Eating Disorders in Everyday Life
A Qualitative Approach
Noraly Ilahi 10541926 Master Medical Anthropology and Sociology
University of Amsterdam Patrick Brown & Kristine Krause
August 2019 in Amsterdam 22927 words
ABSTRACT
Many studies have researched the effect of eating disorders on everyday life. However, most of these studies approached the topic from a biomedical framework. In this study, the effect of the eating disorder label on everyday life was studied from a medical anthropological framework that centred personal narratives of those who suffered from an eating disorder. Special consideration was given to processes of labelling and meaning making. The research utilized semi-structured interviews. The focus lay on three areas. The first – personal
understanding of the eating disorder label – illustrated that while societal beliefs played a role in the meaning that was attributed to the eating disorder label, participants deemed their underlying problems more significant. The second – intersubjective relationships – illustrated that an eating disorder has the potential to change intersubjective relationships in different ways. It affected the people with an eating disorder, and those who have someone in their close vicinity with an eating disorder. The third – changes throughout time – illustrated that the meaning of an eating disorder changed over time. Firstly, an eating disorder was
experienced as a set of disordered acts that disrupt everyday life, later its disruptions were something that needed to be rehabilitated. Some participants stated that it was something for which they always had to be vigilant, even when they felt secure in their recovery. The final section explored the theoretical implications that the findings of this thesis had on, for example Link and colleagues’ modified labelling theory (1989).
CONTENTS
Abstract ... 1
1. Introduction ... 4
2. Eating disorders in the Netherlands ... 7
Eating disorder treatment ... 7
Biopsychosocial model ... 9
The sociocultural perspective ... 10
Construction of the stereotype ... 11
Online representation of eating disorders ... 13
Summary ... 13
3. Theoretical framework ... 14
Phenomenologically inspired ... 15
Labelling ... 17
Everyday life ... 20
Making sense through narratives ... 21
Summary ... 22
4. Methodology: A bottom-up approach ... 24
Sampling and design ... 24
Multi-media research ... 27
Ethical considerations ... 28
Data analysis ... 30
5. Life prior to an eating disorder ... 31
Stereotypical image of eating disorders ... 32
Underlying problems ... 36
Summary ... 39
6. Life with an active diagnosis ... 40
Effect on everyday life ... 41
Everyday social surroundings when ill ... 43
Online community ... 46
Summary ... 49
7. Life during Recovery ... 51
Motivation for recovery ... 53
Barriers of recovery ... 56
Summary ... 58
8. Conclusion ... 60
Meaning of the eating disorder label ... 60
Changes to intersubjective relationships ... 61
Eating disorders through time ... 63
Theoretical implications ... 64
Literature ... 66
Appendices ... 74
Appendix 1 – list of abbreviations ... 74
1. INTRODUCTION
In this thesis, I focus on the experiences of people with an eating disorder. In my approach, I centre their descriptions. Therefore, I start with a quote from Marya Hornbacher that explains her views on eating disorders.
It is, at the most basic level, a bundle of deadly contradictions: a desire for power that strips you of all power. A gesture of strength that divests you of all strength. A wish to prove that you need nothing, that you have no human hungers, which turns on itself and becomes a searing need for the hunger itself. It is an attempt to find an identity, but ultimately it strips you of any sense of yourself, save the sorry identity of the ‘sick’. […] These contradictions begin to split a person in two. Body and mind fall apart from each other, and it is in this fissure that an eating disorder may flourish, in the silence that surrounds this confusion that an eating disorder may fester and thrive (Hornbacher, 1998, p. 6).
The quote is illustrative of the intrinsic complexities of living with an eating disorder (ED). Hornbacher eloquently described how she felt trapped in between different desires that ultimately kept her sick for years. She chronicled her struggles in the book Wasted, a memoir of Anorexia and Bulimia. Throughout this chapter, I use fragments from this book to illustrate the topic of eating disorders.
The title reveals her ED diagnoses: Anorexia Nervosa (AN) and Bulimia Nervosa (BN). These are part of the taxonomy found in the DSM-V. The chapter on eating disorders features five other disorders: Binge Eating Disorder (BED), Unspecified Feeding or Eating Disorder, Avoidant/Restrictive Food Intake Disorder (ARFID), Rumination Disorder, and Pica
(American Psychiatric Association, 2013)1. The DSM-V describes each disorder in accordance with disordered practices and duration. It bases the severity of an ED on the occurrence rate of disordered practices. For AN, it is based on BMI. The taxonomy as described in the DSM received critical reviews (Becker, Thomas & Pike, 2009; Gordon, Holm-Denoma, Crosby & Wonderlich, 2010; Mancuso et al., 2015). An example is invalid diagnostic boundaries due to frequent cross-overs between diagnoses (Gordon, Holm-Denoma, Crosby & Wonderlich, 2010). Hornbacher’s experience illustrates this:
1
The last three disorders were previously stated in a different chapter in the DSM because they were viewed as children’s diseases (Grohol, 2013). In the DSM-V, they have been added into one chapter to indicate that the diagnosis can be made for people of all ages.
I became bulimic at the age of nine, anorexic at the age of fifteen. I couldn’t decide between the two and veered back and forth from one to the other (Hornbacher, 1998, p. 2).
A model that is not exclusively biomedical, is the biopsychosocial model (Collier & Treasure, 2004). This model includes knowledge from multiple knowledge constellations. Within this model, the socio-cultural component is given equal consideration. I explain this model more in-depth in the next chapter. The added value of incorporating socio-cultural knowledge lies in the notion that anthropological research can tease out often-overlooked aspects of culture, that structure the experience of those with an ED, and influence treatment success (Eli & Warin, 2018). However, medical anthropological research regarding EDs does not exclusively research the origin of EDs, it also researches its effect on a life.
Previous research into the ED-experience has looked at, for example, the personal meaning that is attributed to symptoms (Fox, Larkin, Leung, 2011), how recovery is defined (Bowldy, Anderson, Hall, Willingham, 2015), how an ED influences identity (Eli, 2018b) or its effect on personal relationships (Buser, Parkins, Gelin, Buser & Kearney, 2016). Horbacher’s next quote is illustrative of the informative nature of personal narratives.
What I am about to say is tricky, and it is a statement about my own relationship with bulimia and anorexia. Bulimia is linked, in my life, to periods of intense passion, passion of all kinds, but most specifically emotional passion. Bulimia acknowledges the body explicitly, violently. It attacks the body, but it does not deny it. […] the bulimic impulse is more realistic than the anorexic because, for all its horrible nihilism, it understands that the body is inescapable. The anorectic operates under the astounding illusion that she can escape the flesh, and, by association, the realm of emotions (Hornbacher, 1998, p.93).
Personal ED narratives are a rich well of information. However, research into the effect of an eating disorder on everyday life, is often strongly coloured by biomedical frameworks
(Warin, 2010). The aim of my research is to add to the body of knowledge pertaining to EDs in everyday life from a medical anthropological framework. An understanding of EDs does not have an explicit answer. Therefore, I do not uncover causal relationships. Instead, I research how certain experiences came to be and expose the processes that enabled them. To do this, my central research question is: In what way does an eating disorder label impact on everyday life?
