Skip to Content

Office of the Vice President for Research

Banner Image

Response Interference and Activity-Based Anorexia in Rats

Nicole Streeb



Anorexia nervosa is a severe eating disorder characterized by extreme emaciation due to the inability to maintain a healthy weight. The disorder can result in many behavioral and physiological changes. Activity-based anorexia (ABA) in animals is a potentially viable model of anorexia nervosa, reproducing many of the core behavioral components of the disorder such as restriction of food consumption, dramatic weight loss, and increased drive for physical activity (i.e. wheel running). Some of the behavioral and physiological factors involved in this phenomenon have been examined by previous research; however, investigators have yet to determine the mechanisms by which this behavior is maintained and reinforced.  The present study examines the effects of response prevention and response interference on the maintenance of ABA. Data analysis revealed that response prevention somewhat reduced subsequent behavior, while interference resulted in significant increases in later activity. These findings support the notion that learning and motivational variables affect behavioral processes involved in ABA.



According to the Diagnostic and Statistical Manual of Mental Disorders (DSM-TR-IV), anorexia nervosa has several characteristics that distinguish the disorder.  These symptoms consist of an inability to stay at a weight which is above the minimally normal weight for the individual’s age and height, and an intense fear of gaining weight or becoming fat. Despite being severely underweight, anorexics have inaccurate perceptions of true body shape or weight (American Psychiatric Association, 2000).  Due to the prevalence and often severity of the disorder, variables involved in anorexia nervosa merit further analysis to strengthen the scientific community’s understanding of the development and maintenance of eating disorders.

Researchers have known for some time that many patients with anorexia nervosa also have dysfunctional exercise routines (Cook, Hausenblas, Tuccitto, & Giacobbi, 2011) and individuals with anorexia have been shown to increase activity during periods in which they are underweight (Davis, Kennedy, Ravelski, & Dionne, 1994). Excessive exercise is present in eating disorders in the range of 30-70% (Mond & Calogero, 2009) and 10-40% of patients with anorexia meet the diagnostic criteria for obsessive-compulsive disorder (Milos, Spindler, Ruggiero, Klaghofer, & Schnyder, 2002; Sallet et al., 2010). Eating disorders and obsessive-compulsive disorder share many characteristics, namely psychopathologic similarities (e.g. obsessions and compulsions related to food), but also personality traits, neuropsychological impairments, epidemiology, and therapeutic responses (Sallet, et al., 2010).

A potential animal model of anorexia has been demonstrated by a phenomenon called activity-based anorexia (ABA; Hall & Hanford, 1954; Routtenberg & Kuznesof, 1967). Albino rats given access to a running wheel and put on a restricted food diet significantly increase their running over days and decrease food consumption. Weight eventually decreases dramatically, ending in death if not stopped (Hall & Hanford, 1954).  Although it is not clear what specific variables cause ABA, research has identified some potential components. For example, Brown, Avena, & Hoebel (2008) found that a high-fat diet was able to reverse or prevent the onset of ABA. Investigations of physiological and neuroendocrinological factors have discovered a relation between hypoleptinaemia and high activity levels (e.g. van Elburg, Kas, Hillebrand, Eijkemans, & van Engeland, 2007), and Hillebrand, Koeners, de Rijke, Kas, and Adan (2005) concluded that a Leptin treatment was successful in decreasing running wheel activity, indicating a role for that protein.

Behavioral conditioning processes may also play a role in ABA. Evidence has been provided which suggest pre-exposure to the wheel can result in an augmentation of ABA and food exposure resulted in the amelioration of ABA behaviors (Ratnovsky & Neuman, 2011). When rats were placed in a distinct chamber before being placed into an activity wheel, a conditioned place aversion to the chamber developed (Masaki & Nakajima, 2008), whereas when rats received placement in a distinct context after wheel running, the context cues developed positively reinforcing properties (e.g. Lett, Grant, Byrne, & Koh, 2000).


