Review research and develop research questions
Chekwube00
Eriksonetal.2017articleexampleROL.pdf
Journal of Midwifery &Women’s Health www.jmwh.org
Review
CEUBreastfeeding Outcomes After Oxytocin Use During
Childbirth: An Integrative Review Elise N. Erickson, CNM, MS, Cathy L. Emeis, CNM, PhD
Introduction:Despite widespread use of exogenous synthetic oxytocin during the birth process, few studies have examined the effect of this drug
on breastfeeding. Based on neuroscience research, endogenous oxytocin may be altered or manipulated by exogenous administration or by block-
ing normal function of the hormone or receptor. Women commonly cite insufficient milk production as their reason for early supplementation,
jeopardizing breastfeeding goals. Researchers need to consider the role of birth-related medications and interventions on the production of milk.
This article examines the literature on the role of exogenous oxytocin on breastfeeding in humans.
Methods: Using the method described by Whittemore and Knafl, this integrative review of literature included broad search criteria within the
PubMed, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Cochrane, and Scopus databases. Studies published in English
associating a breastfeeding outcome in relation to oxytocin use during the birth process were included. Twenty-six studies from 1978 to 2015 met
the criteria.
Results: Studies were analyzed according to the purpose of the research, measures and methods used, results, and confounding variables. The
26 studies reported 34 measures of breastfeeding. Outcomes included initiation and duration of breastfeeding, infant behavior, and physiologic
markers of lactation. Timing of administration of oxytocin varied. Some studies reported on low-risk birth, while others included higher-risk ex-
periences. Fifty percent of the results (17 of 34measures) demonstrated an association between exogenous oxytocin and less optimal breastfeeding
outcomes, while 8 of 34 measures (23%) reported no association. The remaining 9 measures (26%) had mixed findings. Breastfeeding intentions,
parity, birth setting, obstetric risk, and indications for oxytocin use were inconsistently controlled among the studies.
Discussion: Research on breastfeeding and lactation following exogenous oxytocin exposure is limited by few studies and heterogeneousmethods.
Despite the limitations, researchers and clinicians may benefit from awareness of this body of literature. Continued investigation is recommended
given the prevalence of oxytocin use in clinical practice.
J Midwifery Womens Health 2017;62:397–417 c© 2017 by the American College of Nurse-Midwives.
Keywords: active management third-stage labor, breastfeeding, drug effects, lactation, labor (induced), labor (obstetric), labor stage (third),
oxytocin
INTRODUCTION
While increasing numbers of women are breastfeeding their newborns at birth, the ability to maintain breastfeeding may be affected by factors contributing to maternal milk pro- duction. This is reflected by the Centers for Disease Con- trol and Prevention (CDC) 2016 Breastfeeding Report Card, which shows that while 81.1% of women initiate breastfeed- ing after birth, only 44.4% of women are still exclusively breastfeeding at 3 months, falling to 22.3% of infants by the 6-month target.1 Common reasons for early cessation of ex- clusive or any breastfeeding is the perception of insufficient milk supply2,3 and the early introduction of formula.4,5 There- fore, factors that may influence physiologic milk production are compelling targets for translational research.
Understanding possible causes of suboptimal breastfeed- ing may have implications for improving maternal and in- fant health. Infants receiving formula or solid foods before 6 months of age are at increased risk for acute and chronic illnesses, as well as sudden infant death syndrome.6 The number of infant deaths potentially preventable by meeting breastfeeding goals are estimated upwards of 700 annually.7,8
Furthermore, a growing body of literature is examining the
Address correspondence to Elise N. Erickson, CNM, MS, Oregon Health & Science University, School of Nursing, 3455 SE US Veterans Hospital Rd, Portland, OR 97239-2941. E-mail: [email protected]
long-term effect of breastfeeding on maternal health. Women who have no breastfeeding history have poorer indices of car- diovascular health in later life.9 Another study used a simula- tionmodel to estimate the impact of suboptimal breastfeeding on many maternal health outcomes, reporting a potential an- nual excessmortality of 3340 deaths andmore than $14 billion in costs in the United States due to premature death.7
Milk production and successful breastfeeding require oxytocin-driven neuroendocrine pathways that are primed by pregnancy and the process of childbirth.10 Endogenous oxytocin function is essential for onset of lactation and milk ejection in mammals.11 Manipulation of oxytocin in experi- mental animal models can lead to deficits in lactation, ma- ternal behavior, and abnormal behavioral development of offspring.12,13 Oxytocin is commonly administered inmodern maternity care for labor augmentation, induction of labor,14
and to minimize or treat uterine bleeding in the third stage of labor.15 There is evidence that exogenous oxytocin can pass through the placenta and into fetal circulation.16 Therefore, depending on the timing of administration, this synthetic hor- mone and neurotransmitter could affect neonates as well as women.
The significance of these questions relate to the extensive use of oxytocin in practice. Estimates of induction of labor, typically involving exogenous oxytocin, range from 23% to 29% of births17,18 but may be in the range of 31% to 42% in
1526-9523/09/$36.00 doi:10.1111/jmwh.12601 c© 2017 by the American College of Nurse-Midwives 397
✦ Oxytocin administration during childbirth is widespread; few studies have investigated the effects of this on breastfeeding, and most of these have not directly studied the relationship.
✦ The effect of exogenous oxytocin on breastfeeding has been measured through infant breastfeeding behavior, physiologic lactation, maternal initiation, and duration or exclusivity of breastfeeding.
✦ While oxytocin administration has an important role in modern maternity care, potential effects on lactation should be explored more, as the research on breastfeeding outcomes is incomplete.
some settings, based onUSdata.19,20 Amongwomenwho start labor spontaneously, augmentation of labor with oxytocin due to slow progress is also frequent,20 though exact national rates are not published. Epidural analgesia is also associated with induced and augmented labor, with more than 75% of women using epidural analgesia undergoing induction or augmenta- tion, according to 2008 CDC data.21 During cesarean birth, accounting for 32.7% of births,18 oxytocin is administered af- ter extracting the placenta to slow bleeding.15 Finally, to help minimize bleeding, the World Health Organization (WHO) promotes prophylactic administration of oxytocin as the stan- dard of care following vaginal birth.15 It is also a mainstay treatment for postpartum hemorrhage.
Despite widespread use of oxytocin and the importance of the physiology of oxytocin for successful lactation, clinical studies have rarely explored long-term effects on women and infants, such as breastfeeding outcomes.22,23 The purpose of this integrative review is to understand 1) what breastfeeding outcomes (maternal or infant) have been reported following any clinical oxytocin administration and 2) any patterns in the published results to better inform future research.
METHODS
An integrative approach described by Whittemore and Knafl informed the procedure for this review, as a preliminary litera- ture search revealed significant heterogeneity in methods and outcomes among relevant studies.24 We were unable to iden- tify articles synthesizing the body of literature regarding oxy- tocin administration in humans and breastfeeding outcomes. The complexity of this question is owed to both the various indications and timing of oxytocin use during the birth pro- cess and the multifactorial nature of breastfeeding and lacta- tion research outcomes. In an effort to capture all possible oxy- tocin administration during the birth process, our review in- cluded intrapartum oxytocin and/or third-stage labor admin- istration. Breastfeeding outcomes were defined as any mater- nal and infant breastfeeding-related measure.
Literature Search
Due to the exploratory nature of this investigation, the ap- proach included broad search terms and no limits on pub- lication date. We performed a Boolean search (as shown in Table 1) of PubMed Medical Subject Heading (MeSH) terms including: 1) “oxytocin,” “labor (induced),” “labor (obstet- ric),” “labor stage (third),” or “epidural analgesia”; and 2) “breastfeeding,” “feeding behavior,” “lactation,” or “lactation
(disorder),” yielding 1847 results after limiting to human studies. A duplicate search in the Cumulative Index to Nurs- ing and Allied Health Literature (CINAHL) yielded 268 cita- tions (“infant behavior” substituted for “feeding behavior”). A total of 2115 abstracts (including duplicates) were scanned for inclusion by 1) data-based studies published in English and 2) noting oxytocin administration and a breastfeeding outcome (maternal or infant). If a potential match did not mention oxytocin administration in the abstract, the full text was reviewed in detail. Induction of labor studies not evaluat- ing oxytocin specifically were excluded, as well as studies as- sessing infant bottle feeding. The resulting group consisted of 26 studies published between 1978 and 2015.
Data Evaluation
Significant heterogeneity in the study objectives, design, and outcomes complicated the evaluation of this body of litera- ture. Themajority of the studies were descriptive or secondary analysis reports (either prospective or retrospective); how- ever, one randomized controlled trial, 2 quasi-experimental studies, and 2 case-control studies also made up the sample.
