What do medical sources say about Lithium (Cibalith-S, Eskalith, Lithane, Lithobid, Lithonate) during pregnancy?

1st Trimester: MotherToBaby says it is not known whether lithium increases miscarriage risk. Poels et al. found miscarriage in 16 of 77 lithium-exposed pregnancies versus 40 of 366 unexposed pregnancies in a retrospective cohort of 443; the adjusted odds ratio was 2.94 (95% CI 1.39 to 6.22). Patorno et al. found cardiac malformations in 2.41% of 663 exposed live births versus 1.15% of unexposed births (adjusted risk ratio 1.65, 95% CI 1.02 to 2.68). No exposed infant was coded specifically with Ebstein anomaly, and a six-cohort meta-analysis found no statistically clear increase in major cardiac defects; the current label says to consider fetal echocardiography between 16 and 20 weeks after first-trimester exposure. MotherToBaby says stopping lithium quickly can bring symptoms back, so discuss treatment changes with a healthcare provider. Ask your provider about your situation. 2nd Trimester: MotherToBaby reports studies linking lithium use during pregnancy with preterm birth, while findings on birth weight differ; it says lithium has not been associated with low birth weight. In the Swedish cohort, Hastie et al. associated lithium use during pregnancy with spontaneous preterm birth and large-for-gestational-age birth, and a 2026 Australian cohort reported similar associations. Neither study gives a second-trimester-only estimate. The Swedish study found no statistically clear link with newborn low blood sugar in the whole group, although selected subgroups had higher rates. The current label says to monitor lithium blood levels and adjust the dose if treatment continues during pregnancy. Ask your provider about your situation. 3rd Trimester: MotherToBaby reports that lithium use near delivery, especially with high maternal blood levels, has been followed by temporary symptoms in some newborns, including low muscle tone, sleepiness, breathing or feeding problems, and jaundice. The current label warns that late-pregnancy exposure can cause neonatal lithium toxicity and lists hypotonia, respiratory distress, cyanosis, lethargy, feeding difficulties, apnea, and bradycardia; it says to monitor newborns and provide supportive care until the signs go away, which may take up to 14 days. The label also says to monitor maternal blood levels and adjust the dose if treatment continues, and to decrease or stop lithium two to three days before expected delivery to reduce toxicity risk. MotherToBaby says not every exposed baby has symptoms, and the studies do not give a trimester-specific risk estimate. Ask your provider about your situation.

What do medical sources say about Lithium (Cibalith-S, Eskalith, Lithane, Lithobid, Lithonate) while breastfeeding?

MotherToBaby says lithium passes into milk and discusses watching for restlessness, low muscle tone, or feeding problems, with infant lithium, thyroid, and kidney checks especially for very young or premature infants. LactMed says milk and infant levels vary widely and does not treat lithium as an absolute reason to avoid breastfeeding for selected healthy full-term infants, especially those older than two months, when monitoring is possible; NHS SPS requires extreme caution, specialist supervision, and strict infant monitoring. The current DailyMed label says breastfeeding is not recommended, while the Lithobid label says nursing should not be undertaken except in rare and unusual circumstances. In a 2022 Swedish series of 30 infants, two had unexpectedly high lithium levels in the first month, and poor weight gain was reported. A 2023 study of 25 infant-mother pairs found more symptoms at birth in infants with higher lithium levels, but the difference was not statistically clear and most infants were also exposed before birth, so effects from milk alone could not be separated. Ask your provider about your situation.

Where do sources disagree on Lithium (Cibalith-S, Eskalith, Lithane, Lithobid, Lithonate) in pregnancy or breastfeeding?

