Zuranolone (Zurzuvae): Postpartum Depression Evidence & Risks

Table of contents

• Evidence summary
• What zuranolone and brexanolone are
• How they work
• What zuranolone is prescribed for, by country
• What works and what does not
• Benefits by claim
• Risks and all side effects
• All interactions
• Who should avoid zuranolone
• Dosage and how it is taken (label doses only)
• Follow the money: who makes it and who funded the evidence
• Related research
• Frequently asked questions
• Sources and funding notes

Evidence summary

Independent evidence and credibility scorecard

Independence tiers: 1 = regulator or government body; 2 = academic or Cochrane team with no declared commercial funding for the work, or no funding statement found; 3 = independent authors analysing mostly manufacturer-run trials, or minor declared industry ties; 4 = manufacturer-funded or manufacturer-authored. Credibility runs from A (highest) to D.

Note on clustering: all six Cochrane trials were run in the US and sponsored by two drug makers (Sage for five, Marinus for one). The independent appraisals come from the UK (Cochrane, MHRA), the EU (EMA) and the US (FDA and the NIH-linked label data). None of them ran a placebo-controlled trial of their own in this review.

What zuranolone and brexanolone are

Zuranolone (brand name Zurzuvae) is a synthetic neurosteroid that acts on GABA-A receptors. It is taken as a capsule once a day for 14 days. The US label lists 20 mg, 25 mg and 30 mg capsules (FDA label).

Brexanolone (Zulresso) is a neurosteroid given as a continuous intravenous infusion in hospital, with close monitoring because of the risk of excessive sedation (Cochrane 2025). Its US commercial supply ended on 31 December 2024 (Sage 10-K).

Legal and regulatory status (checked October 2026)

How they work

Both drugs are positive allosteric modulators of GABA-A receptors, the brain's main inhibitory ("braking") system (Cochrane 2025).

• The theory: the natural brain steroid allopregnanolone is thought to fall sharply after birth, and this drop may contribute to postnatal depression (Cochrane 2025). This is a hypothesis. It is not proven to explain who gets depressed.
• What is shown in humans: in trials, the drugs reduce depression scores more than placebo, and their most common side effect is sleepiness (FDA label).
• Not a standard antidepressant: the label says zuranolone can be used alone or with an oral antidepressant (FDA label). It works through GABA-A rather than the serotonin system used by SSRIs.

What zuranolone is prescribed for, by country

Licensed uses

In the US, the licensed use is postpartum depression in adults. The FDA called it the first oral medication indicated for that use (FDA). The trials enrolled women with severe postpartum depression. In the UK, the licence covers moderate or severe postnatal depression (MHRA). In Japan, the 30 mg capsule is approved for "depression and depressive state" (Shionogi).

Where guidelines place it

We could not check the current NICE position or the 2026 ACOG practice update in this review. Both are listed in the gaps section. Do not treat either as settled.

What works and what does not

Benefits by claim

Does zuranolone work for postpartum depression?

Yes, better than placebo, by a modest margin. In SKYLARK, 196 women were randomised to zuranolone 50 mg or placebo, and 170 (86.7%) completed the 45 days. Scores fell 15.6 points on zuranolone and 11.6 points on placebo, a difference of 4.0 points (95% CI −6.3 to −1.7) at day 15 (Deligiannidis 2023; sponsored by Sage and Biogen). ROBIN, with 153 women and an earlier capsule, found −17.8 vs −13.6, a difference of 4.2 points (−6.9 to −1.5) (Deligiannidis 2021; sponsored by Sage).

On a 0 to 52 scale, a 4-point gap is modest, and most of the improvement in both arms happened on placebo too. Cochrane's pooled analysis of the two oral trials found response about a quarter higher (RR 1.26) and remission about two-thirds higher (RR 1.65) at day 45, rated moderate certainty (Cochrane 2025).

Does it work faster than antidepressants?

No trial has compared zuranolone directly with an SSRI or with therapy, so "faster than antidepressants" is an inference, not a tested result (Cochrane 2025). The trials recorded early separation from placebo. We did not extract those early figures in this review.

