Factors associated with breastfeeding continuation for 12 months or more among working mothers in a general hospital

This article aims to identify factors associated with breastfeeding continuation for at least 12 months among working mothers in a hospital in the south of Brazil. We conducted a cross-sectional study, interviewing 251 women who breastfed after returning to work. Eligibility criteria included non-twin biological children aged between 12 and 36 months and the absence of an illness (mother and/or child) that could affect breastfeeding. The association between breastfeeding continuation and the exposure variables was tested using Poisson multivariate regression. Only one work-related variable showed a significant association with the outcome. Working only during the day increased the prevalence of BF continuation for at least 12 months by 37%. The following non-work-related factors showed a positive association with the outcome: mothers without a college degree; mothers with at least 12 months’ prior breastfeeding experience; child not given milks other than breast milk when the mother returned to work, and not using a pacifier. The following variables showed a negative association with the outcome: older maternal age; older gestational age; mother receiving support from the child’s caregiver; and mother receiving professional breastfeeding support. Non-work-related factors had a greater influence on breastfeeding continuation for at least 12 months among working mothers.


Introduction
In view of the benefits of breastfeeding (BF) for the health of children and mothers [1][2][3] , the prevention of early breastfeeding cessation is essential, even when the mother expresses the wish to breast feed for longer.
One of the multiple reasons for early interruption of BF, particularly exclusive breastfeeding (EBF), is maternal employment [4][5][6] . In this regard, studies show that measures supporting the maintenance of breast feeding after women's return to work have a positive impact on BF continuation. These measures include flexible working hours, breastfeeding or milk expression breaks, appropriate facilities for expressing and storing milk, support from employers and work colleagues 7 , and child care near or at the workplace 6 .
A nationwide survey in Brazil 8 showed that 26.8% of working women with children under six months exclusively breastfed, compared to 53.4% of women on maternity leave and 43.9% of women who did not work. With regard to women with children under 12 months, 81.2% of women who did not work were breastfeeding, compared to 65% of working women, thus demonstrating the impact of maternal employment on BF rates. Support for breastfeeding at the work place has been shown to be an important factor influencing BF continuation after return to work. Despite the fact that maternal employment is an important risk factor forearly breastfeeding cessation, few studies on this topic have been published, especially in Brazil. Little is therefore known about the barriers and facilitators for BF continuation after return to work. Nevertheless, our under standing of the factors that influence breastfeeding by working mothers is very relevant to the planning of actions aimed at extending the duration of BF among this population. Therefore, the aim of this study was to identify factors associated with BF continuation for at least 12 months among working women in a large general hospital, with emphasis on work-related factors.

materials and methods
We conducted a cross-sectional study with 251 female staff of a large university hospital in Porto Alegre in the south of Brazil. The data were collected between June 2016 and July 2017.
The hospital has around 6,000 staff, most of whom are women of child bearing age. The hospital offers the following conditions and facilities for workers with infants: 180days of maternity leave; breastfeeding breaks of up to one hour per working day until the child's first birthday; crèche located close to the hospital; human milk bank (HMB), breastfeeding helpline; and trained lactation consultants in the neonatal inpatient unit.
The inclusion criteria for participation in the study were as follows: women with nontwin biological children aged between 12 and 36 months born while the mother was employed at the hospital; women breastfeeding after returning to work; absence of an illness (mother and/or child) that could significantly affect BF, such as orofacial malformations or neurological diseases in the infant.
Infant minimum age was set at 12 months because the outcome was defined as BF continuation for at least 12 months, while maximum age was defined as 36 months to minimize recall bias.
In the case of women with more than one child in the stipulated age group, the youngest child was selected.
The data were collected using a structured questionnaire prepared by the researchers and administered during pre-scheduled interviews held at the hospital. Each interview lasted an average of 30 minutes.
Sample size was calculated using an alpha of 5%, power of 80%, prevalence of BF for at least 12 months of 50% 9 , prevalence of the main independent variables of between 50 and 60% (based on a pilot study with seven women), and odds ratio (OR) of 2.0, resulting in a minimum sample of between 296 and 317, depending on the frequencies of the independent variables.
The data were entered twice and checked for mistakes. Statistical analysis was performed using SPSS version 23. The quantitative variables were described using means and standard deviations or medians and interquartile ranges, while the categorical variables were presented using absolute and relative frequencies.
Due to the large number of variables used to explain the outcome, we used a hierarchical regression model 10 in which the variables were grouped into blocks according to their proximity to the outcome 11 , forming three hierarchical levels: distal, intermediate and proximate. The levels and respective variables are shown in Figure 1.
First, we tested the association between the outcome and the variables in each block using univariate Poisson regression. The variables with a significance level of p<0.20 in the univariate analysis were included in the multivariate analysis in their respective blocks in the Poisson regression (intra-block analysis). The variables with a significance level of p <0.20 in the intra-block analysis remained in the model until the end of the analysis, adjusting for potential confounding factors. The strength of association was determined using crude and adjusted prevalence ratios (PR and PRa, respectively) and their respective 95% confidence intervals (95% CI), adopting a significance level of p<0.05. The re-search protocol was approved by the hospital's research ethics committee and registered on the Plataforma Brasil. All participants signed an informed consent form.

