The Effects of Caffeine on the Wheel Activity of House Mice (Mus domesticus)


Cynthia Sonntag and Emily Zawikowski




            In this experiment we decided to test the effects of caffeine on the wheel running activity of house mice, Mus domesticus, also know as Mus musculus.  For the purpose of this paper we will be refer to house mice as Mus domesticus.  Each of the mice was tested on a regular water diet and a caffeinated water diet for a period of one week each.  We found that caffeine did not have a significant difference on the wheel running activity of house mice (p-value 0.37).


Keywords: caffeine, Mus domesticus, Mus musculus, wheel, activity, mice, wheel activity, automated wheel counter





            As college students, we consume caffeinated beverages on a regular basis. As a drug, caffeine can affect the central nervous system, frequently increasing heart rate, breathing, blood pressure, and can cause sleep irregularities (Monroe, 1998). Many people start their day off with a cup of coffee so the caffeine pills used in this experiment were 200 mg each, a dosage which is approximately equal to about one cup of coffee. To properly administer the correct amount of caffeine, it was necessary to know the average body mass of the mice used. Naturally, mice vary between 10 and 20 grams, in comparison to the heavier weight of lab mice, some having a mass up to 50 grams (Austad, 2002).

            Since house mice are nocturnal animals, we inferred that activity would be less during the day than at night. Based on this we attempted to run the mice once a day, as late in the day as possible, instead of several times a day. We chose to run our experiment using wheels with automated counting systems because Eikelboom (2001), and Koteja and Garland (2001) (as cited by Girard et. al 2001), made reference to the use of wheels with automated counting systems in wheel activity experiments. Being that small mammals often burrow in order to breed, forage and escape predation, we made sure to supply ample bedding in order to accommodate this natural behavior (Schmid-Holmes et. al., 2001). 

Knowing that house mice are sometimes used in medical research with results that can be applied to humans, we wanted to see if caffeine would have the similar effect on house mice as it has on humans. This experiment was designed to quantify the activity of house mice on a caffeinated diet versus the activity of house mice on a non-caffeinated diet. Our hypothesis was that caffeine increases the wheel running activity of house mice.


Materials and Methods






            To begin this experiment we wrote a research proposal and had it approved by Assistant Professor of Biology at Alverno College, Rebecca Burton, Ph.D., who found our proposal both practical and ethical.  We purchased 6 white house mice (Mus domesticus) from Pet World in Greenfield, Wisconsin.  Three males and three females were randomly selected from a communal tank.  For the duration of the experiment each mouse was kept in its own cage in the animal laboratory at Alverno College.  The mice were caged individually so readings of how much food and water each mouse was consuming could be taken on a daily basis.  Each cage contained enough pine bedding to completely cover the bottom of the cage.  A food dish was placed in the back left hand corner of each cage, directly opposite from where the water bottle would hang (Figure 1).






Overhead view of cage setup.


            Before placing the mice in their cages they were contained in a small mammal holding device so they could be massed on an electronic scale.  While in the holding device, three of the mice were marked with a green Sharpie permanent marker.  The green markings on these three mice were used to identify the mice that would receive the caffeine treatment during week one.  Unmarked mice received regular water during week one.  Each group of three mice consisted of at least one mouse of each sex.  After massing the mice, they were placed in the cages with tags naming them as Mouse A through Mouse F.  Each mouse was provided with 25 g of food and 100 mL of regular water and allowed to habituate to their new environment for a 24-hour period.

            After the habituation period we again massed each mouse using a tared 1000 mL beaker on an electronic scale.  We found that the average body mass for the mice to be 25 g.  Using a 100 mL graduated cylinder we then measured how much water each mouse consumed.  We found that the average amount consumed was approximately 15 mL.  Next we made a stock of caffeinated water that would be used for the experiment for administering the caffeine to the mice.  The caffeine tablets that we used came in 200 mg doses.  Using a proportion, we calculated how much caffeine per milliliter of water would be appropriate for our mice.  To do this we assumed that the average body mass of an adult was 68 kg.  The recommended dosage for an adult was one 200 mg pill.  Using the average body mass of our mice we set and calculated a proportion and found that the appropriate dosage for our mice was 0.06 mg.  Setting up another proportion we applied this amount to the average water consumed and found that the mice should receive 0.06 mL per 15 mL of water.  One 200 mg caffeine pill crushed to a fine powder and dissolved in 5000 mL of regular water gave us the correct ratio of caffeine to water for our mice.  Once the stock was made on Day 2, the food of each mouse was replenished to 25 g.  The marked mice were then given 100 mL of the caffeine solution and the unmarked mice had their regular water supply replenished to 100 mL.  The mice were then allowed to habituate for 24-hours before daily testing would begin.



Daily Procedures (this was done between 1530 and 2100)


            Each day we recorded the mass of the mice, how much food was eaten, and how much water the mice drank.  The food was then replenished to 25 g and the appropriate water type replenished to 100 mL.  The mice were then transported in their cages to another lab where they would be run in wheels to measure their activity.  The wheels we used had counters on them so we could keep track of the number of wheel revolutions, thereby quantifying the activity of the mouse.  Using a spring scale we found that the average amount of force needed to turn the wheels was 3.17 g.   Six running wheels with counters were placed individually on six tables (Figure 2a, 2b).  This would allow for few distractions while the mice were in the wheels.  Each mouse was then placed in the holding cage attached to the wheel and allowed to habituate as we took readings of the starting numbers on the wheel counters.  This usually took about a minute or two.  After the readings were taken we opened the door that led from the cage to the wheel so the mice could enter.  Upon entering the doors were closed and the mice were then allowed to run for 30 min.  At then end of 30 min we took the ending readings from the wheel counters and transferred the mice back to their cages and then returned them to the animal lab.  The starting reading from each counter was subtracted from the final reading to obtain the number of revolutions during the 30 min run time.



