|
RETURN
TO THE WORLD SCIENCE HOME PAGE
New advice to scientists: get
to know your lab animals
Posted May 4, 2005
Special to World Science
Scientists have labored for years to standardize their lab animals — breeding armies of near-clones in some cases — in order to get more reliable, consistent scientific results.
But now, a pair of scientists is advising those who toil in animal-research laboratories to
go a step further.
If the pair’s findings are
correct, to get the best results, scientists might have to spend some time with the little
creatures that they might later have to kill for science.
The point of all these measures is to ensure that the same test give the same result each time it is performed. Otherwise no one could verify the accuracy of the findings.
Animals who are familiar with their experimenters may help achieve this goal, according to the research team, which published the findings in the April issue of the journal
Behavioral Brain Research. Familiarity can “increase consistency in results from animal tests,” wrote the researchers, Katja S. van Driel and Janet C. Talling of the Animal Welfare Team at the Central Science Laboratory, York, U.K.
The pair had rats from the same breeding colony undergo a commonly used laboratory anxiety test. They found the identity of the experimenter had a significant effect on the results if the experimenter was unfamiliar to the animals, but not if the experimenter was familiar.
The research comes on the heels of a previous, controversial review study suggesting that animals are usually scared when research is done on them, even during relatively mild or harmless-seeming procedures.
The authors of this previous work also warned that the fear might throw off the test results, because frightened animals release stress hormones that can change their internal chemistry and responses.
It’s not clear whether the newer study gave the results it did because familiarity eases the animals’ fear, van Driel said. “Familiarity does not necessarily reduce anxiety,” she wrote in a recent email. “How exactly it does influence results is not clear.”
The findings do show that the human-animal interactions are more important than many researchers
realize, van Driel and Talling wrote. Although a 1992 book, “The inevitable bond : Examining scientist-animal interactions,” edited by Hank Davis and Dianne Balfour, did spark debate over the issue, it hasn’t been studied systematically before now, they explained.
“Human-animal interactions are generally regarded as being inconsequential,” van Driel and Talling wrote. “Our results clearly show that they are not. To minimise effects of experimenter identity on results from the EPM test at least, the experimenters should, wherever possible, be persons ‘familiar’ to the animals.”
In modern laboratories, the scientists leading a study often don’t handle the animals directly but leave this work to students, post-doctoral researchers, technicians and others.
In their study, van Driel and Talling assigned a set of female caretakers to look after rats for at least six weeks. By the end of this period, each rat had spent about 14 hours in the presence of one of these caretakers, although this included only a few minutes of direct physical contact.
The rats were then subjected to the test after a brief handling by either their assigned caretaker, or a stranger.
The contact with the stranger didn’t make the results significantly different on average — but it did make them more inconsistent, the researchers found.
Although the authors didn’t conduct other types of tests besides anxiety assessments, they left open the possibility that the familiarity effect might contaminate results from other types of tests as well.
Indeed, their study isn’t the first to have found that an experimenter’s identity can influence results. A University of Illinois study published in the November 2002 issue of the research journal Nature Neuroscience found that mice respond differently to pain tests depending on several factors related to the laboratory environment, including experimenter identity.
But van Driel’s and Talling’s study is the first to systematically examine the differing effects of familiarity and unfamiliarity, they said. However, they cautioned that their study is far from the final word.
“Further trials are needed to determine which aspect of the person the rats are responding to. The rats may not have been responding to the people per se but rather to the smell of the laboratory coats they were wearing,” they wrote. “If this is the case then there is the potential to ‘fake’ familiarity.”
—EJL
* * *
Send us a comment
on this story, or send
it to a friend
|