A break buys time on the diet, not extra fat off it#
A diet break is a planned return to maintenance calories — typically one to two weeks — taken in the middle of a deficit rather than at the end of one. The evidence for it is better than the skeptics allow and much narrower than the sales pitch. Pooled across 12 randomized trials and 881 participants, intermittent dieting with break periods produced no significant advantage over continuous restriction in body mass, fat mass, BMI, body-fat percentage or waist circumference. What it did produce was a significantly smaller reduction in resting metabolic rate1.
So the honest summary is not "breaks help you lose more." It is: a break appears to cost you nothing in fat, protects a measure most people cannot feel, and — in the trial that looked hardest at behavior — changes how the diet feels to run. That makes it a tool for continuing rather than a tool for accelerating, which is a less exciting claim and a more useful one. What follows is where the famous positive result came from, what actually recovers during a pause, what the psychological data show, and how to run one without misreading the scale afterward.
The trial everyone cites, and the detail that changes it#
The reputation of the diet break rests almost entirely on the MATADOR trial, in which men with obesity ran a 33% deficit either continuously or in alternating two-week blocks, and the intermittent group lost more fat (12.3 ± 4.8 vs 8.0 ± 4.2 kg) with a smaller composition-adjusted fall in resting energy expenditure (−360 ± 502 vs −749 ± 498 kJ/day, P < 0.05)3. That result, and the resistance-trained trial that failed to replicate it, are worked through in how to break a real plateau, so they get one paragraph here rather than a section.
One feature of its design deserves more attention than it usually gets. The continuous arm was on the program for 16 weeks; the intermittent arm was on it for 30. Both accumulated the same 16 weeks of actual restriction, but one group spent nearly twice as long inside the trial to do it — with all the contact, structure and supervision that implies. Nineteen and seventeen participants respectively completed. It is a genuinely striking result from a small trial with unequal exposure, which is exactly the profile of a finding that shrinks when the field catches up with it.
And the field has. Poon's meta-analysis includes MATADOR among its twelve trials, and the pooled body-composition estimate lands on no difference. A longer, larger trial points the same way: when 332 adults were randomized to continuous restriction, week-on-week-off restriction, or 5:2 for a full year, twelve-month weight loss came out at −6.6 kg, −5.1 kg and −5.0 kg respectively (p = 0.2), with no differences in body composition or dropout4. If breaks reliably improved fat-loss efficiency, a year of week-on-week-off should have been the place to see it.
What a pause plausibly does, and for whom#
The surviving effect — smaller RMR reduction — is worth taking seriously and worth sizing correctly. Poon's review reports it as a between-group difference in favor of the break protocols, and finds it larger in people with overweight or obesity than in resistance-trained participants1. That subgroup split is the same population pattern that separates MATADOR from its failed replication: the more adaptation a person has accumulated, the more there is for a pause to relieve.
There is a mechanism consistent with this, and it is worth stating as a hypothesis rather than a fact. A large part of what gets measured as adaptation depends on being in ongoing negative energy balance: it is largest during active loss and recedes once weight stabilizes, which is the pattern in how metabolic adaptation causes plateaus. A diet break is a deliberate, temporary stabilization — you remove the negative balance for a fortnight, and the component of the suppression that tracks it relaxes. That predicts exactly what the pooled data show: a real effect on measured metabolic rate, and no consequent effect on fat loss, because the suppressed component was never large enough to be the thing limiting the diet.
Running a break well therefore means actually reaching maintenance, not drifting to a smaller deficit and calling it a pause. That number is lower than it was when you started, and it needs recalculating for the body you have now — the method is in finding your maintenance calories.
The effect that showed up most clearly was psychological#
The most informative single trial on this question measured everything the metabolic story predicts and found the difference somewhere else. Thirty-eight resistance-trained women ran a ~25% energy deficit: one group continuously for six weeks, the other for the same six weeks of restriction with one-week breaks at energy balance inserted after weeks two and four, spread over eight weeks total. Fat mass, fat-free mass and resting metabolic rate showed no group-by-time interaction — RMR came in at p = 0.99, which is about as null as a result gets2.