While studies in this review considered oxytocin exposure during birth with at least one breastfeedingmeasure, most did not set out to study this relationship. Many noted the associ- ation between oxytocin and breastfeeding as a subanalysis of the primary aim or as a covariate or control for another objec- tive. We identified 3 groups of research objectives within the sample studies. Only 9 studies examined the effect of oxytocin use on breastfeeding. Four studies examined factors (general health and obstetric) associated with delayed lactogenesis and poor breastfeeding generally. In these reports, use of oxytocin was among many variables considered. The largest group of studies, however, sought to understand broad outcomes of specific obstetric interventions: epidural analgesia (n = 4), medication use (n = 3), active management of third-stage la- bor (AMTSL) (n = 1), or as part of an induction of labor (n = 5). These studies included a breastfeeding measure among other outcomes.
Time point of oxytocin administration varied among the studies, illustrated in Figure 1. The majority considered intra- partum oxytocin administration only. Four of these assessed the postpartum dose of oxytocin as well.25–28 Another 3 stud- ies mention that oxytocin was routinely given postpartum but was not included in the analysis in terms of exposure.29–31
Three other studies addressed the third-stage issue generally by reporting “increased need for postpartum uterotonics”
398 Volume 62, No. 4, July/August 2017
Table 1. Search Strategy for Oxytocin Use During Birth and Breastfeeding
Database Search Terms (MeSH and Keyword) Results
Unique Studies
Included
PubMed Oxytocin, labor (induced), labor (obstetric), labor stage (third),
epidural analgesia AND breastfeeding, feeding behavior, lactation,
lactation disorder
598 14
Lactogenesis (keyword) 131 3
Labor (induced) AND oxytocin 1118 4
CINAHL Oxytocin, labor (induced), obstetric care, labor stage (third), epidural
analgesia AND breastfeeding, infant behavior, lactation, lactation
disorder
89
Lactogenesis 54 1
Labor (induced) AND oxytocin 125 0
Cochrane Induced labor AND breastfeeding 13 0
Active management (third stage) labor 1 1
Scopus 1
Hand check of reference
lists
1
Total 26
Abbreviation: CINAHL, Cumulative Index to Nursing and Allied Health Literature.
Figure 1. Number of Studies by Time Point of Oxytocin Exposure
and Type of BreastfeedingMeasures Reported
MaternalMeasures of Breastfeeding (Initiation orDuration
of Breastfeeding, Physiology of Lactation).
Infant Measures of Breastfeeding (Infant Feeding
Behavior).
Both Maternal and Infant Measures.
Abbreviations: AOL, augmentation of labor; IOL, induction of labor; PP, postpartum prophylaxis.
(ie, oxytocin and other medications),32–34 or commenting on the relationship of postpartum hemorrhage and breastfeeding outcomes.35
Breastfeeding outcomes included maternal behaviors like initiation, duration of breastfeeding, measures of physio- logic milk production (eg, hormones, lactogenesis), and in- fant breastfeeding behavior. A total of 34 measures in the 26 studies were examined in relationship to oxytocin use as il- lustrated in Figure 2. Some studies reported more than one outcome in the findings. Due to the variety of study objec- tives, methods, and outcomes used in the sample, rigor of the studies was not evaluated by a standardized rubric or score. Instead, we addressed quality of the studies by assess- ing and synthesizing themes that may introduce bias or limit generalizability.
RESULTS
Breastfeeding Outcomes
No primary study outcome associated oxytocin use with a more favorable breastfeeding outcome. Data were arranged into 3 categories: 1) use of oxytocin (intrapartum and/or post- partum) and a less optimal breastfeeding outcome, 2) no association, or 3) having mixed findings. Results were la- beled mixed if they were the subanalyses of the primary aim of the study or significance was seen in certain subgroups of the sample (ie, primiparas). Of the 34 measures reported in the studies, 50% found oxytocin use was associated with a less optimal breastfeeding outcome (n = 17). Mixed or qual- ified support of less optimal outcomes was reported by 26% (n = 9), and 23% showed no differences in breastfeeding out- comes with oxytocin use or not (n = 8). Table 2 lists the mea- sures, statistical data, and information about the study design and limitations.
Initiation of Breastfeeding
Eleven studies examined associations between breastfeeding initiation and oxytocin administration; 7 studies reported on initiation only.28,32,33,36–39 Initiation of breastfeeding was de- fined by various time points ranging from 10 minutes after birth through 7 days postpartum. An additional 4 studies re- ported duration measures as well as initiation measures of breastfeeding.30,40–42
Four of these 11 studies were generated from large data sets and controlled for multiple covariates in their analyses.28,32,33,36 Two noted delay in initiation of breastfeed- ing following induction of labor and elective induction of la- bor in Latin American countries.32,33 Another reported lower breastfeeding rates at hospital discharge following AMTSL in
Journal of Midwifery &Women’s Health � www.jmwh.org 399
Figure 2. Number ofMeasures by Direction of Findings Reporting Relationship Between Oxytocin Use and BreastfeedingOutcomes
Measures Showing Less Optimal Breastfeeding Outcome With Oxytocin Use.
Measures Reporting Mixed Findings: Less Optimal Outcome With Oxytocin Use in Subgroup Analysis.
Measures Reporting No Association Between Oxytocin and Breastfeeding.
theUnited Kingdom.28 In this study, after controlling formul- tiple intrapartum factors and examining a subgroup ofwomen with low-risk, physiologic labors, AMTSL was still associated with an approximate 7% reduction in breastfeeding at 2 days postpartum.
However, the study by Prendiville,39 the only random- ized controlled trial in the sample, did not find an as- sociation between AMTSL and breastfeeding at hospital discharge. This study is limited by a lack of fidelity to the ran- domization; only 403 of 849 participants allocated to physio- logic management had it performed. In addition, the physio- logic group was also more likely to put the newborn to breast 10 minutes after birth per midwives’ recommendation.
Brown and Jordan42 also did not find thatAMTSL affected rates of breastfeeding initiation in a self-report study of breast- feeding and administration of postpartum oxytocin.42 How- ever, they did report a reduction in duration of breastfeeding at both 2 and 6 weeks postpartum among participants who had AMTSL. The most often reported reasons for cessation were pain, difficulty, and embarrassment compared towomen who had physiologic management. This study did not control for prenatal intentions to breastfeed.
Altogether, the definition of initiation of breastfeeding was variable but appeared to reflect the first several postpar- tum days. Five papers associated delayed initiation of breast- feeding with induction or augmentation of labor compared to spontaneous labor or no augmentation (postpartum use not reported)30,32,33,37 or postpartum administration of oxytocin compared to expectant management.28 Mixed findings were reported in 3 studies.36,40,41
Duration of Breastfeeding
Eight studies examined duration of breastfeeding. This was defined as the time of breastfeeding cessation,25 report of ex- clusive breastfeeding at 3 months after birth,30,31 at 6 weeks postpartum,42,43 or breastfeeding at 8 weeks.26,40,41 Shorter duration or exclusivity of breastfeeding was associated with intrapartum oxytocin use by 4 studies compared to sponta- neous labor25,26,30,31 and with postpartum use in the study by Brown and Jordan.42 Two reports hadmixed findings ondura- tion of breastfeeding.40,43 One paper reported no difference.41
The total dosage of oxytocinwas examined in terms of du- ration of breastfeeding by 2 authors. Both Gu et al26 andOlza- Fernandez31 noted that higher levels of exposure to oxytocin during the birth process were associated with reduced exclu- sive breastfeeding at 2 and 3months postpartum, respectively. Additionally, the participants in the study by Dozier et al25
most likely to cease breastfeeding by one month postpartum were those with both epidural analgesia and oxytocin expo- sure during labor (HR, 1.34; 95% confidence interval [CI], 1.00-1.79).25 Women with epidural analgesia in this study were more likely to have oxytocin administered during labor (58.8% vs 38.3%, P � .01). Breastfeeding was not analyzed by total dosage specifically in this study, but this may imply that women with epidural analgesia had more need for oxytocin administration, possibly representing higher total dosage.
