Article Text

Download PDFPDF
Early lactation failure and formula adoption after elective caesarean delivery: cohort study
  1. Vincenzo Zanardo1,
  2. Anna Pigozzo2,
  3. Gary Wainer3,
  4. Diego Marchesoni4,
  5. Antonella Gasparoni5,
  6. Sandra Di Fabio6,
  7. Francesco Cavallin2,
  8. Arturo Giustardi2,
  9. Daniele Trevisanuto2
  1. 1Pediatric Department, Padua University, Padua, Italy
  2. 2School of Medicine, Padua University, Padua, Italy
  3. 3Neonatal-Perinatal Medicine, St. Joseph Mercy Hospital Ann Arbor, Ann Arbor, Michigan, Houston, USA
  4. 4Institute of Obstetrics and Gynecology and Reproductive Sciences, University of Udine, Udine, Italy
  5. 5Department of Pediatrics, University of Brescia, Brescia, Italy
  6. 6Department of Pediatrics, University of L'Aquila, L'Aquila, Italy
  1. Correspondence to Vincenzo Zanardo, Padua University, Pediatric Department, Via Giustiniani 3 Padua, 35128, Italy; zanardo{at}


Objective To investigate the effects of elective primary and elective repeat caesarean deliveries on lactation at hospital discharge.

Design Cohort study.

Setting Four Italian teaching hospitals – Padua, Brescia, L'Aquila and Udine.

Interventions Deliveries were classified as vaginal, elective caesarean (primary and repeat) or emergency caesarean. A total of 2296 (24.7%) infants born by caesarean section (CS), 816 of which (35.5%) classified as primary elective CS and 796 (34.7%) as repeat elective CS, were studied. Moreover, 30.2% of the elective CS deliveries took place before 39 weeks.

Main outcome measures Feeding modalities at discharge: formula, complementary and breastfeeding.

Results At discharge, 6.9% of the vaginal delivery mothers, 8.3% of the emergency CS mothers, 18.6% of the elective CS mothers, 23.3% of the primary CS mothers and 13.9% of the repeat CS mothers were using infant formula exclusively. Multivariate analysis (OR; 95% CI) identified primary elective delivery (3.74; 3.0 to 4.60), lower gestational age (1.16; 1.10 to 1.23), and place L'Aquila versus Udine (1.42; 1.01 to 2.09) and of Brescia versus Udine hospitals (6.16; 4.53 to 8.37) as independent predictors of formula feeding at discharge.

Conclusions These findings provide new information about the risks of breastfeeding failure connected to elective CS delivery, particularly if primary and scheduled before 39 weeks of gestation.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic:

  • Few studies have evaluated the effect of elective caesarean delivery on subsequent breastfeeding.

  • This is relevant because there has been an increase in the rate of elective caesarean deliveries over the past 30 years.

What this study adds:

  • This study shows that elective primary caesarean delivery does affect short-term breastfeeding success in about 1:4 of mother-infant pairs.


The benefits of breastfeeding are well established for the mother and infant.1 The WHO has published strategies to promote the initiation of breastfeeding.2 Despite the widely documented health benefits of maternal milk, breastfeeding rates in many developed countries remain consistently lower than the American Academy of Pediatrics has recommended.3 It is important to identify factors that may interfere with the initiation of breastfeeding during the newborn's initial hospital stay.4 These factors may not be the same as those that ultimately determine the duration of breastfeeding.5

