Breast or bottle? Eating disordered childbearing women and infant‐feeding decisions

Abstract

Debates about infant‐feeding methods have intensified in recent years with increasing pressures on women living in industrialized nations to breastfeed their infants. This paper, based on a qualitative study of 16 childbearing women with a pre‐existing eating disorder living in the north of England, examines participants' motivations for, and understandings of, infant‐feeding decisions and practices. In this study, a small number of participants reported being ‘desperate’ to formula feed in order to resume practices underpinning their eating disorder and thereby to shed the weight accumulated during pregnancy. These participants anticipated an early return to restrictive eating, heavy exercise regimes and/or bingeing/purging behaviours. Most participants, however, reported being ‘desperate’ to breastfeed because this implied ‘good’ mothering and prolonged the time during which they could consume ‘naughty’ treats. Women who opted to breastfeed generally believed this would accelerate weight loss. This study contributes to research on the subjective experiences of a particular group of women living with chronic illnesses and problematic relationships with their bodies. Negotiating individual transitions to motherhood required participants to confront their own, often longstanding, disrupted eating patterns and to make important decisions about infant‐feeding methods. Findings from this study raise questions about some of the assumptions underpinning infant‐feeding activities and articulate some of the complexities surrounding these issues. By highlighting ways in which women may compromise their own well‐being by prioritizing their baby's needs, for example by persisting with breastfeeding when they were ‘desperate’ to re‐engage with their disordered eating practices, an individualized cost‐benefit framing is outlined.

Introduction

It is generally acknowledged, at least in the medical and public health literature on the subject, that women who breastfeed do so primarily to confer health benefits on their babies (Murphy 1999; Hamlyn et al. 2002). A desire to lose the weight gained in pregnancy, and to recover an approximation of pre‐pregnancy shape and size, is also known to influence women's decisions, at least in Western cultures (Ryan 1998). In this study, concerns about body image, which tend to be overlooked in the moral hierarchy of motivational factors influencing women's infant‐feeding decisions, were significant factors in participants' decisions to both formula feed ¹ and breastfeed their infants.

Until fairly recently, the research‐driven literature on eating disorders has tended to be clinically oriented, employing quantitative methods to classify, and elaborate, (female) pathology. Indeed, prior to the early 1980s, gender was either absent altogether or was theorized in essentialist terms (Bordo 1993) which closes off the possibility of change or variation. Within the framework of essentialism ‘a belief in the real, true essence of things’ (Fuss 1989, pp. xi–xii) is endorsed so that variables such as gender, culture and class are either discounted or seen as fixed, solid and unchanging. Most importantly insofar as the central thesis of this paper is concerned, ‘essentialism is typically defined in opposition to difference’ (Fuss 1989, pp. xi–xii) and hence the production of alternative meanings and understandings of individual experience is disallowed.

Following on from critical discourses advanced by feminist scholars on discrete aspects of maternity transitions, debates on infant feeding have become more prominent in recent years. Increasing social, medical, and more recently political, pressures on women in industrialized nations to breastfeed (and women's strategies of avoidance) has tended to polarize debates on the circumstances underpinning infant feeding (Dykes 2005). A lack of attention to the socio‐cultural conditions surrounding infant‐feeding practices has encouraged a misperception that breastfeeding, as it is practised by women living in ‘traditional’ societies, can be regarded as an homogenous event to be exported as an exemplar of good practice for Western women (another homogenous category) to emulate (Maher 1995). This overlooks the fact that countervailing pressures, including lower social class and younger age (Wiggins et al. 2005), and a lack of privacy (Hoddinott & Pill 1999), also influence decisions not to breastfeed.

Many women make concerted attempts during pregnancy to cease, or at least reduce, behaviours considered deleterious to their own health and that of their unborn baby, including those associated with disordered eating practices (Lacey & Smith 1987; Morgan et al. 1999). Despite these efforts, symptoms often recur early in the post‐natal period among eating disordered women who elect not to breastfeed (Lewis & Le Grange 1994). Pre‐pregnancy attitudes to body image and weight act as a significant predictor of post‐natal discontent (Fairburn & Welch 1989) and a rapid return to pre‐existing disordered eating practices following childbirth is not uncommon (Stein & Fairburn 1996; Morgan et al. 1999). There is also some evidence of the actual genesis of an eating disorder around this time (Tiller & Treasure 1998; Mitchell‐Gieleghem et al. 2002).

Given the continued ambivalence which surrounds mothering and the maternal role in contemporary Western societies (Hollway & Featherstone 1997), it is perhaps unsurprising that most study participants were generally apprehensive about this transition. Furthermore, research with pregnant women who do not declare an eating disorder reveals that they nonetheless experience difficulties adapting to their changing body shapes and sizes during pregnancy and some use smoking and dieting as a means of weight control (Abraham et al. 1994). Little is known about the strategies pregnant women currently use to maintain their body weight and shape within acceptable (to them) limits although research in New Zealand (Longhurst 2005) and the UK (Earle 1998; Beale 2002) confirms that many pregnant women are very concerned about their visual appearance throughout the transition to motherhood. Lucy Bailey's (2001) work conceptualizes female embodiment explicitly within a maternity setting and discusses how first‐time mothers' perceptions of (gendered and classed) self‐identity are shaped by the physicality of pregnancy.

