How Does Anorexia Affect the Brain?

Anorexia nervosa reshapes the brain in measurable, widespread ways. Prolonged starvation shrinks both gray matter and white matter, disrupts the chemistry of mood-regulating neurotransmitters like serotonin, rewires the reward system, and alters how hunger hormones communicate with the hypothalamus. Many of these changes begin reversing with weight restoration, but some linger for months or years, and in chronic cases, certain differences persist even after recovery.

Gray Matter and White Matter Shrinkage

The most immediately visible effect of anorexia on the brain is a loss of volume. A large multicenter neuroimaging study found significant gray matter reductions in people with anorexia across ten distinct brain clusters, including the cerebellum, the frontal lobe, the temporal lobe, the parietal lobe, the insula, the cingulate cortex, and the thalamus.¹ In other words, the shrinkage is not limited to one area. It touches regions involved in movement, emotion, decision-making, memory, and sensory processing.

White matter, the insulated nerve fibers that connect different brain regions, also takes a hit. In adolescent patients, both cortical and subcortical gray matter and cortical white matter were associated with low weight at the time of hospital admission.² The overall picture is one of a brain that has been materially reduced in size by malnutrition.

Alongside the tissue loss, the fluid-filled spaces inside the brain, called ventricles, expand. Researchers have documented this for decades. In one study, total ventricular volume in patients with anorexia was significantly elevated on admission and dropped back into the normal range after refeeding.³ Because the tissue loss reverses with adequate nutrition in many cases, researchers sometimes call it “pseudoatrophy,” distinguishing it from the permanent atrophy seen in neurodegenerative diseases.⁴ That label captures the optimistic side: much of the damage is reversible. But “pseudo” can be misleading, because for as long as the malnourishment continues, the brain is genuinely smaller and functioning differently.

Serotonin and the Chemistry of Anxiety

Anorexia does not just physically shrink the brain. It alters the chemical signaling systems running through it. Serotonin, the neurotransmitter most associated with mood regulation and anxiety, is one of the most consistently affected. Research using brain imaging with radioactive tracers has found dysregulation of serotonin pathways in both cortical and limbic structures, implicating changes in specific serotonin receptor types and the serotonin transporter.⁵ These alterations are linked to the anxiety, behavioral inhibition, and body image distortion that characterize the disorder.

What makes the serotonin story especially interesting is that the disruption does not vanish immediately when weight is restored. In women who had fully recovered from restricting-type anorexia, a specific serotonin receptor in the temporal and cingulate regions of the brain was still abnormally active, and the degree of that activity correlated with how much anxiety and harm avoidance the person reported.⁶ This raises a genuine chicken-or-egg question: did the serotonin abnormality precede the eating disorder, or did starvation create a lasting chemical scar? The honest answer is that researchers are still trying to untangle it, and it may be both.

Rewired Reward Circuits

Eating is supposed to feel rewarding. Your brain’s reward system, centered on a loop of structures including the striatum, insula, and orbitofrontal cortex, is designed to make food appealing and to reinforce eating behavior. In anorexia, that system goes haywire. Structural and functional changes in the insula and frontal cortex appear in people with current or past eating disorders compared to healthy controls, and researchers have proposed that these changes may both predispose someone to developing the disorder and then deepen once starvation begins, promoting illness behavior and making recovery harder.⁷

There is an important nuance here. People with anorexia do not seem to have a general inability to experience pleasure. They can enjoy music, social connection, achievement. The problem is more targeted: food-related and taste-related stimuli get reappraised as aversive, while cues compatible with weight loss and restriction become rewarding. Evidence supports the idea that this pattern shares neural overlaps with addiction, where the “habit” of restriction gets reinforced through the brain’s striatal reward system in a self-perpetuating cycle.⁸ The person is not choosing to starve through pure willpower; their brain has literally restructured what it finds rewarding.

