Progesterone

Pregnancy

MotherToBaby reports that progesterone or progestin use does not raise miscarriage chances and that most studies found no increase in birth defects; it notes older hypospadias reports and study-design limits. These are broad summaries and are not separated by route or trimester. In PRISM, 4,153 women with early pregnancy bleeding received vaginal micronized progesterone 400 mg twice daily or placebo through 16 weeks; live birth was 75% versus 72%, a difference that was not statistically significant (P=0.08). In PROMISE, 836 women with unexplained recurrent miscarriage received vaginal micronized progesterone 400 mg twice daily or placebo through 12 weeks; live birth was 65.8% versus 63.3% (relative rate 1.04, 95% CI 0.94 to 1.15), with no significant difference. The 2025 Cochrane review included 9 trials and 1,426 randomized participants; 8 trials with 1,276 participants contributed to analysis, which found little to no difference in miscarriage rates (RR 0.91, 95% CI 0.76 to 1.07). The review excluded threatened miscarriage and IVF trials and combined different progestogens, routes, doses, and durations. NICE limits its recommendation to vaginal micronized progesterone 400 mg twice daily for bleeding with a scan-confirmed pregnancy in the uterus and a prior miscarriage; if a heartbeat is confirmed, NICE says to continue through 16 completed weeks. NICE found no evidence of benefit for other preparations or doses. ACOG says vaginal progesterone may be considered for a prior preterm birth with a singleton pregnancy and shortened cervix, but not offered without shortening; SMFM calls for shared decision-making if cervical length is unavailable or at least 25 mm. FDA withdrew approval of 17-OHPC in 2023 after a confirmatory study did not verify benefit. EMA recommended suspension of 17-OHPC in 2024 because it found no prevention benefit and a possible in-utero cancer risk that could not be confirmed; EMA states that action does not affect progesterone itself. The Endometrin label, for ART support through 10 weeks, reports 7 neonatal defects among 203 live births (3%) in the twice-daily group and 7 among 223 live births (3%) in the three-times-daily group. The primary 1,211-person active-comparator trial abstract describes the groups but does not report those defect counts. A 2025 prospective cohort of 17,713 women with early bleeding found no association between treatment by injection or orally and malformations compared with no treatment (HR 1.05, 95% CI 0.65 to 1.72); the study also considered folic-acid use. A 2025 retrospective case-control study of 610 cases and 610 controls associated prenatal exogenous progesterone with higher odds of undescended testicles (aOR 1.85, 95% CI 1.42 to 2.40). A separate 2025 retrospective study of 256 children found lower language and personal-social scores in the exposure group. Another 2025 case-control study found no overall ADHD association (97 cases and 194 controls), though continuous use across all trimesters was reported more often among cases. These observational findings cannot establish cause and differ in route, population, and outcome. A 2026 retrospective analysis of the 1959 to 1976 US Collaborative Perinatal Project included 49,507 mothers, 952 of whom used progesterone; use was associated with a 135 g lower birth weight (95% CI 106 to 163 g lower). Because the data are historical and observational, they do not establish cause or describe current products. A 2025 randomized-trial follow-up assessed 206 children of 104 mothers, 35.4% of the original trial participants, and found no significant behavior or cognitive differences at ages 6 to 9 after prenatal vaginal progesterone. A 2026 cohort of 890 singleton pregnancies with no prior preterm birth and a cervix shorter than 25 mm diagnosed after 24 weeks found no significant reduction in preterm birth with vaginal progesterone. A 2026 neonatal meta-analysis included 12 randomized studies and 1,557 participants, and its search covered both vaginal progesterone and 17-OHPC, so its pooled results do not identify the effect of natural progesterone alone.

Breastfeeding

LactMed reports that progesterone levels in milk are low and that the amount an infant ingests would not be expected to cause adverse effects. Its human evidence includes implants and vaginal rings, including studies that found no differences in infant growth and development. MotherToBaby also reports low milk amounts and does not expect harm to a nursing infant. Roy et al. (2020) compared a progesterone vaginal ring with a copper IUD in a nonrandomized study of 789 lactating women at 20 Indian centers; infants in both groups grew at expected rates. A coauthor was an employee of the Population Council, which developed the vaginal ring. The Prometrium capsule label says detectable progestin has been identified in milk and advises caution. The Endometrin label says detectable progesterone has been identified in milk and that the effect on the nursing infant has not been determined; it also cites a published study that reported no effect on milk production or infant growth during the first postpartum year. The reviewed human evidence does not directly establish outcomes for oral Prometrium capsules.

What was tested

The PRISM and PROMISE trials tested vaginal micronized progesterone 400 mg twice daily in different early-pregnancy populations and for different time periods. The Endometrin label concerns a 100 mg vaginal insert used two or three times daily in an assisted-reproduction program through 10 weeks. Prometrium is an oral capsule whose label says it should not be used during pregnancy. Lactation studies mainly examined progesterone implants or vaginal rings; Roy et al. studied a vaginal ring, not oral capsules. FDA and EMA actions discussed here concern synthetic 17-OHPC, not natural progesterone.