The fashion and media industries are frequently identified as primarily responsible for the continued (global) rise in eating disorders among women because their advertising campaigns emphasize unachievable ideals of slenderness (Stice et al. 1994; Groesz et al. 2002). Personal agency has tended not to figure in these explanations. Social constructionists have criticized this stance, arguing that interpretations of the body are neither ahistorical nor acultural, but are contextually situated (Benveniste et al. 1999). Furthermore, feminist critiques have highlighted the role of Biomedicine in the construction and production of gendered reproductive identities and normative constructions of femininity in Western (Moulding 2006) and non‐Western (Lee 1999) cultures. Feminist theorists have thus progressively deconstructed the eating disordered ‘subject’ and have begun to theorize (her) position in a more discursive manner (Malson 1998; Saukko 1999).

Childbearing and rearing activities may thus be viewed as focal points in the life cycle when, in addition to positive feelings, potentially significant, and negative, body‐related feelings may also be generated. Our findings support the positioning of the pregnant body as a site of potential resistance to the wholesale objectification and commodification of females in Western societies (Earle 2003) and for participants in this study, pregnancy and breastfeeding were rare zones of exclusion from normative preoccupations with dieting, bingeing, purging and an obsession with slenderness.

Materials and methods

This was a small, internally funded, qualitative study, ² exploratory and descriptive in design, employing a feminist (Reinharz 1992) ethnographic approach (Hammersley & Atkinson 1983) to data generation. An inductive approach was necessary as little is known about how women with eating disorders manage transitions to motherhood. A pilot interview with one eating disordered mother of two children preceded the study proper and provided an opportunity to refine the interview guide.

The study itself was undertaken in two phases and comprised two cohorts of women (n = 16 in total), five of whom were pregnant at the time of recruitment while the remaining 11 were already mothers with at least one child under the age of 2 years. The original aim of 20 participants (10 pregnant and 10 non‐pregnant) was revised following recruitment difficulties. This was particularly the case for pregnant women, most of whom were not accessing specialist eating disorder facilities at the time pregnancy was confirmed and nor had they necessarily disclosed their eating disorder to a health professional. Hence, they could not be specifically targeted and invited to join the project. Although considerable effort was made to increase the diversity of the sample by placing posters and adverts in non‐National Health Service (NHS) environments including community newsletters, supermarkets, gyms, sports centres and alternative health clinics, the majority of participants were recruited through NHS facilities including a specialist eating disorders service, antenatal clinics and General Practitioner surgeries. Some women contacted the researchers directly in response to information they had read about the study in the public domain, others responded to information given to them by their health professionals.

Participants ranged in age from 23 to 44 years and included first‐time mothers and those with more than one child. All participants were white and from varied socio‐economic backgrounds; all described themselves as heterosexual and most were living with male partners. All participants self‐defined as having an eating disorder; bulimia and restrictive eating were the most widely reported practices. As the aim of the research was to report what childbearing women themselves had to say about their eating disorder, no attempt was made to assess participants' according to established mental health criteria (American Psychiatric Association 1994) and nor was any attempt made to impose a diagnosis on participants. From outward appearances, all participants were of average size. We did not succeed in recruiting very thin or very large women. Having commenced their disordered eating career in adolescence, all participants could be classed as chronic sufferers and all but one had sought, and received, medical treatment, including psychotherapy and prescribed medication.

Ethical approval was obtained and project information sheets were made available to all participants. Written consent to participate in the study and to audio‐record interview(s) was obtained from all participants; women recruited to the longitudinal phase of the research re‐consented prior to each episode of data collection. The first author (H.S.), who has a midwifery background but has not been in clinical practice for some years, interviewed all participants by prior arrangement in their own homes. Interviews typically lasted between 90–120 min. An interview guide was drafted and subsequently modified in light of emergent themes. Prior to follow‐up interviews with the longitudinal cohort of women, the interviewer read individual transcripts in order to probe issues raised in the previous interview. All respondents were asked to provide baseline demographic details including age, relationship status, number of children, a brief history of the eating disorder and any treatment received. First interviews with pregnant women typically focused on experiences of pregnancy and the maternity services. Subsequent interviews explored participants' ongoing relationships with their changing bodies, experiences of childbirth/transitions to motherhood and the current status of their eating disorder. Interviews with non‐pregnant women investigated their experiences of motherhood and the impact of this transition on their eating disorder. Interviews were transcribed verbatim and subsequently verified as a legitimate record by the first author (H.S.). Anonymized quotes from interview transcripts are presented here to illustrate key theoretical points.

