Pocket worthyStories to fuel your mind

Let’s All Stop Holding Out for Science to Find the Perfect Diet

Nutrition science’s next project is personalized diets based on genetics. It sounds good, and sure, it might work, but don’t hold your breath.


Read when you’ve got time to spare.


Photo illustration by Slate. Photo by Milkos/iStock/Getty Images Plus.

Nutrition advice is often all over the map, even contradictory: Red wine is good, all alcohol is bad, eat breakfast, skip breakfast, eat a million small meals, go vegetarian, eat lots of meat. One explanation for why it’s all so confusing? Maybe there is no right diet for everyone. Maybe the best diet is slightly different for everyone, dependent on a combination of our DNA, lifestyles, and the microbes hanging out in our guts. The science of how we each individually process and respond to food is just getting going. And it’s being turned into a consumer product almost as quickly as it’s being produced.

That we’d all be healthier following food plans created for us as individuals is the premise behind Zoe, a company founded by Tim Spector, a professor at King’s College London, and a pair of entrepreneurs who partnered with the scientist to help the company sell the results. Zoe is getting attention thanks to the cutting-edge research it’s funded: In June 2019, Spector presented preliminary results from “the world’s largest and most comprehensive experiment to look at individual responses to food,” according to a write-up by Cynthia Graber and Nicola Twilley (who themselves participated in the study) in the New York Times. Other researchers say it’s important work: Geneticist Eric Topol told the Times that it’s a “major milestone” toward personalized nutrition, which Amy Miskimon Goss, a researcher at the University of Alabama at Birmingham, told me “is the holy grail.”

While surely lots of useful medical insights and new research avenues will tumble out, as often happens when you collect mountains of data, the end goal for Spector’s Zoe-funded work is a consumer app that can tell you, as in you specifically, exactly what your body wants to consume. “Eat the way your body loves,” says Zoe’s website. There’s a mock-up of the app that shows how it will help you rate the utility of “your breakfasts,” giving health scores to your options: fat-free yogurt might get an orange 3.5, avocado toast a green 7.1. Spector can’t say that much more about how Zoe might work— “we can’t be specific about the product, because it’s going to be driven by the science” and that science isn’t done yet—though he envisions recommendations as fine-grain as what time of day to eat and even if you’d be better off relaxing with a can of beer or a glass of wine based on how your body responds to hops versus grapes. A press person for the company told me that the goal was to launch the consumer product in 2020, though that’s contingent on how the research progresses.

I’m intrigued by what experts see as genuinely exciting research, but skeptical about what impact Zoe will have on the average user’s life. We are likely to spend many of our days on earth being told that the perfect diet is on the other side of the swipe of a credit card, but you shouldn’t spend too much time thinking personalized diets are about to change everything—that’s unlikely.

The premise of the study, called Predict, is this: Over a thousand participants ate a series of carefully logged meals while regularly forking over bodily waste, information on sleep and stress, and samples of blood. Many of the participants were twins, which will allow researchers to also suss out the role that genetics play in how we respond to food. Spector is specifically interested in how gut microbes—which can vary even between twins—that process the stuff we eat affect our health. The results are preliminary but, on the surface, striking: The researchers found that nutrition labels could account for less than half of how subjects’ blood sugar, fat levels, and insulin increased after meals—factors that, Zoe notes, are linked to things like weight gain and heart disease. One cool take-home message is that counting calories might not be all that helpful for maintaining health.

But as Twilley and Graber point out, there’s not yet clear evidence that the findings could be implemented as advice that could make the average person meaningfully healthier. What if, for example, it’s simply hard for some people to eat in a way that doesn’t spike their blood sugar, no matter what they eat? Zoe might give good advice, but the evidence that it will be better than following the direction of a nutritionist or spending some time tracking how you feel in a food diary—tools that are available right now—isn’t yet there. It’s much more likely that any advances “are going to be around the margins,” argues Traci Mann, who runs an eating laboratory at the University of Minnesota. If this research turns out to assist most people in losing and keep off weight (which, it’s worth remembering, is not synonymous with health), that would be flat-out astonishing. “People’s bodies have a weight that they tend to defend,” says Mann. Even with cutting-edge advice, chances are “your body is going to veer back to that range.”

The reality that it can be near impossible to lose weight is perhaps exactly why personalized nutrition sounds so alluring. It’s also part of the reason Zoe could take off whether or not the science turns out to be all that useful to the masses. Zoe’s offering will join other services that already provide personalized recommendations based on your microbiome, like DayTwo (based on a smaller, narrower nutrition study than Zoe’s) and Viome (which just boasts about its microbe DNA sequencing prowess). A man who tried the latter wrote on Medium that much of the advice he received from Viome was either common sense (avoid beer and white flour), counter to his preferences (eat apples, which he dislikes; avoid pears, which he does like), or just a pain to follow (cook “baked eggplant parmigiana crusted with walnuts and garlic,” which he did once before realizing “I don’t have time to cook so I never made it again.”) It’s perhaps this last example that puts the endeavor of “personalized nutrition” into perspective. Zoe may be based on more sophisticated science, but it’s hard to see how ever more accurate advice could skirt the basic reality that eating well, especially if you are strapped for time or cash, is simply challenging.

Like countless diet plans before it, Zoe will likely not offer a guaranteed way for us to finally become our best, thinnest selves. That anything could is a central myth of our culture, one that is wildly good at selling things like … diet apps. There is no magic bullet for nutrition or health, just a set of recommendations that we are continuously refining, incrementally, slowly, and slightly. It’s not something science should be leaned upon to deliver, even though that dream is, in part, precisely what’s powering this science. We’d probably be happier tossing the idea of a holy grail, and just doing as well as we can with the guidance we have already. Eventually, Spector’s research might be part of that. But just a part.

Shannon Palus is a staff writer for Slate.

How was it? Save stories you love and never lose them.

Logo for Slate

This post originally appeared on Slate and was published June 18, 2019. This article is republished here with permission.

Want more Slate in your life?

Get the daily newsletter.