Progesterone

What progesterone is and how it works

Progesterone is the steroid hormone secreted by the corpus luteum after ovulation and by the placenta in pregnancy. It acts at nuclear progesterone receptors, and its best-characterised job is converting an estrogen-primed endometrium from a proliferative to a secretory state — which is why the approved US indication that matters most is not a treatment for anything a reader feels, but protection of the uterine lining in a woman taking estrogen. Progesterone metabolites also act at GABA-A receptors, which is the mechanism behind its sedative reputation and the reason its trials give it at bedtime.

The single most important boundary on this page is an identity one, and the most-cited "progesterone" risk finding in the world falls on the wrong side of it. Micronized progesterone is not medroxyprogesterone acetate. It is not norethindrone, not levonorgestrel, not drospirenone, and it is not "progestins" as a class. Medroxyprogesterone acetate is a synthetic derived from 17-alpha-hydroxyprogesterone with glucocorticoid and partial androgenic activity that progesterone does not have. The Women's Health Initiative's estrogen-plus-progestin arm administered oral conjugated equine estrogens 0.625 mg daily with medroxyprogesterone acetate 2.5 mg daily — in one tablet, to 16,608 women aged 50 to 79 with an intact uterus — and it contained neither estradiol nor progesterone.^[15] A sentence that attaches one of that trial's hazard ratios to "progesterone" is describing a different molecule.

Preparation and route are the second boundary. Oral micronized progesterone is a capsule suspended in oil and reaches the circulation after first-pass metabolism; vaginal preparations achieve high uterine and lower systemic exposure; and compounded transdermal creams are a third thing again. A systematic review of endometrial outcomes states plainly that transdermal micronized progesterone does not provide endometrial protection,^[2] which is a specific, testable claim about a specific preparation, and is the reason "bioidentical progesterone cream" cannot be treated as a substitute for the approved capsule.

What the research actually shows

The approved indication has the strongest evidence, and it is about the uterine lining rather than about how anyone feels. In the PEPI trial, postmenopausal women given conjugated equine estrogens alone developed simple hyperplasia in 27.7% of cases against 0.8% on placebo, complex hyperplasia in 22.7% against 0.8%, and atypical hyperplasia in 11.8% against 0%, all P < .001. Women given one of three estrogen-plus-progestogen regimens — cyclic medroxyprogesterone acetate, continuous medroxyprogesterone acetate, or cyclic oral micronized progesterone 200 mg daily for the first 12 days — had hyperplasia rates statistically indistinguishable from placebo, P = .16.^[1] A systematic review adds the boundaries that the trial does not: sequential oral micronized progesterone at 200 mg a day for 12 to 14 days a month protects the endometrium for up to 5 years, vaginal administration is off-label and may protect, and transdermal micronized progesterone does not.^[2]

For assisted reproduction the evidence is broad and low-quality at the same time. A Cochrane review of 94 randomized trials in 26,198 women found live birth or ongoing pregnancy higher with progesterone than with placebo or no treatment, OR 1.77, 95% CI 1.09 to 2.86 — from five trials in 642 women, graded very low quality, in a review whose authors record that most included studies carried unclear or high risk of bias. The same review found no evidence that route of administration improved outcomes.^[4]

The claim most readers arrive with is sleep, and the randomized evidence for it is two studies with single-digit and low-double-digit samples. A randomized placebo-controlled study in 8 postmenopausal women without sleep complaints, given 300 mg orally at 23:00 for three weeks, found progesterone had no effect on undisturbed sleep; it reduced wake time and increased slow-wave sleep only on the night when sleep was deliberately disrupted by 15-minute blood sampling, which its authors read as physiologic regulation rather than a hypnotic effect.^[7] A randomized double-blind crossover study in 10 healthy postmenopausal women given the same amount for 21 days found decreased intermittent wake time and increased REM sleep in the first third of the night, with no significant differences in EEG power spectra and no effect on cognitive performance.^[8] The largest relevant randomized trial did not set out to test sleep at all: a 4-month Canada-wide Phase III trial in 189 perimenopausal women MISSED its primary vasomotor endpoint, third-month score difference -1.51, 95% CI -3.97 to 0.95, P = 0.222, and reported improved sleep quality, P = 0.005, as a participant-perceived secondary outcome.^[10]

On hot flushes the two trials disagree, and they come from one group. A 12-week randomized placebo-controlled trial in 133 healthy postmenopausal women found a mean adjusted difference of -4.3, 95% CI -6.6 to -1.9, favouring oral micronized progesterone 300 mg at bedtime.^[9] The later Phase III trial in 189 perimenopausal women missed its primary endpoint, and its authors state it was underpowered and that its confidence interval could not exclude a minimal clinically important difference.^[10] Comparing the two papers' author lists by exact match on PubMed's own "Surname Initials" strings, both authors of the 2012 trial appear on the 2023 trial — so these are not two independent groups converging, and the indication is off-label on every approved label read for this page.

