Encyclopedia of Animal Cognition and Behavior

Living Edition
| Editors: Jennifer Vonk, Todd Shackelford

Activity-Based Anorexia

  • Ana de Paz
  • Pedro Vidal
  • Ricardo PellónEmail author
Living reference work entry
DOI: https://doi.org/10.1007/978-3-319-47829-6_2088-1



Activity-based anorexia (ABA) is a robust phenomenon that takes place in many animal species in which excessive physical activity is developed to a deleterious level when access to food intake is restricted to a short period of time a day (e.g., 1–2 h), resulting in reduced eating and substantial body weight loss. Paradoxically, the increment in locomotor activity takes place when energetic balance is negative, so that continued exposition of an animal (usually rats or mice) to intermittent food delivery plus almost permanent access to exercise could drive the animal to die if it is not withdrawn from the procedure. ABA constitutes the most recognized animal model of anorexia nervosa disease in humans due to the similarities between the behavior observed when animals are submitted in the laboratory to the above conditions and the characteristics of clinical patients suffering the disorder.

Historical Background

The first scientific observations of the effect of long periods of starvation on different species of birds, reptiles, and mammals in the mid-nineteenth century by Charles Chossat (1843) reflected that, during much of the process, the animals remained to a greater or lesser extent in a state of agitation that lasted until the moments near before their death. Although those studies, which would hardly meet current ethical standards, detailed the physiological processes that took place in the animals, no significant importance was conferred to the high physical activity developed. It was not until the early twentieth century when the work of Richter (1922), sponsored under the direction of J.B. Watson, delved into the biological basis of spontaneous activity in laboratory rats, bringing to experimental study the locomotive behavior developed by the animals in constant conditions of light and temperature and introducing the use of boxes and running wheels to make precise measurements of the activity developed. In those experiments, it was revealed the existence of circadian rhythms in the activity of the subjects, as well as the concentration of this activity around the meal period when it was supplied once a day, among other observations of the factors that influence the level of activity and that are still valid today, such as the sex of the animal or the ambient temperature.

However, the systematization of studies on the failure of maintaining stable weight in rats living in activity wheels under a food deprivation schedule has its milestone in the work of Routtenberg and Kuznesof (1967), who named the phenomenon “self-starvation.” Those authors introduced a control group, with the same food schedule but without access to the running wheel, that ate more and stabilized its loss in weight compared with the activity group who “self-starved” and died.

In the 1980s of the twentieth century, the influential work of Epling, Pierce, and Stefan (1983) highlighted the importance of locomotor activity in times of food shortage in the phylogeny of species as a form of survival, relating the behavior observed in the laboratory animals with that developed in many cases of anorexia nervosa in humans. These instances were called activity-based anorexia, linking since then the study of its causes and the development of treatments for this disorder to a valid animal model given that it reproduces many of the characteristics present in the anorexic patients.

Experimental Protocol

The basic ABA protocol includes the use of two groups of animals, usually named Active and Sedentary, or simply ABA and FR (food restricted). The procedure in the ABA (Active) group starts with the removal of food and allowance of free access to a running wheel for 23 h before the next meal opportunity (see Fig. 1). This day is known as day 0, and the weight data obtained prior to food restriction is the initial weight against which its loss is calculated the remainder days. Animals in the ABA group usually have 1–2 h food exposition and 22–23 h of wheel access each subsequent day until the end of the procedure. In order to prevent death and reduce animal suffering, each subject is removed from the procedure when reaching 20–25% of weight loss for two consecutive days. Animals in FR (Sedentary) group are subjected to the same food restriction schedule but without access to the running wheel. All animals in both groups are weighed each day before the meal period. As seen in Fig. 2, FR animals (dotted line) show initial reduction in body weight and then stabilization around 85% of their free-feeding weights. However, ABA animals (straight line) fail to maintain their weights and progress with a dramatic reduction. Accompanying weight loss, ABA animals show a gradual increase of running along the days in the procedure (Fig. 3, straight line). Comparing both groups, it can be suggested that weight loss relates to the increased activity expressed in ABA animals. It can be added an optional extra control group for activity (unrestricted); here animals are maintained in ad libitum food schedule and have the same time access to a running wheel than ABA animals. In this case, animals do not show the excessiveness in activity observed in the ABA group (Fig. 3, dotted line). These results suggest that the increase in activity relates to the food schedule. See Carrera, Fraga, Pellón, and Gutiérrez (2014) for an extensive description of the protocol.
Fig. 1

