Printer Friendly

Use of a time-restricted access to food diet to achieve and maintain weight reduction in mice.

Recent years have seen the development of mouse models of a wide range of psychological disorders and afflictions to the central nervous system, including Alzheimer's disease, anorexia nervosa, abnormal anxiety, attention deficit hyperactivity disorder, autism, major depression, Fragile X Syndrome, Huntington's disease, obsessive-compulsive disorder, Parkinson's disease, schizophrenia, and Tourette's syndrome. Mouse models are used chiefly to study neurophysiological processes, but behavior is sometimes studied as well. One reason to study the behavior is to help establish the validity of the model: Do the mice manifest the expected behavioral abnormalities? Furthermore, the study of behavior may provide new insights into underlying neurophysiological processes. Behavior also affords a means to measure the efficacy of treatments designed to prevent, mitigate or reverse neurophysiological abnormalities.

Behavior analysts have taken an increasing interest in mice as research subjects to judge from recent publications in The Journal of the Experimental Analysis of Behavior (e.g., Belke & Garland, 2007; Leslie, Shaw, Gregg, McCormick, Reynolds, & Dawson, 2005; McKerchar, Zarcone, & Fowler, 2005; Papachristos & Gallistel, 2006; Zarcone, Chen, & Fowler, 2007). However, behavior analytic animal research traditionally has been conducted with pigeons and rats, and unfamiliarity with mice may deter some from pursuing research with mouse models. To help remedy this concern, we examined whether time-restricted feeding procedures would provide a simple means of allowing researchers to achieve and maintain an appropriate level of food deprivation in mice.

There are two basic strategies used to establish food deprivation in operant research. One strategy is to restrict the amount of food subjects are allowed to consume and the other is to restrict the length of time subjects are allowed to eat (as shorthand, these are referred to below as "restricted-amount" and "restricted-time" diets, respectively). Under the former approach, subjects are weighed daily and the size of the food ration is adjusted to maintain weights at a given level (and by extension, ensure a constant level of food deprivation). Hurwitz and Davis (1983) observed that this procedure quickly predominated behavior-analytic animal research, although it is not without disadvantage. The restricted-amount procedure requires that subjects be housed individually to ensure that all subjects receive the appropriate of food. This, in turn, limits research on social processes. The restricted-amount procedure is also impractical to implement because the weight of both the subject and the food ration must be carefully monitored. The difficulty of calculating an appropriate amount of food is compounded if the amount of food subjects earn during data collection is inconstant or if data collection is not performed daily.

In view of the shortcomings of the restricted-amount procedure, Hurwitz and Davis recommended that researchers use instead the restricted-time procedure. Hurwitz and Davis demonstrated the relative advantages of this approach using six, male, Long-Evans rats. Subjects had food access for 1 or 2 hours each day (1 hr when the food was placed inside the subjects' cages, 2 hrs when the food was available through an external wire-mesh hopper). They found that when the procedure was introduced, subjects lost approximately 20% of their free-feeding weights, achieving in a few days the 80% level of weight reduction used in many animal studies. Continued use of the procedure, however, did not result in further weight loss, and by the end of 20 days of observations, mean weights had stabilized to 78.5% of free-feeding levels. Hurwtiz and Davis found that the procedure was effective even when subjects could obtain additional food during experimental sessions with a variable-interval schedule of reinforcement. Evidently, food obtained during data collection depressed food intake during the feeding period with the result that net food intake (and weights) remained constant. In other words, the restricted-time procedure appears to obviate the need to calculate the size of the food ration.

A disadvantage of a restricted-time diet is that the researcher sacrifices some control over the degree of deprivation that occurs, and subjects may differ somewhat over time and from each other in the exact amount of weight loss that occurs. In the Hurwitz and Davis study, the six rats ranged from between 75% to 84% of ad lib weights. However, the loss of control over the exact deprivation level is at least partly compensated for by greater control over other aspects of food deprivation. Under a restricted-time diet (but not under a restricted-amount diet), the researcher has control over how much time elapses between the daily feeding and the start of the subsequent data collection session.