Where large trials have tested progesterone against hard clinical endpoints, the results have mostly been null. PRISM randomized 4,153 women with bleeding in early pregnancy to vaginal progesterone 400 mg twice daily or placebo; live birth at or after 34 weeks was 75% against 72%, relative rate 1.03, 95% CI 1.00 to 1.07, P = 0.08, a missed primary endpoint.^[11] Two phase 3 trials in traumatic brain injury, published the same day, both failed: PROTECT III stopped for futility after 882 of a planned 1,140 patients, relative benefit 0.95, 95% CI 0.85 to 1.06,^[13] and SYNAPSE, in 1,195 patients, returned an odds ratio of 0.96, 95% CI 0.77 to 1.18, favourable outcomes 50.4% against 50.5% on placebo, its authors noting the contrast with robust preclinical data.^[14] The one large positive signal is route- and population-bounded: an individual-participant meta-analysis of 31 trials in 11,644 women found vaginal progesterone reduced birth before 34 weeks in high-risk singleton pregnancies, RR 0.78, 95% CI 0.68 to 0.90, with no benefit in twins, RR 1.01, 95% CI 0.84 to 1.20; the oral estimate in that same analysis was numerically favourable, RR 0.60, 95% CI 0.40 to 0.90, but rests on two trials in 181 women, and the authors' own verbatim conclusion is that "Evidence for oral progesterone is insufficient to support its use."^[12]

Where the evidence is weak

The read-across problem is this page's central weakness and it cannot be solved by more reading. The largest randomized safety dataset about "estrogen plus progestin" administered medroxyprogesterone acetate, and its 13-year follow-up shows why that matters: the estrogen-plus-progestin arm's breast-cancer excess persisted at HR 1.28, 95% CI 1.11 to 1.48, while the estrogen-alone arm's cumulative breast cancer was significantly reduced, HR 0.79, 95% CI 0.65 to 0.97.^[16] The progestogen is where those two arms differ. Whether micronized progesterone would have behaved like medroxyprogesterone acetate has not been tested in a trial of that size, and the evidence that it would not is observational: the E3N cohort, 2,354 invasive breast cancers among 80,377 postmenopausal women, reported RR 1.00, 95% CI 0.83 to 1.22 for estrogen with progesterone against 1.69, 95% CI 1.50 to 1.91 for estrogen with other progestagens — and 1.29, 95% CI 1.02 to 1.65 for estrogen alone, which is the part usually dropped.^[5]

"Bioidentical" is described here, not endorsed. The structural-identity argument has specific empirical support on three endpoints and no general support as a safety claim. The three are endometrial protection at defined oral amounts,^[2] a lower observational breast-cancer risk than other progestagens,^[5] and a venous thromboembolism odds ratio of 0.7, 95% CI 0.3 to 1.9, against 3.9, 95% CI 1.5 to 10.0 for norpregnane derivatives — a null and a signal, not a demonstration of safety.^[6] No FDA approval has been granted to any product on the basis of bioidentity.

Hormonal effect and clinical outcome are not the same thing. Progesterone changes endometrial histology, sleep architecture on an EEG, and hormone secretion profiles — and it did not change live births in 4,153 women with early-pregnancy bleeding,^[11] nor outcomes in 2,077 patients across two brain-injury trials.^[13][14] A measurable biological effect is a reason to run a trial, not a substitute for its result.

The sleep and mood evidence is small, short and mostly not primary. The two randomized sleep studies enrolled 8 and 10 participants and ran three weeks each.^[7][8] A PubMed search on 1 September 2026 for the term "micronized progesterone" AND (sleep OR insomnia) AND randomized controlled trial[pt] returned 11 records; reading all 11, the largest with sleep as its primary outcome enrolled 100 women complaining of insomnia and compared two progestogens given alongside estradiol valerate, with no placebo arm — so it cannot establish that progesterone improves sleep, only that it differed or did not differ from another progestogen. On anxiety, a search on the same date for the term "micronized progesterone" AND anxiety AND randomized controlled trial[pt] returned 9 records, and for the term "micronized progesterone" AND (libido OR "sexual desire") AND randomized controlled trial[pt] returned 0. For scale, progesterone[MeSH] returned 75,938 records in PubMed on that date and progesterone[MeSH] AND randomized controlled trial[pt] returned 2,868 — a very large literature with almost nothing in it on the specific claims that sell this molecule.

