2026-10-14T18:18:20Z
  
  Source effective date
  Not reported
  
  Source state
  Corrected
  
  Snapshot hash
  12c6ed30532a24be00a480e7f5dcb6792d5416f1ad5788ee1cf27d8f2a0c0b35
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
4. Supports Zuranolone for the Treatment of Postpartum Depression
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. PubMed lists a published erratum for this paper (PMID 39659161). PubMed holds no abstract for the erratum, and its text was not read, so any figure taken from this abstract is stated as printed in the abstract and may have been corrected. · Published 2023-09-01 · DOI: 10.1176/appi.ajp.20220785 · PMID: 37491938
  
  Locator
  Abstract, RESULTS, sentence 2, fragment 1 of 3
  
  How it applies
  Statistically significant improvement at day 15.
  
  Retrieved
  2026-10-14T18:18:20Z
  
  Source effective date
  Not reported
  
  Source state
  Corrected
  
  Snapshot hash
  12c6ed30532a24be00a480e7f5dcb6792d5416f1ad5788ee1cf27d8f2a0c0b35
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
5. Supports Zuranolone for the Treatment of Postpartum Depression
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. PubMed lists a published erratum for this paper (PMID 39659161). PubMed holds no abstract for the erratum, and its text was not read, so any figure taken from this abstract is stated as printed in the abstract and may have been corrected. · Published 2023-09-01 · DOI: 10.1176/appi.ajp.20220785 · PMID: 37491938
  
  Locator
  Abstract, RESULTS, sentence 2, fragment 2 of 3
  
  How it applies
  Mean changes and the difference with its interval.
  
  Retrieved
  2026-10-14T18:18:20Z
  
  Source effective date
  Not reported
  
  Source state
  Corrected
  
  Snapshot hash
  12c6ed30532a24be00a480e7f5dcb6792d5416f1ad5788ee1cf27d8f2a0c0b35
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
6. Supports Zuranolone for the Treatment of Postpartum Depression
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. PubMed lists a published erratum for this paper (PMID 39659161). PubMed holds no abstract for the erratum, and its text was not read, so any figure taken from this abstract is stated as printed in the abstract and may have been corrected. · Published 2023-09-01 · DOI: 10.1176/appi.ajp.20220785 · PMID: 37491938
  
  Locator
  Abstract, RESULTS, sentence 2, fragment 3 of 3
  
  How it applies
  Improvement also reported at days 3, 28 and 45.
  
  Retrieved
  2026-10-14T18:18:20Z
  
  Source effective date
  Not reported
  
  Source state
  Corrected
  
  Snapshot hash
  12c6ed30532a24be00a480e7f5dcb6792d5416f1ad5788ee1cf27d8f2a0c0b35
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
7. Background A Study to Evaluate the Efficacy and Safety of SAGE-217 in Participants With Severe Postpartum Depression (PPD)
  
  Registry · U.S. National Library of Medicine (ClinicalTrials.gov) · Edition ClinicalTrials.gov API v2 record, last update posted 2023-11-30. The record holds a results section; the atlas does not copy registry result tables. · NCT: NCT04442503
  
  Locator
  protocolSection.sponsorCollaboratorsModule.leadSponsor.name
  
  How it applies
  Registry lead sponsor.
  
  Retrieved
  2026-10-14T18:18:23Z
  
  Source effective date
  2023-11-30
  
  Source state
  Current
  
  Snapshot hash
  fdd893c7dccd1ed53eea5282f35f5e206c6ec2da60ddc2338db12152aff132dd
  
  Reuse basis
  Registry record. The raw API JSON is committed. The ClinicalTrials.gov terms page is rendered by JavaScript and its text was not read by a script, so the redistribution basis is unverified; any quotation stays short. The raw file is committed at research/snapshots/mj-ctgov-NCT04442503.json.
8. Background Correction to Deligiannidis et al
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. This is a published erratum (a correction notice). PubMed holds no abstract for it; only its bibliographic record was read. · Published 2025-03-01 · DOI: 10.1176/appi.ajp.20220785correction · PMID: 39659161
  
  Locator
  PublicationType
  
  How it applies
  PubMed lists a published erratum for this paper (PMID 39659161). It has no abstract and was not read.
  
