The Effects of Intrapartum Administration of Synthetic Oxytocin on Breastfeeding in the First 9 Months Postpartum: A Longitudinal Prospective Study

Introduction

Oxytocin (OT) is a peptide hormone produced in the hypothalamic nuclei and released into the bloodstream by the posterior pituitary gland. ¹ OT is also synthesized in the placenta and uterus during labor and birth ² and plays a key role during pregnancy and perinatal period. As pregnancy advances, the number of oxytocin receptors that are located throughout the body (including uterus and breasts) increases.^1,3 During labor, dilation of the cervix stimulates OT release, which enhances uterine contractions. The endogenous OT effects also include reduction of maternal anxiety and stress, as well as facilitation of bonding between the mother and the child. ⁴ Importantly, OT is known to be essential for initiation and continuation of breastfeeding. ⁵

In maternity care settings, synthetic oxytocin (synOT) is frequently administered to induce or accelerate labor and to prevent postpartum hemorrhage. However, while endogenous OT is essential for the initiation of lactation and milk ejection reflex, ⁴ some research suggests that synOT administered during peripartum period may be associated with negative breastfeeding outcomes such as delayed or difficult onset and shorter duration of breastfeeding.^6–10

A majority of the studies on the association between peripartum synOT and breastfeeding focused on the early postpartum period, reporting prevailingly negative effects of synOT on breastfeeding outcomes in the first hours or days after the child's birth.^7–9,11,12 While Morillo et al. ¹³ found no relationship between synOT and breastfeeding in the early postpartum period, there is no study reporting positive effects of synOT on breastfeeding status soon after birth.

The few studies investigating the effects of intrapartum synOT on breastfeeding status several months after the child's birth have provided conflicting results. Gu et al. ¹⁴ and Fernandez et al. ⁹ observed a negative dose-dependent effect of synOT on exclusive breastfeeding at 2 and 3 months postpartum, respectively, and García-Fortea et al. ⁶ reported that synOT predicted a higher risk of bottle feeding and breastfeeding withdrawal at 3 months. However, Gomes et al. ⁷ found significant impact of synOT on breastfeeding at 3 months in the nonadjusted, but not in the adjusted analysis. Similarly, Morillo et al. ¹³ found no difference in the exclusive breastfeeding rates at 3 and 6 months between the women who received intrapartum synOT and those who did not. The same research group published a study based on a larger sample size 2 years later, ¹⁵ concluding that although the synOT dose administered intrapartum is not associated with a risk of exclusive breastfeeding cessation at 3 and 6 months, women who gave birth vaginally without synOT were exclusively breastfeeding at 3 and 6 months more frequently compared to those with synOT who had either vaginal or cesarean delivery. Importantly, most of the existing studies included only healthy mother-child pairs and accounted for gestational age at birth. However, only Gomes et al. ⁷ and Fernandez et al. ⁹ reported on early mother-child contact, including only mother-child pairs who had skin-to-skin contact immediately after the birth. No study controlled for maternal breastfeeding self-efficacy that is considered to be an important predictor of breastfeeding outcomes. ¹⁶

The previous studies examining the link between synOT and breastfeeding usually followed the mother-child pairs during the first postpartum days, with few studies extending the time period until several postpartum months. To the best of our knowledge, the longest time period of the follow-up was 6 months postpartum.^13,15 Moreover, only few studies adopted a prospective design,^9,13–15 whereas the others relied on the retrospective assessments of breastfeeding status.^6,7 Another limitation of the existing research can be seen in the relatively small sample size of the previous studies, where only some of them reported the results based on a sample consisting of more than 100 mother-child pairs.^6–8,15 It was therefore the aim of this study to extend the current knowledge on the association between intrapartum synOT and breastfeeding by using prospectively collected data from a relatively large sample of mother-child pairs, covering the time span of the first 9 postpartum months.

Materials and Methods

Results

SynOT was administered intrapartum to 100 women (22.8%) of those for whom data at T1 and T2 were available, and to 65 women (23.7%) of those for whom data at T3 were available. The proportion of women who were breastfeeding exclusively was 89.1% at T1, 84.5% at T2, and 14.6% at T3. Only 2.7% of women were not breastfeeding at T1, 9.8% at T2, and 24.8% at T3. The breastfeeding status of the women with and without synOT at all three time points is shown in Table 2.

Intrapartum synOT was associated with a lower rate of exclusive breastfeeding at T1 in a bivariate analysis (odds ratio [OR] = 0.33; confidence interval [CI] = 0.17–0.61; p < 0.001) (Table 3) and this association remained significant after adjusting for maternal age, parity, educational level, marital status, child's sex, delivery mode, and intrapartum analgesia/anesthesia (OR = 0.37; CI = 0.18–0.75; p = 0.006) (Table 4). No association between synOT administration and breastfeeding status was found in the bivariate or adjusted analyses at T2 and T3 (Tables 3 and 4).

The pairwise comparisons showed no significant effect of synOT on breastfeeding at T2 coded as exclusive versus partial (OR = 1.11; CI = 0.37–3.38; p = 0.850) or as exclusive versus none (OR = 1.67; CI = 0.63–4.41; p = 0.299) in multivariate models adjusted for maternal age, parity, educational level, marital status, child's sex, delivery mode, and intrapartum analgesia/anesthesia.

Discussion

The aim of this study was to examine the effects of intrapartum synOT on breastfeeding status several days, 6 weeks, and 9 months postpartum. Our data indicate that synOT administered intrapartum may prevent successful initiation of breastfeeding in the first postpartum days, but we found no significant association between synOT and breastfeeding status at later time points.

