The perceptions and experiences of women with a body mass index ≥ 30 kg m2 who breastfeed: A meta‐synthesis

Abstract Breastfeeding has copious health benefits for both mother and child, but rates of initiation and maintenance among women with a body mass index (BMI) ≥ 30 kg m2 are low. Few interventions aiming to increase these rates have been successful, suggesting that breastfeeding behaviour in this group is not fully understood. Therefore, this review aimed to systematically identify and synthesise the qualitative literature that explored the perceptions and experiences of women with a BMI ≥ 30 kg m2 who breastfed. The search identified five eligible papers, and a meta‐ethnographic approach was taken to synthesise the findings. One theme was identified: “weight amplifies breastfeeding difficulties,” revealing that women with a BMI ≥ 30 kg m2 experience common breastfeeding difficulties to a greater degree. In particular, women with a BMI ≥ 30 kg m2 struggle with the impact of medical intervention, doubt their ability to breastfeed, and need additional support. These findings can inform understanding of breastfeeding models, future research directions, intervention development, and antenatal and post‐natal care for women with a BMI ≥ 30 kg m2.

. Moreover, in contrast to interventions targeting all women (Sinha et al., 2015), most were unsuccessful at increasing breastfeeding duration. The one intervention that reported an increase in breastfeeding duration was not representative of all women with a BMI ≥30 kg m 2 (Carlsen et al., 2013). Together, this suggests that more understanding of the facilitators and barriers to breastfeeding behaviour in women with a BMI ≥30 kg m 2 is needed in order to better inform targeted and effective interventions.
Psychological factors can facilitate and prevent breastfeeding behaviour (de Jager, Broadbent, Fuller-Tyszkiewicz, & Skouteris, 2014). A recent systematic review that investigated the role of psychological factors in breastfeeding behaviour in women with a BMI ≥ 30 kg m 2 (Lyons, Currie, Peters, Lavender, & Smith, 2018) found that planning to breastfeed, believing in the nutritional adequacy and sufficiency of breast milk, believing that important others approve of breastfeeding, knowing others who have breastfed, and having positive body image were associated with better breastfeeding outcomes. However, this review included only quantitative studies, and therefore, the conclusions are limited to those factors, which had been investigated in studies using psychometric measures, and do not include factors emerging from women's lived experiences.
Therefore, the current review aims to systematically synthesise the qualitative literature that explores the perceptions and experiences of women with a BMI ≥30 kg m 2 who breastfed.

| Design
A systematic approach was taken to identifying and reviewing all qualitative research that has explored the perceptions and experiences of women with a BMI ≥30 kg m 2 who report engaging in breastfeeding behaviours. The synthesis was informed by Noblit and Hare's (1988) method of meta-ethnography. This method was appropriate for the analysis as it can be used to synthesise studies that have employed different qualitative methods (Campbell et al., 2011;Elmir, Schmied, Wilkes, & Jackson, 2010). Included studies were identified by conducting a systematic search. Multifield search builders were used to combine keywords in accordance with the SPIDER framework (see Table 1; Cooke, Smith, & Booth, 2012). Authors and journals of included studies were also hand-searched to minimise the possibility of missing relevant articles.

| Eligibility criteria
This synthesis included studies that investigated the perceptions and experiences of women with a BMI ≥30 kg m 2 who reported breastfeeding behaviours. Therefore, included studies focused on women's perceptions and/or experiences of breastfeeding who had a BMI ≥30 kg m 2 at the start of their pregnancy. All studies had to include qualitative methods of both data collection and analysis.
Studies that focused on health professionals' perceptions of these women's experiences were included; in this case, only relevant themes were extracted (i.e., those themes that depicted factors that influenced the women's breastfeeding behaviours). Studies that included subsamples were included if their views were reported separately from those of women with a BMI ≥30 kg m 2 . Included studies were restricted to those written in English due to a lack of funding to translate non-English papers, with no date restrictions set.

