What do medical sources say about Doxycycline during pregnancy?

1st Trimester: The current FDA label says first-trimester studies have not shown a difference in major birth-defect risk, but it says doxycycline miscarriage data are unavailable. In 2025, Shitrit et al.'s cohort of 2,696 first-trimester exposures found no higher overall major-malformation risk; an older case-control study found more use among case mothers overall, but its matched analyses found no higher use in months two or three for any defect group. A Norwegian registry study, first online in March 2025 and published in a 2026 issue, reported a miscarriage association after doxycycline dispensing; its authors said infection may explain it. MotherToBaby says tetracycline-class miscarriage risk is unknown and notes that filled prescriptions do not prove the medicine was taken. Ask your provider about your situation. 2nd Trimester: The current FDA label warns that use in the second or third trimester may reversibly slow bone growth, and that use during tooth development in the last half of pregnancy may permanently discolor baby teeth. MotherToBaby says bone growth seems to return after tetracycline is stopped, while tooth staining can last. Shitrit et al.'s 2025 cohort included 134 second-trimester exposures; the authors did not calculate relative risks for late outcomes because the second- and third-trimester groups were small. CDC lists doxycycline in specific RMSF and anthrax regimens for pregnant patients, while UK malaria guidance generally avoids it but says not to withhold it before 15 weeks if other options are unsuitable. Ask your provider about your situation. 3rd Trimester: The current FDA label warns that use in the second or third trimester may reversibly slow bone growth and that use in the last half of pregnancy may permanently discolor baby teeth. MotherToBaby says tetracyclines are generally avoided after month four unless there is a reason to prescribe them. In Shitrit et al.'s 2025 cohort, very-low-birthweight occurred in 6 of 112 third-trimester exposures and 3,397 of 262,990 unexposed pregnancies; the exposed group was small, and the authors did not calculate an adjusted risk for that late outcome. CDC's anthrax review says doxycycline was not associated with fetal growth delays or infant tooth staining in reviewed studies, but this is specific to treatment of anthrax. Ask your provider about your situation.

What do medical sources say about Doxycycline while breastfeeding?

The current FDA label says breastfeeding is not recommended during treatment and for five days after the last dose. It says there are no data on milk levels or infant effects, while LactMed cites small milk-level studies and reports no published infant levels or effects as of May 2024. LactMed and MotherToBaby say short-term use is unlikely to cause infant problems because milk levels and absorption are low; LactMed advises against repeat courses or courses longer than 21 days. UK malaria guidance prefers other options while breastfeeding but allows doxycycline if none are suitable. A 2025 French Lyme guideline lists doxycycline as a first-line option for pregnant or breastfeeding patients with Lyme disease. Ask your provider about your situation.

Where do sources disagree on Doxycycline in pregnancy or breastfeeding?

First-trimester birth-defect findings Czeizel found a higher proportion of maternal doxycycline treatment among case mothers overall, and Muanda reported small doxycycline-specific cardiac and circulatory signals based on 8 or 9 exposed cases. The current DailyMed label reports no difference in major birth-defect risk in first-trimester studies. Shitrit's larger cohort found no increase in overall major malformations or adjusted organ-system groups. The studies used different outcomes and populations. Czeizel's overall treatment comparison was not trimester-specific, Muanda's individual signals had very few exposed cases and included only liveborn singletons, and Shitrit assessed a larger cohort that also included pregnancy terminations and stillbirths. The larger first-trimester cohort and label do not show an overall increase, while rare organ-specific signals remain uncertain. The live entry's two-case neural-tube wording is not supported by the saved abstract.

