Extracts from my report on animal research flaws

It may be some time before I can put together a coherent blogpost on animal research related to the gut microbiome and ME (I'm always pressed for time) so here is some more general info originally posted in the 'Other Health News and Research' forum.

The problems

Basic research into human conditions usually involves very crude ‘models’...which are not only in the wrong species but also do not accurately model the conditions.

Chronic illnesses take many years to develop and have complex causes. Heart disease and stroke generally arise in older humans due to lifestyle factors, so cannot be reproduced by, for example, damaging the hearts of young, healthy dogs. Arthritis usually develops slowly and is due to wear and tear or the body’s immune system attacking the joints. Injecting a laboratory animal to create arthritis over a few weeks cannot possibly reproduce the human disease and give reliable findings.

Genetically modifying animals is another crude tool which cannot mimic the human condition. Genes do not act in isolation – they interact with other genes, biochemicals and the external environment – so changing one or two genes in an animal is not likely to produce useful data.

The sheer illogicality of much animal research is starkly illustrated in this quote by Auer et al.: “the repair of articular cartilage defects has been studied in many species...none of these species replicate the anatomical, cellular, and biomechanical properties of the human knee.” The quote is from a 2005 US study which aimed to determine whether patients should rest or exercise after a knee injury treatment - by injuring the kneebones of monkeys in the assumption that this would provide more relevant results than the other species. The rational option would be to analyse clinical data for some of the many sportspersons who had had such treatments and thus ascertain what had worked best!

Auer et al. continue: “Such dilemmas are frequently faced in the selection of appropriate animal models for biomedical research. Relevance to human conditions varies widely...Several bone and connective tissue human diseases do not naturally occur in animals (e.g., osteoporosis appears to be human-specific) and if they do, such diseases behave quite differently...”

The authors refer to differences between species in cell biology, immunology, biochemistry, wounds and healing, inflammation, complications, tissue structure, susceptibility to infection, bones, posture and the effectiveness and safety of painkillers. They describe the crudeness and inaccuracy of the experimental models and the problems added by unnatural laboratory conditions, and point out: “Many effective animal studies show no direct correlation to results in the most significant test – the human – and this should be continually emphasised.”

Scientific denial

Yet the authors, whilst conceding so many shortcomings of animal models, advocate refinements of animal research. As many authors, including themselves and van der Worp et al., have found that scientists are already poor at following existing guidelines in designing, carrying out and reporting their work, it is unrealistic to expect them to want to bother with further, inconvenient refinements and calculations, or follow new guidelines better than they do existing ones. Like most ‘life’ scientists, the authors ignore the elephant in the room: the fact that no amount of refinement will make animals reliable predictors of the human condition.


Because animal use is so widely established and accepted as the norm, scientists who use animals still recommend refining the ‘models’ even whilst acknowledging their many shortcomings (e.g. see Auer et al., 2007). Dr Graham Lappin said in a discussion on human microdosing “I think that the pharmaceutical industries need to be a lot more courageous in adopting a lot of this technology. They are conservative by nature and some of them are very risk averse and new technologies always come with a risk.” (Goozee, 2007)


It has long been difficult to gain life science qualifications without having to harm animals or collude in such harm. Thus the life science education system tends to screen out the more compassionate would-be life scientists. Those who start out with a level of compassion can become desensitised by repeated exposure to animal abuse and suffering. Others, who have little compassion to start with, can complete their studies unhindered by qualms.

Fear and resistance

  • Many scientists are reluctant to ‘come out’ and advocate the complete replacement of animal experiments rather than simply refine and reduce them, which is now widely accepted as desirable. This probably has a range of origins, including habit and reasons cited below.
  • Scientists are reluctant to appear to be siding with ‘animal rights extremists’ – that term so beloved of sensation-seeking journalists. (In fact, such extremism is very rare in relation to the large numbers of law-abiding people opposed to animal experimentation and other abuses.)
  • Scientists are afraid of alienating and/or being ostracised by less-enlightened colleagues.

Inadequate infrastructure

It is unnecessarily difficult to register one’s wishes to donate one’s body, tissues or organs for medical research. This is compounded by the fact that many researchers use human tissue in animal experiments, which wastes the valuable resource and potentially deters many people from donating. Some of the problems are summarised here: http://www.scienceroom.org/reviews_1 (sorry - this link is now dead)

A lot of human material obtained in hospitals and other clinical settings, for example during surgery, is simply thrown away.