Key messages
• Women with a body mass index (BMI) ≥ 30 kg m 2 are less likely to breastfeed than women with a BMI ≤ 30 kg m 2 .
• Women with a BMI ≥ 30 kg m 2 experience common breastfeeding difficulties to a greater degree due to their weight.
• Women with a BMI ≥ 30 kg m 2 struggle with the impact of medical intervention, doubt their ability to breastfeed, and need additional support.
• It is recommended that health professionals empower women with a BMI ≥ 30 kg m 2 to breastfeed despite medical intervention, address women's beliefs about their ability to produce nutritionally adequate milk, and fully explain and discuss the purpose of breastfeeding support.

| Study selection and quality appraisal
Search results were entered into EndNote ×7, and duplicates were removed. Titles and abstracts were screened by one researcher (S. L.), and those that did not meet the eligibility criteria were excluded. At this stage, an inter-rater reliability assessment was conducted, with a second researcher (D. M. S.) also reviewing 10% of title and abstracts (Kitchenham, 2004 (2017) was used. This involves assessing the aims, design, recruitment strategy, data collection, researcher bias, ethical issues, data analysis, findings, and implications of each study and awarding "yes" if the item is met, "no" if it is not, or "unclear" if the necessary information is not available. As the quality appraisal did not form part of the eligibility criteria, a scoring system was not applied.

| Data extraction and synthesis
Data extraction and synthesis followed the meta-ethnography process described by Noblit and Hare (1988 ; Table 2). This first involved scoping the literature and identifying an appropriate area of interest and research question. Eligibility criteria were created and applied to determine which studies were included in the synthesis. Included studies were read and re-read by two researchers, with key phrases and themes highlighted and combined into lists. Two researchers completing this step reduced the risk of excluding relevant studies (Kitchenham, 2004). These lists also included information on the study aims, sample, methods, and form of analysis. The list for each study was then compared and contrasted with the list for each other study by three researchers, and the relationships between studies were determined. This resulted in the translation of studies into one another, and a line of argument was developed. Three researchers FIGURE 1 PRISMA flow diagram of study selection completing this step prevent bias and improve conceptual development (Morse, 2015). The results of the synthesis are expressed in this paper.

| RESULTS
The systematic search identified 1,115 papers, with 1,074 remaining after duplicates were removed ( Figure 1). The titles and abstracts were then reviewed, resulting in the exclusion of 1,066. Eight papers were reviewed at full text, with five included in the synthesis. For four of the five studies, women with a BMI ≥ 30 kg m 2 were interviewed about their perceptions and experiences of breastfeeding; for the remaining study (Massov, 2015), health professionals were interviewed on their experiences of providing care for BMI ≥ 30 kg m 2 who breastfeed. Included studies employed either content or thematic analysis. All studies were high quality. Study characteristics are summarised in Table 3, and individual quality item scorings for each study can be found in Table S1.
The synthesis identified one theme: Weight amplifies breastfeeding difficulties. The way in which the included papers mapped on to the theme and subthemes is illustrated in Table 4.
Quotes from included papers are provided to support the synthesis findings.  (Garner et al., 2014;Garner et al., 2016;Keely et al., 2015;Lyons et al., 2019;Massov, 2015). Therefore, the theme had three subthemes: (a) psychological reactions to medical intervention, (b) perception of body's ability to breastfeed, and (c) additional need for support.

| Psychological reactions to medical intervention
This subtheme describes how psychological reactions to medical intervention during pregnancy and birth can have a negative impact on breastfeeding initiation and maintenance. Women with a BMI ≥30 kg m 2 are often labelled as "high risk" in pregnancy (Lyons et al., 2019) due to their higher risk of experiencing long labours, which result in caesarean section, subsequent infections, longer hospital stays, reduced mobility, delayed skin-to-skin contact, and even periods of separation from their infant (Garner et al., 2014;Garner et al., 2016;Keely et al., 2015). During this, women often feel disappointed about their breastfeeding experience and unmotivated to continue, as there are already numerous barriers making breastfeeding difficult (Garner et al., 2014;Lyons et al., 2019). In particular, women with a BMI ≥ 30 kg m 2 may not feel comfortable to breastfeed in the busy hospital environment, where privacy can be difficult to achieve (Keely et al., 2015). Due to this, women feel that their choice of infant feeding method is out of their control, which has a negative impact on breastfeeding initiation and maintenance (Lyons et al., 2019). As the likelihood of experiencing medical intervention during pregnancy and birth is higher among women with a BMI ≥ 30 kg m 2 (Weiss et al., 2004), these difficulties are amplified within this group.
Actively seeking professional and social support can help women to overcome these barriers and breastfeed (Lyons et al., 2019).  Method of data collection

Method of analysis
Garner et al. (2014) (1) • To understand how health professionals across the continuum of care perceive breastfeeding among obese women.
• To understand how health professionals experience providing care to obese women to identify potential barriers and ways to improve breastfeeding-related care.

