Anorexia nervosa (AN) is characterized by emaciation, hyperactivity, and amenorrhea. To what extent AN-related symptoms are due to food restriction or neuronal dysfunction is currently unknown. Thus, we investigated the relevance of food restriction on AN-related symptoms. Disrupted circadian rhythms are hypothesized to contribute to the pathophysiology of AN. Starvation was induced by restricting food access in early adolescent or adolescent mice to 40% of their baseline food intake until a 20% weight reduction was reached (acute starvation). To mimic chronic starvation, the reduced weight was maintained for a further 2 weeks. Locomotor activity was analyzed using running wheel sensors. The circadian-rhythm-related activity was measured using the tracking system Goblotrop. Amenorrhea was determined by histological examination of vaginal smears. All cohorts showed an increase in locomotor activity up to 4 h before food presentation (food-anticipatory activity, FAA). While amenorrhea was present in all groups except in early adolescent acutely starved mice, hyperactivity was exclusively found in chronically starved groups. Adolescent chronically starved mice showed a decrease in circadian-rhythm-related activity at night. Chronic starvation most closely mimics AN-related behavioral changes. It appears that the FAA is a direct consequence of starvation. The circadian activity changes might underlie the pathophysiology of AN.

Anorexia nervosa (AN) is associated with hyperactivity, amenorrhea, and brain atrophy. The underlying pathophysiology is mostly unknown, and new targets for therapeutic interventions are needed. This study aimed to systematically establish a murine AN model with the parameter extent of starvation, animal age, and length of starvation for functional studies. The activity-based anorexia (ABA) model combines food restriction with running wheel access. Early adolescent and adolescent mice received 40% of their baseline food intake until a 20% or 25% weight reduction was reached (acute starvation). To mimic chronic starvation, body weight loss was maintained for another two weeks. Running activity was examined using wheel sensors, while amenorrhea was investigated by analysis of vaginal smears. Brain sections were used to analyze cerebral cortex volumes. Acute starvation did not lead to either AN-related symptoms, whereas chronic starvation led to hyperactivity and amenorrhea except in the adolescent cohort with 20% weight reduction. Only ABA mice with 25% weight reduction revealed a cortex volume reduction. The optimal parameters to mirror AN-related symptoms included a 25% weight reduction, early adolescent or adolescent mice, and chronic starvation. The ABA model enables functional analysis of the impact of chronic AN on the underlying hormonal, behavioral, and brain pathophysiology.