EmbodyDNA and the False Hope of Personalized Weight Loss

by Brian Rigby, MS, CISSN

3 Replies


Have you heard of EmbodyDNA yet? It’s been making some rounds on the internet. The basic idea is that it’s genetic testing specifically aimed at nutrition and weight loss, and it promises to be able to help you “fine tune” your dietary plan to lose weight (or be healthier, perhaps). Sounds great, but there’s a problem: it won’t tell you anything particularly helpful. Let’s break it down.

Genetic Testing & Weight Loss

The biggest premise is that genetic testing will help a person lose weight easier, but how exactly is a bit of a mystery. The test can tell you if you are more prone to having a high body mass index, for example—presumably by identifying “thrifty genes” or their ilk—but once known about, nothing changes; you cannot turn off genes you don’t like. Is this helpful information for those seeking to lose weight? To learn that everything might be a bit harder? I would argue not, and could actually have a negative effect because simply identifying a single such gene tells us very little about a whole person’s predisposition to higher weight (which will be based on numerous other genes, social, and environmental factors) but could very really discourage someone into believing it’s just their lot in life.

EmbodyDNA also purports to be able to tell you about your individual macro- and micronutrient needs (here it’s a bit fuzzy and I could be wrong, but that appears to be their intention). If this is indeed their intention, it seems exceedingly unlikely they would be able to do more than give you the barest of hints about how much of any particular nutrient you need. For certain nutrients, like folate, there are known genes that affect need (the methyltetrahydrofolate reductase TT polymorphism), but for most nutrients all we can do is measure a spectrum of need society-wide because there’s no single, easily identifiable way to measure individual need. That’s a big reason why we use the RDA, which is set high enough to cover the needs of 97.5% of people—97.49% of those people will need less than the RDA to remain healthy and 2.5% will need more.

Genetic Testing, Tastes, and Sensitivities

These topics seem more plausible, but if anything, I think they actually overreach even further. Let’s start with tastes.

There are a few identifiable genes associated with taste, such as TAS2R38 which has several variations that either allow or disallow the carrier from tasting certain bitter compounds in foods. Presumably, you could identify this gene and many more like it and draw a map of what flavors you are more or less sensitive to. The problem, however, is that this only tells you what you can and cannot taste, not what you prefer.

Taste is extraordinarily cultural. Foods like kimchi which are devoured in Korea cause the gag reflex to those not accustomed to them; bitter vegetables can be either savored for their complex taste or rejected as disgusting. Much more goes into what we eat than simply being able to taste certain chemicals in food (though that information does illustrate how different foods may taste even in a single culture to different people).

The case of “sensitivities” may be a bit more interesting, but I wish EmbodyDNA had used a more appropriate term. It claims to (soon) be able to identify lactose tolerance and gluten tolerance—the first is rightly called “tolerance”, the second relates to the likelihood that a person has celiac disease. In neither case are we discussing “sensitivities”, which is a catchall term that is used to describe when a food is subjectively thought to cause harm in the absence of objective evidence (such as blood or stool markers). There are genes associated with both lactose intolerance and celiac disease, so these may guide someone in the right direction, but the risk is (again) that it’s really too little information to draw a conclusion from.

For example, 98% of celiac patients have a gene that encodes for a specific serotype (HLA-DQ2), but the reverse is not true as well because 55% of the population also carries those genes. In other words, if you already know you’re celiac you can be almost certain you’re a carrier for that gene, but if you only know you’re a carrier of that gene you can’t know whether or not you’re celiac. In a perfect world, of course, this is still great information because you’ll go see a doctor and get further testing (assuming you have at least some semblance of symptoms)—but more often than not, I suspect, the person who finds out they’re a carrier will simply give up gluten without further investigation.


It’s (Almost) Never the Details

Humans tend to zero in on the details without paying much mind to the foundations. EmbodyDNA is a great example of this. For people struggling with weight loss, getting some individualized recommendations based on their own genetics sounds like a magical solution to solve all their problems. The unfortunate reality, however, is that fine-tuning only works when a plan is 95% good—and even then, only minimally.

All good weight loss diets start with the same foundations of caloric restriction, increased activity, reduced junk intake, increased whole food intake, and social support. This isn’t to say any of it is easy, but rather that these are the necessary components that when present will lead to reduced weight. If you’re not losing weight (and there is weight to lose), you can’t change the fine details to compensate for the lack of a solid foundation.

For the price of $190.00, there isn’t enough to EmbodyDNA to make it worthwhile. It can only inform you about some of the fine details, and even then it can’t tell you much more than your predispositions (which may or may not be accurate past the genetic level). If you need help losing weight, focus on the foundation first—the fine details will resolve themselves as you progress.


  1. Diane

    Your posting about EmbodyDNA is a comprehensive, well-written analysis and evaluation. There is no magic-bullet for losing weight, but I suppose, like Ponce de Leon, humans always search for their “fountain” …whatever that might be.

    A further concern I have is the potential for genetic data to be used for nefarious gains. Without any regulation, their promises are as empty as the vacuum of space.

  2. nick emery

    Great post. I can’t see how EmbodyDNA can claim to identify specific genes responsible for dietary needs of nutrition. I’ve just finished reading a great book by Robert Sapolsky (Behave; The Biology of Humans at our Best and Worst). He talks of genonewide association studies (GWAS). I assume EmbodyDNA would use a similar techniques? In essence these are little more than fishing expeditions. In one study looking at specific genes for height 183K + people had there genome examined and 200 + genes identified as being partly responsible for height and the one gene with the great significance explained 0.4% of variation in height.

    In a 2013 study on academic attainment 126K subjects were tested and the most predictive gene found accounted for 0.02% variance. As one of the authors stated ‘In short, educational attainment looks to be a very polygenic trait’.

    As Sapolsky states there is a gene environment interaction, looking at genes in isolation doesn’t answer all the questions and to assert that diet can be determined by genetics is misleading at best.

  3. Brian Rigby, MS, CISSN Post author

    I would suspect it’s even less predictive, except in the few cases (such as for folate) where we know there is at least one polymorphism that is strong. Beyond those cases, I don’t think we know enough about dietary need on an individual-to-individual basis to tie genes to it—in other words, we can use GWAS to associate certain genes with height because height is objective and easily measured, but if we took the same 183K people and tried to tie certain genes to dietary needs we’d be in the dark because we wouldn’t know which of those individuals have greater or lesser needs for a certain nutrient without further, expensive, in-depth testing. And then, we’d probably still have the same problem you mentioned of there being no single gene responsible, but rather a complex interaction between genes that is far harder to use predictively. Thanks for the comment!

Leave a Reply

Your email address will not be published.