Antidepressants and Breastfeeding

Selective Serotonin Reuptake Inhibitors

Selective serotonin reuptake inhibitors (SSRIs) are generally considered first-line drugs in treating postpartum depression, but important differences exist among them in pharmacokinetics and clinical experience during breastfeeding. Sertraline has an RID that is usually ∼0.5% and the RID of paroxetine is 1–1.5%. Most breastfed infants have undetectable serum levels of these two drugs when taken by their mothers. Sertraline has a metabolite with minor activity that is undetectable in infant serum. Their favorable pharmacokinetics make these two drugs the preferred SSRIs in women with no history of antidepressant use. However, paroxetine is generally not preferred during pregnancy because of possible teratogenicity.

At the other end of the spectrum, fluoxetine has pharmacokinetics that are the most unfavorable for breastfeeding among the SSRIs. The average RID is ∼6%, which usually would be considered acceptable; however, a computer simulation indicated that some infants might receive doses of >20% of their mother's dose. Fluoxetine has a half-life of 4–6 days and its main active metabolite, norfluoxetine, has an average half-life of 9.3 days during long-term use. These long half-lives mean that drug accumulation will occur in infants and some infants will have serum levels as high as 59% of the maternal levels that decrease very slowly after discontinuation of the drug or breastfeeding. If the mother was taking fluoxetine during pregnancy, most experts recommend against changing medications during breastfeeding. In fact, if another antidepressant is substituted postpartum, the infant will be exposed to two drugs for days to weeks until norfluoxetine is eliminated by the infant. Adverse effects such as colic, fussiness, and drowsiness have been reported in some breastfed infants, but no adverse effects on development have been found in a few infants followed for up to a year.

Intermediate between fluoxetine and the preferred drugs is citalopram and its S-isomer escitalopram. The RID of citalopram ranges from 4–8% and is ∼3.5–4% for escitalopram. Citalopram and escitalopram have half-lives of 27–35 hours and have active metabolites with somewhat lesser activity. Drowsiness and irritability in breastfed infants have been reported occasionally for both drugs. Some guidelines inexplicably consider citalopram a first-line agent, even after noting the higher amounts in milk and increased levels of side effects. In one guideline, the body of the article lists sertraline and paroxetine as first-line agents, but the abstract of the article lists sertraline and citalopram, somewhat conflating pregnancy and lactation guidance. ¹⁰ Newer articles that cite this article often reference the abstract information rather than the body of the article.

Tricyclic Antidepressants

These older drugs have fallen out of favor because of their side effect profile. Excretion into milk is consistent among the drugs that have been studied, with RIDs of 1–2%. Infant serum levels are usually low to undetectable. Amitriptyline has the most information. Extreme drowsiness with a positive rechallenge unexpectedly occurred in a breastfed infant whose mother was taking only 10 mg/day of amitriptyline, but other infants have tolerated maternal doses of 175 mg/day well. Nortriptyline, which is both a marketed drug and an active amitriptyline metabolite, is a good choice because it has fewer adverse reactions and no active metabolites. Some guidelines list it as an alternative first-line agent.

The one tricyclic with serious adverse reactions reported is doxepin. An 8-week-old breastfed infant was found pale, limp, somnolent, and almost not breathing 4 days after the maternal doxepin dosage had been increased from 10 mg daily to 25 mg three times daily. Another 9-day-old breastfed infant had poor sucking and swallowing, hypotonia, vomiting, and weight loss with a maternal dosage of 35 mg of doxepin at bedtime. Because of these cases, doxepin is generally avoided during breastfeeding.

Serotonin Norepinephrine Reuptake Inhibitors

Venlafaxine is the oldest member of this group of drugs and thus has the most information on use during breastfeeding. It is metabolized to the active metabolite, desvenlafaxine, which often has a greater concentration in milk than venlafaxine itself because its half-life is 11 hours compared with 5 hours for venlafaxine. When both are measured, the combined RID is ∼6.5%. Desvenlafaxine has been detected in infant serum at variable levels, ranging from undetectable to 37% of maternal serum levels. Most infants exposed to venlafaxine and desvenlafaxine in breast milk have no adverse reactions and develop normally, although a few cases of drowsiness and agitation have been reported.

Duloxetine excretion into milk is very low, with an RID <1% in the few studies that have been done. In two infants, serum concentrations were undetectable in one and 0.8% of maternal serum levels in the other. No adverse reactions in breastfed infants have been reported, but experience is limited. Levomilnacipran is in this class, but has no breastfeeding information, although the racemic form, milnacipran, has an RID of <5%.

Bupropion

Bupropion is metabolized to three active metabolites with about half the activity. Data from a few mothers have found RIDs of 1.4–10.6%, mostly as active metabolites. Low levels of drug and metabolites have been found in infant serum, but published experience in newborns is minimal. Bupropion is noteworthy for its dose-related seizure potential. Two breastfed infants developed seizures during maternal bupropion therapy, both at ∼6 months old. One was also exposed to escitalopram in milk. Other antidepressants are preferred.

Triazolopyridines

This class consists of trazodone and nefazodone. Trazodone is not often used as a sole antidepressant, but is sometimes used in low doses for sleep, alone or with other antidepressants. In a few reported cases, milk levels have been low or undetectable and no adverse reactions have occurred. Since doses used for sleep are low, they are unlikely to cause problems in nursing infants.

Nefazodone has been studied in only a few mothers. The excretion of the drug and active metabolite is variable, with RIDs of <1–6.2%. No measurements of infant serum concentrations have been reported. Drowsiness, lethargy, poor feeding, and low body temperature occurred in a breastfed 9-week-old preterm infant whose mother was taking 300 mg/day of nefazodone. Until more data become available, other drugs may be preferred, especially while nursing a newborn or preterm infant.

Monoamine Oxidase Inhibitors

Older monoamine oxidase inhibitors (MAOIs) have always been considered risky to use in nursing mothers, so no information on nursing is available on isocarboxazid, phenelzine, or tranylcypromine. Selegiline is an MAOI initially used in Parkinson's disease that now has an indication as an antidepressant as a transdermal patch. The manufacturer recommends against breastfeeding during treatment with transdermal selegiline and for 5 days after the final patch is removed. However, one woman with severe depression used a 6 mg/day patch during pregnancy and postpartum while exclusively breastfeeding her infant. A blood sample taken from the infant on day 12 postpartum found no selegiline or its metabolite. The infant was developing normally at 5 months of age.

Moclobemide is a reversible selective monoamine oxidase-A inhibitor used for depression and social phobias. It is not available in the United States but is available in Canada and several other countries. Two articles comprising 14 mothers found RIDs of 1–4%. Thirteen infants have reportedly been breastfed during maternal moclobemide therapy with no adverse effects and normal development.

Conclusions

Most antidepressants can be used in nursing mothers. In general, the mother should be treated with a drug that she has responded well to before. If she has not taken an antidepressant previously, sertraline, paroxetine, and possibly nortriptyline are currently considered the drugs of choice because of the low amounts in milk, few adverse effects in infants, and extensive clinical experience. Antidepressants to avoid include doxepin and most MAOIs. Caution should be used with bupropion, fluoxetine, nefazodone, venlafaxine, and some newer drugs with no information on their use in breastfeeding, including levomilnacipran, vilazodone, and vortioxetine.

Mothers with depression or taking an antidepressant at delivery may have delayed lactogenesis II and often require more assistance with establishing nursing. The infants of mothers who were taking an antidepressant during pregnancy may display transient signs of a neonatal discontinuation syndrome.