Disinhibition — the tendency for eating to run away from you once control slips — moved sharply and in opposite directions. It fell in the break group (6.80 to 6.05) and rose in the continuous group (4.91 to 6.17), p < 0.01. One caution belongs on that: the break group started nearly two points higher, so some of the convergence could be regression toward the mean rather than the intervention. Take the direction seriously and the effect size loosely. The authors' own framing is the practical one — that breaks "may be employed for those who desire a short-term break from an energy-restricted diet without fear of fat regain."
That sentence is the whole case, stated by people who went looking for a metabolic benefit and did not find one. A break is permission with a return date. Its value is that it makes a long deficit survivable, and the evidence says you do not pay for it in fat.
Running one without misreading the aftermath#
| Decision | What the evidence supports |
|---|---|
| Length | 1–2 weeks; both the pooled trials and the psychological trial used breaks in this range |
| Intake | Actual maintenance, recalculated for your current weight — not a softer deficit |
| Frequency | Every 4–8 weeks of restriction, or when adherence rather than the scale is failing |
| Who benefits most | People carrying more accumulated adaptation: longer, steeper diets, higher starting body fat |
| What to expect on the scale | A rise of 1 kg or more within days, mostly glycogen and its bound water |
| What not to expect | More fat lost over the whole diet — twelve pooled trials found none |
That fifth row is the one that ends most people's relationship with diet breaks. Eating at maintenance refills glycogen, and glycogen is stored wet, so the scale goes up quickly for reasons that have nothing to do with fat — the mechanics are in why the scale fluctuates. If you did not expect that rise, the break reads as a failure and you never take another one. Expect it, wait a week after resuming, and read the trend rather than the fortnight.
One framing note, since it is the actual argument for planning a pause rather than having one: a scheduled break is a fortnight of maintenance eating with a date on it. An unscheduled one is a fortnight of maintenance eating you did not choose, followed by a story about willpower. The energy is often similar; the difference is that only one of them ends when you say it does — and only one leaves the week's arithmetic intact.
FAQ#
How long should a diet break be?#
One to two weeks at true maintenance is the range used across the trials that have tested it, including the pooled set of 12 randomized trials1 and the trial that inserted one-week breaks after every two weeks of restriction2. Shorter than a week is a high day rather than a break; longer than two weeks is a maintenance phase, which is a legitimate choice but a different one.
What should you eat during a diet break?#
Your recalculated maintenance calories, with protein and training held where they were. The point is to remove the energy deficit, not the structure — the food quality, the protein target and the resistance work are what protect lean tissue, and none of them is what you are taking a break from. Setting the number against your current, lighter body matters, because the maintenance figure from the start of the diet is now a surplus.
Will I regain weight during a diet break?#
Some scale weight, almost certainly — often a kilogram or more within a few days — and very little of it fat. Eating at maintenance refills muscle and liver glycogen, which is stored with several times its own weight in water. Across 12 randomized trials there was no body-composition penalty to taking breaks1, and the trial that measured it concluded people could take one "without fear of fat regain"2. Give it a week after resuming before you judge anything.
Sources#
- Poon ET, Tsang JH, Sun F, Zheng C, Wong SH. Effects of intermittent dieting with break periods on body composition and metabolic adaptation: a systematic review and meta-analysis. Nutr Rev. 2025;83(1):59-71.
- Siedler MR, Lewis MH, Trexler ET, et al. The Effects of Intermittent Diet Breaks during 25% Energy Restriction on Body Composition and Resting Metabolic Rate in Resistance-Trained Females: A Randomized Controlled Trial. J Hum Kinet. 2023;86:117-132.
- Byrne NM, Sainsbury A, King NA, Hills AP, Wood RE. Intermittent energy restriction improves weight loss efficiency in obese men: the MATADOR study. Int J Obes (Lond). 2018;42(2):129-138.
- Headland ML, Clifton PM, Keogh JB. Effect of intermittent compared to continuous energy restriction on weight loss and weight maintenance after 12 months in healthy overweight or obese adults. Int J Obes (Lond). 2019;43(10):2028-2036.
