Can DNA Testing Offer Personalized Diet Advice? Nope, Not Yet.

Related articles

A lot has been written about the strengths and weaknesses of using DNA testing to customize individual diets. It's a promising idea, but our knowledge of genetics isn't yet good enough to pinpoint what each of us should eat.

You're familiar with the phrase "diet wars." Some people swear low-carb is the way to go. Others say such hi-fat programs will kill you; what you really need is a vegetarian diet. Actually, still other commentators claim, eat whatever you like, just cut your caloric intake, and you'll be fine. After following this ongoing fight with great interest for years, I got bored and came to the underwhelming conclusion that individuals respond differently to various dieting methods; what's right for me may not be suitable for you. Research has shown that genetic variation is partially responsible for the divergent outcomes people experience on a given diet. 

DNA testing companies have turned this observation into a business that generates almost $200 million annually by helping customers identify what they should and shouldn't eat. One apparently popular advocate of this approach calls herself the “DNA Dietician”:

The DNA Dietitian has helped hundreds of clients become the healthiest version of themselves by providing them with 100 per cent personalized diet and fitness recommendations formulated specifically for their DNA. Her proven method – ‘Eating The DNA Way’ – ensures you understand exactly why and what you need to eat to be your healthiest self.

DNA-based dieting has been widely critiqued in the last few years, so I dug through the literature and spoke to an expert to see where things stand in 2021. For now, it appears that our knowledge of the human genome is too crude to reveal customizable weight-loss plans. Like so many varieties of internet tomfoolery, “Eating The DNA Way” mixes a little bit of science with a whole lot of marketing. The result is an overpriced and underdeveloped product that may or may not actually work as advertised.

What do genes really tell us?

The fundamental problem is this: while our genes undoubtedly influence how we respond to different nutrients, DNA testing only gives us a statistical correlation between a gene variant and a particular outcome; it can't (yet) tell us with much certainty how particular mutations influence our individual responses to different nutrients.

There are some exceptions. For example, the inability to metabolize the sugar galactose is controlled by mutations in the GALTGALK1, and GALE genes; individuals with these variants can thus be easily diagnosed with galactosemia. A similar case could be made for caffeine insensitivity and some other food tolerances. But things are typically more complicated, as molecular geneticist Mariam Sajid noted earlier this year:

Many companies focus on macronutrients such as metabolism of lipids and carbohydrates. For that they use FTO (fat mass and obesity associated) genetic variants. The FTO gene variant is universally considered as one of the variants that has an association with our BMI. However, outside of this one gene, nearly 900 other gene variants also exist that have a phenotypic association with BMI. One identified variant (or even a few) cannot explain a person’s genetic predisposition to weight loss.

There's even more to it than that. Consider this February 2020 study (the source of the 900 variants figure) of more than 40 firms that offer nutrition-related DNA testing. Roughly one-third of the companies in the study offered an analysis of FTO gene variants, but there's no evidence that individuals with these variants should follow a particular diet. Moreover, the best-studied of these mutations, rs9939609, accounts for less than one percent of the variation in body fat percentage and BMI among Europeans.

Research has also found that calorie restriction and exercise appear to modulate the effects of FTO gene variants, which leads us to an important observation: if following standard weight-loss advice can limit the expression of obesity-promoting gene variants, an expensive test that tells consumers they carry those variants isn't very helpful. The authors of a November 2020 review summed up the situation this way:

It can be concluded here that although genetic factors predispose to increased risk of obesity the environmental, behavioral, socioeconomic, psychological factors that contribute to overall life style determine the overall phenotype of an individual. Physical activity in this regard plays a vital role. Several studies have shown the importance of diet coupled with physical activity in maintaining a healthy weight.

DNA-based diets don't work

Given our discussion thus far, it seems that diets based on your genetics shouldn't yield especially impressive results, and indeed they don't. Some research has shown that individuals with specific gene variants may lose slightly more weight following a low-fat instead of a high-fat diet. However, the results weren't statistically significant, and subsequent studies haven't replicated them.

In a well-known 2018 study, The DIETFITS Randomized Clinical Trial, all individuals assigned to a healthy low-fat (HLF) diet or healthy low-carbohydrate (HLC) diet for a year lost weight (11.68 lbs for the HLF group vs.13.22 lbs for the HLC group). But there was no significant difference in individuals matched to either diet based on their genotype pattern. The authors concluded:

The finding of no significant difference in weight loss in genotype-matched vs mismatched groups in the current study highlights the importance of conducting large, appropriately powered trials such as DIETFITS for validating early exploratory analyses. Analyses of all the genomic data obtained are under way to evaluate whether other genetic signatures may demonstrate effect modification.


Some people will insist that DNA-based dieting changed their lives. Perhaps it did, but I suspect their success can be attributed to the benefits of personalized nutritional advice often sold alongside genetic testing as a package deal. Weight-loss programs based on an individual's food preferences and activity level are generally more effective than the standard “eat less, move more” recommendations. Adding genotype information to these customized programs appears to make little difference.

Biologist Dr. Josh Rappoport, author of Mapping Humanity: How Modern Genetics Is Changing Criminal Justice, Personalized Medicine, and Our Identities, helpfully summed up the science in an email to ACSH:

The world of metabolism, even just at the level of macronutrients, is really complicated, and while there are dietary and analytical observations that do seem to hold up - for example dietary omega-6 vs. omega-3 fatty acids or serum HDL vs. LDL cholesterol levels - whether subtle impacts of gene variants are actionable in the individual is an entirely other matter. 

Furthermore, a great deal of the data that went into identifying the statistically significant trends correlating specific gene variants with different phenotypes in large populations comes from generally very homogeneous (i.e., mostly white European) sampling.

If you still want to spend a few hundred bucks on DNA diet testing, knock yourself out. Just know that you paid for advice your grandmother probably would have given you for free.