Diet Breaks: Does a Week at Maintenance Help Fat Loss?

What a diet break is supposed to fix

A diet break, in the sense the trials use, is a planned stretch of eating at maintenance in the middle of a calorie deficit. Usually one or two weeks, then back to the deficit. A refeed is the shorter cousin: a day or two of higher intake, usually from carbohydrate, every week or so.

The claims for it are mostly physiological: that it slows the fall in resting energy expenditure that comes with dieting, and that it protects muscle. The other claim, that it makes the diet easier to live with, is about hunger and morale. It sounds softer, and it turned out to have the better evidence.

I went looking because the plateau article on this blog weighs diet breaks on a single trial in 36 men. That trial is the famous one, but it is not the only one, and the others turned out to be more interesting.

The trial everyone cites, and what it bought

MATADOR randomised 51 men with obesity to 16 weeks of continuous energy restriction, or to the same 16 weeks of restriction split into eight two-week blocks with seven two-week blocks at energy balance in between, 30 weeks in total. In both arms, intake while dieting was 67% of maintenance. Among the 36 who completed per protocol, the intermittent group lost 14.1 kg against 9.1 kg, and 12.3 kg of fat against 8.0 kg. Fat-free mass loss did not differ.¹

On metabolism the result is narrower than the headline. The fall in absolute resting energy expenditure did not differ between the groups. Only after adjusting for the change in body composition was it smaller in the intermittent arm.¹ I went through the limits in the weight loss plateau piece: all men, all with obesity, under 20 per group, and the intermittent arm took 30 weeks to get its result against 16. So MATADOR showed more fat lost per week of dieting. It did not show faster fat loss by the calendar.

Two trials in people who lift

Most people asking about diet breaks online are tracking macros for fat loss and lifting weights, which is not who MATADOR enrolled. Two trials since have tested that reader.

ICECAP took 61 resistance-trained adults, 32 of them women, with an average body fat of 25.5%. Half did 12 weeks of continuous moderate restriction. The other half did four three-week blocks of the same restriction with a week at energy balance after each of the first three, 15 weeks in total. Analysed by intention to treat, the breaks did not produce lower fat mass, lower body weight or more fat-free mass. Fat mass at the end was 15.3 kg against 18.0 kg, a gap the confidence intervals swallow (P = 0.321). Resting energy expenditure, strength and endurance, and the hormones they measured, including leptin and testosterone, did not differ either.²

What did differ was appetite. The group with breaks reported less hunger and less desire to eat, and more satisfaction, and had higher levels of peptide YY, a gut hormone tied to fullness.²

The second trial was smaller and shorter. Thirty-eight resistance-trained women, average age 22, cut intake by 25% for six weeks, either continuously or with a week at energy balance after every two weeks of dieting, eight weeks in total. Everyone was told to eat 1.8 g of protein per kilogram and trained under supervision three times a week. Body composition and resting metabolic rate changed the same way in both groups.³

One questionnaire score did diverge. Disinhibition, roughly a measure of losing control around food, rose in the continuous group from 4.91 to 6.17 and fell in the break group from 6.80 to 6.05.³ Look at where those numbers ended up, though. The two groups finished within about a tenth of a point of each other, and the break group had started well above. It was one of eight eating-behaviour measures, and the other seven did not differ. The authors' own summary is modest: breaks do not appear to improve body composition or metabolic rate over six weeks of dieting, but can be used by someone who wants a short pause without fear of fat regain, at the cost of a longer diet.³

Weekend refeeds, and a result that shrank on reanalysis

The refeed version has its own trial, and its own cautionary tale. Twenty-seven resistance-trained men and women spent seven weeks on a diet about 25% under their needs, with four training sessions a week. One group ate at that level every day. The other ate higher-carbohydrate meals on two consecutive days each week and restricted on the other five. Both groups lost similar amounts of weight and fat. The authors concluded that the two-day refeed preserved fat-free mass and resting metabolic rate.⁴

Then another group of researchers, including one of the ICECAP authors, reanalysed the published data. Their objection was a common statistical mistake: the paper found a significant drop from baseline in the continuous group and no significant drop in the refeed group, and treated that as a difference between the groups. Compared directly, the groups did not differ significantly in fat-free mass or in resting metabolic rate. Only dry fat-free mass, a measure that leaves out body water, came out different. They also pointed out that only 27 of the 54 people who went through baseline testing finished the trial. Counting everyone, with the dropouts assumed to have changed nothing, the dry fat-free mass difference fell from 1.7 kg to 0.9 kg.⁵

I am telling this at length because the refeed claim usually reaches people in the original form, and because it is a good example of why I read the comments on a paper before citing it. If your weekends already run higher than your weekdays, that is a different subject, and the weekend calorie tracking piece works out what two days at maintenance cost a five-day deficit.