In the following chapter, I discuss the context of eating disorders in everyday life. I focus on the treatment of EDs in the Netherlands, the biopsychosocial model, societal beliefs
pertaining to EDs, and the online subculture of ED communities. In the third chapter, I structure my research using phenomenologically inspired concepts (Schutz, 1967; Eli, 2016; Eli, 2018a). I also explore the significance of personal narratives and the use of symbolic language (Kirmayer, 1992; Garrett, 1998). The research is embedded in the larger theoretical fields of labelling and meaning making (Link, Cullen, Struening, Shrout & Dohrenwend, 1989; Corrigan & Watson, 2002). Based on the theoretical framework, I constructed 3 sub-questions to guide my research. In the fourth chapter, I justify my methodological approach. An investigation into the everyday experience of an ED, requires that I talk to people who have first-hand knowledge. To that end, I conducted semi-structured interviews with people who were currently, or in the past, diagnosed with an ED. I also discus sampling, research design and data analysis. In chapters five through seven, I discuss the findings in relation to my theoretical framework. The last chapter is dedicated to answering the central research question. I end with the theoretical implications of my research.
2. EATING DISORDERS IN THE NETHE RLANDS
The goal of this chapter, is to provide contextual knowledge regarding EDs in the Netherlands. I focus on how EDs are experienced in the Netherlands, and how they are treated in the health care system. In the second section I describe an explanatory model that is based on knowledge of different knowledge constellations. I give special consideration to social research. In the last section, I discus societal and online representations of EDs.
EATING DISORDER T REATMENT
The prevalence of EDs is highly debated, with many studies reporting a different incidence rate (Hoek & Van Hoeken, 2003). Based on numbers published by the Dutch Youth Institute – Nederands Jeugdinstituut – the numbers have been rising (NJI, 2014). Eating disorders with the highest prevalence rate2 are AN, BN and BED. The average duration of an ED in the Netherlands was 9,2 years3 (Op ‘t Landt-Slof, de la Torre y Rivas, van Furth, 2015). The average age of diagnosis was 19,6 years. Of those people, 27 per cent had multiple ED diagnosis during their lifetime. Many also had comorbid disorders4. This is common with EDs (Herpertz-Dahlmann, 2009). Of those who received health care, 85 per cent had
outpatient care, 47 per cent was admitted during the day and 45 per cent was committed to a residential treatment facility. Within biomedical psychiatric research, AN and BN are studied most because they have the highest mortality and chronicity among all psychiatric disorders (Eetstoornissen, 2006). After 10 years, 20 per cent of those who have been diagnosed, are still ill and 30 per cent still show symptoms to some degree. It is estimated that 20-25 per cent of people with AN or BN never (fully) recover. This trend is not exclusive to the Netherlands as EDs are on the rise across the world (Agras, 2010; Keel, 2010). In 2003, the WHO declared EDs a priority disorder because of its high prevalence (WHO, 2003). These
2 The lifetime prevalence for women is 1-4 per cent for AN, 1-2 per cent for BN, and 1-4 per cent for BED. For
men, the lifetime prevalence is 0.16-0.3 per cent for AN, 0.1-0.5 per cent for BN and 1-3 per cent for BED (Samenvattingskaart Zorgstandaarden Eetstoornissen, 2017).
3 These numbers are based on a questionnaire conducted by the Dutch Eating disorder register (Op ‘t
Landt-Slof, de la Torre y Rivas, van Furth, 2015).
4
Mood disorders (68%), Personality disorders (34%), Anxiety/compulsion (31%), PTSS/trauma (29%), ADD/ADHD (12%)
problematic numbers and percentages might, in part, be due to the numerous problems that people encounter in the Dutch health care system.
The first problems are caused by the ‘stepped care principle’ (Brandt-Dominicus, 2006, p. 78). This entails that people seeking care go through different steps of health care
professionals5. Each step functions as a gateway to the next step. The design ensures that people start with the least burdensome, cheapest, and shortest form of care. Access to the next step is granted only when care in the present step is deemed insufficient by a health care professional. Within this system, multiple concurrent referrals for specialized second step health care destined for the same disorder, are impossible (Omgaan met meerdere diagnoses, 2017). This causes two problems.
The first problem is the long waiting lists for specialized ED care (Oosterom, 2019). Many people wait months on end to receive the care they need. While waiting, they cannot receive other specialized care. The second problem is that disorders that are categorized under the same DSM axis, cannot be treated concurrently (Omgaan met meerdere diagnoses, 2017). As a result, someone who has both an ED and a different psychiatric disorder that falls under the same axis, can only be treated for their other psychiatric disorder, if the treatment for their ED has been completed.
The second problem that people with an ED encounter in the Dutch health care system, is in relation to the recommended treatment option. The treatment recommended for EDs, is some form of cognitive behavioural therapy, CBT (Fijten & Noordenbos, 2017; Linardon, Wade, de la Piedad Garcia & Brennan, 2017; Zorgstandaarden Eetstoornissen, 2017).
The idea is that ED psychopathology may be likened to a house of cards with the strategy being to identify and remove the key cards that are supporting the ED, thereby bringing down the entire house (Murphy, Straebler, Cooper & Fairburn, 2010, p. 616-617).
However, many other psychiatric disorders, or ailments, are contraindications6 for the beginning of CBT (Murphy, Straebler, Cooper & Fairburn, 2010, p. 617). This means that Murphy and colleagues predict that CBT will not work when someone has both an ED and a
5 The 0th step are family, partners and self-care groups. The 1st step are general practitioners, dieticians,
homecare, basis certified psychologists and the Youth Care Office. The 2nd step is specialised health care (Brandt-Dominicus, 2006).
6
Murphy and colleagues mention severe clinical depression, significant substance abuse, life crises, competing commitments as contraindications for CBT (Murphy, Straebler, Cooper & Fairburn, 2010, p. 617).
different psychiatric disorder. In those cases, as they would say, the entire house will not go down because the other psychiatric disorder could be a key card. The problems with the health care system unnecessarily prolong an ED, and as such, prolong its effect on everyday life.
BIOPSYCHOSOCIAL MODEL
There is no consensus as to what exactly causes an ED (Oberlin & Morrisey, 2018b). However, there is considerable support that the biopsychosocial model, a multifactorial model, offers different possibilities as the cause of an ED (Collier & Treasure, 2004). This model brings together perspectives from the biomedical, the psychological and the social knowledge constellations. While I do not take a biomedical or a psychological approach, I discuss all perspectives because the participants whom I interviewed were in contact with either/or biomedical/psychological professionals. Consequently, the explanatory narratives within these fields are important as they were offered to the participants to make sense of their own experience. Because this is a medical anthropological study, the social aspect of the model is discussed more in-depth.
Within the biomedical knowledge constellation, EDs are framed within the medical model of disease and treatment (Oberlin & Morrisey, 2018a). Research is conducted within the fields of genetics, neurology and pharmacology, and focusses on the biological factors that play a role in the development and maintenance of EDs (Collier & Treasure, 2004). The
psychological knowledge constellation focusses on finding correlations between EDs and someone’s psychological make-up.
