Anorexia Nervosa: Diagnosis, Neurobiology, Medical Complications, and Evidence-Based Treatment

Anorexia nervosa (AN) is a severe psychiatric illness characterized by restriction of energy intake leading to significantly low body weight, intense fear of gaining weight, and disturbance in the way one experiences body weight or shape. Among all mental disorders, AN carries the highest mortality rate — a distinction that underscores the urgency of early detection and intervention.

The standardized mortality ratio (SMR) for anorexia nervosa is approximately 5.86, meaning individuals with AN die at nearly six times the rate of the age-matched general population. Roughly one in five deaths in AN is attributable to suicide, while the remainder result from the cumulative medical toll of prolonged starvation, cardiac arrhythmias, and organ failure. Despite these statistics, AN remains widely misunderstood — frequently trivialized as a lifestyle choice or phase of adolescence rather than recognized as a biologically grounded, psychologically entrenched illness with identifiable neurobiological substrates.

What makes AN particularly challenging from a clinical standpoint is its ego-syntonic nature. Unlike most psychiatric conditions, where symptoms are experienced as distressing and unwanted (ego-dystonic), many individuals with AN perceive their restriction and weight loss as achievements rather than symptoms. This fundamental feature shapes every aspect of treatment engagement — from initial help-seeking (often delayed by years) to therapeutic alliance, motivation for weight restoration, and relapse prevention. Clinicians treating AN must work within this paradox: the illness feels protective and identity-defining to the very person it is killing.

This article provides a detailed clinical overview of AN — its diagnostic classification, epidemiology, neurobiological underpinnings, psychological maintaining factors, medical complications across organ systems, evidence-based treatment approaches, and long-term prognosis.

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR) defines anorexia nervosa according to three core criteria:

1. Restriction of energy intake relative to requirements, leading to a significantly low body weight in the context of the individual's age, sex, developmental trajectory, and physical health. "Significantly low weight" is generally operationalized as a BMI below 18.5 kg/m² in adults, though clinical judgment is required — a BMI of 19 in someone who previously maintained a BMI of 28 may represent clinically significant weight loss even if it falls within a "normal" range.
2. Intense fear of gaining weight or of becoming fat, or persistent behavior that interferes with weight gain, even though the individual is at a significantly low weight. This criterion captures not only the subjective dread of weight gain but also the behavioral patterns — compulsive exercise, food rituals, body checking — that maintain the disorder.
3. 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.

The DSM-5-TR specifies two subtypes:

• Restricting type: During the last three months, the individual has not engaged in recurrent episodes of binge eating or purging behavior (self-induced vomiting, misuse of laxatives, diuretics, or enemas). Weight loss is accomplished primarily through dieting, fasting, or excessive exercise.
• Binge-eating/purging type: During the last three months, the individual has engaged in recurrent binge-eating or purging episodes. This subtype carries distinct medical risks, including electrolyte derangements from purging and dental erosion, and tends to be associated with greater psychiatric comorbidity, including impulsivity and substance use.

Severity is coded by BMI in adults: Mild (BMI ≥ 17), Moderate (16–16.99), Severe (15–15.99), and Extreme (BMI < 15). These thresholds are guidelines rather than rigid cutoffs; clinical severity also depends on the rate of weight loss, medical stability, and functional impairment.

The lifetime prevalence of AN among women is estimated at 0.9% to 4%, depending on the diagnostic threshold applied and the population studied. Among men, lifetime prevalence is approximately 0.3%, though male cases are likely underdiagnosed due to clinical bias and differences in symptom presentation — men with AN may focus on muscularity and leanness rather than thinness per se. The peak age of onset is between 14 and 19 years, though AN can develop at any age, including in children under 12 and in adults over 40.

Historically, AN was characterized as a disorder predominantly affecting young, white, affluent women. This stereotype has been systematically dismantled by epidemiological research showing that AN occurs across racial, ethnic, and socioeconomic groups. A large U.S. study by Marques and colleagues (2011) found that lifetime prevalence of AN among Latinos was comparable to that among non-Latino whites. Diagnostic and treatment disparities — not true prevalence differences — largely account for the historical demographic skew.

Mortality data are sobering. Arcelus and colleagues (2011), in a meta-analysis of 36 studies covering over 17,000 patients, reported a weighted annual mortality rate of 5.10 deaths per 1,000 person-years for AN — substantially higher than for bulimia nervosa (1.74) or eating disorders not otherwise specified (3.31). The aggregate SMR was 5.86. Causes of death include cardiac arrest, multi-organ failure, and suicide. Among those who die by suicide, the risk is elevated approximately 18-fold compared to age-matched women in the general population.

These mortality figures place AN alongside substance use disorders and major depressive disorder as one of the deadliest psychiatric conditions, but with the critical distinction that AN has the highest mortality rate when examined per capita among those diagnosed.

Anorexia nervosa is not simply a disorder of willpower or vanity. A growing body of neuroscience research has identified specific neurobiological vulnerabilities that predate the illness, are exacerbated by starvation, and often persist after weight restoration.

