Imagine you are a monkey in a laboratory and a person dressed in a white coat walks into the room with a catching net. How do you think you would react? You would probably not be surprised to learn that monkeys in this situation immediately show significant distress.
What may surprise you, however, is that the distress that animals in laboratories experience is one of the main reasons why animal experimentation doesn't work.
In the first part of my series on animal experiments, I discussed how more scientists now acknowledge that animal experiments are ineffective for understanding human biology and predicting human outcomes. Here, I discuss the first of the top three reasons why. There are other potential reasons for animal experimental failure, such as bias in reporting animal experiments and lack of standardized protocols.
However, the key problem with the three causes that I will discuss is that they are fundamental flaws of animal experimentation itself. They simply can't be fixed with better experimentation, reporting or protocol.
Let's look at the first main reason why animal experimentation doesn't work: stressed animals affect research results.
Animals in laboratories are involuntarily placed in artificial environments; usually in windowless rooms. Routinely, these animals are caught from their cages, placed in restraining chairs, have blood taken from them, and subjected to experiments that cause significant pain. Dogs, cats, monkeys, rats, mice and other animals are often denied food, water and even pain medications. They are burned, irradiated, crushed, shocked, poisoned and infected.
It's important to understand how animals live in modern laboratories because what we do to them -- how we house them, how we handle them, how we experiment on them -- all can cause misleading research results that eventually harm us.
Arguably, everything we do to animals in laboratories causes them substantial distress. It does not matter if they were captured from the wild, seized from pounds or bred for experiments; just keeping animals in laboratories causes them to develop abnormal behaviors such as rocking back and forth or injuring themselves.
Mice in laboratories commonly show clear signs of distress and, at the New England Regional Primate Research Center, almost 90% of monkeys show behavioral abnormalities. These abnormal behaviors indicate that animals are psychologically distressed.
As described in my book, there is a phenomenon in laboratories known as "contagious anxiety". Blood pressure and heart rates shoot up in rats watching other rats being decapitated. Cortisone levels rise in monkeys watching other monkeys being restrained for blood collection.
Just the routine laboratory procedures, such as catching an animal and removing him from his cage, in addition to the experiments themselves, cause significant and prolonged elevations in animals' stress markers, which can affect research results.
What these examples show is that animals don't habituate to the laboratory environments and the procedures--they don't get used to it. Fear and anxiety are daily parts of their lives. Their distress causes changes in their physiology that affect research data in very unpredictable ways.
For example, New Scientist article "Too Stressed to Work" shows how stressed rats develop lasting inflammatory conditions, which causes their intestines to leak. As expressed in the article:
"This inflammation adds uncontrollable variables to experiments....confounding the data".
How we house and test animals add unknown variables that can also affect research results. Conditions in the laboratory are shown to cause unpredictable changes in neurochemistry, genetic expression and nerve regeneration. In one study, mice were genetically altered to develop defects in their hearts. But when the mice were put in larger cages, those defects almost completely disappeared (1). It's been shown that the noise levels in laboratories damage blood vessels in animals (2). Even the type of flooring on which animals are tested in spinal cord injury experiments can affect whether a drug shows a benefit or not! (2)
Some might argue that we can control these factors and standardize the laboratory conditions and procedures. That's exactly what one team tried to do. They found that despite all attempts to standardize the environments across three labs, there were systematic differences in test results. What's more, different mouse strains showed markedly different results.
No matter how hard we try, we cannot fully standardize criteria like how often and in what manner animals are handled. More importantly, we cannot control and predict how animals will react, whether physiologically or behaviorally, to the procedures and settings in the laboratories.
Some might also argue that stress affects humans and therefore the animals in laboratories are not that different from us. One major difference is that these animals have no control over the situations that cause them stress. They can't avoid being caught and put into restraining chairs. They can't avoid being experimented on. They simply can't relieve their pain. And we subject animals to these stressful situations over the duration of their lives in the laboratory.
How these conditions and procedures affect research results is unpredictable and usually not obvious. That's because each species and even different strains of mice and rats may respond in different ways physiologically to the same stressors.
The cumulative effect is that these stressors cause animals to be less reliable and representative of human biology.
The very nature of animal experimentation adds unknown variables that lead to misleading research results. The only way to avoid this and improve the accuracy of research data is to avoid using animals. We can do better. Cut out the animal experiments and move science forward.
Stay tuned. In the next article, I will discuss the second major reason why animal experimentation doesn't work: animals don't get human diseases.
1. Baldwin, Bekoff. New Scientist 2007; 194: 24
2. Akhtar et al. Reviews in the Neurosciences 2008; 19: 47-60