Physiology of Lactation
Eight studies examined breastfeeding as a measure of phys- iologic milk production. Six of these examined lactogenesis
400 Volume 62, No. 4, July/August 2017
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Gu et al,26
2015
Canada
Oxytocin Time Point
Intrapartum and postpartum
Design
Prospective longitudinal
Baby-Friendly settinga
Mixed parity sample
Self-report: Exclusivity of
breastfeeding at 2 months
Plasma oxytocin levels at 2
months
N = 386
92% of women received oxytocin
Duration
Exclusively breastfeeding mothers at 2 months
postpartum had received significantly less oxytocin
during labor (33 units) when compared to formula
(44 units) or mixed feeding mothers (43 units)
(after controlling for education level) (P � .0001)
Physiology of Lactation
Circulating oxytocin at 2 months postpartum was
positively correlated to dosage given during birth
(Pearson, 0.16, P � .01)
Did not specify the rates of analgesia,
mode of birth, indication for
oxytocin use, or neonatal problems
Breastfeeding intention not reported
Did not control for parity or other
neonatal or obstetric issues in
breastfeeding outcomes
Brimdyr et al,49
2015
United States
Oxytocin Time Point
Intrapartum
Design
Prospective Comparative
Baby-Friendly setting
Mixed parity sample
Widström’s 9 instinctive
stages of neonatal
behavior
N = 63
84% of women having oxytocin with or without
epidural analgesia
Infant Behavior
Infants born after exposure to oxytocin were less
likely to suck in the first hour after birth (P = .03).
Dose dependent response.
Groups examined with use of epidural analgesia,
which also exhibited a main effect by dosage and
was frequently interrelated with oxytocin use
Breastfeeding intention not reported
Duration of oxytocin exposure not
analyzed in relation to infant
behavior
Duration of labor overall not
controlled
Maŕın-Gabriel et al,29
2015
Spain
Oxytocin Time Point
Intrapartum
Design
Prospective cohort
Baby-Friendly setting
Mixed parity sample
Breastfeeding intentions
reported (inclusion
criteria)
Primitive neonatal reflexes
related to feeding on days
1-2 postnatal
N = 98
53 women received oxytocin, 45 women did not
Infant Behavior
Fewer reflexes noted in newborns exposed to
oxytocin infusion compared to nonexposed,
(�, −12.7; 95% CI, −25 to −0.5)
Adjusted for parity, labor difficulty, epidural
analgesia use
Nulliparas and epidural analgesia were
more common in the oxytocin
group, though this was controlled in
the analysis
Dose of oxytocin not reported
(Continued)
Jo u rn alo
f M id w ifery
& W o m en’s
H ealth
� w w w .jm
w h .o rg
401
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Mauri et al,47
2015 Italy
Oxytocin Time Point
Intrapartum
Design
Prospective longitudinal
descriptive
Mixed parity sample
Self-report: timing and
intensity of
lactogenesis-related breast
symptoms
N = 366
62.8% of women received oxytocin
Physiology of Lactation
No association between oxytocin infusion alone and
onset of lactation symptoms (HR, 1.06; 95% CI,
0.77-1.45)
Epidural analgesia related to oxytocin infusion (P �
.001) and suboptimal breastfeeding at 20 days (P =
.02)
Baby-Friendly not reported
Skin-to-skin not reportedb
Rooming-in not protocol
Breastfeeding intention not reported
Intrapartum oxytocin protocol lower
than other studies: 5 units/500 mL
Oxytocin dose not recorded/reported
Brown & Jordan,42
2014
United Kingdom
Oxytocin Time Point
Postpartum
Design
Retrospective descriptive
Mixed parity
Self-report: feeding method
at birth, duration of
breastfeeding
N = 288
84.1% of sample reported postpartum oxytocin
administration
Initiation
No differences between active and physiologic third
stage on breastfeeding after birth (OR, 0.57; 95%
CI, 0.23-1.42)
Duration
AMTSL associated with reduced levels of
breastfeeding at 2 weeks (OR, 0.35; 95% CI,
0.18-0.71) and 6 weeks (OR, 0.38; 95%
CI,0.19-0.78), but not at birth
90.2% of the formula-feeding group at 2 weeks
received AMTSL compared to 76.3% of the
breastfeeding group
Relationship held when women with epidural
analgesia and gestational age �41 weeks were
removed from analysis (to control for possible
intrapartum exposure)
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention not reported
Self-report of labor procedures subject
to recall bias
Could not control for all intrapartum
synthetic oxytocin use
(Continued)
402 V o lu m e 62,N
o.4,Ju ly/A
u gu st 2017
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Garćıa-Fortea et al,30
2014
Spain
Oxytocin Time Point
Intrapartum
Design
Retrospective descriptive
cohort (randomly selected)
Parity not reported
Self-report breastfeeding
status and duration of
breastfeeding
N = 316
59.8% women received oxytocin
Initiation
Synthetic oxytocin was associated with fewer reports
of breastfeeding (63.5% of exposed group vs 92.1%
nonexposed) (RR, 1.45; 95% CI, 1.288-1.635)
Duration
For duration (n = 237), use of synthetic oxytocin
(120/237) associated with average of 33 fewer days
of breastfeeding.
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention not reported
Parity not reported
Medical record used for clinical
variables; self-report (5 years prior):
breastfeeding status (study does not
report which time point this report
represents) and duration of
breastfeeding (reported in days)
Duration not specified as exclusive or
partial breastfeeding
Large proportion of sample was twin
gestation (30.7%)
Bell, White-Traut, &
Rankin,48
2013 United States
Oxytocin Time Point
Intrapartum
Design
Prospective descriptive
Mixed parity
Prefeeding behaviors
Neonatal Behavioral
Assessment Scale 45
minutes after birth
N = 47
76.5% of women received oxytocin
Infant Behavior
Newborn behaviors in the exposed group were more
likely to show low levels of feeding behavior
compared to unexposed who had more high-level
prefeeding behavior (OR, 11.5; 95% CI, 1.8-73.3)
Adjusted for labor length and epidural analgesia use
Newborns went to a warmer following
birth per hospital routine,
skin-to-skin not routine
Breastfeeding intention not reported
Vogel, Souza, &
Gülmezoglu,36
2013
16 Africa/Asian
Countries
Oxytocin Time Point
Intrapartum
Design
Retrospective descriptive
WHO Global Survey
Initiation of breastfeeding
�24 h – 7 days
N = 192,538
11,700 (6%) induction with oxytocin
Initiation
Increased odds of not breastfeeding in first 24 hours
in Asian sample (OR, 2.17; 95% CI, 1.27-3.73); also
associated with increased risk of low Apgar, birth
weight, and ICU admission
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention not reported
Oxytocin effect not examined with
controls for obstetric complications
(per aim of study)
(Continued)
Jo u rn alo
f M id w ifery
& W o m en’s
H ealth
� w w w .jm
w h .o rg
403
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Olza Fernández et al,31
2012
Spain
Oxytocin Time Point
Intrapartum
Design
Prospective longitudinal
descriptive
Mixed parity
Skin-to-skin noted (not
Baby-Friendly)
Duration exclusivity at 3
months
Primitive neonatal reflexes
on second day of life
N = 20
100% received oxytocin for induction or
augmentation
100% had epidural analgesia; 30% forceps rate
Infant Behavior
Negative association between rate of newborn
sucking reflex after birth and dosage of oxytocin
administered (P = .03).
Duration
Women exclusively breastfeeding at 3 months were
exposed to significantly less oxytocin during birth
(P = .04).
Breastfeeding intention not reported
Small sample, pilot study
All women had epidural analgesia,
effect of epidural analgesia could not
be controlled statistically
Dozier et al,25
2012
United States
Oxytocin Time Point
Intrapartum and postpartum
Design
Prospective cohort
Baby-Friendly in part of
sample (controlled for in
analysis)
Breastfeeding goals and
confidence reported
Secondary analysis of
self-report and medical
record data: duration at 2
months postpartum
N = 727
50% of women received intravenous oxytocin
14.8% had intramuscular oxytocin.
Duration
Combination of epidural analgesia and intrapartum
oxytocin had increased early cessation (HR, 1.34;
95% CI, 1.00-1.79); absence of epidural analgesia
and oxytocin were most protective of ongoing
breastfeeding
Women giving birth in a Baby-Friendly hospital who
had oxytocin IV were less likely to have early
breastfeeding cessation (HR, 0.67; 95% CI,
0.53-0.86)
Women giving birth in non-Baby-Friendly hospitals
who had oxytocin IV were more likely to have early
breastfeeding cessation (HR, 1.50; 95% CI,
1.25-1.80)
Postpartum dose not included in
oxytocin exposure for analyses
Indication for oxytocin use was not
specified
(Continued)
404 V o lu m e 62,N
o.4,Ju ly/A
u gu st 2017
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Guerra et al,33
2011
8 Latin American
countries
Oxytocin Time Point
Intrapartum
Design
Retrospective descriptive
(secondary analysis)
WHO Global Survey:
Initiation of breastfeeding
�24 h – 7 days
N = 37,597
Subset of elective induction of labor compared to
low-risk spontaneous labor
4.4% oxytocin exposure for elective induction of
labor
Initiation
Increased risk of delayed initiation (compared to first
hour after birth) of breastfeeding adjusting for
parity, mode of birth, etc. (RR, 1.59; 95% CI,
1.24-2.05).
