Pharmacologic management is often a part of the comprehensive treatment plan for perinatal mood and anxiety disorders (PMADs).  Antidepressants are present in human milk to some extent and data to inform clinical decisions are derived primarily from case reports or case series. Prescribers should engage patients in shared decision making regarding the decision to use pharmacologic treatments and the choice of medications for the treatment of PMADs, including an individualized risk–benefit discussion of the treatments.

For more detailed information and references on specific medications, please refer to LactMed, e-lactancia, Infant Risk, or Mother to Baby.

SSRIs (sertraline/Zoloft, fluoxetine/Prozac, citalopram/Celexa, escitalopram/Lexapro, fluvoxamine/Luvox, paroxetine/Paxil) are used to treat depression and anxiety (as well as other conditions such as obsessive compulsive disorder/OCD, panic disorders, and posttraumatic stress disorder/PTSD).¹ They are often used in combination with psychotherapy, such as cognitive behavioral therapy (CBT). All SSRIs have been detected in human milk. Generally, if the lactating individual has been stable on an SSRI, they should be continued on the medication that has been effective for them. Available data on long-term effects on neurobehavior and development of breastfeeding infants are overall reassuring although infants should be monitored for side effects with certain medications in this class as noted below. While milk production and infant weight gain should be monitored and infants should be monitored for side effects with specific agents as noted below, there is no absolute indication to pump and dump with this class of medications.

Sertraline/Zoloft is commonly used in the management of perinatal mood and anxiety disorders (PMADs) and is widely used due to the larger amount of available data. This medication has low to undetectable levels in milk and is undetectable in infant serum from available data. A weakly active metabolite of sertraline, norsertraline, has been found in very low levels in infants. There are small 2 case series where decreased milk production and delayed lactogenesis II were noted postpartum with the use of this medication.^2,3 While lactating parents should be monitored for low milk production and onset of lactogenesis II with early intervention if needed, there is no absolute indication to pump and dump.

Paroxetine/Paxil is commonly used in lactating individuals. This medication has low to undetectable levels in milk and is undetectable in infant serum. There are case reports of insomnia, restlessness, and increased crying among breastfeeding infants whose lactating parents are on paroxetine so infants should be monitored for these symptoms. There is no absolute indication to pump and dump.

Citalopram/Celexa has a relatively high relative infant dose (RID) or high milk levels compared to other SSRIs so it is not preferred if a new agent is being prescribed for a lactating individual.  If a lactating individual has been stable on this medication and continues on it, then the breastfeeding infant should be monitored closely for side effects including poor sleep, bruxism, irritability, colic, poor feeding, and drowsiness. There is one recent case report of bruxism (teeth grinding) while sleeping in a 9 month old infant after their lactating parent started citalopram which resolved with discontinuation of the citalopram.⁴ While infants should be monitored closely for side effects, there is no absolute indication to pump and dump.

Available data on escitalopram/Lexapro suggests doses up to 20 mg/day produce low levels in milk. This medication has a favorable side effect profile and tight therapeutic range for the lactating individual so may be a good choice for some lactating individuals. There is a case report of a seizure with the use of escitalopram in combination with bupropion ⁵ Infants should be monitored closely for side effects including drowsiness, agitation, restlessness, poor feeding, and poor weight gain especially with doses greater than 20 mg/day, with concurrent use with bupropion, and in younger infants. While infants should be closely monitored for possible side effects, there is no absolute indication to pump and dump.

Fluoxetine/Prozac is not a preferred agent as it tends to have a higher RID and milk level as well as a longer half-life with detectable levels of its metabolite, norfluoxetine, in breastfed infants’ serum (especially in newborns and younger infants). Infants should be closely monitored for colic, fussiness, drowsiness, poor feeding, and poor weight gain. If this agent was effective for the lactating individual during pregnancy/prior to initiation of lactation, a risk benefit discussion should be had with close monitoring of the breastfed infant for side effects if this agent is continued. While alternatives are preferred when possible and infants should be closely monitored for side effects when the lactating individual is on this medication, there is no absolute indication to pump and dump. Shared decision making should be used regarding the use of milk from lactating individuals who are taking fluoxetine.

There is limited data on Fluvoxamine/Luvox during lactation. According to Lactmed, doses up to 300 mg daily produce low levels in breastmilk and would be unlikely to cause side effects in infants, particularly after 2 months of age. A recently published 2 dyad case series found low levels of fluvoxamine in breastmilk which the authors stated was consistent with previously published data.⁶ There is one case report of an infant with elevated blood levels with fluvoxamine.⁷ Another case report noted diarrhea, vomiting, and stimulation in an infant after their breastfeeding mother started on this medication.⁸ While alternatives are preferred due to limited data on this medication during lactation and infants should be closely monitored for side effects including diarrhea, vomiting, decreased sleep, and agitation, there is no absolute indication to pump and dump. Shared decision making should be used regarding the use of milk from lactating individuals who are taking fluvoxamine.

