Nature has already solved the problem of aging in species all around the world. While most of these species are long-lived trees, there are quite a lot of animals too. To be upfront, most of these animals are primitive, less complex than we humans are. But just like we replicated the technology of manufacturing penicillin from a tiny fungus just trying to defend itself from bacteria, we can achieve the same if only we could understand how these animals escaped senescence.
The first step in doing that is to study these animals and the way reporting on animal research is done is crucial for future human clinical trials. While the latter has improved a lot, I can’t say the same about animal research. When it comes to ethics, progress has been made. But papers are still accepted where basic information about the experiments that were done are missing. Hence this blog post where I outline what must be reported in scientific papers on animal research so that we could glimpse what makes these animals tick. And tick again for decades and sometimes centuries. Not millennia, that is the province of plants. Sorry. Or could this be an excuse to study how plants age too?

Coming back to human clinical trials, these depend on the existence and reporting of animal research studies. The latter can introduce both false positive results by catering to researchers’ biases or false negative results by including a sample size too small for obtaining any meaningful results. I can totally understand the desire to use as few animals as possible. My compassion towards animals is the reason for which I chose medicine instead of biology. And it is one of the major reasons for which I quit my PhD five years ago. It is one of my weaknesses which made me prefer theoretical work as well as human clinical work. But I am grateful to the animals that made so many efficient and safe drugs possible. The road to hell is paved with good intentions and doing animal experiments with too few subjects is unethical whether we’re talking about basic research or applied research – no reliable results can be obtained when the sample size is too small.

Human clinical trials used to be incompletely reported, making them difficult to replicate. This problem has been mostly solved. It is now time to do the same with animal trials because these pave the way towards future drugs and devices for human use.
If you are involved or plan to be involved in animal research, here are the 3 main areas you need to take into consideration for obtaining meaningful results, whether positive or negative:
1. The experimental design
2. Interpreting results
3. Reporting results

The experimental design

This  is the scary part. You have no idea how will everything turn up. The first question you need to answer is whether you are attempting a hypothesis-generating experiment or a hypothesis-testing one. The number of animals you use matters most in experiments that test hypotheses.
You also have to decide which data to include and which one to exclude. More importantly, you have to decide when to stop collecting data right at the beginning. You will include bias if you will keep on doing intermediate analyses on small groups of data and keep on collecting data depending on what you obtain. This is the fastest way to publication because your bias will confirm your hypothesis, but on the long term, human clinical trials may rely on your unusable and inaccurate research.

As long as there will be research done on animals, the section on ethics must exist in published reports. It could be that in the future we could test hypotheses on synthetic organs. But many things can be done today to minimize the suffering of animals and it all starts out with designing the experiment. And gerontology is one of those fields where the minimum degree of suffering is inflicted in animal subjects. Because in order to prove that an intervention prolongs lifespan, you must keep those animals alive instead of killing them. There are many species out there that display interesting patterns of aging, but few of them are used in labs. Measuring the lifespan of wild populations could potentially inflict a slight nuisance on their part, but it could provide huge benefits to the advancement of practical gerontology. Sequencing their DNA could mean one blood drawing per animal which would pose no damage to them on the long term. Things could get tricky when comparing the expression of RNA in closely related animals with differential aging like Hydra oligactis and Hydra vulgaris, but that’s a story for another blog post.

The most reliable results are obtained from double-blind experiments on randomized subjects. This is the target to achieve in animal research too, although it rarely happens.
In order to test a hypothesis, one must group animals in two parts at a minimum and apply one variable only to one group and maintain the other one as control. In order to get reliable results, those two groups (at a minimum) must be randomly assigned. You also want to specify exactly how randomization was done and whether there were any criteria limiting it.
The double-blind part is more difficult to put in practice, but it can be done with the help of numbered tags. The person caring for the animals must be different from the person doing the measurements which must be different from the person doing the statistical analysis. The person collecting the data must not know which animal belongs to which group so that bias is not unwittingly added.

Interpreting results

It so happens that animals are removed or die during the time it takes to collect the measurements. This can happen for a multitude of reasons, some of them having nothing to do with the experiment itself. But you must report on the amount of data you have left from what you initially expected.
It is also important to avoid doing any intermediary analyses on your data before finishing collecting it because you will introduce bias in your results. But if you did this, you must report it in the subsequent paper.

Reporting results

You’d be amazed how many papers omit to report the number of animals that were used in experiments. From the scientific point of view, it is best to use a sample number as high as possible to get reliable results. At the same time, ethics panels limit the number of animals that can be used to a number as low as possible, especially in pilot projects. And the easiest way to hide the unreliability of your data is to omit mentioning the sample number. Only that in so doing, future human clinical trials could be just as unreliable. Just my two cents with added inflation.

The second thing inaccurately or incompletely mentioned in papers all over the world is the method(s) used. Publishing original research is only the beginning. Since any study has its bias and errors, results must be replicated. And how could they be replicated when the methods used are so vaguely described? Or when the animals used were in-house genetic breeds? Or when no mentioning is done regarding the source of the animals or the criteria through which certain individuals were selected for the experiments? All these must be reported so that other people can replicate your experiments as reliably as possible. Ideally, each experiment must be repeated several times before publication.

Many publications nowadays archive datasets and I think it’s a step forward. This wasn’t always how things were done, but nowadays many researchers upload their data so that other people can analyze the same thing and see whether they skipped anything. This takes time, of course, with many researchers doing their best to publish original research instead of replicating others’, but with the advent of AI, these analyses could be done automatically before a paper is even accepted for publication.

Finally, design your experiment, interpret your results with care and finally report them completely and accurately so that basic science and medicine progress even further. If you want to do animal research, do it right and make it worthwhile!

If you have any other tips on ways to improve animal research, I’d love to hear from you in the comments below!

Bibliography:

Here is one guideline for reporting animal research if you want those studies to create a tangible benefit in human health and longevity:

Landis, Story C., et al. “A call for transparent reporting to optimize the predictive value of preclinical research.” Nature 490.7419 (2012): 187-191.

Anca Ioviţă is the author of Eat Less Live Longer: Your Practical Guide to Calorie Restriction with Optimal Nutrition ,The Aging Gap Between Species and What Is Your Legacy? 101Ways on Getting Started to Create and Build One available on Amazon and several other places. If you enjoyed this article, don’t forget to sign up to receive updates on longevity news and novel book projects!

Don’t miss out on the Pinterest board on calorie restriction with optimal nutrition where she pins new recipes every day.
https://www.pinterest.com/longevityletter/eat-less-live-longer/
Or the Comparative Gerontology Facebook Group where you can join the discussions on how species age at different speeds and what could be the mechanisms underlining these differences!
https://www.facebook.com/groups/683953735071847/