The first model that offers an explanation as to what makes someone psychologically inclined to develop an ED is that an ED functions as a coping strategy (Nolen-Hoeksema & Watkins, 2011). The second is that an ED is the result of trauma (Brady, Killeen, Brewerton &
Lucerini, 2000). The third is that an ED functions as a coping strategy for developmental transitions within the stages of a lifetime (McNichols, Lydon, Lennon & Dooley, 2009). The fourth is that people with an ED have attachment issues that stem from early childhood (Zerbe, 2008). While it is beyond the scope of this thesis to comment on the cause of EDs, these perspectives are important as they speak to the narratives that participants are offered
within the psychological knowledge constellation to make sense of their ED experience. The next section discusses the sociocultural perspectives of the biopsychosocial model.
THE SOCIOCULTURAL PERSPECTIVE
The sociocultural perspective is concerned with:
‘how the cultural norms and expectations of the society in which a person lives, influence their thinking and behaviours in relation to the development and maintenance of their ED’ (Oberlin & Morrissey, 2018a, p. 21).
The influence of culture on EDs, is described by the ‘tripartite model’ (Markey, 2004). The model posits that culture influences eating disorders through three distinct mediating pathways. The first, eating behaviours, is the transmission of food preferences and restrictions. The second, body image ideals, is the definition of body ideals and what is deemed ideal, but also, what people are expected to work for. The third, perceptions of health, are perceptions regarding health and disease.
Figure 1: the tripartite model (Markey, 2004, p.
143)
The autobiographical account of Tatjana Almuli – pretty for a fat girl – offers an example of the model in real life. Almuli chronicles what it is like to have a fat body in Dutch society, but also, what it is like to ascertain a slim body in Dutch society. As she described it:
Because I am almost 60 kilograms lighter, most people assume that I have become a different person, and most of all, that everything is going well. For most, losing weight and being slim are equated with feeling good about yourself, being healthy and being happy. […] I feel very
uncomfortable with these compliments that are actually not compliments at all. They make me think about my own reasons for wanting to lose weight and the unrealistic ideal that
accompanied that (Almuli, 2019, p. 118).
Almuli suffered from BED for many years and the above comment was written when she lost a significant amount of weight. The first part highlights the notion that slim bodies are
viewed more positively than fat bodies. The second part highlights Almuli’s contemplation of whether her longing for a slim body was due to her own desire, or societal expectations. Her experience is in line with academic research regarding normative views on the body. In our society, certain bodies are associated with certain qualities, that come with (dis)advantages (Hamermesh, 2011, Gruys, 2012, Vandebroeck, 2016; O’Neill & Silver, 2017). Slim bodies are associated with positive traits (Couniha, 1999; O’Neil & Silver, 2017), and produces a certain amount of privilege (Hamermesh, 2011; Vandebroek, 2016). Fat bodies are associated with negative traits (Gruys, 2012; O’Neill & Silver, 2017). The juxtaposition between slim and fat bodies results in slim bodies being viewed as the ideal body image. Furthermore, dieting as a practice has become fairly normalized as people mind their eating to a certain extent to ascertain a slim body (Tyler & Abbott, 1998; Grimlin, 2007; Holla, 2018). Accordingly, body size and food become a moral matter (Coveney, 2006; Holla, 2018). Consequently, eating practices and the shape of a body are viewed as a reflection of character. While norms regarding eating behaviour, body image ideals and perceptions of health do not cause EDs, they do create a soil in which EDs thrive.
However, societal beliefs that influence the ED-label are not limited to eating behaviours, body image ideals and perceptions of health. Societal beliefs regarding EDs themselves, also play a role in the personal process of meaning making. In the next section I discuss ED stereotypes.
CONSTRUCTION OF THE STEREOTYPE
The way in which women give meaning to their own ED-label, is influenced by the societal portrayal of EDs. Therefore, I researched media portrayals, personal beliefs and Dutch media portrayals.
In the US media, EDs were often presented as a disease with which mostly young white women are afflicted (O’Hara & Smith, 2007). The illness itself was presented as simple, and straightforward with an optimistic recovery rate. Its development was attributed to bad
parental figures. These findings built on a study done 10 years prior (Levitt, 1997). This study found that, in the US, people associated the following terms with eating disorders: low self-esteem, obsessiveness; body, food, female, teenager/young, perfectionistic, media figures, wanting to please others, narcissistic, needs/lack of control (Levitt, 1997, p. 175). However, when the same question was asked regarding ‘mental disorders’, the following concepts were associated with it: depression, schizophrenia, withdrawn, aggressive,
confusion, poor (Levitt, 1997, p. 175). Levitt concluded that ‘mental disorders’ were viewed from a pathological framework while ‘EDs’ were viewed from a moral framework. He stated that:
More responsibility is placed on the eating disordered for their behaviour. The terms used to describe eating disorders suggest that this disorder is considered to be a value choice, whereas mental illness is seen as happening to a person (Levitt, 1997, p. 176).
These findings were supported by a poll commissioned by the National Eating Disorders Association (Insight, 2005). However, they expanded knowledge regarding developmental beliefs to include dieting and the media as root causes of EDs. In the last 15-20 years, NEDA attempted to change representations of ED in the media to encourage depictions of its
complex nature (NEDA, 2018).
While there have been no academic studies on the media representation, or the personal understanding of EDs in the Netherlands, there have been two non-fiction series this past year that have received different responses. ‘Anorexia eetclub’ – which loosely translates to
Anorectic dining club – people suffering from AN who came together weekly to learn how to cook and eat well. The program received a lot of backlash (Nieber, 2019b; Irene, 2018). First, because it displayed the participants in a disrespectful manner and second because it
displayed AN only in terms of problems with food. There were also many problems with casting as they only allowed participants who were ‘skinny enough’ (Nieber, 2019a). It is highly problematic that this was broadcasted as it might have added a flat description (Finlay & Gough, 2008) of EDs to the ED discourse.
On the other hand, a positive addition to the discourse regarding EDs was ‘Eetstoornissen, masker af’ - Eating disorder behind the mask. This mini-series showed four women who each suffer from a different ED (Beerekamp, 2018). These women and their therapists were
followed for an extensive period. Each episode presented the complex and differing nature of each of their EDs. Representations of EDs in the media are not only produced by popular
media outlets, they are produced by individuals themselves as well. In the next section, I discuss representations of EDs on social media.