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention not reported
Oxytocin effect not examined with
controls for obstetric complications
(per aim of study)
Nommsen-Rivers
et al,35
2010
United States
Oxytocin Time Point
Intrapartum
Design
Prospective longitudinal
descriptive
Primiparous
Breastfeeding intention:
inclusion criteria
Onset of lactogenesis—
maternal report
N = 431
56.6% of women received oxytocin for induction or
augmentation
Overall delayed lactogenesis rate 44.3%
Physiology of Lactation
Delayed lactogenesis not associated with oxytocin
exposure
Shorter labor predicted less delayed lactogenesis but
only for non-oxytocin group
Baby-Friendly not reported
Duration of labor reported but not
duration of oxytocin exposure—only
if it were part of the labor
Indications for labor induction or
augmentation not reported
Matias et al,45
2009
Peru
Oxytocin Time Point
Intrapartum
Design
Prospective longitudinal
descriptive
Baby-Friendly
Primiparous
Onset of lactogenesis-
maternal report
Researcher observation of
breastfeeding behavior
with Infant Breastfeeding
Assessment scale; infant
weight loss
N = 156
2.3% induction of labor rate
15% augmentation of labor with oxytocin rate
Physiology of Lactation
Of the augmented group, 30.4%, reported delayed
onset of lactogenesis compared to 15% of the
nonaugmented group (P = .1); not associated with
excess weight loss or suboptimal breastfeeding
behavior
Breastfeeding intention not reported
Breastfeeding outcomes of women
with labor induction not reported in
table
Low number of women with oxytocin
exposure for labor augmentation (n
= 25)
(Continued)
Jo u rn alo
f M id w ifery
& W o m en’s
H ealth
� w w w .jm
w h .o rg
405
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Guerra et al,32
2009
8 Latin American
countries
Oxytocin Time Point
Intrapartum
Design
Retrospective descriptive
WHO Global Survey:
Initiation of breastfeeding
�24 h – 7 days
N = 97,095
87% of inductions used oxytocin (11,077 total
inductions)
Initiation
Induction associated with delayed initiation of
breastfeeding until after the first day (RR, 1.31; 95%
CI, 1.22-1.43) adjusted for multiple risk factors
*See Guerra33
Jordan et al,28
2009
United Kingdom
Oxytocin Time Point
Intrapartum and postpartum
Design
Prospective data collection,
secondary analysis
Mixed parity
Medical record: Initiation of
breastfeeding by 48 hours
N = 48,366
79% of women received uterotonic medication
(oxytocin and/or ergometrine) in the third stage of
labor
10% were induced with oxytocin
Initiation
Third-stage labor uterotonic associated with reduced
breastfeeding at 48 hours postpartum in all women
(P � .001) and primiparous subset (P � .001) for
IM or IV oxytocin and ergometrine; this controlled
for other medications in labor, social class, parity,
age, and deprivation rank
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention not reported
Classification of women breastfeeding
at 48 hours included women partially
breastfeeding and excluded women
who were expressing milk
Outcome variable of breastfeeding at
48 hours unclear if referring to entire
48 hours or just the last feeding at
that time (ie, discharge feeding
diagnosis).
Jonas et al27
2009
Sweden
Oxytocin Time Point
Intrapartum and postpartum
Design
Prospective descriptive
comparative
Skin-to-skin reported,
number of feeds during
first 2 days not different
between groups
Breastfeeding intention
reported
Oxytocin and prolactin
levels during
breastfeeding on second
day postpartum
N = 63
Physiology of Lactation
Prolactin levels peaked earlier (10 minutes) (P = .01)
and were higher in the oxytocin intrapartum
groups (P = .006) for up to 60 minutes (P = .001)
Negative correlation between amount of oxytocin
during labor and median level of oxytocin in blood
on second postpartum day (rs = −.495, P = .02)
No clinical measures of breastfeeding
outcomes were linked to the
hormone data to correlate clinical
significance
(Continued)
406 V o lu m e 62,N
o.4,Ju ly/A
u gu st 2017
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Kong & Bajorek,46
2008
Australia
Oxytocin Time Point
Intrapartum
Design
Prospective descriptive
Breastfeeding intention was
reported
Onset of
lactogenesis—maternal
report
N = 75
6.7% of the sample received oxytocin for induction of
labor; postpartum use not reported
Physiology of Lactation
Average (SD) time to onset of lactogenesis was 77.0
(34.7) hours for induction of labor with oxytocin (n
= 5), compared to 68.1 (22.8) hours for
spontaneous labor (n = 28) (P = .66).
Baby-Friendly not reported
Skin-to-skin not reported
Sample receiving oxytocin small,
underpowered for this comparison
Wiklund et al,37
2007
Sweden
Oxytocin Time Point
Intrapartum
Design
Comparative retrospective:
matched control
Mixed parity (analysis did
control for parity, length of
labor in regression
analyses)
Initiation after birth,
formula supplementation
N = 702
54% of the women received oxytocin during labor
Initiation
Oxytocin administration associated with delayed
initiation �4 hours of breastfeeding (OR, 3.28; 95%
CI, .1.57-6.84) and giving artificial milk
supplement (OR, 2.15; 95% CI, 1.28-3.61).
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention not reported
(Continued)
Jo u rn alo
f M id w ifery
& W o m en’s
H ealth
� w w w .jm
w h .o rg
407
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Dewey et al,34
2003
United States
Oxytocin Time Point
Intrapartum (postpartum?)
Design
Prospective longitudinal
descriptive
Breastfeeding intention:
inclusion criteria
Self-report onset
lactogenesis
Infant behavioral
observation Infant
Breastfeeding Assessment
Tool
N = 280
31% of the women received oxytocin for labor
augmentation; no data on induction of labor
Physiology of Lactation
32% of augmented group had delayed onset
lactogenesis compared to 18% of nonaugmented
group (P � .05)
64% of the sample received “postpartum hemorrhage
medications,” which may have included oxytocin,
and 26% of this group had delayed onset of
lactogenesis compared to 16% (P � .1)
Multiple regression analysis was not significant for
oxytocin
Infant Behavior
No differences in suboptimal infant breastfeeding
behavior scores or weight loss of infant
Baby-Friendly not reported
Skin-to-skin not reported
Duration/dosage of oxytocin
augmentation not reported;
comparison of lactogenesis outcomes
from augmentation include women
who had scheduled cesarean births
(n= 11), which may affect the results
Radzyminski,50
2003
United States
Oxytocin Time Point
Intrapartum
Design
Prospective comparative
Multiparous only
Preterm Infant
Breastfeeding Behavior,
Neurologic and Adaptive
Capacity Score
N = 56 dyads
Unknown percentage of sample receiving oxytocin
Infant Behavior
6 infants scored below the mean for breastfeeding
behavior; these had a higher incidence of labor
induction
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention: not reported
Data outcomes on breastfeeding
behavior incomplete: percent not
reported, no descriptive statistics
(Continued)
408 V o lu m e 62,N
o.4,Ju ly/A
u gu st 2017
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Chapman & Perez-
Escamilla,44
1999
United States
Oxytocin Time Point
Intrapartum
Design
Longitudinal prospective
descriptive
Mixed parity
Self-report onset
lactogenesis
N = 192
Physiology of Lactation
Induction with oxytocin was not associated with
delayed onset of lactogenesis in chi-square test
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention not reported
Number of women induced with
oxytocin not reported; cannot make
comparison to those not exposed
Rajan,43
1994
United Kingdom
Oxytocin Time Point
Intrapartum
Design
Descriptive retrospective,
secondary analysis
Self-report breastfeeding at
6 weeks
N = 1064
18% of the sample reported oxytocin for
induction of labor
Duration
Chi-square analysis showed relationship between
oxytocin use and shorter duration of second
stage (�1 hr) was associated with reduced
exclusive breastfeeding compared to women
who had a longer second stage or were not
receiving oxytocin (P = .04)
Baby-Friendly not reported
Skin-to-skin not reported
Breastfeeding intention not reported
Statistical analysis not robust; no
regression analysis; multiple
chi-square tests cannot control for
confounding variables
Out, Vierhout, &
Wallenburg,41
1988
Netherlands
Oxytocin Time Point
Intrapartum
Design
Prospective
quasi-experimental with
control group
Mixed parity
Intention to breastfeed
recorded at 36 weeks of
pregnancy
Nursing staff report “any
serious attempt” and
self-report 3-4 days
postpartum and at 6
months
N = 185
26% of the sample received oxytocin for
induction and 16% for augmentation
Initiation & Duration
More women decided not to breastfeed in the
elective induction of labor group than the
others; rates of breastfeeding beyond initiation
did not differ over the reported 1 and 2 month
postpartum time points
Skin-to-skin not reported
Statistical analysis not robust
Did not control for confounding
factors: duration of labor or parity
(Continued)
Jo u rn alo
f M id w ifery
& W o m en’s
H ealth
� w w w .jm
w h .o rg
409
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Prendiville et al,39
1988
United Kingdom
Oxytocin Time Point
Postpartum
Design
Randomized trial
Medical records:
breastfeeding at discharge
N = 1695
74% of sample received active management
