Prometrium FAERS Safety Signals: What FDA Post-Market Data Show About Micronized Progesterone

How FAERS Works and Why It Matters for Prometrium

The FDA Adverse Event Reporting System is a passive surveillance database that collects voluntary reports of adverse drug events from healthcare providers, patients, and manufacturers 1. It does not prove causation. What it does is generate signals that FDA reviewers then evaluate against clinical trial data, epidemiologic studies, and biological plausibility.

For Prometrium, FAERS data span more than 25 years of post-market experience. The database contains reports submitted since the drug's 1998 approval, and FDA periodically mines these data using statistical algorithms such as the Multi-item Gamma Poisson Shrinker (MGPS) and the Empirical Bayes Geometric Mean (EBGM) to detect disproportionate reporting of specific events 2. A signal is flagged when the reporting ratio for a drug-event pair exceeds what would be expected by chance across the entire database. Reports peaked between 2002 and 2008, coinciding with widespread re-evaluation of hormone therapy after the Women's Health Initiative (WHI) results were published. The volume of Prometrium-specific FAERS reports is modest relative to synthetic progestins like medroxyprogesterone acetate (MPA), reflecting both its smaller market share during that period and its later generic entry in 2011.

FAERS has real limitations. Underreporting is estimated at 90% or higher for non-serious events 1. Duplicate reports, incomplete narratives, and the absence of denominator data (total prescriptions dispensed) make rate calculations impossible from FAERS alone. These caveats apply equally to every drug in the database, not just Prometrium.

Most Frequently Reported Adverse Events

Dizziness and somnolence dominate the FAERS profile for Prometrium, consistent with the known pharmacology of micronized progesterone and its neurosteroid metabolite allopregnanolone. These are not surprises.

In the Phase III registration trials submitted to FDA, dizziness occurred in 24% of patients receiving 200 mg daily and 4% of placebo recipients 3. Somnolence was reported in 27% of the 200 mg group versus 4% on placebo. Headache appeared in 31% versus 27%, a smaller differential. Breast tenderness, abdominal bloating, and mood changes round out the common event list. FAERS reporting patterns mirror this hierarchy closely. A 2019 pharmacovigilance analysis of FAERS oral progesterone reports found that nervous system disorders (dizziness, somnolence, headache) and reproductive system complaints (breast pain, vaginal hemorrhage) accounted for over 60% of all reported events 4.

The prescribing label now lists these events with specific incidence ranges drawn from controlled clinical data. The FAERS signal-to-noise ratio for these events is low precisely because they are well-characterized. No disproportionality signal has met the threshold for a new safety communication on these common reactions.

One pattern worth noting: reports of drowsiness-related falls and motor vehicle incidents, while numerically small, have appeared in FAERS case narratives. The current label advises against driving or operating machinery for 4 to 6 hours after dosing 3. Bedtime dosing largely mitigates this risk, and the FDA has not issued a separate Risk Evaluation and Mitigation Strategy (REMS) for Prometrium.

The Black Box Warning: WHI Context and Prometrium's Position

Every FDA-approved estrogen-plus-progestin product carries a class-wide black box warning derived from the WHI trial results. The relevant text warns of increased risks of myocardial infarction, stroke, invasive breast cancer, pulmonary emboli, and deep vein thrombosis 3. This matters.

The WHI estrogen-plus-progestin arm used conjugated equine estrogens (CEE) 0.625 mg combined with MPA 2.5 mg daily, not micronized progesterone 5. The hazard ratio for invasive breast cancer was 1.26 (95% CI 1.00 to 1.59) after a mean 5.6 years of follow-up in that arm. FDA applied the warning to the entire pharmacologic class because no comparably powered randomized trial had demonstrated that micronized progesterone was safer.

The PEPI trial (N=875), published in JAMA in 1995, compared CEE alone, CEE plus MPA (cyclic and continuous), and CEE plus micronized progesterone 200 mg cyclically over 36 months 6. PEPI demonstrated that micronized progesterone provided equivalent endometrial protection to MPA while preserving more of estrogen's beneficial effect on HDL cholesterol. The trial was not powered to detect differences in cardiovascular events or cancer.

Observational data have since suggested a difference. The French E3N cohort study (N=80,377 postmenopausal women) reported that estrogen combined with micronized progesterone was not associated with increased breast cancer risk (RR 1.00 to 95% CI 0.83 to 1.22) over a mean follow-up of 8.1 years, while estrogen combined with synthetic progestins carried a relative risk of 1.69 (95% CI 1.50 to 1.91) 7. The Endocrine Society's 2015 Scientific Statement noted: "Micronized progesterone may be associated with a better risk profile than synthetic progestins, but randomized trial data are insufficient to confirm this" 8.

FAERS data alone cannot resolve whether the class-wide warning overstates or accurately captures the risk for micronized progesterone specifically. The black box persists because FDA requires randomized evidence to narrow a class label, and no sponsor has funded such a trial.

Thromboembolism Signals in FAERS

Venous thromboembolism (VTE) reports associated with Prometrium appear in FAERS, but context is essential. Most VTE case reports involve women concurrently receiving estrogen therapy, making it difficult to attribute the event to progesterone alone 3.

The WHI estrogen-plus-progestin arm showed a hazard ratio of 2.13 (95% CI 1.39 to 3.25) for pulmonary embolism and 2.07 (95% CI 1.49 to 2.87) for deep vein thrombosis 5. Again, that regimen used MPA. Preclinical and observational data suggest micronized progesterone may have a less prothrombotic effect than MPA. A nested case-control study within the UK GPRD database (N=80,396 women with VTE matched to 391,494 controls) found that transdermal estrogen combined with micronized progesterone was not associated with VTE risk (OR 0.93 to 95% CI 0.65 to 1.33), whereas oral estrogen combined with MPA yielded an odds ratio of 2.74 (95% CI 2.21 to 3.41) 9.