ONLINE REPRESENTATION OF EATING DISORDERS
Online communities are used to connect and collaborate (Haimson, Brubaker, Dombrowski & Hayes, 2015), share inspiring media (Fleming-May & Miller, 2010) and find like-minded people (Boero & Pascoe, 2012). ED communities are often found on social networking sites like Tumblr (De Choudhury, 2015) and Instagram (Pater, Haimson, Andalibi & Mynatt, 2016) and either centres recovery or pathology. These communities allow participants to both disclose a part of themselves which they might keep a secret from their offline social
surroundings, while at the same time, getting support for their intended goal (Brotsky & Giles, 2007). Pater and colleagues found that the media shared within these communities can be divided into five media archetypes: thinspiration, ED journey, diet, suicidal ideation and self-harm. These archetypes represent different aspects of the online representations of EDs. Thinspiration, or any of the other lexical variations7, is content that shows slim bodies. It represents what the individual in question aspires to. ED journey is content that showcases before and after pictures as a motivation for people to ‘keep going’. Diet content showcases content relating to dieting practices. Suicidal ideation content conveys the depressive
symptoms that an individual might experience due to their ED. Self-harm content speaks for itself. Content is often accompanied by a message through which the content is personalized. For some, their online representation becomes an integral part of their ED as it is the only way in which they feel they are connected to others (Yeshua-Katz & Martins 2013). These representations have the potential to define and redefine what EDs mean and how they are to be enacted. They also speak to how individuals might present and copy ED-related
behaviour.
SUMMARY
7 Due to censorship, hashtags, or handles that allude to ED jargon, have been banned (Chancellor, Pater, Clear,
Gilbert & De Choudhury, 2016). Lexical variations were created to surpass this ban. For example ‘thinspiration’ was written as thinspo, thinspooo(o), thinspire, thynspo, thynspiration (Pater, Haimson, Andalibi & Mynatt, 2016, p. 1191).
To answer the central research question – in what way does an ED-label impact on everyday life – I provided contextual knowledge in this chapter. This started with EDs in the
Netherlands and how they are treated in the Dutch health care system. Within the health care system, systemic problems arise that prolong suffering in three different ways. The first is that waiting lists are too long. The second shortcoming is that the principle of stepped care denies access to multiple forms of treatment and the recommended therapy. The third – CBT – is often insufficient to treat complex pathologies.
The second section discussed an explanatory model (Collier & Treasure, 2004). This model prescribes that EDs are cause by biological, psychological and social factors. The social contributes to the development of EDs through three pathways: norms regarding eating behaviours, body image ideals and perceptions of health (Markey, 2004). The first pathway regarding eating behaviour prescribes that people need to eat healthy (Tyler & Abbott, 1998; Grimlin, 2007; Holla, 2018). Eating is judged from a moral framework – those who eat healthy are viewed as morally superior to those who eat unhealthy (Coveney, 2001). The second pathway regarding body image ideals prescribes that slim bodies are favourable to fat bodies (Hamermesh, 2011, Gruys, 2012, Vandebroeck, 2016; O’Neill & Silver, 2017). The third pathway regarding perceptions of health prescribes that slim bodies are viewed as healthy and fat bodies as unhealthy (Coveney, 2001). In the last section, I discussed representations of EDs. This started with the societal image pertaining to EDs. This
illustrated the flat description of EDs (Levitt, 1997; O’Hara & Smith, 2007; Finlay & Gough, 2008). However, within Dutch society strides are being made to adjust this narrative with many books and documentaries that convey the complexity of the topic (Beerekamp, 2018; Almuli, 2019). Those with an ED sometimes seek refuge online. The online content that they post, represents their ED-label in different ways. The next chapter discusses the theoretical concepts that are used to investigate the way in which an ED-label impacts on everyday life. Important for this discussion, are concepts regarding meaning making and labelling.
The research question – in what way does an ED-label impact on everyday life – immediately raises questions regarding the theorization of meaning making, labelling and everyday life. In reference to labelling, these questions pertain to which knowledge constellations are used when attributing meaning to the ED-label and the way in which people reproduce or deviate from these knowledge constellations. In reference to everyday life, these questions pertain to how everyday life is structured, which parts of everyday life are relevant to these structures, and how these relate to each other. The most important question, is how labelling and everyday life are brought together.
In this chapter, I discuss the theoretical concepts that I chose to structure these questions. The chapter starts out with an account of phenomenology and why phenomenology offers
valuable sensitizing concepts. In the next section, the process of labelling is theorized and related to phenomenological concepts. The last section focusses on theoretical understandings of how labelling impacts on everyday life and the role of the personal narratives in the
process of meaning making.
PHENOMENOLOGICALLY INSPIRED
Phenomenology is a broad theoretical tradition that offers different approaches. First, I discuss some approaches that were adopted by other researchers to understand different aspects of EDs. These studies demonstrate the merit of a phenomenological approach. Next, I discuss why I chose Schutzian phenomenological concepts.
Buser and colleagues researched how social surroundings experienced a loved one with an ED (2016). They adopted a transcendental phenomenological approach (Moustakas, 1994 in Buser et al., 2016). The aim of that approach is to analyse experiences without the
researchers’ personal biases, assumptions and preconceptions (Buser et al., 2016, p. 328). Thus, they had to put aside their own subjectivity when approaching the participants and the subsequent data. Within phenomenological research, it is highly debated what place the researcher’s subjectivity holds within the research, and to what extent it is important (Finlay, 2012). Within transcendental phenomenology, it is believed that the subjectivity of the researcher should be bracketed of as much as possible. However, a different school of
the solution (Girogi, 1994, p. 205 in Finlay, 2012, p. 12). The underlying theoretical idea is that data is the result of a reality that is created by the interaction between the participant and the researcher as a subject. Ignoring the researcher’s subjectivity, means denying the effect that the researcher has on the creation of the data.
Fox and colleagues used a phenomenological approach to understand the way in which participants gave meaning to their ED symptoms (Fox, Larkin & Leung, 2011). However, their phenomenological approach was limited to using interpretative phenomenological analysis (IPA) to analyse the data. Phenomenology was not used as a framework from which the data collection was designed. This is a problem that is often seen with academic studies – the pretence that phenomenology is used while only IPA, or an emphasis on the process of meaning making is used (Finlay, 2012). In accordance with Finlay’s typology of
phenomenological approaches, their work does not count as phenomenology.
In her body of work, Eli uses embodied phenomenology to understand different aspects of ‘the eating disordered experience’ (Eli, 2016; Eli, 2018a; Eli, 2018b). She categorizes her work as embodied phenomenology. Within embodied phenomenology, culture and individual perception are inextricably linked, with sensory experience and the social environment always-already implicating one another (Csordas, 1993 in Eli, 2016 p. 73). Her approach emphasized lived emotional experience and the way in which the body is central in that experience. Using embodied phenomenology, the body becomes an experiencing subject, but what is neglected, is the body as an object (Crossley, 1995). Because of this, embodied approaches often leave out how people make sense of their lives and bodies in relation to societal meaning frameworks.
The phenomenological studies that I discussed, offered valuable insights into experiences related to an ED. Nonetheless, I adopt Schutzian concepts (Schutz, 1967). Due to its contributions to the methodology of the social sciences in grasping concrete experiences in the social world (Nasu, 2008). Within Schutzian phenomenology, the body is mostly approached as an object (Crossley, 1995). This means that his work foregrounds the way in which meaning is constructed on a personal and societal level, and the way in which these levels interact with one other. An important Schutzian contribution to this thesis, is his concept of ‘life-world’. According to Schutz, the life-world is:
In principle a field of human actions […]. In addition, the life-world is never the world of a single person. There are always other people in the world. That is, the life-world is from the beginning an intersubjective social world (Schutz, 1967 in Yu, 1999, p. 159).