First-trimester cardiac malformation findings and Ebstein anomaly Patorno et al. found cardiac malformations in 2.41% of 663 exposed live births versus 1.15% of unexposed births (adjusted risk ratio 1.65). The 2026 Australian cohort also reported an association with cardiac malformations after lithium exposure during pregnancy, but did not provide trimester-specific results. Munk-Olsen et al. found a pooled first-trimester difference in major cardiac malformations that was not statistically significant (2.1% versus 1.6%, adjusted odds ratio 1.54, 95% CI 0.64 to 3.70), and no Ebstein anomaly cases occurred in participating sites. The current label says the increased risk is likely to be small but that data are insufficient to establish a drug-associated risk. The studies used different exposure windows, reference groups, and outcome definitions. Cardiac outcomes are rare, the lithium-exposed groups were small, and the meta-analysis had only 16 first-trimester exposed cardiac cases. The 2026 Australian cohort did not separate results by trimester. The evidence supports a possible increase in cardiac malformations after lithium exposure, but it does not establish causation or provide a verified lithium-specific risk for Ebstein anomaly. Patorno et al. found right ventricular outflow defects, but none were coded specifically as Ebstein anomaly. The label's fetal echocardiography language should remain visible.

Back to search

Lithium

Medication

Cibalith-S, Eskalith, Lithane, Lithobid, Lithonate

Mood stabilizer

Pregnancy

Patorno et al. (2017) analyzed 1,325,563 Medicaid pregnancies ending in live birth, including 663 infants with first-trimester lithium exposure. Cardiac malformations occurred in 16 of 663 exposed infants (2.41%) and 15,251 of 1,322,955 unexposed infants (1.15%); the adjusted risk ratio was 1.65 (95% CI 1.02 to 2.68). Right ventricular outflow tract obstruction defects occurred in 0.60% of exposed infants and 0.18% of unexposed infants (adjusted risk ratio 2.66, 95% CI 1.00 to 7.06). None of the exposed infants had a defect specifically coded as Ebstein anomaly, although most identified outflow defects were consistent with defects that can co-occur with it. This study used filled prescriptions, included live births only, and could not rule out residual confounding. Munk-Olsen et al. (2018) pooled six cohorts with 22,124 live-born singleton pregnancies, including 727 lithium-exposed pregnancies and 21,397 pregnancies in people with mood disorders who were not exposed to lithium. First-trimester exposure was associated with major malformations overall (7.4% versus 4.3%, adjusted odds ratio 1.71, 95% CI 1.07 to 2.72). The pooled difference in major cardiac malformations was not statistically significant (2.1% versus 1.6%, adjusted odds ratio 1.54, 95% CI 0.64 to 3.70); no Ebstein anomaly cases occurred in the study sites. Only 16 cardiac cases were observed in the first-trimester exposed group, and the study included live births only. In the 1992 prospective multicentre cohort by Jacobson et al., major congenital malformation rates did not differ between 148 first-trimester lithium-exposed pregnancies and age-matched controls (2.8% versus 2.4%). One lithium-exposed pregnancy was terminated after prenatal detection of Ebstein anomaly. The abstract does not report a miscarriage comparison. For miscarriage, Diav-Citrin et al. (2014) prospectively followed 183 lithium-exposed pregnancies and reported higher adjusted odds of miscarriage compared with a nonteratogenic-exposure group (adjusted odds ratio 1.94, 95% CI 1.08 to 3.48). Poels et al. (2020) reviewed 443 pregnancies among 241 women with bipolar I disorder. Miscarriages occurred in 16 of 77 lithium-exposed pregnancies (20.8%) and 40 of 366 unexposed pregnancies (10.9%); the adjusted odds ratio was 2.94 (95% CI 1.39 to 6.22). The Poels study relied on retrospective self-reported medication and outcomes, had no lithium dose or serum-level data, and lacked information on smoking, body mass index, alcohol, substance use, and other possible confounders. Hastie et al. (2021) studied 854,017 women, including 434 who used lithium during pregnancy. Lithium use was associated with spontaneous preterm birth (8.7% versus 3.0%, adjusted relative risk 2.64, 95% CI 1.82 to 3.82) and large-for-gestational-age birth (9.0% versus 3.5%, adjusted relative risk 2.64, 95% CI 1.91 to 3.66). In the whole cohort, neonatal hypoglycaemia was not statistically clearly associated with lithium exposure (5.5% versus 2.5%, adjusted relative risk 1.46, 95% CI 0.89 to 2.40), but an association appeared in the subgroup with a diagnosed psychiatric illness. The study used dispensed prescriptions and could not confirm medication use. A 2026 Australian statewide cohort included 867,454 births, of which 234 were exposed to lithium during pregnancy. It reported spontaneous preterm birth in 8.1% versus 2.4% (adjusted relative risk 2.18, 95% CI 1.45 to 3.30), large-for-gestational-age birth in 13.7% versus 6.4% (adjusted relative risk 1.94, 95% CI 1.36 to 2.76), and cardiac malformations in 3.0% versus 0.8% (adjusted relative risk 2.64, 95% CI 1.26 to 5.53). The abstract reports exposure during pregnancy and does not provide trimester-specific estimates.