Vortioxetine is a serotonin reuptake inhibitor and receptor modulator. Levels in breastmilk appear to be low and no adverse events have been reported in infants.⁹ While there is no absolute indication to pump and dump, this medication should be used with caution and monitoring of the infant for side effects due to limited data.

SNRIs (duloxetine/Cymbalta, venlafaxine/Effexor, desvenlafaxine/Pristiq) are commonly used in the management of depression as well as other conditions, such as chronic pain and neuropathic pain. The infant should be monitored for sedation and poor weight gain, especially among newborns and premature infants, as a recent case of venlafaxine toxicity in an infant has been reported and a small amount of milk transfer may occur.¹⁰ While alternative medications are preferred and the infant should be closely monitored for side effects with use of this class of medications by the lactating individual, there is not an absolute indication to pump and dump. Shared decision making should be used regarding the use of milk from lactating individuals who are taking SNRIs.

Heterocyclic antidepressants, including TCAs (amitriptyline/Elavil, nortriptyline/Pamelor, doxepin/Silenor), are one of the oldest classes of antidepressant medications. The anti-cholinergic properties of tricyclic antidepressants have been observed to decrease milk production, especially with higher doses although data are limited. There are also reports of infant side effects, including sedation, with certain medications in this class.

In general, breastfed infants should be closely monitored for side effects (such as sedation) with use of this class of medication by the lactating individual and parents should be informed of the risk to milk production and the risk for side effects in their infants. There is no absolute indication to pump and dump, but providers should engage lactating individuals in shared decision making regarding the use of TCAs (with the exception of doxepin, which should be avoided).

There has been one case report of infant sedation with low dose amitriptyline.¹¹ Breastfed infants should be closely monitored for side effects (such as sedation) with use of this medication by the lactating individual and parents should be informed of the risk to milk production and the risk for side effects in their infants. There is no absolute indication to pump and dump, but providers should engage lactating individuals in shared decision making regarding the use of this medication.

This TCA is less anticholinergic than amitriptyline and theoretically carries a lower risk of side effects. Breastfed infants should be monitored for side effects (such as sedation) with use of this medication by the lactating individual and parents should be informed of the risk to milk production and the risk for side effects in their infants. There is no absolute indication to pump and dump, but providers should engage lactating individuals in shared decision making regarding the use of this medication.

This medication should be avoided due to the significant milk transfer with a high risk for side effects, including lethargy, sedation, hypotonia, poor feeding, emesis/vomiting, and respiratory depression, amongst others.

The NDRI, bupropion/Wellbutrin, is found in small amounts in breastmilk for maternal doses up to 300 mg daily. Breastfed infants of lactating parents taking bupropion and an SSRI together should be closely monitored for vomiting, diarrhea, jitteriness, or sedation. There has been one case report of a possible seizure in a 6-month-old partially breastfed infant whose mother was taking bupropion with escitalopram.⁵ Alternatives are preferred for high-risk dyads and parents on multiple psychoactive medications and infants should be monitored closely for side effects. The use of this medication is not an absolute indication to pump and dump and patients and physicians should engage in shared decision making, especially if the patient is on higher doses or other psychoactive medications.

Allopregnanolone is a naturally occurring neuroactive steroid derived from progesterone. There are several new agents designed specifically to treat postpartum depression.

Brexanolone/Zulresso is a synthetic allopregnanolone and it is used in the management of postpartum depression. Brexanolone is given via a continuous 60 hour IV infusion in a certified, monitored healthcare setting. It results in a rapid clinical response with improvement in postpartum depression immediately with treatment (compared to the slower progress seen with oral medications). Allopregnanolone is detectable in human milk of lactating individuals receiving IV brexanolone although levels rapidly decline after completion of the 60 hour infusion and is undetectable by day 3 after the infusion. The maximum RID is 1-1.5% of the maternal weight adjusted dose and there was no accumulation of allopregnanolone in plasma or breastmilk.¹² Although data is limited on the use of this medication in lactation, due its low levels in milk and poor oral bioavailability, brexanolone is unlikely to cause adverse effects in a breastfeeding infant. Due to the risk of parental sedation and sudden loss of consciousness, another caregiver should be present to help care for the infant during the use of this medication. While an additional caregiver should be available to care for the infant, there is no absolute indication to pump and dump with the use of brexanolone.