results
We contacted 300 women, 49 of whom were excluded because they were not breastfeeding after their return to work, resulting in a final sample of 251womenwho met the inclusion criteria. Table 1 shows the characteristics of the study population.
Median duration of EBF was 150 days and 95% of the children were given milk other than DISTAL LEVEL Mother's age, education level, skin color, parity, living with partner, living with mother, prior breastfeeding experience INTERMEDIATE LEVEL Place of antenatal care, intention regarding breastfeeding duration, intention regarding breastfeeding after return to work, place of delivery, type of delivery, gestational age, child's sex, exclusive breastfeeding at hospital discharge, breastfeeding in the first hour of life PROXIMATE LEVEL Not related to mother's work: Duration of EBF, use of milks other than breast milk when mother returned to work, breastfeeding support from partner, breastfeeding support from caregiver, breastfeeding support from health professionals, use of a pacifier, child's age when mother returned to work. Related to mother's work: woman's position/function in the hospital, work shift, work week, breastfeeding during work, milk expression during work, use of HMB, use of the services of a lactation consultant, breastfeeding support from colleagues, breastfeeding support from management, perception of difference in support between male and female colleagues, perception that work performance is negatively affected by breastfeeding, guidance on continuing breastfeeding after return to work provided by the organization, breastfeeding agreement with employer, Knowledge of the existence of a human milk bank in the hospital, knowledge of the existence of a lactation consultant in the hospital, knowledge of entitlement to take 1 hour off work each day for breastfeeding, Use of the hospital crèche, breastfeeding in the crèche, use of expressed milk in the crèche, feeling of work overload in conciliating breastfeeding and work, perception of difficulty of breastfeeding upon returning to work, degree of satisfaction with employer support for the continuation of breastfeeding. OUTCOME: Breastfeeding for at least 12 months breast milk when the mother returned to work, which occurred between 4 and 11 months (median of 6 months).
Most of the women reported that they intended to breastfeed for at least 12 months and practically all the sample said that they wanted to continue breastfeeding after returning to work; 62.3% of the children were breastfed for at least 12 months. Table 2 shows data related to return to work and breastfeeding. Most of the women worked with patient care and only during the day. The average work week was 35.5 hours. Approximately three-quarters of the women reported that they were not informed by the organization about matters relating to breastfeeding after return to work, and level of unfamiliarity with facilities and conditions for workers with infants varied according to each facility and condition.
Around one-third of the workers expressed breast milk during work and/or used the HMB. A little under one-quarter of the children went to the crèche and 13.5% were breastfeding the crèche. More than half the women said that the place for breastfeeding in the crèche was inadequate. Table 3 presents the results of the final hierarchical multivariate model run to determine the association between continuation of breastfeeding for at least 12 months and the study variables.
The following variables showed a positive association with breastfeeding continuation for at least 12 months: mothers without a college degree; mother with previous experience of breastfeeding for at least 12 months; child not given milk other than breast milk when the mother returned to work; child had not used a pacifier; and mother works only during the day. The following variables showed a negative association with the outcome: maternal age (3% reduction in prevalence for each year of age); gestational age (7% reduction in prevalence for each week of gestation); mother received breastfeeding support from the child's caregiver and professional breastfeeding support.

