502092 2013

ANP48210.1177/0004867413502092ANZJP PerspectivesAllison et al.

Caest iur, dem popte tercerbi sena, ductum ompopubli, vem ressa nit fatque publis, ublis? At aris. Ac omnequo merit, sideat, ut auceps, nost pernihillem faudes mentro hint, vicisses iaciem perei per locchic aesuloc, Ti. Sum ommordi probusc eporum eorum sed faciis? Abestam locaede

Viewpoint

Anorexia nervosa and social contagion: Clinical implications

Australian & New Zealand Journal of Psychiatry 2014, Vol. 48(2) 116­–120 DOI: 10.1177/0004867413502092 © The Royal Australian and New Zealand College of Psychiatrists 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav anp.sagepub.com

Stephen Allison1, Megan Warin2 and Tarun Bastiampillai1

Introduction During adolescence, female friendship cliques can develop an unhealthy focus on body image, dieting and extreme weight loss (Eisenberg and NeumarkSztainer, 2010; Fletcher et  al., 2011; Paxton et al., 1999). These peer group processes contribute to the prevalence of eating problems amongst young women in westernised cultures (Becker, 2004; Becker et  al., 2011). Community studies provide robust evidence for peer influence as a significant mediator in harmful eating practices amongst young women (Fletcher et  al., 2011; Quiles Marcos et  al., 2013). Social pressures regarding body weight and shape make recovery from eating problems much more difficult (Murray et  al., 1995). Anthropological studies describe how peer influence intensifies within inpatient wards, residential units and day hospitals, which can become ‘proana’, with patients actively promoting the practices of anorexia nervosa (Warin, 2006, 2010). The power of the peer group becomes greater as young people are aggregated into homogeneous groups in tertiary care. Mutual peer influence serves to intensify ‘deviant’ eating practices in a process termed peer contagion (Dishion and Tipsord, 2011). Social media and online networks greatly increase the ability of peer groups formed in tertiary care to remain in contact within larger informal pro-ana communities. There is scant literature on the clinical management of peer contagion in eating disorders (Vandereycken, 2011). This contrasts with the study

of delinquency where empirical studies of peer contagion are advancing (Dishion and Tipsord, 2011). Further research is required into the peer system, especially during the adolescent phase of development when peer dynamics are crucial. Up to this point, the peer group has been a relatively neglected topic in the clinical treatment of eating disorders. The peer sub-system could usefully be added to the current discussion on the essential components of early intervention alongside individual, parent, family and multi-family therapy (Hay, 2012; Murray et  al., 2012). Peer influence might play a role in the course of each of these forms of early intervention. Equally, later in the progression of the disorder, social media and online networks of pro-ana peers can influence the course of tertiary treatment. It is well recognised that the cultural context for this peer influence is the high status of thinness as sanctioned, desired and valued by western culture. It is the currency of cultural practices around food and bodies as understood by groups of young women that carry power (Warin, 2006). Moral values attributed to thinness are firmly embedded in the habitus of young people: their everyday, taken-for-granted practices. Young people (and young women in particular) learn from an early age that their bodies require constant surveillance, modification and improvement. Weight and food are the main focus of this attention, and they are important markers of competition and connection within female peer groups. Under the guide (and guise) of ‘healthy

Australian & New Zealand Journal of Psychiatry, 48(2)

living’, attention to one’s body demonstrates a moral virtue that should be suitably rewarded. Within female peer groups, this cultural value is transmitted through a variety of processes including friends modelling dieting practices, body size comparisons, teasing about weight and ‘fat talk’ to encourage weight loss. Imitation and peer modelling tend to play the most important role (Quiles Marcos et al., 2013). Within the peer group, nothing is more contagious than a behavioural example from a popular and influential friend. Eating disorder behaviours can begin within these adolescent peer networks where friends routinely congratulate one another on weight loss or demanding physical exercise routines. Young people actively and competitively pursue crafted, thin bodies, which bestow unprecedented levels of symbolic capital in the contemporary peer group. By succeeding where many others fail (at dieting, self-discipline and thinness), the progression towards an eating disorder offers the appeal of a new adolescent identity and social distinction in the group. ‘Anorexia’ is not something a young person catches or a force that 1Department

of Psychiatry, Flinders University, Bedford Park, Australia 2Discipline of Gender Studies and Social Analysis, University of Adelaide, Adelaide, Australia Corresponding author: Stephen Allison, Department of Psychiatry, School of Medicine, Flinders University, Bedford Park, SA 5042, Australia. Email: [email protected]

117

Allison et al.

