Fetal Risk Summary
Fluoxetine, a selective serotonin reuptake inhibitor (SSRI), is used for the treatment of depression. The chemical structure of fluoxetine is unrelated to other antidepressant agents.
All of the antidepressant agents in the SSRI class (citalopram, fluoxetine, fluvoxamine, paroxetine, and sertraline) share a similar mechanism of action although they have different chemical structures. These differences could be construed as evidence against any conclusion that they share similar effects on the embryo, fetus, or newborn. In the mouse embryo, however, craniofacial morphogenesis appears to be regulated, at least in part, by serotonin. Interference with serotonin regulation by chemically different inhibitors produces similar craniofacial defects (1). Regardless of the structural differences, therefore, some of the potential adverse effects on pregnancy outcome may also be similar.
Reproduction studies in rats and rabbits revealed no evidence of teratogenicity using up to 1.5 and 3.6 times the maximum recommend human daily dose on a body surface area basis [MRHD], respectively, throughout organogenesis (2,3). In rats, however, doses of 1.5 times the MRHD during gestation or 0.9 times the MRHD during gestation and lactation were associated with an increase in stillbirths, a decrease in pup weight, and a decrease in pup survival during the first 7 days postpartum (2). The no-effect dose for pup mortality was 0.6 times the MRHD (2). There was no evidence of developmental neurotoxicity in the surviving pups exposed to 1.5 times the MRHD during gestation (2).
Using uterine rings from midterm (gestation day 14) and term pregnant rats, fluoxetine, and two other antidepressants (imipramine and nortriptyline), were shown to attenuate the activity of serotonin-induced spontaneous uterine contractions (4). Although a direct myometrial role could not be demonstrated for these monoamine reuptake inhibitors, the investigators discussed several other possible pathways that fluoxetine could induce preterm delivery (4).
Administration of fluoxetine to pregnant rats produced a down-regulation of fetal cortical 3H-imipramine binding sites that was still evident 90 days after birth (5). The clinical significance of this finding to the development of the human fetal brain is unknown.
In a study to determine if fluoxetine increased the bleeding risk in neonates, pregnant rats were administered fluoxetine (5.62 mg/kg/day) from day 7 of gestation until the delivery (6). The dose was approximately 5 times the maximum recommended human dose. Compared to controls, fluoxetine-exposed pups had a significantly higher frequency of skin hematomas. The mechanism was thought to be related to the inhibition of serotonin uptake by platelets (6).
Both fluoxetine and the active metabolite, norfluoxetine, cross the placenta and distribute within the embryo or fetus in rats (7). No reports describing the placental transfer of fluoxetine early in gestation have been located. The molecular weight (about 346 for the hydrochloride salt) is low enough, however, that passage should be expected. Moreover, two studies (cited below as References #17 and #18), have documented the human placental transfer of the antidepressant and its active metabolite at term.
During clinical trials with fluoxetine, a total of 17 pregnancies occurred during treatment, even though the women were required to use birth control, suggesting lack of compliance (3). No pregnancy complications or adverse fetal outcomes were observed.
A prospective evaluation of 128 women treated with a mean daily dose of 25.8 mg of fluoxetine during the 1st trimester was reported in 1993 (8). Two matched control groups were selected, one with exposure to tricyclic antidepressants (TCAs) and the other with exposure only to nonteratogens. No differences were found in the rates of major birth defects (2, 0, and 2, respectively) among the groups. An increased risk was observed, although not statistically significant, in the rate of spontaneous abortion when the fluoxetine group was compared to those in the nonteratogen group, 14.8% vs. 7.8% (relative risk 1.9; 95% confidence interval [CI] 0.923.92). Because only 74 TCA 1st trimester exposures were available for matching, comparisons between the three groups were based on 74 women in each group. The rates of miscarriage from this analysis were 13.5% (fluoxetine), 12.2% (TCAs), and 6.8% (nonteratogens), again without reaching statistical significance. Because of the increase in the number of spontaneous abortions observed in both antidepressant groups, additional studies are needed to separate the effects of the psychiatric condition from that of the drug therapy (8). The authors also concluded that exposure to fluoxetine during the 1st trimester was not associated with an increased risk of congenital defects, but that long-term studies were warranted to evaluate the potential neurodevelopmental toxicity of the antidepressant (8).
