The significance of the father-daughter relationship to understanding and treating Bulimia Nervosa: a Hermeneutic Phenomenological Study

Abstract Bulimia nervosa (BN) is a highly researched eating disorder, yet real world recovery rates remain poor and incidence continues to rise. This study provides a focused exploration of the father-daughter relationship where BN emerges, in order to explore this relationship’s significance to the aetiology of BN and to BN’s resistance to CBT-based treatment. A hermeneutic phenomenological study of six women in recovery from BN was undertaken. Unstructured interviewing gathered detail-rich information, which was interpreted using the hermeneutic phenomenological method of multiple-level repeated readings, thematic comparisons and contextualisation. Findings were confirmed and validated using the hermeneutic circle and peer consultation. Fathers of daughters with BN were found to be a source of fear, control, abuse, emotional and physical avoidance and gender diminishment. This was a key source of complex traumatic experience in the family setting, with BN emerging in daughters to provide distraction and soothing. Furthermore, BN acts as a survival mechanism from early childhood and is a logical embodied response to the lived experience of complex trauma. The presence of trauma in the aetiology of BN, makes sense of why cognitive-based therapeutic protocols provide for limited treatment success. The research suggests greater potential lies in adopting the individualised, multi-modal complex trauma treatment model for BN, as this more appropriately addresses outcomes of relationship trauma.

treatment model for BN, as this more appropriately addresses outcomes of relationship trauma.

Introduction
To date, bulimia nervosa (BN) has been the subject of extensive research by academics, the medical community and practitioners. The individual with BN equates self-worth with body size, weight or shape. This results in an obsession with losing weight or avoiding weight gain, ultimately leading to the adoption of a cycle of bingeing on objectively large quantities of food, followed by compensatory action such as vomiting, compulsive exercising, ingesting laxatives and diuretics, or fasting. Once this cycle is established, it becomes resistant to treatment in most cases, regardless of worsening physical consequences and psychological pain (Yu et al., 2013).
Historically, it has been thought that around 90% of all those with eating disorders (EDs) were female; but research over the last 40 years or so has revised that view and it is now estimated that around 25% of those living with BN may be male (Murray et al., 2017). It appears that the gender bias attributed to BN, and all EDs, is the result of applying too narrow a lens to detection and reporting; and adhering to a feminised social construction of EDs in general (Sweeting et al., 2015).
Without dismissing the variously gendered reality of BN, the subject for this study is specific to the father-daughter relationship, as it affects the aetiology and treatment of BN in females. This fills a gap in the extant body of knowledge and answers the call of previous researchers (Attili et al., 2018;Horesh et al., 2015;Berge et al., 2014;Gale et al., 2013;Jones et al., 2006;Johnson et al., 2002;Rorty et al., 2000) to explore this specific dynamic, which is known to be significant, but remains under-researched.

Importance of the family in BN
Existing research has revealed the primary importance of family dynamics in the aetiology of BN (Lecomte et al., 2019). BN tends to emerge in families where nurturing by caregivers is poor and parenting style involves heightened levels of conflict and control (Hampshire et al., 2022;Jáuregui Lobera et al., 2011;Stice & Van Ryzin, 2019). In response, bulimia nervosa develops in offspring as a means of emotion regulation (McEwan & Flouri, 2009), where exists a profound fear of vulnerability, driven by pervasive low self-esteem. The push-pull dynamic of BN absorbs and distracts the sufferer from these underlying emotional challenges (Meule et al., 2021).
BN is also often accompanied by other difficulties, such as substance abuse, self-harm or alcoholism (Carbaugh & Sias, 2010;Duncan et al., 2006;Sagiv & Gvion, 2020). These can emerge as attempts to ameliorate the symptoms of BN, or as additional strategies to help regulate emotion. In these cases, treatment can prove particularly challenging (Gregorowski et al., 2013).
Previous research has highlighted the significance of attachment difficulties in such families, with studies showing that up to 100% of those with BN experience insecure attachment (Tasca & Balfour, 2014). These attachment challenges can result from poor parental dyadic adjustment (Espina et al., 2003) leading to stressors such as triangulation or parentification. Moreover, parenting in families where BN arises tends to be critical, and emphasis is given to the importance of appearance and achievement (Goodman et al., 2014;Haworth-Hoeppner, 2000). This can leave daughters unable to internally justify their own validity: aspiring to unrealistic standards for themselves and adopting perfectionistic thoughts and behaviours (Striegel-Moore & Bulik, 2007).
The discourse in these families tends to place significance on body shape and weight in relation to success and personal value, which negatively impacts development of self-esteem (Stoeber & Yang, 2015). Further, the poor attachment may be an outcome of parental psychopathology, such as the presence of alcoholism or depression (Anastasiadou et al., 2014), which can contribute to inconsistent or unreliable parenting. The dynamics in such families are complex, however, and it is known that the presence of an ED in a member may affect the behaviour of other members in response (Hedlund et al., 2003).