Perhaps the main problem in using a restricted-time feeding procedure is that the ideal feeding period duration is often unknown. If the duration is too short, subjects will steadily lose weight and fall into poor health; if the duration is too long, weights will instead increase to the detriment of reinforcer efficacy. The goal of the present investigation was to determine whether a time-restricted feeding procedure could be used to achieve and maintain an appropriate level of food deprivation in mice and to establish the approximate food-access period that is most effective. Data were collected with C57 BL/6 mice, a mouse strain used widely in biomedical research.

Experiment 1

We initially repeated the Hurwitz and Davis (1983) restricted-time feeding procedure with seven C57 BL/6 mice that were 12 weeks of age at the beginning of the investigation. In the Hurtwitz and Davis study, a daily feeding period of 1 hr caused weights to stabilize at approximately 80% of ad lib levels. Others have also used this procedure, apparently with success (e.g., Molina-Hernandez, Tellez-Alcantara, Garcia, Lopez, & Jaramillo, 2003). However, some researchers have used instead somewhat longer feeding periods (e.g., Williams, 1999). In the present experiment, subjects were subjected to daily feeding periods of both 1 and 2 hrs. Observations with each diet were planned for 20 days; observations were suspended earlier if mouse weights fell below 75% of ad lib levels.


Subjects were individually housed and fed ad lib (rodent chow was kept in an external food bin built that was part of the cage lid). Throughout the study, the mice were weighed at the same each day. Subjects were weighed for 8 consecutive days before the diet was imposed (i.e., the baseline). The average weight from the last 5 days was the value against which subsequent comparisons were made. After the final baseline weighing, the food was removed.

During the restricted-time diet, two 4-g pieces of food were placed on the floor of the cage immediately after the daily weighing. This was an approximately twice the amounts subjects ate daily while fed ad lib. After 1 hr, the food pellets were removed. Observations were continued until the weight of any individual mouse fell below 75% of ad lib levels (i.e., below the lower range obtained by Hurwitz and Davis). Otherwise, observations continued for 20 days. Once observations with a daily 1-hr feeding period were complete, subjects were returned to an ad lib diet, and once weights returned to former levels, the procedure was repeated (including an 8-day baseline) but with a 2-hr feeding period in effect.


The upper panel of Figure 1 shows weight changes of each of the subjects during the final five baseline days (Days B4 through B8) and the subsequent days with the diet in effect (D1+). Weights on each day are expressed as a percent of the mean free-feeding weight. The graph shows that weights rapidly decreased when daily food access was limited to 1 hr. By the third day, the weights of two of the subjects fell below 75% of baseline weights while the weights of others were trending in the same direction. Under a 2-hr feeding period (Figure 1, lower panel), weights decreased more slowly, but by Day 6 one of the subjects fell below the 75% of baseline weights. At the time, other subjects were also trending towards the suspension criterion.



The failure to obtain stable weights at the 80% level with 1-hr daily feeding periods indicates that the restricted-time feeding procedure Hurwitz and Davis found to be effective for rats is ineffective for mice. Some researchers have used longer daily feeding periods with rats, but in this case doubling subjects' access to food still did not produce stable weights at the 80% level. Neither outcome shows that a restricted-time diet is ineffective for mice. It may be that a somewhat longer access to food would produce the desired outcomes. However, an important advantage of the Hurwitz and Davis procedure is that, at least with rats, it is relatively practical to use. If relatively long feeding durations are required to maintain stable weights in mice then at least some of that advantage is lost.

Experiment 2

Following the unsuccessful attempt to replicate the study by Hurwitz and Davis (1983) with mice, a variant of their procedure was used in which weights were first gradually reduced to 80% of free-feeding weights before a restricted-time diet was introduced. During the first phase of the experiment, subjects were assigned to one of three food-access periods to determine what approximate duration would best maintain weights. Then, during a second phase, the procedure was repeated with subjects receiving what the results suggested was the ideal feeding period duration.