A 1992 prospective multicenter study evaluated the effects of lithium exposure during the 1st trimester in 148 women (9). One of the pregnancies was terminated at 16 weeks’ gestation because of a fetus with the rare congenital heart defect, Ebstein’s anomaly. The fetus had been exposed to lithium, fluoxetine, trazodone, and L-thyroxine during the 1st trimester. The defect was probably caused by lithium exposure.
In a surveillance study of Michigan Medicaid recipients involving 229,101 completed pregnancies conducted between 1985 and 1992, 142 newborns had been exposed to fluoxetine, 109 during the 1st trimester (F. Rosa, personal communication, FDA, 1994). Two (1.8%) major birth defects were observed (five expected), but details of the abnormalities were not available. No anomalies were observed in eight defect categories (cardiovascular defects, oral clefts, spina bifida, polydactyly, limb reduction defects, hypospadias, brain defects, and eye defects) for which specific data were available. These data do not support an association between the drug and congenital defects.
A 1993 letter to the editor from representatives of the manufacturer summarized the postmarketing database for the antidepressant (10). Of the 1103 prospectively reported exposed pregnancies, 761 of which had potentially reached term, data were available for 544 (71%) outcomes, including 91 elective terminations. Among the remaining 453 pregnancies, there were 72 (15.9%) spontaneous abortions, 2 (0.4%) stillbirths, and 20 (4.4%) infants with major malformations, 7 of which were identified in the postperinatal period. Details of the aborted fetuses and stillbirths were not given. The malformations observed in the perinatal period were abdominal wall defect (in one twin), atrial septal defect, constricted band syndrome, hepatoblastoma, bilateral hydroceles, gastrointestinal anomaly, intestinal blockage, macrostomia, stubbed and missing digits, trisomy 18, trisomy 21, and ureteral disorder (2 cases). The postperinatal cases included an arrhythmia, pyloric stenosis (2 cases), tracheal malacia (3 cases), and volvulus. An additional 28 cases of major malformations reported retrospectively to the manufacturer were mentioned, but no details were given other than the fact that the malformations lacked similarity and, thus, were not indicative of a pattern of anomalies (10).
A review that appeared in 1996 (before Reference #12) examined the published data relating to the safety of fluoxetine use during gestation and lactation in both experimental animals and humans (11). Using previously published criteria for identifying human teratogens, the authors concluded that the use of fluoxetine during pregnancy did not result in an increased frequency of birth defects or effects on neurobehavior (11).
A prospective study published in 1996 compared the pregnancy outcomes of 228 women who took fluoxetine with 254 nonexposed controls (12). The rates of spontaneous abortion in the two groups were 10% (exposed) and 8.5% (controls), but 13.6% (23 of 169) among those who were enrolled in the study during the 1st trimester and who had 1st trimester exposure. Major structural anomalies were observed in 5.5% (9 of 164) of live-born infants exposed to fluoxetine during the 1st trimester compared to 4.0% (9 of 226) of live-born infants among the controls (p =0.63). No patterns were evident in either group (12). A total of 250 infants (97 study, 153 controls) were examined (by a physician who was unaware of the infant’s drug exposure ) for minor anomalies and among those with three or more, 15 (15.5%) were exposed and 10 (6.5%) were not exposed (p =0.03). In comparison to those infants who were exposed during the 1st trimester to fluoxetine or not exposed at all, infants who were exposed late to the drug had a significant increase in perinatal complications, including prematurity (after excluding twins), rate of admission to special-care nurseries (after excluding preterm infants), poor neonatal adaptation, lower mean birth weight and shorter length in full-term infants, and a higher proportion of full-term infants with birth weights at or below the 10th percentile (12). Moreover, two (2.7%) of the full-term infants who were exposed late had persistent pulmonary hypertension, a complication that is estimated to occur in the general population at a rate of 0.07%0.10% (12). Although the authors concluded that the number of major structural anomalies and the rate of spontaneous abortions were not significantly increased by fluoxetine exposure in this study, the increased rate of three or more minor anomalies, an unusual finding, is indicative that the drug does affect embryonic development and raises the concern of occult malformations, such as those involving brain development (12). Moreover, the use of fluoxetine late in pregnancy was related to an increase in perinatal complications.