Significance of paternal parenting in BN
To date, there are researchers who have concluded that fathers play a pivotal role in families where EDs emerge,but their studies do not focus on fathers specifically (Haslam et al., 2008;Johnson et al., 2002;Miller-Day & Marks, 2006;Penelo et al., 2011;Rorty et al., 2000). Current knowledge of the impact of fathers in the aetiology of BN includes fathers engaging in unwanted sexualised behaviour (Castellini et al., 2012); exerting overprotection and unreasonable psychological control, thus preventing their daughters developing autonomy (Gale et al., 2013); presenting unstable, fickle personality traits, which contributes to a daughter's internal stress (Fassino et al., 2009); and being highly critical, poor at listening, emotionally withdrawn and passive (Botta & Dumlao, 2002). Bullying is also a risk factor for the development of eating disorders in general (Lunde et al., 2006) and weight-based teasing can lead to the desire to be thin (Rojo-Moreno et al., 2013). Fathers have been found to specifically play a significant role in modelling behaviour around teasing; and where fathers routinely use teasing, siblings are more likely to follow (Ata et al., 2007). Penelo et al. (2011) have asserted that it is fathers' unhealthy attitude towards food that plays the most significant role in the development of BN;and May et al. (2006) concluded that the correlation between fathers, conflict and weight concerns amongst daughters requires further attention by researchers.

Treatment challenges
The current status of help for BN is that treatment success for those seeking long-term recovery remains elusive for the majority, particularly amongst adolescents and young adults (Hail & Le Grange, 2018). The protocol most often followed in the USA, Canada, Germany, the UK and Australia as a first-line treatment for BN, is to offer Cognitive Behavioural Therapy (CBT), CBT-E (eating disorder focused) or TF-CBT (trauma focused), with or without medication. Even in the UK, where family-based treatment for BN in adolescents is recommended as a first option (National Institute for Health and Care Excellence, 2017), the fundamental building blocks of family treatment are CBT-oriented.
These protocols have achieved dominance as a result of the significant contribution made by CBT practitioners and researchers to the body of quantitative, evidence-based literature on BN (Atwood & Friedman, 2020;MacDonald et al., 2021;Pellizzer & Wade, 2016). However, the paradigm of the predominant conventional medical model preferences evidence-based research over qualitative research designs or anecdotal, practitioner reporting, so this has arguably resulted in an obscuring of the potential of other treatment approaches informed by alternative value systems that do not fit into the current evidence-based research model. Importantly, despite increased awareness of eating disorders amongst the mainstream, and the widespread availability of CBT and medication as a treatment protocol, incidence of BN continues to rise (Galmiche et al., 2019).
The theoretical justification behind the choice of CBT and medication is that if mood is stabilised and maladaptive thinking adjusted by testing out new behaviour, recovery becomes possible. However, in reality recovery is not a state encapsulated by one universal definition, but personal to each individual (Kenny et al., 2020), so research outcomes can be misleading; and the experience of enduring recovery, as personally defined and including symptom management, has been suggested to be as low as 15.5% when treated by this method (Yu et al., 2013). In fact, CBT has been criticised for some time by numerous sources for its inability to help effect meaningful positive long-term change (Keski-Rahkonen et al., 2009;Raykos et al., 2013;Steinhausen & Weber, 2009;Turner et al., 2015;Walsh et al., 2000). One concern about reported successes with CBT is that those with eating disorders have elevated levels of perfectionism (Bardone-Cone et al., 2010) and can be highly compliant during treatment. Patients may report reduction in symptoms short-term, or in order to please the therapist, yet these "gains" fall away once treatment ceases. This emerges as a result of the attachment insecurity present in the one with symptoms of BN.
Since the father-daughter relationship is known to be significant, but under-researched, it was hypothesised that there may be insights arising from a closer examination of this relationship that could help explain BN's resistance to cognitive-based treatment and perhaps offer direction for treatment options with greater hope of success.

Towards a research design
To date, the vast majority of studies into BN have adopted quantitative research designs. Moreover, they have largely avoided a specific focus on the father-daughter dynamic. There is also limited qualitative research conducted on resistance to treatment in the eating disorder field, despite it being known that subjective meaning-making amongst those seeking help plays a role in motivation to attain long-term recovery (Abbate- Daga et al., 2013). The limitation, therefore, arising from the existing body of knowledge, is that whilst BN has been mapped to a large degree, recovery rates from BN remain poor and current research has not yet facilitated the move from knowledge of the problem to its solution.
This research stands in deliberate contrast to the majority of extant studies and seeks to fill a gap. In order to uncover new understandings of BN and provide greater clarity into the development of BN and its resistance to treatment, this study uses a qualitative methodology to explore the father-daughter relationship where BN has arisen in the daughter. Hermeneutic phenomenology was selected as the most appropriate methodology for this study because of its ability to capture detail-rich information and delve deep into the phenomenon. It emphasises and values the subjectivity of knowledge, which allies respectfully with the subjectivity of relationship experience and of recovery from BN.