USA
• Thirty-four health professionals in a range of roles (e.g., lactation consultants, nurse midwife, and nurses), and with experience ranging from 5 to 30 years.
• Recruited participants by sending emails through listservs to health professionals in obstetrics, midwifery, family medicine, and paediatric practices.
Face-to-face, semistructured interviews Content analysis Garner et al. (2016) (2) • To understand obese women's experiences and perceptions longitudinally, with a normalweight comparison group, to identify key experiences and barriers that are unique to or more common among obese women. I was just shaking all over, like complete shaking so I was like, "I don't feel like I can hold him" (Eve; 3) And then I got really bummed because she was only able to stay on me for a couple of minutes. The nursery came and got her because her blood sugar levels were low (Allison; 1)

| Perception of body's ability to breastfeed
This subtheme describes how lacking belief in your ability to breastfeed and, in particular, your body's ability to produce adequate and sufficient milk can have a negative impact on breastfeeding initiation and maintenance. For this reason, women either did not plan to breastfeed or did so only for short periods of time to avoid disappointment, which negatively impacted initiation and maintenance (Garner et al., 2016). Furthermore, difficulty latching and positioning due to larger breasts and bodies led women to feel unprepared and fear smothering their infant and reinforced their lack of belief in their ability to breastfeed (Garner et al., 2014;Garner et al., 2016;Keely et al., 2015;Lyons et al., 2019;Massov, 2015). Women also lacked belief in their ability to breastfeed discreetly, as their baby would cover less of their larger breasts and body, which led them to give their infant formula milk when family members or friends were visiting their homes, or they were out in public (Garner et al., 2014;Keely et al., 2015;Lyons et al., 2019;Massov, 2015). Because of their size, women also worried about the nutritional quality and sufficiency of their breast milk; women believed their diet quality may be poorer than women with lower BMIs and that this would translate directly to their breast milk quality, meaning many women introduced formula milk early due to concerns that their infant was still hungry after breastfeeding (Keely et al., 2015;Lyons et al., 2019;Massov, 2015). Other barriers, such as poor availability of nursing bras and tops in larger sizes, also reinforced women's lack of belief in their ability to breastfeed (Garner et al., 2016;Lyons et al., 2019). Seeking professional support to assist with latching and positioning, social support to normalise breastfeeding for women with a BMI ≥ 30 kg m 2 , and information to correct beliefs about their ability to produce nutritious and sufficient milk helps women to overcome these barriers and initiate and maintain breastfeeding (Lyons et al., 2019   •Women with a BMI > 30 kg m 2 needed more latching support over a longer period of time but had less access to social support.
• Partner support was viewed as important when making infant feeding decisions. • Few women with a BMI > 30 kg m 2 sought professional support because they misunderstood its purpose, but those that did described it as a positive experience. • Women with a BMI > 30 kg m 2 wanted support at home and valued support from friends who breastfed.
• Seeking support helped women with a BMI > 30 kg m 2 to overcome medical, social, and practical barriers. • Women with a BMI > 30 kg m 2 wanted support beyond the initial latching stage to normalise and support maintenance.
• Women with a BMI > 30 kg m 2 sometimes found professional help distressing, as they were not comfortable with health professionals touching their bodies and infant. • Women with a BMI > 30 kg m 2 felt uncomfortable asking for professional help in hospital as they felt staff were too busy.