Serotonin dysregulation: The serotonergic system is consistently implicated in AN. Walter Kaye and colleagues at the University of California San Diego have demonstrated that individuals recovered from AN show elevated cerebrospinal fluid levels of 5-hydroxyindoleacetic acid (5-HIAA), a serotonin metabolite, suggesting trait-level serotonergic overactivity. This excess serotonin tone may produce a baseline state of anxiety, harm avoidance, and behavioral inhibition. Paradoxically, dietary restriction reduces tryptophan availability (the amino acid precursor to serotonin), which may temporarily lower serotonin activity and produce a transient reduction in anxiety — effectively making starvation self-reinforcing as a form of anxiolysis.

Altered interoception: Interoception — the ability to perceive internal bodily signals such as hunger, satiety, and heartbeat — is impaired in AN. Functional neuroimaging studies reveal decreased activation in the anterior insula, a brain region central to interoceptive processing, during tasks requiring participants to attend to bodily states. This deficit means that individuals with AN may genuinely have diminished awareness of hunger cues, making claims of "not feeling hungry" neurologically plausible rather than purely volitional.

Reward processing differences: The dopaminergic reward circuitry in AN is atypical. While food is normally rewarding, individuals with AN show reduced activation in the ventral striatum in response to food stimuli and, in some studies, increased activation in the dorsal striatum — a region associated with habit formation rather than pleasure. This suggests that restrictive eating becomes a habitual, compulsive pattern rather than a consciously chosen behavior. Additionally, research by Fladung and colleagues (2010) demonstrated that underweight images of the self activated reward circuits in individuals with AN, suggesting that the experience of emaciation itself becomes neurobiologically rewarding.

While neurobiology provides the substrate for vulnerability, a set of well-characterized psychological factors maintain AN once it is established. Understanding these maintaining factors is essential for effective psychotherapy.

Cognitive rigidity: Neuropsychological testing consistently shows that individuals with AN demonstrate poor set-shifting — the ability to flexibly move between tasks, mental sets, or strategies. This cognitive rigidity manifests clinically as rigid food rules, black-and-white thinking about eating ("good" versus "bad" foods), and difficulty adapting to new information or circumstances. Importantly, set-shifting deficits appear to be a trait marker rather than a consequence of starvation alone, as they persist in weight-restored individuals and are found at elevated rates in unaffected first-degree relatives.

Perfectionism: Clinical perfectionism — the tendency to base self-worth almost entirely on striving for and achieving demanding standards — is both a risk factor for developing AN and a maintenance factor once the illness is established. Weight control and dietary restriction become domains in which perfectionist standards can be rigorously applied and "success" clearly measured. Fairburn and colleagues have highlighted how perfectionism operates transdiagnostically across eating disorders as a core maintaining mechanism in the cognitive-behavioral model.

Fear of weight gain: This fear is not ordinary body dissatisfaction. It is intense, pervasive, and often described in phobic terms — patients report visceral dread, panic, and disgust at the prospect of gaining even small amounts of weight. This phobic response drives avoidance behavior (restriction, exercise, body checking) and typically intensifies rather than diminishes during weight restoration, making the early phases of treatment psychologically brutal for the patient.

Ego-syntonic nature: Perhaps the single greatest obstacle to treatment is that AN is experienced as aligned with the person's values and identity. The discipline of restriction feels like strength; the visible thinness feels like accomplishment. Unlike depression or psychosis, where the sufferer typically wants relief, many individuals with AN are ambivalent about recovery or actively resistant to it. This ego-syntonicity explains the high treatment dropout rates (up to 50% in some outpatient studies) and the frequent need for external motivation — from family members, treatment teams, or medical crisis — to initiate and sustain treatment.

Chronic caloric restriction and malnutrition affect virtually every organ system. Clinicians managing AN must monitor for a range of medical complications, many of which can be life-threatening.

Cardiac: Cardiovascular complications are the leading cause of death in AN. Bradycardia (resting heart rate < 60 bpm, often < 40 bpm in severe AN) results from reduced metabolic demand and is nearly universal in underweight patients. QT prolongation — an ominous ECG finding — increases the risk of torsades de pointes and sudden cardiac death, particularly in the context of electrolyte disturbances (hypokalemia, hypomagnesemia). Mitral valve prolapse occurs in up to one-third of patients due to loss of cardiac muscle mass disproportionate to valve size. Pericardial effusion and reduced left ventricular mass are also common.

Endocrine: Hypothalamic suppression leads to functional hypothalamic amenorrhea in females, with cessation of menstruation resulting from low gonadotropin-releasing hormone pulsatility. Loss of estrogen (and testosterone in males) contributes to bone density loss — osteopenia is found in up to 92% and osteoporosis in 38% of women with AN, with fracture risk elevated years after recovery. Cortisol is characteristically elevated due to HPA axis activation in starvation, while thyroid function shows a sick euthyroid pattern (low T3, normal or low TSH). Growth hormone resistance is common, with elevated GH but low IGF-1.

Gastrointestinal: Gastroparesis — delayed gastric emptying — produces early satiety, bloating, and nausea, paradoxically reinforcing food avoidance. Constipation is nearly universal due to reduced gut motility and decreased dietary volume. Superior mesenteric artery syndrome, in which the duodenum is compressed between the aorta and SMA due to loss of the intervening fat pad, occurs in severely underweight patients. Refeeding syndrome — the most dangerous complication of nutritional rehabilitation — is discussed in the treatment section.

Hematological: Pancytopenia (low red cells, white cells, and platelets) results from bone marrow hypoplasia due to gelatinous transformation of marrow fat. Leukopenia may be marked, yet clinical infections remain relatively uncommon until very severe malnutrition supervenes. Anemia is typically normocytic and normochromic.

Treatment of AN requires a multidisciplinary approach integrating nutritional rehabilitation, psychological therapy, medical monitoring, and — for adolescents — family involvement. No psychiatric medication has demonstrated robust efficacy for the core symptoms of AN.

Family-Based Treatment (FBT / Maudsley Approach): For adolescents with AN, FBT is the treatment with the strongest evidence base. Developed at the Maudsley Hospital in London and manualized by Lock and Le Grange, FBT proceeds through three phases: Phase 1 empowers parents to take charge of their child's eating and restore weight; Phase 2 gradually returns eating autonomy to the adolescent; Phase 3 addresses broader developmental issues. Randomized trials have shown that 50–75% of adolescents treated with FBT achieve full remission by end of treatment, with gains largely maintained at 5-year follow-up. FBT's superiority over individual adolescent therapy was demonstrated in the landmark study by Lock and colleagues (2010).

Enhanced Cognitive-Behavioral Therapy (CBT-E): Developed by Christopher Fairburn at Oxford, CBT-E is a transdiagnostic treatment for eating disorders that has been adapted for underweight patients. The AN version (CBT-E for low weight) includes an initial phase focused on motivation and engagement, followed by targeted work on the overvaluation of shape, weight, and control, along with addressing maintaining factors such as perfectionism, interpersonal difficulties, and low self-esteem. Evidence for CBT-E in adults with AN is promising, though effect sizes are more modest than in bulimia nervosa, and dropout rates remain high.

Specialist Supportive Clinical Management (SSCM): SSCM combines clinical management of eating and weight with supportive psychotherapy techniques. In the multisite ANTOP trial and other studies, SSCM performed comparably to more structured psychotherapies, suggesting that a warm therapeutic relationship combined with nutritional counseling and psychoeducation can be effective for some adults with AN.

Inpatient criteria and refeeding protocols: Hospitalization is indicated for medical instability (heart rate < 40 bpm, blood pressure < 90/60 mmHg, temperature < 36°C, electrolyte derangement, BMI < 15 with rapid decline, or suicidality). Refeeding syndrome — characterized by dangerous shifts in phosphate, potassium, and magnesium when nutrition is reintroduced to a starved body — requires careful management. Current protocols typically begin at 1,200–1,500 kcal/day (or higher in some recent "higher calorie refeeding" protocols with close electrolyte monitoring), with phosphate supplementation and twice-daily electrolyte checks during the first week. The Junior MARSIPAN and MARSIPAN guidelines (Royal College of Psychiatrists) provide structured frameworks for medical risk assessment.

Long-term outcome in AN has traditionally been summarized by the "rule of thirds" — a heuristic derived from decades of follow-up studies. Approximately one-third of individuals with AN achieve full recovery (normalization of weight, eating behavior, and psychological relationship with food and body). Another third show significant improvement but retain residual symptoms — partial weight restoration, persistent body image disturbance, or subclinical dietary restriction. The final third follow a chronic, unremitting course, often cycling through repeated hospitalizations, and face the highest risk of premature death.

More recent longitudinal data have refined this picture. Eddy and colleagues (2017), following a cohort for over 20 years, found that approximately 63% of those with restricting-type AN and 68% with binge-purge type eventually achieved full recovery — more optimistic figures than the rule of thirds would suggest — though recovery often took a decade or longer. The median time to recovery was approximately 9 years, emphasizing that AN is frequently a long-duration illness even in those who ultimately recover.

Prognostic factors favoring recovery include younger age at onset, shorter duration of illness before treatment, higher BMI at presentation, absence of binge-purge behaviors, and strong family support. Poor prognostic indicators include later onset, longer illness duration, lower nadir BMI, comorbid personality disorder, and poor social adjustment.

The chronicity of AN in a substantial minority of patients has prompted growing interest in harm-reduction models for individuals with severe and enduring AN (SE-AN), generally defined as illness lasting more than 7 years with failure to respond to multiple evidence-based treatments. These approaches prioritize quality of life, medical stability, and psychological well-being over aggressive weight restoration — a pragmatic if controversial shift in treatment philosophy.

Regardless of the outcome trajectory, early intervention remains the single most consistent predictor of favorable prognosis. Each year of untreated AN reduces the probability of full recovery and increases the medical and psychological burden of the illness.