Initiation
No difference between groups in breastfeeding at
discharge (OR, 0.96; 95% CI, .77-1.19)
Skin-to-skin: not specifically reported;
women in control group encouraged
to put baby to breast in first 10
minutes after birth more than
AMTSL group (225/849 vs 63/846)
Breastfeeding intention not recorded
Lack of fidelity to treatment group:
only 403/849 in physiologic
management had this performed
compared to 840/846 in the
treatment group
Breastfeeding outcome not examined
by parity, oxytocin intrapartum
exposure
Yudkin et al,38
1979
United Kingdom
Oxytocin Time Point
Intrapartum
Design
Retrospective case control
Breastfeeding intention:
recorded at first prenatal
visit
Mixed parity
Breastfeeding at discharge N = 400
185/200 induction group received oxytocin
18 of the spontaneous group had oxytocin
augmentation
Initiation
Of the women intending to breastfeed during
antenatal care, 86% of the spontaneous group were
breastfeeding at “discharge” compared to 82% of
induction group
Skin-to-skin not reported
Inconsistent outcome variable;
discharge outcome was “when
records stop,” which include some
follow-up postpartum care
(Continued)
410 V o lu m e 62,N
o.4,Ju ly/A
u gu st 2017
Table 2. Studies Reporting an Association Between Synthetic Oxytocin Use and a BreastfeedingOutcome
Author, Year, Location Design Measures Results Limitations
Ounsted et al40
1978
United Kingdom
Oxytocin Time Point
Intrapartum
Design
Prospective longitudinal
quasi-experimental with 3
induction methods and
control group
Primiparous only
Breastfeeding intention
recorded
Breastfeeding self-report at
birth and 4 days later and
at 2 months postpartum
N = 184
26% of women received oxytocin for induction of
labor
Intention to breastfeed ranged from 66% to 71% of
each comparison group
Initiation
Fewer women changed to bottle feeding in
spontaneous labor group compared to all induction
methods
Duration
Oxytocin group alone were breastfeeding 37.1% at 2
months compared to 68% of the spontaneous
group (NS P � .1)
Skin-to-skin not recorded
Statistical methods limited analysis of
oxytocin group alone due to high
number of cells in the chi-square
analysis; did not control for multiple
confounding variables like length of
labor or neonatal issues
Abbreviations: AMTSL, active management of third-stage labor; HR, hazard ratio; IM, intramuscular; IV, intravenous; NS, nonsignificant; OR, odds ratio; RR, relative risk; SD, standard deviation; WHO, World Health Organization. aBaby-Friendly Initiative certification noted in study for research site. bSkin-to-skin: the practice of mothers holding their infants skin-to-skin after birth.
Jo u rn alo
f M id w ifery
& W o m en’s
H ealth
� w w w .jm
w h .o rg
411
onset, consistently defined by maternal report of breast fullness by 72 hours postpartum.34,35,44–47 Three studies reported no association between lactogenesis and synthetic oxytocin use during labor.44,46,47 Three papers reportedmixed findings.34,35,45 None of these studies’ primary aim was to examine the role of synthetic oxytocin on lactogenesis; thus, these findings were the result of subanalyses or covariate data. All of these studies were prospectively conducted and sam- pledmixed populations regardingmodes of birth (eg, vaginal, cesarean, instrument assisted) and use of analgesia. None reported information on postpartum oxytocin exposure.
Augmentation of labor with exogenous oxytocin (compared with no oxytocin) was associated with de- layed lactogenesis in a bivariate analysis by Dewey et al34
(P � .05) but not in regression analysis. Matias et al45 found a marginal association with labor augmentation in bivariate analysis as well (P � .10), but adjusted analyses found only low Apgar score predicted delayed lactogenesis. Nommsen- Rivers et al35 found no difference in delayed lactogenesis with oxytocin administration for induction or augmentation compared to women who had none. Postpartum oxytocin use was not considered by these studies, except as implied by Dewey et al, noting that women receiving “postpartum hemorrhage medications” were more likely to have delayed lactogenesis (26% compared to 16%, P � .10).
The final 2 maternal studies examined physiologic re- sponse by measuring hormone levels in maternal plasma in relation to oxytocin use. Jonas et al27 examined phys- iologic response to exogenous oxytocin during birth via blood samples collected during a breastfeeding session 2 days postpartum.Maternal oxytocin and prolactin levels were measured; however, this was not reported in relationship to any clinical marker of lactation (ie, lactogenesis). They further demonstrated an inverse relationship (r = −.495, P = .02) between the total dosage administered during la- bor and level of oxytocin found in women’s blood at 48 hours during breastfeeding (n = 61). Prolactin levels in women who received third-stage prophylaxis with oxytocin (n = 13) were lower compared to the 20 women who received no oxytocin.
Gu et al26 measured exclusivity of breastfeeding as well the level of plasma oxytocin in maternal blood at 2 months postpartum. The authors found higher levels of oxytocin in women exposed to higher collective dosages of oxytocin (intrapartum and postpartum), which were also linked to higher likelihood of formula or nonexclusive breastfeeding at 2 months.
Infant Behavior
In relationship to oxytocin administration, authors exam- ined prefeeding behaviors,48 Primitive Neonatal Reflexes,29,31
the Widström 9 stages of instinctive newborn behavior after birth,49 suboptimal infant breastfeeding behavior as measured by the Infant Breastfeeding Assessment Tool,34,35,45
and finally, the Premature Infant Breastfeeding Behavior Scale.50
Four infant studies reported a significant negative rela- tionship between oxytocin used for induction or augmenta- tion of labor and infant behaviors or feeding-related reflexes
in healthy newborns. Three of these found that higher dosages of oxytocin predicted lower infant behaviors,31,48,49 while one did not.29 Radzyminski50 reported that term infants undergo- ing oxytocin-induced labor scored below themean for breast- feeding behavior on the Premature Infant Breastfeeding Be- havior Scale; however, no statistics were provided. In contrast, oxytocin exposure did not associate with differences between groups on the Infant Breastfeeding Assessment Tool when as- sessed during the first week.34,45
Questions about the generalizability of the infant-focused studies arise from the variation in the measurement of neona- tal behavior. The Primitive Neonatal Reflex tool has not been widely used in clinical breastfeeding assessment,29 but these innate reflexes (eg, hand to mouth, finger flexion and ex- tension, gazing, head turning, bobbing, sucking, swallowing) relate to behaviors necessary to locate the maternal breast, latch, and suckle unassisted. The study by Bell et al48 recorded “prefeeding” behaviors, which are a subset of reflexes more associated with feeding (eg, hand to mouth, rooting, suck- ing on hand). Brimdyr et al49 video-recorded the first hour of skin-to-skin contact following birth and reported theWid- ström stages, which lead to unassisted suckling at the breast by the newborn when placed skin-to-skin with the mother during this period. Conversely, the Infant Breastfeeding As- sessment Tool is a validated measure that assesses an in- fant’s breastfeeding mechanics.34 This measure evaluates 4 behaviors—readiness, rooting, latching, and sucking—on a 12-point scale; these measures were used for infants be- yond the immediate birth period. While it may imply neu- robehavioral organization, it is also influenced by position- ing and maternal efforts to assist her infant, as the infants are not assessed for unassisted latching as during Widström stages.
Overall, the body of literature reports breastfeeding out- comes from birth through several months postpartum in- cluding mothers’ and infants’ experiences. Notably, only 3 studies31,34,45 measured both maternal and infant factors. While the results do demonstrate various statistical associa- tions, generalizability of these findings may be affected by the aim of the study or limitations of study setting, sample, and control of confounding variables.
Setting
The majority of studies originated in Western Europe and Australia (n = 15) and the United States and Canada (n =
8). A minority of studies were in the developing world (n
= 4). Three of these utilized large international data sets from the WHO Global Survey,32,33,36 2 of which, conducted by Guerra et al32-33, addressed 2 different questions within Latin America (induction of labor and elective induction of labor).
Five studies described “baby-friendly” or early skin-to- skin practices following birth.25,27,29,31,49 In the report by Bell et al,48 neonates went to a warmer after birth, per hospital routine. This study utilized an open crib for observation of prefeeding behavior at 40 minutes of life, in contrast to the other 3 early infant behavior studies that reported observa- tions while the neonate remained in physical contact with the mother. Despite these differences, the infant behavior studies
412 Volume 62, No. 4, July/August 2017
did report similar diminished feeding-related behavior asso- ciated with oxytocin use.
