Why Every Diet Works Until Month Three

Three diets in five years. Each one started with a new name and ended the same way — weight dropped for a few weeks, progress stalled around month three, and the number crept back to where it started. Low carb. Then intermittent fasting. Then calorie counting. The names changed. The timeline didn’t.

The internet has three explanations for why most diets fail.

All three are wrong.

Why Do Most Diets Fail?

Keto, paleo, Atkins, Zone, low-fat, Mediterranean — every plan with a brand name has been tested against the others. They all produce the same weight loss. Around 4.5 kg at six months. Every single one.

The gap between the best-performing plan and the worst was so small a bathroom scale couldn’t reliably detect it. The diet-shopping loop — trying a new name every January — was never going to solve anything, because nothing was wrong with the last one.

If the plan itself doesn’t matter, something else does. And that something has been measured down to the month.

A mathematical model built from tracked weight-loss data found what the diet industry doesn’t advertise. The six-month plateau isn’t your metabolism fighting back. It’s compliance quietly declining, one skipped day at a time.

In the first month, adherence sits around 80%. By month three, it drops to somewhere near 40%. Not because anyone gives up in one dramatic moment. Because small, intermittent lapses stack — a missed meal here, an extra portion there — until the calorie deficit that was driving the results has nearly vanished. The model’s compliance numbers are derived from population-level data rather than individual day-by-day tracking, but the curve they produce matches what free-living weight-loss trials consistently show: early discipline eroding on a schedule.

The metabolism story everyone reaches for? Even if metabolic adaptation dropped energy expenditure by 10% more than physics predicts, the math shows it wouldn’t change when the plateau arrives. It shifts where you end up, slightly. The timing stays the same regardless of what your metabolism does, because the variable controlling the timeline was never metabolic.

Sticking with the plan predicted weight loss 8.5 times more strongly than which plan someone picked. That ratio came from a trial tracking four diets head-to-head, measuring both compliance and results. Nearly nine to one. Every argument about which diet is “best” is arguing about a variable that barely registers compared to the one nobody wants to talk about.

Which means the question “which diet works best?” was the wrong question from the start.

The profile of someone who keeps weight off long-term has nothing to do with demographics. Age, gender, income, education, starting weight: none predicted success. What separated the maintainers was what they did, not who they were.

Self-monitoring — stepping on the scale, tracking food intake — was the strongest behavioral predictor. Increasing physical activity came next. Not how active someone already was. The change itself. People who moved more than before kept the weight off, regardless of where they started.

One variable reliably predicted failure. Not willpower, not a slow metabolism, not a history of yo-yo dieting. Emotional eating — losing control of intake in response to feelings rather than hunger. That was the only strong negative predictor across the entire evidence base.

The plan never failed. Compliance decayed on a schedule so predictable it can be graphed — and the explanation you were given pointed at the wrong variable entirely. What actually keeps the weight off has nothing to do with which plan you followed and everything to do with the behaviors you build around it.