Effects of Increased Interaction Between Research Rodents and Their Handlers

Does Touch Lower Anxiety Levels in Both, Producing More Conclusive Test Results?

At the University of Notre Dame we have a system in place that allows the principle investigators to utilize our trained laboratory animal technicians and registered veterinary technicians to perform routine animal procedures such as blood sampling. As such, staff is often called upon to take blood samples for a variety of experimental protocols. Mice and rats in these experiments are handled minimally by laboratory animal technicians, yet these are the people who take the blood samples.

Because it has been documented that handling rats is a stressor and that plasma glucocorticoid levels increase within 2-3 minutes of capturing an animal, we feared that the parameters measured during experiments, such as blood chemistries, heart rate, blood pressure, and drug or test component interactions, were being skewed by the stress associated with the handling of the animals. We were also concerned that anxiety caused by blood sampling had a negative impact on the well-being of the animals. Our goal in this study was to show if an increase in human interaction would improve the quality of life for the rodents, because they would not experience high levels of anxiety during routine experimental procedures.

We chose two rodents most commonly used in our facility, the C57Bl/6 female mouse and the LOBUND-Wistar male rat. Studies have revealed that mammals develop social and adaptive skills during the adolescence period of development. In view of that, animals used were obtained from in-house breeding colonies and placed in the experimental groups at weaning age, 3 weeks of age for the mice and 4 weeks of age for the rats. Animals were taken from several litters and randomly placed in control and experimental groups. A total of 12 mice and 12 rats were used, six of each species for the experimental animals and six for controls.

Experimental animals were handled five times per week for three-minute periods (a total of 15 minutes a week excluding cage changing). Control animals were handled only during routine cage changes, biweekly for mice and once weekly for rats. At scheduled times throughout the day, handling was done by a trained undergraduate student and me, a registered veterinarian technician and laboratory animal technologist. Handling consisted of initially grasping the animal by the base of the tail to remove the animal from the cage. They were then held for three minutes in the handler’s palm. Animals were petted and allowed to roam around on the handler’s palm and arm. Observations were noted at the following times:

  • as we first entered the cubicle room (rats only as the mice were on a ventilated rack in an open room);
  • as we removed the cage from the rack;
  • as we removed the cage top from the cage;
  • as we reached into the cage; and
  • as the animals were being held.


Observations were noted as "no reaction," "curious exploration," or "random movements" for the first three time points. For time points four and five, observations were noted as "no reaction," "curious exploration," and "attempts to avoid or escape handler."

Over the first four weeks, it was obvious that the rats were acclimating to the frequent handling. The animals approached the front of the cage as the top was removed. It was not necessary to remove the rats by the base of the tail as they would readily climb into the handlers’ palms. During sample taking, these rats were easily restrained as they did not struggle. The control group did not anticipate the removal of the cage top nor did they climb into the hands of the handler. They were more difficult to restrain as they would not relax as those who had been handled did.

The experimental mice, though slower to acclimate to frequent handling, became much calmer over time. They would actively seek the handler as the cage top was removed and climb on the handler’s palm without the need to grasp them by their tail. During restraint for blood sampling, they were easily scruffed and, like the rats, did not struggle while being restrained.

To quantify our observations, we measured corticosterone (the glucocorticoids present in mice and rats) at day one, day 28, day 63, and day 101. We also recorded body weights and pack cell volumes (PCV) for each animal on those days. Blood samples were taken by trained laboratory animal technicians with assistance from a student.

In a previous experiment, we took only one blood sample from each of the mice and rats to test the corticosterone levels and found no significant difference between control animals who were not handled and experimental animals who had been handled. It was concluded that because the bleeding procedure is done with minimal restraint, the time it took to obtain the sample was less than the time it takes for the activation of the hypothalamus/pituitary/adrenal axis (HPA axis). Without HPA activation, there would not be a spike in the corticosterone levels. However, we then hypothesized that the hormonal reaction would be a delayed reponse that would be detected with a second bleed 20 minutes after the initial bleed. For the rats this was accomplished by taking samples at T=0 and T=20. However, for the mice, the amount of blood required to analyze corticosterone levels at both time points would have resulted in too high a blood loss. Therefore, we used the data from the original experiment as our T=0 time point for the mice. To have a T=20 minute time point, we first did a sham bleed on the mice to simulate the bleed at the initial time point, T=0. This was accomplished by handling the mice in the exact manner as if they would be bled: We restrained them, touched their check with the lancet and held a hematocrit tube to their cheek for 30 seconds. We then did the 20-minutes response test by actually taking a sample at the 20-minute time interval.

Analysis of the results for rats was unexpected. All of the animals maintained their normal PCV level, gained weight at the same rate, and did not show a significant difference in the corticosterone levels at either time point. Analysis of the results for mice also showed that the control and experimental animals did not differ in weight gain or PCV levels. However, corticosterone response to blood collection was significantly lower in handled mice (mean = 588 ng/ml) than in not-handled mice (mean = 818 ng/ml). This indicates that the regularly handled mice had acclimated and no longer were stressed when they were handled by a person.

Although the experiment revealed a reduction of corticosterone response in mice only, it does not reduce the importance handling has on rats. Rats are always more easily handled during cage change than mice, and typically display more interest in and less fear of the care staff. Additional handling of the rats further reduces their anxieties during procedures which was evident in the reduction of struggling during restraint.

We conclude that handling experimental animals more often than only at the time their cage is changed is beneficial to both animals and technicians. Animals experience less anxiety and are more easily handled. This in turn makes the collection of samples less stressful for technical staff because animals are not struggling against restraint, nor are they trying to bite their handlers.

 - Kay Stewart RVT, RLATG, CMAR, Associate Director

Freimann Life Science Center, University of Notre Dame