Setting of the studies is important as likelihood of use of exogenous oxytocin during birth, and the promotion of early breastfeeding best practices, would affect outcomes related to this study question. Studies observing low rates of induction or augmentation of labor,32,33,36,43,45,46 using lower volumes of oxytocin for induction of labor (ie, 5 units/500mL),47 or those that do not report the percentage of the sample exposed44,50
would bemore difficult to compare to populationswith higher rates. Newborns that had no or interrupted skin-to-skin time following birth may also have a different breastfeeding course than others. Standardizing these study elements would be im- portant for interpreting the findings.
Sample
Parity
Many studies in this review did not control for parity, and 2 did not report parity.30,43 Parity predicted not only breastfeed- ing differences34,37,44,47 but also risk of oxytocin exposure.29
Dewey et al34 noted that use of oxytocin was greater for prim- iparous women than multiparous (38% vs 23%), though the variable was not included in the regression model of delayed lactogenesis with interactions of parity. Interestingly, of the studies that foundno association between exogenous oxytocin and suboptimal breastfeeding, all used a sample of women of mixed parity. However, 2 studies reported a significant ef- fect of oxytocin on decreased expression of primitive neonatal reflexes29 and breastfeeding initiation28 even after controlling for parity.
Intention to Breastfeed
Three studies linking oxytocin administration to poor breast- feeding outcomes did not report intentions to breastfeed among their samples, only initiation and duration.26,30,42 This factor introduces study bias, as women with strong inten- tions to breastfeed may persist if difficulties arise. Of the 4 studies that reported risks for delayed lactogenesis, only 2 recorded maternal intentions to breastfeed, which were in- clusion criteria.34,35 A minority of studies examining inter- ventions during birth on breastfeeding reported maternal in- tention to breastfeed38,40,41,46,47 or breastfeeding confidence25; as such, the risk of bias in the findings for breastfeed- ing attrition should be considered with this limitation in mind.
Obstetric Risk Level
Twelve studies focused on a lower-risk sample (eg, vaginal birth, healthy newborns) versus higher risk (eg, cesarean birth, preterm birth, neonatal intensive care unit [NICU] admission). Seven of the 12 low-risk studies’ samples ex- amined the role of synthetic oxytocin on breastfeeding as a primary aim, highlighting the outcomes of healthy, lower-risk women and neonates born vaginally in relation to oxytocin exposure specifically. For example, the 4 infant be- havioral studies examining feeding reflexes included healthy neonates (normal Apgar score and no NICU admission)
born vaginally; all studies controlled for epidural analgesia use, which was not significantly related to the neonatal behaviors except for the study by Brimdyr et al.49 While using a lower-risk sample reduces the risk of confounding variables contributing to the breastfeeding outcomes, it limits generalizability to women with more complex courses of care and surgical birth. However, differences noted among lower-risk women in breastfeeding strengthen the pos- sible association of exogenous oxytocin and suboptimal breastfeeding.
In contrast, the studies examining delayed lactogenesis, those using the Infant Breastfeeding Assessment Tool, and outcomes of obstetric interventions included varied levels of obstetric risks for breastfeeding problems. The effect of this single intervention of oxytocin is therefore difficult to dis- cern from the rest. Only 3 studies in these categories focused on low-risk vaginal birth.25,37,47 Several other studies in these groups reported low rates of oxytocin use34,45,46 or did not report the proportion of sample exposed,44 which limit the interpretation of the findings.
Indications for Synthetic Oxytocin Use
Despite studies in this review stating that healthy or lower-risk women participated, authors did not routinely report the in- dications for the use of oxytocin. Various labor-related factors may drive the use of oxytocin, such as use of epidural anal- gesia or length of labor. Eight studies examined labor dura- tion in relation to breastfeeding outcomes. Four of the 8 as- sociated longer labor with less optimal breastfeeding.34,35,37,50
Notably, 3 of these studies grouped primiparous and mul- tiparous women together for this analysis, and multiparous women are more likely to have shorter labors as well as less difficulty breastfeeding.
Epidural analgesia and oxytocin use are often correlated.21
This finding may be due to the potential for epidural anal- gesia to lower endogenous oxytocin levels in maternal cir- culation, which may slow second-stage labor51,52 or lead to other factors (eg, fetal malposition) that may contribute to augmentation.51,53 Oxytocin-induced or augmented labor may be perceived asmore painful, therebywomenopt for neu- roaxial analgesia.54,55 However, some research has not consid- ered the specific role of oxytocin when studying the effect of epidural analgesia on lactation.56
Exogenous oxytocin may be useful in reducing risks as- sociated with prolonged labor. Breastfeeding problems may also be associated with longer labors, but researchers should try to tease apart the role of oxytocin from labor duration. For example, Nommsen-Rivers et al35 reported an association between length of labor and prevalence of delayed lactation; women with spontaneous labors less than 14 hours in dura- tion had significantly less delayed lactogenesis, 35.7% com- pared to 57% of women with labors longer than 14 hours. In contrast, womenwith oxytocinwho labored less than 14 hours had 47.1% delayed lactogenesis compared to 40.1% of those who labored greater than 14 hours. However, the authors did not report the dose of oxytocin nor proportion of labor ex- posed to oxytocin which limits the analysis. Finally, Matias et al,45 looking only at primiparas, did not find a relationship between long labor and delayed lactation.
Journal of Midwifery &Women’s Health � www.jmwh.org 413
Second-stage labor was also examined by 6 studies.34,35,37,43,44,46 Three reported less delayed lactation in women who pushed for less than 60 minutes compared to longer second stage34,44,46; however, they included multi- parous women in their analysis. Data by Rajan43 contrasted with other study findings; administration of oxytocin was associated with higher bottle-feeding at 6 weeks postbirth but only when second stage was less than one hour compared with greater than one hour when receiving oxytocin.
The primiparas in the Jonas study that evaluated levels of endogenous oxytocin and prolactin27 had augmentation of labor due to slow or stalled labor. Therefore, the differ- ences seen in blood levels of oxytocin may be attributable to other physiologic differences in thewomenwho required oxy- tocin administration. However, in this small sample, third- stage administration was prophylactic, and changes in pro- lactin following oxytocin administration in this group could be more directly linked to the drug itself. It is unknown if women requiring induction or augmentation of labor are innately different physiologically, which may also impact breastfeeding.
DISCUSSION
The purpose of this review was to conduct a thorough exploratory search for research on synthetic oxytocin and breastfeeding outcomes. No 2 studies were similar enough to provide results at the level of meta-analysis. Given the varia- tions in study design, we cannot conclude that oxytocin use during the birth process contributes to altered breastfeeding outcomes. However, because many of the studies did show associations between exogenous oxytocin and less optimal breastfeeding outcomes, especially in lower-risk samples, this question deserves more research before ruling out the possi- bility of an effect.
Exposure to Synthetic Oxytocin
Augmentation of labor tends to occur when labor is already prolonged. Oftentimes synthetic oxytocin can be infused for many hours or days during a lengthy induction process. The availability of the oxytocin receptor in uterine tissue may be a function of duration and/or the level of oxytocin in circulation.57,58 Whether oxytocin receptors located in breast tissue respond similarly to those in uterine tissue has not been researched directly. However, use of oxytocin in this review was often reported as a binary outcome rather than a contin- uous outcome of dosage or duration. Study participants with minimal augmentation would have been grouped together with those having significantly longer exposure. Furthermore, study designs that do not adequately sample women exposed to oxytocin have more limited generalizability or power to detect a difference between groups. Consideration of the duration and dosage of oxytocin rather than a binary outcome may be more relevant to this line of research.
Measurement of Breastfeeding
As illustrated by this review, the measure of breastfeeding varies greatly. The only outcome reported with consistency was the maternal report of timing of lactogenesis. This mea-
sure has been found to be linked to the increased likelihood of continued breastfeeding.59 Maternal report of breast full- ness is considered reliable and valid.60 However, significant variation in the initiation and duration outcomes were a func- tion of the design, feasibility of the studies as well as the ori- gin of the data (ie, medical records). The binary nature of the breastfeeding variable in many of the studies also cannot consider the women who are partially breastfeeding and sup- plementing formula or donor milk. Several studies measured breastfeeding duration via maternal report, one occurring 5 years after birth, leaving room for recall bias.30 While some research has noted that early exclusive breastfeeding may pre- dict longer-term outcomes,4 many of these studies did not include any longitudinal data.