Images showing a home cage with attached activity wheel that can be used to run the activity-based anorexia (ABA) procedure in the laboratory with rats

Fig. 2

Expected results in weight loss during ABA procedure

Fig. 3

Expected results in wheel running during ABA procedure

Main Interpretations

Regarding the hypotheses to explain ABA, they can be grouped in two: those that focus on food consumption and those that suggest a central role for activity.

Within the first group, authors (e.g., Kanarek and Collier 1983) suggest that running interferes with the adaptation to the meal schedule preventing ABA animals from maintaining their weights, while the subjects in the control group with same food exposition but with lack of access to a running wheel can maintain their weights. Results have shown that prior exposure to the food schedule causes retardation in the development of ABA, but it does not prevent it (Lett et al. 2001).

Within the group of authors that give a crucial role to activity, there are theories that suggest that wheel running is linked to foraging behavior. Animals increase their activity during shortages of food as a natural selection mechanism to search for alternative sources, activity being possibly induced by the intermittency of food occurrence (Epling et al. 1983). The fact that the intensity of wheel running frequently increases in anticipation to food when it is given periodically (Bolles and Moot 1973) suggests that running is maintained by its consequences as it is explicitly operant running or another schedule-induced behavior. It is well documented that ABA animals distribute their running in two peaks around food occurrence, just before and immediately after meal, according to the control exerted by food availability.

Influencing Factors

There are some factors that are involved in the vulnerability or resistance to ABA. As commented before, prior exposure to the feeding schedule retards the development of ABA. On the contrary, prior exposure to a running wheel facilitates its development. It is known that high ambient temperature prevents the phenomenon. Regarding individual variables, older and heavier rats are more resistant than young and light rats to develop ABA. In the case of sex differences, there are conflicting results, but data seems to point a greater vulnerability to ABA of female rats. Factors that could increase stress, such as social isolation or early maternal separation, contribute to a greater vulnerability; however, environmental isolation (e.g., sound of other animals running) can prevent excessive activity. Early handling experience also reduces the vulnerability to ABA. Some studies have found that ABA can be developed with only 2–4 h of wheel running access per day, mainly those in proximity to next food availability (a phenomenon known as food-anticipatory activity – FAA). Other modulating factors are related to the type of food used (e.g., its dryness or quality).

Neurobiological Basis

The ABA model has not only been applied for the analysis of environmental and individual variables that influence the phenomenon and/or the study of the functional relationships established between starvation and hyperactivity during its development. In recent decades, it has been used to try to unravel neurobiological and physiological aspects involved, as well as possible vulnerabilities. The ABA protocol allows the control of psychosocial variables that are difficult to isolate in studies of this sort carried out with clinical patients. In these experiments there have been used different strains of rats and mice with specific phenotypic characteristics or genetically modified for a gene that encodes both central and peripheral substances, or receptors, that regulate feeding and/or physical activity. These subjects are submitted to the ABA conditions (i.e., restricted food delivery and free access to a running wheel) allowing the identification and understanding of the biological substrates and processes that drive behavioral characteristics exhibited along the procedure under tightly controlled conditions.