Method
Hermeneutic phenomenology, as drawn from the writings of Heidegger (2013) and Gadamer (2013), is an interpretive qualitative methodology, based on the assumption that there may be multiple, contingent realities (Laverty, 2003). It seeks to capture the meaning-making in experience, taking into account context and time (Tuohy et al., 2013). It requires in-depth conversation between researcher and participant, because language is considered to be the key to understanding experience (Gadamer, 2013). Research rules and structures are kept to a minimum with this method (Ricoeur, 2016). This allows the researcher to engage with participants' viewpoints and for the most appropriate interpretation of experience to surface.
The methodology seeks to maintain the context of experience. BN, as a set of behaviours, thoughts and consequences, is situated in, not divorced from, its world; and it was hypothesised for this study that the father-daughter dyad had the potential to be contextualised in ways specific to each participant. Objectivity and measurement are not the goals of this methodology. In fact, it aims to uncover meaning overlooked by "scientific" methodologies that seek to objectify lived experience (Plager, 1994). Subjectivity is sought out and embraced, whilst also acknowledging the researcher's impact on what emerges. Since no-one sits outside of a context, researcher reflexivity is part of the research process.
Hermeneutic phenomenology requires approaching another's experience with an open mind (Finlay, 2012). It does not look to generate or test theory, or provide generalisable results, because the level of interaction necessary between participant and researcher implies a small sample size. Instead, it seeks to produce a defensible account and clear understanding of the phenomenon under research, through deep investigation and interpretation (Brown et al., 2003).
Interpretation emerges from analysis of textual information. Dialogue between researcher and participant is captured via unstructured interviews, which are then transcribed into texts. These texts are not considered factual accounts, but initial, encultured, transformations of experience into language, which may then be reflected upon (Gadamer, 2013). Interpretations are derived from repeated readings set against relevant contexts.
A hermeneutic phenomenological methodology does not prescribe a method, but it does provide guidance. The method for this study was guided by the six steps detailed by Van Manen (1990). Thus the father-daughter relationship, as expounded by the participants and captured as text was subject to a multi-faceted involvement by the researcher (reading, researching, talking, listening); deep reflection on its essential themes, both in constituent parts and as a whole; writing and rewriting in order to develop an understanding of the phenomenon; and use of the hermeneutic circle, in which participants are given the opportunity to provide feedback regarding the validity of the researcher's interpretations.
For this study, a sample of six was recruited using the snowball method. A small sample size is common with hermeneutic phenomenology (Ferch, 2000;Kierski, 2014;Nourian et al., 2016;Rossetto, 2012) as this facilitates in-depth gathering of information. All those approached to participate keenly agreed to do so and provided informed written consent. Since this study focused on the father-daughter relationship, an all-women sample was selected.
Due to ethical considerations, all participants were over the age of 21 and had a minimum of 18 months' continual recovery from BN. The definition of recovery was considered to be a cessation of the binge/compensation cycle and significant reduction in the obsessive thinking about body shape and weight together with its negative impact on self-worth, as experienced on an individual, self-assessed basis. For an individual to consider themselves in recovery, complete removal of all aspects of BN does not appear to be necessary, as recovery is a process and not a fixed point (Cogley & Keel, 2003). Moreover, some remaining measure of negative evaluation of self in relation to appearance is in line with the broad-brush response to exposure to American and European media influences experienced in any population (O'Garo et al., 2020).
All participants spoke fluent English in order to share the same language as the researcher. All participants had grown up in households where both the biological mother and father had been present throughout childhood. This allowed for sufficient contact with father to allow ample gathering of detailed information on the relationship.
None of the participants were active in any other eating disorders or addictions and all participants had to be active members of 12-Step food fellowships, such as Overeaters Anonymous, in order to ensure all participants had established networks of support. This was because the subject matter of the interviews was considered to be sensitive.
Each participant was interviewed for one hour. Interviews were conducted face-to-face and audio-recorded. Participants were offered the choice of whether to be interviewed in their homes or in a consulting room. This was to reduce the balance of power inherent in the interviewer/ interviewee relationship. Four out of the six women chose to be interviewed at home. Interviewees were encouraged to provide detail, anecdote and examples to illuminate their accounts. Any follow up questions were conducted by telephone call. All the participants selected their own pseudonyms by which they are referred throughout the research.
As recommended by Kahn (2000), field notes were written by the researcher after each interview, noting elements such as body language and home surroundings that could not be captured by audio-recording. The researcher also maintained a reflexive diary throughout the research process, as recommended by Sloan and Bowe (2014), noting pre-judgements, observations, hunches, thoughts, reflections and points of identification with the participants. A hermeneutic phenomenological approach regards the participant as the expert in their experience (Morgan, 2011), but it is recognised that knowledge is co-constructed with the researcher through the interview (Brinkmann & Kvale, 2015). A further source of information brought to the interpretive process was research conducted post-interview around cultural references raised by the participants during their interviews, such as music, books and religious matters.
Once the interviews were transcribed into texts, they were read and re-read, searching for detailed understandings of the father-daughter relationship in the lives of the participants. Van Manen (1990) recommends three approaches to reading texts: wholistic, selective and detailed, moving between readings to uncover insights and themes. Interpretations made by the researcher in response to the textual readings were contextualised using all the information gathered during the research process.
Validity and reliability are not terms closely associated with this methodology, but the issues of research rigour, scholarly integrity and credibility of findings remain essential and pertinent. Whilst generalisability in the quantitative sense is not possible, Moules et al. (2015) suggest that transferability is a term with a better fit for hermeneutic research. Van Manen (2014) and Stenbacka (2001) agree that despite statistical generalisation not being possible, analytic generalisation is, in terms of recognising recurring aspects of the meaning of a phenomenon. There is a vital requirement for research of this kind to be reflexive and transparent, as well as for evidence to be credible and interpretations recognisable to others. For this study, the methodology and method were painstakingly laid out and justified; the researcher's preconceptions made overt; findings discussed with coresearchers; and interpretations shared with co-researchers and participants prior to final writing. This is referred to as the hermeneutic circle (Heidegger, 2013), which allows for feedback to act as a contributor to the robustness of the research outcomes.
The hermeneutic phenomenological methodology facilitated a deep and focused examination of the father-daughter relationship in families where BN had emerged and a raft of new findings were able to surface.