  Retrieved
  2026-10-14T18:19:05Z
  
  Source effective date
  Not reported
  
  Source state
  Current
  
  Snapshot hash
  7f4a25189d34399d741dc149248c7d6967002b6a87356a099568b41ae8b38ad0
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
Safety Interventional Certainty not assessed

In SKYLARK the most common adverse events (10% or more) with zuranolone were somnolence, dizziness and sedation. The abstract reports no loss of consciousness, no withdrawal symptoms and no increased suicidal ideation or behavior in the trial. The US label for Zurzuvae carries a boxed warning that it causes driving impairment, and FDA says it may cause fetal harm.

Source problem. A cited source is corrected. Check the sources before relying on this claim.
Applies to Maternal
- Women with severe postpartum depression who were randomized 1:1 to zuranolone 50 mg a day or placebo for 14 days. The registry gives the age range as 18 to 45 years and the sex as female. The abstract says severe postpartum depression but prints no score cut-off or onset window.
- Not stated in the abstract. The registry lists study sites in the United States, Spain and the United Kingdom.
- United States, Spain and United Kingdom (registry locations)
- Postpartum. The abstract and the registry fields read give no postnatal age limits.
- Gestational age Not reported
- Postnatal age Not reported
Follow-up 45-day study: 14 days of treatment, primary endpoint at day 15, secondary endpoints at days 3, 28 and 45 (abstract).
Certainty method
Not assessed

Limitations
The abstract gives no adverse event counts, and the trial was small (196 randomized) and short (45 days), so rare harms cannot be excluded. The label statements come from today's label text in openFDA, not from the trial. Only the abstract was read, and an erratum exists that was not read.

Last checked
2026-10-14

Next evidence check due
Not scheduled
Sources and citations (5)
1. Supports Zuranolone for the Treatment of Postpartum Depression
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. PubMed lists a published erratum for this paper (PMID 39659161). PubMed holds no abstract for the erratum, and its text was not read, so any figure taken from this abstract is stated as printed in the abstract and may have been corrected. · Published 2023-09-01 · DOI: 10.1176/appi.ajp.20220785 · PMID: 37491938
  
  Locator
  Abstract, RESULTS, sentence 4
  
  How it applies
  Most common adverse events in the trial.
  
  Retrieved
  2026-10-14T18:18:20Z
  
  Source effective date
  Not reported
  
  Source state
  Corrected
  
  Snapshot hash
  12c6ed30532a24be00a480e7f5dcb6792d5416f1ad5788ee1cf27d8f2a0c0b35
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
2. Supports Zuranolone for the Treatment of Postpartum Depression
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. PubMed lists a published erratum for this paper (PMID 39659161). PubMed holds no abstract for the erratum, and its text was not read, so any figure taken from this abstract is stated as printed in the abstract and may have been corrected. · Published 2023-09-01 · DOI: 10.1176/appi.ajp.20220785 · PMID: 37491938
  
  Locator
  Abstract, RESULTS, sentence 5
  
  How it applies
  Events the abstract reports as not observed.
  
  Retrieved
  2026-10-14T18:18:20Z
  
  Source effective date
  Not reported
  
  Source state
  Corrected
  
  Snapshot hash
  12c6ed30532a24be00a480e7f5dcb6792d5416f1ad5788ee1cf27d8f2a0c0b35
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.
3. Background ZURZUVAE drug label (openFDA record, set id f18e53b0-d0bb-422d-8de7-ab64b7292b29)
  
  Label · U.S. Food and Drug Administration (openFDA drug label endpoint) · Edition openFDA drug label record, label effective 20260615, dataset last updated 2026-10-13. · SPL set id: f18e53b0-d0bb-422d-8de7-ab64b7292b29
  
  Locator
  results[0].boxed_warning[0]
  
  How it applies
  Boxed warning in the current label.
  
  Retrieved
  2026-10-14T18:20:51Z
  
  Source effective date
  2026-06-15
  
  Source state
  Current
  
  Snapshot hash
  f3b10768b79c29b67094f1486e3cc7023e968b4ebb6dea5247b472eedc40a869
  
  Reuse basis
  openFDA data is public domain under a Creative Commons CC0 1.0 Universal dedication unless noted (https://open.fda.gov/terms/, https://open.fda.gov/license/). The label text is the manufacturer's labeling as FDA publishes it; check the openFDA documentation for a not-in-the-public-domain warning before quoting beyond short passages. The raw file is committed at research/snapshots/mj-openfda-label-f18e53b0-d0bb-422d-8de7-ab64b7292b29.json.
4. Background FDA approves first oral treatment for postpartum depression (FDA press announcement)
  
  Regulator · U.S. Food and Drug Administration · Edition Press announcement dated August 04, 2023. The page says its content was current as of 08/22/2023. · Published 2023-08-04
  
  Locator
  Paragraph 6, sentence 3
  
  How it applies
  FDA press statement at approval.
  