Essentially, it is the sum of the world we perceive to live in, the actions that are enacted within it and the people with whom our perception of that world and the actions within in are created. An important part of our life-world, is that it is taken-for-granted. This means that we are not consciously aware of every meaning that is constructed or the ripple effect of every action. The meaning that is given to aspects within our life-world, are not constructed solely by ourselves. Instead, they are subject to the society in which we live and the people with whom we share our life-world. I use the concepts that Schutz puts forth as ‘sensitizing concepts’ (Blumer, 1954). I use this concept to theoretically ground ‘everyday life’. In relation to my central research question, this means that, if I want to understand the effect of the ED-label, I need to research the societal meaning framework in which participants reside, how these frameworks contribute to their personal process of meaning making, and how their intersubjective relationships are influenced by their intersubjective relationships. Based on this, I supplement my question with two sub-questions.
1. How do people with an eating disorder understand their label? 2. How does an eating disorder change intersubjective relationships?
However, the theoretical approach is Schutzian phenomenologically inspired rather than Schutzian phenomenology per se. This allows for the experiences shared by the participants, not to be altered. Instead the concepts give guidance in the approach towards the experience. In the next section, I theorize how the process of labelling and meaning making take place.
LABELLING
A personal experience is affected by the society in which it takes place. According to Schutz, the social world is constitutive of meaningfully interlocking activities of actors (Zaner, 1961, p. 71). One of the ways in which the activities of actors ascertain meaning, is through the process of labelling. I theorize the process of labelling using Corrigan and Watson (2002). I start the discussion of this framework from a macro-perspective, and work towards the micro for the sake of clarity.
Actions are often induced by certain beliefs. Whether these beliefs are positive or negative, influences whether the actions are positive or negative. As mental illness is often associated
with negative beliefs, the actions towards the mentally ill, often are negative (Levitt, 1997; O’Hara & Smith, 2007). Corrigan and Watson’s theory focusses on how stigma towards the mentally ill occurs. In the previous chapter, I illustrated that EDs are often associated with negative beliefs (Levitt, 1997; O’Hara & Smith, 2007; Ebneter & Latner, 2013; Griffith, Mond, Murray & Touyz, 2015). This makes Corrigan and Watson’s theory suitable to analyse the way in which the participants gave meaning to their ED-label.
Corrigan and Watson’s model theorizes the way in which mental illness is labelled by society in a negative manner, and the way in which those who are ill, navigate the process of
labelling. They found three layers, both for society and the individual, that show pathways through which the labelling process takes place. From the societal standpoint, labelling occurs in the following manner. The first layer, stereotypes, means that there are certain negative beliefs about a group. The second layer, prejudice, is the agreement with the stereotypes. The third layer, discrimination, is a behavioural response to the prejudice. From an individual standpoint, they theorised that there are three mediating factors in the process of labelling. The first is whether the individual sees the collective representation as accurate. In the case of EDs, this raises the question of whether the participants see EDs as disorders that exclusively affect young white girls. It also calls into question whether they associate EDs exclusively with AN. The second mediating factor is whether the individual deems the stigma that ensues as legitimate. This raises the question of whether the
participants agree with the notion that EDs are viewed from a moral, instead of a pathological framework (Levitt, 1997). The third is whether the individual identifies with the group of people with whom they share their diagnosis. This raises two questions. The first is whether they know other people who are afflicted with an ED. The second is whether they feel that those people conform to or deviate from societal beliefs. Depending on how people feel towards these mediating factors, people deal with the societal label in one of three ways. In the first instance, they internalize the label and stigmatize themselves based on it. This would mean that they view their ED as a choice instead of a disorder. In the second instance, they reject the stigma and feel anger towards its existence. In the third instance, people neither internalize the stigma or become enraged by it. They would feel as though it does not apply to them.
While this framework offers tangible tools with which I can approach the labelling process, there are tensions with their work and the ED experience. As such, I supplement the process
of labelling with two theoretical frameworks (Link et al., 1989; Eli, 2018). To facilitate a full account of the ED experience, I add the concept ‘liminality’ (Eli, 2018, p. 477). Eli defines ‘liminality as:
the embodied being-in- the-world effected through eating disordered practices, upon its subjective and intersubjective visceral, sensory, perceptual, felt, and cognized aspects (Eli, 2018, p. 477).
The concept moves away from defining ‘sickness’ in exclusively biomedical terms. Instead, it acknowledges that the effect of an ED can be thusly profound, that it alters the course of a life, securing people in between the realms of ‘sickness’ and ‘health’. It does not negate the possibility of recovery, rather, it emphasizes that, even in recovery, people can still struggle with their ED past in the present. This concept forms the basis for the last sub-question that guided my research:
3. How does the impact of an eating disorder change over time?
There is a slight tension between Schutz’s operationalisation of phenomenology, and Eli’s. Eli’s body of work focusses on the embodied experience while Schutzian phenomenology focusses more on the mind, intersubjective meaning and its layers of assumptions. The body and the experiences of the body are not central in this thesis, but they do have a place. First, because bodily knowledge influences our knowledge base. Second, because the disordered body influences the potentiality of the subject within the world. Within sociological research, Crossley (1995) sees a difference in how the body is treated. He theorizes that, in some studies, the body is treated as an object that resides in discourses and is subject to regimes of practices. The emphasis lies on what is done to the body. In other studies, the body is treated as an active subject that consciously takes part in social life. In these studies, the emphasis lies on what the body does. Crossley criticises both approaches and makes a case for studies that focus both on what is done to the body, but also what the body does. Combining
Schutzian concepts with concepts proposed by Eli, allows for an approach by which the body is mostly approach as an object, but still treated as a subject (Crossley, 1995, p. 43). This allows for an examination of the process of meaning making in which lived experience is not neglected. In the next section, I continue the theorization of labelling, and connects it to everyday life.
EVERYDAY LIFE
In the previous section, I described the different impacts of labels on different individuals, and the processes and factors which help explain this. In this section, I built on these concepts to theorize ways in which individuals interact with societal labels to further illustrate what the effect of the label is on everyday life. To understand the effect on everyday life, it is
important to not only understand how the individual interacts with it, but also, how it effects an interaction between individuals, and how that in turn, shapes their own subject position (Link et al., 1989).
Important for the process of labelling, is what Schutz calls ‘typification’ (Schutz, 1967). This concept posits that everything encountered by an individual, is recognized as having a certain meaning. For example, the word ‘tree’ has certain properties by which we recognize a tree. EDs also have certain properties by which they are recognized as an ED. These have been discussed in the previous chapter. Before contracting an ED, these properties, or beliefs, stem from socialization, but after its development, the beliefs stem from experience (Link et al., 1989). During the development of an ED, the entire life-world changes. Disordered acts and thoughts fill the day and become normalized, thus changing the taken for granted reality. As these change, intersubjective dynamics potentially change as well.