What 12 trials add up to

A 2025 meta-analysis pooled 12 randomised trials, 881 participants in all, comparing diets interrupted by break periods with continuous restriction. Both approaches improved body mass, fat mass, BMI, body fat percentage and waist circumference, with no significant difference between them. Resting metabolic rate fell significantly only with continuous restriction, and the drop was significantly smaller with breaks. That protective effect on resting metabolic rate was stronger in people with overweight or obesity than in resistance-trained people.⁶

I could read only the abstract, which does not give the size of the metabolic rate difference, so I cannot tell you whether it comes to 30 calories a day or 150. The pattern fits the individual trials, though: MATADOR, in men with obesity, found a smaller adjusted fall,¹ while ICECAP and the trial in women, both in lifters, found none.²³ On fat loss, which is the number most people are tracking, the pooled answer is the same either way.

What a week at maintenance does to the scale

If you log your weight, a break can look like a failure when it is not one.

ICECAP measured what happened across one of its diet-break weeks. Intake went up to predicted maintenance, mostly as extra carbohydrate. Body weight rose by 0.6 kg on average and fat-free mass by 0.7 kg, while fat mass did not change significantly. Resting energy expenditure went from about 7,000 to 7,200 kJ a day, which is roughly 48 calories. Hunger, irritability and the urge to eat dropped, and fullness, satisfaction and alertness went up.⁷

So the scale goes up, and what comes back is mostly not fat. Fat-free mass on a body composition scan includes water and stored carbohydrate, so most of that 0.7 kg is likely glycogen and the water held with it rather than new muscle. The authors did not break it down that far, so I am inferring that part. MATADOR, with two-week blocks, recorded an average weight change of 0.0 kg across its balance periods,¹ so a break does not have to show on the scale at all. If yours does, it is most likely this.

So do not judge a break by the weigh-in at the end of it. Compare the trend two or three weeks after you go back to the deficit with the trend before the break.

Can you stop logging during a diet break?

People ask whether the break means a break from calorie tracking too. There is one trial that comes close to answering it, and it was designed to find harm.

In 2003, researchers at Brown wanted a way to make people relapse on purpose, so they could study relapse without waiting six months for it. They gave 142 people a 14-week behavioural weight-loss programme. One group had no breaks. One took a six-week break halfway through. One took three two-week breaks. In the breaks, people were told to stop all weight-loss efforts.⁸

It did not work, as a way to cause relapse. People did drop their self-monitoring, their diet and their self-weighing during the breaks, though not their exercise, and then picked them all back up quickly afterwards. Weight loss slowed or reversed slightly during the breaks, but overall losses at five and eleven months were the same in all three groups. The authors concluded that prescribed breaks did not seem to do harm.⁸

That is a reassuring result, and an old one, in a supervised programme with weekly sessions to come back to. A break you take on your own has nobody waiting for you at the end of it. I think you can stop logging for a fortnight, but I would put the restart date in the calendar before stopping.

What the evidence does not settle

The trials are small. The largest body-composition trial here randomised 61 people, the refeed trial finished 27, and MATADOR's headline comparison rests on 36. They are short as well. MATADOR did follow people afterwards: six months on, among the 28 men still available, the intermittent group was on average 8.1 kg further below its starting weight than the continuous group.¹ That is encouraging and it is 28 men. The Brown trial ran to 11 months but tested stopping effort, not eating at a measured maintenance.⁸ Nobody has tested whether a diet break changes regain a year or two later, which is the outcome most people care about.

Everyone in them was supervised. MATADOR provided all main meals and snacks for the whole study,¹ the others worked out maintenance for their participants, and in the lifting trials the training was done in front of a researcher. Whether someone eating at maintenance by app estimate, with a target that is itself an estimate, gets the same result is untested.

The appetite findings are self-report from people who knew which diet they were on. That does not make them wrong, and the peptide YY result in ICECAP points the same way,² but it is the kind of outcome that knowing your group can move.

And the cost is real in every trial. MATADOR took 30 weeks against 16, ICECAP 15 against 12, and the trial in women 8 against 6.¹²³ A diet break makes the diet longer, and whether that is worth it depends on how much the hunger is costing you.

How to take a diet break while tracking macros for fat loss

If you want one, this is how I would run it, kept as close as I can to what the trials actually did.

Take one or two weeks, not a weekend. Those are the lengths the body-composition trials tested.²³ Eat at maintenance, not above it. A break in these trials was energy balance, and the weight did not move much because of that. It was not a free fortnight.

