A few days ago, I came across this podcast and thought “Maaan people need to hear this”. Almost everything you’re about to read can be listened to (and probably better explained) here, but just in case you’d like the short version of what I thought was a very important thing to understand, stay with me! I am just going to focus on a small part of the podcast, which talks about studies in the nutrition field. The whole podcast delivers amazing information and covers the big and notorious debates : Vegan VS Omnivore. I won’t cover that part, but I really invite you to listen to it, with a pen and a notebook 😉 (in my case I had only post-its handy. My whole room got redecorated in a way that would give an interior designer a heart attack (Hi Francis, if you happen to read this! :-)).
Being a science enthusiast, I used to think that something is either good or bad for you, and could never really understand why they were so many debates around nutrition. To me, trying to wrap my head around why doctors/nutritionists are claiming something to be good for us, and other ones claiming it to be bad, doesn’t make any rational sense. Turns out, thanks to this podcast I understood a bit better how conclusions can be so off and/or unclear.
I bet you are familiar with : “I read it ‘somewhere on the internet’, it must be right then!”. I think it’s pretty safe to say that 99% of the population knows that’s not true. However, when we’re reading something on a website that looks legit, and that’s backed up by a documentary we might have seen on Netflix and by one of our friends who’s a doctor, chances are, we’re going to believe it’s true. And yet, sometimes, we couldn’t be further away from the truth! Not because we are stupid and believe everything we read, but just because, despite good intentions, the truth is well hidden and requires us to dig deeper.
Let’s break it down a little bit so hopefully at the end of this article, you’ll understand this better than quantum physics.
In science, in order to draw any conclusions from an experience, it needs to be:
- Observable
- Measurable
- Repeatable
- Falsifiable
That’s basic science that we learn in high school (remember when we didn’t have smartphones and we were actually listening?)! When conducting an experiment, (dig into your high school memories for me), we always have to isolate what we want to study, and change only one variable at a time. Are you with me so far?
This is why in medicine, we call the Randomized Control Trials (RCT’s), the “Gold Standard Study”. It checks all the boxes. RCT’s divides (randomly) a target study group into two (or more); one group gets the control variable being assessed, the other one the gets a placebo, and all other conditions are carefully controlled and monitored. At the end of the trial you observe and draw conclusions. Easy peasy.
Easy peasy, but not so much in nutrition. If we want to see if there is a correlation between red meat and colon cancer, getting two groups of people, one eating only fruits and veggies and the other one eating the same but adding red meat, and keeping this going for years… Let’s just say it would be easier to get my dad and his friends talking about politics and remain calm. And that’s not even talking about the fact that those people would have to have the exact same lifestyle (exercise, sleep, work, social activities,…). Sounds like a bit of a challenge to me.
On top of it, there is also an ethical dimension : If we wanted to prove the link between smoking cigarettes and lung cancers in a Randomized Control Trial, that would probably be comparable to a remake of concentration camps => not the best idea! Having said that, some studies in nutrition can be done with RCT’s.
Now that this is cleared up, it leaves us with 80% of all nutritional studies being observational studies. What does this mean? Exactly what it says : you observe, and from the observation, you draw conclusions! There are four basic types of those described here . Observational studies are used most of the time when it comes to nutrition, but there are three major problems with them. I will use the example of the consumption of red meat and colon cancer to illustrate my point (actually to be 100% accurate, it was processed red meat).
1. Data Collection
When conducting a study, the data we is going to collect are going to be as good/bad as the tools we used to collect them. In other words : you get what you pay for!
Often times, the main tool used to collect data is a basic questionnaire. As far as I’m concerned, I got a little scared learning that most of our nutrition research (and therefore conclusions) relies on the same memory as some guys I dated that couldn’t even remember my birthday (maybe this is another debate). But you get the point! If you get the hiccups now and I’ll ask you what you ate last Thursday, you’ll think for 3 minutes, start laughing nervously because I’m still waiting for an answer, and finally say “I have no idea how it is possible, but I don’t remember”, with a little blushing smile (but you will notice that your hiccup went away ;-)). And yet, 80% of all our research in nutrition is based on what people’s remember of what they ate for the last few days/weeks/months. S-c-a-r-y.
On top of that, people tend to lie and underreport what they are actually eating. There was a 39 year-long study (Nurses’ Health Study), that reported that the amount of calories consumed per day by the participants wouldn’t keep a small, frail, elderly woman alive. Yet, most of the people in the study were either overweight or obese. Seeing any red flags yet?
Basically, it is based on shaky memories and/or lies (I’m starting to think that environmental studies could be compared to a bad dating situation).
2. Healthy User Bias
This states that if you are engaging in a behavior that is perceived as unhealthy, you’re more likely to engage in other unhealthy behaviors. With our example of red meat consumption (perceived as unhealthy), it has been observed that people who eat a lot of processed red meat tend to eat a lot of sugar, processed food, to smoke, drink alcohol, not exercise, and the whole “How To Die Younger” package.
The opposite is valid as well, people who engage in perceived healthier habits (aka no red meat), tend to have a healthier lifestyle and avoid processed foods altogether.
3. “The relative risks in nutrition are so low that they are indistinguishable from chance.” Chris Kresser
Yes, I know what you’re thinking, this is Chinese, but in English, it means that the percentage of increased chances of risk is almost non-conclusive. For instance, in this particular study for red meat consumption, they found that the risk of getting colon cancer increased by 18%. This translates into :
“4,5 people out of a hundred were getting colon cancer without eating processed red meat to 5,3 people out of a hundred getting colon cancer if they were eating it.”
To put things in perspective, there is a 3000% increase for the link between smoking cigarettes and lung cancer. Any medical papers require at least a 200% increase to be published. And yet in nutrition we’re allowed to throw red meat under the bus for an 18% increase? I was never the best at math, but I’d challenge anyone who thinks those numbers add up.
There you have it! Hopefully you learned something new today, so next time someone tells you “this or that causes this or that disease”, you can take that info with a grain of salt.
Don’t get me wrong, we need observational studies, but not all of them aren’t worth taking a look at. First of all because as stated before, we can’t have RCT’s on everything when it comes to nutrition, and furthermore, some of the observational studies are conducted in a much more intelligent way. It’s all about identifying the difference between the ones that have been done to avoid the healthy user bias as much as possible, and interpreting the final data. Remember, medical papers and scientific reviews do not publish anything under a 200% increase in risk. Anything under this number is highly questionable.
A last word on this, studying nutrition has taught me that even though this is a science, nothing is black and white. Keep in mind that everyone is different. The way our bodies react to food can drastically change from one person to another. Let’s take the very simple example of a vegan diet. I won’t get into the big debate vegan VS meat eaters, but I will say that someone might literally thrive on a vegan diet, and someone else might be miserable. That only applies to real food though, when it comes to processed crap, science is really black and white ;-).
Additional resources:
If you liked this article, you might find those very valuable as well:
https://chriskresser.com/why-you-should-be-skeptical-of-the-latest-nutrition-headlines-part-1/
https://chriskresser.com/how-to-read-and-understand-scientific-research/
http://cisncancer.org/research/how_cancer_is_studied/epidemiological/study_types.html
https://academic.oup.com/advances/article/5/1/7/4558018
https://chriskresser.com/what-the-is-wrong-with-what-the-health/
https://www.vox.com/2015/1/5/7482871/types-of-study-design
https://kresserinstitute.com/the-fundamental-problem-with-most-nutrition-research/
The Belgian Palate 
1 thought on “Trust Science, Not Opinions.”