Cognitive Flexibility and Detail-Focused Thinking

People with anorexia tend to show specific patterns in how they think. Two of the most well-documented are difficulty with cognitive flexibility, the ability to shift between tasks or mental frameworks, and a processing style called weak central coherence, which means focusing heavily on fine details at the expense of the bigger picture.⁹ In practical terms, this can look like extreme rigidity around food rules, perfectionism, difficulty adapting to new situations, and trouble seeing the forest for the trees.

These cognitive patterns appear in studies using standardized neuropsychological tests, and people with eating disorders consistently perform worse on measures of decision-making and cognitive flexibility compared to healthy controls.¹⁰ The weak central coherence pattern has been observed across all eating disorder subtypes and even in unaffected relatives of people with anorexia, suggesting it may be at least partly a trait-level vulnerability rather than purely a consequence of starvation.¹¹ The fact that family members who have never been malnourished show similar tendencies is one of the stronger pieces of evidence that anorexia is not purely an illness of starvation: the brain may have been wired differently before the disorder ever began.

That said, the picture is not perfectly consistent. One study of adolescents with anorexia found no significant differences from controls on set-shifting tasks, despite the well-established findings in adults.¹² This could mean the cognitive rigidity worsens with longer illness duration rather than being present from the start, or it could reflect the fact that adolescent brains are still developing and harder to compare against a single benchmark. Either way, weaker cognitive flexibility and central coherence are correlated with poorer quality of life in people with anorexia, especially the restrictive subtype.¹³

The Hypothalamus and Hunger Hormones

The hypothalamus is the brain’s thermostat for hunger, metabolism, and energy balance. It receives hormonal signals from the body, like leptin (which signals how much fat you have stored) and ghrelin (which signals that you need to eat), and adjusts behavior and metabolism accordingly. In anorexia, this system breaks down. People with anorexia have been found to have reduced volume in the hypothalamus overall, with specific shrinkage in the posterior and inferior tuberal subregions, compared to healthy controls.¹⁴

Leptin levels are extremely low in people with anorexia, because they have very little body fat. But the relationship between leptin and the hypothalamus appears to go beyond just low levels. In people with anorexia, lower leptin concentrations were associated with decreased volume in the anterior-inferior subregion of the hypothalamus, suggesting a disrupted feedback loop between the brain and the hormonal system that is supposed to drive eating behavior.¹⁴ Researchers have characterized this as evidence that anorexia may be a “metabo-psychiatric disorder,” meaning the metabolic and psychiatric components feed each other in ways that are difficult to separate.

The hippocampus, a region critical for memory and stress regulation, is also affected. Patients with anorexia show structural alterations in the hippocampus alongside higher perceived stress and poorer coping ability compared to healthy controls.¹⁵ Since the hippocampus is a key part of the brain’s stress response system, changes there could amplify the emotional distress that makes recovery so difficult.

Why Adolescent Brains Are Especially Vulnerable

Anorexia most commonly begins during adolescence, and this timing matters enormously for the brain. The adolescent brain is still under construction. White matter, in particular, is still being insulated with myelin, the fatty coating that helps nerve signals travel quickly and efficiently. When anorexia strikes during this period, the impact on white matter appears to go beyond simple volume loss. A study of adolescent girls with anorexia found reduced myelin content across a large number of white matter tracts, measured using an imaging method that specifically targets myelin rather than just overall tract volume.¹⁶ This is concerning because myelin development during adolescence lays the groundwork for efficient brain communication in adulthood.

Gray matter is also hit early. A study of adolescent girls experiencing their first episode of anorexia found significant reductions in total and cortical gray matter volume compared to typically developing peers, with the frontal and parietal cortices most affected.¹⁷ The frontal cortex is responsible for planning, impulse control, and flexible thinking; the parietal cortex handles spatial awareness and integrating sensory information. That these regions are affected even in the early stages of a first episode underscores how rapidly malnutrition can alter a developing brain.