In an accompanying editorial (14) and subsequent letters (15,16), various investigators cited perceived problems with the above study and were addressed in a reply (13).
A 1993 case report described possible fluoxetine-induced toxicity in a term 3580-g male newborn (17). The infant’s 17-year-old mother had taken the antidepressant (20 mg/day) throughout most of her pregnancy for severe depression and suicidal ideation (17). The infant was initially alert and active with mild hypoglycemia (33 mg/dL). Oral 5% dextrose was given and hourly blood glucose samples over the next 4 hours were within normal limits. At 4 hours of age, marked acrocyanosis was noted and the infant became jittery. Tachypnea developed with a respiratory rate of 70. His condition continued to worsen with symptoms peaking at 36 hours. The symptoms included continuous crying, irritability, moderate to marked tremors, increased muscle tone, a hyperactive Moro reflex, and emesis (17). An extensive diagnostic work-up, including a drug screen, was negative. The cord blood fluoxetine and norfluoxetine levels were 26 ng/mL and 54 ng/mL, respectively, both within a nontoxic range for adults (17). The infant was asymptomatic at 96 hours of age at which time the serum levels of the parent drug and metabolite were
A case report in 1997 described a 3020-g male newborn who was delivered at term from a 34-year-old woman with obsessive-compulsive disorder treated with fluoxetine 60 mg/day (18). Apgar scores were 7 and 8 at 1 and 5 minutes, respectively. After birth, the infant was jittery and hypertonic with mild grunting, flaring, and retracting (18). Scattered petechiae on the face and trunk and a cephalohematoma were noted. A right, nondisplaced clavicular fracture was noted on chest x-ray. The remainder of the examination and diagnostic work-up was nonrevealing. On the second day of life, the serum fluoxetine and norfluoxetine concentrations were 129 ng/mL and 227 ng/mL, respectively, both in the normal adult range (18). The mother did not breast feed the infant. Marked improvement in the jitteriness was observed by 2 weeks of age and by 5 months of age, the infant was considered normal. Because of similar symptoms observed in newborns (see References #11 and #13 above) and in adults, the authors attributed the baby’s symptoms to fluoxetine. Moreover, the bruising and bleeding were also thought to have been caused by the antidepressant (18).
Using data from the manufacturer’s prospective pregnancy registry, postnatal complications that had been reported in 112 pregnancies (115 infants) exposed to fluoxetine during the 3rd trimester were tabulated in a 1995 Reference (19). Maternal doses ranged from 10 to 80 mg/day, but were not reported in 20 pregnancies. Postnatal complications, unrelated to congenital malformations, were noted in 15 singleton term infants, including two infants with jitteriness and three with irritability. Irritability was also observed in one premature infant (gestational age not given). Except in one infant, the complications were considered mild and transitory. No relationship to the maternal dose was observed, but plasma drug levels were not measured in any of the infants (19).
A study published in 1997 described the outcomes among 796 pregnancies with confirmed 1st trimester exposure to fluoxetine that had been reported prospectively to the manufacturer’s worldwide fluoxetine pregnancy registry (20) (this is an update of the data presented in Reference #10). Of the total number, 37 pregnancies were identified during clinical trials and 759 from spontaneous reports. Spontaneous abortions occurred in 110 (13.8%) cases, and for the remaining 686 pregnancies, malformations, deformations, and disruptions occurred in 34 (5.0%). No consistent pattern of defects was observed and only one minor malformation was reported. Moreover, no recurring patterns of malformations, increase in unusual defects, or adverse outcomes were observed in 89 infants from 426 retrospectively reported pregnancies (20). Based on these data, the authors concluded that it was unlikely that the drug was related to an increased risk of malformations (20). Others, however, have previously pointed out that under-reporting and documentation of outcomes are problems with these types of surveillance (13).