Results
Of essential interest here are two themes in the findings. The first is the prominence of complex traumatic experience in the aetiology of BN, arising from dynamics in the father-daughter relationship. That is, father is the key perpetrator of repetitive, distressing, threatening interpersonal experiences throughout his daughter's childhood. The second theme is that BN emerges as a survival strategy: a functional solution to the fear and insecurity inherent in growing up in a home where traumatic experience is a quotidian reality. These findings provide a clear direction for treatment, which will be detailed in the discussion.

Fathers may be a source of fear
Existing literature on the aetiology of BN, points to BN arising in households that are more critical and conflictual than most (Botta & Dumlao, 2002;Jáuregui Lobera et al., 2011;Kent & Clopton, 1992). Where this study adds insight, is to uncover a more extreme dynamic at play: in the households of five of the women in this study, genuine fear was a persistent presence throughout childhood, and father was the key source of that fear.
Anne described her home as a frightening place. Her father's moods would change very quickly and she described living " . . . on tenterhooks a lot of the time". The first question she would ask on returning home from school was whether her father was in a good mood or a bad mood, because she feared his "outbursts of emotional intensity" that could be triggered by something as innocuous as the rustling of a newspaper.
Claire also grew up in fear. She likened her father to Dracula and described him thus: " . . . angry, he raged, volatile, inconsistent, and just explosive, explosive fury". Her father would smash dinner plates against walls; he killed the dog for "whining"; he beat Claire if she misbehaved; and she feared for her life when he was driving her in his car.
Kate's mother was an active alcoholic. Her drinking intensified in the years following the accidental death of Kate's youngest brother. Kate's mother's behaviour spiralled to the point where she would throw knives in rage. Kate's father failed to protect the children from this; and chose to leave the family home, abandoning the children to face their mother alone.
Louise's parents had a very intense and volatile dynamic operating between them. They were both alcoholics and quick to anger. Her father became increasingly violent when drunk and would attack his wife physically, something witnessed by Louise. In one particular incident, he hurt Louise as well.

Fathers may engage in the emotional abuse of their daughters
A strong theme running through the accounts of the participants was that their fathers played a role in denying their reality. As father was a primary attachment figure for these women, this action comes under the banner of emotional abuse. It is a common factor in cases of complex traumatic experience (Sanderson, 2006). In three of the households there was sexual abuse. In five of the households there was physical violence. In all these cases, the participants reported that their fathers had minimised what had occurred and in three cases that they had experienced blanket denial.
Claire remembered confronting her father as an adult, telling him about the impact on her of his behaviour. His response was not what she'd hoped for. "I . . . wrote out a list of things . . . how he had affected me . . . but he just disregarded it and said that it never happened . . . he doesn't know what I'm talking about".
Louise tried to talk to her father about the times he attacked her mother. "I saw you do this" she told him, "I saw it with my own eyes". Yet her father refused to acknowledge his actions. Louise found it deeply upsetting that her father seemed to be protecting himself at her expense.