Does the benefit last?

Up to day 45 in the trials, there was still a gap: −4.1 points in ROBIN (95% CI −6.7 to −1.4) (ROBIN). Beyond that, the data are not in the Cochrane review, so the duration of benefit is unknown.

Trial design and the speed of review

Sleepiness is common on the drug, so participants and raters may guess who is on it. The label reports that 16% of study-1 patients needed a dose reduction, mostly for somnolence (FDA label). This is our reasoning, not a finding of the trial authors. In 2026, regulatory-science researchers argued that priority review and fast-track designation, which took about 7 months here, lowered the bar for evidence of benefit (Cosgrove 2026). That is an argument about process, not new trial data.

Brexanolone

Two Sage-sponsored brexanolone trials were published in 2018. We did not open them in this review, so we do not quote their effect sizes. Cochrane's pooled analysis of intravenous neurosteroids at 30 days found low-certainty evidence of little or no difference in response (RR 1.24, 95% CI 0.74 to 2.06) or remission (RR 1.18, 0.59 to 2.38) (Cochrane 2025).

Major depression outside the postpartum period

The FDA refused the major depressive disorder application in August 2023, and Sage and Biogen agreed not to pursue further US development for that use (Sage 10-K). A 2024 meta-analysis reported day-15 differences of −2.40 (95% CI −3.07 to −1.63) for major depression and −4.06 (−4.25 to −3.87) for postpartum depression (Winslow 2024). Japan approved the 30 mg capsule for "depression and depressive state" in December 2025 (Shionogi).

Risks and all side effects

Common side effects (trials)

• Somnolence (sleepiness): 36% vs 6% in study 1 (50 mg vs placebo); 19% vs 11% in study 2 (about 40 mg marketed equivalent vs placebo)
• Dose reduction needed for an adverse reaction: 16% in study 1, mainly for somnolence (10%) and dizziness (6%); 4% in study 2

Source: FDA label. The full adverse-event table was not extracted in this review, so other common effects are not listed here.

Serious and rare risks

• Central nervous system depression: sedation is the main serious risk, and the label links it to the driving warning and to dose reduction (FDA label).
• Suicidal thoughts (post-marketing signal): a 2025 FAERS analysis of 154 reports found a disproportionality signal for suicidal ideation, plus signals for insomnia, vertigo and blurred vision that the label does not list. Spontaneous reports cannot prove cause (FAERS analysis).
• Harm to an unborn baby: the EU and UK say it must not be taken in pregnancy (EMA; MHRA).

Misuse, dependence and withdrawal

Zuranolone is a controlled drug (CIV). In a human abuse-potential study, its abuse potential was "comparable to alprazolam" (FDA label). We did not extract the label's withdrawal section in this review.

Overdose

Seek emergency medical help for any suspected overdose. We did not extract the label's overdose section in this review.

Brexanolone risks

Brexanolone was given with close monitoring because of the risk of excessive sedation (Cochrane 2025).

All interactions

Source: FDA label (dose adjustments and drug-interaction sections). Many common medicines and some herbal products affect CYP3A4, so a pharmacist should check every medicine and supplement before a course starts.

Who should avoid zuranolone

Contraindications

The EU and UK contraindicate zuranolone in pregnancy (EMA; MHRA).

Use with care

• Anyone who must drive, care for a baby alone at night, or operate machinery within 12 hours of a dose. The MHRA asks prescribers to weigh sedation in a new mother.
• People with a history of substance misuse, because misuse potential is documented (see Misuse above).
• Bipolar disorder or psychosis (our reading): the trials studied severe postpartum depression. Postpartum psychosis is an emergency that needs specialist care, not a 14-day pill.

Pregnancy, contraception and breastfeeding

Women who could become pregnant should use effective contraception during treatment and for one week after the last dose (FDA label). In a study of 14 healthy breastfeeding women on 30 mg for 5 days, the mean relative infant dose was 0.357% (FDA label). The label says there are no data on effects in breastfed infants, so the benefit and risk should be weighed with a prescriber.

