Eating disorders: anorexia nervosa, bulimia nervosa, binge-eating disorder, and the biology of starvation

Intuition Beginner

Eating disorders are not lifestyle choices, vanity, or willpower failures. They are serious psychiatric illnesses with the highest mortality rate of any mental disorder — anorexia nervosa kills roughly $5$ to $6$ in every $100$ people diagnosed, mostly through cardiac arrest or suicide. Three core patterns dominate. Anorexia nervosa (AN) is severe food restriction that drives body weight to dangerous lows, paired with an intense fear of weight gain and a distorted body image. Bulimia nervosa (BN) is cycles of binge eating followed by purging — vomiting, laxatives, exercise. Binge-eating disorder (BED) is binges without compensation.

The biology matters because starvation changes the brain. When the body is starved, it responds with obsessive thoughts about food, bizarre eating rituals, depression, social withdrawal, and — most strikingly — binge eating during refeeding. These are not character flaws. They are the mammalian response to deprivation. The cultural layer — the Western thin ideal propagated through magazines, fashion, and social media — shapes who is vulnerable and how symptoms present, but eating disorders are documented across cultures and through history. A London physician, William Gull, described anorexia in 1873.

Why does this unit exist? Because eating disorders affect roughly $5$ to $9$ percent of women and $2$ to $3$ percent of men across a lifetime, because they kill more often than any other psychiatric condition, and because evidence-based treatment — especially Maudsley Family-Based Treatment for adolescent anorexia, which reaches roughly $70$ percent recovery — saves lives when delivered early. Understanding the biology of starvation is the difference between blaming the patient and treating them.

Visual Beginner

The picture shows the DSM-5 eating-disorders spectrum at a glance — anorexia nervosa (severe restriction, low weight), bulimia nervosa (binge plus compensation, often normal weight), binge-eating disorder (binge without compensation), and ARFID (avoidance unrelated to body image) — alongside the binge-starve cycle and the design of the Keys 1950 Minnesota Starvation Experiment.

The three rows compress the unit's three load-bearing claims: the disorders are clinically distinct (top), they share a common dynamical engine in the binge-starve cycle (middle), and that engine is empirically grounded in a starvation study run on healthy volunteers (bottom).

Worked example Beginner

In 1944 the University of Minnesota recruited 36 healthy male conscientious objectors for a study of starvation and refeeding, intended to guide relief efforts after the Second World War. Ancel Keys and his team published the results in 1950 as The Biology of Human Starvation.

Step 1: a 12-week control phase established each man's baseline. Subjects ate roughly $3,200$ kilocalories per day and maintained a stable weight, while researchers measured body composition, metabolic rate, and psychological functioning.

Step 2: a 24-week semi-starvation phase cut intake to about $1,800$ kilocalories per day — close to a $50$ percent reduction. The men lost an average of $25$ percent of body weight, dropping from about $150$ pounds to $113$ pounds. Heart rates slowed. Faces hollowed. The men reported cold intolerance, dizziness, and muscle soreness.

Step 3: refeeding. The psychological effects were profound. The men became obsessed with food, invented bizarre eating rituals, grew depressed and socially withdrawn, and — most strikingly — binged uncontrollably when restrictions eased. Two subjects were hospitalised; one chopped off three fingers with an axe in a stress response.

What this tells us: semi-starvation itself produces eating-disorder-like symptoms in previously healthy men. Many features of anorexia nervosa are consequences of starvation, not its antecedent causes.

Check your understanding Beginner

Formal definition Intermediate+

The DSM-5 (American Psychiatric Association 2013) ^[DSM5_2013] recognises four principal feeding and eating disorders in adults: anorexia nervosa, bulimia nervosa, binge-eating disorder, and avoidant/restrictive food intake disorder (ARFID). Each is defined by behavioural, cognitive, and (for AN) physiological criteria. The diagnostic thresholds carry clinical weight: they determine treatment access, insurance coverage, and the architecture of clinical trials.