Fathers may hold traditional values that marginalise their daughters' attachment needs
It appears from this research that where fathers rank gender roles in a hierarchy and view the female gender role as lower down the hierarchy than the male gender role, that this may contribute to the development of bulimia in daughters.
The women in this study encompassed a broad band of ages, from early 20s to late 50s. Regardless of their generation, however, they all grew up in families that clearly demarcated traditional gender roles and devalued the female one. This appears to have caused significant attachment disruption between father and daughter, noticeably affecting the daughters' selfesteem and feelings of safety.
Rebecca recalled, "there's rules for boys and there's rules for girls". For Rebecca, this meant being compelled to help cook dinner, whilst her brother got to sit at the table with her father talking as they waited to be served. She realised early on that she would be treated differently from her brother and this caused her a good deal of emotional difficulty. Rebecca remembered crying about it and telling her parents that it was not fair. Time with Dad was a rare and precious resource and Rebecca felt her allocated role as a girl meant she was side-lined and deprived of the attention she craved from her father.
Rebecca's father worked in women's fashion and had strongly held attitudes about how women should look and what body size was preferable. Rebecca described her father as "a hypocrite" in this regard because he was overweight himself with "a big tummy", yet she was adamant that growing up "I knew that I couldn't be fat". There seemed to have been quite a lot of discussion around women's bodies at home, in terms of who was and who was not "fat", and Rebecca grew up believing that, ideally, women should be skinny, but men could carry extra weight.
Kate was often given the role of looking after her younger siblings and, like Rebecca, starting dinner before her mother got home from work. She remembered thinking "My brother never had to help. My dad never helped. I felt like a slave".
Preparing food was an activity that separated her from her father and because he didn't participate in it, Kate saw herself as of lower status and less deserving of privilege in her father's eyes than her brother. Connecting with her father was a real challenge and experiencing any barrier to closeness and approval was a blow to Kate and the support she craved.

Fathers' psychopathology may set the tone of family mealtimes
For all the women in this study, it was at dinner time, over shared meals, with the family gathered together, that trauma and dysfunction in the family system played out in full. It appears that the tone and direction of this was driven by their fathers' particular unresolved mental health concerns, such as alcoholism, depression or anti-social behaviour.
In Claire's case, dinner was a time when her father would turn his rage against any unfortunate member of the family. She remembered her disabled brother spilling a pea onto the floor and it resulting in a tremendous explosion of anger and her father storming out of the house.
Nina's memories of family mealtimes had all but disappeared, a lapse of memory she could not explain. It had been made clear repeatedly to Nina that she was not a family priority and neither were her food preferences. Her father was the main cook in the house, but he was an "orthorexic" who was very controlling around food. He could cause a tremendous fuss if Nina ate anything without his permission; and his obsession with "clean" eating meant the food lacked salt, and there was little prospect of sugar. Family members would serve themselves from shared plates where quantity was abundant, but the food rarely offered sensorial pleasure. Nina was left with a feeling of deprivation meal after meal, and that her needs had been ignored.
In Anne's emotionally unsafe home, meals were eaten in silence. At these meals her father, a depressed, unknowable and volatile man, typically would not show any interest in her. The focus of each mealtime was the television. "When I see families talking around the table, I think that's really odd. What are they all doing? Why are they talking?" she said. But for Anne, silence did not offer the assurance of safety, or even neutrality. Silence was often used by her father as a punishment, or it could be the precursor to long stretches of time exposed to her father's anger.
Louise's father was a larger-than-life character, an alcoholic, who could switch between being charming and loving, to violent and bullying. He encouraged Louise and her siblings to acquire a sophisticated appreciation of a variety of cuisines from a very young age, but Louise had adopted a vegetarian diet and made this known to her parents. One evening Louise's father erupted at dinnertime when she protested having to eat fish. I looked at him and I was like, 'Dad, I don't eat fish.' And he was like, 'Well, you're going to eat it.' And I was like, 'No, no, no. I actually hate it. I can't eat it.' And he shouted at me and made me eat it. And I remember sobbing and the cream of the pasta splattering everywhere because I was crying into it, and he made me eat it. And Dad, I mean, that just . . . it's traumatic.

