Keto versus low-carb versus intermittent fasting versus whatever's trending this month — the diet debate online is mostly people arguing past each other about which macronutrient ratio is "correct." The largest head-to-head trial ever run on the question suggests that's the wrong argument entirely.
The DIETFITS trial, published in JAMA in 2018, randomized 609 adults to either a healthy low-fat diet or a healthy low-carb diet for a full 12 months — no crossover, no short-term crash period, just a year of each. At the end, the two groups had lost almost the same amount of weight on average: 5.3 kg on low-fat versus 6.0 kg on low-carb, a gap too small to be statistically meaningful.
That roughly 90-pound spread inside each arm is the real finding. The researchers had also hypothesized that a person's genotype pattern or their baseline insulin secretion would predict which diet worked better for them — the idea that some people are biologically "low-carb responders" and others aren't. Neither panned out: the diet-by-genotype interaction wasn't statistically significant, and neither was the diet-by-insulin-secretion interaction. People matched to their "ideal" diet by genotype didn't do any better than people mismatched to it.
So if it wasn't the macronutrient ratio and it wasn't biology, something else was driving who lost 60+ pounds and who gained weight on paper-identical rules. The trial's authors pointed to execution — how consistently each person actually followed the plan they were assigned, not which plan it was.
A separate 2014 JAMA meta-analysis pooling 48 randomized trials and over 7,000 dieters found the same shape of result at a larger scale: differences between popular named diets (Atkins, Zone, South Beach, and others) were real but small — small enough that the authors' practical conclusion was to recommend whichever diet a person is most likely to actually stick with, rather than chasing the diet with the marginally better average.
This is why OriginLeap's diet habit isn't "eat low-carb" or "do intermittent fasting." It's: pick your own plan, and follow it every day for 75 days. The app has no opinion on keto versus Mediterranean versus anything else, because the research doesn't give it one — it tracks whether you did the thing you said you'd do, which is the variable that actually showed up in the data.
Curious what the other six habits are backed by?
See the rest of the research