What do medical sources say about Lorazepam (Ativan, Loreev) during pregnancy?

1st Trimester: MotherToBaby previously said it was not known whether lorazepam increases miscarriage chance. A 2026 meta-analysis by Picot et al. found higher miscarriage odds for early benzodiazepine use, including a lorazepam estimate from three studies, but these studies were observational and cannot prove the medicine caused the losses. MotherToBaby says a large increase in birth defects is unlikely but describes possible links to anal atresia and pulmonary valve stenosis; a 2026 U.S. cohort found no substantial increase in major birth defects for benzodiazepines overall. Because the recent results and earlier defect signals do not fully agree, the evidence remains uncertain. Ask your provider about your situation. 2nd Trimester: MotherToBaby reports that some studies link lorazepam exposure in the second half of pregnancy with preterm birth or low birth weight, while others do not; it says other factors may explain the findings. A 2026 Taiwan study of benzodiazepines as a group found a higher relative risk of preterm birth and a small, method-sensitive link with babies being smaller than expected for pregnancy week; associations were stronger after second-trimester exposure. The Taiwan results do not isolate lorazepam, and an older cohort could not verify this timing claim for lorazepam specifically. The ACOG-endorsed 2025 joint guideline says clinicians should weigh maternal and fetal risks when considering continued benzodiazepine use or tapering. Ask your provider about your situation. 3rd Trimester: MotherToBaby reports that lorazepam can cause temporary withdrawal-like symptoms in some newborns, including crying, tremors, breathing trouble or weak muscle tone; symptoms usually ease within weeks. The Ativan and Loreev XR labels also warn that late-pregnancy use can cause newborn sedation or withdrawal and call for monitoring after exposure during pregnancy or labor. Whitelaw et al. followed 53 newborns, reporting delayed feeding in 7 of 29 full-term babies after oral lorazepam, mostly with larger doses; results for intravenous treatment and preterm babies were not directly comparable. MotherToBaby says lorazepam-specific behavior or learning studies have not been done, and two recent studies of benzodiazepines or Z-drugs reported conflicting class-level child-development findings without isolating lorazepam. The ACOG-endorsed 2025 guideline says clinicians should weigh maternal and fetal risks when considering continued treatment or tapering. Ask your provider about your situation.

What do medical sources say about Lorazepam (Ativan, Loreev) while breastfeeding?

LactMed reports that lorazepam reaches milk in small amounts and recommends watching for sleepiness, poor feeding and poor weight gain. Kelly et al.'s 2012 study of 124 mothers taking benzodiazepines found sedation in 2 infants, but did not report whether lorazepam caused those cases; LactMed separately reports none of 64 lorazepam-exposed infants were reported sedated. A small study reported sedation in one infant exposed to both lorazepam and risperidone, so lorazepam alone cannot be blamed. A 2026 study measured milk after one 1 mg dose in one mother and estimated low infant exposure, but it does not establish effects from repeated use or in newborns. NHS SPS lists lorazepam as a preferred benzodiazepine for full-term healthy infants when clinically appropriate and calls for monitoring, while the Loreev XR label states, "breastfeeding is not recommended during treatment with LOREEV XR." Ask your provider about your situation.

Where do sources disagree on Lorazepam (Ativan, Loreev) in pregnancy or breastfeeding?

Does early pregnancy lorazepam exposure increase miscarriage risk? Picot et al. (2026) reported a lorazepam-specific pooled odds ratio of 1.48 from 3 observational studies with more than 896 exposed pregnancies. Sheehy et al. (2019) also reported an adjusted lorazepam-specific association. MotherToBaby's fact sheet says it is not known whether lorazepam increases miscarriage chance. PRAC's 2026 minutes describe the DARWIN EU work as a feasibility assessment, not a conclusion about the medicine's risk. The meta-analysis appeared after the MotherToBaby page and included drug-specific analyses. Its estimates are observational, and studies differ in exposure timing, comparison groups and residual confounding. The 2026 meta-analysis makes the older statement of unknown risk incomplete. It found an association for lorazepam, but observational findings do not prove that lorazepam caused the losses. Keep the uncertainty and the newer signal visible.