The pregnant cohort of women (n = 5) provided an important longitudinal element which enabled the researchers to map their ongoing experiences of pregnancy and early motherhood. They consented to in‐depth interviews on four separate occasions: twice during pregnancy (around 24 and 36 weeks) and twice within the first few months following birth. Interviews were undertaken at these points in the childbearing/rearing trajectory in order to capture women's reactions to changes in their weight, shape and body image, and their infant‐feeding intentions and actual practices. Established mothers (n = 11) participated in a one‐off, in‐depth, interview which focused on their (sometimes multiple) experiences of child bearing and rearing, and the effect of these events on the overall trajectory of their eating disorder.

The analysis of interview transcripts was preceded by an examination of the relevant literature. Although interviews followed on after this stage of research activity and hence theoretical discussion partly preceded analysis, it was nonetheless grounded in the narratives of women's everyday lives and their experiences of living with a chronic, disabling, stigmatized but largely invisible, condition (Moss & Dyck 2002; Broussard 2005).

The empirical data and the theoretical material thus worked together, informing the process of analysis in a continuous cycle. In this way, the data elaborated key theoretical concepts such as the role of power and (self)control. The authors collectively analysed a random selection of transcripts (n = 6) to identify, and agree, key themes. Following this process, a framing code was devised and subsequently modified by the first author in light of newly emerging material. Changes to the framing code were subsequently agreed among all authors before new themes were added.

Results

Infant‐feeding decisions: is formula feeding all right?

In defying normative understandings of ‘good’ parenting, women who opt to formula feed risk being constructed as morally suspect and deviant mothers who wantonly disregard their mothering responsibilities (Murphy 1999). Hence, infant‐feeding decisions may be regarded as a ‘highly accountable matter’ (Murphy 1999, p. 205) because decisions are weighted with significant ‘moral baggage’ (Murphy 1999, p. 205) and set against broader ideologies and conceptualizations of ‘appropriate’ parenting. Regardless of the method of infant feeding eventually chosen then, women may expect their decisions to be challenged in both lay and professional arenas. Breastfeeding tends to be widely promulgated as the morally superior choice despite the fact that less than 50% of UK mothers are still breastfeeding 6 weeks following birth and only just over 20% are doing so at 6 months (Department of Health 2002). What is more, despite considerable investment in breastfeeding initiatives, these figures have remained largely unchanged over the past decade. Although well informed about the benefits of breastfeeding, some study participants were unwilling to consider this option because they were ‘desperate’ to resume purging and strenuous exercise regimes in order to lose the weight they had gained in pregnancy. These women believed that a swift return to such practices would effect a more rapid weight loss than they could hope to achieve by breastfeeding.

In common with many women who live with socially constructed limitations of a disabling condition (Moss & Dyck 2002), participants in this study were very knowledgeable about their own capacity and their own limitations. The small number of women who opted to formula feed their babies from birth made this choice knowing that breastfeeding was the preferred option insofar as promoting infant health and well‐being was concerned. The re‐emergence of symptoms associated with their eating disorder, however, worked against them making this choice. Most participants believed that frequent cycles of bingeing and purging were incompatible with the production of good quality milk, in sufficient amounts, to satisfy the baby and hence breastfeeding was discounted as a possible feeding option. Such women could thus be described as making an ‘ethically informed’ choice (Edwards 2004) which was grounded in a profound concern for the baby's well‐being but which competed with a more urgent need to protect their own (mental) health.

A number of participants, who were of the opinion that they had no option but to formula feed, reported being reprimanded because their choice of feeding method did not accord with the prescribed norms of contemporary midwifery practice:

I couldn't breastfeed. I just couldn't. I was desperate to get rid of the weight. I just wanted some reassurance from the midwives that bottle‐feeding was all right but all they did was tell me off for not breastfeeding. (Margaret, age 26, two children)

I know that yes, of course they've (midwives) got to encourage you to breastfeed, but they've also got to acknowledge that sometimes you just can't. I couldn't. I couldn't bear eating proper food anymore. (Susan, age 42, four children)

Formula feeding mothers wanted reassurance from midwives that formula milk would not compromise their baby's development but most failed to receive the encouragement they sought. This is perhaps unsurprising in societies where mothers who formula feed tend to be associated with ‘failure’ and report feeling unsupported and very much ‘like second class citizens’ (Battersby 2006, p. 202). This is not to suggest that breastfeeding does not suffer from something of an ‘image’ problem, especially among adolescents and disadvantaged women who lack role models and a cultural tradition. For eating disordered women who are already exceedingly image‐conscious, but who generally project a negative body image, it is suggested that the prospect of breastfeeding may stimulate uncomfortable, and unmanageable, feelings about identity and experience.

Contemporary midwives are in a difficult position with regard to advising childbearing women about infant feeding. On the one hand, they must strive to be ‘with’ women (and the ‘bad’ choices they make) but on the other hand, they ‘are caught up in the disciplinary technologies to which they contribute’ (Murphy 2003, p. 458) and, as such, must vigorously promote breastfeeding (Royal College of Midwives 2004). Within this contested territory, the preferences of individual childbearing women may be easily overlooked as midwives struggle to be facilitators, rather than quiet coercers (Foucault 1991) of choice. The following participant was comforted when she finally encountered a midwife who provided her with much‐needed support in the form of a personal story:

One midwife was really nice. She said ‘Don't be so stupid – my mother never (breast) fed me and I've got two degrees’. But the others tried to pressure. [. . .] All you want is that reassuring voice telling you it will be all right. (Wendy, age 28, one child)

Midwives' reluctance to share ‘positive impact’ stories, based on personal experience and/or professional observations with their clients has been previously remarked upon (Kirkham & Stapleton 2001).