What Recovers and What Does Not

The encouraging news is that much of the structural damage improves with weight restoration. The general consensus from decades of research is that gray and white matter reduction tied to acute malnutrition is mostly reversible with weight gain, at least in patients whose illness has not become chronic.¹⁸ Cortical thickness shows significant global restoration during treatment, with the fastest gains occurring in the early phases of weight recovery.¹⁹

But “mostly reversible” is not “completely reversible.” Even after significant weight gain, some regions remain thinner than in people who never had the disorder. One study found that after treatment, a residual patch of cortical thinning persisted in the right superior frontal cortex of patients compared to healthy controls.¹⁹ In women who had been weight-recovered for some time, one specific region, the right pars orbitalis (a part of the frontal lobe involved in language and decision-making), still showed reduced cortical thickness compared to healthy controls, while other differences had normalized.²⁰

Outcome seems to track with how completely the brain normalizes. A machine-learning study found that brain structural alterations were more pronounced in patients who were only partially weight-restored and went on to have worse long-term outcomes, while these alterations were not detectable in individuals who achieved full long-term weight recovery.²¹ In other words, the degree of brain restoration may actually help predict who will stay well and who will relapse, which opens the door to using brain scans as a clinical tool rather than just a research measure.

Habitual Behavior and the Frontostriatal Loop

One of the most frustrating features of anorexia, both for patients and their families, is how automatic and entrenched the restriction behaviors become. A person can intellectually understand they need to eat and still find it nearly impossible to do so. Neuroscience is starting to explain why. Research suggests that in people with anorexia, dorsal frontostriatal circuits, the brain pathways that govern habit rather than goal-directed choice, play a larger role in guiding decisions about what to eat than they do in healthy individuals.²²

There is also imaging evidence that the white matter connections in these habit-associated circuits are stronger in people with anorexia. One study reported greater connectivity in premotor and putamen circuits among individuals with anorexia compared to controls, and this connectivity was linked to the severity of ritualistic eating behaviors.²³ Essentially, the brain shifts control of eating behavior from the flexible, deliberate parts of the frontal lobe to the more automatic, habitual parts of the striatum. This helps explain why anorexia is so treatment-resistant: you are not just fighting a belief system, you are fighting a well-worn neural groove.

Resting Brain Networks and Social Processing

Beyond the reward and habit circuits, anorexia also alters broader brain networks that are active even when a person is not doing any particular task. Resting-state brain imaging has revealed reduced functional connectivity in areas associated with processing facial information and social cognition, alongside increased connectivity in regions linked to sensory stimulation and social anxiety.²⁴ Major brain networks including those involved in self-referential thought, executive control, and emotion regulation all show alterations.

This may help explain some of the social difficulties people with anorexia often experience, like difficulty reading other people’s emotions or heightened self-consciousness in social situations. These are not just psychological effects of feeling unwell. They have a measurable neurological substrate: the brain networks that handle social information are literally communicating differently.

Inflammation and the Gut-Brain Connection

An emerging area of research is the role of inflammation and the gut microbiome in anorexia’s brain effects. A case-control study of adolescent girls with anorexia found significantly elevated levels of several inflammatory markers, including interleukin-1β, interleukin-6, and tumor necrosis factor-α, compared to healthy controls.²⁵ This low-grade inflammatory state could contribute to brain changes, since many of these signaling molecules cross the blood-brain barrier and affect neurons directly.

The gut microbiome adds another layer. In a rat model of anorexia, chronic food restriction dramatically altered gut microbial diversity and composition and reduced brain volume. Certain gut bacteria were positively associated with brain matter volume while others were negatively associated, suggesting the microbiome may actively influence whether and how much the brain shrinks during starvation.²⁶ This is still animal research and far from clinical application, but it points toward a future where supporting gut health might be part of protecting the brain during refeeding and recovery.

Taken together, the inflammation and microbiome findings suggest that starvation does not affect the brain through a single mechanism. It works through cascading systems: hormonal disruption, inflammatory signaling, altered gut ecology, and direct nutrient deprivation all converge on the brain simultaneously. Treating any one pathway in isolation likely misses the others, which is part of why a comprehensive approach to recovery, including adequate nutrition, psychological support, and time, remains the standard of care.