Posted: Tue, April 02, 2013 | By: Biohack
by Winslow Strong
This is the second post in the Biohacking 101 series. The first post gives an overview of what biohacking is all about and a compendium of sources for identifying potential biohacks. Here I will discuss the reasons for self-experimentation. In particular, I will detail how self-experimentation supplements both the results of population-average trials and the “collect data and seek out patterns” approach that is now popular in the Quantified Self movement.
This essay first appeared in Winslow’s blog, Biohack Yourself, HERE
In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. -Galileo Galilei
A self-experiment is an experiment performed on yourself and by yourself.
That is, you are both the object of the experimental treatment and the experimenter conducting it.
Self-experimentation has a long history in science, especially in medicine.
It’s a natural manifestation of human curiosity, ingenuity, and the desire for self-improvement.
As I expressed in my bio, I have found personally that self-experimentation in career, lifestyle, and health, has been incredibly rewarding for me in terms of growth in these areas. As a bonus, it’s often quite fun. However, these days many of us have become accustomed to asking Google for best practices and assuming that what we find is the distillation of human science and wisdom. How could little old you hope to arrive at better practices without becoming an expert in a field that you currently don’t know much about?
Big science today: population average trials
Almost all big scientific trials involving humans as test subjects address the question of whether an experimental treatment affects a certain quantified objective on average in a certain population. In the coarsest sense, this yields information along the lines of: if you select a random human being on earth, and feed him or her 100g of French fries per day, then on average he or she will outlive those not receiving this treatment. I made up this example, but it’s plausible given that much of the world is malnourished, and providing more food of nearly any type to a malnourished person is likely to benefit his or her survival.
Adding specificity to the population group can drastically change the results
Population average studies often don’t highlight the level of variation between individuals within the study, although they typically record this information. It would be easy to see how the favorable result from the hypothetical french fry treatment could in fact be unfavorable for some segments of the world population. On average, would forcing Americans to eat 100g of French fries a day improve their longevity? I doubt it. Americans consume sufficient food, and too much of some types. French fries cooked at high temperatures in processed, omega-6 rich vegetable oils have a pathological effect on the health of nourished subjects. The original experiment is not sensitive to the important variable of the nourishment level of the test subjects.
The ultimate specificity for you is so-called n=1
Biohacks are ideally carried out as self-experiments. When selecting a biohack, it’s nice to know that it has been effective for a segment of the population, even better if that segment resembles you. This narrows down the search process from a huge universe of potential biohacks to the ones that most plausibly might work for you. However, the biohack is only useful if it actually does work for you. The only way to become very sure of this is to perform an experiment on yourself.
No one is identical to you. Even identical twins, still in the womb, have differences due to epigenetic factors resulting in different gene expression despite identical DNA base pairs. In theory, self-experiments can be sensitive to all of the uniqueness that is you. In practice, repeated measurements must be carried out over time, so self-experiments are reproducible in so far as their outcomes are unaffected by the properties of you that change over the experimental measurement process. If we can nail down our certainty to all but those changes in us over time, that would be quite strong evidence to draw conclusions from. We would still fool ourselves occasionally, but not very often.
But n > 1
Self-experimentation is now popularly being referred to as n=1 experimentation, because instead of many test subjects, like there are in population-average trials, there is just one subject, yourself.
But this n properly refers to the statistical parameter representing the number of (independent) observations. In a self-experiment, n=1 if you, for example, pop a happy pill one day and feel happy that day. Not because there is just one subject, but because there is just one observation of one trial (one pill, one day). It would be poor logic to conclude from one trial that a treatment is effective (unless perhaps your happiness was well outside the bounds of anything that you had ever experienced before). I.e. you would be fooling yourself.
In contrast, if each day you independently have a 50% chance of being happy overall and a 50% chance of being unhappy, but you try taking a particular pill a day for 10 days and notice that you are happy every day, then , and this is a highly significant result. Even better if you blind yourself and randomly take the treatment pill or a placebo.
Personalization in modern medicine
Establishment medicine is gradually warming to the idea of greater personalization and patient empowerment. Our genetic data is now at our fingertips, useful in such things as personalizing cancer treatment 1. Some of the more enlightened doctors are recommending measurement, tracking, and self-experimentation to their patients. In other words, they are encouraging their patients to become biohackers.
A great example of patients self-biohacking against a headwind of their doctors’ advice is the story of the eventual recognition of non-celiac gluten sensitivity.