Purpose of this study

Inasmuch as response prevention has been viewed as a viable way to treat obsessive-compulsive disorders (McAllister & McAllister, 1995), and given the previous research suggesting similarities in anorexia and obsessive-compulsive behavior, the present study attempted to replicate the ABA effect and investigate the impact of response prevention. It was hypothesized that response prevention may be effective in reducing the magnitude of the ABA effect. A second variable investigated was response effort. Lydall, Gilmour, & Dwyer (2010) observed that rats valued a reward more when it followed a high effort response than when the same reward followed a low effort response. In the present experiment effort was manipulated by altering the drag resistance of the wheel. It was hypothesized that this response interference/ effort manipulation would decrease operant wheel running behavior and attenuate the ABA effect.




All procedures were approved by the USC Aiken Institutional Animal Care and Use Committee.  The subjects were 36 male and female Sprague-Dawley rats (n=9), approximately 120 days old, from the USC Aiken Psychology Department Animal Vivarium. Subjects were counterbalanced by sex and randomly assigned to one of four experimental conditions, differing in the availability of wheel responding during one phase of the experiment: Full Response (FR), Response Interference (RI – wheel resistance increased), Response Prevention (RP – Wheel accessible but locked), and Response Prevention Wheel Blocked (RPWB – No access to wheel).


The primary apparatus consisted of four Med-Associates wheel boxes (Fig. 1), each having an attached running wheel. Access to the wheel could be allowed or blocked by inserting a manual door. The home cage measured 19.0″ L x 10.5″ W x 8.0″ D. The attached wheel is 14.0″ in diameter with standard 0.1875″ stainless steel grid rods with a standard 0.625″ spacing. The wheel cage was housed in a light and sound attenuating chamber with a houselight that provided illumination on a 12 hour light:dark cycle. Wheel running was recorded in 30-minute intervals for 14 consecutive days with a PC and Med-Associates software.  Animals were kept individually in the apparatus during the experiment and had continuous water access through a bottle inserted into the cage.


The general procedure followed that used by other research to produce the ABA effect (e.g., Routtenberg & Kuznesof, 1967):  A habituation phase (Days 1-4) was followed by food restriction and wheel responding (Days 5-14).  During habituation, all subjects received 24-hour access to food and water in the apparatus.  Access to the running wheels was allowed for all groups and wheel resistance was set on the lowest value (12 grams). Beginning on Day 5, and continuing through the remainder of the experiment, all subjects were placed on a restricted feeding schedule. Food was administered for a 90 minute period daily, beginning one hour after the start of the light cycle. Wheel running was prevented during the feeding period by locking the wheel. Wheel running was available during the remaining 22.5 hour period.

The major experimental manipulation occurred on Days 8-11. Two groups (RP, RPWB) received a response prevention manipulation in which the wheel was locked during these four days and no running could occur. For Group RP, the subjects could enter the wheel but not turn it, and for Group RPWB access to the wheel was completely blocked. A third group (RI) was allowed to run in the wheel, but the resistance was increased to 40 grams, thus requiring more effort to turn the wheel. For the final group, FR, full responding at normal resistance (12 grams) was permitted.  For the last three days of the experiment (12-14), all groups had equal access to the wheel (as in Days 5-7) with normal resistance (12 grams).

The total number of complete wheel rotations each day and daily weight were analyzed using parametric statistics. An alpha level of p<.05 was used for all analysis.



It was anticipated that overall responding would increase while weight decreased, and that both response prevention and response interference would decrease subsequent running behavior.  The performance of Group FR replicated the ABA effect, in that daily wheel responses significantly increased from Day 4 to Day 14 (Ms=3558, 7908), t(9)=2.20, p=.05. Across these same days, average body weight for the group significantly decreased (Ms=315.8g, 267.2g), t(9)=9.22, p<.01.