Children, older adults, liver and kidney disease

The label sets lower doses for severe liver impairment and for moderate or severe kidney impairment (eGFR below 60) (FDA label). We did not extract the label's wording on children and older adults in this review.

Dosage and how it is taken (label doses only)

These are the doses stated on the US label, not personal advice. Only a prescriber can decide whether, and at what dose, zuranolone is appropriate.

• Label dose: 50 mg once daily in the evening for 14 days, taken with fat-containing food. If CNS depressant effects occur, the label allows a reduction to 40 mg once daily (FDA label).
• Label-specified lower doses: 30 mg once daily for 14 days in severe liver impairment, moderate or severe kidney impairment, or with strong CYP3A4 inhibitors (FDA label).
• Missed doses and stopping: discuss with the prescriber. We did not extract the label's wording on these points.
• Trial doses: SKYLARK used 50 mg; ROBIN used 30 mg of an earlier capsule, equivalent to about 40 mg of the marketed product (FDA label).
• Brexanolone (historical): given as an intravenous infusion in hospital with close monitoring (Cochrane 2025).

Follow the money: who makes it and who funded the evidence

Who made it, and who sells it now

• Sage Therapeutics (US) developed both drugs and sells zuranolone with Biogen (Sage 10-K).
• Biogen (US) entered a collaboration effective 31 December 2020. It paid an $875 million upfront fee and bought 6,241,473 Sage shares for $650 million, about $104.14 each, a 40% premium. Sage and Biogen split US operating profits and losses 50/50. Biogen holds rights outside the US except Japan, Taiwan and South Korea, and Sage received a $75 million milestone for the first US commercial sale (Sage 10-K). The EU authorisation holder is Biogen Netherlands B.V. (EMA).
• Shionogi (Japan) holds rights in Japan, Taiwan and South Korea. It paid $90 million upfront in 2018, with up to $485 million in further milestones (Sage 10-K).
• Supernus Pharmaceuticals (US) completed its acquisition of Sage on 31 July 2025. It paid $8.50 a share in cash plus a contingent value right of up to $3.50, tied to a Japanese sale and to US Zurzuvae net sales in a calendar year of $250 million, $300 million and $375 million. Supernus receives 50% of the total US net revenue Biogen records for Zurzuvae (Supernus release).
• Later change: on 1 August 2026 Supernus agreed a "merger of equals" with Indivior. Indivior shareholders would own about 56.5% of the combined company, to be named Supernus, Inc., subject to shareholder and antitrust approvals (SEC 8-K). We did not check whether the merger has closed.

Price

The US launch wholesale acquisition cost was $15,900 for a 14-day course, according to a November 2023 trade report (Drug Discovery & Development). We did not verify a current US price or a UK price.

Who funded the evidence

• All efficacy trials in the Cochrane review: the manufacturers sponsored all six included trials, five by Sage and one by Marinus, and Cochrane says the companies "appear to have had a considerable role" in design and conduct. Many trial authors declared employment or stock in the manufacturers (Cochrane 2025).
• SKYLARK and ROBIN: sponsored by Sage, with authors listed with Sage and Biogen affiliations (SKYLARK; ROBIN).
• Independent appraisals: the Cochrane review's funding statement could not be read in this check. The FAERS analysis had a university teaching-reform grant and declared no conflicts. The Cosgrove critique lists the Bioethics and Human Rights Lab as funder. We found no funding statement for the Winslow meta-analysis.
• Regulators' own funding: the EMA and the FDA both receive industry fees. These are structural ties, not payments for a particular decision. We did not verify their exact shares in this review.

Documented integrity and regulatory events

• Split FDA decision (2023): approval for postpartum depression, refusal for major depression, for lack of substantial evidence of effectiveness in that indication (Sage 10-K).
• Brexanolone supply ended: commercial availability in the US was discontinued as of 31 December 2024 (Sage 10-K). We could not confirm the company's stated reasons or any cost figure.
• Review-speed critique (2026): regulatory-science researchers argued the approval set a low evidence bar (Cosgrove 2026).