Due to socialization, individuals might be privy to negative beliefs. The awareness of these beliefs influences the way in which people deal with their label in relation to others. Link et al. theorized that, concerning a negative label, individuals can take three approaches that influence their social life: education, secrecy and withdrawal. The first strategy, education, was also proposed by Corrigan and Watson (2002) to combat the negative label. By doing this, the individual in question allows those in their social surroundings to acquire first-hand knowledge regarding EDs. Furthermore, it minimalizes typification in an interaction as people have an expanded frame of reference from which they can approach someone. The second strategy, secrecy, entails that the individual in question does not share their struggle. This results in one of two things. Either, the pathology fits with the beliefs pertaining to an ED, or the pathology does not. In the first instance, the ED is typified as such, and
relationship potentially changes. The third strategy is social withdrawal. This, in turn, allows to ED to take up even more of the life-world.
The most notable difference between Link and colleagues, and Corrigan and Watson, is that Corrigan and Watson propose that negative effects are mostly felt when people internalize the negative label. Link and colleagues go further to argue that internalization of the label is not necessary for it to have a negative effect. For them, the extent to which the label has a negative effect on the social surroundings is not dependent on the negatively labelled group, but firstly on the degree of belief that they could be devalued based on their label, and secondly the degree of belief that they could be discriminated against based on their label. These in turn are dependent on both awareness of possible beliefs, but also on negative experiences throughout the illness. I propose that personalizing beliefs and making sense of the illness experience, happens through narratives (Kirmayer, 1992; Garrett, 1997). In the next section, I theorize how participants utilize narratives.
MAKING SENSE THROUGH NARRATIVES
Illnesses themselves can be viewed as irrational in the sense that the rational mind cannot fully comprehend what is happening. This is illustrated by the notion that sickness raises the questions ‘why me’ and ‘what can be done’ (Kleinman, 1988). Consequently, illness drives us to put into words what is happening, and through this, create a reality in which what is happening makes sense (Kirmayer, 1992). The use of metaphors closes the gap between the rational and irrational as it puts the irrational experience into words that the rational mind can comprehend. Kirmayer theorizes that metaphors present themselves in two ways. The first is that they operate as a cognitive tool and the second as communicative acts or gestures
(Kirmayer, 1992, p. 337). In both cases, they fashion new meanings. They are situated both in bodily experience and the social surrounding in the sense that:
Metaphors embody the situational knowledge that constitutes culture. While conventional metaphors reflect common sense, the idiosyncratic metaphors of individuals are local contributions to the cultural situation (Kirmayer, 1992, p. 339).
The narratives that result from these experiences, are textual – subjective and lived – rather that flat – objective picture of the phenomenon – description (Finlay & Gough, 2008, p. 32).
The body plays an integral part in the narrative as sickness is both what is done to the body, and something that makes the body act (Crossley, 1995). The narratives of those with an ED are vital for understanding their experience as those who suffer from an ED themselves report that not therapy, but events and people outside of therapy were most relevant to their
recovery (Beresin, Gordon & Herzog, 1989 in Garrett, 1997, p. 263). Garrett continues that since the account of causes change through time change, it is the narrative of these causes which is most relevant (Garrett, 1997, p. 262). She concluded:
the coherence of the story creates the coherence of the experience. The coherence of the story often makes it possible to transcend the gaps and inconsistencies in experience. A coherent experience, like a coherent story, is also a connected experience; one which has internal agreement and a sense of unity and which can be communicated to others in narrative form (Garrett, 1997, p. 267).
Viewing the narratives of the participants in the way Kirmayer and Garrett theorized, facilitates acknowledgement of the interactive relationship between reality and meaning making. Simultaneously, it validates the constructive nature of both.
SUMMARY
To theorize the way in which the ED-label impacts on everyday life I have sketched a framework of sensitizing concepts regarding the process of labelling and its connection to everyday life. The theoretical framework outlined in this chapter, is inspired by Schutzian phenomenological concepts regarding meaning making and the life-world. Attributing meaning to the ED-label, is theorized through concepts proposed by Corrigan and Watson (2000). Their theory offers the possibility to theorize three processes. First, their concepts regarding societal stigma offers insights into the creation of stereotypes. Second, they offer explanatory factors to understand different interpretations of societal labels. Thirdly, they offer different strategies that people utilize to navigate labels. However, their theory offers insufficient concepts to connect the process of labelling to the effect on the life-world. As such, I supplemented the theory with two theoretical frameworks. The first theoretical framework is Eli’s work within the embodied paradigm on the ED experience. The
connection of Eli’s work to that of Corrigan and Watson allows this thesis to approach the body both as a subject and an object (Crossley, 1995). The second theoretical framework with which I supplement Corrigan and Watson’s work, is Link et al.’s modified labelling theory. This framework illustrates that while people with an ED have first-hand knowledge
on the topic of EDs, they also have preconceptions dating from socialization. Furthermore, the framework illustrates that internalization of the negative label is not necessary for it to have a profoundly negative effect on the life-world of those afflicted with an ED. The beliefs pertaining to EDs potentially affect the validation that people receive regarding their
suffering. The absence of validation has the potential to make them feel that they are not suffering at all. Based on this theoretical framework, I constructed three sub-questions that guided the research
1. How do people with an eating disorder understand their label? 2. How does an eating disorder change intersubjective relationships? 3. How does the impact of an eating disorder change over time?
I discuss how participants experience this in chapter five through seven. In the next chapter I discuss the method that I utilized to investigate in what way an ED-label impacts on everyday life.
4. METHODOLOGY: A BOTTOM-UP APPROACH
To answer the central research question – in what ways does an eating disorder label impact on everyday life – I developed a qualitative approach with an explorative nature. The
research is meant as bottom-up. The research is built on philosophical assumptions. I discuss these first, before I discuss the chapter outline.
The findings of this thesis are the product of the interaction between me as a researcher and the participants during the interview (Finlay, 2012). The research presented here is not
neutral or value-free, and my role as researcher requires a discussion (Gergen, 2009). My role as a researcher is influenced by the many other social roles that I play: Dutch, educated, black, middle-class and woman (Schutz, 1967). It is also influenced by the preconceptions that I have towards the subject of EDs, both from a societal and a personal perspective. The societal preconceptions have been discussed in the second chapter. My personal
preconceptions stem from watching others around me struggle with EDs. Their struggle has shown me that EDs are all encompassing diseases that go far beyond food. While it is impossible to escape my preconceptions altogether, it is important to take those into consideration as it is impossible to bracket those roles (Smith, 2004). Furthermore, my presentation of self, influences how the participants presented themselves. The bodies in this research are treated mostly as objects, but also as a subject (Crossley, 1995). In the remainder of this chapter, I discuss the sampling methods and the research design, the ethical
considerations and data-analysis. These sections further illustrate how my presentation of self, guided the presentation of the participants.
SAMPLING AND DESIGN
I interviewed 16 women for this research project. The participants were recruited based on three criteria. The first criterion was that participants had to be diagnosed by a professional. This was to ensure that I was not the first person with whom they discussed their disorder, as I am not qualified to have such a conversation. The second criterion was that participants themselves had to state whether they had an active diagnosis (AD) or that they were in
recovery8 (IR). There was no set definition for what it means to be IR or AD as this is personally defined (Fox, Larkin & Leung, 2011; Bowlby, Anderson, Hall & Willingham, 2015). The third criterion was that they lived in the Netherlands.