Our finding that use of synOT may be negatively associated with breastfeeding rates in the first postpartum days is consistent with prior studies focusing on the immediate postpartum period.^7–9,11,12 One of the underlying mechanisms that has been proposed to explain such finding is alteration of maternal oxytocinergic system. ²⁰ While endogenous OT is secreted in a periodic bolus manner, ²¹ synOT is administered as a continuous infusion. SynOT may interfere with the pulsatile secretion of mother's own endogenous OT and desensitize oxytocin receptors, ²² which may weaken the milk ejection reflex.^1,23

Another pathway through which synOT may affect breastfeeding outcomes is the alteration of the fetal oxytocinergic system. ¹⁰ SynOT may cross the placental barrier and the fetal blood–brain barrier ²⁴ and affect the function of the fetal nervous system oxytocin receptors.^10,22 Although the research on behavioral effects of synOT in the offspring is still emerging, several studies reported inhibition of breastfeeding-related reflexes^9,11,12 and fewer prefeeding cues^10,23 in newborns exposed to synOT during birth.

As we excluded newborns with serious health issues from our sample to avoid confounding, it seems unlikely that adverse newborn health status could explain the link between synOT and the lower rate of exclusively breastfeeding women in the early postpartum in our sample. However, although not uniformly observed in the studies,^25,26 intrapartum synOT has been linked with a higher risk of neonatal jaundice,^27–30 which is in turn associated with weak sucking and less optimal breastfeeding outcomes. ³¹ This could explain why we detected synOT-related breastfeeding problems in the immediate postpartum period, but not at the later time points during the first postpartum year.

Our finding that there is no association between intrapartum synOT administration and breastfeeding status at the later time points (6 weeks and 9 months after birth) is in line with several studies that found no effect of synOT on breastfeeding outcomes at 3^7,13,15 and 6 months postpartum. ¹³ Nevertheless, it is inconsistent with others reporting longer-term negative influence of synOT on breastfeeding. For example, Gu et al. ¹⁴ reported a dose-dependent negative effect of synOT on exclusive breastfeeding at 2 months postpartum, adopting a prospective design and using a relatively large sample. However, while Gu et al. only excluded preterm-born children, thus using a medically diverse sample, our sample consisted of healthy mother-child pairs. Another study based on a relatively large sample ⁶ reported adverse synOT effects on both onset and duration of breastfeeding, although the authors assessed breastfeeding status retrospectively, 5 years after childbirth, which might have led to a recall bias. In addition, Fernández et al. ⁹ found that mothers exclusively breastfeeding versus those not exclusively breastfeeding at 3 months received a lower average dose of synOT. Nevertheless, this was a pilot study, including 20 participants who all received synOT, and those results should therefore be considered preliminary.

Another explanation why our results differ from the previous studies may be seen in the variability of national policies to support mothers in the postpartum period. In the Czech Republic, women generally breastfeed for a relatively long period of time, which is enabled by paid maternity or parental leave that may last up to 4 years, with most women being on maternity leave during the first year of child's life. In the countries where mothers return to work soon after childbirth and the breastfeeding rates are thus lower in general, the effects of synOT on breastfeeding duration might be more apparent. Nevertheless, it must be pointed out that we found no effect of synOT on breastfeeding status not only at 9 months but also at 6 weeks postpartum, suggesting that other factors than long-term maternity leave (such as risk status of the women included in the sample) should also be considered while interpreting our results.

This study has several strengths, including its prospective design spanning the first 9 postpartum months and a relatively large sample size. However, several limitations must be noted. Most importantly, as data regarding synOT dosage were not available, we only considered whether synOT was administered or not. In addition, the attrition rate from T2 to T3 was relatively high. Although this is a common limitation of longitudinal studies, it might affect the generalizability of our findings. Also, we excluded mothers and children with compromised health status, including diabetes and hypertension in pregnancy, and preterm birth to avoid confounding, even though it is an additional limitation, reducing the generalizability of our results that may only apply to healthy populations. Also, it is a limitation of this study that we did not control the analyses for early mother-infant contact and maternal breastfeeding self-efficacy, the variables that are linked with breastfeeding outcomes in the literature.^16,32 Another limitation concerns the time points of data collection. While collecting data about breastfeeding status in the first days, at 6 weeks, and at 9 months postpartum, we did not collect data at 6 months, which made it impossible to determine whether synOT administration affected the length of exclusive breastfeeding (a period of 6 months is recommended as optimal for exclusive breastfeeding duration). ³³ Moreover, although we controlled for the effect of delivery mode and intrapartum pain medication in our models, it cannot be ruled out that the effects of synOT on breastfeeding might be mediated by birth outcomes. Nevertheless, the mediating effect of birth outcomes was not the topic of our study.

Conclusion

This study is the first to prospectively investigate the effects of perinatal administration of synOT on breastfeeding outcomes beyond the 6 months postpartum. We found a negative association between the synOT use and the initiation of breastfeeding in the early postpartum period, but no association was observed at later time points. Nevertheless, even though the effect of synOT on breastfeeding was only observed in the early postpartum, our findings are clinically relevant as this period is critical for breastfeeding success. Future studies on the associations between synOT and breastfeeding outcomes should take into account synOT dosage, while using ethnically and medically diverse samples and including early mother-child contact and breastfeeding self-efficacy as potential confounders.