Fathers consolidate their emotional unavailability by engaging in physical distancing
It is known that fathers of girls who develop bulimia are less likely to be emotionally available than those of girls who do not (Rorty et al., 2000). This study adds to that knowledge by revealing that physical unavailability may feature prominently in these relationships as well. In five of the father-daughter relationships reported in this research, physical distancing was used by the fathers to control their accessibility to their daughters.
Kate's father travelled for work and would be away for at least half of the year. Even when he was home, " . . . he just disappeared. He was either in his study or sailing all weekend, but he was absent". Kate attributed his physical unavailability to marital difficulties; and to his grief at the loss of Kate's younger brother, who had drowned age four. She suspected his emotional resources were depleted. For Kate, however, her father's physical absences were devastating, particularly in her adolescent years when her mother's anger and alcoholism were building in intensity. Kate was left in fear and with a sense of abandonment and rejection that she remembered soothing with bingeing, compensating and isolation.
In contrast, Anne's father was home a lot. He took care of her and her brother after school and was around at weekends and in the evenings.
Anne believed her father was suffering from PTSD and depression when she was growing up and that this influenced the amount of attention and emotional availability he had to offer her. But her father also carved out spaces for himself at home that were for him alone and into which no-one else was welcome. Whenever he wanted distance from the family, Anne's father would absent himself and go alone to one of these spaces. " . . . he would go off into his bedroom . . . or into one of his many sheds.
Anne's father would remain in his sheds for hours at a time, for long periods when the weather was mild, and Anne recalled it as "withdrawing". Once there, the understanding was that he was not to be disturbed, and life in the house would have to go on without him until he returned. He even had a piano in one of the sheds, which he would play alone.
Exploring the relationship between fathers who are perpetrators of complex traumatic experience and bulimia arising in their daughters leads onto the second theme of findings: BN as a survival strategy. The research revealed insights into how BN functions to resolve the fear and insecurity inherent in a childhood shaped by complex traumatic experience and that far from being "disordered", BN is a logical, pragmatic solution to circumstances in which the daughter finds herself.

Bulimia begins in early childhood
Existing research has shown that BN enables distraction from psychological overwhelm and provides a way to regulate emotion (McEwan & Flouri, 2009;Mendes et al., 2017). This research provides compelling evidence to suggest that because of the presence of interpersonal trauma in the father-daughter relationship, the ability of BN to achieve emotional and psychological relief means it develops from when it is first needed: early childhood.
This challenges the DSM-5-TR's conceptualisation of BN, with its assertion that BN emerges almost always in adolescence (American Psychiatric Association, 2022). It appears from this research that the DSM-5-TR (American Psychiatric Association, 2022) is capturing a stage in BN's progression and may not be reflecting a comprehensive understanding of BN: as a purposeful condition that morphs over time.
Evidence of this was found in the participants' narratives, where they recalled feeling out of control around food and experiencing negative body image in early childhood. They only later discovered the binge/compensation mechanic described in the DSM-5-TR (American Psychiatric Association, 2022). It also appears that these negative feelings about their bodies were influenced by family dynamics.
Rebecca remembered having difficulties with and "abnormal reactions" to food at the age of five or six. Kate stated, "Look, my dysfunctional relationship with food, I can't ever remember not having it . . . " and continued, "I had a mental obsession around food from a very early age. I overate". She recalled feeling "chubby" compared to her siblings before the age of eight, despite being able to assess in retrospect that she was of normal weight.
Claire, too, recalled feeling overweight in primary school. "I would notice that the other girls were perhaps thinner than me, and I wanted to be thinner than I was". From memory, she believed she was feeling this way by the age of 10, years before she began to binge and compensate with exercise or vomiting. Claire theorised that it related to dynamics within her family at the time, including the success of a cousin who was modelling in New York. This attracted her father's admiration. Claire's father took photographs of her to send to modelling agencies; photographs Claire considered much later to be too highly sexualised. However, Claire never did achieve modelling success. Instead she was left feeling exposed and vulnerable at a young age in the adult world of work, to which her father had introduced her, but from which he had not protected her. She remembered eating sweet foods quickly and secretively: "there was always a feeling of just wanting . . . it was never sated, like never enough".

Bulimia is a tool for building resilience
Contrary to the medical perspective of BN as a mental disorder, the research provides strong evidence to support BN being viewed more usefully as a source of strength and an adaptive tool for building resilience in the face of emotional suffering.
It emerged that all the women in the research had endured childhoods where their needs had been inconsistently met. Their attachment bonds were insecure and home was not experienced as a safe haven. It has previously been documented that the binge/compensation cycle of BN has the ability to provide distraction from difficult emotions and serve as a form of self-soothing (Meyer & Gillings, 2004;Polivy & Herman, 2002). This research adds the finding that BN may be viewed as an adaptive way of shoring up emotional resources and increasing resilience in circumstances where parenting fails to do so. BN may be an indicator of strength, not weakness.
Five of the participants detailed triangulated relationships with their parents. For example, Kate described herself as "surrogate spoused" and that food and isolation provided her with respite from the war at home. Food for her provided a boundary and some much-needed nurturing. ". . .my go-tos were food, for sure, and isolation. I would go up to my room and lock myself in. . ." Anne also felt as though neither of her parents could be trusted with her emotional needs. She frequently found herself placed in the middle of her parents' marital disputes, on occasion being asked who she would live with in the case of a divorce. It became her role to protect herself and her younger brother from the violent raging that would intermittently explode in the house. She rarely let her parents witness her tears. Instead of the sweetness of nurturing parents, Anne discovered the sweetness of sugar, and used it to inure herself to the burden of her volatile childhood.
Louise and Claire both had experience of intervening in their parents' physical fights, leaving them both feeling unprotected, frightened and burdened with family secrets. They used to retreat and fill themselves up with food to block out their fear and sadness. Claire recalled her father's response to her showing him she was upset: "He would tell me to shut up . . . to get over it . . . ". She expanded and toughened herself up with excess food, and later distracted herself with purging and exercise. Claire didn't recall her father telling her he loved her until she was in her 20s.