The recruitment took place, both through a snowball method and purposive sampling (Hennink, 2011; Silverman, 2015). I conducted purposive sampling through Instagram. I knew of an ED Instagram influencer who had a rather large following. In her posts, she chronicled what life was like with an ED and many people who themselves had an ED followed her. I used the hashtags that she used in her posts to find other ED-accounts9. The participants to whom I reached out on Instagram, were selected based on one more criterion; their biography10 had to state that they had an ED.
The participants were diagnosed with ARFID (n=1), AN (n=11), BED (n=2), BN (n=4), EDNOS11 (n=3). Of those participants, five had multiple ED diagnosis during their lifetime. Six participants stated that they had an AD12 and ten people who stated that they were IR. When I use an interview excerpt, I reference a participant with their pseudonym and their status. The age at which disordered eating started was early childhood to late adolescence. The duration of the illness ranged from 1 to 30 years. The duration of therapy ranged from 8
8 From a psychiatric perspective, ‘recovered’ means that people were officially in remission. Remission entails
that someone has a significant decrease of symptoms (Oberlin & Morrissey, 2018b). However, with this thesis, I wanted to facilitate space for participants to provide their own meaning of what it means to have an ‘active diagnosis’ or to be ‘in recovery’.
9 The choice to not state those hashtags here is based on two reasons. The first is privacy of the individual.
Stating the hashtags will make it possible to find the participants that I interviewed. The second reason is privacy of the community. Many of the participants stated that the Instagram community was one of the few places in which they could share their experiences unencumbered. Laying bare the community might reduce the feeling of safeness.
10A ‘biography’ is a short description on your Instagram profile in which you can write something about
yourself. 11
EDNOS was an ED diagnosis that was removed from the DSM-V (Call, Walsh & Attia, 2013). To reduces the incidence rate of EDNOS, other diagnosis in the DSM-V were changed (American Psychiatric Association, 2013; Machado, Gonçalves & Hoek; 2013). It has been replaced by Unspecified Feeding and Eating Disorder.
12 There was one participant who stated that she was IR, but she had an AD for EDNOS. While this made an
interesting piece of data, conform the pre-set conditions, I respected her personal label and the narrative that she had constructed (Applequist, 2014). As such, I adhered to 6 participants who were AD instead of 7.
months to upward of 25 years. Interviews took place between February and April 2019. The interviews were conducted using a medium of the participants choosing. Five interviews took place face-to-face, while the other interviews took place via an online medium. In the
remainder of this section, I broadly sketch the procedure of the interviews. In the next section, I zoom in on challenges that I encountered in the online interviews. I conducted semi-structured interviews (Bryman, 2012). To this end, I devised an interview guide. The interview questions were based on studies on the experiences of people with EDs, both academic and auto-biographical. While I had an order of questions in mind, I often strayed from that structure if that better accommodated the flow of the interview.
The interview was divided into 4 parts. In the first part, I asked introducing questions in the form of vignettes (Kvale, 1996; Barter & Renold, 2012). My focus was to get a sense of the participants’ values and beliefs (Bryman, 2012). An added advantage was that this gave participants a chance to get used to the interview and me without having to talk about their own experiences too much (Barter & Renold, 2012). In the second part of the interview, I focussed on their oral history (Bryman, 2012). Through my questions, we reconstructed their illness narrative starting at their earliest memories (Garrett, 1997). In the third part, I focussed on how their ED changed throughout the time of their illness. This allowed for a deeper understanding of the personal meaning of the eating disorder (Fox, Larkin & Leung, 2011; Stockford, Kroese, Beesley & Leung, 2018). The fourth part of the interview focussed on the intersubjective relationships of the participants. The goal of this part was to understand the way in which they felt their relationships were changed due to their ED(s) (Buser et al., 2016). The interview ended with a section where the participants could comment on the interview – whether they felt I should have done are asked something differently. Most of the suggestions were incorporated into subsequent interviews. An example of an added question was whether people felt some sort of intrinsic motivation to get better. Throughout all parts, I used different probing techniques to encourage the flow of the interview (Bernard, 2011). I used silent probing, tell-me-more probes and echo probes.
MULTI-MEDIA RESEARCH
At the beginning of my fieldwork period, most of the interviews I scheduled were either rescheduled many times or cancelled. During my fourth interview, this phenomenon was explained.
If you look on a scale from 1 to 10, then today is a 5 or something. If it had been a 1, I would have cancelled [the interview] (Dani, Active Diagnosis).
Something that I had not considered, was that EDs thus influence everyday life, that sometimes the participants did not feel well enough to leave their house. Because of this, I took a very eclectic approach to the interviews. I did the interviews using different mediums: face-to-face (n=5), e-mail (n=2), skype (n=1), Instagram messenger (n=2), WhatsApp video call (n=1), by phone (n=2) and WhatsApp (n=3). When given the choice, most participants chose to use an online medium to do the interview. The verbal interviews (face-to-face, skype, WhatsApp video call, phone) took between 1 and 2,5 hours. The text-based interviews (e-mail, Instagram messenger, WhatsApp) took between 2 and 4 hours. An upside to my eclectic approach was that almost no interviews were cancelled. However, it did come with some difficulties.
I used Markham’s concerns regarding online ethnography, to think through the difficulties I encountered doing non-face-to-face interviews (Markham, 2005, p. 800). While my
methodology does not constitute an ethnography, Markham’s work provides useful
considerations. The first concern she encourages researcher to think through, is defining the boundaries of the field. In my research, this is geographically defined. Discursively, I draw the line at those participants who mention in their biography that they have an ED. My underlying assumption was that people who posted their diagnosis, would be more open to discussing it. The second concern proposed by Markham was in relation to determining what constitutes data. I only counted things directly communicated to me as data. I viewed content posted on the profiles of the participants off limits because I wanted their community to remain a safe space13. The third concern was interpreting the other through text. Markham explains:
13
I did contact one participant with a question regarding her profile. Before looking at her profile, I sought consent to do so.
When the participant, researcher and context are nothing but text and everything beyond mere language, our perceptual filters must be adjusted to accommodate complexities of human expression (Markham, 2005, p. 806).
As stated earlier, the interviews included a biographical account of their ED, thus, it was up to the participant to name the cultural labels that they deemed important for their narrative. I did not include anything that was not mentioned by them to ensure that my perception did not alter their narrative. The fourth concern is using embodied sensibilities to interpret textuality. Markham urges researchers, who use online mediums to expand their reach, to consider how well people express themselves in text. This was resolved by letting participants choose with which medium they felt most comfortable. The last concern that Markham raises, is how to ethically present the other in research reports. She postulates that any method decision is an ethics decision because it contributes to how these are understood, framed, and responded to by those who read it (Markham, 2005, p. 811). I discuss this concern thoroughly in the next section.