What this means: The main efficacy evidence comes from trials run and sponsored by the makers. Independent bodies (Cochrane, the EMA and the MHRA) have reviewed it and accepted a short-term benefit with safety conditions. No independent placebo-controlled trial appears in our sources. The open questions (how long the benefit lasts, how it compares with standard treatments, and whether it is worth the price) are the ones the company trials were not designed to answer.

Related research

For wider, evidence-graded context, see these Pure City Research guides:

• Brain stimulation for depression
• Cognitive behavioural therapy (CBT)
• Esketamine (Spravato): independent evidence
• Bupropion: independent evidence
• Escitalopram: independent evidence
• Citalopram: independent evidence
• Alprazolam (Xanax): independent evidence
• Antidepressant withdrawal and dependence
• Bipolar depression
• Depression relapse prevention

Frequently asked questions

Sources and funding notes

What we could not verify: NICE's draft recommendation, its committee dates and its expected final guidance, because NICE's pages returned errors in this review. We therefore make no claim about the English NHS decision. We also could not open the full 2023 FDA integrated review (the file was too large to read), the full text of the SKYLARK and ROBIN papers beyond their abstracts and the PubMed records, the 2018 brexanolone trial papers, or the ACOG 2026 practice update. We could not confirm the Cochrane review's own funding statement, the full competing-interest statement of the Cosgrove critique, the funder of the breastfeeding milk study, or whether a completed ICER assessment exists. We could not confirm the April 2025 brexanolone licence step, any brexanolone cost figure, a current 2026 US list price, the Scottish or EU reimbursement position, the Japanese milestone payment, or whether the Supernus–Indivior merger has closed. We did not search exhaustively for FDA warning letters or lawsuits, and we report none.

1. ZURZUVAE (zuranolone) US prescribing information, DailyMed (revised April 2026): regulator-approved label; the sponsor drafts the wording.
2. FDA: FDA approves first oral treatment for postpartum depression (4 August 2023): US regulator; FDA funded partly by industry user fees.
3. FDA integrated review, NDA 217369 (2023): cited for the existence of the review only; not opened in this review. US regulator.
4. EMA: Zurzuvae EPAR: EU regulator; partly funded by industry fees.
5. MHRA: approval of zuranolone (27 August 2025): UK regulator.
6. Wilson CA et al. Brexanolone, zuranolone and related neurosteroid GABAA receptor positive allosteric modulators for postnatal depression. Cochrane Database Syst Rev 2025 (CD014624): independent systematic review; sponsor-funded trials; funder not confirmed in this review.
7. Deligiannidis KM et al. Zuranolone for the Treatment of Postpartum Depression (SKYLARK). 2023: Sage and Biogen; authors with company affiliations.
8. Deligiannidis KM et al. Effect of Zuranolone vs Placebo in Postpartum Depression: A Randomized Clinical Trial (ROBIN). JAMA Psychiatry 2021: Sage; authors with company affiliations.
9. Post-marketing safety assessment of zuranolone for postpartum depression using FDA Adverse Event Reporting System data. Front Psychiatry 2025: academic; Gannan Medical University grant; no conflicts declared.
10. Cosgrove L et al. Zuranolone: a case study in (regulatory) rush to judgement? Br J Clin Pharmacol 2026: academic critique; Bioethics and Human Rights Lab listed as funder.
11. Winslow M et al. Efficacy of zuranolone versus placebo in postpartum depression and major depressive disorder: systematic review and meta-analysis. 2024: academic; no funding statement found in the record we read.
12. Sage Therapeutics Form 10-K for 2024: company filing (interested party); source of the deal terms and the FDA refusal wording.
13. Supernus: completion of Sage acquisition (31 July 2025): company release.
14. Supernus Form 8-K: merger agreement with Indivior (1 August 2026): company filing.
15. Shionogi: approval in Japan of ZURZUVAE 30 mg capsules (22 December 2025): company release.
16. Drug Discovery & Development: Zurzuvae launch wholesale price (7 November 2023): trade press.

Educational research reviewed 6 October 2026. This article is not medical advice; talk to a qualified clinician or pharmacist before starting, stopping or combining treatments.