The suggestion to formula feed was not infrequently proposed by male partners:

From the beginning (partner) wanted me to bottle‐feed because he said then he could take his turn. But I think it was really that he didn't trust me to eat properly. I think that was really the reason. (Maureen, age 38, two children)

He (husband) didn't want me to breastfeed because he thought I wasn't eating enough to feed her (baby) properly. [. . .] He was on and on about me giving her the bottle. He even dragged my sister in to try and get her to talk me round. (Emma, age 31, one child)

I think he thought right, pregnancy's over and now the baby's out I can take over. I can make sure he's getting the food he needs. [. . .] That's why he wanted him (2^nd baby) to have a bottle from the start. (Jill, age 27, two children)

It is beyond the scope of this study to offer cogent analyses for why male partners might prefer their babies to be formula fed. A number of participants reported, however, that those partners who were cognizant of the nature and severity of the eating disorder (and many were not), and who had contained their powerlessness to intervene during pregnancy, viewed the post‐natal period as an opportunity for potential reparation. Some male partners were reported as being angry when they were unable to exert control over their partner's disordered eating and related practices during pregnancy. Others expressed concerns about the baby's health and well‐being and had threatened to contact social services with a view to obtaining custody of the baby if participants resumed bingeing and purging in the post‐natal period. Hence, it is reasonable to suggest that formula feeding enabled male partners to wrest control at a point where the welfare of the baby was no longer the mother's exclusive preserve. The rationale advanced for male partners in this study becoming more involved in infant‐feeding decisions confirms earlier research which suggests that men's motivation for greater involvement is based on altruistic concerns, in this case for the welfare of the baby rather than a desire to share the parenting workload (Earle 2000).

Breastfeeding as a strategic (and ‘selfish’) practice

The majority of participants ³ (n = 11) successfully initiated breastfeeding and a considerable number (n = 6) breastfed until 6 months, the UK Government's recommended minimum period for exclusive breastfeeding (Department of Health 2005). All participants had ceased breastfeeding by 7 months except for Louise and Patricia, both of whom continued beyond 12 months. Participants were motivated to breastfeed primarily because they believed this would help them to lose weight and/or resume their pre‐pregnancy body shape more quickly. Women also breastfed because they reasoned that the caloric expenditure associated with this activity meant that they could consume additional food, especially ‘naughty’ ⁴ treats such as chocolate and ice cream. This acted as something of a compensatory mechanism for purging and vomiting practices which most women had suspended throughout pregnancy, and which those who were breastfeeding did not intend resuming until after they had completed the weaning process.

Breastfeeding was like a calorie muncher. I was burning up calories so I could eat more. (Emma, age 31, one child)

I think it (breastfeeding) was partly about me behaving selfishly. It was knowing that it brought your figure back more quickly so I kept putting off and off weaning him because I knew the weight was still dropping off me. (Tina, age 32, one child)

The ‘selfish’ drive to recover the pre‐pregnant (eating disordered) figure seemed to be at least as powerful a motivator for women to breastfeed as were altruistic desires to privilege the welfare of the baby. It should be stressed, however, that breastfeeding is not necessarily synonymous with maternal weight loss and indeed one prominent researcher in the field reports that ‘there is surprisingly little evidence that breastfeeding makes women lose more weight after pregnancy’ (Linne & Rossner 2003, p. 320). Research from the USA and Sweden suggests that women not infrequently report modest weight gains following childbirth but that such gains coexist on a continuum, with a few women actually becoming quite obese (Walker 1998; Linne et al. 2002). Despite weight gain being an important, and perhaps inevitable, aspect of childbearing, the extent to which women may suffer actual weight‐related distress following this event is unknown.

Many participants in this study looked on breastfeeding as a coping strategy which enabled them to maintain an increased calorie intake:

I was told all the things about why breastfeeding is important for the baby but I was doing it because it meant I could eat [. . .] I allowed myself to eat a lot of naughty things when I was breastfeeding [. . .] The calories went straight through me. I could feel it. I was using up so much energy to feed him. (Louise, age 29, one child)

The freedom to eat in a relatively unrestrained way, and to retain what they had eaten, was a very new experience and hence it is perhaps unsurprising that some participants capitalized on the opportunity to extend the breastfeeding period for longer than was initially anticipated.

I ended up breastfeeding till (baby) was over a year old. One of the things about breastfeeding that I loved was the fact that I could eat more and still not get fat. That was one of the things that kept me hooked in. (Louise, age 29, one child)

Extending breastfeeding for longer than originally intended, with the concomitant suspension of some of the more extreme practices associated with the eating disorder, provided a rare opportunity for women to experience a relatively stable body weight and size in the absence of purging and vomiting.

