Drugs that Suppress Lactation,Part 1

Antidepressants

In a small prospective study, 8 primiparous women who were taking a serotonin reuptake inhibitor (SRI; 3 taking fluoxetine and 1 each taking citalopram, duloxetine, escitalopram, paroxetine, or sertraline) were compared with 423 mothers who were not taking an SRI. Mothers taking an SRI had an onset of milk secretory activation (lactogenesis II) that was delayed by an average of 16.7 hours compared with controls (85.8 hours postpartum in the SRI-treated mothers and 69.1 hours in the untreated mothers), which doubled the risk of delayed feeding behavior in the untreated group.

A case control study compared the rate of predominant breastfeeding at 2 weeks postpartum in mothers who took a selective serotonin reuptake inhibitor (SSRI) antidepressant throughout pregnancy and at delivery or an SSRI during pregnancy only (n = 117) to a control group of mothers who took no antidepressants. SSRIs taken by the subjects included citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, and sertraline. Among the women who took an SSRI, the breastfeeding rate at 2 weeks postpartum was 27–33% lower than the breastfeeding rate of mothers who did not take an antidepressant. A recent observational study of 1648 women who took an antidepressant during pregnancy found it decreased the likelihood of breastfeeding at hospital discharge by 35 to 75% compared to women who had taken an antidepressant during the 2 years prior to pregnancy, but not during pregnancy.

Data are insufficient to indicate whether any one agent is more likely to affect lactation than another. In general, the adequacy of the treatment of depression is the overriding concern with the use of antidepressants during pregnancy.

Insulin

An older study showed that mothers with type 1 insulin-dependent diabetes are at increased risk for delayed lactogenesis II. ¹ In a more recent study, 883 women with gestational diabetes were interviewed at 6–9 weeks postpartum. Those who had been treated with insulin more frequently reported having a delayed onset of lactogenesis II (>72 hours) postpartum than those not treated with insulin, independent of other maternal risk factors. The odds ratio of having delayed lactogenesis II was 3.1 among insulin-treated mothers compared with mothers with gestational diabetes who did not receive insulin. Other risk factors for delayed lactogenesis II were prepregnancy obesity and suboptimal in-hospital breastfeeding. This study implicates exogenous insulin, or perhaps insulin resistance, as a risk factor for delayed lactogenesis II, independent of diabetes itself.

Terbutaline

Terbutaline is sometimes used as a tocolytic agent to delay premature labor. A small retrospective survey found that mothers who received a beta agonist pharmacologically similar to terbutaline (fenoterol or hexoprenaline) as a tocolytic breastfed for a shorter period of time than those who received no tocolytic (4.5 months vs. 9.5 months). It is not known what the mechanism might be or whether terbutaline has a similar effect.

Betamethasone

Betamethasone is often used to induce fetal lung maturation in preparation for premature delivery. A study of 46 women who delivered an infant before 34 weeks of gestation found that a course of betamethasone given between 3 and 9 days before delivery resulted in delayed lactogenesis II and lower average milk volumes during the 10 days after delivery. Milk volume was not affected if the infant was delivered less than 3 days or more than 10 days after the mother received the corticosteroid.

Magnesium

Some, but not all, studies have found a trend toward decreased sucking in infants or increased time to the first feeding of mothers treated with intravenous magnesium sulfate during labor, probably because of placental transfer of magnesium to the fetus.

Maternal and neonatal outcome endpoints were compared in mothers receiving continuous oral magnesium aspartate supplementation for at least 4 weeks before delivery for various conditions versus nonsupplemented controls. Fewer women who received magnesium supplements breastfed their infants exclusively at discharge (63% vs. 85%).

Oxytocin

Currently oxytocin, the most commonly used oxytocic, has come under scrutiny for its effects on lactation. Numerous studies suggest that oxytocin given during labor can negatively affect breastfeeding. Both a delayed onset of lactation and a greater risk of bottle feeding have been found in various studies. In one fairly large retrospective study, mothers who received oxytocin during the first and second stages of labor had a 45% increased risk of bottle feeding and a 129% increased risk of breastfeeding discontinuation by 3 months of age. Effects were most pronounced in women less than 27 years of age. ²

The decrease in lactation appears to be related to neonatal behavior rather than a direct effect on milk production. Some studies have found alterations in sucking behavior in the newborn as a possible cause. One study found that all rhythmic reflexes, the antigravity reflex, and total primitive neonatal reflexes were inhibited by intrapartum oxytocin administration, which could adversely affect breastfeeding. ³ An interesting hypothesis supported by one prospective study is that use of oxytocin increases maternal labor pain, thereby requiring more analgesia that, in turn, depresses the infant's breastfeeding behavior. ⁴

Dinoprostone

A nonrandomized prospective study compared women who had spontaneous deliveries with those who had elective induction using dinoprostone vaginal gel. At hospital discharge, exclusive breastfeeding rates were similar between the two groups (88% and 89%). However, at 1 and 3 months postpartum, exclusive breastfeeding rates were significantly lower in mothers who had dinoprostone induction than in those who delivered spontaneously. Exclusive breastfeeding rates were 54% and 85% at 1 month and 46% and 59% at 3 months postpartum, respectively. Rates of supplemental and exclusive formula feeding were also higher in the induced mothers at both time points. ⁵

Ergot alkaloids

A retrospective review of obstetrical records of 18,165 records of mothers giving birth in Wales found that use of intravenous or intramuscular ergonovine (no longer available in the United States) during the third stage of labor as a uterotonic reduced the odds of the mother breastfeeding at 48 hours postpartum. The reduction was 36% in the overall sample and 49% for primiparous mothers.

In contrast to ergonovine, effects of methylergonovine are mixed, although the relevant studies have small numbers of women. A single intramuscular injection of methylergonovine 0.2 mg was compared with placebo given during the first 1.5 hours postpartum. At 80–90 minutes after the injection, the normal postpartum rise in serum prolactin was 56% in the women who received methylergonovine compared with a 285% in serum prolactin in women who received a placebo injection. Six treated women had no increase in serum prolactin compared with two of the control women. In another study, women who delivered full-term infants received a single intramuscular dose of methylergonovine 0.2 mg after delivery, followed by oral ergotamine 1 mg three times daily for 6 days. Compared with women who delivered full-term infants and received no ergot derivatives, there was no difference in the milk production, as measured by weight differences before and after nursing, between the two groups during the first 6 days postpartum.

Pain control

A national survey of women in the United States and their infants from late pregnancy through 12 months postpartum compared the time of lactogenesis II in mothers who did and did not receive pain medication during labor. Categories of medication were spinal or epidural only, spinal or epidural plus another medication, and other pain medication only. Women who received medications from any of the categories had about twice the risk of having delayed lactogenesis II (>72 hours) compared with women who received no labor pain medication.

The effects of epidural medication for pain control during labor on breastfeeding initiation and duration have been the subject of numerous studies and a recent systematic review. ⁶ Unfortunately, the results are mixed and confusing, because of the many different combinations of drugs, dosages, and patient populations studied as well as the variety of techniques used and deficient designs of many of the studies. Many studies do not specify the dosages of drugs or even which drugs were used! One small, but carefully performed prospective study that was not included in the systematic review found an inverse correlation between the amount and duration of exposure to epidural fentanyl and the likelihood of achieving suckling during the first hour after a vaginal birth, with a threshold total dosage of about 150 mcg. ⁴ Overall, it appears that commonly used local anesthetics (e.g., bupivacaine) with or without epidural opiates (e.g., fentanyl) have little effect on breastfeeding success when good breastfeeding support is provided to the mother, but as all self-respecting reviewers state, more study is needed.