Anorexia nervosa is officially classified as a psychiatric disorder, not a neurological one. But that clean distinction is becoming harder to defend. Brain imaging, genetics, and neuroscience research over the past decade have revealed that anorexia involves measurable changes in brain structure, reward processing, and neural circuitry that go far beyond willpower or body image. The emerging picture is of a condition with deep neurobiological roots, even if the medical system still files it under mental health.
How Anorexia Is Currently Classified
The DSM-5, the diagnostic manual used by mental health professionals, categorizes anorexia nervosa as a feeding and eating disorder. Severity is graded by BMI: mild (BMI of 17 or above), moderate (16 to 16.99), severe (15 to 15.99), and extreme (below 15). That weight-based scale has drawn criticism, since it leaves out cognitive, emotional, and biological dimensions that also drive severity.
The Academy for Eating Disorders has pushed to reframe eating disorders as “biologically-based, serious mental illnesses,” placing them alongside schizophrenia and bipolar disorder. Their position, endorsed by the director of the National Institute of Mental Health, is that genetic and neuroimaging evidence gives anorexia a biological basis comparable to other conditions widely accepted as brain disorders. The key phrase in the field right now is “biomedical as well as behavioral,” meaning that treating anorexia as purely psychological misses a significant part of what’s going on.
What Happens to the Brain During Anorexia
People with anorexia show measurable reductions in brain volume. In adolescents with the restricting type of anorexia, researchers have found significantly lower gray matter volumes in both frontal lobes and the left insula, a region critical for sensing what’s happening inside the body. Reduced volume in the anterior cingulate cortex, which helps regulate emotion and decision-making, has also been documented in early-stage illness.
These aren’t subtle findings. The frontal lobes govern planning, impulse control, and flexible thinking. The insula processes hunger, fullness, heartbeat, and gut sensations. When these regions shrink, the brain’s ability to accurately read internal signals and make flexible decisions about food is physically compromised. This helps explain why people with anorexia can genuinely not feel hungry, or feel full after eating very little. It’s not just psychological denial; the hardware that processes those signals is altered.
Reward Circuits Respond Differently to Food
One of the more striking findings involves the brain’s reward system. In a study comparing women with anorexia, women with obesity, and healthy controls, researchers found that the dopamine-driven reward circuits in the brain responded more strongly to food-related stimuli in the anorexia group, not less. Areas including the ventral striatum, insula, and prefrontal cortex all showed heightened activation during a reward-learning task involving taste. Women with obesity showed the opposite pattern: dampened reward responses.
This counterintuitive finding challenges the assumption that people with anorexia simply don’t enjoy food. Instead, food may be intensely stimulating to their reward circuits, which could make eating feel overwhelming rather than pleasurable. The avoidance of food, in this model, is partly an attempt to manage an overactive neural response, not the absence of one.
Faulty Body Signals and the Insula
The insula plays a central role in interoception, the brain’s ability to sense internal body states like hunger, heart rate, and stomach fullness. Research using brain imaging in weight-restored anorexia patients found that two distinct regions of the insula responded abnormally during tasks that required paying attention to stomach and heart sensations. The dorsal mid-insula showed increased activation during anxious rumination, and activity in this region during stomach-focused attention correlated with anxiety levels and eating disorder severity.
Critically, these differences persisted even after weight restoration, suggesting the altered insula function isn’t simply a consequence of starvation. It may be a trait that predates the illness or one that starvation triggers but doesn’t fully reverse. Either way, it points to a neurological mechanism underlying the difficulty people with anorexia have in trusting or interpreting their own body’s signals.
How Starvation Becomes a Habit
One of the most compelling neurological explanations for anorexia’s persistence involves the brain’s habit circuits. Normally, behaviors start as goal-directed (consciously chosen) and gradually become habitual (automatic) through repetition. Research has identified dysfunction in the frontostriatal circuits, the connections between the frontal cortex and the striatum, that govern this transition in people with anorexia.