Subjects were 12 male C57 BL/6 mice that were 20 months of age at the beginning of the investigation. The subjects had previously served in studies of operant behavior (some of the data are reported in Derenne, Arsenault, Austin, & Weatherly, 2007) and they had a prior history of 80% weight deprivation. The subjects were free-feed for the 2 months preceding the beginning of observations. As was the case in Experiment 1, weights were recorded for 8 consecutive days prior to the beginning of weight reduction and the mean weight from the final 5 baseline days was used as a comparison against subsequent weight loss. Then, over the course of 7 days, weights were gradually reduced using a restricted-amount diet. During this phase, subjects were fed between 1 and 3 g each day, depending on the rate of weight loss on preceding days.

By the end of the 7 days, all subjects were at the designated 80% deprivation level. The subjects then were divided into three groups of four and each group was assigned one of the following food-access durations: 1 hr (the feeding period used by Hurwitz and Davis), 2 hrs, or 4 hrs. In Experiment 1, observations were suspended once the weight of one subject fell below 75% of free-feeding levels. In the present case, however, only individual subjects were returned to an ad lib diet if their weights fell below the 75% criterion. After 8 days, the remaining subjects were returned to an ad lib diet and weights were allowed to recover to previous levels.

The second phase of the procedure began 1 month later with 10 of the mice. The procedure for calculating baseline weights and achieving 80% of baseline weights was the same as that for the first phase. Target weights during the second phase were usually equal to those during the first phase, and were within 1 g in all cases. In light of the results from the first phase (described below), the daily food access period was 2.5 hrs. This restricted-time diet remained in effect for 20 days, although individual subjects were removed earlier if their weights fell below the 75% level.


Figure 2 shows the results of the first phase of Experiment 2 when subjects were divided into three restricted-time groups. The graph includes the last 5 days of baseline observations (B4-B8) and the subsequent days in which the diet was in effect (D1-D8). Weight changes during the 1-week transition from a free-feeding diet to a restricted-time diet are not shown. The figure indicates that none of the groups maintained exactly 80% of ad lib weights under the selected feeding times. Subjects in the 1-hour feeding group (upper panel) rapidly lost weight under the restricted feeding schedule, recapitulating the results of Experiment 1. Weights of all subjects in this group fell below the 75% criterion by the fourth day of observations. Two of the subjects in the 2-hr feeding group (middle panel) also had to be removed from the study because of excessive weight loss; the remaining two, however, were close to the target 80% level (on Day 8 they were at 77% and 82% of ad lib levels). Under the 4-hour feeding group (lower panel), no subjects were removed from the study for excessive weight loss, and on Day 8 all subjects were above the target level (weights ranged from 83% to 90% of ad lib levels).

Figure 2 indicates that a feeding period between 2 hrs and 3 hrs would be most effective in maintaining weights at the 80% level. Therefore, during the second phase of the experiment, the mice were given food access for 2.5 hrs each day.


Figure 3 shows the results of the second phase of Experiment 2, including weight levels during the last 5 days of baseline observations and the 20 days with a 2.5-hr restricted-time diet. The figure shows that this second restricted-feeding procedure was generally effective in maintaining the appropriate level of food deprivation. Mean weights per day ranged from 78% to 81% of ad lib levels under the restricted-time diet, and weights were not trending upwards or downwards at the time the experiment concluded. However, two of the 10 subjects had to removed from the study early because their weights fell below 75% of ad lib levels.



A 2.5-hr daily feeding period was effective in maintaining weights at the 80% level for most subjects (at the end of the study, mean weights were at 80.13% of ad lib weights). However, two of the 10 subjects exposed to this procedure were found to lose an unacceptable amount of weights (at one point they reached a deprivation level of 74% or less, which resulted in removal from the study). This result might indicate that a slightly longer feeding-period duration would better maintain an 80% deprivation level for all subjects (perhaps the optimal duration would be 2.75 or 3 hrs). Under a 2.5-hr feeding period duration, two subjects were at deprivation levels of 83% and 84% at the end of observations, and a longer feeding period duration would likely have resulted in greater deviance from the target level. It may be the case that no single feeding period duration maintains weights at approximately the 80% levels for all subjects.