Previous research has led to conflicting results when evaluating the association between the mode of delivery and the initiation of breastfeeding. Some studies have reported that caesarean delivery has a negative effect on breastfeeding while others have shown no association between operative delivery and the initiation of breastfeeding.6 ,7 Fewer studies have evaluated the effect of elective caesarean delivery on subsequent breastfeeding and, to our knowledge, none have directly compared the different types of elective caesarean delivery, namely elective primary and elective repeat caesarean delivery.8 This is relevant because there has been an increase in the rate of caesarean deliveries over the past 30 years.9 In particular, there has been an increase in the incidence of elective primary caesarean deliveries at term. In many Italian hospitals, the caesarean delivery rate is greater than 40%, which is higher than the WHO guidelines (5–15%) for acceptable caesarean delivery rates.10 ,11 Primary elective caesarean delivery is not unusual, and there is a risk that it may be performed earlier than is appropriate.12,,14 Recent surveys of obstetricians reveal that a substantial minority of physicians would choose that mode of delivery for themselves or their spouses and that a higher percentage would honour a patient's request for an elective caesarean delivery.15 Population-based evidence evaluating an association between primary elective caesarean delivery and breastfeeding initiation is important for policy-makers. If certain modes of delivery hinder successful breastfeeding, health providers could target women who have experienced such deliveries for extra support during their hospital stay.

We undertook this study of a large assembled cohort of women in Northern and Central Italy to investigate the effects of emergency, elective primary and elective repeat caesarean deliveries on the likelihood of breastfeeding at hospital discharge.

Materials and methods

This observational cohort study was undertaken between January and December 2009 in four Italian teaching hospitals (Padua, Brescia, L'Aquila, Udine) covering a geographical area with a stable population and over 10 000 deliveries annually. All infants who were born at term gestation (±37 0/7 weeks of gestation by best obstetrical dates) and admitted to the normal postpartum unit were eligible for inclusion. Newborns admitted to the Neonatal Intensive Care Unit (NICU) were excluded.

Written consent was obtained from all study participants and the local ethics committees granted approval.

Hospital records for mothers and newborns were examined from the time of admission until the time of discharge. Discharge was selected as the cut-off to maintain the focus on the initiation of breastfeeding rather than the ultimate duration of breastfeeding. Deliveries were classified as vaginal, elective primary caesarean, elective repeat caesarean, or emergency caesarean. Operative deliveries were classified as ‘elective’ if they were performed before the onset of labour and ‘emergency’ if they were performed after labour began.12 Labour was defined as regular uterine contractions resulting in cervical dilatation. The method of feeding at discharge was classified as either (1) exclusive breastfeeding (breastfeeding), (2) breastfeeding with formula supplementation (complementary), or (3) exclusive formula feeding (formula) according to the WHO classification.2

Within all four hospitals, it was standard practice to optimise immediate skin-to-skin contact, offer to initiate breastfeeding in the delivery room, encourage mothers to ‘room-in’ with their newborn, provide information about the advantages of breastfeeding and give practical instruction on how to initiate breastfeeding in the postnatal ward.

The vaginal delivery group was considered to be the reference group, since this delivery route would be primarily considered for most women, and the caesarean section group, including emergency, elective primary or iterative deliveries, was the comparison group.

Maternal indications for elective caesarean delivery were repeat caesarean section or other cause of uterine scarring (ie, antecedent myomectomy), medical conditions (severe myopia) or contraindications to vaginal delivery, such as malpresentation, placenta praevia, multiple gestation, macrosomia, tocophobia and patient choice.

The main indications for emergency caesarean delivery were fetal distress, dystocia, pre-eclampsia and clinical chorioamnionitis.

The caesarean section (CS) mothers started drinking water about 6 h after surgery, where then placed on a liquid diet, followed by a soft diet for three subsequent meals and thereafter were given a regular hospital diet. Intravenous fluids were stopped when the liquid diet was well tolerated.

The four groups were compared for infant feeding outcomes at discharge. Categorical data were expressed as number and percentage, whereas continuous data as mean and SD. Categorical data were compared among groups using Fisher's test and continuous data using analysis of variance. A logistic regression model was estimated to identify the effect of delivery on formula feeding at discharge, adjusting for potential confounding factors identified by univariate analysis (maternal age, gestational age, birth weight, residential hospital). A p value less than 0.05 was considered significant. Statistical analysis was performed using R 2.12 software (R Development Core Team 2010. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria).