Breastfeeding as a coping strategy and a delaying tactic

Although breastfeeding did prolong the suspension of binge‐purging behaviours in most participants, this was not always easy. A significant number of pregnant women volunteered that they struggled with not restricting their food intake and/or ceasing to self‐induce vomiting and some admitted that they failed to completely stop these practices. The following quotations are illustrative:

I kept trying to stop eating, to go on a diet or something, but I couldn't really do it with the breastfeeding. When he was about 5 months old I went through a really bad period where I didn't feel like eating and of course that affected my milk supply quite a bit. (Patricia, age 25, one child)

I didn't need to make myself sick so often (when breastfeeding) but that wasn't because I didn't want to! [Laughs] I had to fight with myself all the time to control the urge. I thought breastfeeding would take that urge away but it didn't. It eased a bit but I was still vomiting all the time I was breastfeeding. (Lorraine, age 28, two children)

Some participants drew parallels between pregnancy and breastfeeding and self‐imposed restrictions on ‘eating properly’:

I had to eat properly when I was breastfeeding because I had a baby to think about. The baby needs nutrition. I thought whatever I eat the baby is going to get it. So I had to eat properly. Like when I was pregnant I made myself eat properly. (Maureen, age 38, two children)

As suggested, a significant number of women volunteered that they had been motivated to breastfeed because they understood that this would effect early, and substantial, weight loss. Indeed, many women reported having been told this by midwives and some were upset when they did not lose the weight they had gained in pregnancy at the rate they anticipated. There appears to have been surprisingly little research undertaken on the subject of maternal weight loss during lactation but the results of available studies have failed to demonstrate any consistent relationship between mode of feeding and postpartum weight loss (Lawrence & Lawrence 1999; Walker et al. 2005). A few women stated that they would have ceased breastfeeding much earlier in order to resume their binge/purge activities had they known that weight loss did not automatically follow from breastfeeding. As is attested by quotations from participants throughout this paper, however, breastfeeding women often reported difficultly in balancing pressures to eat ‘healthily’ in order to produce ‘healthy’ milk, while simultaneously denying themselves access to familiar coping mechanisms, such as restricted eating, bingeing and purging and/or strenuous exercise.

Breastfeeding as a signifier of good and competent mothering

Breastfeeding appeared to act as a powerful reinforcer of maternal competence and nurturing ability. It also acted as an external signifier that, despite the eating disorder, participants were endowed with a ‘good’ maternal body (Stearns 1999) and were therefore embodied with the necessary expertise and capacity for their new role. Breastfeeding, then, seemed to assuage some of the guilt women carried throughout pregnancy for their eating disorder to potentially ‘damage’ the baby.

It wasn't my instinct to want to breastfeed him but in the end I did. In some ways it made up for all the damage I thought I'd done to him because of my eating disorder. (Fiona, age 24, one child)

I'm glad I did (breastfeed). It's the one thing good I did for him. (Patricia, age 25, one child)

Breastfeeding was also seen as a ‘fair exchange’, signifying a degree of reciprocity between the eating disordered mother and her infant:

When I was breastfeeding it was like well, I'm holding off bingeing and vomiting and I'm eating well so I'll make the milk he needs. He's eating the calories so I'm not putting on any weight even though I'm eating way more than I normally would. So in a way I'd say it was a fair exchange really. [Laughs] (Carlene, age 36, one child)

From a Foucauldian (Foucault 1991) perspective, biomedical constructions tend to frame breastfeeding in terms of (milk) production; as a means of supplying the infant with the necessary nutrients to achieve efficient outputs or developmental ‘milestones’. Such constructions reduce breastfeeding to an agendered, purely physical, process which emphasize the more mechanical elements involved in the transference of milk from mother to infant (Ryan 1998). Advocates of breastfeeding also tend to emphasize the benefits to either the mother or the baby but many women in this study stressed the relational aspect of this activity which some understood as an exchange based on mutual need and symbiotic association.

There's that kind of connection [. . .] which is very precious [. . .] I carried on (breastfeeding) as long as I could not necessarily because the breast milk was more nutritious [. . .] but because of the intimacy of that relationship, because of the bond. [. . .] there's still a very necessary and vital connection, which we've established through the breastfeeding. (Maureen, age 38, two children)

Unlike some mothers in this study whose motivations for breastfeeding were driven more by consumerist values and a need to be seen as ‘performing’ (Shaw 2004, p. 99) a duty until such time as they could reclaim their bodies for themselves, participants such as Maureen valued the pleasurable, sensory and intimate aspects of breastfeeding.