“Some people eventually fail to be diagnosed with celiac disease because they don’t fit the criteria, but because they were desperate because nothing else explained their symptoms, they decide, despite the negative results, to try the diet no matter what. And some of them, sure enough, had their symptoms improved or completely resolved.
. . .So as typically happens in these situations, it was from the grassroots that the problem really became a problem, because when we saw this critical mass of people come into our clinic, at the beginning we sent them away. We said, you know, you don’t have celiac disease. You have no reason to be on a gluten-free diet. But when we saw this phenomenon to take great proportion, we asked ourselves: Is that possible that all these people are nuts? Are they all responding as a placebo effect? So we started to dig into this situation a little bit more, and sure enough, we discovered that there is another form of gluten reaction that we don’t call gluten intolerance anymore because we went through a revision of nomenclature, but we call it gluten sensitivity. And it turns out to be an immune response to gluten not on an autoimmune basis like in celiac disease, not even on an allergic basis because we know that sometimes wheat can induce an allergic reaction like any other foodstuff.” – Dr. Alessio Fasano 2
The trend towards greater personalization and proactive patient involvement seems likely to continue to grow. But there’s no reason to restrict self-experimentation to disease treatment. It is a powerful methodology for enhancing almost any aspect of your life.
Unstructured self-tracking is not enough to gain insight
The Quantified Self (QS) movement is a popular trend describing those who take data on themselves and track changes over time. Interest in QS has fueled demand for a profusion of new technologies – from activity monitors to brain-wave (EEG) detectors – for individuals to take data on themselves. I am an active proponent of the movement, having started the Zurich chapter of the meetup group in November 2012. QS empowers the individual to take control and responsibility for their health, performance, and well-being. This is in contrast to most medical systems in the world that encourage individuals to passively rely on their doctors for health evaluations, diagnosis, and prescriptions. In this sense I think that QS is a giant step in the right direction.
However, many of the QSers that I have met go only so far as taking data on themselves. It’s well-known that even this simple act of measuring can have profound influence on behavior. For example, when individuals track their activity level it typically leads to increases in it. However, when one wants to cause a change in a health or performance objective that is not as directly under our control, it’s generally insufficient to merely measure the objective and hope that it improves. Testing well-thought-out treatments is more efficient and effective. In order to truly gain insight from these tests it’s best to use a scientifically valid process, namely a properly run self-experiment. Ideally such an experiment would incorporate: repetition (of observations), randomization (of the experimental treatment), controlling (for other variables), self-blinding (you don’t know when you are receiving the treatment), and statistical analysis of the results. All of these aspects of experimental design will be discussed in more detail in future posts. Not all will be possible or practical in each case, but knowing the ideal allows us to better approximate it in reality, leading to more efficient and accurate learning of what does and does not improve our lives.
Some results of biohackers from self-experimentation
You don’t have to look very hard to find the potential fruits of biological self-experimentation. A few examples:
Me - I hacked: a repetitive-strain injury using mind-body techniques, my mindstate with meditation, my working memory with dual-N-back training, and fact acquisition with spaced repetition software (SRS), like Anki. I have also done a lot of dietary experimentation with varying results.
Dave Asprey -“He upgraded his brain by >20 IQ points, lowered his biological age, and lost 100 lbs without using calories or exercise.” http://www.bulletproofexec.com/author/daveasprey/
Seth Roberts - Hacked acne, sleep, mental acuity, weight. http://blog.sethroberts.net/
Brian Kerr - Checkout his story here: http://www.quantifiedscience.com/sample-page/
Plan your own self-experiment
The Biohacking 101 series of posts is designed to provide exactly the knowledge that you need to plan and execute self-experiments to improve your performance or health. If you already have a biohack in mind that you would like to test out (see the previous post for some inspiration), then take few minutes to think about how you might setup a self-experiment to determine whether it is successful. In future Biohacking 101 posts I will give you ideas to help you refine your approach.
This essay was originally posted in Winslow’s blog, BioHackYourself, HERE
Note: For some of the links in this article, if you use them to make a purchase, then I will earn a small commission. I only recommend products that I have used myself. Your purchases help me dedicate more time to this website.
- http://www.nytimes.com/interactive/2012/07/08/health/conventional-cancer-therapy-and-whole-genome-sequencing.html ↩
- http://chriskresser.com/pioneering-researcher-alessio-fasano-m-d-on-gluten-autoimmunity-leaky-gut ↩