Chapter 1: Food Should Make You Healthy

Covering this chapter will be somewhat brief as there was only 1 citation. The rest of the chapter was a short intro to the program and the story of the Hartwig’s journey back to health using the Whole30. There are many anecdotes littered throughout the book that I will not be covering as they are irrelevant to reviewing the science behind It Starts With Food. Anecdotes can be a powerful mechanism for persuading people but as the saying goes, the plural of anecdote is not data.


“The food you eat either makes you more healthy or less healthy. Those are your options.”

In the foods they consider to be "less healthy" we have alcohol, added sugars, sweeteners, grains, legumes, seed oils, dairy, and select food additives (MSG, sulfites, and carrageenan - reviewed here). At first glance, this seems to be pretty black and white, disregarding the effect that dose will play. Take water for example. Everyone knows water is "more healthy" when ingested in moderate doses and there is a clear benefit (like staying alive). But if you were to chug 6+ liters at once you begin to see signs of harm like dilutional hyponatremia or death [2-4]. I’d say death is pretty harmful. Since future chapters take a look at each of these foods in depth, that’s all I'll say about that for now. 


“We are going to teach you how to turn yourself into a scientific experiment of one, so you can figure out for yourself, once and for all, whether the foods you are eating are making you more healthy or less healthy. And that’s worth more than any scientific findings you read about—because there hasn't been a single scientific experiment that includes you as a subject.”

The citation provided for this paragraph is a link to a blog post by Epistemocrat, a medical student at The Johns Hopkins University School of Medicine and a fan of the Paleo Diet [1]. In his post, he talks about applying Nassim Taleb’s Black Swan logic to n=1 health such as creating meta-rules for eating like “Don't eat anything that causes a negative physiological reaction”. This seems all well and good (if not a bit too black and white), but Epistemocrat goes on to apply this rule in a way that demonstrates a shocking oblivity to basic chemistry concepts for someone who is a med student. 


“I don't need to test HFCS again on my own body--don't want to eat something that is one chemical step away from plastic and then watch my body produce alien tissues as a result (our metabolisms simply can't handle that stuff).”


Two issues here. Preying on people’s chemophobia and an ignorance of metabolism. Let’s tackle the chemophobia first. I don't know if HFCS is one chemical step away from plastic. I don't care and neither should you. Here’s why. 

Sodium explodes when you put it in water [5]. 

Chlorine, an important ingredient in mustard gas, is poisonous [6,7].

Combine the right forms of those together and they form the deadly substance known as sodium chloride, otherwise known as table salt. 

You see what I'm getting at here? Saying HFCS is one step away from plastic is just a fear-mongering ploy that is used to prey on the general public and their lack of knowledge about chemistry. 

The second issue is the metabolism of fructose from “natural” versus “unnatural” sources. Let’s play a game. On the left, we have some fructose from an orange. On the right, we have some from a Coca-Cola sweetened with HFCS. Can you tell the difference?

No? Neither can your body (Hint: There is no difference). It does not matter what the source of fructose is, your body is going to digest it the exact same way [8]. I would love to see some data, any data, which demonstrates fructose from HFCS is magically diverted to an alternative metabolic pathway that will probably give you diabetes and makes you infertile because your body just knows that you have ingested some unholy HFCS. I suspect I’ll be waiting a while. Or I'll become an infertile diabetic. I'll let you know. 

If Epistemocrat's blog post is representative of the quality of sources ISWF will be citing in the rest of the book we're going to have a long journey ahead of us. 

Getting back to Whole30: 

“We are going to teach you how to turn yourself into a scientific experiment of one.”

They even have a pyramid for this, indicating that the science should be layered with clinical experience that should culminate in self-experimentation which should be your primary informer.

Not to be outdone, I've gone and made my own pyramid. Mine looks a little different, though.

On the Pyramid of Evidence, the n=1 or case report provides very low-quality information. 

I don't really have a problem with attempting a n=1 trial. Self-experimentation can be a useful tool but the experiment needs to be well designed and you need to be aware of the plethora of potential sources of bias, namely:

  • Confirmation bias
  • Expectation bias
  • Wishful thinking
  • Self-fulfilling prophecy
  • Confusing correlation with causation
  • Regression to the mean
  • Selective memory
    • Selective perception
    • Selective retention
  • Multiple Treatments
  • Subjective Outcome Measures
  • The Placebo/Nocebo Effect
  • Type I Errors (False Positive)
  • Type II Errors (False Negative)

Randomized controlled trials can be susceptible to some of these too, but RCT’s usually have better mechanisms in place to prevent bias from seeping in, like using a control group. 

One of the benefits touted by ISWF is that by using their diet plan, you will be able to identify some food allergies and sensitivities through your n=1 trial. Ideally, when attempting to figure out food sensitivities or allergies, you would perform a Double-Blind, Placebo-Controlled Food Challenge (DBPCFC), the gold standard in allergy diagnostics [9-11]. In the DBPCFC, the patient would receive an increasing dose of either the suspected allergen or a placebo. The delivery method of the allergen and placebo would have to be indistinguishable from each other, which can be tricky. Most people will not be able to do this in a real world setting, so to ensure an accurate result you will need to remove and re-add the food back into your diet multiple times. Here is what an n-of-1 trial design might look like over the course of 8 days [12].

It will be tough to pull off and time-consuming but to help eliminate potential sources of bias from a trial method that is highly susceptible to it, it’s your best bet. 

For more information than you ever wanted to know on how to set up n-of-1 trials, go nerd out at the below link from the US Department of Health & Human Services [13].