Cyclobenzaprine

Pregnancy

Human evidence is limited. Fisher et al. (2023) combined two US population-based case-control studies. The study included 33,615 cases and 13,110 controls; 51 case participants and 9 control participants reported use from one month before conception through the third month of pregnancy. The researchers reported increased odds for seven defects, with 3 to 6 exposed cases per defect: cleft palate (OR 4.79, 95% CI 1.71 to 13.44), cleft lip (OR 2.50, 95% CI 0.89 to 7.02), anorectal atresia or stenosis (OR 6.91, 95% CI 1.67 to 28.65), d-transposition of the great arteries (OR 6.97, 95% CI 2.17 to 22.36), coarctation of the aorta (OR 5.58, 95% CI 1.88 to 16.58), pulmonary valve stenosis (OR 4.55, 95% CI 1.10 to 18.87), and secundum atrial septal defect (OR 3.08, 95% CI 0.83 to 11.45). The researchers said the estimates were unadjusted and imprecise and called the results hypothesis-generating; they noted that chance and confounding may explain some findings. This study does not answer questions about use later in pregnancy. Moreira et al. (2014) described one full-term infant with respiratory distress, high pulmonary artery pressure, right ventricular enlargement, and no open ductus arteriosus after the mother used cyclobenzaprine chronically for back pain late in pregnancy. The authors described early ductal closure as a possible association; the report does not establish cause. The immediate-release label says that adequate, well-controlled studies in pregnant women are unavailable and limits use to clearly needed situations. Its statement is about a lack of controlled data, not a confirmed outcome. The 2026 extended-release label says case reports have not identified a drug-associated risk of major birth defects, miscarriage, or adverse maternal or fetal outcomes. FDA's 2025 Tonmya approval letter requires a descriptive study of pregnancy, maternal, fetal, neonatal, and infant outcomes. No separate cyclobenzaprine estimate of miscarriage risk was identified in the human study or label sources reviewed. The current TONMYA sublingual label is formulation-specific: based on animal data, it advises avoiding TONMYA from two weeks before conception through the first trimester, recommends a pregnancy test before starting, and advises effective contraception during treatment and for two weeks after the last dose. It reports neural tube defects in a rabbit study and decreased pup body weight and survival in rats after dosing during pregnancy and lactation; these are animal findings, not human outcome estimates. The FDA approval letter requires a prospective and retrospective pregnancy-outcomes study and a milk-only study. The letter says spontaneous reports and FDA's active postmarket system were not sufficient to assess the specified maternal, fetal, infant, and milk-exposure risks; this requirement is data collection, not a finding that harm has occurred. A 2026 MarketScan analysis of 1,048,726 pregnant women identified a cyclobenzaprine-associated tubulo-interstitial nephritis signal. It did not report signal-specific event counts or an effect estimate; 18.5 per 1,000 is prescription frequency, not nephritis incidence. The signal is maternal, not a fetal outcome, and does not establish causality.

Breastfeeding

Burra et al. (2019) measured cyclobenzaprine in milk samples from two lactating mothers using liquid chromatography mass spectrometry and reported a calculated relative infant dose of 0.5%. The authors recommended regular clinical assessment of infants because of the medicine's sedating properties; the abstract does not establish long-term effects in infants. LactMed (last revised January 15, 2026) says milk amounts appear very small, breastfeeding need not be stopped if the mother requires cyclobenzaprine, and infants should be monitored for drowsiness, weight gain, and development, especially newborn or preterm infants and when other sedating medicines are used. LactMed describes a 16-day-old infant exposed through milk to cyclobenzaprine, acetaminophen, and oxycodone who was hospitalized with a slow heart rate, low blood pressure, and breathing arrest and survived; dose and extent of breastfeeding were not reported, so cyclobenzaprine's role cannot be separated. Beauchamp et al. (2019) reviewed 76,416 information calls and 2,319 exposure calls to US poison centers from 2001 through 2017; cyclobenzaprine was among the substances involved in calls reporting major effects, but the abstract does not establish cause for the individual infant case. The immediate-release label says human milk excretion is unknown and advises caution. The 2026 extended-release label says there are no data for milk transfer, infant effects, or milk production for that formulation. FDA's 2025 approval letter also requires a milk-only study for Tonmya. The current TONMYA label also says that the small number of published milk-transfer cases cannot be confirmed and that infant-effect and milk-production data are unavailable. The FDA-required milk-only study has no results in the reviewed sources.

What was tested

Human pregnancy evidence includes one population-based case-control analysis of reported use from one month before conception through the third month, plus one late-pregnancy case report. The case-control study reported birth-defect associations but was exploratory, unadjusted, and based on very few exposed cases for each defect. Human milk evidence includes a drug-level study of two mothers and poison-center call analyses; neither provides controlled infant outcome follow-up. The FDA required a pregnancy-outcomes study and a milk-only study for Tonmya, and the reviewed sources do not contain results from those studies. A 2026 MarketScan claims screen of pregnant women also flagged tubulo-interstitial nephritis with cyclobenzaprine, but the paper did not provide event counts or an effect estimate for this signal. The reported 18.5 per 1,000 figure is prescription frequency, not an outcome rate.