Zuranolone/Zurzuvae is a synthetic allopregnanolone used in the management of postpartum depression. It is a relatively new medication and was recently approved to be given orally for 14 days to acutely manage postpartum depression. The package insert for zuranolone notes that the relative infant dose (RID) was less than 0.5% in a small study of 14 lactating parents. In this study, zuranolone could no longer be detected in breastmilk by 4-6 days after the final dose. Data is limited given that there is little evidence. Of note, this medication carries a risk of sedation and sudden loss of consciousness for the parent so another caregiver should be available to care for the infant.¹³ While the lactating parent is taking this medication, infants should be closely monitored for any side effects with an additional caregiver available at all times. There is no absolute indication to pump and dump given the extremely low RID in the limited available data in lactating parents and parents should be engaged in shared decision making regarding the use of this medication during lactation.

Commonly used mood stabilizers include lithium and certain antiepileptic drugs as noted below. While infants should be closely monitored for adverse events as noted below, there is no absolute indication to pump and dump and shared decision making should be utilized with the use of these mood stabilizers.

Lithium concentrations in breastmilk and in infant serum are extremely variable. There are many case reports of breastfeeding during lithium monotherapy without signs of toxicity or long term developmental problems (see LactMed linked above for specific references). There are also reports of possible adverse effects in infants when its elimination is impaired (such as with dehydration), or with newborn or premature infant exposure, so special care should be taken with these populations with close monitoring for side effects. Consider monitoring serum lithium, serum creatinine, BUN, and TSH in breastfed infants while the lactating individual is on lithium therapy (see LactMed for possible monitoring intervals and data behind them) especially for infants at risk including newborns, premature infants, or infants at risk of dehydration. While infants should closely be monitored for side effects, possibly with laboratory monitoring, lithium use is not an absolute indication to pump and dump and patients should be engaged in shared decision making with close follow-up for dyads who wish to continue breastfeeding while on lithium therapy.

Anticonvulsants that may be used for mood and anxiety disorders include divalproex sodium/Depakote, valproic acid, carbamazepine/Tegretol, lamotrigine/Lamictal. Due to theoretical risks, these medications should be used with close monitoring of infants for side effects as noted below. Generally, there is no absolute indication to pump and dump, but alternative medications should be considered for some medications in this class due to the concerns noted below. For more information on antiepileptic medications, see the section on Antiseizure Medications.

Little is known about divalproex, but it is quickly metabolized into the active drug valproic acid. Valproic acid levels are low in breast milk. There is a theoretical risk of liver damage and low platelets in the infant with valproic acid use; therefore, infants should be monitored for jaundice and abnormal bruising or bleeding. When valproic acid is used in combination with other sedating psychotropic medications, the infant should also be monitored for sedation. Alternative medications should be considered due to the risks to the infant. If alternatives are not available, providers should engage in shared decision making with patients if this medication must be used and infants should be monitored for jaundice, bruising, or bleeding.

Carbamazepine’s active metabolite is found in relatively high concentrations in breastmilk; however, it is normally below an anticonvulsant therapeutic window. While infants should be monitored for side effects (sedation, poor sucking, withdrawal symptoms, liver abnormalities), there is no absolute indication to pump and dump.

Most infants exposed to lamotrigine via breastmilk have not demonstrated adverse reactions. Rare adverse reactions have been reported including apnea, rash, drowsiness, poor sucking. Breastfed infants should be closely monitored for these side effects with measurement of serum lamotrigine levels and interruption of breastfeeding if there is concern for toxicity. There may also be benefit to monitoring platelet counts, liver function, and lamotrigine levels in infants after an increase in lamotrigine dosing in the lactating parent. Additionally, lamotrigine levels should be monitored in lactating parents after delivery. While alternatives should be considered and infants should be closely monitored for side effects as well as with blood work if clinically indicated, there is no absolute indication to pump and dump.

Buspirone is an anxiolytic with reports of increasing prolactin levels and causing galactorrhea in non-lactating individuals. This effect on prolactin has not been studied in lactating individuals, but it should not negatively impact lactation if prolactin levels increase. Milk levels were undetectable in one parent who was on buspirone 15 mg 3 times daily.¹⁴ Two case reports showed no issues attributable to buspirone in infants.^14,15 Due to the lack of data on this medication during lactation, alternatives are preferred particularly when the nursling is a newborn or preterm infant. There is no absolute indication to pump and dump.

Benzodiazepines (ex: clonazepam/Klonopin, diazepam/Valium) may be used as an adjunct treatment in managing postpartum anxiety acutely. These medications should be used with caution in lactating parents with close monitoring of the breastfed infant for sedation, poor feeding, and/or decreased respiration. Agents with shorter half-lives and lower milk transfer are preferred.

Lorazepam has low milk transfer, a relatively short half-life compared to other benzodiazepines, and is used in the direct treatment of infants. Available data from a few small studies suggest breastfed infants do not have side effects with use of this medication during lactation. However, infants should be closely monitored for side effects such as sedation, poor feeding, and decreased respiration. There is no absolute indication to pump and dump.