The neurodevelopment of children between the ages of 16 and 86 months, who had been exposed in utero for varying lengths of duration to fluoxetine (N=55) or tricyclic antidepressants (N=80), were described in 1997 (21). A control group (N=84) of children not exposed to any agent known to adversely affect the fetus was used for comparison. Assessments of neurodevelopment were based on tests for global IQ and language development and were conducted in a blinded manner. No statistically significant differences were found between the three groups in terms of gestational age at birth, birth weight, and weight, height or head circumference at testing. The mean global IQ scores in the fluoxetine, tricyclic, and control groups were 117, 118, and 115, respectively (differences not significant). Moreover, there were no significant differences in the language scores, or assessment of temperament, mood, arousability, activity level, distractibility, or behavior problems (21). In addition, no significant differences between the three groups were found with analysis of the data by comparing those exposed only during the 1st trimester to those exposed throughout pregnancy.
A 1998 noninterventional observational cohort study described the outcomes of pregnancies in women who had been prescribed one or more of 34 newly marketed drugs by general practitioners in England (22). Data were obtained by questionnaires sent to the prescribing physicians one month after the expected or possible date of delivery. In 831 (78%) of the pregnancies, a newly marketed drug was thought to have been taken during the 1st trimester with birth defects noted in 14 (2.5%) singleton births of the 557 newborns (10 sets of twins). In addition, two birth defects were observed in aborted fetuses. However, few of the aborted fetuses were examined. Fluoxetine was taken during the 1st trimester in 52 pregnancies. The outcomes of these pregnancies included 2 ectopic pregnancies, 6 spontaneous and 6 elective abortions, 11 cases lost to follow-up, 25 normal newborns (1 premature), and 2 infants with major malformations (22). The birth defects were spina bifida with hydrocephalus (mother also took dothiepin, sodium valproate, and carbamazepine) and congenital hypothyroidism. In addition, one newborn with a normal chromosome pattern had a minor congenital anomaly (single palmar creases).
In a 1998 case report, a 32-year-old woman with bipolar disorder took fluoxetine, buspirone, and carbamazepine (see Breast Feeding Summary for doses and further details) throughout gestation (23). At 42 weeks’ gestation she gave birth to a healthy 3940-g, normally developed female infant. The mother continued her medications for 3 weeks while exclusively breast feeding the infant. She reported seizure-like activity in her infant at 3 weeks, 4 months, and 5.5 months of age.
In 1999, the Swedish Medical Birth Registry published the results of a study on the use of antidepressants in early pregnancy and delivery outcome for the years 19951997 (24). During the period, 281,728 infants were registered, 531 of whom had been exposed in utero to SSRI antidepressants, 15 to SSRIs plus a non-SSRI antidepressant, and 423 to non-SSRI antidepressants. Of the 16 women who used fluoxetine, 15 used it alone and 1 used it in combination with clomipramine. There was no significant differences in relative risk (RR = observed/expected) for birth defects between those exposed to any depressant (total 39; RR 1.13), SSRIs only (total 21; RR 1.12), and non-SSRIs only (total 18; RR 1.15). Similarly, no significant differences in infant survival were observed among the groups. A shorter gestational duration (
The use of fluoxetine in two women was associated with the induction of ovulation that had previously been resistant to clomiphene (25). Although no pregnancies occurred, ovulation continued after fluoxetine was discontinued.
In summary, the available animal and human experience with fluoxetine appears to indicate that the antidepressant is not related to major congenital malformations (26,27 and 28). The increased rate of three or minor anomalies found in one investigation, however, may be evidence that the drug does adversely affect embryonic development. However, this study did not control for depression and, thus, the incidence of fluoxetine-induced minor anomalies remains unknown. The other studies cited above lacked the sensitivity to identify minor anomalies because of the absence of standardized examinations. The perinatal complications, observed in three case reports and one study when fluoxetine exposure occurred late in pregnancy, are probably uncommon and a causative association with the antidepressant has not been proven. Moreover, these toxic complications, even if some have been drug induced, must be weighed against the potential maternal risks of discontinuing antidepressant therapy. In addition, the incidence of spontaneous abortions in exposed pregnancies, which has been increased over controls in several studies, needs further investigation. Finally, although one study failed to measure any effects of in utero exposure to fluoxetine on human central nervous system development, more investigations of this potentially serious toxicity are required. A published admonition (29), relating to an earlier work (by one of the authors of the above negative study) emphasizing that their data at that time could not be interpreted as excluding the possibility of behavioral teratology, still seems to be appropriate. Because at least one animal study has shown that fluoxetine can produce changes, perhaps permanently, in the fetal brain, the maternal benefits must be carefully weighed against the potential embryo and fetal risks before exposing a pregnancy to this drug.