Change the target, rather than ignoring it. In MyPlate that means updating your weight in Profile and then setting Goal to Maintain on the same screen, because logging a weigh-in does not recompute the target on its own. The maintenance calories piece goes through why the order matters. Take a 35-year-old woman, 165 cm and 75 kg, who exercises moderately. Her weight-loss target is 1,904 calories, with 133 g of protein, 214 g of carbohydrate and 57 g of fat. At maintenance it becomes 2,240, with 140 g, 263 g and 69 g. Most of the extra 336 calories goes to carbohydrate, which is close to how ICECAP ran its breaks.⁷ Protein barely moves, and there is no reason for it to.

Keep logging if you can, even loosely. A break is a good week to find out whether your estimated maintenance is right: if your weekly average holds steady across a fortnight at the new target, you have measured it. That is useful whatever the break does for your fat loss.

Then switch the goal back to Lose weight on a set date, and give it two or three weeks before you compare your trend with where it was. Expect a small rise during the break and do not act on it.⁷

A diet break is not a reason to cut harder either side of it. The deficit in MATADOR was a third below maintenance and run by a research team.¹ MyPlate's weight-loss target is 15% below maintenance, and it will not show a target below 1,200 calories a day for women or 1,500 for men, the floor in the NIH/NHLBI guidelines for anyone dieting without medical supervision.⁹ If you find the break is the only thing making the diet bearable, or logging food is making you anxious, stop and talk to a doctor or a registered dietitian.

Common questions

Take the break for the hunger, not the metabolism

The case for a diet break that you usually hear is metabolic. On fat loss, the pooled trials found the same result with or without one, a smaller drop in resting metabolic rate that may matter more in people with obesity, and a refeed finding that shrank when someone checked the arithmetic. The case that holds up better is the one about living with the diet. People with breaks reported less hunger, and stopping for a while did not undo their progress.

So if counting calories to lose weight has started to grind, a week or two at maintenance is a reasonable thing to try, as long as you know the diet will take longer and the scale will move a little. Set the new target, keep logging, and put the end date in the calendar before you start.

Sources

Every figure above traces to one of these. If you find a number that doesn't match the source it claims, tell me and I'll correct it.

1. Byrne NM, Sainsbury A, King NA, Hills AP, Wood RE (2018). Intermittent energy restriction improves weight loss efficiency in obese men: the MATADOR study. International Journal of Obesity, 42(2):129-138. Read on PubMed →
2. Peos JJ, Helms ER, Fournier PA, et al. (2021). Continuous versus intermittent dieting for fat loss and fat-free mass retention in resistance-trained adults: the ICECAP trial. Medicine & Science in Sports & Exercise, 53(8):1685-1698. Read on PubMed →
3. Siedler MR, Lewis MH, Trexler ET, et al. (2023). 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. Journal of Human Kinetics, 86:117-132. Read on PubMed →
4. Campbell BI, Aguilar D, Colenso-Semple LM, et al. (2020). Intermittent energy restriction attenuates the loss of fat free mass in resistance trained individuals. A randomized controlled trial. Journal of Functional Morphology and Kinesiology, 5(1):19. Read on PubMed →
5. Peos J, Brown AW, Vorland CJ, Allison DB, Sainsbury A (2020). Contrary to the conclusions stated in the paper, only dry fat-free mass was different between groups upon reanalysis. Comment on: "Intermittent energy restriction attenuates the loss of fat-free mass in resistance trained individuals. A randomized controlled trial". Journal of Functional Morphology and Kinesiology, 5(4):85. Read on PubMed →
6. Poon ET, Tsang JH, Sun F, Zheng C, Wong SH (2025). Effects of intermittent dieting with break periods on body composition and metabolic adaptation: a systematic review and meta-analysis. Nutrition Reviews, 83(1):59-71. Read on PubMed →
7. Peos JJ, Helms ER, Fournier PA, Krieger J, Sainsbury A (2021). A 1-week diet break improves muscle endurance during an intermittent dieting regime in adult athletes: a pre-specified secondary analysis of the ICECAP trial. PLoS One, 16(2):e0247292. Read on PubMed →
8. Wing RR, Jeffery RW (2003). Prescribed "breaks" as a means to disrupt weight control efforts. Obesity Research, 11(2):287-291. Read on PubMed →
9. National Heart, Lung, and Blood Institute, NIH (1998). Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. NIH Publication 98-4083. Read the guidelines →

The formulas and limits behind MyPlate's own numbers, with their citations, are on the health sources page.