Oxazepam has low milk transfer and a relatively short half-life compared to other benzodiazepines. While data on infant side effects of this medication are limited, the risk for side effects seems to be low based on available pharmacologic properties of this medication. Infants should be monitored for side effects given the lack of data while using this medication, but there is no absolute indication to pump and dump.

Clobazam has a longer half-life with a greater risk for infant side effects compared to other benzodiazepines. Infants should be closely monitored for sedation, poor feeding, and delays in developmental milestones (especially in younger infants, exclusively breastfed infants, or breastfed infants of lactating parents on multiple psychotropic medications). Alternatives are preferred due to clobazam’s long half-life. If this medication is given in infrequent low doses, there is no absolute indication to pump and dump. Parents should be engaged in shared decision making regarding the use of breastmilk while taking clobazam.

Clonazepam, like clobazam, has a longer half-life with a greater risk for infant side effects compared to other benzodiazepines. Infants should be closely monitored for sedation, poor feeding, and delays in developmental milestones (especially in younger infants, exclusively breastfed infants, or breastfed infants of lactating parents on multiple psychotropic medications). Alternatives are preferred due to clonazepam’s long half-life. If this medication is given in infrequent low doses, there is no absolute indication to pump and dump. Parents should be engaged in shared decision making regarding the use of breastmilk while taking clonazepam.

Diazepam has significant milk transfer with a longer half-life. Its metabolite, nordiazepam, also has a long half-life. Timing breastfeeding around doses would have limited benefit to reduce the total infant exposure to this medication due to this long half-life. Breastfed infants (especially with newborns or preterm infants) should be closely monitored for hyperbilirubinemia, poor feeding, and sedation. A single higher dose of IV diazepam given perioperatively in the first few days after birth during lactation was associated with significant increase in hyperbilirubinemia and weight loss in a group of Asian neonates, although this effect was not seen with a 5mg oral dose of diazepam.¹⁶ Parents may consider waiting 6-8 hours after a higher one time dose to resume breastfeeding, especially if the lactating parent is on other psychoactive medications. While alternatives are preferred (particularly with newborn and premature infants) and  infants should be closely monitored for side effects, there is no absolute indication to pump and dump. Parents should be engaged in shared decision making regarding the use of breastmilk while taking diazepam.

This section covers antipsychotics commonly used in the management of PMADs. For more information on antipsychotics, see the section on Antipsychotic Medications.

This medication is commonly used as adjuvant therapy for depression in combination with SSRIs or on its own. Currently available data suggests that up to 15 mg daily produces low milk levels, but data is limited.¹⁷ There is strong evidence that aripiprazole decreases prolactin, and many breastfeeding medicine specialists have observed very low milk production with aripiprazole.¹⁸ Alternative medications are preferred due to the limited data on it use during lactation and the significant risk of insufficient milk production. There is no absolute indication to pump and dump and the lactating individual should be engaged in shared decision making regarding the use of this medication. For more information, see the section on Antipsychotics.

Herbal and natural medicines are commonly used by lactating parents. Evidence on these medicines is limited, as they are not regulated like pharmaceutical agents in the United States. These therapies can create medication interactions and possible side effects. The following are commonly used for the management of depression. While there is generally no absolute indication to pump and dump, families should be aware of the lack of data and potential risk for interactions and side effects.

St. John’s Wort is a widely used herb in Europe for mild to moderate depression. While the metabolites of St. John’s Wort are poorly excreted into breastmilk, there is some data suggesting increased frequency of colic, drowsiness, and lethargy in breastfed infants although side effects were mild in one study without need for treatment.¹⁹ This medication can increase the metabolism of oral contraceptives, which can interfere with effective contraception and result in short interval pregnancies (which would result in decreased milk production). While the infant should be closely monitored for adverse effects and the parent should be counseled on the risks of interactions with pharmaceuticals, there is no absolute indication to pump and dump.

There is limited evidence that there is benefit with supplementation with omega-3 fatty acids predominantly composed of eicosapentaenoic acid (EPA) based on a meta-analysis. Per the meta-analysis, omega-3 fatty acids EPA greater than or equal to 60% at dosages of less than or equal to 1 g/day may be beneficial for depression.²⁰ There is no absolute indication to pump and dump.

One small study shows no increase in milk tryptophan levels with the use of this supplement.²¹ There is no absolute indication to pump and dump.

This compound has poor oral bioavailability and is generally well tolerated in adults with the main side effect being nausea. While data is limited, there is no absolute indication to pump and dump especially with infants who are at least 2 months old per LactMed.

B complex vitamins are naturally occurring in human milk. Supplementation with B vitamins will increase the vitamin levels in breastmilk  to some degree but there is limited passage of B vitamins into breastmilk.  Vitamin B levels are not expected to  cause excessive vitamin B levels or breastmilk or cause harmful to the breastfed infant.²²  There is no absolute indication to pump and dump.