Definition (Anorexia nervosa, AN). AN is defined by (i) a restriction of energy intake relative to requirements, leading to a significantly low body weight in the context of age, sex, developmental trajectory, and physical health (clinically, a body-mass index below $18.5kg/{\mathrm{m}}^2$, with severity specifiers from mild ($\ge 17$) to extreme ($<15$)); (ii) an intense fear of gaining weight or becoming fat, or persistent behaviour that interferes with weight gain, even though at significantly low weight; and (iii) a disturbance in the way one's body weight or shape is experienced, undue influence of body weight or shape on self-evaluation, or persistent lack of recognition of the seriousness of the current low body weight. Two subtypes are specified: the restricting subtype (no recurrent binge-eating or purging in the past three months) and the binge-eating/purging subtype (recurrent episodes in the past three months).

Definition (Bulimia nervosa, BN). BN is defined by (i) recurrent episodes of binge eating — eating, in a discrete period, an amount of food definitely larger than most people would eat under similar circumstances, with a sense of loss of control; (ii) recurrent inappropriate compensatory behaviour to prevent weight gain (self-induced vomiting; misuse of laxatives, diuretics, or enemas; fasting; excessive exercise); (iii) both occurring on average at least once per week for three months; and (iv) self-evaluation unduly influenced by body shape and weight. The disturbance does not occur exclusively during episodes of AN.

Definition (Binge-eating disorder, BED). BED is defined by recurrent binge-eating episodes (as in BN criterion i) that are not accompanied by regular inappropriate compensatory behaviour. Binge episodes are marked by at least three of: eating rapidly, eating until uncomfortably full, eating large amounts when not hungry, eating alone because of embarrassment, or feeling disgusted with oneself afterwards; marked distress is present; episodes occur at least once per week for three months. BED was promoted to a full diagnosis in DSM-5 (having been in Appendix B of DSM-IV as a research diagnosis), reflecting evidence that the condition is common, persistent, and clinically impairing even without compensatory behaviour.

Definition (ARFID). Avoidant/restrictive food intake disorder is defined by an eating or feeding disturbance (apparent lack of interest, avoidance based on sensory qualities, or concern about aversive consequences) resulting in persistent failure to meet nutritional or energy needs, without body-image distortion or fear of weight gain. ARFID is distinct from AN: the restriction is not driven by weight or shape concerns. It commonly presents in childhood and can persist into adulthood.

Definition (OSFED). Other Specified Feeding or Eating Disorder captures clinically significant presentations that do not meet full criteria for the above — atypical AN (all criteria except low weight), BN or BED of lower frequency or shorter duration, purging disorder, and night-eating syndrome. The OSFED category is not a minor diagnosis: it carries substantial morbidity and mortality.

Counterexamples to common slips

• "You can tell someone has an eating disorder by looking at them." No. BN and BED typically occur at normal or elevated weight. AN-restricting subtype requires low weight, but the wider spectrum does not. Body weight is a poor screen; behaviour and cognition are the diagnostic targets.

• "Anorexia is vanity taken to an extreme." No. AN carries a standardised mortality ratio of approximately $5.9$ (Arcelus 2011) ^{[Arcelus2011]}, with deaths driven by cardiac arrest and suicide. Vanity does not kill at this rate; severe physiological and cognitive disruption does. The "vanity" framing is the principal lay misconception that delays presentation and treatment.

• "Refeeding is straightforward — just feed the patient." No. Refeeding syndrome is potentially fatal. The acute insulin surge during refeeding drives intracellular shifts of phosphate, potassium, and magnesium, leading to hypophosphataemia, arrhythmia, and heart failure. Modern protocols (Heruc et al. 2020) ^[Heruc2020] require phosphate and electrolyte monitoring for the first one to two weeks of refeeding, with calorie titration starting below the patient's estimated needs.

• "Force-feeding cures anorexia." No. Weight restoration is necessary but not sufficient. Without psychological treatment — typically CBT-E (Fairburn 2008) ^{[Fairburn2008]} for adults or Maudsley Family-Based Treatment (Lock and Le Grange 2012) ^{[LockLeGrange2012]} for adolescents — relapse rates after weight restoration alone are high (roughly $30$ to $50$ percent within one year in older studies).

• "Eating disorders are a modern Western problem." No. Gull (1873) ^[Gull1873] and Lasègue (1873) ^{[Lasègue1873]} gave independent medical descriptions before mass media. Non-Western AN exists, though incidence rates are lower; cases meeting full DSM-5 criteria for eating disorders have been documented on every continent.