Breastfeeding

LactMed (revised June 15, 2026) says lithium concentrations in milk and infant serum are highly variable. It describes selected healthy full-term infants as a group for whom lithium is not an absolute contraindication to breastfeeding, particularly during maternal lithium monotherapy and after two months of age, when infant monitoring is available. LactMed also notes that infants with high lithium levels at birth may have more side effects, and that levels acquired before birth can decline more slowly in some breastfed infants. Heinonen et al. (2022) retrospectively reviewed 30 breastfed infants. Infant serum lithium levels varied with age, two infants had unexpectedly high levels during the first month, and poor weight gain was reported; the abstract reported no other adverse effects. In the postpartum chart review by Kummerlowe et al. (2024), 18 mothers used lithium, 7 breastfed, and only 4 infant charts were available; no adverse events were reported in those records. These small, selected samples do not establish risk for all infants. Whaites Heinonen et al. (2023) studied 25 mother-infant pairs with infant lithium levels measured at birth or around two days of age. All but three infants were also exposed to lithium in utero. Symptoms at birth were reported in 85.7% of infants with a high lithium level at birth and 41.2% of those with a lower level (p=0.08); at follow-up the rates were 28.6% and 11.8% (p=0.55). Two infants in the high-level group had therapeutic lithium levels at follow-up. Because the study had no unexposed breastfed comparison group and most infants had late in-utero exposure, it cannot separate effects of lithium in milk from effects of exposure before birth. The authors said that polypharmacy may also contribute and that larger studies are needed.

First-trimester cardiac malformation findings and Ebstein anomaly

Source A says: Patorno et al. found cardiac malformations in 2.41% of 663 exposed live births versus 1.15% of unexposed births (adjusted risk ratio 1.65). The 2026 Australian cohort also reported an association with cardiac malformations after lithium exposure during pregnancy, but did not provide trimester-specific results. (Patorno et al., 2017 cardiac malformation cohort, Australian statewide lithium pregnancy cohort, 2026)
Source B says: Munk-Olsen et al. found a pooled first-trimester difference in major cardiac malformations that was not statistically significant (2.1% versus 1.6%, adjusted odds ratio 1.54, 95% CI 0.64 to 3.70), and no Ebstein anomaly cases occurred in participating sites. The current label says the increased risk is likely to be small but that data are insufficient to establish a drug-associated risk. (Munk-Olsen et al., 2018 international six-cohort meta-analysis, DailyMed: lithium carbonate label)
Why they differ: The studies used different exposure windows, reference groups, and outcome definitions. Cardiac outcomes are rare, the lithium-exposed groups were small, and the meta-analysis had only 16 first-trimester exposed cardiac cases. The 2026 Australian cohort did not separate results by trimester.
What the evidence shows overall: The evidence supports a possible increase in cardiac malformations after lithium exposure, but it does not establish causation or provide a verified lithium-specific risk for Ebstein anomaly. Patorno et al. found right ventricular outflow defects, but none were coded specifically as Ebstein anomaly. The label's fetal echocardiography language should remain visible.