  Retrieved
  2026-10-14T18:20:21Z
  
  Source effective date
  2023-08-22
  
  Source state
  Current
  
  Snapshot hash
  ed96e61675291b5318abb26ded7e0d06866745024b486e763b58e9b91ef997ad
  
  Reuse basis
  FDA website policies page (fetched as mj-fda-website-policies.html, sha256 59b7c85417532ac45d5c62ed35646d399324edfc2cee023684995302e4eeb9b3) says that, unless otherwise noted, the contents of www.fda.gov are not copyrighted, are in the public domain, and may be republished without permission. The page bytes are not committed. Short quotations with attribution.
5. Background Correction to Deligiannidis et al
  
  Publication · The American journal of psychiatry · Edition PubMed record (abstract and bibliographic data) from NCBI E-utilities. The full text and the version of record were not fetched. This is a published erratum (a correction notice). PubMed holds no abstract for it; only its bibliographic record was read. · Published 2025-03-01 · DOI: 10.1176/appi.ajp.20220785correction · PMID: 39659161
  
  Locator
  PublicationType
  
  How it applies
  PubMed lists a published erratum for the SKYLARK paper (PMID 39659161). It was not read.
  
  Retrieved
  2026-10-14T18:19:05Z
  
  Source effective date
  Not reported
  
  Source state
  Current
  
  Snapshot hash
  7f4a25189d34399d741dc149248c7d6967002b6a87356a099568b41ae8b38ad0
  
  Reuse basis
  PubMed record fetched through NCBI E-utilities. Only the hash of the bytes is kept, because publisher copyright may apply to the abstract (https://pubmed.ncbi.nlm.nih.gov/disclaimer/). Short factual quotations under 25 words, with the citation.

Clinical links

Each link reports what one cited claim says about one population. "Evaluated for" means studied, not shown to work. These links do not show that anything is effective or right for you.

Claim, not a settled fact SKYLARK: HAM-D change at day 15, zuranolone 50 mg against placebo, severe postpartum depression: Postpartum depression (mother) (Claimed to treat)
Applies to Maternal
- Women with severe postpartum depression who were randomized 1:1 to zuranolone 50 mg a day or placebo for 14 days. The registry gives the age range as 18 to 45 years and the sex as female. The abstract says severe postpartum depression but prints no score cut-off or onset window.
- Not stated in the abstract. The registry lists study sites in the United States, Spain and the United Kingdom.
- United States, Spain and United Kingdom (registry locations)
- Postpartum. The abstract and the registry fields read give no postnatal age limits.
- Gestational age Not reported
- Postnatal age Not reported
Follow-up 45-day study: 14 days of treatment, primary endpoint at day 15, secondary endpoints at days 3, 28 and 45 (abstract).
In SKYLARK, a double-blind phase 3 trial, 196 women with severe postpartum depression were randomized 1:1 to zuranolone 50 mg a day or placebo for 14 days (98 in each group), and 170 (86.7%) completed the 45-day study. At day 15, the primary endpoint, the average change from baseline in the 17-item Hamilton Depression Rating Scale (HAM-D) total score was -15.6 with zuranolone and -11.6 with placebo, a least-squares mean difference of -4.0 (95% CI -6.3 to -1.7). The authors report significant improvement at days 3, 28 and 45 too.
Source problem. A cited source is corrected. Check the sources before relying on this claim.
Limitations. Only the PubMed abstract was read, not the full paper. PubMed lists an erratum for the paper (PMID 39659161) that has no abstract and was not read, so these numbers are as printed in the abstract and may have been corrected. The registry names an industry lead sponsor (Biogen). The trial enrolled women with severe postpartum depression, so the result does not apply to milder illness. A rating-scale change at day 15 is not remission or relapse, and follow-up ended at day 45. The registry records 200 enrolled while the abstract prints 196 randomized; both counts are kept on the study page.

Record details

Record id: maternal:zuranolone
Content hash: 208e7c08f7c15177e410e3c32e55ab198d52dd7c7c17cb7a9c065c6a0211d382