• "Binge eating and overeating are the same thing." No. A binge episode requires both an objectively large amount of food and a subjective sense of loss of control. Holiday overeating without loss of control is not a binge; a small amount of food eaten with marked loss of control can still count clinically.

Key experiment: the Minnesota Starvation Study Intermediate+

Experiment (Keys, Brožek, Henschel, Mickelsen, Taylor 1944–45; published 1950). Thirty-six healthy male conscientious objectors, ages 20 to 33, were studied at the University of Minnesota Laboratory of Physiological Hygiene under a four-phase protocol: a 12-week control (baseline) phase on approximately $3,200$ kcal/day; a 24-week semi-starvation phase on approximately $1,800$ kcal/day, designed to produce a $25$ percent loss of body weight; a 12-week restricted refeeding phase; and an 8-week unrestricted refeeding phase. Physiological, behavioural, and psychological measurements were taken throughout.

Key results.

1. Body weight. Subjects lost an average of $24$ to $25$ percent of initial body weight (mean $69$ to $52\mathrm{kg}$), with a corresponding $40$ percent reduction in basal metabolic rate. Heart rate dropped to a mean of approximately $35$ beats per minute.

2. Psychological effects. All subjects developed intense preoccupation with food, including collecting recipes, studying cookbooks, and ritualising eating. Many reported depression, irritability, social withdrawal, and impaired concentration. Two subjects were hospitalised for acute psychiatric decompensation; one self-mutilated by chopping off three fingers with an axe.

3. Binge refeeding. During the refeeding phase, subjects ate continuously, often far beyond assigned intake. Episodes of massive overconsumption (reported above $8,000$ kcal in single binges in some cases) were observed in previously disciplined eaters. The pattern matched the diagnostic profile of bulimia nervosa decades before Russell formally defined the syndrome in 1979 ^{[Russell1979]}.

4. Body image. Several subjects reported subjective distortions of body image, including feelings that their bodies were grotesquely thin yet still too fat — the cognitive signature later formalised as AN criterion (iii).

Key argument. The Minnesota data establish that many features used diagnostically to characterise anorexia nervosa — food preoccupation, ritualised eating, depression, social withdrawal, binge refeeding, body-image distortion — appear in healthy adults under conditions of experimentally induced semi-starvation. The inferential claim is that these features are consequences of starvation, not antecedent personality traits or cultural inheritances. This is the empirical foundation for the bio-psychosocial model: starvation is sufficient to produce the disorder's behavioural and cognitive signature in a non-clinical population, which means the disorder's maintenance is driven by the starvation state itself and not solely by the cultural or psychological factors that may have triggered its onset.

Limitations. The Minnesota subjects were not selected for AN vulnerability, did not exhibit body-image distortion before starvation, and the cultural context of postwar conscientious objection differs from contemporary diet culture. The experiment therefore speaks to the maintenance and amplification of eating-disorder symptoms, not to the full aetiology of AN. Body-image disturbance in particular appears to require antecedent vulnerability that the Keys study did not induce.

Key analysis. The clinical implication is that weight restoration is a precondition for any psychological treatment of AN to succeed. Treating the cognitive distortions of an actively starved brain is empirically unreliable, because the starvation state itself is generating much of the pathology. CBT-E and Maudsley FBT both operationalise this insight: Maudsley FBT begins with weight restoration as phase one, and CBT-E prioritises the normalisation of eating patterns before cognitive restructuring.

Bridge. The Minnesota experiment builds toward the modern bio-psychosocial model of anorexia nervosa by separating antecedent vulnerability from starvation-driven maintenance, and this is exactly the distinction that motivated Fairburn's transdiagnostic CBT-E framework. The central insight — that starvation produces eating-disorder-like symptoms in healthy adults — generalises across the diagnostic spectrum: restriction, binge, and purging behaviours all escalate under calorie deficit, and the pattern appears again in 29.09.01 as a recurring failure mode of the diagnostic classification system itself, where rigid category boundaries break down precisely because the underlying starvation-maintenance mechanism is shared.

