Why exercise alone rarely melts fat, and what the research says works

The body pushes back

The arithmetic everyone learns is simple. A treadmill session burns 300 calories, so do it every day and you are 2,100 calories down by Sunday. People who try this usually find the scale moves far less than the sum promised, and conclude they did something wrong.

They did not. The sum is wrong. In 2016 the anthropologist Herman Pontzer and colleagues measured total daily energy use in 332 adults from five populations, using a precise laboratory method rather than estimates. Up to a point, more active people burned more energy. Beyond it, the line went flat. The body appeared to protect its daily budget by spending less elsewhere.

What the 2026 analysis found

This February, a larger analysis in Current Biology pulled together exercise trials and population studies to test the idea properly. In human trials of aerobic exercise, total daily energy use rose by only about 30 percent of what simple addition predicted. Most of the calories the exercise burned were offset over the rest of the day.

The authors found the effect was strongest in the combination most people try first: aerobic exercise paired with dieting. Cut food and add long cardio together, and the body defends itself hardest. Part of the offset came from a fall in resting metabolism, and part from changes the studies could not fully explain.

None of this makes exercise useless. It makes it a poor tool for creating a calorie deficit on its own, and an excellent tool for almost everything else.

Weights behave differently

The same 2026 paper found compensation appeared smaller with strength training. The paper does not establish why, and I will not guess. What it does mean is that the case for lifting during fat loss does not rest on calories burned in the session, which is just as well, because its real value lies elsewhere.

What strength training does to body fat

A 2022 meta-analysis of 58 randomised trials looked at strength training on its own, with no change in diet, against doing nothing. Body fat percentage fell by about 1.5 points, fat mass by about half a kilo, and visceral fat, the fat packed around the organs that drives diabetes and heart disease, came down too.

Half a kilo is modest, and I would rather say so than oversell it. The bigger effect of strength training during fat loss is what it protects. A review of 52 studies in adults over fifty found that 81 percent of diet-only groups lost at least 15 percent of their weight as lean tissue, against 39 percent of groups that also exercised. Lose weight by dieting alone and a large part of it can be muscle. Lift while you diet and much more of it is fat.

What actually works

Put the research together and the roles are clear. Food does most of the work of creating a deficit. Strength training two or three times a week decides whether what you lose is fat or muscle. Walking and other aerobic exercise are worth doing for your heart and your blood pressure, just not as a way to earn dessert.

• Do not eat back your machine calories. The body has already offset part of that number.
• Do not pair a crash diet with hours of cardio. That is the combination the body resists hardest.
• Lift through the deficit. It is the best-supported way to keep muscle.
• Change the food you already eat rather than a second set of meals. Losing fat on a Gujarati diet shows where the calories in a normal day actually sit.

If you want to compare the two kinds of training directly, cardio against weight training sets out a week that uses both. And if your weight has come back before, why weight comes back after dieting explains the pattern.