Breast Feeding Summary
Fluoxetine is excreted into human breast milk (23,30,31 and 32,34,35 and 36). A 1990 case report described a woman, 3 months postpartum, who was started on fluoxetine, 20 mg every morning, for depression (30). No drug-related adverse effects were noted in the infant by the mother or the infant’s pediatrician. However, the woman’s husband, also a pediatrician, thought the nursing infant showed increased irritability during the first 2 weeks of therapy. Two months after treatment had begun, plasma and milk samples were obtained from the mother (time in relationship to the dose was not specified). Plasma concentrations of the antidepressant and its active metabolite, norfluoxetine, were 100.5 and 194.5 ng/mL, respectively. Similar measurements in the milk were 28.8 and 41.6 ng/mL, respectively. The milk:plasma ratios for the parent compound and the metabolite were 0.29 and 0.21, respectively (30).
A 1992 report described a woman treated for postpartum depression with fluoxetine, 20 mg at bedtime, 10 weeks after delivery (31). The dosing time was chosen just before the infant’s longest period of sleep to lessen his exposure to the drug. After 53 days of therapy, milk and serum samples were collected 8 hours after the usual dose and 4 hours after a subsequent dose administered to approximate peak concentrations of fluoxetine. Serum concentrations of fluoxetine and the active metabolite at 4 hours were 135 and 149 ng/mL, respectively, and at 8 hours 124 and 141 ng/mL, respectively. The variation in the milk samples was greater, with values at 4 hours of 67 and 52 ng/mL (hand-expressed foremilk), respectively, and at 8 hours of 17 and 13 ng/mL (hand-expressed hindmilk obtained after nursing), respectively. The authors speculated that the variation in milk levels was more likely because of differences in milk composition (foremilk being high in protein and low in fat; hindmilk having a higher fat content) rather than a reflection of maternal serum concentrations. Assuming that the milk contained a steady concentration of 120 ng/mL of fluoxetine and norfluoxetine, and the infant was ingesting 150 mL/kg/day of milk, the authors calculated that the maximum theoretical dose that the infant had received was 15 to 20 g/kg/day. No adverse effects were observed in the nursing infant’s behavior, feeding patterns, or growth during the treatment period (31).
A case study that appeared in 1993 described colicky symptoms consisting of increased crying, irritability, decreased sleep, vomiting, and watery stools in a breast-fed infant whose mother was taking fluoxetine, 20 mg/day (32). The mother had begun breast-feeding the infant immediately after birth and began taking fluoxetine 3 days later. The baby began to show the symptoms noted above at 6 days of age. The mother was enrolled in a study of infant crying at 3 weeks postpartum and at 6 weeks, the infant was switched to a commercial formula for 3 weeks. The mother continued to pump her breasts during this time. She noted a marked change in the infant’s behavior shortly after the change to formula feeding. Under an approved study protocol, the mother’s breast milk concentrations of fluoxetine and norfluoxetine were measured (by a commercial laboratory), revealing levels of 69 ng/mL and 90 mg/mL, respectively. After 3 weeks of bottle-feeding, feeding with the mother’s milk from a bottle was resumed, and within 24 hours the colic returned and she restarted feeding with the commercial formula. Drugs levels of fluoxetine and metabolite, determined by a commercial laboratory, in the infant’s serum on the second day after the return to mother’s milk were 340 ng/mL and 208 ng/mL, respectively. The authors associated the symptoms of colic with the presence of fluoxetine in the mother’s milk (32).
The very high infant serum levels of fluoxetine and metabolite, similar to therapeutic range in adults, are difficult to explain based on the mother’s low dose. A 1996 review suggested that one possible explanation was laboratory error (33).