ETHICAL CONSIDERATIONS
I used four ethical principles – autonomy, benefice, non-maleficence and justice – to
safeguard ethical practice (Beauchamp & Childress, 2001). The first, autonomy, requires that my research does not violate the right to make free and duly informed choices. I contacted most participants (n=9) personally through their social media. Appendix 2 outlines the outreach message. With this message, I introduced myself and my research. I also stressed that they were not obligated to participate and that they could change their minds at any time. Before the interview, I encouraged each participant to ask questions regarding the research so they could make an informed decision regarding the interview. After each part of the
interview, I sought explicit consent again14.
The second principle, benefice, requires that I reflect on who benefits from my research. I myself benefit from the research as completing it means that I complete my master. However, I also hope that the those with an ED benefit from it as my research contributes to the
14 After each part, I explicitly asked them if they wanted to stop, or continue the interview. Beforehand, I had
explained that if they said that they wanted to stop, I would immediately stop the interview and the recording.
If they said nothing, I would do the same. Only when they said that they wanted to continue, would I ask my next question.
faceted and complex nature of the experience. Benefice also requires that I reflect on who is excluded from the research. My aim was to include people with different diagnosis and gender identifications. While I tried to prevent exclusion, it did happen on two fronts. The first was that I did not interview people who identified as men. While it was never the intention to write only about women, or the generalize my findings, this thesis offers no insights to experiences of men. This means that, unfortunately, my research falls into the tradition of ED research that excludes men (Crosscope-Happel, Hutchins, Getz & Hayes, 2000; Gila, Castro, Cesena & Toro, 2005; Rodgers et al., 2015). The second shortcoming to inclusion was the unequal distribution of diagnosis in my sample. Both AN and BN were overrepresented in my sample. Nonetheless I tried to foreground all diagnoses equally.
The third principle, non-maleficence, requires that I reflect on how my research could potentially hurt the participants. The first way through which I could safeguarded that, was by anonymizing all the data. I changed their names and identifying features. The audio files of the interviews were saved on my (password protected) laptop and deleted after
transcription15. The second way in which the interview could be harmful, was that it might trigger their ED past. There were two ways in which the participants took that. Ilze stated that she found it very difficult to talk about the past. She was one of the participants whom I knew personally. After the interview, we spent the entire evening together. I found that it was very important to – if there was space, and if the participant was open to it – wind down after the interview. On the other hand, Jane stated that:
From my side, I want to thank you for listening to my story, I think, I don’t know, it was kind of nice I guess, yeah, it has been so long, yeah, I never talk about it with anyone, and it was really nice to really, yeah I don’t know, it was nice, so thank you (Jane, In Recovery). An important part of the benefice of the interview was that it provided an opportunity to discuss the past.
The last principle, justice, requires that I provide a fair treatment for all participants. I
safeguarded this by including all interviews in the analysis equally. Another important aspect, was the risk that the research posed to my own health. I found that it was important to wind down after an interview, whether this was with the participant or someone who was close to me. I also kept a detailed log, in which I discussed everything regarding the interview and
15
how it made me feel. This was done in word document that I kept on my (password protected) laptop.
There was, however, one unexpected ethical problem that arose during the research; the death of a participant. While it certainly gave me a lump in my throat, it also left me with the question of what to do with her data and whether I should or should not include it.
In consultation with my supervisor, I decided to include the data because the participant was very active in combatting the stigma against mental health problems. I wanted to honour her wish in telling her story. For the sake of anonymity, I do not share her pseudonym.
DATA ANALYSIS
There are many ways to do interpretative phenomenological analysis. I adopted the IPA method as postulated by Smith, Jarman and Osborn (1999). In various ways, this resembles classic grounded theory. Both methods work from the ground up, however grounded theory foregrounds general views and perspectives while IPA foregrounds the depth and nuance of each individual case (Brown et al., 2015). It further deviates in its emphasis on meaning making and the underlying framework that is necessary for the negotiation of meaning. To use IPA, it is important the recognize the central role of the researcher, and to not bracket away that role (Smith, 2004; Markham, 2005).
After I had conducted and transcribed the interviews, the first step was to look for
preliminary themes in the first transcript. As I looked at two groups, AD and IR, I initially did this for both groups. However, there was no added value so I merged the preliminary theme list. Next, they propose to look for connections between themes. I did this per case and between cases. All themes were ordered into a table – the master-theme list – that included an explanation of their meaning and an example. I used the master-theme list to analyse the remaining transcripts. Themes that arose after this, were added into the list as well.
Smith and colleagues stress that analysing the data is a cyclical process in which one goes back and forth between different stages of the emerging thematic frameworks. When all the transcripts were analysed, I grouped them to find patterns. The data analysis exposed multiple themes – underlying problems, the role of the online ED community, both EDs and recovery
as a liminal space, barriers of the health care system – and their influence on the meaning that participants attributed to the ED-label. The form and salience of these themes differed based on the timeframe in which they took place – before the ED, during the ED or in recovery. Due to this, I chose to present the findings chronologically and in accordance with the timeframe in question. Important to note, is that all narratives were shared in retrospect. This means that some participants had years, even decades, to interpret and reinterpret their situation. Within the chapters, I give due consideration to this. In this paragraph, I gave a short description of the structure of the data chapters. The next chapter contains a detailed description of the data chapters and continues with the discussion of the first findings.
The central research question of this thesis asks, in what way does an ED-label impact on everyday life? Many participants mentioned that they had participated in multiple studies and they often gave the same critique: the interviews were confined to food and diet practices. The goal of this thesis was to move beyond the analysis of EDs exclusively in relation to food. Because of the challenge of conveying experiences that are highly complex, and that can seem contradictory at times, I choose to present the experiences of the women I
interviewed, in chronological order. This facilitates the emphasis on the narrative basis of the data. Thus, this thesis contains 3 data-chapters.
This first chapter therefore focusses on the situation of the participants prior to their ED. Important for the understanding of contributing factors to the construction of the ED-label, is the way in which participants understand the societal ED-narrative, and the way in which they interact with this narrative. In the second chapter I focus on the experiences of the participants when their ED diagnoses were ‘active’. Important for this chapter, was the way in which their ED-label influenced their everyday life and their social surroundings. The third chapter focusses on the experiences of the participants when they were in recovery. Here it becomes evident how participants define recovery and the obstacles they faced in recovery. Before I dive into the findings of this first data chapter, a point needs stressing in relation to the chronological presentation of data. Most of the participants I talked to where IR. This means that most, had been through years of therapy, the goal of which was to make sense of their own story. What appears as a stylized version of events, obfuscates years of hard work in therapy and other forms of self-help. Discussing the influence of therapy throughout hopefully does justice to the complexity and inherent contradictory nature that is left out when presenting it in chronological order.
STEREOTYPICAL IMAGE OF EATING DISORDERS
To explicate the process by which participants gave meaning to their ED, it is important to understand how they viewed societal beliefs regarding EDs (Corrigan & Watson, 2002). The most mentioned characteristics of someone with an ED were: extremely skinny, young, girl, barely eats and AN. This description is in line with the description of EDs in US media