Satisfying the baby

In common with many breastfeeding women, some participants in this study experienced difficulties with ‘satisfying’ the baby. It is suggested that this is unsurprising given women's own troubled relationship with satisfying their own needs, especially those which are food‐related. On the occasions women sought advice from health professionals for such problems, they did not always find it helpful:

He'd just cry and cry but I couldn't satisfy him. He didn't seem to be getting enough from me. The health visitor told me to increase my fat intake to see if that would help. I felt really guilty but I couldn't do that. I'd put on so much weight in pregnancy already there was no way I could do that. (Emma, age 31, one child)

She (baby) started losing weight and I panicked. The health visitor came and said ‘Get some Mars bars down you’– which of course I wasn't going to do. But it was just a glitch. It was just for a week where she didn't put weight on. I'm glad I didn't listen to the health visitor or I'd have been back into bingeing and vomiting. (Lorraine, age 28, one child)

For various reasons, few women disclosed their eating disorder to a health professional involved with their maternity care and it is possible that lack of knowledge may have hindered professionals in their desire to offer more appropriate, and individualized, advice.

Participants reported feeling intensely proud when their (breastfeeding) efforts resulted in weight gain in the baby (and continued weight loss in themselves); such positive feedback reinforced participants' commitment to continue. Breastfeeding also, of course, emphasizes the moral certitude inscribing mothering practices in many contemporary societies.

I remember looking at him growing and thinking that's totally down to me. He's growing totally because of me. I'm giving him all this nourishment and that's all he's getting and that's making him grow. It's very, very powerful feedback. It makes you feel important, like you're the main person. (Carlene, age 36, one child)

The approval thing was a big factor. Everyone was telling me how well I'd done to keep breastfeeding. All that approval made me feel really good about myself, and that I was being a good mother to (baby). I wasn't thinking negative thoughts about myself, I was feeling very positive really. (Lorraine, age 28, one child, 18 weeks post‐natal)

Positive feedback was especially important for women whose primary relationship with their own bodies tended to vacillate between extreme ambivalence and/or self‐hatred.

Weaning

Weaning is defined here as the cessation of breastfeeding. The age at which weaning occurs varies but it is likely to reflect local cultural norms and the socio‐political environment in which infant‐feeding practices occur. Weaning was widely reported as a difficult transition and participants were open and eloquent when describing their attachment to continuing breastfeeding, often for the simple reason that this delayed the inevitable, but nonetheless frightening, confrontation with their eating disordered personae. Delaying the cessation of breastfeeding enabled women to sustain their belief in their ability to adequately provide for the needs of their growing infant; it also provided an extended window of protection from re‐engagement with practices underpinning their eating disorder. That said, no participant volunteered that they delayed weaning in order to delay the responsibility they would face for providing (healthy) foods on a regular basis.

Many participants reported feeling pressured to wean their babies and such pressures were sometimes internally generated. As is evidenced by the following quotation, some participants feared that ceasing to breastfeed would trigger a swift return to their disordered eating practices:

[Baby] was about 10 months old and I was starting to feel really bad about it [breastfeeding]. I was feeling almost guilty. I was starting to wonder whether there was some unhealthy pattern developing. You know like was I just breastfeeding because I was so afraid of what would happen with my eating when I stopped. (Louise, age 29, one child)

It is interesting to consider Louise's concern that breastfeeding her 10‐month‐old baby may be ‘read’ as indicative of an ‘unhealthy’ (psychologically dependent) pattern developing, when this is compared with the damaging consequences to her own health in maintaining her eating disorder. Her statement does, however, support psychological theorising (Chassler 1997) which suggests that eating disordered women may suffer greater disturbance in the relationship they have with themselves and their own physical bodies than they do with external objects – including perhaps their own children.

A desire to reclaim the (pre‐pregnant) body in order to resume regimes which had previously been employed to control and shape the body, was a significant motivator for some participants ceasing to breastfeed:

I wanted my body back and I knew I wouldn't get it back until I'd stopped breastfeeding. I knew the minute that I stopped feeding him I could control my food again and that's what I wanted. When I was feeding I needed to eat properly because he needs the nutrients. (Lorraine, age 28, one child)

Concerns were expressed, however, that weaning would inevitably be accompanied by personal weight gain:

I was very conscious about weaning him because I knew he wasn't going to be taking out as much from me so of course I was worried I'd put on weight. But at the same time I was glad that I wouldn't have that responsibility any more of having to feed him and having to eat properly. (Maureen, age 38, two children)

Breastfeeding and weaning may thus be understood as enormously conflicted activities which required participants to adopt an entirely new regime of eating (‘healthily’ and ‘regularly’) without recourse to purging. Many volunteered that this was the first time in their adult lives that they had accomplished this goal.

Some participants reported that they waited for a signal from the baby before they initiated weaning:

I'd rather her [baby] put an end to it than me. It's quite a big thing for me to have to do so I'll wait for her to do it. (Maureen, age 38, two children)

Given participants' expressed needs to control the circumstances underpinning decisions about food and related matters, it is suggested that the ability to relinquish decision‐making control to the baby represented a significant accomplishment. Further research is needed to establish the impact of such decisions, including whether the autonomy conferred on the baby had an enduring, and positive, effect especially on new mothers.

