As if paleo weren’t ridiculous and harmful enough, the latest fad diet book takes the trend to an extreme that includes lacing one’s morning coffee with butter and coconut oil.

The Bulletproof Diet recommends 50-60% of calories from fat (or as the author Dave Asprey puts it, “healthy fats”), 20% from protein, and the rest (as in 20-30%) from vegetables.

The idea is that since our brains require fat to function, the more fat we give them, the better they will function.

Under that line of reasoning, I now have a new method of fueling my car: Once the gas tank is full, I continue pumping gas into the oil chamber, the wiper fluid container, and the cup holders. Then I’ll top off the glove compartment and the storage unit between the seats.

Undeterred by centuries of evidence that a high carb, plant-based diet is the human default and best suited to overall health, Asprey boldly includes the “scientific proof” behind his deadly diet.

You can read about the diet and see the 40 cherry-picked references here: https://www.bulletproofexec.com/the-complete-illustrated-one-page-bulletproof-diet/

Before I show you a few of the references, I want to put that number – 40 references – in context.

I just finished helping Dr. Garth Davis write Proteinaholic, a book that comes to the opposite conclusion as The Bulletproof Diet. We ended up with 699 references, all from peer-reviewed journals.

They range from short-term lab studies to randomized clinical trials to case studies to multi-decade, large-scale epidemiological studies to comprehensive meta-analyses. And the preponderance of evidence is undeniable, unless you are constitutionally unable to do anything but deny evidence (see “there’s no such thing as climate change, and if there is, it isn’t manmade” for another example of this mental disorder).

Now, back to Asprey’s comprehensive list of 40 references designed to overwhelm you with the truth of his position. Let's take a look at three of them – a representative sample of mischief and misrepresentation.

Reference #13

Wow, that’s shocking and alarming. Let’s read the actual study.

Before we even look at the results, we’re staring Asprey’s sloppy approach to fact in the face: the study didn’t look at vegan or vegetarian children, but those fed a macrobiotic diet up to the age of 6. Not vegan. Macrobiotic.

Isn't that the same thing, though? The above abstract does describe the macrobiotic diets as “vegan type.” Are they?

As you can see from a description of the diets of the three groups (macrobiotic kids with low cobalamin (B12), macrobiotic kids with normal cobalamin, and the control group of omnivores, the supposed “vegans” got 3% of their calories from animal protein. Back of the napkin calculation suggests that they were getting somewhere between 6-10% of total calories from animals; hardly the vegan diet Asprey is attempting to disparage.

Second, the number of study participants is tiny. At best, this study is suggestive, and you could make the argument that larger studies should be done to examine the issue. The problem of small numbers is made much worse by one of the oldest statistical tricks in the book: measuring so many outcome variables that something is bound to come up as significant.

Here's a list of the cognitive tests conducted on the children.

The only statistically significant result? The macrobiotic children with normal B12 levels underperformed compared to the omnivorious control group on a test of picture completion.

In essence, we're looking at entirely random data distribution. In Word fluency (K), for example, the low B12 group did much better than the omnivorous controls. In Word fluency (A), by contrast, the macrobiotic group with normal B12 did the worst. And in most of the tests, there was no significant difference.

Recently, a Harvard researcher demonstrated the misleading power of a small study with lots of variables when he “proved” that chocolate contributes to weight loss. Read his story of how he fooled millions with bad science here.

Third, there's no description in the article about how the researchers chose the control group. This is very important, because one way you can manipulate the results of a study is to bias the groups. For example, drug companies frequently put “perfect patients” in the groups receiving their drugs by setting inclusion criteria that exclude anyone whom they think would respond poorly.

Fourth, we have no idea what the different groups actually ate. Dietary recall is notoriously inaccurate even for last week, let alone for dietary patterns followed years in the past. Furthermore, it's not uncommon for participants to “improve” upon their diets in the telling to impress the researchers.

In summary, does this study support Asprey's claim that “kids who eat a vegan diet are deficient in B12 and have impaired brain function” which reverses when they start eating animal products?

In the immortal words of Wallace Shawn's Vizzini in The Princess Bride: “Not remotely.”

Reference #11

Another reference with a shocking statistic:

50%? The researchers must have conducted a large-scale representative sample, using randomization and other statistical best practices, for Asprey to so characterize the results of their inquiry. Right?

Well, sadly and predictably, no. Not exactly.

Check out the actual numbers of study participants:

Is it possible to generalize to millions of people from a study of 88 subjects?

No, not even if they are selected randomly from the population. A tiny case-control study such as this one can be used to suggest that vegans may have lower B12 levels, but certainly can't be enlisted to state the absolute percentage of B12 deficiency in all vegetarians.

And this study says not a word about how the participants were chosen, so we have no idea how representative they are, and of whom.

In the discussion section of the paper, the authors spend a great deal of time lamenting the dangerous and deficient vegetarian diet. Then, puzzled, they toss in this incidental comment:

Meaning, the vegetarian diet looks bad in our tiny little study, but people who adhere to it tend to live longer.

Hmmm…

Reference #8

OK, one more study that Asprey claims nails the lid on the coffin of plant-based diets. Did you realize that brown rice is a silent killer?

Before we look at the study upon which that startling “fact” is based, stop and consider what type of study could provide this information. What are you expecting to see?

I’ll bet it wasn’t the following (unless you’re as cynical as I am):

Five young men. Eating pretty much nothing but rice. 14 days of white rice and 8 days of brown rice.

Using that logic, we should not drink water because drowning.

This is the dictionary definition of a straw man argument. In this case, it's rice straw.

Summary

I’m not writing this simply to pick on The Bulletproof Diet (which my son was disappointed did not include kevlar), but rather to highlight the blurred line between deceptive marketing and real scientific evidence.

To summarize the techniques employed to confuse us:

1. Straw man
2. Small sample size
3. Too many variables
4. Misrepresentation of data
5. Illogical conclusions

Also, most of Asprey's references are years or decades old, which means they are exempt from modern reporting requirements of funding and potential conflicts of interest. So I can't say if the studies were funded by industries with a horse in the race.

Speaking of money, I didn’t have to pay for any of the abstracts I found online. They were all freely available. I did need a university affiliation to read the full article for reference #11, but the rest weren't blocked by paywalls.

Which means that there don’t have to be barriers between you and the truth. Simple curiosity and a little bit of time can save you from many dangerous alleys disguised as highways to health.

Bulletproof References

Cognitively impaired vegan kids: http://www.ncbi.nlm.nih.gov/pubmed/10966896

50% of vegetarians are B12 deficient: http://www.ncbi.nlm.nih.gov/pubmed/12011576

Brown rice: http://www.ncbi.nlm.nih.gov/pubmed/2822877