Exercises Intermediate+

Interpretive debates and developments Master

Result 1 (Gull 1873–74, Lasègue 1873: the first medical descriptions). William Gull's 1873 lecture to the Clinical Society of London, published in 1874 ^[Gull1873], and Ernest-Charles Lasègue's concurrent paper in the Archives Générales de Médecine ^{[Lasègue1873]}, gave independent descriptions of the disorder Gull named apepsia hysterica and then anorexia nervosa. Gull's contribution was to firmly separate the condition from tuberculosis and other organic wasting diseases, attributing it to a "morbid mental state" and demonstrating that it could be reversed by feeding under medical supervision. The Lasègue–Gull syndrome is the historical root of the modern diagnosis.

Result 2 (Bruch 1973: the psychodynamic reframing). Hilde Bruch's Eating Disorders ^[Bruch1973] reframed anorexia nervosa from a behavioural curiosity into a disorder of self: a disturbance in body image, interoception, and the sense of personal effectiveness. Bruch's emphasis on the patient's distorted inner experience — not the visible weight loss — set the conceptual foundation for the cognitive-behavioural interventions that would follow, even where her psychodynamic framework was set aside.

Result 3 (Keys 1950: the empirical foundation for starvation as maintenance). The Minnesota Starvation Experiment ^[Keys1950] established that semi-starvation in previously healthy men produces eating-disorder-like symptoms: food preoccupation, ritualised eating, depression, social withdrawal, body-image distortion, and binge refeeding. The inferential claim is that many of the diagnostic features of AN are consequences of the starved state rather than antecedent personality traits. This is the empirical anchor for the bio-psychosocial model and the clinical principle that weight restoration precedes psychological treatment.

Result 4 (Russell 1979: bulimia nervosa as a distinct disorder). Gerald Russell's paper in Psychological Medicine ^{[Russell1979]} named bulimia nervosa and distinguished it from anorexia nervosa on the basis of body weight (normal or near-normal) and the cyclic pattern of binge and purge. The DSM-III (1980) included bulimia as a diagnosis; the modern BN criteria crystallised in DSM-IV (1994) and were refined in DSM-5 (2013). Russell's 1979 paper remains the single most cited clinical paper on bulimia.

Result 5 (Garner and Garfinkel 1980: the cultural dimension). Garner and Garfinkel's Psychological Medicine paper ^{[GarnerGarfinkel1980]} demonstrated that the cultural context — particularly the post-1960s Western thin ideal propagated through fashion and media — modulates the prevalence and presentation of eating disorders. The paper established the sociocultural layer of the bio-psychosocial model without reducing the disorder to it. Subsequent cross-cultural work has confirmed that eating-disorder prevalence rises in populations undergoing Westernisation, but that AN-like presentations exist across cultures and historical periods.

Result 6 (Fairburn 2008: the transdiagnostic CBT-E model). Fairburn's Cognitive Behavior Therapy and Eating Disorders ^{[Fairburn2008]} unified the treatment of AN, BN, and BED under a single protocol (CBT-E) targeting the over-evaluation of shape and weight as the shared core pathology. The transdiagnostic model treats the DSM-5 categories as different stable equilibria of the same maintenance system, and the unified protocol has become the first-line evidence-based treatment for adults with any eating-disorder presentation.

Result 7 (Lock and Le Grange 2012: Maudsley FBT as first-line for adolescent AN). The Maudsley Family-Based Treatment manual, refined by Lock and Le Grange ^{[LockLeGrange2012]}, established the family as the primary treatment agent for adolescent anorexia nervosa. Phase one refeeds the adolescent under parental control; phase two returns eating autonomy; phase three addresses adolescent development. Randomised trials support approximately $70$ percent full recovery, substantially outperforming individual therapy for adolescent AN. Maudsley FBT operationalises the Keys insight — weight restoration first — by handing the feeding task to parents rather than clinicians.

Result 8 (DSM-5 2013: BED promoted, ARFID formalised). DSM-5 ^[DSM5_2013] made two structural changes of clinical importance: it promoted binge-eating disorder from a research diagnosis (DSM-IV Appendix B) to a full diagnosis, recognising the substantial morbidity of binge eating without compensatory behaviour; and it formalised avoidant/restrictive food intake disorder (ARFID), renaming and broadening the DSM-IV "feeding disorder of early childhood" to capture children and adults whose food restriction is driven by sensory sensitivity, fear of aversive consequences, or apparent lack of interest rather than by body-image concerns. Both changes broadened the diagnostic net without lowering the threshold for AN or BN.