Figure 1.  Social network map of body mass index (BMI). BMI is clustered within social networks of the Framingham Heart Survey cohort (n = 2200). The size of each circle is proportional to the person’s BMI (with green circles indicating normal or low weight, and yellow circles indicating obesity). Used with the kind permission of Nicolas Christakis and James Fowler.

simply sweeps over passive (and gendered) bodies; it involves personal agency, especially in the early stages of the disorder. Literary critic René Girard (1996) perceptively noted that the ‘authoritative voice’ of anorexia initially ‘emanates from the people who really count in our adolescence and who are our peers and contemporaries’ (Girard, 1996: 3). He pointed to the active adoption of western cultural ideals by peer culture: ‘The individual models of young people reinforce the authority of the collective models which are the media, Hollywood, and television. The message is always the same: we have to get thinner, regardless of the cost’ (Girard, 1996: 3). Mimetic rivalry was the term that he employed for this complex peer dynamic of desire, imitation and competition. He suggested that the competitive process of mimetic rivalry was inherently contagious and helped spread eating disorders through female friendship networks. In an early longitudinal study of peer contagion, Crandall (1988) tracked the ‘epidemic’ of bulimia

beginning among female American college students. At the beginning of the students’ first year in a college sorority, friends’ eating behaviour was no more similar than average, but by the end of the year, a sorority member’s binge eating could be predicted by her friends’ level of binge eating. More recently, Becker et  al. (2004, 2011) studied the social changes in Fiji following the introduction of mass media portraying idealised images of western femininity (which differ from Melanesian ideals of rounded bodies). Mass media exposure was associated with the onset of eating disorder pathology amongst young Fijian women. The analysis suggested that western cultural influence could be transmitted either directly through the media or indirectly through social networks of close friends and classmates. A friend’s family buying a television had the potential to indirectly influence her entire friendship group. Social influence on eating behaviour can extend up to three degrees of separation in the wider peer network (i.e. a friend of a friend of a friend)

with body mass index (BMI) being related across three social ties (Christakis and Fowler, 2007: see Figure 1). These latter findings were derived through the seminal Framingham Heart Study (FHS) cohort. The FHS identified contagion across the web of relationships within broader social networks for obesity, smoking rates, alcohol use, depression, loneliness and happiness. Indeed, in another Viewpoint, we specifically focused on depression and the potential for peer contagion based on the data obtained from the FHS (Bastiampillai et al., 2013).

Clinical implications of peer influence It is important to recognise that adolescents are not all equally susceptible to peer influence. Individual vulnerability depends on a wide variety of biopsychosocial characteristics, including the patient’s defensive structure, which need to be fully addressed in a psychiatric treatment plan. Peer influence helps create social environments

Australian & New Zealand Journal of Psychiatry, 48(2)

118 where restrictive dieting is reinforced but the serious consequences of these social environments occur for more susceptible individuals. Genetically sensitive studies reveal the multifactorial aetiology of eating disorders, with a large genetic component being partly expressed through personality traits (Wade et  al., 2008). Epigenetic research suggests the possibility of a complex interaction where the social environment might modify genetic risk. The secondary changes from malnutrition induce powerful neuroendocrine changes that strongly perpetuate anorexia nervosa. These major biological forces are likely to combine with the social and cultural environment in the progression of eating disorders. Clinical responses to peer influence in eating disorders are a workin-progress; however, some general principles can be applied. As in other domains of eating disorders, early intervention may hold the key (Hay, 2012; Murray et al., 2012). An important goal of early intervention in the peer group is to maintain any existing pro-recovery friendships. While young people with anorexia nervosa have well-recognised social difficulties, they have classmates and may have close friends, especially in the early stages of the disorder. These friendships can be lost rapidly, leaving the person socially isolated. Our conceptualisation of peer-focused therapy began with Amy, who was an adolescent receiving individual psychotherapy for anorexia nervosa. Her overall health was just beginning to improve when she had an acute relapse requiring admission for medical rescue. On discharge Amy agreed to resume psychotherapy but she also insisted on setting up her own form of social activation that she called ‘friend therapy’ to improve her morale. Her peer activation meant that she spent a great deal more time with her existing friendship group, which subsequently improved her mood and motivation. Fortunately, Amy’s ‘friend therapy’ was successful, largely because her

ANZJP Perspectives friendship group did not focus on body image concerns and dieting. For Amy, their collective stance was predominantly pro-recovery. Peer network sociograms are useful for tracking pro-ana and prorecovery influences within a patient’s friendship network (Bastiampillai et al., 2013). Sociograms can explore the peer group structure, the quality of connections, interpersonal dynamics and their influence on the patient. Psychiatrists need to ask specifically about close friends’ exercise and dieting patterns (Eisenberg and NeumarkSztainer, 2010). Over the course of psychotherapy, sociograms are useful for discussing the changing and dynamic nature of the friendship group. The form of these network conversations is a social narrative about group members and their relationships. Young people might seem resistant initially to discussing their friendships (as they can be with adults generally), but it is actually a form of expression that comes more easily than talking about intimate feelings. The workings of the peer group can be of great interest to them and salient to their present goals and desires. Therapy sessions can rehearse the patient’s responses to particular peer group situations that trigger strong emotions and eating disorder responses. The Maudsley Model of Anorexia Nervosa Treatment for Adults (MANTRA) includes a useful module for selecting helpful influences and eliminating unhelpful influences in the social environment (Schmidt et al., 2012). Within a recovery strategy, patients can deliberately decide to increase their contact with prorecovery friends and contacts in their social network or reduce the amount of contact with friends who promote a strongly pro-ana stance.