Bulimia exists in the space between people
This research found that, particularly in the case of the father-daughter relationship, recovery from BN in the daughter led to significant changes in all her relationships, even where attachment previously had been poor. It seems that where she is able to change her relationship to herself, the person recovering from BN is able also to transform her relationship to food, to her family and to the world about her.
Anne and Nina found empathy, understanding and forgiveness for their fathers' past behaviour, through their personal transformation in recovery. Anne was keen to have this understood: " . . . when I tell people about my dad I say 'he wasn't all bad, that's the thing' . . . I can feel compassion for him now".
Louise described her relationship with her father as unrecognisable from the one in which she grew up.
I would say talking about that part of my life, it does feel like I'm telling a story . . . I don't carry any of that today. And it doesn't reflect where we are, me and my father . . . we can almost be two different people from the people I talked about.

Discussion
Whilst the sample was small for this study, the detail of experience gathered was very rich. The methodology allowed the participants to share their accounts in their own ways, which made room for emotional subtlety, multi-layered explanation and reflective reasoning to illuminate their lived experience for the researcher. This also facilitated the researcher's ability to connect to it and interpret it through multiple lenses. Ultimately, this resulted in a highly nuanced understanding of ways in which the father-daughter dynamic impacts the emergence of BN, and justifies responding to previous calls for a study that would focus on this relationship. This research uncovered evidence to suggest that the major significance of the father-daughter relationship to BN is that it is experienced by the daughter as a source of ongoing interpersonal trauma throughout childhood. BN emerges in early childhood as a functional survival strategy, providing for avoidance of the psychological, emotional and physical overwhelm resulting from such a problematic dyadic dynamic. This adds insight into the resistance of BN to CBT-based treatment and offers direction for new protocols that may bring about better treatment outcomes.
The evidence for BN being an outcome of complex trauma in which father is the key perpetrator was extensive in every case. This research strongly indicates that BN is a response to repeated emotional overwhelm, something also noted by Lo Sauro et al. (2008). The soothing and distracting function known to be provided by active BN (Sansone & Sansone, 2007;Trottier & MacDonald, 2017) supports the idea of BN being an embodied process of emotional regulation in response to complex trauma.
It was already known that BN arises in households that are conflictual and lack compassionate nurturing (Jáuregui Lobera et al., 2011;Levine & Mishna, 2007), but this research went beyond existing understandings to contextualise the ways in which this plays out. The women in this research offered up stories of paternal rage and emotional instability; emotional neglect and abuse; physical distancing and abandonment; and marginalisation of personal needs within the family context. This research uncovered deep feelings of fear, powerlessness and vulnerability, emerging at the level of dyadic attachment and family system.
Existing literature has indicated that in families where eating disorders appear, insecure attachment plays a role (Illing et al., 2011;Keating et al., 2013). But the relationship of insecure attachment to the development of BN is not one that has so far been clearly understood (Jewell et al., 2016). This research provides new insight into how emotional distress and breaks in attachment between father and daughter may contribute specifically to BN developing in the daughter. For example, it was seen how mealtimes with father were highly stressful; fathering lacked sweetness, gut instincts had to be pushed down and ignored, and the kitchen was offered up as a female space. In addition, since slimness was presented as an indicator to father of the successful achievement of femininity, weight control suggested the potential for paternal approval. In the context of poor self-esteem and a yearning for greater attachment with father, a focus on controlling food intake and the body becomes a logical, accessible pursuit.
Contextualised thus, BN may signal a profound form of self-care related to the deepest of attachment needs. It also supports the finding that BN develops from early childhood and that the binge/compensation cycle characteristic of BN is a stage in the condition, not a sign of its emergence. This aligns with prior findings (Le Grange & Loeb, 2007;Stice et al., 2021;) and this study confirms that BN has a progressive system of development, from feeling states and meaning-making to behaviours. This is an important and noteworthy divergence from the DSM-5-TR's (American Psychiatric Association, 2022) nosological approach to the diagnosis of BN. The DSM-5-TR (American Psychiatric Association, 2022) currently performs a vital role in dictating when and whether treatment is offered by those in the medical community, and whether health insurance funds will finance treatment. Yet by the time BN reaches the DSM-5-TR's (American Psychiatric Association, 2022) clinical threshold, it is already very difficult to treat successfully (McClelland et al., 2020). This research reinforces the view expressed by others (Cutinha et al., 2017;Gorrell et al., 2019;Loeb et al., 2004) that early intervention, at pre-clinical thresholds, may offer greater chance of treatment success. Identifying issues of complex trauma within a family in early childhood, through schools, GPs or social services, and offering solutions aimed at improving attachment and reducing family system stress could be far more propitious than proposing CBT-oriented treatment at a later stage.
The presence of complex trauma in the aetiology of eating disorders has been acknowledged in existing literature (Gordon et al., 2016;Hicks White et al., 2018;Leraas et al., 2018). This research sheds light on the currently limited success of a CBT-based first-line treatment approach specifically for BN. BN has been evidenced in this research to be linked to the presence of trauma in the family, particularly perpetrated by fathers. This means BN is rooted in a non-verbal, felt-sense of abandonment and danger to the Self, and thus is largely inaccessible to analytical processing (Van der Kolk, 2014).
The research suggests that far from BN indicating dysfunction, as the CBT treatment approach implies, BN is evidenced here to possess a functional rationale. It has a logic and a felt sense of purpose. That purpose is survival. This is why lesser consequences, such as worsening health, social difficulties or relationship failures, do not provide sufficient barriers to maintaining BN. Attempts to remove the symptoms of BN at a cognitive level inevitably risk failure long-term, because very little of BN operates in that domain. The survival response from the limbic area of the brain connects poorly to reason (Levine, 1997) and the adrenal system gets triggered as soon as there is a threat of the survival strategy (BN) being removed. This heightens ambivalence towards recovery at a visceral level and motivates intransigence in treatment.
Insecurely attached individuals are less likely to respond well to treatment protocols overall (Kowal et al., 2015). But short-term compliance with treatment is common amongst those with eating disorders, often because primary insecure attachment results in a search for an alternative attachment relationship, and this can manifest as a desire to please the therapist (Williams, 2006). This does not, however, translate into meaningful, long-term change. Long-term success relies upon prioritising the emergence of agency and a felt-sense of safety, whilst respecting and accessing non-verbal, somatic understandings of experience (Van der Kolk, 2014).
It is the suggestion here that where an individual seeks help for BN, a far deeper, longer-term, relational approach to recovery is necessary to shift this most profound of survival mechanisms. The recommendation arising from this research is therefore to recognise BN as an outcome of complex trauma and offer help that follows the individualised, staged, multi-modal, complex trauma treatment model. This focuses on establishment of safety, followed by unpacking of traumatic experience and then the building of relational and other essential skills for more successful living (Van der Kolk, 2014). A combination of treatment approaches focusing on mind, body and spirit are elements of this model, such as trauma-focused yoga, Somatic Experiencing, meditation and mindfulness practices, 12-Step programmes and Eye Movement Desensitisation and Reprocessing (EMDR), amongst others, alongside psychotherapy. There is already growing evidence as to the promising nature of this approach (Freudenberg et al., 2016;Hicks White et al., 2018;Munchel, 2013;Schaffner & Buchanan, 2008); and this longer-term, more comprehensive strategy of care offers a more effective and appropriate way forward than CBT techniques and medication alone.
Finally, since this research highlights the relational nature of BN, BN should not be considered to be situated in one person. It has a systemic scaffolding, it is rooted in connection. BN arises between individuals and within an environmental context. Martin Buber's conceptualisation of existential connectedness with others was the I-Thou relationship: that we are always in relationship with another, never existing in isolation (Buber, 1996). It is worth considering how this spiritual element of connection may be more fully understood, embraced and harnessed by treatment practitioners.