Figure 2a.  Holding area of wheel setup.


Figure 2b

Side view of wheel setup.





            This procedure was repeated for one week (7 trials).  For week two, the marked mice were taken off of the caffeine solution and put on 100 mL of regular water a day.  The unmarked mice were then switched from regular water to the caffeine solution.  They were also supplied 100 mL a day.  As before when we put the marked mice on the caffeine solution, we again allowed a 24-hour habituation period before testing would resume for another week (7 trials). The data were then analyzed using a t-test on Microsoft Excel© for Windows 2000©.




The data from this experiment did show that the average number of wheel revolutions made by the mice on the caffeine solutions was slightly greater than the revolutions made by the mice on the regular water (Figure 3).  Although there was a difference, our P-value of 0.37 indicated that difference was not statistically significant.  The standard error for our data was 91.64.

            Fig. 3. Average amount of wheel activity.  Error bars calculated based on standard error using 6 mice.





            Our experiment yielded results that did not support our hypothesis that caffeine increases the wheel running activity of house mice.  However, there were several situations that occurred during our experiment that may have contributed to our results.  First of all, during the first week of testing we ran into a problem with one of our wheels.  We had noted that the revolutions produced by mice on one particular wheel were consistently lower than revolutions produced by mice on other wheels.  By consistently lower we mean that there was little to no change in the beginning and ending number on the counter in a 30 minute testing period.  We eventually realized that there was mechanical problem with the wheel, which caused it to require more force than the mouse could apply in order to make it turn.  Because this wheel had been only used three times at this point, and no mouse had been tested in it more than once, we decided to continue to use the wheel for the remainder of the week and then swap it out for a correctly functioning wheel after each mouse had been tested in it once. 

            We also noticed a particular running behavior of the mice. Many times the mice would run for periods of time and come to an abrupt stop while maintaining a grip on the wheel. This would then cause the wheel to still revolve without the mouse actually running. Therefore the mechanical readings were not directly representative of mouse wheel-running activity. This behavior however is consistent with a natural occurring behavior in mice. In a study performed by Kenagy and Hoyt et. al., 1989; Weinstein, 1995; McAdam et. al., 1998, (as cited by Edwards et. al., 2001), intermittent behavior is defined as periods of activity occasionally containing short breaks that don’t provide recovery time for the mice.

            Near the end of week one we noticed a dramatic increase in the body mass of Mouse A.  Eventually we came to the conclusion that Mouse A was pregnant.  As time went on, her body mass continued to increase, and as her abdominal area noticeably grew, we realized that this, in fact, was the case.  We decided to continue to use Mouse A until she gave birth because she was still able to run on the wheel.  Upon giving birth she would be removed from the experiment to prevent strain.  On Wednesday March 31st, 2004 around 1630 Mouse A went into labor and delivered approximately 10 babies.  At this point she was placed separately from our experimental mice and given only regular water.

            Our testing was also interrupted on March 27th, 2004.  We were not able to run the mice this day because the testing facility at Alverno College was closed early preventing us from running the first trial of week two.  The above-mentioned instances should be taken into consideration when reviewing the results of this experiment.

            Should we conduct this experiment again there are several changes we would make.  First of all, we would have liked to test more than six mice.  However, due to time, space, and equipment limitations, we were not able to do this.  Ideally we would like to use at least 12 mice at a time having 6 mice on the caffeine solution and 6 on regular water.  We would also attempt to make sure that the mice we were using were not pregnant.  We could do this by keeping the mice on a regular water diet for a long enough time to ensure that pregnancy in our female mice would not be an issue, then testing would begin.  We would also lengthen the duration of the experiment by keeping the mice on the caffeine solution for two weeks instead of one to allow the mouse more time to become accustomed to the caffeine.  However, prior to lengthening the time on caffeine, we would first have to see if the mice would develop a tolerance to the caffeine during this time period.  The mice would be allowed to habituate to the caffeine solution for an entire week instead of only 24-hours before testing would begin.  Finally, we would again allow the mice an entire week to habituate after being switched from a caffeine solution to a regular water solution and vice versa.

Literature Cited


Austad, S. 2002. A mouse’s tale. Natural History 111: 1-6. Retrieved February 4, 2004 from EBSCOhost database.


Edwards Baker Emily and Gleeson T. Todd. 2001. The Journal of Experimental Biology 204: 59-605. Retrieved February 4, 2004 from the EBSCOhost database.


Girard, I., McAleer, M., Rhodes, J., and Garland Jr., T. 2001. Selection for high voluntary wheel-running increases speed and intermittency in house mice (Mus domesticus). The Journal of Experimental Biology 204: 4311-4320.


Monroe, J. 1998. Caffeine’s hook. Current Health 2 24: 1-4. Retrieved September 26, 2003 from EBSCOhost database.  


Schmid-Holmes Sabine, Drickamer C. Lee, Robinson Sessions Ami, and gillie L. Lynn. 2001. The American Midland Naturalist 146: 53-62. Retrieved February 4, 2004 from the EBSCOhost database.