Experiment 1 and the first phase of Experiment 2 both used 1 and 2-hr daily feeding periods. A noteworthy outcome is that in both cases, subjects reached the criterion for removal from the study in a similar timeframe. This finding suggests that the age of the subjects, the past experience with food deprivation, and the manner in which weight loss is established, all of which varied between the two experiments, may be less important in determining the effects of a restricted-time diet than the duration of the daily feeding period.


The present study suggests that restricted-time diets can be effective with mice. Under this method, the researcher does not need to weigh each day the subjects' food rations, nor is the researcher prohibited from using social housing of subjects. However, use of the Hurwitz and Davis (1983) restricted-time feeding procedure for achieving and maintaining weight reduction with rats was not found to be effective with C57 BL/6 mice. We examined variants to the Hurwitz and Davis procedure to determine whether any restricted-time diet would have the desired effects. A procedure that generally was effective involved first gradually reducing weights to target levels (80% of ad lib weights in this case) and then providing access to food for 2.5 hrs each day.

A shortcoming of the 2.5-hr restricted time diet was that not all subject weights stabilized at close to the 80% level under this procedure. Variability in weight loss was also a feature of the findings reported by Hurwitz and Davis and may be common to restricted-time diets. This outcome indicates that the researcher must carefully attend to subject weights, and be prepared to take steps to ensure that some subjects do not deviate too greatly from target weights. Two strategies that may be effective in this regard are to use different feeding period durations for different subjects, and to occasionally provide supplemental food to subjects whose weights fall below their targets.

Author's Note

Correspondence should be sent to first author at the Department of Psychology, University of North Dakota, P.O. Box 8380, Grand Forks, ND 58202-8380 (E-mail:


Belke, T. W. & Garland, T., Jr. (2007). A brief opportunity to run does not function as a reinforcer for mice selected for high daily wheel-running rates. Journal of the Experimental Analysis of Behavior, 88, 199-213.

Derenne, A., Arsenault, M. L., Austin, D. P., & Weatherly, J. N. (2007). Weaver mutant mice exhibit long-term learning deficits under several measures of instrumental behavior. Physiology & Behavior, 92, 1002-1009.

Ferster, C. B. (1953). The use of the free operant in the analysis of behavior. Psychological Bulletin, 50, 263-274.

Hurwitz, H. M. B. & Davis, H. (1983). Depriving rats of food: A reappraisal of two techniques. Journal of the Experimental Analysis of Behavior, 40, 211-213.

Leslie, J. C. Shaw, D. Gregg, G. McCormick, N. Reynolds, D. S. & Dawson, G. R. (2005). Effects of reinforcement schedule on facilitation of operant extinction by chlordiazepoxide. Journal of the Experimental Analysis of Behavior, 84, 327-338.

McKerchar, T. L. Zarcone, T. J. & Fowler, S. C. (2005). Differential acquisition of lever pressing in inbred and outbred mice: Comparison of one-lever and two-lever procedures and correlation with differences in locomotor activity. Journal of the Experimental Analysis of Behavior, 84, 339- 356.

Molina-Hernandez, M. Tellez-Alcantara, N.P. Garcia, J.P. Lopez, J.I.O. & Jaramillo, M.T. (2003). Anti-conflict-like actions of intralateral septal infusions of allopregnanolone in Wistar rats. Pharmacology, Biochemistry, and Behavior, 75, 397-404.

Papachristos, E. B. & Gallistel, C. R. (2006) Auto-shaped head poking in the mouse: A quantitative analysis of the learning curve. Journal of the Experimental Analysis of Behavior, 85, 293-308.

Williams, B. A. (1999). Value transmission in discrimination learning involving stimulus chains. Journal of the Experimental Analysis of Behavior, 72, 177-185.

Zarcone, T. J., Chen, R., & Fowler, S. C. (2007). Effects of differing response-force requirements in CD-1 mice. Journal of the Experimental Analysis of Behavior, 88, 381-393.


COPYRIGHT 2008 Behavior Analyst Online
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Derenne, Adam; Cicha, Rebecca J.; Flannery, Kathryn A.; Manley, Rachel M.
Publication:The Behavior Analyst Today
Date:Mar 22, 2008
Previous Article:Washoe--Child of the CWU community dies.
Next Article:Defining frequency: a natural scientific term.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters