A Comparison of Factors Associated with Cessation of Exclusive Breastfeeding at 3 and 6 Months

Introduction

Exclusive breastfeeding (EBF) during the first 6 months of life provides many benefits to both mother and child. Breast milk promotes sensory and cognitive development and protects the infant against infectious and chronic diseases. It is also an important source of energy and nutrients to infants between 6 and 23 months. ¹

The World Health Organization (WHO) recommends EBF during the first 6 months of life, the introduction of food that is safe and appropriate after that age, and to maintain breastfeeding until 2 years or more. ²

No official monitoring system for breastfeeding exists in Spain; however, several reports published in the last decade indicate that the frequency and duration of breastfeeding do not reach the levels recommended by the WHO.^3,4

According to the last Spanish Health Survey conducted by the National Statistical Institute between 2011 and 2012, the percentage of breastfeeding is 51% at 3 months of life and 26% at 6 months of life. ⁵ A study carried out by Rius et al. in 2005, in a region in eastern Spain, showed that the maintenance of breastfeeding at 3 and 6 months of age was 39% and 21%, respectively. ⁴ In other European countries, EBF rates at 6 months are 34% (Portugal), 17.2% (Denmark), 14% (Sweden), and 5% (Italy). ⁶

To improve EBF rates in Spain, we sought to identify factors that might lead to cessation of breastfeeding and the introduction of formula feeding. Previous studies have shown that some factors related to labor may affect breastfeeding, including administration of synthetic oxytocin, mode of delivery, and epidural analgesia.^7,8 Other studies have found correlations between the abandonment of breastfeeding and several sociodemographic factors, such as maternal age, no college degree, primiparity, gestational age, absence of prenatal education, gender of the newborn, pacifier use, not having attended a breastfeeding support group, and reintegration into the workplace.^9,10

The aim of the current study is to evaluate the association of labor and sociodemographic factors with the cessation of EBF at 3 and 6 months of life.

Materials and Methods

A prospective cohort study (n = 529) was carried out in a tertiary hospital in Madrid, Spain, with the Baby-Friendly Hospital Initiative (BFHI) award and with more than 3,000 deliveries per year.

Inclusion criteria were single full-term gestation, Apgar score >7 at 5 minutes, and prenatal intention or wish to breastfeed.

Exclusion criteria were the following: preterm delivery, fetal chromosomopathies and other anomalies that could affect adaptation to the extrauterine environment, admission of the mother or the newborn to the intensive care unit in the first 48 hours of life, prenatal intention or desire to administer formula feeding, general anesthesia, and having an insufficient understanding of Spanish. Only two mothers (0.3%) refused to participate in the study.

The Ethics and Clinical Research Committee of the Hospital Universitario Puerta de Hierro-Majadahonda approved the study, and informed consent was obtained from all participants before delivery.

The following labor factors were investigated: administration of synthetic oxytocin, mode of delivery, epidural anesthesia, gestational age, and birth weight. In addition, the following sociodemographic factors were recorded at delivery: maternal age, mother's nationality, educational attainment (college degree or not), marital status, parity (primiparous versus multiparous), and prenatal education (maternal education program supported by midwives in primary care). All data were registered by midwives.

For deliveries with a spontaneous onset, exogenous oxytocin was not administered unless the uterine dynamic was poor. Labor was induced with intravenous oxytocin when there was a medical indication to intervene.

To administer oxytocin, a vial of 10 UI (Syntocinon^®; Sigma Tau) was diluted in 500 mL of saline solution. It was then given to the patient at a rate of 1 mUI/min, and the dose was doubled every 20 minutes until achieving at least three contractions every 10 minutes. The maximum recommended dose was 32 mUI/min (Cardiff method ¹¹ ).

Active management measures were applied in the third phase of labor. After the delivery of the anterior fetal shoulder, 5 UI of oxytocin was administered through a venous catheter. Oxytocin administration was maintained at a rate of 2 UI/h for the first 2 hours of the postpartum period.

In the case of cesarean section, 3 UI of oxytocin was administered through a venous catheter after fetal expulsion, and an additional 10 UI of oxytocin was administered during the first hour of the postpartum period.

When the patient requested epidural anesthesia, an epidural catheter was inserted after receiving informed consent. Anesthesia was induced with either 0.125% levobupivacaine (Normon) plus fentanyl or 0.2% ropivacaine (Inibsa) plus fentanyl.

In the case of vaginal deliveries, when the newborns did not need cardiopulmonary resuscitation, they were dried, identified, and placed in skin-to-skin contact with their mothers for 90 minutes.

After this period, routine measurements were performed (weight, size, cephalic perimeter, eye prophylaxis, and antihemorrhagic prophylaxis). If the newborn had not latched onto the breast spontaneously, both mother and child were oriented to begin breastfeeding.