For all remaining analysis, difference (change) scores were calculated by subtracting the total wheel responses on a baseline day (the last day of habituation, Day 4) from the total responses on each subsequent day (5-14), with higher values reflecting increased wheel running. Figure 2 presents the mean change in wheel responses at the end of the response interference phase (Day 11), comparing the interference group (RI) to the normal resistance group (FR). Although Group RI responded at a higher rate than Group FR, this difference was not statistically significant, t(16)=1.28, p >.05.

Figure 3 presents the mean change in responding on the three recovery days (Days 12-14) after the response prevention or interference treatment. Response prevention with the wheel blocked (RPWB) appeared to be somewhat effective in reducing responding compared to Group FR, with group RP responding between these two groups. The increased responding observed in Group RI during the interference sessions not only continued, but increased further on all three days.

These impressions were generally confirmed with statistical analyses.  Multivariate analysis of variance (MANOVA) tests were conducted on the daily response difference scores, with Groups as the between-subjects variable and Days as the within-subjects factor.  An overall MANOVA analyzing responses before and after the response prevention/interference manipulation revealed a significant Days effect, F(4,29) = 4.36, p=.007,  a marginal Groups effect, F(3,32)=2.40, p=.086, and a significant Days x Groups interaction, F(12,83)=2.16, p=.021.  The significant interaction effect reflects the diverging patterns of responding seen in Figure 3, with Group RI increasing at a higher rate than the others and the response prevention groups starting to decrease their responding by Day 14.  Fisher LSD multiple comparison tests at the end of training indicated that on Day 13 Group RI was responding at a higher rate than Groups RP and RPWB (p’s =.056 and .037).  On the final day of training Group RI was significantly higher than RPWB (p =.046), and marginally higher than Groups FR and RP (p’s = .068 and .078).   Analysis of weight loss over the 14 days indicated no significant differences among groups, with all groups losing between 14-17%.



The present findings add to an understanding of the variables that affect behavioral processes involved in ABA. The present experiment successfully replicated the ABA phenomenon (Hall & Hanford, 1954; Routtenberg & Kuznesof, 1967), demonstrated by a significant increase in running and decrease in weight in Group FR. The response prevention manipulation was not as effective as anticipated, given that the two prevention groups did not differ significantly from Group FR. There was a trend for the group that had the wheel blocked (RPWB) to decrease responding compared to the other groups by Day 14, but the difference was not significant. It is possible that this effect would have been greater had additional days of running been given.

One of the most interesting and unanticipated outcomes of the present study was the effect of the effort (interference) manipulation. It was expected that increasing the effort to run in the wheel by increasing wheel resistance would decrease responses. However, the opposite effect occurred. Group RI significantly increased running and persisted throughout the experiment. A follow-up experiment in our lab has replicated this effect and we are currently investigating possible explanations of it.

Throughout this experiment, the researchers noticed a consistent (non-quantified or analyzed) behavior: Most rats brought food into the wheel during the feeding period. They would also bring pine bedding into the wheel, and some carried all the bedding from the chamber into the wheel. This phenomenon suggests the wheel context may have developed a positive association as a result of the rewarding effect of wheel running (e.g. Lett et al., 2000). We have plans to further explore these observations regarding food and bedding in the wheel area, perhaps with video recording. 




About the Author


Nicole StreebNicole Streeb

I am a senior Psychology and Business Management major at the University of South Carolina Aiken. I was awarded the Magellan Scholar AWARD for Spring 2012 to continue research IN the USC Aiken psychology lab on Response Interference and Activity-Based Anorexia in Rats. I have been the sole researcher in this project, supported by Dr. Edward Callen. I was responsible for all experimental procedures and running all subjects in this project and Dr. Callen and I worked together on data analysis. This study has given me the experience of working with a respected faculty member on a research experiment and has given me the opportunity to be a more competitive and confident applicant for graduate schools. I am currently applying to Ph.D. Programs in Clinical Psychology and this research experience conducting the experiment is an important part of my resume and having Response Interference and Activity-Based Anorexia in Rats accepted for publication in Caravel will significantly enhance that resume. In graduate school, I will have to work on original research and this experience was great training and preparation for that. A portion of this research was presented as part of a poster session at the annual meeting of the Southeastern psychological Association (SEPA) in New Orleans last February, as well as USC Aiken Research Day and USC Columbia Discovery Day. I have discussed my research with the Neuroscience Education and Research-Valued Experience (C-NERVE) group at USC Aiken, made up of students and faculty. I have also submitted a proposal for acceptance at SEPA 2013 conference based on some follow-up research. My academic and research activities contributed to me being the recipient of the USC Aiken 2012 Outstanding Psychology Student Award and the USC Aiken Psychology Undergraduate Research Award.



