The idea is that experiments are first performed in the lab, are then performed in animals, and these experiments inform the eventual human studies. As a (seemingly) necessary step in this chain, animal experiments are (rightly or wrongly) tolerated based on their eventual benefit to humans. Animal studies however, are not good predictors of human trials, often do not inform human trials, and are methodologically inferior to human trials, so much so, that the results from animal studies are unreliable and biased. In other words, animal studies are often of no benefit to humans. Arguably, they do not benefit humans at all, let alone enough to justify their use. We either need to fix the problem or get out of the animal research game.
Problem 1: Animal research not translating to humans
Often, research just fails to make the cross-species jump to humans. For example, animal studies showing an association between stress and coronary heart disease were not replicated in humans. Biologically, there are many other reasons why findings in one species are not applicable in another (different immune systems, drug tolerances, behavioural traits, etc.). Also, animal studies often assume ‘ideal’ situations that do not take into account the complexities (concurrent diseases, social aspects, concurrent treatments, etc.) of modern human life and healthcare.
The overall failure of animal research to provide benefit to humans is covered in this 2014 BMJ article (here).
Problem 2: The lack of consideration given to animal studies
Examples exist where animal studies were done after clinical (human) trials had already concluded that the treatment was of no benefit. Other examples exist of animal studies being done simultaneously with (human) clinical trials. A good review of the lack of consideration given to animal research can be found in this 2004 BMJ review (here).
In these cases, human benefit cannot be derived from animal research. Therefore, the research is unethical as the animals have been harmed without providing gain to humans. Oh, and by “harmed”, I usually mean killed. And by “killed” I don’t mean sacrificed, as this implies that the death has been traded for some benefit – I just mean killed.
Problem 3: The lack of quality of animal studies
We often consider studies done in a laboratory to be scientifically superior; to be ‘pure’ or ‘basic’ research’, as opposed to applied or clinical research done in humans, which is (supposedly) complex and harder to control. The opposite is true. Given the regulatory and ethical oversight of clinical research and advances in research methodology, clinical research is now of a very high scientific standard. It isn’t always, but it is getting harder to do bad research, and the overall standard continues to rise.
Animal research is methodologically inferior to human clinical research. The 2004 BMJ review (here) showed that the standards demanded of clinical research are not routinely applied in animal research. For example, animal studies are often not randomised, not blinded, not registered, underpowered (too small), and prone to selective reporting bias and publication bias. Consequently, they are ripe for biased interpretations and p-hacking from the researchers. If human research is meant to be informed by animal research, then the humans had better watch out.
To be blunt, the results of animal research are more likely to be wrong than human research. There is considerable room for improvement in the quality of animal research, but these recent reviews (here and here) tell us that despite efforts to improve animal research, things are still bad.
The bottom line
Animal research either needs to improve or stop. In saying this, I have not considered animal ethics, partly because it is a difficult area, and partly because my argument doesn’t need it. Animal research is methodologically poor, the results unreliable, often not transferable to humans and largely ignored but despite this, it still gets funded because it is considered ‘pure’ research. Animal research should be reduced and refined, and be replaced where possible (the “3 Rs”) Otherwise it is just another WOFTAM whose benefits are, you guessed it, overestimated.
Note: this post is about animal research in which animals are harmed, because in order to balance that harm, you need to have a potential benefit. My argument is that the benefit is either non-existent or much less than we supposed. I have no problem with non-harmful animal research.