Back to search

Doxycycline

Medication

Tetracycline antibiotic

Pregnancy

Czeizel and Rockenbauer (1997) reported doxycycline treatment in 56 of 18,515 case mothers (0.30%) versus 63 of 32,804 controls (0.19%; P=.01). However, matched analyses found no significantly higher doxycycline use in gestational months two or three for any abnormality group; the saved abstract does not support the live entry's specific two-case neural-tube claim. Muanda et al. (2017) reported doxycycline-specific associations for circulatory, cardiac, and ventricular or atrial septal defects (adjusted odds ratios 2.38, 2.46, and 3.19, based on 9, 8, and 8 exposed cases). That cohort included liveborn singletons. Shitrit et al. (2025; online 22 Aug 2025, issue Dec 2025) studied 265,686 pregnancies, including 2,696 first-trimester exposures. The adjusted relative risk for overall major malformations was 1.07 (95% CI 0.93-1.23), with no adjusted organ-system association. The study included 134 second-trimester and 112 third-trimester exposures; the authors did not calculate relative risks for late outcomes because these groups were small. Very-low-birthweight occurred in 6 of 112 third-trimester exposures (5.4%) versus 3,397 of 262,990 unexposed pregnancies (1.3%; p=.003), without an adjusted relative risk for that outcome. Boissiere-O'Neill et al.'s Norwegian registry study was first published online 28 Mar 2025 and appeared in a 2026 issue. It included 704,082 pregnancies and 3,233 with doxycycline dispensed; 587 of the exposed pregnancies ended in miscarriage. The overlap-weighted hazard ratio was 1.56 (95% CI 1.44-1.70). The authors' bias analysis found that the estimate approached the null only in scenarios where the underlying infection carried a high miscarriage risk, and they said infection could be the source of the observed association. Dispensing records do not prove the medicine was taken. Muanda et al. (2017) also found an association between tetracyclines as a class and miscarriage (adjusted odds ratio 2.59; 67 exposed cases), but this is not doxycycline-specific. Ding et al. (2026) studied 2,387 mother-child pairs and reported associations between urine-detected doxycycline in the first or second trimester and moderate- or high-risk preschool ADHD symptom trajectories. Children were assessed at ages 3, 5, and 6; the outcomes were symptom trajectories, not ADHD diagnoses, and the abstract did not give a doxycycline-specific effect estimate.

Breastfeeding

LactMed's 15 May 2024 review says short-term doxycycline use is unlikely to harm a nursing infant because milk levels are low and calcium in milk limits absorption. It advises avoiding prolonged courses over 21 days or repeat courses as a theoretical precaution and monitoring for rash, diarrhea, or candidiasis. LactMed reported no published infant levels or effects as of its revision. MotherToBaby describes tetracycline milk levels and infant absorption as low, says problems are not expected with short-term use, and cites the WHO working group's low-risk view for use lasting 7 to 10 days; these statements are class-level. Matsuda's 1984 study measured 30 antibiotics, including tetracyclines, in milk after one dose in 70 postpartum women and reported small overall transfer, but its abstract does not provide a doxycycline-specific estimate. The current FDA label differs from these sources and says breastfeeding is not recommended during treatment or for five days after the final dose.

First-trimester birth-defect findings

Source A says: Czeizel found a higher proportion of maternal doxycycline treatment among case mothers overall, and Muanda reported small doxycycline-specific cardiac and circulatory signals based on 8 or 9 exposed cases. (Czeizel and Rockenbauer, Teratogenic study of doxycycline, Muanda et al., Use of antibiotics during pregnancy and risk of major congenital malformations)
Source B says: The current DailyMed label reports no difference in major birth-defect risk in first-trimester studies. Shitrit's larger cohort found no increase in overall major malformations or adjusted organ-system groups. (DailyMed: Current doxycycline hyclate label, pregnancy section, Shitrit et al., Doxycycline safety during pregnancy: a large population-based cohort)
Why they differ: The studies used different outcomes and populations. Czeizel's overall treatment comparison was not trimester-specific, Muanda's individual signals had very few exposed cases and included only liveborn singletons, and Shitrit assessed a larger cohort that also included pregnancy terminations and stillbirths.
Other sources: MotherToBaby says overall tetracycline information has not shown an increased chance of birth defects. (MotherToBaby: Tetracycline fact sheet, pregnancy)
What the evidence shows overall: The larger first-trimester cohort and label do not show an overall increase, while rare organ-specific signals remain uncertain. The live entry's two-case neural-tube wording is not supported by the saved abstract.

Miscarriage association

Source A says: The Norwegian registry study reported a doxycycline-dispensing association with miscarriage, with an overlap-weighted hazard ratio of 1.56. The Canadian study found a class-level tetracycline association. (Boissiere-O'Neill et al., Exposure to systemic antimicrobials and miscarriage in Norway, Muanda et al., Use of antibiotics during pregnancy and risk of spontaneous abortion)
Source B says: The current label says doxycycline miscarriage data are unavailable. MotherToBaby says it is not known whether tetracyclines increase miscarriage risk and notes that filled prescriptions cannot confirm use. (DailyMed: Current doxycycline hyclate label, pregnancy section, MotherToBaby: Tetracycline fact sheet, pregnancy)
Why they differ: The label's risk summary does not reflect the later-published Norwegian study. The registry study used dispensing records, and its authors said infection may explain the association under high-risk infection scenarios.
What the evidence shows overall: The Norwegian study adds doxycycline-specific observational evidence, but it does not establish that doxycycline caused miscarriage. The label's no-data sentence should be updated.