In the case of cesarean deliveries, when maternal and/or fetal conditions permitted, skin-to-skin contact took place during the surgical procedure and afterward at the postoperative care unit. If skin-to-skin contact was not possible, such contact took place after surgery at the postoperative care unit (in the first 60 minutes of life). Under no circumstances was the newborn placed in a separate room from the mother.

During the stay in the hospital, the nursing staff offered education in breastfeeding, encouraging proper positioning of the newborn to latch onto the mother's breast and answering questions about the process.

Likewise, after discharge from hospital, mothers were able to call a nursing advice line 24 hours a day, 7 days a week, for breastfeeding consultation.

To track the duration of breastfeeding, the use of a pacifier at any time, attendance at a breastfeeding support group, and reintegration into the workplace, telephone calls were made at 3 and 6 months of life.

The study considered EBF at 3 months of life, when the infants had been fed only with mother's own milk, and at 6 months, when they had been fed with mother's own milk with or without the introduction of solid foods (fruits, vegetables, etc.). Partial breastfeeding was considered when the infant was receiving breast milk and formula.

A loss of follow-up was considered after five unsuccessful telephone attempts. After 3 months, 6 participants (1.1%) were considered as dropouts and 17 (3.2%) after 6 months.

Descriptive analyses were performed through absolute and relative frequencies for categorical variables and mean (standard deviation) or median (percentiles 25 and 75) for numerical variables.

Univariate analysis was performed with the chi-square test or with Fisher's exact test for categorical variables.

To identify risk factors associated with the cessation of breastfeeding at 3 and 6 months postpartum, two multivariate logistic regression models were developed. All significant variables in the univariate analysis were entered into the models. The maximum model included cesarean delivery, the absence of a college degree, not having attended a breastfeeding support group, pacifier use, and reintegration into the workplace. A nonautomatic backward procedure was followed to reach the final model according to the parsimonious principle (criteria to eliminate variables with p > 0.05). Goodness of fit was measured by assessing calibration and discrimination models; the Hosmer–Lemeshow test ¹² was used to assess calibration and the area under the receiver operating curve was used to assess discrimination.

The statistical significance level was established at 0.05. Statistical analysis was carried out using Stata/IC v.14.1. (StataCorp LP, College Station, TX).^13–15

Results

The sample characteristics are described in Table 1. The proportion of infants who initiated EBF was 90.4%. At 3 months of life, 64.4% of mothers maintained EBF and this fell to 31.4% at 6 months of life.

Univariate analyses of labor and sociodemographic factors that could be associated with EBF cessation at 3 and 6 months of life are shown in Tables 2 and 3, respectively. Factors influencing the cessation of EBF at 3 months were cesarean delivery, the absence of a college degree, the use of a pacifier, not having attended breastfeeding support groups, and reintegration into the workplace. Factors influencing the cessation of EBF at 6 months were being primiparous, the use of a pacifier, and reintegration into the workplace.

The final adjusted model is shown in Table 4. Cesarean delivery, the absence of a college degree, not having attended a breastfeeding support group, and pacifier use were all associated with cessation of EBF at 3 months. Calibration and discrimination were good, with a p-value in the Hosmer–Lemeshow test of 0.778 and the area under the receiver operator curve of 0.695.

The candidate variables to predict EBF cessation at 6 months were primiparity, reintegration into the workplace, and the use of a pacifier. All these variables remained significant in the final model (Table 4). Calibration and discrimination were good, with a p-value in the Hosmer–Lemeshow test of 0.998 and an area under the receiver-operator curve of 0.748.

Discussion

The present study shows that some labor and sociodemographic factors are associated with premature cessation of EBF; specifically, cesarean delivery, use of a pacifier, not having a college degree, not having attended a breastfeeding support group, being primiparous, and reintegration into the workplace.

We noted that EBF at 3 months was lower with cesarean delivery than with vaginal delivery (67.6% and 51.5%, respectively). These data agree with the results published by Oves Suárez et al., ¹⁶ who found an association between vaginal delivery and a higher probability to maintain breastfeeding after 4 months (83.3% versus 16.7% for cesarean delivery). Additionally, a systematic review by Prior et al., ¹⁷ which included 53 studies, and a meta-analysis, including 48 studies, concluded that elective cesarean section was related to low rates of breastfeeding.

Regarding the administration of synthetic oxytocin, a previous study suggests that an active management of the third phase of labor may reduce the rates of breastfeeding at 2 and 6 weeks. ¹⁸ Likewise, in a study investigating the relationship between the administration of oxytocin during the first and second phases of labor and initiation and duration of breastfeeding, García-Fortea et al. ¹⁹ found a negative association in both the initiation and continuation of breastfeeding at 3 months.