The absorption question for compounded creams is real and published. A systematic review states that transdermal micronized progesterone does not provide endometrial protection, which means a cream cannot be assumed to be delivering what a capsule delivers merely because the molecule is the same.^[2]

And long-term safety is bounded by its own sources. The favourable breast-cancer conclusion for oral micronized progesterone stops at five years by its authors' own wording, and the same authors record limited evidence of increased risk beyond that.^[3] Both of the systematic reviews carrying the favourable molecule-specific conclusions on this page were written by the same three-author expert panel — compared by exact match on PubMed's "Surname Initials" author strings, all three authors are shared — which is worth knowing when they are read as two sources agreeing.

Legal and regulatory status

Progesterone is FDA-approved, and this is the compound on which the distance between the badge and the reason people arrive is widest.

Queried on 1 September 2026 in openFDA's Drugs@FDA database, the field products.active_ingredients.name for the quoted term "PROGESTERONE" returned 26 applications, and enumerating the ingredient names actually present in those records shows all 26 carry an ingredient spelled exactly PROGESTERONE — this molecule is clean of the token-contamination that inflates the equivalent estradiol query, which is precisely why the check has to be run rather than assumed. The second field on the same date and database, openfda.generic_name for "progesterone", returned 16 applications, a strict subset of the 26 with no rows unique to it; the ten it misses are not absent from the record, their indexed generic-name block is simply unpopulated. That is a property of the index, not a regulatory fact, and writing "16 applications exist" would be false. A same-session control on the ingredient-name field returned 88 applications for TESTOSTERONE and an invalid ingredient name returned HTTP 404 NOT_FOUND. The 26 are 10 NDAs and 16 ANDAs; product routes are 23 oral, 9 injection, 7 vaginal and 1 intrauterine; marketing status is 29 Prescription and 11 Discontinued.

The approved indications are narrow and their wording is the whole point. The oral capsule under NDA 019781 is, verbatim, "indicated for use in the prevention of endometrial hyperplasia in nonhysterectomized postmenopausal women who are receiving conjugated estrogens tablets. They are also indicated for use in secondary amenorrhea." The vaginal product under NDA 022057 is indicated "as part of an Assisted Reproductive Technology (ART) treatment program, to support embryo implantation in infertile women less than 35 years of age by supplementation of corpus luteal function and to maintain pregnancy after implantation up to 10 weeks." Read the first one closely: the approved use is endometrial protection in women receiving conjugated estrogens tablets — not in women receiving transdermal estradiol. Combining transdermal estradiol with oral micronized progesterone, which is the regimen much of contemporary menopausal practice uses, is therefore off-label on the progesterone label's own terms. That is a statement about labeling, not a judgment about the practice, and this page states it because softening it would misrepresent the badge.

Whether progesterone "carries a boxed warning" depends on which product is being discussed, and both answers are true of something. Queried on openFDA's SPL labeling endpoint on 1 September 2026, openfda.generic_name for "progesterone" returned 98 label records, of which 38 have a populated boxed_warning field; a same-session warfarin control on the identical field pair returned a populated boxed warning, so the field is real and the query is sound. The record read for the reference oral capsule, NDA 019781, has no boxed_warning field at all, and neither does the vaginal product under NDA 022057. A generic oral progesterone capsule under ANDA 200900 does carry one — and it is headed "WARNING: CARDIOVASCULAR DISORDERS, BREAST CANCER and PROBABLE DEMENTIA FOR ESTROGEN PLUS PROGESTIN THERAPY" and recites the Women's Health Initiative conjugated-equine-estrogen-plus-medroxyprogesterone-acetate results. It is a warning about a combination regimen using a different progestogen, imported onto a progesterone-only label. Naming the product is therefore not pedantry: "progesterone carries a boxed warning" is true of some approved labels and false of others.

A labeling fact worth raising with a prescriber: the reference oral capsule's own contraindications section states that it "contains peanut oil and is contraindicated in patients allergic to peanuts," alongside contraindications for undiagnosed abnormal genital bleeding, known or suspected or prior breast cancer, active or prior deep vein thrombosis or pulmonary embolism, active or prior arterial thromboembolic disease, and known liver dysfunction or disease. The labeled dosing schedules that accompany these indications are facts about a regulated document rather than research ranges, which is why they are described here rather than listed in the protocol block.

