Understanding Anorexia Nervosa Through the Behavioral Learning Lens of Activity-Based Anorexia

        Activity-based anorexia (ABA) is a behavioral phenomenon that was first investigated by Routtenberg and Kuznesof (1967), who found that rats given restricted access to food, typically 1.5 hours per day, while also having continuous access to a running wheel, began to exhibit extreme levels of physical activity alongside a marked decrease in overall food consumption. This ultimately produced a dangerous feedback loop, where the more the rats ran, the less they ate, and the less they ate, the more they ran. Within a week, many rats were running the equivalent of 12 miles per day and consuming almost little to no food, resulting in emaciation, and if the experiment wasn’t stopped, death would have resulted.  More importantly, this effect is only seen when food restriction and physical activity are explicitly combined, in other words, rats undergoing food restriction without a wheel, or a wheel without food restriction, do not exhibit these behaviors and maintain normal health and eating behaviors (Epling & Pierce, 1991). This model, which has been termed activity anorexia, has become a valuable resource and model for understanding anorexia nervosa in human beings, particularly the relationship between dietary restriction and compulsive physical activity. Studying activity anorexia in animals provides researchers with an ethical, as well as very clear picture of how specific environmental conditions can trigger dangerous, and continuous self-reinforcing behaviors. 

        The behavioral patterns observed in activity anorexia closely mirror many features of anorexia nervosa in humans. Like the rats in the original experiments, people with anorexia nervosa often begin with voluntary food restriction, usually aimed at weight loss, which is then accompanied by a significant increase in physical activity. This can involve structured workouts or persistent restlessness, both of which suppress appetite and reinforce continued restriction. Epling and Pierce (1991) describe this as a self-reinforcing cycle where decreased eating fuels increased activity, which in turn suppresses hunger further. More recently, Spadini et al. (2021) highlighted that the ABA model captures this feedback loop in a highly consistent as well as very measurable way. Their review emphasized that ABA is not just a behavioral phenomenon, instead, it should be considered a valid scientific tool that shows how limited food access, when combined with the opportunity for exercise, can create compulsive activity patterns that override basic survival needs. They also noted that this shift in behavior is linked to changes in the brain’s reward system, where physical activity becomes more reinforcing than eating, ultimately a behavioral mechanism that closely aligns with what many people with anorexia nervosa report experiencing.

        Building on this, Foldi (2024) argues that the compulsive nature of running in the ABA model stems from impaired decision-making, describing it as “paradoxical hyperactivity.” The overarching work supports the idea that excessive exercise may be a driving force in anorexia, not just a byproduct. She also points out that not all animals develop the ABA pattern, which opens the door for studying and better understanding vulnerability, resilience, as well as other key factors in human anorexia. Certain high-risk groups, like ballet dancers and wrestlers, who regularly combine dietary restriction with intense physical training, are known to experience elevated rates of anorexia symptoms. While laboratory rats in ABA experiments are exposed to externally and explicitly imposed food limits, humans, on the other hand, often face internal psychological restrictions driven by body image ideals, social pressure, and a fear of weight gain. Foldi’s work reinforces the idea that the ABA model offers more than just behavioral resemblances, it can help uncover and lay the groundwork for how environmental triggers, reward-based learning, and brain function interact to ultimately sustain anorexic behavior over time.

References

Spadini, S., Ferro, M., Lamanna, J., Malgaroli, A. (2021). Activity-based anorexia animal model: a review of the main neurobiological findings. Journal of Eating Disorders, (9)123. https://doi.org/10.1186/s40337-021-00481-x 

Foldi, J. C. (2024). Taking better advantage of the activity-based anorexia model. Trends in Molecular Medicine, 12(4), 330-338. https://doi.org/10.1016/j.molmed.2023.11.011