The use of oxytocin during labor and its consequences for breastfeeding is our main field of research. In a preliminary study, ²⁰ we observed that the dose of oxytocin used in mothers who did not maintain EBF at 1 and 3 months of life was higher than that used in mothers who did maintain EBF. However, in a subsequent investigation, ²¹ we failed to find any relationship between the administration of synthetic oxytocin during labor and the preservation of EBF at 3 and 6 months of life. Our findings in the present study are in agreement with these latter findings.

Epidural analgesia is another labor factor that has been related to breastfeeding. In a systematic review conducted in 2015, which included 23 experimental studies, the authors found that half of the studies reported a negative association between epidural analgesia and breastfeeding. ²² They concluded that more research was needed to obtain meaningful conclusions. In the present work, we found that the rates of EBF at 3 and 6 months were higher in those mothers who did not receive epidural analgesia, but this did not reach statistical significance.

Among the sociodemographic factors related to cessation of breastfeeding, pacifier use has been widely investigated and has been linked to a shorter duration and exclusivity of breastfeeding. After reviewing the evidence, the Breastfeeding Committee of the Spanish Pediatric Association recommended against using a pacifier until breastfeeding was established, usually until after the first month of life. ²³ According to Jaafar et al., ²⁴ the use of a pacifier once breastfeeding has commenced does not affect its prevalence at 4 months of life; however, the authors concluded that not enough information was available about the potential negative consequences of pacifier use in newborns. In the present study, we found a negative relationship between pacifier use and breastfeeding at 3 and 6 months of life. Accordingly, at 3 months, 81.6% of infants who did not use a pacifier maintained EBF compared with 57.7% for those who did. Additionally, at 6 months, 51.1% of infants who did not use a pacifier maintained EBF against 24% who did.

Another factor to consider is the level of education of the mother. According to a study conducted by Machado et al., ²⁵ mothers with a lower educational level withdrew from EBF more frequently 4 months after labor. Similarly, Jessri et al. ²⁶ studied EBF indicators and found that mothers with a college degree and multiple children had higher rates of EBF after 6 months. Therefore, primiparity might be negatively associated with breastfeeding. The results of our study relate the lack of a college degree to the cessation of EBF at 3 months of life and primiparity to the cessation of EBF at 6 months.

Regarding the attendance at support groups during breastfeeding, we observed that at 3 months of life, the percentage of EBF was 74.8% for mothers who attended a support group compared with 61% for mothers who did not. Similarly, Britton et al. ²⁷ reviewed 34 trials from 14 countries and concluded that the support offered by professionals could help to extend breastfeeding duration, especially during the first 2 months. They also added that personal support was more effective than support over the telephone.

Finally, we found that reintegration into the workplace is related to the duration and exclusivity of breastfeeding. Bai et al. ²⁸ performed a study with 1,738 mothers, finding that only one-third of them maintained breastfeeding after 2 weeks of having returned to the workplace. In our study, reintegration into the workplace was also related to cessation of EBF, both at 3 and 6 months of life; however, at 3 months, only 3.6% of mothers had reintegrated into the workplace, while at 6 months, it was 57.4% of the mothers.

This study has some limitations that must be considered when interpreting the results. The first limitation is one of sample selection bias since all participants expressed their desire to practice breastfeeding before labor and only due date gestations were included. However, we consider it relevant to establish which factors are associated with the abandonment of EBF when mothers were willing to breastfeed, given that they were motivated to begin or maintain breastfeeding. Furthermore, the research took place in a hospital with a BFHI award. These characteristics could minimize the impact of the studied factors over cessation of EBF and reduce its external validity. Additionally, mothers attending breastfeeding support groups may be more predisposed and motivated to breastfeed longer and therefore this may be the reason for the longer duration of breastfeeding in this group. There could be a recall bias that may limit the validity of the study as the information on the type of feeding was retrospectively collected at 3 and 6 months from birth. Nevertheless, there is evidence that this type of questionnaire has high reliability. ²⁹ Loss of follow-up could limit the research findings, but as these were minimal (1.1% and 3.2% in each of the periods studied) and the sample size was suitable, we believe that this did not modify the observed outcomes.

Another limitation to take into account is that EBF at 6 months was considered if the mothers' own milk was provided with or without solid foods. In addition, rates of EBF at 6 months may be influenced by the legislation in force in Spain (16 weeks of maternity leave). Finally, there may be other factors related to cessation of EBF that may have not been analyzed in the present study, such as pre-existing maternal health problems, lifestyle behavior, socioeconomic status, and postnatal depression, among others.

Conclusions

Several risk factors are related to the premature cessation of EBF: cesarean delivery, the use of a pacifier, the lack of a college degree, nonattendance at breastfeeding support groups, being primiparous, and reintegration into the workplace.

Based on these findings, it would be desirable to define strategies to correct the modifiable factors as well as to promote the protective factors with the aim of improving the success rate of maternal breastfeeding to reach the recommendations of the WHO.