You can't out-exercise bad eating habits: what trials found

Researchers had people train five days a week and left their food alone. One person lost 14.7 kg and another gained 1.7 kg on the same program.

On this page
An open metal tin of biscuits resting on a folded white towel on a wooden table.
A workout happens a few times a week. The tin is open every day, and over a week the tin usually wins.
Summary in 5 points
  • Across 14 trials, aerobic exercise alone produced about 1.7 kg of average loss at 12 months.
  • In E-MECHANIC, eating rose about 124 kcal a day, roughly 865 a week, with no change in resting burn or other activity.
  • One 12-week program ranged from a 14.7 kg loss to a 1.7 kg gain, with food intake moving in opposite directions.
  • High-volume supervised exercise can work: 4 to 6% loss over 10 months with over 90% attendance.
  • Fitness and blood pressure improved even in people who lost almost no weight, so keep training for that.

What supervised exercise trials found when diet was left alone#

You can exercise five days a week and still not lose weight. Researchers have tested exactly this, and the results explain why a recurring eating habit usually beats a training plan.

The usual assumption goes like this. Had a big dinner? Go for a run. Ate too much over the weekend? Add a gym session. It sounds reasonable, because exercise does burn calories. Can you outrun a bad diet? does the snack-versus-miles arithmetic for a single indulgence. This article is about the habit that repeats, and what happened when people trained for months while nobody changed their food.

A systematic review pooled 14 randomized trials and 1,847 participants doing aerobic exercise as the isolated intervention. Average weight reduction was 1.6 kg in the 6-month trials and 1.7 kg in the 12-month trials1. Those are pooled estimates from trials of different lengths, not one group measured twice. They aren't zero, but they're far below what most people expect from a year of training.

A direct comparison gives the same picture. In a year-long trial of 439 sedentary postmenopausal women with overweight or obesity, the diet group lost 8.5% of body weight, the exercise group 2.4%, and the combined group 10.8%, while controls lost 0.8%2. Exercise helped, and adding it to diet beat either alone. On its own, it was the weaker lever by a wide margin, at least in this population.

So where did the calories go? Often, back onto the plate. In the E-MECHANIC trial, adults did supervised exercise at either 8 or 20 kcal per kg per week for 24 weeks. Intake rose by about 91 kcal/day at the lower dose and 124 kcal/day at the higher one, while controls barely changed. Non-exercise activity and resting metabolic rate didn't differ, so eating accounted for it3.

An extra 124 calories a day sounds like nothing. By our arithmetic it's about 865 calories a week. Your legs worked for half an hour. Your mouth needed four minutes to negotiate a refund.

One program, a result from −14.7 kg to +1.7 kg#

The average also hides how much people differ. In one study, 35 overweight or obese adults completed 12 weeks of supervised exercise, five times a week. Mean loss was 3.7 kg, and individual results ran from a 14.7 kg loss to a 1.7 kg gain4.

The researchers split the group by whether people lost what the exercise predicted.

Measure, 12 weeks Compensators Noncompensators
Weight change −1.5 kg (±2.5) −6.3 kg (±3.2)
Change in reported daily intake +268 kcal (±455) −130 kcal (±485)
Change in resting metabolic rate −69 kcal (±269) +14 kcal (±243)

That's about a 400-calorie daily gap in the direction people's eating moved, and the spreads are wide. These subgroups were defined after the fact rather than randomized, and the study was small, so the figures aren't universal averages. They do show why two people can follow the same program and get very different scales.

E-MECHANIC also looked at what went along with compensating. High compensators reported more appetite and more belief that healthy behaviors offset unhealthy ones, along with sweet cravings, sleep disturbance and pain3. Eating back exercise calories goes further into why this happens.

The weekly ledger: three workouts against seven days of eating#

Here's an illustrative week, not a clinical result. You complete three workouts, each burning 400 additional calories. You also have a recurring 250-calorie daily extra: a sweetened coffee, some snacks, slightly larger portions.

Weekly line Calories
3 workouts × 400 kcal 1,200 burned
250 kcal extra × 7 days 1,750 eaten
Net +550 surplus

The asymmetry is the whole point. Exercise happens three or four times a week. Eating decisions happen every day, often several times a day. A small repeating surplus has more turns at the table, and the workout numbers it's competing with are shakier than they look: among seven wrist devices, none achieved an energy-expenditure error below 20%5.

Enough exercise can work, and the price is high#

The title of this article shouldn't be read as "exercise doesn't work." In the Midwest Exercise Trial 2, 141 young adults ate ad libitum and did supervised exercise expending 400 or 600 kcal per session, five days a week, for 10 months. Those who attended at least 90% of sessions lost 3.9 kg (4.3%) and 5.2 kg (5.7%), while controls gained 0.5 kg. The two doses weren't significantly different6.

So exercise alone can produce meaningful loss. Look at what it took: five supervised sessions a week for ten months, in participants averaging 22.6 years old, with attendance above 90%. That's very different from assuming two gym visits will cover a nightly habit.

Dose can matter, too. In a smaller 12-week trial, people doing 300 or 600 kcal per session five days a week compensated by 943 and 1,007 kcal a week. Doubling the exercise barely moved compensation, so only the higher dose lost fat7. Only 29 people completed it, which makes this a lead rather than a rule, but it's a reason not to read compensation as a fixed tax. A 2024 meta-analysis of 116 trials similarly found roughly linear returns up to 300 minutes a week, at about 0.52 kg per 30 weekly minutes8.

