OBJECTIVE: Postpartum major depression, a frequently (10%) occurring complication of childbirth, adversely affects the mother's functioning, the mother-infant relationship, and the child's subsequent development and propensity for later psychopathology. Although selective serotonin reuptake inhibitors (SSRIs) are effective in treating postpartum depression, concerns have been raised regarding their use in lactating women. Although plasma drug levels of infants who are exposed to SSRIs through breast milk are low compared with those typically seen in patients, infant levels in some reports do seem to be at or near the drugs' reported affinities (K(D)s) and IC(50)s for inhibition at the serotonin (5-hydroxytryptamine [5-HT]) transporter. The impact of central serotonin 5-HT modulation by SSRIs during critical periods of brain development is unknown. These concerns have led our group to examine whether exposure through breast milk has a discernible effect on platelet 5-HT uptake. Taking advantage of the similarities between platelet and neuronal serotonin transporters, we previously used measurements of platelet 5-HT before and during maternal sertraline treatment to determine the degree of 5-HT transporter blockade in breastfed infants. We found that infants who were exposed to sertraline through their mothers' breast milk experienced little to no change in platelet 5-HT levels, suggestive of minimal effects on peripheral and central 5-HT transporter blockade. Compared with sertraline and most other SSRIs, fluoxetine and its active metabolite, norfluoxetine, have substantially longer plasma half-lives, and both compounds have been found in measurable quantities in plasma of nursing infants. Thus, to extend our previous work in this area, we measured platelet 5-HT levels and plasma drug levels in breastfeeding mother-infant pairs before and during maternal treatment with fluoxetine. METHODS: Maternal and infant transporter blockade was assessed by measurement of platelet 5-HT in 11 breastfeeding mother-infant pairs before and after 4 to 12 weeks of maternal fluoxetine (20-40 mg/d) treatment for postpartum depression. The study was approved by the Human Investigation Committee of Yale University School of Medicine, and each mother (mean age: 34.5 years; standard deviation [SD]: 5.3) gave written informed consent. Whole-blood 5-HT levels and plasma fluoxetine and norfluoxetine levels were determined by high-performance liquid chromatography. RESULTS: Five mothers were taking 20 mg of fluoxetine daily, 4 were taking 30 mg daily, and 2 were taking 40 mg daily. Mean infant age at the start of the study was 16.8 (SD: 8.8) weeks. All infants except 1 were <6 months of age and 4 were <3 months of age when their mothers began treatment. Six infants were breastfed exclusively; the remaining were breastfed between 3 and 8 times daily and were given supplemental feedings. Mean maternal postexposure 5-HT levels of 22.9 ng/mL (SD: 12.5) were markedly lower than mean preexposure (baseline) levels of 156.6 ng/mL (SD: 71.4; paired t = 6.9, df = 10). In contrast, the mean infant pre- and postexposure 5-HT concentrations of 217.1 (SD: 66.5) and 229.9 (SD: 83.5) ng/mL, respectively, were similar (paired t = -0.24, df = 10). However, the 1 infant with measurable plasma fluoxetine had a substantial decline in 5-HT to 40% of baseline. In samples obtained from the same infant 4 months later, plasma drug levels were undetectable (<1 ng/mL) and the platelet serotonin levels were no longer reduced (12% increase from baseline). CONCLUSIONS: The marked declines (to 9%-28% of baseline) in platelet 5-HT concentrations seen in mothers who were treated with the SSRI fluoxetine were similar to those observed in our study of sertraline in breastfeeding and other previous studies. In contrast, all but 1 infant experienced little or no decline in whole-blood (platelet) 5-HT concentrations after exposure to fluoxetine through breast milk. The substantial drop in platelet 5-HT seen in 1 infant and the coupling of this drop with measurable plasma fluoxetine leves drop with measurable plasma fluoxetine level raises some concern. Possible reasons for the infant's measurable plasma fluoxetine level include his mother's high plasma drug level and his being breastfed exclusively. However, the observations may be coincidental, and the infant experienced no discernible adverse effects. These data suggest that most infants may continue to breastfeed without experiencing meaningful changes in platelet 5-HT transport while their mothers are treated with 20 to 40 mg of fluoxetine daily. Given the limited data regarding occurrence and extent of SSRI exposure and the uncertainties concerning the possible effects of exposure, it is premature to propose treatment guidelines. Our own advice to women who are thinking of combining breastfeeding and SSRI treatment will weigh a range of factors, including severity of postpartum depression, any demonstrated preferential response to a specific SSRI, and the mother's commitment to breastfeeding. Additional research is needed to establish more definitively the frequency of physiologically meaningful infant SSRI exposure during breastfeeding and to determine the behavioral consequences of such exposure.