Our review identified 9446 potentially eligible maternal-infant pairs. One hundred and thirty-five pairs were excluded because the infant was admitted to the NICU. Delivery data were also missing from five eligible subjects. The final study population included 9306 maternal-infant pairs who gave birth at term over the 12-month study period.

A total of 7010 (75.3%) infants were born by vaginal delivery and 2296 (24.7%) were born by caesarean delivery. Of the caesarean deliveries, 685 (29.8%) were classified as emergency, 816 (8.8%) as primary elective and 796 (8.5%) as repeat elective (table 1). Mothers in the vaginal delivery group were younger (32.0±5.2 years) when compared with the elective primary caesarean group (33.7±5.3 years) (p=0.01). There were no differences in maternal age between any of the caesarean delivery subgroups. Gestational age at delivery was similar between the two elective caesarean delivery groups; however, the gestational age for infants in these groups was lower when compared with either vaginal delivery or emergency caesarean delivery (p<0.0001). Birth weight was similar between all study groups.

Table 1

Mother-infant characteristics and feeding method for the total sample at discharge by delivery mode

At the time of discharge, a total of 7713 (83.7%) infants were exclusively breast feeding, 669 (7.3%) were complementary feeding, and 836 (9%) were formula feeding (table 1). When stratified by delivery method, the proportion of infants exclusively breastfeeding at hospital discharge was 85.4% (vaginal delivery), 81.6% (emergency caesarean), 71.6% (primary elective caesarean) and 83.8% (repeat elective caesarean). The proportion of infants exclusively formula feeding was 6.8% (vaginal delivery), 8.3% (emergency caesarean), 23.3% (primary elective caesarean) and 13.9% (repeat elective caesarean). The highest proportion of formula feeding was found among those infants delivered by primary elective caesarean (23.3%; p<0.0001).

Maternal-infant characteristics and discharge feeding practices for each participating hospital are described in table 2. Differences were noted in subject demographics, the frequency of caesarean deliveries, and discharge feeding methods. Maternal age was higher at Padua (p<0.0001) and lower at Brescia (p<0.0001). Gestational age (p<0.0001) and neonatal birth weight (<0.0001) were higher at L'Aquila and lower at Brescia (p<0.0001). Vaginal deliveries were more frequent at L'Aquila (p<0.0001), while caesarean deliveries were more common at Udine (p<0.0001). Exclusive breastfeeding was more common at L'Aquila (<0.0001) and lower at Brescia (<0.0001), while formula feeding was more common at Brescia (<0.0001) and less common at Udine (<0.0001).

Table 2

Mother-infant characteristics, delivery mode and feeding method by delivery hospital

Univariate analyses (table 3) showed significant associations between formula feeding at discharge and the following variables: delivery mode (p<0.0001), delivery hospital (p<0.0001), higher maternal age (0.01), lower gestational age (p<0.0001) and neonatal birth weight (p<0.0001).

Table 3

Feeding modalities at discharge: univariate analysis

Multivariate logistic regression analysis (table 4), controlling for these variables, identified the following independent predictors of formula feeding at discharge (OR; 95% CI): primary elective caesarean delivery (3.74; 3.03 to 4.60), lower gestational age (1.16; 1.10 to 1.23), delivery at L'Aquila versus Udine hospital (1.42; 1.01 to 2.09) and delivery at Brescia versus Udine hospital (6.16; 4.53 to 8.37).

Table 4

Multivariate logistic regression: independent predictors of formula feeding at discharge


This study provides, for the first time, new information about the risks of breastfeeding failure connected to elective CS delivery, particularly if primary and scheduled early.