The presence of fluoxetine and its active metabolite, norfluoxetine, were measured in the breast milk of 10 women and in serum or urine of some of the 11 (one set of twins) nursing infants (median age, 185 days) (34). The women had been taking fluoxetine at an unchanged dose for at least 7 days (9 women for at least 14 days, 1 for 7 days) before the study. The mean maternal dose of fluoxetine was 0.39 mg/kg/day (range 0.170.85 mg/kg/day). Milk concentrations of fluoxetine, determined from milk samples collected at 2, 5, 8, 12, and 24 hours after a dose (separate samples collected from both breasts of a woman with twins) ranged from 17.4 to 293 ng/mL, whereas those for norfluoxetine ranged from 23.4 to 379.1 ng/mL. In 3 women, the mean milk:plasma ratios for the two agents were 0.88 (range 0.521.51) and 0.82 (range 0.601.15), respectively. Peak milk concentrations of fluoxetine occurred within 6 hours in 8 women, more than 12 hours in 2 women, and undetermined (because of insufficient samples) in 1 woman. A plasma sample obtained from 1 infant contained no measurable drug or metabolite (limit of detection for both
In a 1998 case report, a 32-year-old woman with bipolar disorder took fluoxetine (20 mg/day), buspirone (45 mg/day), and carbamazepine (600 mg/day) throughout pregnancy and during the first 3 weeks postpartum (23). She reported seizure-like activity in the infant at 3 weeks, 4 months, and 5.5 months of age. Breast milk and infant serum were evaluated for the presence of fluoxetine and metabolite on postpartum days 13 and 21. On day 13, fluoxetine concentrations in breast milk were 45 ng/mL and 68 ng/mL (right and left breasts), whereas norfluoxetine levels were 68 and 57 ng/mL (right and left breasts). On day 21, the milk concentrations of the drug and metabolite were 38 and 28 ng/mL (mixed milk), respectively. The infant’s serum had no detectable (test sensitivity not reported) fluoxetine on day 13 but the level was 61 ng/mL on day 21. Norfluoxetine concentrations in infant serum on day 13 and 21 were 58 and 57 ng/mL, respectively. Maternal serum samples were not obtained for fluoxetine analysis. Similar analyses were conducted for buspirone and carbamazepine (see Buspirone and Carbamazepine for results). A neurologic examination of the infant, that included electroencephalography, was within normal limits. The authors were unable to determine the cause of the seizure-like activity, if indeed it had occurred (none of the episodes had been observed by medical personnel) (23).
Serum and milk concentrations of fluoxetine (maternal dose 2040 mg/day) and norfluoxetine in four breast-feeding women were reported in a 1998 study (35). Maternal fluoxetine serum concentrations ranged from 71 to 250 ng/mL, whereas the levels for the metabolite were 67 to 177 ng/mL. Hindmilk levels of fluoxetine and norfluoxetine (always higher than foremilk) ranged from 37 to 132 ng/mL and 11 to 74 ng/mL, respectively. Neither the parent drug nor metabolite could be detected in the serum or urine samples from the nursing infants. No neurological abnormalities were detected in the infants and all had normal mental and psychomotor performance development up to 12 to 13 months of age as assessed by the Bayley Scales of Infant Development (35).
In 14 breast-feeding women receiving a mean fluoxetine dose of 0.51 mg/kg/day, the mean milk:plasma ratios for the parent drug and metabolite were 0.68 and 0.56, respectively (36). The mean total infant dose (fluoxetine plus active metabolite) was estimated to be 6.81% (range 2.15%12%) of the weight-adjusted maternal dose. In nine infants for which plasma samples were obtained, fluoxetine was detected in five (range 20252 ng/mL) and norfluoxetine in seven (range 17187 ng/mL). In eight cases, the antidepressant had also been taken during pregnancy and three of these infants had the highest plasma concentrations of fluoxetine. Two infants had colic which had resolved in one infant before the study. Two other infants, with the highest plasma levels of fluoxetine, norfluoxetine, or both, exhibited symptoms of withdrawal consisting of uncontrollable crying, irritability, and poor feeding. In one case, however, maternal methadone use may have contributed to the symptoms (36).
