Psychiatric treatment during lactation rarely requires an all-or-nothing decision
Postpartum psychiatric illness is common enough that breastfeeding decisions and medication decisions frequently collide. For many women, the fear is not theoretical: they are asked to choose between maintaining lactation and receiving treatment that can stabilize mood, sleep, anxiety, psychosis, or attention symptoms. In practice, that choice is often framed too starkly. Most psychotropic medications do enter breast milk, yet the degree of infant exposure varies widely, and the available clinical experience is far more reassuring than many patients expect.
A central principle guides the field: the risks of untreated maternal illness must be placed on the same scale as the risks of drug exposure through milk. When depression, anxiety, bipolar disorder, or psychosis remains undertreated, maternal functioning, bonding, and safety can deteriorate quickly. Breastfeeding itself remains an important infant health intervention, but it does not automatically outrank maternal psychiatric stability. The safest path is usually the one that preserves both whenever possible.
How lactation safety is judged in clinical practice
Safety assessment depends on more than a simple yes-or-no label. Clinicians rely on measured drug levels in maternal plasma, breast milk, and, when available, infant plasma. Low or undetectable infant blood levels are reassuring even when a medication can be measured in milk. Pharmacokinetic measures add another layer: the milk-to-plasma ratio estimates how much drug enters breast milk, while the relative infant dose estimates the amount a nursing infant receives relative to the mother’s weight-adjusted dose.
A relative infant dose below 10% is generally considered acceptable. That threshold does not guarantee zero risk, but it helps separate drugs with substantial transfer from those with much more limited exposure. Case reports, case series, and registry data then fill in the clinical picture by identifying sedation, feeding problems, irritability, or sleep disturbance when they occur. Professional recommendations from groups such as ACOG, AAFP, and LactMed synthesize those data into practical guidance for bedside use.
For a more formal appraisal, Uguz proposed a scoring system that rates psychotropic medications in lactation using six parameters: total sample size, maximum relative infant dose, sample size reporting relative infant dose, infant plasma levels, frequency of adverse events, and severity of serious adverse events. Higher scores indicate a more robust safety profile.
Antidepressants and anxiolytics are often compatible with breastfeeding
Among antidepressants, the best-studied and most reassuring options are sertraline, paroxetine, and fluoxetine, along with tricyclic antidepressants. These agents have repeatedly shown low concentrations in breast milk, and infant serum levels are often undetectable or very low. Adverse effects are uncommon. Other antidepressants, including venlafaxine, duloxetine, mirtazapine, and trazodone, have less lactation data but may still be reasonable when clinically indicated, provided the infant is monitored.
Case reports have described jitteriness, irritability, feeding disturbance, and sleep disruption in breastfed infants exposed to antidepressants. Those events are uncommon, and causality is often difficult to assign because maternal illness, infant temperament, and other exposures can blur the picture. More recent consensus statements conclude that evidence of serious harm from antidepressant exposure in breast milk is limited, while the consequences of leaving maternal depression untreated are much more tangible.
Benzodiazepines follow a similar pattern. Lorazepam and clonazepam appear in breast milk at low levels, and clinically significant adverse effects are rarely reported. Sedation and feeding difficulty remain the main concerns, especially with longer-acting agents or higher doses. Short-acting benzodiazepines at the lowest effective dose for the shortest feasible duration are preferred when treatment is needed. Routine infant monitoring for sedation and feeding adequacy is sensible rather than alarmist.
Breastfeeding can continue during short-term benzodiazepine treatment in many cases.
Mood stabilizers and antipsychotics require more careful tailoring
Among mood stabilizers, lithium demands the greatest caution because infant serum concentrations can reach roughly one-third to one-half of maternal levels. If lithium is used during breastfeeding, close monitoring of maternal and infant lithium levels, renal function, and thyroid function is required. Lamotrigine produces variable infant serum levels, reported at 20% to 50% of maternal levels, but serious adverse events have not been reported in exposed breastfeeding infants. Carbamazepine and valproic acid warrant caution as well, with infant liver function monitoring recommended because rare hepatotoxicity has been described.
The antipsychotic data are uneven but clinically useful. First-generation agents such as haloperidol and chlorpromazine are generally considered compatible with breastfeeding in moderate doses, with low milk-to-plasma ratios. Among second-generation drugs, olanzapine remains the best studied. Maternal doses up to 20 mg/day generally yield low breast milk levels and negligible infant serum concentrations, with rare and mild adverse effects such as occasional sedation.
Quetiapine also looks favorable: maternal doses up to 400 mg daily produce milk concentrations that are less than 1% of the maternal weight-adjusted dose, and limited follow-up suggests normal infant development. Risperidone has lower milk transfer but more concern because sedation, failure to thrive, jitteriness, and abnormal muscle movements have been reported, although causality remains uncertain. The most clinically striking issue is aripiprazole, which can lower prolactin in a dose-related fashion and has been linked to lactation failure. In one study of 35 women taking aripiprazole, 26 experienced complete lactation failure and four had insufficient milk production. Infant feeding and weight gain deserve close attention when aripiprazole is used.
Some antipsychotics remain poorly characterized. Lurasidone, ziprasidone, and clozapine have extremely limited lactation data, and clozapine remains contraindicated because of agranulocytosis risk.
Stimulants and other ADHD medications depend on symptom severity and milk supply
For women with mild ADHD symptoms, limiting stimulant use during breastfeeding is often reasonable. Severe symptoms may justify treatment, especially when untreated impairment threatens daily functioning or safety. Methylphenidate is present in breast milk at very low concentrations, with relative infant doses generally below 1%, and case reports have not raised major developmental concerns. Amphetamines also appear at low levels in milk, and infant blood concentrations are typically very low or undetectable.
Non-stimulant options remain less certain. Atomoxetine, guanfacine, and clonidine have limited data, and guanfacine and clonidine may reduce milk supply because of their physiologic effects. Monitoring for infant sedation, feeding problems, irritability, and growth issues is prudent, especially early in the postpartum period. When possible, clinicians generally favor the lowest effective dose and immediate-release formulations, since timing can help reduce peak exposure.
Infant monitoring and long-term evidence gaps still shape care
Routine infant drug-level testing is not usually needed for healthy full-term babies. Premature infants and medically vulnerable infants are different; reduced hepatic or renal reserve makes them less able to clear medication exposure from milk. In those situations, caution rises quickly, and collaboration with the infant’s pediatrician becomes essential.
The most persistent gap is not short-term safety but long-term developmental outcome data. Existing evidence is most informative for immediate and short-term effects; it is thinner on later neurodevelopment, behavior, and cognition. Until that changes, individualized prescribing remains the standard. The practical question is rarely whether psychotropic medication and breastfeeding can ever coexist. The real question is which medication, at what dose, for which mother-infant pair, and with what level of monitoring.