Infant behavioral studies in this review, particularly those examining the primitive and feeding reflex behaviors of healthy newborns, did share similarities in design and find- ings. As explained by the authors, the underpinnings of these designs rest on the potential for oxytocin to cross the pla- centa and act within the brain of the newborn either indi- rectly through feedbackmechanisms (afferent vagus nerve) or directly by possibly crossing the blood-brain barrier itself or as an effect of increased lactate levels,49 all of which are hy- pothesized to alter the behaviors based on animal research models.61,62
Limitations
This review has clear limitations due to high variability within the reviewed studies’ designs. It is also not exhaustive; many elements of statistical analysis and synthesis of other outcomes (eg, role of cesarean birth or postpartum hemorrhage) were outside the scope.
Research Implications
Broadly, this review highlights the paucity of literature on this topic, despite the known physiology of oxytocin and lacta- tion, frequent use of the hormone in childbirth, and growing emphasis on improving breastfeeding. Addressing this gap is possible through 2 main lines of commonmaternal-infant re- search. First, many studies published on lactation outcomes do not address the role of oxytocin use during labor and birth or control for its use.56,63 Second, studies of labor induction or AMTSL are commonly done to compare intervention pro- tocols, yet they rarely report lactation outcomes. These stud- ies often utilize larger sample sizes, more rigorous random- ized designs, and can control for more factors like parity or duration of labor, which would be helpful in addressing this question.
Several specific recommendations stem from this review. First, future lactation research regarding oxytocin should consider neonatal behaviors as well as maternal function. Differences in newborn behavior may manifest as maternal report of decreased milk supply or duration of exclusive breastfeeding. Second, setting and selection bias should be considered, including breastfeeding intentions of the participants and birth practices. Third, measurement of oxytocin used in labor should be more comprehensive, including indicated or elective administration, combined
414 Volume 62, No. 4, July/August 2017
intrapartum and postpartum dosages, and those following cesarean birth. Fourth, better reporting on epidural anal- gesia use and timing of oxytocin administration, including the order and duration of events, would help address the temporal role of the 2 often concurrent interventions on subsequent outcomes. Finally, cumulative pharmacoki- netic effects should be considered (dosage and duration). As research on oxytocin outside of childbirth has shown a dosage response in terms of behavioral and biological effects,61,64 dosage-related (rather than binary) data would be more informative when characterizing exposure to oxytocin.
Clinical Implications
Use of synthetic oxytocin has an important place in modern midwifery and obstetric care, as its use can reduce morbidity or mortality in the setting of prenatal complications or dysto- cia or during postpartum hemorrhage. We have reviewed and organized this body of literature to inform clinicians about ex- isting research.We recommend counseling clients that there is no proven effect of this medication on lactation or breastfeed- ing outcomes while noting that research is incomplete. While the existing research does not provide a clear answer of the ef- fects of oxytocin, care providers may want to be observant for breastfeeding challenges among women and newborns who received oxytocin. Including oxytocin exposure as part of a risk assessment for suboptimal breastfeeding may allow for early intervention.
CONCLUSION
This article is the first known review of literature report- ing synthetic oxytocin administered during childbirth on breastfeeding outcomes. We used a comprehensive and in- tegrative approach including data from studies examining other research questions. This strength, combined with in- clusion of multiple breastfeeding outcomes (maternal and infant), adds needed complexity to the discussion of rou- tine birth interventions and our knowledge about any lasting consequences.
Since oxytocin was first used clinically in the early 1900s,65 research has inadequately addressed the possibility of an impact on the human breastfeeding relationship. As lac- tation is an oxytocin-dependent process, the role of oxytocin administered during birth is worth consideringwhen examin- ing suboptimal breastfeeding outcomes.Women’s perceptions of inadequate milk supply are a leading cause of supplemen- tation or discontinuation of breastfeeding. These perceptions deserve validation by clinicians and researchers by examin- ing the issue through a holistic lens that includes physiologic foundations to this problem.
AUTHORS
Elise N. Erickson, CNM, MS, is a PhD student and adjunct clinical instructor at Oregon Health & Science University.
Cathy L. Emeis, CNM, PhD, is an Assistant Professor and the program director of the Nurse-Midwifery Education Pro-
gram and the Clinical Department Chair of Nurse-Midwifery at Oregon Health & Science University.
CONFLICT OF INTEREST
The authors have no conflicts of interest to disclose.
REFERENCES
1.Centers for Disease Control and Prevention. Breastfeeding Report
Card, 2016. 2016:1-8.
2.Kent JC, Gardner H, Geddes DT. Breastmilk production in the first
4 weeks after birth of term infants. Nutrients. 2016;8(12):756.
3.Stuebe AM, Horton BJ, Chetwynd E, Watkins S, Grewen K,
Meltzer-Brody S. Prevalence and risk factors for early, undesiredwean-
ing attributed to lactation dysfunction. J Womens Health (Larchmt).
2014;23(5):404-412.
4.Chantry CJ, Dewey KG, Peerson JM, Wagner EA, Nommsen-Rivers
LA. In-hospital formula use increases early breastfeeding cessation
among first-time mothers intending to exclusively breastfeed. J Pedi-
atr. 2014;164(6):1339-1345.e5.
5.Semenic S, Loiselle C, Gottlieb L. Predictors of the duration of ex-
clusive breastfeeding among first-time mothers. Res Nurs Health.
2008;31(5):428-441.
6.Section on Breastfeeding. Breastfeeding and the use of human milk.
Pediatrics. 2012;129(3):e827-e841.
7.BartickMC, Schwarz EB, Green BD, et al. Suboptimal breastfeeding in
the United States: Maternal and pediatric health outcomes and costs.
Maternal &Child Nutrition. September 2016.
8.BartickM, Reinhold A. The burden of suboptimal breastfeeding in the
United States: a pediatric cost analysis. Pediatrics. 2010;125(5):e1048-
e1056.
9.Schwarz EB, McClure CK, Tepper PG, et al. Lactation and mater-
nal measures of subclinical cardiovascular disease. Obstet Gynecol.
2010;115(1):41-48.
10.Uvnäs-Moberg K, Prime D. Oxytocin effects in mothers and infants
during breastfeeding. Infant. 2013;9(6):201-206.
11.Crowley WR. Neuroendocrine regulation of lactation and milk pro-
duction. Compr Physiol. 2015;5(1):255-291.
12.Hammock EAD. Developmental perspectives on oxytocin and vaso-
pressin. Neuropsychopharmacology. 2015;40(1):24-42.
13.Nephew B, Murgatroyd C. The role of maternal care in shaping CNS
function. Neuropeptides. 2013;47(6):371-378.
14.Martin JA, Hamilton BE, Ventura SJ, et al. Births: final data for 2009.
Natl Vital Stat Rep. 2011;60(1):1-70.
15.World Health Organization. WHO Recommendations for the Pre-
vention and Treatment of Postpartum Haemorrhage. Geneva,
Switzerland: World Health Organization; 2012:1-48.
16.KenkelWM,Yee JR, Carter CS. Is oxytocin amaternal-foetal signalling
molecule at birth? Implications for development. J Neuroendocrinol.
2014;26(10):739-749.
17.Dublin S, Johnson KE,Walker RL, et al. Trends in elective labor induc-
tion for six United States health plans, 2001-2007. J Womens Health
(Larchmt). 2014;23(11):904-911.
18.Osterman MJ. National Vital Statistics Reports. January 2015;
64(1):1-68.
19.Laughon SK, Zhang J, Grewal J, Sundaram R, Beaver J, Reddy UM.
Induction of labor in a contemporary obstetric cohort. Am J Obstet
Gynecol. 2012;206(6):486.e1-486.e9.
20.Freeman RK, NageotteM. A protocol for use of oxytocin.Am J Obstet
Gynecol. 2007;197(5):445-446.
21.Osterman MJ, Martin JA. National Vital Statistics Reports. March
2011;59(5):1-14.
22.Carter C. Developmental consequences of oxytocin. Physiol Behav.
2003;79(3):383-397.
23.Odent MR. Synthetic oxytocin and breastfeeding: Reasons for testing
an hypothesis.Med Hypotheses. 2013;81(5):889-891.
Journal of Midwifery &Women’s Health � www.jmwh.org 415
24.Whittemore R, Knafl K. The integrative review: updatedmethodology.
J Adv Nurs. 2005;52(5):546-553.
25.Dozier AM,HowardCR, Brownell EA, et al. Labor epidural anesthesia,
obstetric factors and breastfeeding cessation. Matern Child Health J.
2012;17(4):689-698.
26.Gu V, Feeley N, Gold I, et al. Intrapartum synthetic oxytocin and
its effects on maternal well-being at 2 months postpartum. Birth.
2016;43(1):28-35.
27.Jonas W, Johansson LM, Nissen E, Ejdeback M, Ransjö-Arvidson AB,
Uvnas-Moberg K. Effects of intrapartum oxytocin administration and
epidural analgesia on the concentration of plasma oxytocin and pro-
lactin, in response to suckling during the second day postpartum.
Breastfeed Med. 2009;4(2):71-82.