| Additional need for support
This subtheme describes how women with a BMI ≥ 30 kg m 2 need more support, not only while initiating but also while maintaining breastfeeding. This is because women with a BMI ≥ 30 kg m 2 often have more difficulty latching and positioning, meaning they need more professional support to help them successfully initiate (Garner et al., 2014;Garner et al., 2016;Lyons et al., 2019). Women also valued social support from family members and friends, and particularly other women with a BMI ≥30 kg m 2 who were breastfeeding, as they could share their experiences and ask for advice and tips, which improved their confidence (Keely et al., 2015;Lyons et al., 2019). This, in particular, helped women to normalise breastfeeding beyond 6 months, as many women had little social support available to them and had not met women who had breastfeed infants beyond this point (Garner et al., 2014;Garner et al., 2016;Lyons et al., 2019). However, there were barriers that prevented support seeking. For example, many women did not seek professional support because they felt they may be irritating busy health professionals, and they were uncomfortable breastfeeding in front of others, or with health professionals touching their bodies and infant (Massov, 2015). Furthermore, many women did not attend breastfeeding clinics due to a misconception of their purpose; some women believed that groups were only for women who had trouble with their latch, others believed that groups were only for women who were successfully breastfeeding, and some believed that the groups were for social support, but felt uncomfortable attending alone (Keely et al., 2015).
She was like you need to go to the local breastfeeding support group you know you'll get support from women like you and it'll, it'll help your confidence, and it was the best thing I did, definitely (Chloe; 4) I just remember the midwife coming in and almost angry that I was upset because I was having trouble doing it and "I'll show you how to express" but then just pretty much grabbed my breast without asking if that was OK.
And being really, really rough and aggressive with her and I kind of … it's not on … She just ruined the whole stay in hospital, like then it made me apprehensive about any other midwife that walked in the room and what were they going to do to me? (Jess; 5)

