And why it's probably not worth the trouble.
The basic argument is simple: If you run two miles, you'll burn somewhere in the neighborhood of 250 calories. If you go home and eat something from a so-called "bad diet," like a cup of rocky road, you'll take in 300 calories—meaning you're 50 calories over. You can't, conventional wisdom says, easily outrun the amount you can chow down.
There are also plenty of studies that show trying to lose weight without dieting is a losing strategy: In a year-long study published in 2011, those who dieted without exercise lost 10 percent of their body weight, while those who exercised without changing their nutrition lost just 1 percent of their body weight. Summarizing a bunch of research, the Mayo Clinic has said "most studies have demonstrated no or modest weight loss with exercise alone … an exercise regimen is unlikely to result in short-term weight loss beyond what is achieved with dietary change."
But in many of these studies, the amount and type of exercise prescribed to the participants is very basic: In one classic study, for instance, dieters were instructed to walk or jog just two or three times per week for 30 to 45 minutes. James Heathers, a postdoctoral fellow at Northeastern, recently took issue with protocols like these.
"Basically, [they] range between a gussied-up version of 'go for a walk' to 'come on, shift yourself about a bit, I guess.' They have no skill-development component, no increase in work capacity to match progress, no anaerobic work of any kind, no mobility or movement pattern work of any kind, nothing," he wrote in a Medium post. "The idea that this is a 'normative' amount of exercise is a total failure to communicate the correct expectations."
That doesn't mean Heathers thinks "out-training a bad diet" is a slam dunk, or even an easy concept to define: "You need parameters on what 'out-train' is," he wrote. "And then we need to agree on what makes a diet 'bad.'"
For most people, Heathers suggests, out-training a bad diet probably means maintaining low levels of subcutaneous white fat—looking "lean," basically—while eating copious amounts of whatever they want. One of the most egregious examples of "whatever they want" is swimmer Michael Phelps, who famously claimed to eat 12,000 calories per day while preparing for the Beijing Olympics.
But it wasn't the food intake, which Phelps later admitted was exaggerated, that impressed Heathers. Phelps was training a hell of a lot—reportedly up to 50 miles per week.
And unfortunately, that's also the amount of effort it seems to take for mere mortals. In 2015, Spencer Nadolsky, a doctor practicing family medicine and obesity medicine in Olney, Maryland, transitioned from training for a bodybuilding competition to training for a triathlon. To do so, Nadolsky went from "pretty much starving myself" to eating whatever he wanted, including a nightly bowl of ice cream. He gained 8 pounds, but from the looks of his progress pictures, almost none of it was fat—the doctor still had a six-pack, and saw increases in his cardiovascular fitness and speed.
But like Phelps, he also trained a lot: six or seven days a week, averaging seven or eight total hours. On an average Friday, he says, he would do upper-body lifts for an hour, then swim for an hour.
"It was so much training. It took so much time," he says. But he did eat ice cream almost every day and took down, by his estimates, 3,000 to 4,000 calories daily. Aside from a slight uptick in the size of his LDL particles, a marker for cholesterol risk, his bloodwork, weight, and fitness all improved. Still, Nadolsky doesn't recommend this to his patients.
"You can't just go out and work out like a fiend. You have to work your way up to it," he says. For Nadolsky's obesity clients, this amount of exercise is unrealistic when they're starting out. But for the doctor, a former D-I wrestler who just finished an intense cycle of training to compete in bodybuilding, it was within range. "I already had a lot of muscle. I had a big engine," he says.
That's the "you" part of the "you can't out-train a bad diet"—"you" can do it if "you" are someone like Phelps and Nadolsky. Their training volume is also so high that it almost requires a "bad diet."
"These elite-level athletes are eating so much just to keep up with their training," Nadolsky says, and the same went for his experiment: Most of the food he ate was "quality food," but "if I ate just good, clean stuff, I couldn't get it all in."
As a result, at night he'd have to eat high-calorie foods like ice cream and cupcakes to reach his daily caloric needs. "Because I could, but also if I didn't, my running or swimming or whatever the next day was not as good."
"People might ask, 'Why isn't he eating lean meat with vegetables?' Because it's hard to eat 12,000 calories of anything," Heathers said, referring to Phelps' supposed training diet. "So he's got to eat pasta with commercial red sauce on it. He's got to eat things that are palatable."
So can you out-train a "bad" diet? "If you do it strategically, you enjoy it, and you have enough time, it's an option. It is possible," Nadolsky says, depending on who "you" are and what you consider to be a "bad" diet.
But it might not be worth it, as Heathers points out. "If I suddenly think, 'I want to improve everything [about my health],' I can't go out and do one of those five-hour training rides. My legs will fall off."
It's easier, as Nadolsky suggests, to have a healthy relationship with "bad" food like ice cream—have it as a treat, recognize it as a treat, and don't have it every day.
Unless, you know, you have 23 gold medals. Then eat whatever you want.