As a result, it is difficult for scientists to access human bodies and materials that could be used instead of animals and animal materials.

These problems are due to the lack of a properly-coordinated system for collecting and using these valuable resources.

Irrational validation processes

Replacements for live-animal tests are, absurdly, tested for effectiveness against the animal tests (Hoffmann et al., 2008a), despite the facts that they themselves have never been formally tested and that they are very poor models for humans. For example, results from a live-rabbit skin test, used to predict whether a substance will corrode or irritate human skin, have conflicted with those from patch tests on people’s skin about 50% of the time (Hoffman et al., 2008b). Yet the animal test is used as the standard which alternatives such as EpiDerm are expected to match as closely as possible. Reluctance of some governments to sanction human skin patch tests on ethical grounds means that very few chemicals have human data against which to validate alternatives.

Legal and regulatory constraints

To quote Thomas Hartung, professor of pharmacology and toxicology, “It is increasingly recognized that it is neither the lack of new approaches nor their proven reliability by validation, but that translation into regulatory guidelines and use is now the bottle-neck of the process.” (Hartung, 2009)

Until regulatory requirements permit the substitution of animal methods with non-animal alternatives, and until the regulations are made sufficiently clear and communicated to those involved in testing, and as long as the officially-accepted animal methods give manufacturers a degree of protection against prosecution for damage caused by their products, the animal methods will inevitably continue to be used, sometimes irrespective of whether they are legally required or whether they protect those they are claimed to protect – the patient and the general public.

With the large, diverse, versatile and growing range of reliable alternatives available, it is saddening to note the extent to which the regulatory and scientific communities are unwilling to abandon cruel, expensive, primitive and unreliable methods. Science is a discipline based on the quest for knowledge and advancement, but today has become suffused with conservatism and inertia. Fortunately there are still scientists with that original drive for discovery and progress. If they can just convince the regulators, we can start to move forward at the required pace, to the benefit of all.

Auer, J.A., Goodship, A., Arnoczky, S., Pearce, S., Price, J., Claes, L., von Rechenberg, B., Hofmann-Amtenbrinck, M., Schneider, E., Müller-Terpitz, R., Thiele, F., Rippe, K-P. and Grainger, D.W. (2007) Refining animal models in fracture research: seeking consensus in optimising both animal welfare and scientific validity for appropriate biomedical use, BioMedCentral Musculoskeletal Disorders, vol. 8, Article 72, online at http://www.biomedcentral.com/1471-2474/8/72

Goozee, R. (2007) Failing Faster, Next Generation Pharmaceutical Magazine Issue 3, March 2007, online at http://www.ngpharma.eu.com/article/Failing-faster (accessed 13.6.10)

Hartung, T. (2009) A toxicology for the 21st century - mapping the road ahead, Toxicological Sciences, vol. 109, no. 1, pp. 18-23, online at http://toxsci.oxfordjournals.org/cgi/reprint/109/1/18.pdf

Hoffmann, S., Edler, L., Gardner, I., Gribaldo, L., Hartung, T., Klein, C., Liebsch, M., Sauerland, S., Schechtman, L., Stammati, A. and Nikolaidis, E. (2008a) Points of Reference in the Validation Process, The Report and Recommendations of ECVAM Workshop 66a, Alternatives to Laboratory Animals vol. 36, pp. 343-352, downloadable from

Hoffman, S., Gallegos Saliner, A., Patlewica, G., Eskes, C., Zuang, V. and Worth, A.P. (2008b) A Feasibility study developing an integrated testing strategy assessing skin irritation potential of chemicals, Toxicology Letters, vol. 180, pp. 9-20, abstract online at http://www.ncbi.nlm.nih.gov/pubmed/18585875

van der Worp, H.B., Howells, D.W., Sena, E.S., Porritt, M.J., Rewell. S.,O'Collins, V. and Macleod, M.R. (2010) Can Animal Models of Disease Reliably Inform Human Studies? Public Library of Science Medicinevol. 7, Issue 3, e1000245, online at http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.1000245
Likes: Joh


There are no comments to display.

Blog entry information

Last update

More entries in User Blogs

More entries from MeSci