What Are Eating Disorders?

Eating disorders are psychiatric conditions marked by persistent disturbance of eating behavior and of the thoughts and emotions attached to food, body weight, and body shape, severe enough to impair physical health or daily functioning. The principal diagnoses are anorexia nervosa, bulimia nervosa, and binge-eating disorder, joined in current diagnostic manuals by avoidant/restrictive food intake disorder, rumination disorder, and pica. They are classified among mental disorders, but they carry direct medical consequences including cardiac arrhythmia, electrolyte disturbance, osteoporosis, and gastrointestinal injury, and anorexia nervosa has one of the highest mortality rates of any psychiatric illness.

The diagnoses differ in behavior rather than in severity. The National Institute of Mental Health describes anorexia nervosa as persistent restriction of intake leading to low body weight, together with distorted body image and intense fear of weight gain; bulimia nervosa as recurrent binge eating followed by compensatory behavior such as purging, fasting, or excessive exercise; and binge-eating disorder as recurrent binge episodes with loss of control but without regular compensatory behavior. People with bulimia nervosa frequently fall within a typical weight range, which is one reason the condition goes undetected.

Risk Factors and Course

Eating disorders arise from interacting genetic, neurobiological, psychological, and social influences rather than from any single cause. Twin studies indicate substantial heritability, and imaging research points to altered reward processing and cognitive control in circuits involving the insula and prefrontal cortex. Onset clusters in adolescence and early adulthood, and comorbidity with anxiety disorders, depression, obsessive-compulsive disorder, and substance use is common. Population figures compiled in national statistics on eating disorder prevalence show lifetime prevalence in the low single-digit percentages, with rates higher among women, though under-recognition in men and in people at higher body weights is well documented.

Behavioral Measurement and Sensing

Because the defining behaviors occur in daily life rather than in clinics, measurement has become an engineering problem. Ecological momentary assessment prompts participants several times a day for self-reports of affect, urges, and eating episodes, capturing timing that retrospective interviews miss. Passive sensing extends that record: wrist-worn inertial sensors detect hand-to-mouth gestures, acoustic and piezoelectric throat sensors detect chewing and swallowing, and photoplethysmography and electrodermal sensors track autonomic arousal that precedes a binge episode. A scoping review of automatic wearable-based eating detection for public health research catalogs the sensor modalities and the accuracy reported for each. Machine learning models trained on these combined streams have anticipated overeating episodes at rates above chance, which supports the design of interventions delivered at the moment of risk, although reported accuracy varies widely across studies, sensor sets, and populations.

Treatment Approaches

Evidence-based treatment differs by diagnosis. Family-based treatment is first-line for adolescents with anorexia nervosa, while enhanced cognitive behavioral therapy is the most studied approach for bulimia nervosa and binge-eating disorder. Nutritional rehabilitation and medical stabilization take priority when weight or electrolyte status is dangerous. Pharmacotherapy has a supporting role, with fluoxetine approved for bulimia nervosa and lisdexamfetamine for binge-eating disorder. Digital delivery, including guided self-help applications and just-in-time adaptive interventions triggered by sensor data, is an active area of clinical trials aimed at the large share of patients who never reach specialist care.

Applications

Eating disorder research has applications in a range of fields, including:

• Wearable sensor and mobile health system design for behavioral monitoring
• Machine learning methods for episode prediction and digital phenotyping
• School and university screening programs and health education curricula
• Clinical informatics for outcome tracking in specialist treatment services
• Human-computer interaction research on social media exposure and body image
• Public health surveillance and mental health services planning