Result 9 (Treasure et al. 2015; Heruc et al. 2020: the modern neurobiology and refeeding-syndrome synthesis). Treasure et al.'s Nature Reviews Disease Primers review ^{[Treasure2015]} consolidates the modern neurobiological account: altered dopamine and serotonin signalling, oxytocin dysregulation, reward-circuit anomalies, and altered interoceptive processing in insula and anterior cingulate cortex. The Heruc et al. 2020 Australian-New Zealand guidelines ^[Heruc2020] updated refeeding-syndrome protocols away from the older "start high" approach toward conservative initial calorie loads ($15$ to $20kcal/kg/day$) with aggressive phosphate and electrolyte monitoring, because refeeding syndrome is the leading acute cause of death in inpatient AN treatment.

Synthesis. The foundational reason eating disorders have proven so resistant to lay intuition — and so vulnerable to misconception — is that they are not vanity taken to extremes but a self-reinforcing starvation-maintenance system with the highest mortality in psychiatry. The central insight of Keys 1950, that semi-starvation itself produces the disorder's behavioural signature in healthy adults, is exactly what binds the bio-psychosocial model into a single causal architecture: antecedent vulnerability (genetic, temperamental, cultural) triggers onset, but the starvation state maintains and amplifies the disorder through the binge-starve cycle. Putting these together with Fairburn's transdiagnostic framework identifies the over-evaluation of shape and weight as the shared core pathology whose behavioural expression varies across the DSM-5 categories, and the pattern generalises to the modern neurobiological account in which reward-circuit anomalies interact with the physiology of depletion. The bridge is between Gull's 1873 clinical description and the 2020 refeeding-syndrome guideline update: the same clinical observation that starvation produces a distinctive psychopathology has been progressively formalised into diagnostic criteria, evidence-based psychological treatments, and acute medical protocols, and the pattern appears again in 29.09.01 as one of the clearest demonstrations that the DSM-5 categories, however useful administratively, must be read as different stable equilibria of an underlying dynamical process rather than as discrete diseases.

Full proof set Master

Proposition (Binge-starve cycle as positive-feedback maintenance). Let $R_t$ denote the caloric intake minus expenditure at time $t$ (negative for deficit), $P_t=-R_t$ the caloric deficit, $B_t$ the probability of a binge episode, and $C_t\ge 0$ the magnitude of compensatory behaviour at time $t$. Assume the empirically established monotone relations $B_t=f(P_t)$ with $f^{\prime }>0$ (deficit drives binge probability), $C_t=g(B_t)$ with $g^{\prime }>0$ (binge drives compensation), and $R_{t+1}=R_t-\alpha C_t$ with $\alpha >0$ (compensation deepens the next deficit). Then for any initial deficit $P_0>0$, the sequence $P_t$ is strictly increasing, $P_{t+1}>P_t$, and the cycle has no internal equilibrium.

Proof. By hypothesis, $R_{t+1}=R_t-\alpha C_t$ with $\alpha >0$ and $C_t\ge 0$, so $R_{t+1}\le R_t$ for all $t$. Equivalently, the deficit $P_t=-R_t$ satisfies $P_{t+1}=P_t+\alpha C_t$. Strict monotonicity follows from $C_t>0$ whenever $B_t>0$. Suppose an internal equilibrium $P^{\ast }$ exists with $P_{t+1}^{\ast }=P^{\ast }$. Then $\alpha C^{\ast }=0$, which requires $C^{\ast }=0$, which by $g^{\prime }>0$ requires $B^{\ast }=0$, which by $f^{\prime }>0$ requires $P^{\ast }=0$. The unique equilibrium is therefore $P^{\ast }=0$: the homeostatic state of zero caloric deficit. For any $P_0>0$, the monotonicity gives $P_t\to +\infty$ as $t\to \infty$, but this divergence is bounded in practice by clinical intervention, physiological collapse, or death — never by the cycle's internal dynamics. The binge-starve cycle is thus formally a positive-feedback system whose only fixed point is the starvation-free baseline, and the deviation from that baseline grows strictly with each iteration absent external correction. $\square$

Corollary (Weight restoration precedes cognitive restructuring). Under the assumptions of the Proposition, treating $B_t$ or $C_t$ directly via cognitive restructuring while $P_t>0$ cannot move the system to $P^{\ast }=0$, because the cycle's internal dynamics regenerate $B_t$ and $C_t$ at each step from the maintained deficit. The deficit must be closed by external nutritional input before cognitive restructuring can stabilise the system.

This formalises the clinical principle, established empirically by the Minnesota experiment and operationalised in Maudsley FBT phase one and CBT-E behavioural activation, that weight restoration is the necessary precondition for effective psychological treatment of anorexia nervosa. Treating the cognition of a starved brain is empirically unreliable because the starvation state is generating the pathology faster than cognitive techniques can address it.

Connections Master

• Psychological disorders survey 29.09.01 is the chapter anchor and prerequisite: this unit deepens the brief treatment of eating disorders sketched in the survey's diagnostic-categories section into a full bio-psychosocial account grounded in the Keys 1950 experiment. The general framework of 29.09.01 — diagnostic validity, the limits of DSM classification, the harmful-dysfunction analysis — is exactly the framework within which the eating-disorders spectrum appears as a case study in how a shared underlying maintenance mechanism (the binge-starve cycle) produces multiple categorical diagnoses.

• Cellular respiration: glycolysis and CAC 17.04.01 provides the metabolic-physiology substrate for the biology of starvation. The Minnesota subjects' $40$ percent reduction in basal metabolic rate, their bradycardia (mean heart rate of $\sim 35$ beats per minute), and the refeeding-syndrome phosphate crash all derive from the cellular-energy machinery — glycolysis, the citric-acid cycle, oxidative phosphorylation — operating under chronic substrate depletion and then facing an acute insulin-driven substrate shift. The bridge from the psychology of eating disorders to cellular metabolism is the mechanism by which starvation produces the cognitive and behavioural symptoms that the Minnesota experiment catalogued.

• The neuroscience of consciousness: neural correlates, anesthesia, and disorders of consciousness 20.06.04 supplies the cortical-representation framework within which body-image distortion in anorexia nervosa is now understood. Modern neuroimaging (Treasure et al. 2015) ^{[Treasure2015]} has identified altered processing in the insula (interoception), parietal cortex (body schema), and anterior cingulate cortex (salience and reward) in actively starved and recovered AN patients. The insula is the same cortical region whose role in interoceptive awareness anchors contemporary neuroscience of consciousness; its disruption in AN is a clinical instance of the more general principle that conscious body representation is a constructed cortical representation, not a readout of the body's actual state.

Historical & philosophical context Master

William Gull delivered his lecture "Anorexia Nervosa (Apepsia Hysterica)" to the Clinical Society of London on 24 October 1873 and published it in the Society's Transactions the following year ^[Gull1873]; Ernest-Charles Lasègue published an independent French description in the Archives Générales de Médecine the same year ^{[Lasègue1873]}. Both physicians insisted on distinguishing the condition from tubercular wasting and from purely organic disease, attributing it to a morbid mental state while demonstrating that it could be reversed by medically supervised feeding. The Lasègue–Gull descriptions are the historical root of the modern diagnosis, and the question they implicitly raised — whether the disorder is primarily psychological, physiological, or cultural — has structured a century and a half of subsequent debate.

Ancel Keys's Minnesota Starvation Experiment, conducted 1944–45 and published in 1950 as the two-volume The Biology of Human Starvation ^[Keys1950], transformed the conceptual landscape by demonstrating that semi-starvation in healthy adults produces eating-disorder-like symptoms. Russell's 1979 paper in Psychological Medicine ^{[Russell1979]} named bulimia nervosa. Bruch's 1973 Eating Disorders ^[Bruch1973] reframed anorexia as a disorder of self. Fairburn's 2008 Cognitive Behavior Therapy and Eating Disorders ^{[Fairburn2008]} unified treatment under the transdiagnostic CBT-E protocol. Lock and Le Grange's Maudsley Family-Based Treatment manual ^{[LockLeGrange2012]} established the family as the primary treatment agent for adolescent AN. DSM-5 ^[DSM5_2013] promoted binge-eating disorder to a full diagnosis in 2013, and the Heruc et al. 2020 guideline update ^[Heruc2020] codified modern refeeding-syndrome management.