Discussion
This study is the first in Brazil to conduct a detailed investigation of work-related factors influencing BF after return to work and their relation to breastfeeding duration.
Surprisingly, only one of the 22 variables was associated with BF continuation for at least 12 months: work shift. Working only during the day increased the prevalence of BF continuation for at least 12 months by 37%. This association was reported by a previous study in Brazil, suggesting that working night shifts may act as a barrier to breastfeeding due to night-shift fatigue 12 .
Unlike other studies 6,7,13,14 , the current study did not find an association between maintenance of BF after return to work and flexible working Breastfeeding for at least 12 months hours, breaks for breastfeeding or milk expression, appropriate facilities for expressing and storing milk, support from employers and work colleagues, and child care near or at the workplace. This may be partially explained by differences between our study and the other studies, such as study population and workplace characteristics, duration of maternity leave, and definition of outcome. Our findings show that individual and non-work-related factors showed more associations with BF continuation for at least 12 months than work-related factors. Some of the individual factors associated with maintenance of BF after return to work in the present study -such as maternal age, education level, prior experience of breastfeeding, gestational age, use of milks other than breast milk, breastfeeding support, and using a pacifier -were associated with duration of BF by previousstudies 11,[15][16][17][18] . Therefore, these factors should also be taken into account when designing strategies for promoting BF after return to work.
The negative association between professional support and maintenance of BF for longer found in this study is an apparently paradoxical result. A possible explanation is that the women who reported not having received professional it continues support may have had less problems breastfeeding and thus needed less support and were less likely to seek professional help. Despite the importance of information for breastfeeding management upon return to work 17 , our findings reveal that the women in our sample were uninformed. Most of the women reported that they had not received guidance from their employer on breastfeeding upon return to work and were unaware that they could have used the services of lactation consultants. In addition, some were unaware that the hospital had a HMB and others did not even know they were entitled to take one hour off work each day for breastfeeding until the child's first birthday.
Despite offering various conditions and facilities to promote and protect breastfeeding -for example, being a Baby-Friendly certified hospital and the provision of six months of maternity leave, a HMB, breastfeeding specialists, and child care near the workplace -the hospital does not develop specific actions directed at women returning to work who want to continue breastfeeding. This is probably why just one-third of the women expressed milk during work, which is essential for maintaining an adequate milk supply. This may also explain why half of the women who expressed milk discarded the milk, despite the hospital providing the necessary facilities for safely expressing and storing milk, thus depriving the child the chance of receiving BM at home in the mother's absence.
A study conducted in Australia with female health professionals also showed that women were uninformed about aspects related to breastfeeding after return to work 19 . The findings showed that very few women spoke to managers about breastfeeding after return to work and were aware of the existence of breaks for breastfeeding or milk expression. More than half of the respondents reported that they would have probably breastfed for longer after return to work if they had received more information and support at the workplace.    In the current study, a little over half of the respondents reported having received support from colleagues and management for continuing breastfeeding after return to work. Despite this lack of support, the rate in our study was higher than that observed by Weber et al. 7 in Australia, where only 11% of respondents mentioned having received support from management and 13% from colleagues. It is interesting to note that, according to the women in the present study, there was no difference between the support offered by male and female colleagues.
Studies show that mothers who are unable to breastfeed during work are more likely to stop breastfeeding early 20 and that keeping children close to or at the workplace helps promote BF continuation after return to work by providing the opportunity to directly feed the infant from the breast 6 . This did not occur with the women in our study, which may be explained by the fact that most of the children that went to the crèche were not breastfed there. The inadequacy of the place for breastfeeding at the crèche reported by the mothers in the present study may have contributed to the low adherence to breastfeeding in this space.
It is interesting to note that, despite not having a specific program for welcoming women returning to work after maternity leave, most of the respondents reported that they were satisfied with the support they received. This may be due to the general level of worker satisfaction with the organization.
It is also interesting to note that duration of EBF was greater among the women in the present study than in the general population of Porto Alegre (150 days versus 52 days 8 ). One of the reasons for this difference is that the women in our study are professionals working in a Baby-Friend-ly certified hospital. Other factors that may have contributed to this finding include the provision of six-month maternity leave and exclusion of women who weaned their babies before returning to work from the study.
Despite the importance of this study's findings, it is important to highlight some of its limitations. First, the study was conducted in a single location with a population made up predominantly of health professionals. Second, the women in our study have a number of privileges that most working women in Brazil do not have, including six months of maternity leave, high level of education, and access to a HM Band child care near the workplace. Future research should therefore focus on organizations with different conditions and facilities for breastfeeding women and different groups of women in order to better understand work-related factors that act as barriers and facilitators for BF continuation after return to work.
In conclusion, our findings show that work-related factors that would usually be expected to negatively influence the maintenance of BF after return to work were not associated with BF continuation for at least 12 months. This may be partially explained by the benefits offered by the organization, in particular six months of maternity leave. However, it is probable that a specific program designed to welcome women who want to continue breastfeeding after return to work would increase the prevalence of breastfeeding continuation for at least 12 months and make conciliating breastfeeding and working an easier and more pleasurable experience. The factors associated with BF continuation identified by this study, including those not directly related to the workplace, should be taken into consideration when designing strategies to promote BF afterreturn to work.