Studies using food-choice tasks have found that activity in the caudate, a key part of the striatum, predicts both caloric intake and how patients’ food choices change after treatment. In one study, caudate activity during food decisions correlated with scores on a habit measure, suggesting that food restriction may become neurologically encoded as a habit rather than remaining a conscious choice. Animal research supports this: disrupting normal function in the dorsal striatum in mice shifted behavior from goal-directed to habitual, providing a direct mechanistic link between neural dysfunction and compulsive eating patterns.
This matters because it means that at some point, the decision to restrict food may stop being a decision at all. The brain automates it, the way it automates brushing your teeth or taking the same route to work. That’s a neurological process, not a failure of willpower.
Genetics Point to a Metabolic-Psychiatric Hybrid
A landmark genome-wide association study of nearly 17,000 anorexia cases and 55,500 controls identified eight genetic risk locations and revealed something unexpected. Anorexia shares genetic overlap not only with psychiatric disorders like depression and anxiety, but also with metabolic traits involving blood sugar regulation, lipid levels, and body composition. These metabolic correlations held up even after accounting for the effects of low body weight itself.
The researchers described anorexia’s genetic architecture as “metabo-psychiatric,” a term that captures how the disorder sits at the intersection of metabolic and mental health. This genetic profile helps explain why some people seem biologically primed to maintain extremely low weight, and why weight restoration alone doesn’t resolve the illness. Their metabolism and their psychology are intertwined at the genetic level.
Brain Changes Are Largely Reversible
The encouraging news is that most of the brain’s structural damage from anorexia recovers with weight restoration. A study tracking cortical thickness and subcortical volumes during treatment found that brain restoration was most pronounced in the first phase of weight gain, with cortical thickness increasing by an average of 0.08 mm during early treatment and another 0.04 mm in the later phase. Subcortical structures including the hippocampus (memory), amygdala (emotion), caudate (habits), and thalamus (sensory relay) all showed significant volume increases during treatment.
By the time patients reached a BMI of 17.5 or above, most structural changes had reversed. However, one area lagged behind: the superior frontal cortex, where residual thinning remained even after weight restoration. This region is involved in self-monitoring and cognitive control, which may partly explain why recovered patients can still struggle with rigid thinking patterns even after their weight normalizes. Importantly, the duration of illness didn’t determine how well the brain recovered, meaning even patients who had been ill for years showed comparable structural restoration.
Experimental Brain Stimulation Treatments
The neurological evidence has opened the door to treatments that target the brain directly. Deep brain stimulation, which delivers small electrical pulses through implanted electrodes, has been tested in patients with severe, treatment-resistant anorexia. Two brain targets have shown the most promise: the subcallosal cingulate, a region involved in mood regulation, and the nucleus accumbens, part of the brain’s reward circuit.
In a randomized trial of eight patients, five showed at least a 10% increase in BMI after six months of stimulation, along with significant improvements in quality of life. Smaller case series have reported even more dramatic results. Two adolescents who received stimulation to the nucleus accumbens reached a normal BMI within 12 months. In another series of four adolescents, body weight increased by up to 65% over follow-up periods ranging from nine months to over four years.
These are experimental treatments, tested only in people for whom nothing else has worked. But the fact that electrically stimulating specific brain circuits can reverse entrenched starvation behavior is itself powerful evidence that anorexia operates through neurological mechanisms. You don’t treat a purely psychological problem by stimulating the nucleus accumbens.
A Disorder That Crosses Categories
The honest answer to whether anorexia is a neurological disorder is that it doesn’t fit neatly into either the neurological or the psychiatric box. It involves measurable brain changes, altered neural circuitry, disrupted interoception, habit-encoded behavior, and a genetic profile that spans both metabolic and psychiatric traits. At the same time, psychological factors like perfectionism, trauma, and social pressure clearly play a role in who develops it and how it manifests.
What the neuroscience makes clear is that anorexia is not a lifestyle choice, a phase, or a problem of vanity. It is a condition with biological machinery driving it, machinery that can be seen on brain scans, traced through genetic data, and in some cases directly manipulated with electrodes. Calling it purely psychiatric understates what’s happening in the brain. Calling it purely neurological ignores the psychological and social dimensions. The most accurate framing, and the one gaining ground among researchers, is that anorexia is a biologically based mental illness with significant neurological components.