Direct peer involvement Social network therapies include close friends within the therapeutic process (Bastiampillai et  al., 2013).

Australian & New Zealand Journal of Psychiatry, 48(2)

The investigation of social network therapies is most advanced in addiction studies. In the treatment of alcohol use disorders, social network therapy recruits the pro-recovery forces within the patient’s immediate social circle to aid sobriety. A large multicentre trial showed that social behaviour and network therapy (SBNT) was both practical and effective (UKATT Research Team, 2005). The SBNT strategy can be used with eating disorders by identifying the friends and family members who are most supportive of the patient’s efforts to change and bringing them together with the patient in joint therapy sessions to encourage recovery. During the early phases of an eating disorder, social network therapy can be useful in consolidating close friendships and preventing the patient being ‘left behind’ by the rapid progression of adolescent friendships. The therapeutic focus is building supportive relationships and social activities beyond the eating disorder, leading progressively to the ‘Getting a life’ stage of recovery. As young women who present for treatment may have close pro-ana friends who share their eating disorder pathology, clinical programs can discuss with adolescent patients whether anyone in their immediate social circle could benefit from an offer of prompt early intervention. This voluntary public health process needs to be undertaken in an informed, supportive and non-judgemental way. Confidentiality, consent issues and the duty to inform parents are considered carefully. The potential benefit is early detection. Family therapy is an active intervention that can help the majority of young people when delivered during the early stages of an eating disorder (Murray et  al., 2012). There is a ‘window of opportunity’ during the first few years of the disorder but young people often present after the eating disorder is too well established for early intervention. A process of ‘contact tracing’ can detect young women during the early

119

Allison et al. and treatable stages of an eating disorder. This group-focused process can interrupt peer contagion in high-risk clusters and prompt systemic public health-informed responses in social groups and organisations – high schools, gymnastic clubs, modelling agencies and ballet schools.

Managing contagion in tertiary care Recently, anthropological work revealed that extended patient contact in tertiary care units magnified peer contagion as fellow sufferers met and developed pro-ana relationships (Vandereycken, 2011; Warin, 2006, 2010). These peer networks actively resisted health professionals and family members who attempted to intervene (Warin, 2006, 2010). ‘Anorexia’ had agency and power in these patient groups as both an individual and a collective identity. The patient community developed clear boundaries (defined by the Diagnostic and Statistical Manual of Mental Disorders (DSM) clinical categories), ties of relatedness, values and a status hierarchy. Peers worked hard to become ‘better anorexics’ and share tips and tricks as part of their joint allegiance and the strength of connectedness to ‘the anorexic club’. Unlike other illness categories, anorexia nervosa was transformed from a clinical entity into a friend: it became Ana, a comforter – especially during the early ‘honeymoon phase’ of the disorder. Relationships with Ana and anorexic peers were actively sought, and they provided the underpinning cultural logic to everyday practices of anorexia nervosa. ‘Anorexia’ created an inpatient or day patient community with a powerful sense of belonging. When this happened, anorexia became a team activity with its own covert set of game rules (Warin, 2006). Patients who tried to leave anorexia were maligned as not being authentic or ‘true’ anorexics and the peer group marginalised them as being ‘outside anorexics’ (Warin, 2010). When

young people were discharged from the program, their previous friends may have ‘moved on’. By choice and default, they were drawn increasingly towards the new friendships from hospital. These new relationships were readily maintained through proana social media and online networks. Relationships between fellow sufferers were complex – while they were highly supportive emotionally, they were also competitive and served to perpetuate the eating disorder by mimetic rivalry (imitation and competition) long after discharge from tertiary care. Unchecked, these counter-therapeutic peer influences limit the clinical effectiveness of tertiary care and even introduce iatrogenic effects (Vandereycken, 2011). Tertiary programs find ways to manage peer contagion when it occurs and promote an active pro-recovery ethos. Specialist teams accomplish this in a variety of ways, including strong clinical leadership, well-structured ward programs and selective group work. Some units make it a rule to have ‘no eating disorders talk’ or online activity between patients and they are asked to leave if unable to comply. Patients are given the alternative of individual psychotherapy for the expression of pro-ana desires and compulsions. Groupbased programs require participants to be matched for commitment to recovery so they can support each other. Ideally, this allows the power to reside within the peer group avoiding the traditional ‘us and them’ (top down) relationship of professional and patient, which begets greater resistance. Further quality improvement studies are needed to examine the usefulness of these program design features for reducing unhealthy peer contagion and promoting the positive goal of social rehabilitation in anorexia nervosa.