28.Jordan S, Emery S, Watkins A, Evans JD, Storey M, Morgan G.
Associations of drugs routinely given in labour with breastfeeding
at 48 hours: analysis of the Cardiff Births Survey. BJOG: An In-
ternational Journal of Obstetrics & Gynaecology. 2009;116(12):
1622-1632.
29.Maŕın-Gabriel MA, Olza-Fernández I, Malalana-Mart́ınez AM, et al.
Intrapartum synthetic oxytocin reduce the expression of primi-
tive reflexes associated with breastfeeding. Breastfeed Med. March
2015:10(4):150318121046004.
30.Garćıa-Fortea P, González-Mesa E, Blasco M, Cazorla O, Delgado-
Rı́os M, González-Valenzuela MJ. Oxytocin administered during la-
bor and breast-feeding: a retrospective cohort study. J Matern Fetal
Neonatal Med. 2014;27(15):1598-1603.
31.Olza-Fernández I,MaŕınGabrielM,MalalanaMart́ınezA, Fernández-
Cañadas Morillo A, López-Sánchez F, Costarelli V. Newborn feeding
behaviour depressed by intrapartum oxytocin: a pilot study. Acta Pae-
diatrica. 2012;101(7):749-754.
32.Guerra GV, Cecatti JG, Souza JP, et al. Factors and outcomes associ-
ated with the induction of labour in Latin America. BJOG: An Inter-
national Journal of Obstetrics & Gynaecology. 2009;116(13):1762-
1772.
33.Guerra GV, Cecatti JG, Souza JP, et al. Elective induction versus
spontaneous labour in Latin America. Bull World Health Organ.
2011;89(9):657-665.
34.Dewey KG, Nommsen-Rivers LA, Heinig MJ, Cohen RJ. Risk factors
for suboptimal infant breastfeeding behavior, delayed onset of lacta-
tion, and excess neonatal weight loss. Pediatrics. 2003;112(3 Pt 1):
607-619.
35.Nommsen-Rivers LA, Chantry CJ, Peerson JM, Cohen RJ, Dewey KG.
Delayed onset of lactogenesis among first-time mothers is related to
maternal obesity and factors associated with ineffective breastfeeding.
Am J Clin Nutr. 2010;92(3):574-584.
36.Vogel JP, Souza JP, Gülmezoglu AM. Patterns and outcomes of induc-
tion of labour in Africa and Asia: a secondary analysis of the WHO
Global Survey onMaternal and Neonatal Health. Bhutta ZA, ed. PLoS
ONE. 2013;8(6):e65612-11.
37.Wiklund I, NormanM, Uvnas-Moberg K, Ransjo-Arvidson A, Andolf
E. Epidural analgesia: breast-feeding success and related factors.Mid-
wifery. 2007;25(2):e31-e38.
38.Yudkin P, Frumar AM, Anderson AB, Turnbull AC. A retrospec-
tive study of induction of labour. Br J Obstet Gynaecol. 1979;86(4):
257-265.
39.PrendivilleWJ,Harding JE, ElbourneDR, StirratGM.TheBristol third
stage trial: active versus physiological management of third stage of
labour. BMJ. 1988;297(6659):1295-1300.
40.Ounsted MK, Hendrick AM, Mutch LM, Calder AA, Good FJ. In-
duction of labour by different methods in primiparous women I.
Some perinatal and postnatal problems. Early Hum Dev. 1978;2(3):
227-239.
41.Out JJ, VierhoutME,WallenburgHCS. Breast-feeding following spon-
taneous and induced labour. Eur J Obstet Gynecol Reprod Biol.
1988;29(4):275-279.
42.Brown A, Jordan S. Active management of the third stage of labor may
reduce breastfeeding duration due to pain and physical complications.
Breastfeed Med. 2014;9(10):494-502.
43.Rajan L. The impact of obstetric procedures and analgesia/anaesthesia
during labour and delivery on breast feeding. Midwifery.
1994;10(2):87-103.
44.Chapman DJ, Perez-Escamilla R. Identification of risk factors
for delayed onset of lactation. J Am Diet Assoc. 1999;99(4):
450-454.
45.Matias SL, Nommsen-Rivers LA, Creed-Kanashiro H, Dewey KG.
Risk factors for early lactation problems among Peruvian primiparous
mothers.Matern Child Nutr. 2009;6(2):1-14.
46.Kong MS, Bajorek B. Medications in pregnancy: impact on time
to lactogenesis after parturition. J Pharm Pract Res. 2008;38(3):
205-208.
47.Mauri PA, Contini NNG, Giliberti S, et al. Intrapartum epidu-
ral analgesia and onset of lactation: a prospective study in an
Italian birth centre. Matern Child Health J. 2015;19(3):511-
518.
48.Bell AF,White-Traut R, RankinK. Fetal exposure to synthetic oxytocin
and the relationship with prefeeding cues within one hour postbirth.
Early Hum Dev. 2013;89(3):137-143.
49.Brimdyr K, Cadwell K, Widström A-M, et al. The association between
common labor drugs and suckling when skin-to-skin during the first
hour after birth. Birth. 2015;42(4):319-328.
50.Radzyminski S. The effect of ultra low dose epidural analgesia on
newborn breastfeeding behaviors. J Obstet Gynecol Neonatal Nurs.
2003;32(3):322-331.
51.Anim-Somuah M, Smyth RM, Jones L. Epidural versus non-epidural
or no analgesia in labour. Anim-Somuah M, ed. Cochrane Database
Syst Rev. 2011;(12):CD000331.
52.Worstell T, Ahsan AD, Cahill AG, Caughey AB. Length of the second
stage of labor. Obstet Gynecol. 2014;123:84S.
53.Rahm V, Hallgren A, Hogberg H, Hurtig I, Odlind V. Plasma oxy-
tocin levels in women during labor with or without epidural analgesia:
a prospective study. Acta Obstet Gynecol Scand. 2002;81(11):1033-
1039.
54.Glantz JC. Elective induction vs. spontaneous labor associations and
outcomes. J Reprod Med. 2005;50(4):235-240.
55.Henderson J, RedshawM.Women’s experience of induction of labor: a
mixedmethods study.ActaObstet Gynecol Scand. 2013;92(10):1159-
1167.
56.Lind JN, Perrine CG, Li R. Relationship between use of labor pain
medications and delayed onset of lactation. J Hum Lact. 2014;30(2):
167-173.
57.Phaneuf S, Rodriguez Linares B, TambyRaja R, Mackenzie I,
Lopez Bernal A. Loss of myometrial oxytocin receptors during
oxytocin-induced and oxytocin-augmented labour. J Reprod Fertil.
2000;120(1):91-97.
58.Balki M, Erik-Soussi M, Kingdom J, Carvalho JCA. Oxy-
tocin pretreatment attenuates oxytocin-induced contractions
in human myometrium in vitro. Anesthesiology. 2013;119(3):
552-561.
59.Brownell E, Howard CR, Lawrence RA, Dozier AM.Delayed onset lac-
togenesis II predicts the cessation of any or exclusive breastfeeding.
J Pediatr. 2012;161(4):608-614.
60.Chapman DJ, Perez-Escamilla R. Maternal perception of the onset of
lactation is a valid, public health indicator of lactogenesis stage II.
J Nutr. 2000;130(12):2972-2980.
61.Bales K, van Westerhuyzen J, Lewis-Reese A, Grotte N, Lanter J,
Carter C.Oxytocin has dose-dependent developmental effects on pair-
bonding and alloparental care in female prairie voles. Horm Behav.
2007;52(2):274-279.
62.Hashemi F, Tekes K, Laufer R, Szegi P, Tothfalusi L, Csaba G. Ef-
fect of a single neonatal oxytocin treatment (hormonal imprint-
ing) on the biogenic amine level of the adult rat brain: could
oxytocin-induced labor cause pervasive developmental diseases? Re-
prod Sci. 2013;20(10):1255-1263.
63.Bai DL, Wu KM, Tarrant M. Association between intrapartum inter-
ventions and breastfeeding duration. J Midwifery Womens Health.
2013;58(1):25-32.
416 Volume 62, No. 4, July/August 2017
64.Freeman SM, Samineni S, Allen PC, et al. Plasma and CSF
oxytocin levels after intranasal and intravenous oxytocin in
awake macaques. Psychoneuroendocrinology. 2016;66:185-
194.
65.Holmes JM. The use of continuous intravenous oxytocin in obstetrics.
Lancet. 1954;267(6850):1191-1193.
Continuing education units (CEUs) are available for this article as a part of a collection. To obtain CEUs online, please visit www.jmwhce.org. A CEU form that can be mailed or faxed is available in the print edition of this issue.
Journal of Midwifery &Women’s Health � www.jmwh.org 417