Research has pointed out the involvement of the gut-brain axis in reduced food consumption and development of overactivity in ABA. Peripheral orexinergic signal of ghrelin gut hormone, that induces appetite, and the circulating level of leptin, as a satiety signal informing about adipose tissue reserves, are detected in the brain level through the brain stem afferences to hypothalamus, considered a central regulator of feeding behavior. These hormones influence two populations of neurons located in the arcuate nucleus (Arc), the effects of which are antagonistic respect feeding behavior. While the orexinergic group of neurons are activated by ghrelin and promote ingestion, proopiomelanocortin (POMC) system neurons are activated by leptin, secreting anorexinergic substances that inhibit food consumption. Both leptin and ghrelin are implicated in the development of excessive physical activity during starvation, while ghrelin has been specifically associated with anticipatory activity developed when food is supplied under fixed interval schedules. An imbalance of these peptides along the progression of ABA, and their interaction with dopamine projections in the mesolimbic reward pathway, could account for the aberrant rewarding properties of food and physical activity and contribute to the development and maintenance of the process (see Adan and Kaye 2011, for a review). Drugs that exert control over dopamine, key neurotransmitter in reward and motivational systems, and regulatory central neurotransmitter serotonin, mainly antidepressants and antipsychotics, also have been profusely investigated as targets of possible pharmacological treatments.

Other structures and pathways that have been studied as associated with the hallmarks of ABA are cortico-striatal neuro-circuitry, hypothalamic-pituitary-adrenal (HPA) axis, other structures in the limbic system, as hippocampus or amygdala, or the endocannabinoid system (Schalla and Stengel 2019).


Activity-based anorexia is a good animal model of human anorexia nervosa, and the investigation of the behavioral and neurobiological mechanisms involved will help in better understanding the disorder and the development of more efficacious interventions.



  1. Adan, R. A., & Kaye, W. H. (Eds.) (2011). Behavioral neurobiology of eating disorders (Current topics in behavioral neurosciences, Vol. 6). Heidelberg: Springer.  https://doi.org/10.1007/978-3-642-15131-6.Google Scholar
  2. Bolles, R. C., & Moot, S. A. (1973). The rat’s anticipation of two meals a day. Journal of Comparative and Physiological Psychology, 83(3), 510–514.  https://doi.org/10.1037/h0034666.CrossRefPubMedGoogle Scholar
  3. Carrera, O., Fraga, Á., Pellón, R., & Gutiérrez, E. (2014). Rodent model of activity-based anorexia. Current Protocols in Neuroscience, 67, 9.47.1–9.47.11.  https://doi.org/10.1002/0471142301.ns0947s67.CrossRefGoogle Scholar
  4. Chossat, C. (1843). Recherches expérimentales sur l’inanition. Sciences Mathématiques et Psysiques, 8, 438–640.Google Scholar
  5. Epling, W. F., Pierce, W. D., & Stefan, L. (1983). A theory of activity-based anorexia. International Journal of Eating Disorders, 3(1), 27–46.  https://doi.org/10.1002/1098-108X(198323)3:1<27::AID-EAT2260030104>3.0.CO;2-T.CrossRefGoogle Scholar
  6. Kanarek, R. B., & Collier, G. H. (1983). Self-starvation: A problem of overriding the satiety signal? Physiology and Behavior, 30(2), 307–311.  https://doi.org/10.1016/0031-9384(83)90024-0.CrossRefPubMedGoogle Scholar
  7. Lett, B. T., Grant, V. L., Smith, J. F., & Koh, M. T. (2001). Preadaptation to the feeding schedule does not eliminate activity-based anorexia in rats. Quarterly Journal of Experimental Psychology Section B: Comparative and Physiological Psychology, 54(3), 193–199.  https://doi.org/10.1080/02724990042000119.CrossRefGoogle Scholar
  8. Richter, C. P. (1922). A behavioristic study of the activity of the rat. Journal of Comparative Psychology Monographs, 1, 1–55.Google Scholar
  9. Routtenberg, A., & Kuznesof, A. (1967). A. self-starvation of rats living in activity wheels on a restricted feeding schedule. Journal of Comparative and Physiological Psychology, 64(3), 414–421.  https://doi.org/10.1037/h0025205.CrossRefPubMedGoogle Scholar
  10. Schalla, M. A., & Stengel, A. (2019). Activity based anorexia as an animal model for anorexia nervosa – A systematic review. Frontiers in Nutrition, 6, 69.  https://doi.org/10.3389/fnut.2019.00069.CrossRefPubMedPubMedCentralGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.Universidad Nacional de Educación a Distancia (UNED)MadridSpain

Section editors and affiliations

  • Jennifer Vonk
    • 1
  1. 1.Oakland UniversityRochesterUSA