Project Syndicate - Research on animals is performed to gain more knowledge about diseases and how to cure them, and to evaluate drugs for toxicity before testing them on humans. In fact, animal studies have played a vital role in almost every major medical advance.

Although researchers are committed to finding new ways to reduce and replace animal testing, current technology cannot yet replace many types of animal research. The Nuffield Council on Bioethics estimates that between 50 and 100 million animals, from flies to monkeys, are euthanized for research each year worldwide, with roughly 90% of the vertebrate animals used for research being rodents.

Gaining useful knowledge from animal research requires robust experimental findings: different scientists should be able to reproduce them in different locations. This requires a thorough understanding of each animal species and its biology.

Evidence is gradually accumulating that the majority of mammalian research animals, particularly rodents, are mentally stressed by their living conditions. Stress is generally defined as the state that results when the brain instructs the body to make changes in order to adapt to a new or excessive demand and the individual perceives that the demand will exceed the personal resources which he or she has available. The response is fuelled by stress hormones that flow through the body, altering every organ and biochemical function, with wide-ranging effects on metabolism, growth, and reproduction.

Although the environments of laboratory animals are usually well controlled in terms of lighting, temperature, and humidity, there are many uncontrolled sources of noise in animal facilities, most of which derive from human activities. These include high-pressure hoses, cage cleaners, and air-conditioners/heaters; squeaking doors, carts, and movable chairs; and jangling keys.

Rodents, in particular, are sensitive to these noises, and studies show that this sensitivity does not diminish with time, as is commonly assumed. These noises can alter rodents� behavior and even adversely affect their health. Yet, surprisingly, many scientists are unaware that loud noises in their animal facilities can affect research outcomes and compromise their data.

Apart from noise problems, research animals are often housed in small cages with no source of enrichment, such as wheels, shelves, or tubes. Such devices enable animals to exert some control over their environment, such as escaping an attack from a cage-mate by moving to another level in the cage or hiding.

Often, researchers are unwilling to include such items in their animals� cages because other researchers do not. However, rigorous standardization of the environment, particularly if it leads to barren surroundings, increases the risk of obtaining results that, being specific to a narrow set of conditions, cannot be compared with other researchers� results.

If animals are under stress, they can have permanently raised concentrations of stress hormones, reduced concentrations of sex hormones, and compromised immune systems. These uncontrolled variables make the animals unsuitable subjects for scientific studies. To ensure good science, research animals should be healthy and exhibit normal behavior, apart from the specific effects under investigation. Researchers often dismiss questions concerning environmental influences on their experimental data by claiming that such effects �cancel out,� because their control animals are housed under the same conditions. But the conclusions drawn from such experiments are specific to the stressed animals and cannot necessarily be extrapolated to healthy animals.

The increasing use of genetically modified mice since their advent 20 years ago amplifies this problem. Genetically modified mice either lack a specific gene or gene-pair (knock-out mice) or carry a piece of foreign DNA integrated into their own chromosomes (transgenic mice), and are used to deduce the functions of particular genes. Studies are beginning to show that an animal�s environmental conditions can completely change the results of genetic studies.

A more humane and effective way to ensure the validity and usefulness of animal experimentation would be to provide conditions that minimize stress-related activities such as excessive grooming and fighting. In addition, the environment should allow the animals to perform the behaviors normal for their species. Normal and aberrant behaviors for each species could be agreed upon institutionally and a list made available to all investigators. Designing an environment to suit animals� psychological and physiological needs would be far preferable to the minimalism, otherwise known as �standardization,� that is currently employed.

The exact conditions used to achieve these goals would probably vary between laboratories. But the end result would be similar. Both the welfare of research animals and the quality of the science would be markedly improved, leading to data that could be meaningfully applied to our quest for medical knowledge.

Ann Baldwin is Professor of Physiology and Psychology at the University of Arizona.

Copyright: Project Syndicate, 2007.
www.project-syndicate.org