The fair conclusion is that relying on exercise alone usually takes far more time and consistency than changing what you eat. It isn't impossible. It's expensive. The pillar on exercise calories traces where the rest of the number leaks.

Exercise pays even when the scale doesn't move#

Don't do cardio just to burn calories. Do cardio because your heart needs training, too. You can lose weight without running a single kilometer, and you can run a lot of them without losing any weight.

One study makes that split unusually clear. Of 58 adults on 12 weeks of supervised aerobic exercise, 26 lost less than predicted, averaging just 0.9 kg. Those same people still gained aerobic capacity and improved their health markers9.

Change in the 26 who lost less than predicted Average
Aerobic capacity +6.3 mL/kg/min
Systolic blood pressure −6.0 mmHg
Waist circumference −3.7 cm
Resting heart rate −4.8 beats/min

These are within-group changes rather than comparisons against a control, but the waist number hints at something else. The treadmill is not broken just because the scale hasn't moved. Lose 2 kg of fat, gain 1 kg of lean mass and hold 1 kg of extra water, and the scale reads the same while your body has changed. That's an illustration, and muscle gain masking fat loss covers how to tell if it applies to you.

So split the jobs. Use your eating habits to manage intake, and use exercise to build fitness and strength.

  • Find the recurring extra. Look for what repeats most days: a drink, a snack, a large sauce, a big weekend portion. One thing you do daily outweighs one thing you do monthly.
  • Change that first. Removing a 250-calorie daily habit is worth 1,750 calories a week and takes no extra hours.
  • Keep training for what it does best. Don't drop it because the scale is slow, and don't use it as a refund.

Don't exercise to compensate for bad eating habits. Improve your eating habits so your exercise can make you fitter, stronger and healthier.

FAQ#

How much weight do people lose from aerobic exercise with no diet?#

In pooled trials, about 1.6 to 1.7 kg on average at 6 and 12 months. Supervised programs with high attendance and high volume have done better, around 4 to 5 kg over 10 months in young adults, and individual results vary widely.

Why do some people lose nothing on the same exercise program?#

Food intake is the best documented reason. In one 12-week study, people who lost less than predicted reported eating about 268 kcal a day more, while others reported eating about 130 kcal less. It's a small study with self-reported intake, so treat those as approximate.

Does doubling your exercise double the calorie deficit?#

Not reliably. In one 12-week trial, doubling the session dose left weekly compensation almost unchanged, so the bigger dose kept more of its deficit. In a 10-month trial, 400 and 600 kcal sessions didn't differ significantly. The answer depends on the person and the dose.

Is cardio worth doing if your weight stays the same?#

Yes. In a study of people who lost less than predicted, average aerobic capacity, blood pressure, waist size and resting heart rate still improved. Those are within-group changes, but the health case for cardio doesn't depend on the scale.

Sources#

  1. Thorogood A, Mottillo S, Shimony A, et al. Isolated aerobic exercise and weight loss: a systematic review and meta-analysis of randomized controlled trials. Am J Med. 2011;124(8):747-755.
  2. Foster-Schubert KE, Alfano CM, Duggan CR, et al. Effect of diet and exercise, alone or combined, on weight and body composition in overweight-to-obese postmenopausal women. Obesity (Silver Spring). 2012;20(8):1628-1638.
  3. Martin CK, Johnson WD, Myers CA, et al. Effect of different doses of supervised exercise on food intake, metabolism, and non-exercise physical activity: The E-MECHANIC randomized controlled trial. Am J Clin Nutr. 2019;110(3):583-592.
  4. King NA, Hopkins M, Caudwell P, et al. Individual variability following 12 weeks of supervised exercise: identification and characterization of compensation for exercise-induced weight loss. Int J Obes (Lond). 2008;32(1):177-184.
  5. Shcherbina A, Mattsson CM, Waggott D, et al. Accuracy in Wrist-Worn, Sensor-Based Measurements of Heart Rate and Energy Expenditure in a Diverse Cohort. J Pers Med. 2017;7(2):doi: 10.3390/jpm7020003.
  6. Donnelly JE, Honas JJ, Smith BK, et al. Aerobic exercise alone results in clinically significant weight loss for men and women: midwest exercise trial 2. Obesity (Silver Spring). 2013;21(3):E219-28.
  7. Flack KD, Ufholz K, Johnson L, et al. Energy compensation in response to aerobic exercise training in overweight adults. Am J Physiol Regul Integr Comp Physiol. 2018;315(4):R619-R626.
  8. Jayedi A, Soltani S, Emadi A, et al. Aerobic Exercise and Weight Loss in Adults: A Systematic Review and Dose-Response Meta-Analysis. JAMA Netw Open. 2024;7(12):e2452185.
  9. King NA, Hopkins M, Caudwell P, et al. Beneficial effects of exercise: shifting the focus from body weight to other markers of health. Br J Sports Med. 2009;43(12):924-927.

Source: BurnWeek — "You can't out-exercise bad eating habits: what trials found", https://burnweek.fit/blog/cannot-out-exercise-bad-eating/. Licensed CC BY 4.0: free to quote or reuse with a link to this page.

This article was researched and drafted with AI assistance and reviewed for accuracy by the BurnWeek team. It is general information, not medical advice. How we research and correct our articles →