Later-pregnancy tooth, bone, and newborn outcomes

Source A says: The label warns of possible permanent tooth discoloration and reversible bone-growth inhibition. Shitrit reported a very-low-birthweight signal among 112 third-trimester exposures, without an adjusted relative risk for that late outcome. (DailyMed: Current doxycycline hyclate label, tooth and bone warnings, Shitrit et al., Doxycycline safety during pregnancy: a large population-based cohort)
Source B says: The CDC anthrax guideline says its review found no association between doxycycline and fetal growth delays or infant tooth staining. (CDC: Guidelines for the prevention and treatment of anthrax, 2023)
Why they differ: The CDC statement is part of an anthrax treatment and post-exposure guideline and concerns doxycycline-specific evidence. The label describes class warnings that guide general use, while the recent cohort's third-trimester group was small and the late outcome was not adjusted.
Other sources: UK malaria guidance generally avoids doxycycline in pregnancy but says it should not be withheld before 15 weeks if other options are unsuitable. (UK Health Security Agency: Malaria prevention in special risk groups)
What the evidence shows overall: The general later-pregnancy warning remains relevant. CDC and UK guidance describe narrow, indication-specific exceptions and do not establish routine use as the preferred option.

Doxycycline during breastfeeding

Source A says: The current label says breastfeeding is not recommended during treatment and for five days after the last dose, and says data on milk levels and infant effects are unavailable. (DailyMed: Current doxycycline hyclate label, lactation section)
Source B says: LactMed reports small milk-level studies, no published infant levels or effects as of May 2024, and says short-term use is unlikely to harm an infant. MotherToBaby says short-term tetracycline use is not expected to cause infant problems. (LactMed: Doxycycline, MotherToBaby: Tetracycline fact sheet, breastfeeding)
Why they differ: The label uses a more restrictive precaution in the absence of infant-outcome data. LactMed and MotherToBaby weigh small milk-level studies, low infant absorption, and experience with short courses; LactMed cautions against prolonged or repeated courses.
Other sources: UK malaria guidance prefers other options while breastfeeding but allows doxycycline if no suitable alternative is available. A 2025 French Lyme guideline lists it as first-line in most situations, including for pregnant or breastfeeding patients with Lyme disease. (UK Health Security Agency: Malaria chemoprophylaxis guidance, Raffetin et al., French Lyme borreliosis treatment guideline)
What the evidence shows overall: The public summary should show the current label and the less restrictive short-course guidance side by side. There are no direct infant-level or infant-effect data in the reviewed sources.

Psychiatric adverse-event signal

Source A says: FDA's AEMS page lists psychiatric disorders as a potential signal, and current labeling lists depression, anxiety, suicidal ideation, insomnia, abnormal dreams, and hallucination among postmarketing reactions. (FDA AEMS: January-March 2024 potential safety signals, FDA: Safety labeling changes for doxycycline products, DailyMed: Current doxycycline hyclate label, postmarketing reactions)
Source B says: EMA PRAC concluded that evidence was insufficient to establish a causal relationship with suicidality and that no product-information update was warranted; it will continue monitoring. (EMA PRAC: Doxycycline and suicidality review)
Why they differ: FDA's AEMS listing and postmarketing reports identify a signal for further evaluation, while EMA's review assessed additional evidence and did not find enough to establish causation.
What the evidence shows overall: Report the FDA potential signal and label changes with attribution, and include EMA's conclusion. Do not describe the signal as established causation or as an FDA Drug Safety Communication.

Last reviewed 2026-10-16

Independent check 2026-10-16

Not yet clinician-reviewed

Ask your provider about your situation.

Similar medications

Tetracycline

Doxycycline

Doxycycline — Breastfeeding

What the research shows

What's in it / what was tested

Sources disagree

Options sources describe

Beyond pregnancy

All sources

Similar medications