| DISCUSSION
This review aimed to explore the perceptions and experiences of women with a BMI ≥ 30 kg m 2 who breastfeed. Five studies were included. The findings highlight that these women experience similar difficulties to other women but that these are more problematic for women with a BMI ≥ 30 kg m 2 because of their weight. These findings can help health care professionals and interventions to target their support, increase breastfeeding initiation and duration, and ultimately reduce obesity and obesity-related diseases in women with a BMI ≥30 kg m 2 and their children.
All women face the possibility of experiencing medical intervention in pregnancy and labour (Bhattacharya, Campbell, Liston, & Bhattacharya, 2007). However, women with a BMI ≥ 30 kg m 2 are more likely to suffer complications during their pregnancy and labour, such gestational diabetes and hypertension, pre-eclampsia, preterm delivery, stillbirth, and post-partum haemorrhage, resulting in higher rates of hospital stays, inductions, and elective and emergency caesarean sections (Athukorala, Rumbold, Wilson, & Crowther, 2010;Bhattacharya et al., 2007). The findings highlight that these complications and interventions are barriers to breastfeeding and can leave women lacking the motivation and perceived control over their infant feeding behaviours to overcome them, which can partially explain why breastfeeding rates among women with a BMI ≥ 30 kg m 2 are particularly low. Therefore, it is recommended that health care professionals are aware of the impact of these experiences on women with a BMI ≥30 kg m 2 and provide assistance and encouragement where necessary to allow women to feel in control of their infant feeding behaviours and breastfeed. However, it is interesting that one study (Massov, 2015) did not identify the impact of medical intervention, despite recruiting their sample in a country that has low rates of "normal" spontaneous vaginal births (i.e., 33% New Zealand vs. 59% in England, 56% in Scotland, and 58% in the United States; Ministry of Health, 2015;NHS Digital, 2017;NHS Scotland, 2016;Declerq, Sakala, Corry, Applebaum, & Herrlich, 2013). One explanation for this may be that most women in New Zealand choose a midwife to deliver all of their care (i.e., throughout pregnancy and birth), where particular emphasis is placed on creating a partnership for decision making (Page, 2001;Ministry of Health, 2015), meaning these women may feel more in control of any medical intervention they receive, reducing its impact on breastfeeding. However, this approach to care is also taken in the Women often doubt their ability to breastfeed (Moore & Coty, 2006). However, the findings of this study suggest that women with a BMI ≥ 30 kg m 2 , in particular, are concerned about their ability to produce adequate and nutritious milk, which prevents them from planning to breastfeed. Furthermore, this study highlights that women with a BMI ≥ 30 kg m 2 lack confidence in their ability to breastfeed in social situations. Quantitative studies have shown that lacking belief in the quality and sufficiency of breast milk and confidence to breastfeed publically is associated with negative breastfeeding outcomes (Hauck, Fenwick, Dhaliwal, & Butt, 2011;Stuebe & Bonuck, 2011), meaning the study findings can offer some explanation for the low breastfeeding rates in this group. Therefore, it is recommended that these beliefs are addressed in pregnancy, to encourage women to feel confident in their body's ability to breastfeed and to make infant feeding plans. In particular, addressing beliefs about initial infant feeding frequency may reduce women's worry that the breast milk they produce is insufficient. Once breastfeeding, it is recommended that women are signposted to breastfeeding support groups, in order to build their confidence to breastfeed in social situations.
Professional and social support can be beneficial for all breastfeeding women (Hannula, Kaunonen, & Tarkka, 2008;Ingram, Rosser, & Jackson, 2005 Only five studies met the criteria and were included in the synthesis, which may limit the strength of the conclusions drawn. In particular, it is important to note that the studies included were all conducted in highincome, English-speaking countries, and therefore, the findings may not be applicable to other settings. However, this fits with the form of analysis, as meta-ethnography aims to synthesis findings from closely related studies exploring specific phenomena, rather than make generalisations across fields (Britten et al., 2002;Campbell et al., 2003).
This study had several strengths. In particular, relevant studies were searched for systematically and assessed against a predefined inclusion criteria, minimising the risk of missing material and researcher bias during study selection (Moher, Liberati, Tetzlaff & Altman, 2010;Kitchenham, 2004). Furthermore, inter-rater reliability checks were conducted, and included studies were appraised for their quality using a valid tool, increasing the strength of the conclusions drawn (Kitchenham, 2004). Finally, included studies reported the experiences of both women with a BMI ≥ 30 kg m 2 who ceased breastfeeding early and those who continued to breastfeed for as long as they wished, increasing the applicability of the findings.
Several implications are generated from the findings. First, as few studies were eligible for inclusion in this synthesis, more qualitative research of these women's breastfeeding experiences is needed, with particular focus placed on the influence of care. For example, longitudinal qualitative studies conducted throughout women's pregnancies and breastfeeding experiences may highlight aspects of medical intervention and communication, which could be adapted to reduce the negative impact on breastfeeding behaviour. These studies may also highlight how health care professionals can best reassure women with a BMI ≥ 30 kg m 2 that they are capable of producing nutritionally adequate and sufficient milk and that they can overcome barriers due to their body size or shape with professional and social support. The influence of medical intervention could also be investigated quantitatively, by measuring both the instances and types of medical intervention women experience and their motivation to and perceived control over breastfeeding. In order to increase breastfeeding behaviours among this group, intervention development is also necessary; future interventions should consider these findings, particularly those that offer breastfeeding support. Finally, it is recommended that health care professionals also consider these findings during their practice, by taking extra care to empower women with a BMI ≥ 30 kg m 2 to breastfeed despite whether they have experienced medical intervention, address women's beliefs about their ability to breastfeed and produce nutritionally adequate milk in both pregnancy and postpartum, and to fully explain and discuss the purpose of breastfeeding support, including different available options that are acceptable to each individual.

| CONCLUSION
In conclusion, this study explored the perceptions and experiences of women with a BMI ≥ 30 kg m 2 who breastfeed. It was found that experiencing medical intervention reduced women's sense of motivation and control over breastfeeding behaviours, the women doubted their ability to breastfeed, and often misunderstood the purpose of breastfeeding support, or felt uncomfortable to ask for help. These findings can inform understanding of breastfeeding models, future research directions, intervention development, and antenatal and post-natal care.

ACKNOWLEDGMENTS
The authors would like to thank the authors of included papers for supplying additional information where necessary.

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.

CONTRIBUTIONS
All researchers contributed to the design of the study. SL, DS, and SC completed the analysis. The article was drafted by SL. All authors critically revised the article and gave approval for the final version to be published.