The majority of women weaned their babies before returning to work. Leaving the protection of home and the routine of caring for a new baby confronted women with a new set of dilemmas about self and body image. Many women, especially those who were still leaking milk, or who had not lost the weight they had anticipated, or did not fit their pre‐pregnancy (work) clothes, reported feeling distressed about the prospect of returning to the work environment. These feelings appeared to be particularly pronounced in women working in the retail, fashion, or leisure industries:

I breastfed [baby] for about six weeks because at that point I was back at work and I needed to look the part. In my job (fashion) you're a size 10 or less no matter what. Whether you've just had one baby or 10 babies you're never more than a size 10! [Laughs] (Emma, age 31, one child)

I had to go back to work (as a fitness instructor) when [baby] was three months old so I had to stop breastfeeding about a month before that so I could get back in shape. But I was still terrified I'd start leaking milk or something in the middle of a class. [Laughs] (Judy, age 34, two children)

Participants who felt pressured to return to work early in the post‐natal period quickly resumed the customary regulatory activities over body weight and shape which they had ceased, or at least reduced, during pregnancy and early lactation.

The prospect of weaning brought into sharp focus a return to the denying, and restrictive, attitudes embodied in the ascetic practices which had previously supported women in maintaining their eating disorder. Impending change prompted an internalized sense of increased size, or ‘bigness’ even if this was not reflected in an alteration in actual physical size:

As weaning got closer I was starting to feel big again because I'd eaten lots of things regularly that I wouldn't normally let myself eat unless I was going to vomit. I don't think I was any bigger but I felt I was. (Lorraine, age 28, one child)

For the majority of participants, the cessation of breastfeeding was associated with a resumption of the practices associated with the eating disorder. Although a significant number of women reported a reduction in the intensity of such practices, a few women reported the opposite:

I was worried about getting fat when I stopped (breastfeeding). I was worried about putting weight on because I wasn't burning the calories up. My eating really got worse after I stopped breastfeeding. For a while it was worse than before I got pregnant. (Tina, age 32, one child)

Some women reported that they felt pressured to wean an ‘older’ baby because of externally generated pressures including social norms which proscribe breastfeeding children beyond a certain age and/or beyond a certain size.

I did find it quite difficult after (baby) was about five months. People were starting to say ‘Still feeding then?’[. . .] I stopped (breastfeeding) him at about six months because he felt like he was getting too big and it didn't feel right any more. People kept staring when I was feeding him outside and I just felt more and more uncomfortable. (Patricia, age 25, one child)

The worst thing was when he got a bit older (nine months) and people started to look at you and you know they're thinking ‘You breastfeed new‐borns, not toddlers’. (Louise, age 29, one child)

Recent guidelines from the World Health Organization and United Nations Children's Fund currently recommend exclusive breastfeeding for a minimum of 6 months and to continue this practice for at least 2 years while simultaneously offering weaning foods (World Health Organization and UNICEF 2003). Prevalence rates in the UK currently fall well below these targets, however, with only 21% of women still breastfeeding when the baby is 6 months old and only 13% continuing at 9 months (Hamlyn et al. 2002). Although a sustained improvement in breastfeeding initiation has been reported (Hamlyn et al. 2002), official statistics confirm that childbearing women in the UK remain deeply mistrustful in the efficacy of their bodies to nurture and satisfy the food‐related demands of their babies (Dykes 2005).

Given this backdrop, it is of some concern when eating disordered women, whose adult corporeal experience has been so tenuous, report that negative attitudes impinged upon their desire to continue breastfeeding their growing infants. Despite the public approbation experienced, however, some women reported feeling greatly strengthened by negative interactions. In the following quotation, Patricia volunteers that criticism would previously have triggered a change in her behaviour but her determination to succeed with breastfeeding deflected any disapproval, whether this was implied or real.

But I didn't care what anyone else said (about breastfeeding her toddler in public). And that was a surprise to me because normally I do care very much what people say about me. [Laughs] (Patricia, age 25, one child)

Breastfeeding an ‘older’ infant in public was perhaps less threatening to Patricia's sense of self‐identity because this activity was congruent with her role as a mother – as opposed to her objectified, and eating disordered, identity as a sexually scripted female. Her unconcerned attitude about revealing a part of her physique which, in any other circumstance would normally have shamed her, suggests a considerable degree of body confidence. That she is able to permit her body to engage with her infant in the regular production of body fluids which she understands to be useful, and of which she feels proud, is perhaps in direct contrast to the fluids she ejects from her body in response to her eating disorder.

Some women reported being pressured to wean by male partners:

I didn't want to stop (breastfeeding). [Partner] kept saying ‘When are you going to stop?’ I think he wanted me to stop because he wanted me back so when [baby] was just over a year old I pushed myself to stop. (Louise, age 29, one child)

Regardless of whether the motivation to wean was internally or externally derived, extending the feeding period beyond the time originally planned was widely reported. Weaning an ‘older’ baby was generally experienced as a more difficult undertaking because participants had become more accustomed to eating an extensive range of food items, many of which would be outlawed altogether or consumed only during bingeing episodes.

Discussion

This study identified some new, and significant, influences on decision‐making processes which eating disordered, childbearing women, employ with respect to infant feeding. Women are generally understood to involve their male partners in formula feeding in order to share parental roles and hence move towards a more equitable division of responsibilities. Formula feeding in this context may be understood as ‘a somewhat covert shift in the sexual division of labour, as involving the father in parenting, by beginning with its most gratifying aspects’ (Maher 1995, p. 8). The findings from this study suggest, however, an alternative reading of men's role in this activity.