Bibliography Master

@article{Gull1874,
  author = {Gull, William Withey},
  title = {Anorexia nervosa (apepsia hysterica, anorexia hysterica)},
  journal = {Transactions of the Clinical Society of London},
  volume = {7},
  year = {1874},
  pages = {22--28},
}

@article{Lasegue1873,
  author = {Lasègue, Ernest-Charles},
  title = {De l'anorexie hystérique},
  journal = {Archives Générales de Médecine},
  volume = {21},
  year = {1873},
  pages = {385--403},
}

@article{Russell1979,
  author = {Russell, Gerald F. M.},
  title = {Bulimia nervosa: an ominous variant of anorexia nervosa},
  journal = {Psychological Medicine},
  volume = {9},
  number = {3},
  year = {1979},
  pages = {429--448},
}

@book{Keys1950,
  author = {Keys, Ancel and Brožek, Josef and Henschel, Austin and Mickelsen, Olaf and Taylor, Henry Longstreet},
  title = {The Biology of Human Starvation},
  publisher = {University of Minnesota Press},
  address = {Minneapolis},
  year = {1950},
  note = {2 volumes},
}

@book{Fairburn2008,
  author = {Fairburn, Christopher G.},
  title = {Cognitive Behavior Therapy and Eating Disorders},
  publisher = {Guilford Press},
  address = {New York},
  year = {2008},
}

@book{LockLeGrange2012,
  author = {Lock, James and Le Grange, Daniel},
  title = {Treatment Manual for Anorexia Nervosa: A Family-Based Approach},
  edition = {2nd},
  publisher = {Guilford Press},
  address = {New York},
  year = {2012},
}

@book{Bruch1973,
  author = {Bruch, Hilde},
  title = {Eating Disorders: Obesity, Anorexia Nervosa, and the Person Within},
  publisher = {Basic Books},
  address = {New York},
  year = {1973},
}

@book{DSM5_2013,
  author = {{American Psychiatric Association}},
  title = {Diagnostic and Statistical Manual of Mental Disorders},
  edition = {5th},
  publisher = {American Psychiatric Publishing},
  address = {Arlington, VA},
  year = {2013},
}

@article{Treasure2015,
  author = {Treasure, Janet and Zipfel, Stephan and Micali, Nadia and Wade, Tracey and Stice, Eric and Claudino, Angélica and Schmidt, Ulrike and Frank, Guido K. W. and Bulik, Cynthia M. and Wentz, Elisabet},
  title = {Anorexia nervosa},
  journal = {Nature Reviews Disease Primers},
  volume = {1},
  year = {2015},
  pages = {15074},
}

@article{GarnerGarfinkel1980,
  author = {Garner, David M. and Garfinkel, Paul E.},
  title = {Socio-cultural factors in the development of anorexia nervosa},
  journal = {Psychological Medicine},
  volume = {10},
  number = {4},
  year = {1980},
  pages = {647--656},
}

@article{Arcelus2011,
  author = {Arcelus, Jon and Mitchell, Alex J. and Wales, Jackie and Nielsen, Søren},
  title = {Mortality rates in patients with anorexia nervosa and other eating disorders: a meta-analysis of 36 studies},
  journal = {Archives of General Psychiatry},
  volume = {68},
  number = {7},
  year = {2011},
  pages = {724--731},
}

@article{Heruc2020,
  author = {Heruc, Gabriella and Hurst, Kim and Casey, Anthea and Fleming, Kim and Freeman, Julia and Fursland, Anthea and Hart, Susan and Hodgson, Janet and Jovanovska, Marianne and Kohn, Michael and Mann, Phillipa and McMeniman, Julie and Miskovic-Wheatley, Jane and Mond, Jonathan and Paxton, Susan and Robertson, Mark and Rodan, Amy and Sacks, Morgan and Sarian, Paige and Watson, Hunna and Maguire, Sarah},
  title = {ANZAHPE Eating Disorder Treatment Principles and Practice},
  journal = {Journal of Eating Disorders},
  volume = {8},
  year = {2020},
  pages = {65},
}