Back to search

Lorazepam

Medication

Ativan, Loreev

Benzodiazepine

Pregnancy

Miscarriage: Picot et al. (2026) included 10 observational studies with more than 8,035 exposed pregnancies. The pooled odds ratio was 1.68 (95% CI 1.48-1.90) for benzodiazepines overall; the lorazepam estimate was 1.48 (95% CI 1.23-1.79) from 3 studies and more than 896 exposed pregnancies (I-squared 54%). Sheehy et al. (2019) reported a lorazepam-specific adjusted odds ratio of 1.75 (95% CI 1.44-2.14) in a nested case-control study of 442,066 pregnancies. Meng et al. (2024) reported a class-wide association (OR 1.69; 95% CI 1.52-1.87) in a Taiwan case-time-control study of 3,067,122 pregnancies. These observational studies cannot prove cause and may retain confounding by illness or other factors. Kitchin et al. (2022) contains inconsistent wording about whether lorazepam was short- or long-acting, so its individual subgroup estimate needs caution. Birth defects: Bonnot et al. (2003) reported a possible association between lorazepam and anal atresia (OR 6.2; 95% CI 2.4-15.7), based on five lorazepam exposures among six affected infants in a case-control sample; the authors called for confirmation. Tinker et al. (2019) reported a lorazepam and pulmonary valve stenosis signal (crude OR 4.1; 95% CI 1.2-14.2) with few exposed cases. Noh et al. (2022) found a small class-wide increase in major malformations in a South Korean cohort of 3,094,227 pregnancies (adjusted RR 1.09; 95% CI 1.05-1.13), not a lorazepam-only estimate. Fung et al. (2026) studied 4,229,155 U.S. pregnancies and found no substantial overall increase for benzodiazepines (pooled adjusted RR 0.99; 95% CI 0.96-1.03). It reported adjusted absolute risks of 4.0% versus 4.1% in Medicaid and 4.9% versus 4.8% in MarketScan; those named figures come from the primary study and are not lorazepam-specific. Preterm birth and size: Li et al. (2026) analyzed 59,521 benzodiazepine-exposed and 394,956 nonuse pregnancies in an emulated-trial cohort. After accounting for competing pregnancy losses and stillbirths, the class-level relative risk was 1.20 for preterm birth (95% CI 1.18-1.23) and 1.06 for small for gestational age (95% CI 1.00-1.09); effects were stronger after second-trimester exposure. This did not isolate lorazepam. Calderon-Margalit et al. (2009) studied 2,793 pregnancies and reported class-level associations, but did not verify a lorazepam-specific second-half timing effect. Long-term development: MotherToBaby says lorazepam-specific behavior or learning studies have not been done. Oh and Song (2025) found an association between benzodiazepine receptor agonist exposure and a composite of long-term neurodevelopmental diagnoses in a Korean matched cohort (5,949 exposed and 55,015 matched unexposed children; 13.9% versus 11.4%; OR 1.25, 95% CI 1.16-1.35); exposure included benzodiazepines and Z-drugs, not lorazepam alone. Huang et al. (2026) studied 75,477 Taiwan mother-child pairs and found no significant adjusted association with ADHD (HR 1.08; 95% CI 0.99-1.18) or autism (HR 1.08; 95% CI 0.88-1.32) after restricting to mothers with relevant diagnoses. These observational class-level studies conflict and do not answer the lorazepam-specific question.

Breastfeeding

LactMed, revised July 15, 2026, summarizes low milk levels at usual maternal doses and recommends monitoring for infant sleepiness, poor feeding and poor weight gain. In Kelly et al. (2012), 124 women reported infant outcomes while taking benzodiazepines; 2 infants (1.6%) were reported sedated overall. LactMed further reports that 64 mothers took lorazepam and none reported sedation in their infants, a subgroup result not stated in the primary abstract. Sinha et al. (2021) followed 28 infants exposed to atypical antipsychotics; one infant exposed to maternal risperidone plus lorazepam became sedated, and sedation was not seen on the last day after lorazepam was stopped. The combination and small uncontrolled sample prevent assigning the event to lorazepam alone. Shitreh et al. (2026) studied one mother after a single 1 mg dose at 6.5 months postpartum. All six milk samples were below the assay's quantification limit; estimated relative infant dose was 1.35%-1.80% using standard milk-intake assumptions, with an upper-bound estimate up to 3.85%. The authors caution that this single case does not establish effects from repeated doses, higher doses, newborns, premature infants or infants with health problems. Their review of 25 FDA and 1 European adverse-event reports found co-exposure to other central nervous system medicines in all cases and no lorazepam-only reports; spontaneous reports cannot estimate incidence.

Does early pregnancy lorazepam exposure increase miscarriage risk?

Source A says: Picot et al. (2026) reported a lorazepam-specific pooled odds ratio of 1.48 from 3 observational studies with more than 896 exposed pregnancies. Sheehy et al. (2019) also reported an adjusted lorazepam-specific association. (Picot et al., BMC Pregnancy and Childbirth 2026: updated miscarriage meta-analysis, Sheehy et al., JAMA Psychiatry 2019: early pregnancy benzodiazepines and spontaneous abortion)
Source B says: MotherToBaby's fact sheet says it is not known whether lorazepam increases miscarriage chance. PRAC's 2026 minutes describe the DARWIN EU work as a feasibility assessment, not a conclusion about the medicine's risk. (MotherToBaby: Lorazepam fact sheet, EMA PRAC minutes, 9-12 February 2026: benzodiazepine pregnancy study feasibility)
Why they differ: The meta-analysis appeared after the MotherToBaby page and included drug-specific analyses. Its estimates are observational, and studies differ in exposure timing, comparison groups and residual confounding.
Other sources: A 2026 Taiwan cohort also reported an association between benzodiazepine exposure and spontaneous abortion, but it did not isolate lorazepam. (Li et al., JAMA Internal Medicine 2026: benzodiazepines and pregnancy outcomes)
What the evidence shows overall: The 2026 meta-analysis makes the older statement of unknown risk incomplete. It found an association for lorazepam, but observational findings do not prove that lorazepam caused the losses. Keep the uncertainty and the newer signal visible.