Conclusions Adolescents are well aware of their peers’ desires and the pathways to

social distinction. The forces of social imitation and social competition drive adolescent group behaviour. The ‘authoritative voice’ of anorexia derives partly from the expectations of the peer group. Social network studies provide robust evidence for peer influence on the contagion of unhealthy eating behaviour in westernised societies. Social anthropology contributes detailed descriptions of these culturally informed peer dynamics and their impact on the course of treatment in eating disorders. In psychiatric assessment, the pro-recovery and pro-ana alignment of the peer network can be thoroughly investigated. Potentially, peer network assessments can inform psychiatric intervention across all aspects of treatment: individual psychotherapy, family therapy, peer support, tertiary care and a public health approach focusing on social network clusters. Outcome studies are needed to investigate whether peer network assessments and interventions can make a difference across the course of treatment. Keywords Anorexia nervosa, social, contagion, network, peer

Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References Bastiampillai T, Allison S and Chan S (2013) Is depression contagious? The importance of social networks and the clinical implications of contagion theory. Australian and New Zealand Journal of Psychiatry 47: 299–303. Becker AE (2004) Television, disordered eating, and young women in Fiji: Negotiating body image and identity during rapid social change. Culture, Medicine and Psychiatry 28: 533–559. Becker AE, Fay KE, Agnew-Blais J, et al. (2011) Social network media exposure and adolescent

Australian & New Zealand Journal of Psychiatry, 48(2)

120 eating pathology in Fiji. British Journal of Psychiatry 198: 43–50. Christakis NA and Fowler JH (2007) The spread of obesity in a large social network over 32 years. New England Journal of Medicine 357: 370–379. Crandall CS (1988) Social contagion of binge eating. Journal of Personality and Social Psychology 55: 588–598. Dishion TJ and Tipsord JM (2011) Peer contagion in child and adolescent social and emotional development. Annual Review of Psychology 62: 189–214. Eisenberg ME and Neumark-Sztainer D (2010) Friends’ dieting and disordered eating behaviors among adolescents five years later: Findings from Project EAT. Journal of Adolescent Health 47: 67–73. Fletcher A, Bonell C and Sorhaindo A (2011) You are what your friends eat: Systematic review of social network analyses of young people’s eating behaviours and bodyweight. Journal of Epidemiology and Community Health 65: 548–555.

ANZJP Perspectives Girard R (1996) Eating disorders and mimetic desire. Contagion: Journal of Violence, Mimesis, and Culture 3: 1–20. Hay PJ (2012) Avoid being ‘tyrannised’ by the evidence: Reflections on family-based treatment in adolescent anorexia nervosa. Australian and New Zealand Journal of Psychiatry 46: 1102–1103. Murray SB, Thornton C and Wallis A (2012) A thorn in the side of evidence based treatment for adolescent anorexia nervosa. Australian and New Zealand Journal of Psychiatry 46: 1026–1028. Murray SH, Touyz SW and Beumont PJV (1995) The influence of personal relationships on women’s eating behavior and body satisfaction. Eating Disorders: Journal of Treatment & Prevention 3: 243–252. Paxton SJ, Schutz HK, Wertheim EH, et al. (1999) Friendship clique and peer influence on body image concerns, dietary restraint, extreme weight loss behaviors and binge eating in adolescent girls. Journal of Abnormal Psychology 108: 255–266. Quiles Marcos Y, Quile Sebastian MJ, Pamies Aubalat L, et  al. (2013) Peer and family

Australian & New Zealand Journal of Psychiatry, 48(2)

influence in eating disorders: A meta-analysis. European Psychiatry 28: 199–206. Schmidt U, Oldershaw A, Jichi F, et  al. (2012) Out-patient psychological therapies for adults with anorexia nervosa: Randomised controlled trial. British Journal of Psychiatry 201: 392–399. UKATT Research Team (2005) Effectiveness of treatment for alcohol problems: Findings of the randomised UK Alcohol Treatment Trial (UKATT). British Medical Journal 331: 541. Vandereycken W (2011) Can eating disorders become ‘contagious’ in group therapy and specialist inpatient care? European Eating Disorders Review 19: 289–295. Wade TD, Tiggemann M, Bulik CM, et al. (2008) Shared temperamental risk factors for anorexia nervosa: A twin study. Psychosomatic Medicine 70: 239–244. Warin M (2006) Reconfiguring relatedness in anorexia. Anthropology and Medicine 13: 41–54. Warin M (2010) Abject Relations: Everyday Worlds of Anorexia. New Brunswick, NJ: Rutgers University Press.