Conclusions
The aim of this research was to explore the significance of the father-daughter relationship to the development and treatment of BN. It has been evidenced here that the father-daughter dynamic creates a context of complex trauma beginning in early childhood, where father is a key perpetrator. This dynamic may be instrumental in the development of symptoms of weight concern and use of food for emotion regulation. This research makes sense of BN's resistance to treatment by CBT and medication alone and also helps justify and support the growing body of researchers looking to the complex trauma treatment model as the route offering greater hope for treatment success. This approach, as recommended by this research, addresses the underlying stresses and attachment failures for which BN seeks to provide a solution.
Currently, the pre-eminent value attributed to the "expert" medical view and the evidence-based scientific model undervalues qualitative research, practitioner-based evidence, complementary medical treatments and complex trauma treatment protocols (Bryden et al., 2018;Byrne et al., 2016). This, arguably, has led to continued support for CBT as the predominant treatment modality, regardless of legitimate challenges to the latter's claims of success. Whilst useful as a construct, conventional medical wisdom only provides one conceptualisation of BN. The qualitative approach presented here adds value to the current body of research by offering fresh insight together with pragmatic and actionable direction for more fruitful treatment outcomes.

Limitations
Findings with this methodology are indicative but not generalisable, so it would be useful to explore them using a larger sample. Further, this sample constituted heterosexual females only. It would also be helpful to understand the effect of gender-orientation on the development of BN, both within the father-child dyad and external to it. Finally, the father-son relationship requires its own focused exploration, as there is currently limited research on males and bulimia. It would be beneficial to address these limitations in further research.