An abstract of a study published in 1999 examined the effect of maternal fluoxetine therapy on the weight gain of nursing infants (37). A total of 64 women took the antidepressant during pregnancy and 26 continued the drug during breast feeding. The other 38 women, who also breast-fed their infants but who had discontinued the drug, were used as controls. Fluoxetine exposed nursing infants had a growth curve significantly below the controls, averaging a deficit in weight gain of 392 g (95% CI, -5, -780) in measurements taken between 2 weeks and 6 months of age. Although no adverse effects in the exposed nursing infants were reported by the mothers, the reduced growth was thought to be of possible clinical significance if infant weight gain was already of concern (37).
Although only a few of the above reports described toxicity, and the association between the drug and observed effects is uncertain, the long-term effects on neurobehavior and development from exposure to this potent serotonin reuptake blocker during a period of rapid central nervous system development have not been adequately studied. Further, the reduced weight gain identified in one study may have clinical significance in some situations. As reported by the FDA, the manufacturer was advised to revise the labeling of fluoxetine to contain a recommendation against its use by nursing mothers (38). The current labeling contains this revision (2). In contrast, the authors of a 1996 review stated that they encouraged women to continue breast feeding while taking the drug (11). Similarly, a recent review of SSRI agents concluded that if there were compelling reasons to treat a mother for postpartum depression, a condition in which a rapid antidepressant effect is important, the benefits of therapy with SSRIs would most likely outweigh the risks (39). The American Academy of Pediatrics considers the effects of fluoxetine on the nursing infant to be unknown, but they may be of concern (40).
- Shuey DL, Sadler TW, Lauder JM. Serotonin as a regulator of craniofacial morphogenesis: site-specific malformations following exposure to serotonin uptake inhibitors. Teratology 1992;46:36778.
- Product information. Prozac. Dista Products, 2000.
- Cooper GL. The safety of fluoxetinean update. Br J Psychiatry 1988;153(Suppl 3):7786.
- Vedernikov Y, Bolanos S, Bytautiene E, Fulep E, Saade GR, Garfield RE. Effect of fluoxetine on contractile activity of pregnant rat uterine rings. Am J Obstet Gynecol 2000;182:2969.
- Montero D, de Ceballos ML, Del Rio J. Down-regulation of 3H-imipramine binding sites in rat cerebral cortex after prenatal exposure to antidepressants. Life Sci 1990;46:161926.
- Stanford MS, Patton JH. In utero exposure to fluoxetine HCl increases hematoma frequency at birth. Pharmacol Biochem Behav 1993;45:95962.
- Pohland RC, Byrd TK, Hamilton M, Koons JR. Placental transfer and fetal distribution of fluoxetine in the rat. Toxicol Appl Pharmacol 1989;98:198205.
- Pastuszak A, Schick-Boschetto B, Zuber C, Feldkamp M, Pinelli M, Sihn S, Donnenfeld A, McCormack M, Leen-Mitchell M, Woodland C, Gardner A, Hom M, Koren G. Pregnancy outcome following first-trimester exposure to fluoxetine (Prozac). JAMA 1993;269:22468.
- Jacobson SJ, Jones K, Johnson K, Ceolin L, Kaur P, Sahn D, Donnenfeld AE, Rieder M, Santelli R, Smythe J, Pastuszak A, Einarson T, Koren G. Prospective multicentre study of pregnancy outcome after lithium exposure during first trimester. Lancet 1992;339:5303.
- Goldstein DJ, Marvel DE. Psychotropic medications during pregnancy: risk to the fetus. JAMA 1993;270:2177.
- Nulman I, Koren G. The safety of fluoxetine during pregnancy and lactation. Teratology 1996;53:3048.
- Chambers CD, Johnson KA, Dick LM, Felix RJ, Jones KL. Birth outcomes in pregnant women taking fluoxetine. N Engl J Med 1996;335:10105.
- Jones KL, Johnson KA, Chambers CD. Birth outcomes in pregnant women taking fluoxetine. N Engl J Med 1997;336:873.
- Robert E. Treating depression in pregnancy. N Engl J Med 1996;335:10568.
- Cohen LS, Rosenbaum JF. Birth outcomes in pregnant women taking fluoxetine. N Engl J Med 1997;336:872.