American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed,., text rev.). Washington, D.C.: Author.

Brown, A. J., Avena, N. M., & Hoebel, B. G. (2008). A high-fat diet prevents and reverses the development of activity-based anorexia in rats. International Journal Of Eating Disorders, 41(5), 383-389.

Cook, B., Hausenblas, H., Tuccitto, D., & Giacobbi, P. r. (2011). Eating disorders and exercise: A structural equation modelling analysis of a conceptual model. European Eating Disorders Review, 19(3), 216-225.

Davis, C., Kennedy, S.H., Ravelski, E., & Dionne, M. (1994). The role of physical activity in the development and maintenance of eating disorders. Psychological Medicine, 24(04), 957-967.

van Elburg, A. A., Kas, M. H., Hillebrand, J. G., Eijkemans, R. C., & van Engeland, H. H. (2007). The impact of hyperactivity and leptin on recovery from anorexia nervosa. Journal Of Neural Transmission, 114(9), 1233-1237.

Hall, J. F., & Hanford, P. V. (1954). Activity as a function of a restricted feeding schedule. Journal Of Comparative And Physiological Psychology, 47(5), 362-363.

Hillebrand, J. G., Koeners, M. P., de Rijke, C. E., Kas, M. H., & Adan, R. H. (2005). Leptin Treatment in Activity-Based Anorexia. Biological Psychiatry, 58(2), 165-171.

Lett, B. T., Grant, V. L., Byrne, M. J., & Koh, M. T. (2000). Pairings of distinctive chamber with the aftereffects of wheel running produce conditioned place preference. Appetite, 34. 87-94.

Lydall, E. S., Gilmour, G., & Dwyer, D. M. (2010). Rats place greater value on rewards produced by high effort: An animal analogue of the “effort justification” effect. Journal Of Experimental Social Psychology, 46(6), 1134-1137.

Masaki, T., & Nakajima, S. (2008). Forward conditioning with wheel running causes place aversion in rats. Behavioral Processes, 79(1), 43-47.

McAllister, W. R., & McAllister, D. E. (1995). Two-factor fear theory: Implications for understanding anxiety-based clinical phenomena. In W. O’Donohue & L. Krasner (Eds.), Theories of behavior therapy (pp. 145-172). Washington, DC: American Psychological Association.

Milos, G., Spindler, A., Ruggiero, G., Klaghofer, R., & Schnyder, U. (2002). Comorbity of obsessive-compulsive disorders and duration of eating disorders. International Journal of Eating Disorders, 31, 284-289.

Mond, J. M., & Calogero, R. M. (2009). Excessive exercise in eating disorder patients and in healthy women. Australian And New Zealand Journal Of Psychiatry, 43(3), 227-234.

Ratnovsky, Y., & Neuman, P. (2011). The effect of pre-exposure ad recovery type on activity-based anorexia in rats. Appetite, 56(3), 567-576.

Routtenberg, A., & Kuznesof, A. W. (1967). Self-starvation of rats living in activity wheels on a restricted feeding schedule. Journal Of Comparative And Physiological Psychology, 64(3), 414-421.

Sallet, P. C., de Alvarenga, P., Ferrão, Y., de Mathis, M., Torres, A. R., Marques, A., et al. (2010). Eating disorders in patients with obsessive-compulsive disorder: Prevalence and clinical correlates. International Journal Of Eating Disorders, 43(4), 315-325.