Miscarriage evidence

Source A says: Diav-Citrin et al. and Poels et al. reported higher miscarriage odds in observational studies of lithium-exposed pregnancies. (Diav-Citrin et al., 2014 prospective comparative pregnancy cohort, Poels et al., 2020 lithium and miscarriage cohort)
Source B says: MotherToBaby says it is not known whether lithium increases miscarriage risk and reports that one study did not find an increase. The fact sheet does not identify that study. The accessible Jacobson 1992 abstract does not report miscarriage outcomes. (MotherToBaby: Lithium fact sheet, Jacobson et al., 1992 prospective multicentre pregnancy cohort)
Why they differ: The observational studies were small and used different comparison groups and outcome ascertainment. The Poels study relied on retrospective reports and could not adjust for several potential confounders. The third study described by MotherToBaby could not be confirmed from the sources available for this audit.
What the evidence shows overall: Two observational studies reported an association, but they cannot establish that lithium caused miscarriage. The third study and its result remain not verified. The summary therefore describes the uncertainty without repeating an untraceable study count.

Use of lithium during breastfeeding

Source A says: LactMed says lithium is not an absolute contraindication for selected healthy full-term infants, especially older infants, when monitoring is possible. NHS SPS allows use only with extreme caution, specialist supervision, and strict infant monitoring. A 2026 review describes lithium as a reasonable peripartum option, including during lactation, with additional monitoring. (LactMed: Lithium, NHS Specialist Pharmacy Service: bipolar disorder during breastfeeding, Betcher et al., 2026 review of lithium in pregnancy and lactation)
Source B says: The current DailyMed label says breastfeeding is not recommended during maternal lithium use. The Lithobid label says nursing should not be undertaken except in rare and unusual circumstances where a physician judges that maternal benefit outweighs possible infant or neonatal hazard. (DailyMed: lithium carbonate label, DailyMed: Lithobid extended-release label)
Why they differ: The more permissive clinical sources focus on selected healthy full-term infants with monitoring. Label language reflects a more restrictive approach to potential infant toxicity. Human lactation evidence remains limited to small and selected groups, and several infants were also exposed before birth.
Other sources: MotherToBaby says the product label recommends against breastfeeding, while noting that the benefit of treatment and breastfeeding may outweigh possible risks in some situations. (MotherToBaby: Lithium fact sheet)
What the evidence shows overall: The sources do not give one shared recommendation. The more permissive view applies to selected healthy full-term infants with clinical follow-up, while the product labels advise against breastfeeding or reserve it for rare cases. Keep the guidance rating and show the disagreement clearly.

Newborn low blood sugar findings

Source A says: MotherToBaby reports that some studies found low blood sugar in infants after lithium exposure during pregnancy. (MotherToBaby: Lithium fact sheet)
Source B says: Hastie et al. found no statistically clear association in the whole Swedish cohort (adjusted relative risk 1.46, 95% CI 0.89 to 2.40), although an association appeared in selected psychiatric subgroups. The 2023 study of 25 infant-mother pairs detected no cases of hypoglycemia. (Hastie et al., 2021 Swedish population cohort, Whaites Heinonen et al., 2023 neonatal and breastfeeding cohort)
Why they differ: The studies used different populations, outcome definitions, and observation periods. The 2021 study's finding changed in subgroup analysis, while the 2023 cohort was small and assessed selected infants with late pregnancy and breastfeeding exposure.
What the evidence shows overall: The evidence does not support presenting low blood sugar as a settled or universal newborn effect. The summary distinguishes a whole-cohort result from subgroup findings and describes the uncertainty.

Last reviewed 2026-10-14

Independent check 2026-10-14

Not yet clinician-reviewed

Ask your provider about your situation.

Lithium

Lithium — Breastfeeding

What the research shows

What's in it / what was tested

Sources disagree

Options sources describe

Beyond pregnancy

All sources

Similar medications