Unreproducable results from the animal lab

Experiments depend n conditions as well as species selected

Experimenting on animals has one advantage over studying humans.  They stay in the lab.  While humans live individual lives, have individual histories and undertake a range of activities which affect their lifestyle, lab animals are standard.

Unfortunately for defenders of vivisection, this isn't true.  Apart from the massive problem of species difference, there are countless factors which affect the results of an experiment.  They're so many, and often so emphatic, that experimenters don't know whether they're seeing a reaction from the procedure they're testing, or some other factor.

One that some experimenters do acknowledge is the sex of the animal.  Gentamicin, (an antibiotic), causes kidney damage in male rats, but not in females.  (In humans, there's no sex difference.)[1]  Female rats are immune from copper poisoning, but males can be affected.[2]  Hexachlorobenzene is more likely to cause skin and nerve blistering in female rats than mice, but in humans the reverse is true.[3]

Sex also plays a part in the metabolism of drugs and chemicals.  In particular the bile is affected, which is a crucial part of the process which deals with toxins. When rats are given tartrazine, males excrete 17% of the dose they’ve been given in their bile, and 70% on their urine.  Females excrete 40% in their bile and 45% in their urine.  Other standard lab animal species including rabbits and guinea pigs also vary according to sex.[4]  This effect crops up with various different chemicals, turning animal experiments into even more of a lottery.


The type of materials in a lab affect animals, sometimes quite drastically.  Two colonies of mice were held in Australian and American labs.  They were very similar strains, but almost all the American ones developed cancer in the liver and breast, but this was almost entirely unknown in the Australian mice.[5]  Housing was reported in July 2010 to have a large effect on cancer in mice. (Read here)

A similar problem affected a study on a sleeping drug using mice.  Mice in one set were sleeping for only half the duration that the others slept for.  The cause was identified to be the use of bedding, red cedar chips with one group and ground corncob bedding in the other.[6] This dramatic effect prompted an investigation and was eventually identified, but how many other results are more subtly influenced by such trivial factors?


Anaesthetics offer another role of the dice because they affect how the body works.  A mixture of chloralose and urethane is used to anaesthetise dogs, yet it causes the cardiovascular system to decline[7].  This makes the interpretation of heart drugs tests impossible from dog experiments, and the crucial role of the heart means that this in turn will influence countless other factors.

Animals are commonly given too much anaesthetic, which leads to biological effects sufficiently significant to cause the death off the animal in an estimated 9% of cases.[8] How many others are just sufficiently biologically disturbed to create reactions experimenters think is part of the test?.   A lab manual admits "…we are over anaesthetizing animals…this is having a detrimental effect on the well-being of the animals, and possibly the validity of the experimental results which are obtained."[9]

Not giving pain relief to animals increases stress.  Giving pain relief has unknown reactions.  Veterinary surgeons have found that:

“Considerable variation occurs among animal species in their response to drugs used to alleviate pain and distress.  The comparative pharmodynamics and kinetics of most agents are unknown for many species, especially the smaller laboratory animals. Extrapolation of data from one species to another is fraught with error and should be avoided.”.[10]


Experimenters discovered decades ago that mice exposed to excessive noise suffered an increase in the damaging affects of sympathomimetic amines (a body chemical).[11] It has been frequently referred to on animal technician training courses that noise is a factor affecting animals; it's a matter to consider when considering issues such as choice of cleaning equipment.  Rates among lab animals of adult death, infant mortality, and destruction of young by parents are anecdotally believed among lab staff to be likely to change if a member of staff linked to the animals, changes.  Their routines and the amount of noise they make are likely to be a large part of this.  Behavioural research has discovered that “If the experimenter simply enters the animal room without touching the cage, there are immediately strong elevations in HR [heart rate].” [12] Which goes on to affect other aspects of the body.

Temperature and ventilation

Increasing the room temperature by 11°C (20° F) has been found among mice to increase the effect of sympathomimetic amines.[13] Lowering it has been found to cause rats to be more likely to develop ulcers.[14]

When ventilation has been examined, it is found to be a notable effect: ”If given the opportunity, the rats avoided cages with CO2 concentrations of 3% whereas no preference was observed if the CO2 concentration was only 1%”. “A previous study showed that CO2 in unventilated IVC cages in some types of cages, reaches levels about 3% within 20-45 minutes (Krohn & Hansen, 2002). Elder studies indicate that exposure to CO2 levels above 3% may affect the animal in a way comparable to stress.” “…exposure to CO2 affected both preferences and heart rate but not the systolic blood pressure of the rats, and at 5% CO2, the blood pressure was also affected'[15].

Moving animals

Often animals are bred in one area, and are usually moved from that area or room for the first time since birth, immediately before an experiment.  Just moving rats has been found to a change of 10-20% in the heart rate, packed cell volume, haemoglobin and plasma protein content.  Serum glucose levels, pyruvate and lactate concentrations were up 20-100%.[16]  If this were not enough in itself to render experiments useless, consider that if just moving them caused biological reactions, the often traumatic experiments are likely to provide vital reactions on a greater scale.