Male partners who were aware of their partner's eating disordered status were generally reported by participants as being extremely concerned about the welfare of the baby and, while there was little they could do about this during pregnancy, some acted on their concerns once the baby was born. Insisting on the baby being formula fed was one strategy adopted by male partners who understood this to be the only possible means to guarantee the quality and quantity of nutritional input.

Breastfeeding was identified as a strategy which enabled participants to circumvent customary practices associated with their eating disorder and to reassign their identities as ‘good’, as opposed to ‘neglectful’, mothers. Breastfeeding, which was widely construed as positive and beneficial to infant welfare, permitted women to increase their food intake, particularly their consumption of previously forbidden ‘naughty’ treats such as ice cream and chocolate. While formula feeders were not averse to the idea of breastfeeding, they were unable to countenance the prospect of a delayed return to the project of reclaiming of their pre‐pregnant body. Breastfeeding was generally believed to be incompatible with the re‐enactment of eating disordered behaviours because the production of ‘good’ quality milk, in sufficient amounts to nourish a growing baby, could not be guaranteed. To that end, formula feeding was viewed as the only alternative solution for a number of participants.

Although breastfeeders in this study greatly outnumbered formula feeders, both groups of participants were very concerned to be seen doing the ‘right’ thing; indeed, rationales advanced for infant‐feeding decisions by all participants revealed a desire to protect infant well‐being. Women who had disclosed their eating disorder were apprehensive that knowledge about their (mental health) status might render them vulnerable to criticism and/or instigate unwanted surveillance from health and allied professionals. Some used breastfeeding to deflect possible censure and to maximize the opportunities for demonstrating ‘good’ mothering. In this sense, breastfeeding may be conceptualized as a ‘performative accomplishment which the mundane social audiences, including the actors themselves, come to believe and to perform in the mode of belief’ (Butler 1999 [1990], p. 179).

The practice of breastfeeding seemed to afford eating disordered mothers an opportunity to experience their body in a very different way. By discovering resources within themselves which were essential for the well‐being and development of the baby, participants were able to construct alternative representations of themselves. Hence, the maternal body metamorphosed into a figure of pride and accomplishment but one which, nonetheless, retained the capacity for holding deep‐seated and enduring feelings of shame and embarrassment. It is suggested that breastfeeding demonstrated a particular, and highly specific, experience of mutuality and interdependency which enabled women to see themselves not simply as a ‘fat’ maternal body but as body which was performing an essential function.

It could be argued, however, that breastfeeding encouraged women to disassociate from their (eating disordered) bodies and to foster a belief that the pregnant or lactating body was ‘other’ than their own. Future research is needed to investigate whether eating disordered mothers who breastfeed develop a more fluid identity which enables them to embrace contradictory messages while simultaneously retaining their sense of personal agency.

This study contributes to research on the subjective experiences of a particular group of women who acknowledged a chronic, and problematic, relationship with their bodies. Negotiating the transition to motherhood required study participants to confront their own, often longstanding, disrupted eating patterns and to make important decisions about infant‐feeding methods. Women's solutions to the difficulties they encountered were creative and idiosyncratic attesting perhaps to the ‘cross‐currents of complex and sometimes contradictory obligations, which mean that infant‐feeding decisions are as much about morality as they are about nutrition’ (Murphy 1999, p. 206).

Implications for health professionals

The establishment of an agreed pathway of care, which is properly funded and adequately serviced, is urgently required to facilitate the appropriate referral of women who disclose a history of an eating disorder. Such a facility would also support midwives and other health professionals involved with providing care to childbearing women.

Participants' who breastfed (rather than bottle fed) appeared to develop a more fluid identity – at least during the research period – and this seemed to enable some women to embrace contradictory health‐promotion messages, while simultaneously retaining a strong sense of personal agency. Health professionals need help to recognize this window of opportunity for eating disordered women to maximize the opportunities provided by the behavioural changes adopted as a result of pregnancy and motherhood.

Limitations of the study

The study did not seek to involve health professionals, women's partners or other family members and hence their perspectives, while sometimes alluded to by participants, were not available for independent analysis. As a significant proportion (n = 6) of participants had never disclosed their eating disorder to their partners, and an even greater number (n = 9) did not disclose this aspect of their history to maternity staff, however, it is difficult to imagine how the views of these groups might have been accessed.

A second consideration is that the study sample was recruited primarily through NHS facilities providing medically orientated treatment for women with eating disorders. The majority of participants were currently enrolled in, or were on the waiting list for, psychotherapeutic programmes including group, and cognitive behavioral, therapy. Additionally, most participants had been prescribed antidepressant and/or antipsychotic medications at different points in their eating disordered career ⁵ and it is suggested that prolonged exposure to medical practices may have shaped their views in particular ways and rendered their responses to interview questions more uniform than might be expected in a non‐medicalized sample.