Rapid, recent changes in modes of delivery, postnatal practices, parental expectations and feeding options pose new important challenges for the mother-infant dyad. A number of factors seem to affect breastfeeding initiation (the use of some anaesthesias during labour, initiating breastfeeding in the labour room, skin-to-skin contact after birth, rooming-in arrangements).16 But those are not the only considerations; according to Donath et al, mothers' intention to breastfeed is a stronger predictor of breastfeeding initiation and duration than all standard demographic factors combined.17 And the intention to breastfeed made even before pregnancy has begun may be even more predictive.18

Some investigators have reported that CS is a risk factor for not initiating and discontinuing breastfeeding in sample populations.6 ,19 ,20 And we have recently hypothesised that elective CS has a negative effect on breastfeeding initiation and on the length of time mothers nursed.8 This study confirmed that elective primary CS does indeed affect short-term breastfeeding success in about 1:4 of mother-infant pairs.

The impact of surgical birthing practices on breastfeeding is difficult to assess. Emergency CS may follow a long, difficult labour and be associated to numerous factors such as confinement to bed, fasting, analgesia or anaesthetics or both for pain, oxytocin augmentation and anxiety and stress, all having a negative impact on breastfeeding: even following a planned caesarean section, these interventions can be frightening, especially in first-time mothers. And the mother who has undergone a CS and is lacking support and assistance is unable to hold her newborn baby for the frequent periods necessary for breastfeeding. Bottle feeding has become a common clinical practice in these cases.4 ,5 Bottle-feeding milk-based formulas will, however, reduce newborn sucking capacity as well as mother's lactation stimulus.6 ,16 Finally, according to Lagercrantz and Slotkin, it is actually important for infants to undergo the hormonal and physical stress of being born which seems to prepare them for bonding and feeding.14 This is relevant, considering that elective caesarean delivery is frequently performed before 39 weeks.13

There are also many factors affecting a pregnant woman's decision to choose to have a CS and then to breastfeed (social conditions, her or others' personal experiences, economic and family considerations, religious or ethnic beliefs).21 At times mothers are uninformed about the benefits and advantages of breastfeeding, while others are unprepared for the initial difficulties that they may face.

Most studies, including our own, did not distinguish between the experience of never trying to breastfeeding and that of trying but not succeeding or between breastfeeding initiation and duration. Some investigators favour multifaceted programmes to promote and encourage breastfeeding initiation such as that implemented by WHO/UNICEF named ‘The Ten Steps to Successful Breastfeeding’, which is part of the Baby-Friendly Hospital Initiative.22 Unfortunately, the background and training of nurses and midwifes working in those hospitals are not standardised, and this hinders any attempts to interpret and assess breastfeeding differences presented. Moreover, we did not determine the role of fatigue, stress, pain, length of labour and health complications in emergency CS.

Our findings can probably be considered representative of a developed country's situation in which CS deliveries can be and are routinely programmed and most women can choose not to breastfeed, while low-income sectors of the same population are less likely to breastfeed. As stated in the Cochrane Database of Systematic Reviews 2005 concerning interventions to promote breastfeeding: ‘initiation rates remain relatively low in many high-income countries, particularly among women in lower income groups’.23 Additional studies in high-income and low-income populations are of course warranted.

Kruse et al. who attempted to isolate the component of hospital variation in rates of exclusive breastfeeding at discharge dependent on demographic composition found that sociodemographic variables predict about 60% of the variation in hospital-specific rates.24 All potential confounding factors have probably not been identified or accounted for within a cohort design.7 ,19 ,25 ,26

In conclusion, these data provide more information for obstetricians about the effect of elective CS delivery, particularly if primary and scheduled early, on short-term breastfeeding outcome. More information is needed to identify mothers at risk for lactation failure and studies are of course warranted to investigate how hospital or community programmes promoting breastfeeding activities and counselling can turn the tides to more mothers breastfeeding during their hospital stay.


View Abstract


  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval The local ethics committees (University of Padua, Brescia, Udine, and L'Aquila granted approval.

  • Provenance and peer review Not commissioned; externally peer reviewed.