The general definition of an effective experiment is one that gives the same result if you repeat it.  If you repeat it tomorrow, or someone else repeat’s it elsewhere, you’ll get the same result as you did today.  In the case of animal experiments, the myriad of factors influencing the experiment mean that the chances are that you won’t, and if you do, you won’t trust it.

Ironically, this massive spanner in the works of the vivisectionist is a reason why animals remain in labs.  Species difference has provided convenient flexibility:  You can select the right animal to give the right results, and therefore create data to support what you want to say, regardless of whether it’s true or not.  The countless different ways in which test conditions can be altered provide more options for results to be manipulated until the results hoped for are achieved.

It also provides options for more experimenters to more experiments on the same subjects, under slightly different conditions.  But still no-one can be sure whether the results reflect the effect of the test drug, or whether noise, diet, bedding, sex of the animal, the time they were moved, the technician’s deodorant or some other, currently unidentified yet influential factor have changed the results beyond what could happen if they tried again tomorrow.



1 Edward J Calabrese, Toxic Susceptibility: Male/Female Differences, 1985, p11

2 Edward J Calabrese, Toxic Susceptibility: Male/Female Differences, 1985, p109

3 Edward J Calabrese, Toxic Susceptibility: Male/Female Differences, 1985, p140

4 Edward J Calabrese, Toxic Susceptibility: Male/Female Differences, 1985, p12-3

5 Sabine, JR, Horton, BJ, Wick, MD 'Spontaneous tumours in Cs4-Avy and C3H-Avy of B mice' J Nat Cancer Inst. 50: 1237-1242

6 Domer, FR, Animal Experiments in Pharmacological Analysis, 1971, p14

7 Domer, FR, Animal Experiments in Pharmacological Analysis, 1971, p14

8 Laboratory Animals Vol 26, No 3 Jul 1992, p159

9 Laboratory Animals Vol 26, No 3 Jul 1992, p159

10 Journal of the American Veterinary Association 1987; 191:1227-30

11 Domer, FR, Animal Experiments in Pharmacological Analysis, 1971, p15
12 D.Spani, M. Arras, B. Konig and T. Rulicke, 'Higher heart rate of laboratory mice housed individually vs in pairs', Laboratory Animal Welfare, Vol. 37, No. 1, Jan 2003, pp.54,55,60,61.

13 Domer, FR, Animal Experiments in Pharmacological Analysis, 1971, p15

14 Domer, FR, Animal Experiments in Pharmacological Analysis, 1971, p16

15 'The impact of low levels of carbon dioxide on rats' TC Krohn,
AK Hansen, N Dragsted in Laboratory Animal Welfare, January 2003 (Vol. 37, No. 1) p94, 96, 97

16 Gartner, D 'Stress responses of rats to handling and experimental procedures' Laboratory Animals 14, p267


New Scientist. 17 May 2010.
Flab rats: Unfit animals are bad for experiments.

RATS and mice are the most widely used animals in biomedical research. They are hugely important in the early stages of drug testing, before clinical trials in humans can begin. Unfortunately, laboratory rats and mice are not up to the job. They are kept under conditions that give them little opportunity for exercise, and at the same time they have access to as much food as they can eat. As a result they are overfed, unfit and obese. Such "couch-potato" animals develop a range of health problems, including insulin resistance, diabetes, high blood pressure, impaired brain function, increased oxidative stress and inflammation. They also have elevated levels of glucose, triglycerides and "bad" cholesterol in their blood, and are more susceptible to cancer, neurodegenerative disease and kidney disease. Unsurprisingly, they die prematurely.

Obesity is especially problematic. Rodents kept under normal lab conditions gain weight throughout their adult lives. Up to 50 per cent of the body weight of a typical middle-aged laboratory rodent is fat. Some strains of lab rat attain a body weight in excess of 1 kilogram, nearly double that of a healthy rat.

Surprisingly, very few scientists who use rats and mice in their research are mindful of the fact that their animals are so unhealthy.

They ought to be. The widespread use of sedentary, obese animals may be leading to spurious experimental results. Drugs shown to be effective in couch-potato rodents may prove ineffective - or have side effects - in fit and healthy subjects, while promising drugs may be thrown out before ever being tested properly because they don't work in couch-potato animals.

Fiercebiotechresearch. 18 May 2010.
Are flabby rats derailing preclinical research programs?

The lab rats used in biomedical research are often significantly overweight, says one researcher, and that could skew a scientist's work on a new therapy. Mark Mattson, chief of the laboratory of neurosciences at the U.S. National Institute on Aging Intramural Research Program, writes in New Scientist that some lab rat strains weigh in at more than a kilogram, roughly double the weight of a healthy rat. As a result, the "flab rats" suffer from hypertension, cholesterol problems and more that could derail a research
program. "We know that some carcinogens are more potent in overweight animals and that couch-potato rodents have an elevated risk of developing tumors," writes Mattson, a fitness buff himself.

"In addition, many types of tumor grow more rapidly in animals with unlimited access to food, and certain aspects of metastasis--the process by which tumors spread to new sites in the body--appear to differ between obese and slender mice." Mattson's not just fretting about cancer research. Fat rats could be responsible for spoiling biomedical research into cardiovascular, neurodegenerative and renal drugs as well.