Do the data show an increased risk of major birth defects?

Source A says: Small studies reported possible lorazepam-specific signals for anal atresia and pulmonary valve stenosis; Noh et al. (2022) also found a small class-wide increase in overall malformations. (Bonnot et al. 2003: in utero benzodiazepine exposure and anal atresia, Tinker et al., National Birth Defects Prevention Study 2019, Noh et al., PLOS Medicine 2022: first-trimester benzodiazepines and congenital malformations, MotherToBaby: Lorazepam fact sheet)
Source B says: Fung et al. (2026) found no substantial overall increase in major malformations for benzodiazepines across two large U.S. data sources. Ativan and Loreev XR labels say observational studies do not show a clear class-wide association. (Fung et al., Annals of Internal Medicine 2026: early benzodiazepines and congenital malformations, DailyMed: Ativan lorazepam tablet label, DailyMed: Loreev XR lorazepam extended-release capsule label)
Why they differ: The earlier studies examined rare individual defects with few exposed cases, while the 2026 cohort assessed overall malformations across the class and included live-born infants. A null overall estimate does not settle the earlier rare-defect signals, and the signals do not establish a causal effect.
Other sources: MotherToBaby describes a large overall increase as unlikely but still lists the two possible defect signals. (MotherToBaby: Lorazepam fact sheet)
What the evidence shows overall: The latest large cohort does not show a substantial overall class-level increase, while rare lorazepam-specific signals remain unresolved. The summary should state both without implying that either proves a cause.

Do benzodiazepines during pregnancy affect later child development?

Source A says: A 2025 Korean matched cohort found an association between benzodiazepine receptor agonist exposure and a composite of long-term neurodevelopmental diagnoses; the exposure group included benzodiazepines and Z-drugs. (Oh and Song, British Journal of Psychiatry 2025: prenatal benzodiazepine receptor agonists and long-term neurodevelopment)
Source B says: A 2026 Taiwan cohort restricted to mothers with relevant diagnoses found no statistically significant adjusted association with ADHD or autism after accounting for maternal factors. (Huang et al., Archives of Women's Mental Health 2026: prenatal sedative-hypnotics and ADHD or autism)
Why they differ: The studies used different populations, exposure groups and child outcomes. Both were observational and neither reported a lorazepam-specific estimate.
Other sources: MotherToBaby says lorazepam-specific behavior and learning studies have not been done. (MotherToBaby: Lorazepam fact sheet)
What the evidence shows overall: The class-level results are mixed and do not resolve long-term outcomes for lorazepam. State that direct lorazepam-specific evidence is absent rather than treating either class-level study as a lorazepam result.

What guidance applies to lorazepam while breastfeeding?

Source A says: The Loreev XR label states that breastfeeding is not recommended during treatment with Loreev XR because of possible serious infant effects. (DailyMed: Loreev XR lorazepam extended-release capsule label)
Source B says: LactMed reports low milk levels at usual maternal doses and recommends monitoring. NHS SPS lists lorazepam as a preferred benzodiazepine for full-term, healthy infants when clinically appropriate; the Ativan label calls for monitoring rather than stating that breastfeeding is not recommended. (LactMed: Lorazepam, NHS Specialist Pharmacy Service: Using benzodiazepines during breastfeeding, DailyMed: Ativan lorazepam tablet label, Kelly et al., Journal of Pediatrics 2012: benzodiazepine exposure during breastfeeding, Sinha et al., Indian Journal of Psychiatry 2021: infant outcomes with atypical antipsychotics during breastfeeding)
Why they differ: The Loreev XR statement is a brand-specific label warning. NHS SPS guidance concerns clinically appropriate lorazepam use in full-term, healthy infants, and the available infant studies are small and include co-medications.
Other sources: A 2026 single-dose milk-transfer case estimated low exposure after one 1 mg dose but did not assess repeated use or newborns. (Shitreh et al., Frontiers in Pharmacology 2026: lorazepam milk-transfer case and pharmacovigilance review)
What the evidence shows overall: Keep the Loreev XR label warning explicit and also report the more permissive NHS and LactMed guidance with monitoring. The overall rating remains guidance because brand labeling and clinical recommendations differ and the human evidence is limited.

Last reviewed 2026-10-14

Independent check 2026-10-14

Not yet clinician-reviewed

Ask your provider about your situation.

Similar medications

Alprazolam
Oxazepam
Temazepam

Lorazepam

Lorazepam — Breastfeeding

What the research shows

What's in it / what was tested

Sources disagree

Options sources describe

Beyond pregnancy

All sources

Similar medications