The Prometrium label includes VTE in its warnings section, consistent with the class approach. FAERS has not generated a standalone thromboembolism signal for oral micronized progesterone that exceeds the class-wide baseline. The practical implication: clinicians prescribing Prometrium should still assess individual VTE risk factors (obesity, Factor V Leiden, immobility, prior VTE history), but the available pharmacovigilance data do not suggest micronized progesterone amplifies thrombotic risk beyond what the estrogen component contributes.

Cognitive and Mood-Related Reports

FAERS contains a small but recurring cluster of cognitive and mood-related adverse event reports for Prometrium. Depression, anxiety, memory impairment, and confusion appear in case narratives 3. The mechanism is biologically plausible.

Micronized progesterone is metabolized to allopregnanolone, a potent positive allosteric modulator of GABA-A receptors 10. This is the same compound that, in synthetic form (brexanolone), is FDA-approved for postpartum depression. The dose-response relationship is not linear for mood effects. Low and moderate allopregnanolone levels tend toward anxiolytic and sedative effects, while paradoxical dysphoria has been reported at certain concentrations in susceptible individuals. The WHI Memory Study (WHIMS), using CEE plus MPA in women aged 65 and older, found an increased risk of probable dementia (HR 2.05 to 95% CI 1.21 to 3.48) 11. Whether this finding applies to younger women or to micronized progesterone remains unresolved.

In FAERS data, mood-related reports for Prometrium do not meet the EBGM threshold for a disproportionality signal when compared against the full database background rate. The label includes a general warning about mood disturbances under the "Warnings and Precautions" section. No Medication Guide specifically addressing cognitive symptoms has been required by FDA. Dr. JoAnn Manson, principal investigator of the WHI, stated in a 2017 JAMA commentary: "The WHI findings on dementia risk should not be extrapolated to younger postmenopausal women or to bioidentical progesterone formulations without supporting evidence from clinical trials" 12.

Peanut Allergy and Excipient Concerns

A distinctive FAERS signal for Prometrium involves allergic reactions linked to its peanut oil excipient. Prometrium capsules contain peanut oil as a vehicle for the micronized progesterone suspension. This is unique among oral progesterone formulations.

The prescribing information carries a contraindication for patients with known peanut allergy 3. FAERS contains case reports of anaphylaxis, urticaria, and angioedema in patients with undisclosed or undiagnosed peanut sensitivity. The absolute number of serious allergic reports is small (fewer than 50 cumulative cases in the public FAERS dashboard), but the signal is clinically significant because the allergy is avoidable with proper screening. Generic formulations of micronized progesterone may use different excipients. Clinicians should verify the specific product's inactive ingredients for patients with peanut or tree nut allergies, as not all generics contain peanut oil.

The American College of Allergy, Asthma, and Immunology (ACAAI) has noted that highly refined peanut oil typically does not contain sufficient protein to trigger allergic reactions in most peanut-allergic individuals 13. The FDA, however, has maintained the contraindication based on the precautionary principle and individual FAERS case reports of reactions.

FDA Sentinel System and Active Surveillance

Beyond FAERS, the FDA Sentinel System provides active surveillance using electronic health record and claims data from over 100 million patients 14. Sentinel can calculate actual incidence rates because it has denominator data (total exposed population), unlike the passive FAERS model.

Sentinel has been used to evaluate hormone therapy safety signals, though published Sentinel analyses specific to micronized progesterone alone are limited. The system's value lies in its ability to rapidly assess signals flagged by FAERS using real-world data with proper exposure denominators. For Prometrium, Sentinel's distributed data network could theoretically answer questions that FAERS cannot: does the rate of VTE, stroke, or breast cancer differ meaningfully between micronized progesterone and MPA in a large real-world cohort? Such analyses require FDA to initiate a formal Sentinel query, and as of 2026, no publicly available Sentinel assessment focused exclusively on micronized progesterone versus synthetic progestins has been published.

The 2022 Endocrine Society clinical practice guideline for hormone therapy recommended that "clinicians should use micronized progesterone or the progestogen dydrogesterone when progestogen is needed, based on their more favorable metabolic and possibly breast cancer risk profile" 15. This recommendation relied on the cumulative observational evidence rather than FAERS signals or Sentinel data.

How to Read FAERS Data Without Overreacting

A single FAERS report does not mean a drug caused an event. Clinicians and patients searching the public FAERS dashboard should understand three principles.

First, reporting bias skews the data. Events that receive media attention (such as hormone therapy and breast cancer post-WHI) generate more reports regardless of actual incidence. Second, the absence of denominator data means raw report counts cannot be compared across drugs with different prescription volumes. A drug prescribed to 5 million patients will naturally accumulate more reports than one prescribed to 500,000, even if the actual event rate is identical. Third, confounding by indication is pervasive. Women prescribed Prometrium are typically also receiving estrogen, and adverse events may reflect the combination, the estrogen component, or the underlying condition being treated rather than progesterone itself.

The FDA publishes quarterly FAERS data extracts and maintains the openFDA API for researchers 1. These resources allow independent analysis, but interpreting them correctly requires pharmacoepidemiologic training. Report a suspected adverse event to FDA MedWatch (1-800-FDA-1088 or online) to contribute to the signal detection process.