What the approval does not cover. Searching the same SPL endpoint on 1 September 2026, restricted to openfda.generic_name "progesterone" and the indications_and_usage field, returned zero records (HTTP 404 NOT_FOUND) for each of the quoted phrases "insomnia", "sexual desire", "anxiety", "premenstrual" and "anti-aging", while the same field on the same date returned 48 records for "endometrial hyperplasia", 54 for "amenorrhea" and 5 for "assisted reproductive". The phrases "sleep", "libido" and "hot flashes" did return records — 1, 4 and 10 respectively — and reading them is the point: adding the filter _exists_:openfda.application_number reduced all three to zero, against 48 for "endometrial hyperplasia" under the identical filter. Every one of those hits belongs to an unapproved homeopathic listing whose own label text says the claims are based on traditional homeopathic practice and are not FDA-evaluated. No approved application carries sleep, libido or hot flashes as an indication.

The National Drug Code directory makes the same point in one line. Queried on 1 September 2026 on its active_ingredients.name field for "PROGESTERONE", it returned 211 listing records, all 211 naming that exact ingredient — and reading marketing_category on every one of them, only 80 are approved prescription drugs under an NDA or ANDA. The other 131 are bulk ingredient, unapproved homeopathic, drug for further processing, bulk ingredient for compounding, unapproved drug other and export only. A non-empty search result on this endpoint is not an approval, and here most of the rows are not one.

Progesterone is prescription-only in the United States, and legitimate access runs through a licensed prescriber who can order and interpret the monitoring that goes with it. Peptide Health Lab does not sell anything, does not prescribe, and does not tell anyone where to obtain anything. What the published research adds to that boundary is a reason it matters: the endometrial-protection effect that the approval rests on is preparation-specific, and the same review that establishes it for the oral capsule states it is absent for the transdermal form.^[2]

Anti-doping. A full-text search of the 2026 WADA International Standard Prohibited List, in force 1 January 2026, and of the 2026 Monitoring Program — both retrieved on 1 September 2026 as the WADA-hosted PDF assets linked from USADA's prohibited-list page, HTTP 200 to a browser user agent — returned no occurrence of "progesterone", "progestin", "progestogen", "progestagen" or "micronized" in either document, while same-document controls for "Anastrozole", "Clomifene", "Prasterone" and "Testosterone" each returned hits. Section S0, the non-approved-substances class, reaches only a pharmacological substance "which is not addressed by any of the subsequent sections of the List and with no current approval by any governmental regulatory health authority for human therapeutic use"; progesterone has current US approvals, so S0's own trigger condition does not reach it. Absence from the List is never a statement that a substance is permitted, and the List is reissued annually.

A note on the evidence badge above. This page is graded human_rct because randomized controlled trials in humans exist, in quantity, for the claims this molecule is prescribed and used for — endometrial protection, luteal support, pregnancy, preterm birth, brain injury and vasomotor symptoms. The grade records what kind of evidence has been generated. It does not record that the evidence was favourable, and several of the largest trials on this page missed their primary endpoints. It also does not extend to the sleep, mood and anti-aging claims that bring most readers here: the sleep claim is graded on trials of 8 and 10 participants above, and the molecule-specific breast-cancer comparison is graded anecdotal because its evidence is observational rather than randomized.

Questions to bring to a provider

The first question worth asking is which product and which route, because the answer changes what is approved and what any cited number means. It is fair to ask whether a proposed regimen matches an approved indication or is off-label — including the specific case where estrogen is being given transdermally, since the reference capsule's approval names women receiving conjugated estrogens tablets.^[2]

Where the goal is endometrial safety, the questions the evidence can inform are concrete: what is being given, for how many days of the cycle, and for how long, given that the systematic review's protection conclusion is bounded at five years and excludes the transdermal form entirely.^[2][1] A history of breast cancer, venous thrombosis, arterial thromboembolic disease, liver disease, undiagnosed genital bleeding, or peanut allergy is a reason to raise this with a clinician before considering anything.^[5]

Where the goal is sleep or mood, the useful question is what would count as benefit and how it would be judged, because the randomized evidence is two studies of 8 and 10 people plus a self-reported secondary outcome in a trial that missed its primary endpoint.^[7][8][10] A clinician can put that beside whatever else is on the table, including the possibility that nothing on this page addresses the problem being described.