- Goldstein DJ, Sundell KL, Corbin LA. Birth outcomes in pregnant women taking fluoxetine. N Engl J Med 1997;336:8723.
- Spencer MJ. Fluoxetine hydrochloride (Prozac) toxicity in a neonate. Pediatrics 1993;92:7212.
- Mhanna MJ, Bennet JB II, Izatt SD. Potential fluoxetine chloride (Prozac) toxicity in a newborn. Pediatrics 1997;100:1589.
- Goldstein DJ. Effects of third trimester fluoxetine exposure on the newborn. J Clin Psychopharmacol 1995;15:41720.
- Goldstein DJ, Corbin LA, Sundell KL. Effects of first-trimester fluoxetine exposure on the newborn. Obstet Gynecol 1997;89:7138.
- Nulman I, Rovet J, Stewart DE, Wolpin J, Gardner HA, Theis JGW, Kulin N, Koren G. Neurodevelopment of children exposed in utero to antidepressant drugs. N Engl J Med 1997;336:25862.
- Wilton LV, Pearce GL, Martin RM, Mackay FJ, Mann RD. The outcomes of pregnancy in women exposed to newly marketed drugs in general practice in England. Br J Obstet Gynaecol 1998;105:8829.
- Brent NB, Wisner KL. Fluoxetine and carbamazepine concentrations in a nursing mother/infant pair. Clin Pediatr 1998;37:414.
- Ericson A, Kallen B, Wiholm BE. Delivery outcome after the use of antidepressants in early pregnancy. Eur J Clin Pharmacol 1999;55:5038.
- Strain SL. Fluoxetine-initiated ovulatory cycles in two clomiphene-resistant women. Am J Psychiatry 1994;151:620.
- Gupta S, Masand PS, Rangwani S. Selective serotonin reuptake inhibitors in pregnancy and lactation. Obstet Gynecol Surv 1998;53:7336.
- Austin MPV, Mitchell PB. Psychotropic medications in pregnant women: treatment dilemmas. Med J Aust 1998;169:42831.
- Addis A, Koren G. Safety of fluoxetine during the first trimester of pregnancy: a meta-analytical review of epidemiological studies. Psychologic Med 2000;30:8994.
- Koren G, Pastuszak A. Psychotropic medications during pregnancy: risk to the fetus (reply). JAMA 1993;270:2178.
- Isenberg KE. Excretion of fluoxetine in human breast milk. J Clin Psychiatry 1990;51:169.
- Burch KJ, Wells BG. Fluoxetine/norfluoxetine concentrations in human milk. Pediatrics 1992;89:6767.
- Lester BM, Cucca J, Andreozzi L, Flanagan P, Oh W. Possible association between fluoxetine hydrochloride and colic in an infant. J Am Acad Child Adolesc Psychiatr 1993;32:12535.
- Wisner KL, Perel JM, Findling RL. Antidepressant treatment during breast feeding. Am J Psychiatry 1996;153:11327.
- Taddio A, Ito S, Koren G. Excretion of fluoxetine and its metabolite, norfluoxetine, in human breast milk. J Clin Pharmacol 1996;36:427.
- Yoshida K, Smith B, Craggs M, Channi Kumar R. Fluoxetine in breast-milk and developmental outcome of breast-fed infants. Br J Psychiatry 1998;172:1759.
- Kristensen JH, Ilett KF, Hackett LP, Yapp P, Paech M, Begg EJ. Distribution and excretion of fluoxetine and norfluoxetine in human milk. Br J Clin Pharmacol 1999;48:5217.
- Chambers CD, Anderson PO, Thomas RG, Dick LM, Felix RJ, Johnson KA, Jones KL. Weight gain in infants breastfed by mothers who take fluoxetine (abstract). Pediatrics 1999;104:11201.
- Nightingale SL. Fluoxetine labeling revised to identify phenytoin interaction and to recommend against use in nursing mothers. JAMA 1994;271:1067.
- Edwards JG, Anerson I. Systematic review and guide to selection of selective serotonin reuptake inhibitors. Drugs 1999;57:50733.
- Committee on Drugs, American Academy of Pediatrics. The transfer of drugs and other chemicals into human milk. Pediatrics 1994;93:13750.