All diets work short-term. Why do virtually all of them fail long-term?
The Short Answer
That most do is Mann’s finding rather than this page’s assumption: across the 14 observational studies she reviewed with four-plus-year follow-up, dieters lost about 14 kg and kept roughly 3 kg of it; in the 8 studies that reported counterproductive outcomes an average of 41% weighed more at follow-up than before they dieted; and in randomized trials the average maintained loss was 1.1 kg.[1] They fail, on the account this page gives, because they fight the body’s weight-regulation system without changing it. When you restrict calories, your body responds by lowering metabolism, increasing hunger hormones, and defending the weight it had settled at. You lose weight initially, but the physiological pushback eventually overwhelms willpower. The diet ends, but the metabolic adaptations persist, often resulting in regaining more weight than you lost. The carbohydrate-insulin account points instead to hormonal drivers of weight rather than intake alone: a proposal resting on a single source that is a party to the argument.
Something to Sit With
How many diets have you tried? What happened after each one ended?
Were you fighting your body’s defense system, or working with it?
What would need to be different for a change to be permanent?
Learn More
The Initial Success.
Almost any diet works initially:
- Low-fat diets: Weight loss in the first months
- Low-carb diets: Weight loss in the first months
- Calorie counting: Weight loss in the first months
- Meal replacement: Weight loss in the first months
- Mediterranean, Paleo, Whole30: All show early results
This is why diet advice seems so confusing. Study after study shows diet X “works,” because in the short term, most approaches produce weight loss. The body draws on its reserves when intake decreases.
The Metabolic Pushback.
The body doesn’t passively accept weight loss. It fights back:
Metabolic slowdown. When you eat less, you burn less, and the drop outlasts the diet. Biggest Loser contestants were still running a 499 kcal/day metabolic adaptation six years after the competition, including those who regained the weight.[2]
Hunger hormone surge. Ghrelin (the hunger hormone) increases. Leptin (the satiety hormone) decreases.[3] You feel hungrier after losing weight than you did at the heavier weight.
Increased efficiency. Your muscles do more work per calorie, and you expend less through movement and heat.[3] The same meals that maintained your higher weight now cause gain at your lower weight.
This isn’t failure of willpower; it’s biology. On one account your body evolved against starvation, and its defenses answer a diet’s falling fat stores all the same.
The Set Point Problem.
Your body defends against weight loss.[3] The lipostat model reads that as defense of a “set point” for fat mass: one account, and a contested one.[4]
That is the set point model, one of three. Another has weight settling wherever your surroundings put it, with nothing defended, and a third proposes upper and lower boundaries. Neither familiar model is adequate alone: set point explains regain after dieting, settling point explains the environment.[5] On the set point model, weight below the set point is pushed back up and weight above it pushed back down, though the measured defense against gain is weaker than against loss.[2]
On that model, calorie restriction can temporarily push weight below what the body is defending without changing what it defends. You’re holding a ball underwater. As long as you maintain pressure (diet), it stays down. When pressure releases (diet ends), it shoots back up.
On the carbohydrate-insulin model, high-glycemic carbohydrates raise insulin, insulin promotes fat storage, and overeating follows from growing fat stores rather than causing them; the model’s own proponents call the human evidence inconclusive.[6]
The Sustainability Trap.
Most diets require ongoing restriction that becomes unsustainable:
- Constantly counting and limiting calories
- Permanently eliminating entire food groups
- Ongoing hunger and deprivation
- Social difficulty maintaining the restrictions
Eventually, life intervenes. Stress increases, resolve wavers, special occasions accumulate, and the diet breaks. But now you’re returning to previous eating with a slower metabolism and higher hunger, a recipe for rapid regain.
This is why weight cycling (yo-yo dieting) is so common. Diet → lose weight → regain weight → diet again → lose weight → regain more weight. Each cycle can make the next one harder.
What Actually Works Long-Term.
Sustainable approaches share common features:
Address the hormonal environment. If insulin drives fat storage, as the carbohydrate-insulin model proposes, what you eat matters as well as how much; that is this page’s reading of a model still being argued.[6]
Build habits, not willpower. An automatic pattern is a decision you are not making today; a daily rule is one you are. That is the whole argument, and it needs no theory about what daily deciding costs you.
Change the food environment. Removing problem foods from your space leaves nothing there to resist; that this beats resisting them daily is advice, not a measured comparison.
Focus on eating patterns, not just food choices. When you eat is a choice alongside what you eat; whether time-restricted eating changes metabolism compared with grazing is not settled.
Make it identity, not diet. “I’m on a diet” names an end date. “This is how I eat” names none.
Further reading:
- Fothergill E et al., “Persistent metabolic adaptation 6 years after ‘The Biggest Loser’ competition”[7] — Obesity, 2016
- Sumithran P et al., “Long-term persistence of hormonal adaptations to weight loss”[8] — New England Journal of Medicine, 2011
- Ludwig DS & Ebbeling CB, “The Carbohydrate-Insulin Model of Obesity: Beyond ‘Calories In, Calories Out’”[9] — JAMA Internal Medicine, 2018
- Mann T et al., “Medicare’s search for effective obesity treatments: Diets are not the answer”[10] — American Psychologist, 2007
Written by AI, directed and approved by me. How this is made →