Refeeds, diet breaks and carb cycling are three ways of making a deficit uneven in time, and whether they "work" depends on what you expect them to do. None of them changes the arithmetic of a week: a weekly budget that holds a refeed still carries the same deficit, and a diet break simply pauses the deficit for a fortnight. Five randomised trials have now compared pausing a deficit with dieting straight through. Fat loss came out the same or greater with the pauses, but none produced a faster diet, and the longest found more total loss over a much longer study12345. For carb cycling we found no trial, so that part of this page is arithmetic and stated as such.
The three defined
| Strategy | What it is | Typical shape | What it changes in the arithmetic |
|---|---|---|---|
| Refeed | A short rise in calories, mostly carbohydrate, inside a deficit | 1–3 days a week at or near maintenance | Nothing about the week's total, if the other days absorb it |
| Diet break | A planned pause of the deficit at estimated maintenance | 1–2 weeks at maintenance, then back to the deficit | The planned loss pauses for the length of the break |
| Carb cycling | Carbohydrate moved between days, often towards training days | Higher-carb training days, lower-carb rest days; calories flat or shifted | Nothing, if calories and protein are held; a refeed in disguise if calories move too |
The words get used loosely. A "refeed" that adds calories without taking them from other days is a smaller weekly deficit. "Carb cycling" that raises calories on training days is a refeed. It helps to ask one question of any plan: *what happens to the week's total?*
In plain terms: these are schedules for spending calories, not sources of extra loss.
What a weekly budget already gives you
The calorie deficit calculator produces a daily target, and seven of those is the weekly budget. For the 75 kg example used across these pages (female, 168 cm, 35 years, 4–6 sessions a week): TDEE 2,269 kcal, moderate-rate target 1,857 kcal, weekly budget 12,999 kcal, weekly deficit 2,891 kcal.
A refeed is a redistribution of that budget. Two days at maintenance cost 2 × 2,269 = 4,538 kcal, leaving 8,461 kcal for the other five:
| Plan | Five ordinary days | Two refeed days | Week total | Week deficit |
|---|---|---|---|---|
| Even | 1,857 kcal × 7 | — | 12,999 kcal | 2,891 kcal |
| Two refeeds, budget held | 1,692 kcal × 5 | 2,269 kcal × 2 | 12,998 kcal | 2,892 kcal |
| Two refeeds, budget not held | 1,857 kcal × 5 | 2,269 kcal × 2 | 13,823 kcal | 2,067 kcal |
Row two is a refeed. Row three is a smaller deficit that will deliver about 0.27 kg a week instead of 0.38 at the 7,700 kcal/kg planning value — fine if it is chosen, a surprise if it is not. Two checks apply to row two: the five lower days must stay above the calorie floor (1,200 kcal for this profile — they do, with room), and protein is a daily figure (2.0 g/kg, 150 g here) that does not move with the calories. Daily versus weekly deficits covers the same mechanics from the other direction.
A diet break is simpler still: for one or two weeks the target *is* TDEE. On this plan each week at maintenance defers 0.375 kg of planned loss, so a 12-week diet with a two-week break becomes a 14-week diet for the same result. The calculator has no diet-break setting because it does not need one — run it with the maintain goal for the break and with the lose goal after.
Carb cycling with calories held is invisible to the energy arithmetic. Swapping 50 g of fat (450 kcal) for 112 g of carbohydrate on a training day and back again on a rest day leaves every day's total where it was. The macro rules still want the fat minimum met on the low-fat days — max(0.6 g/kg, 20% of calories), so 45 g at 75 kg — which caps how far the swing can go.
What the evidence shows, stated plainly
The longest diet-break trial is MATADOR1: 51 men with obesity randomised to 16 weeks of energy restriction at 67% of weight-maintenance requirements — a 33% cut, close to the calculator's own 35% cap — either continuously, or in two-week blocks alternated with two-week periods at energy balance. The intermittent group therefore took 30 weeks to accumulate its 16 weeks of dieting. Thirty-six men completed per protocol (19 continuous, 17 intermittent).
| Continuous | Intermittent (2 weeks on / 2 weeks off) | |
|---|---|---|
| Weeks of restriction | 16 | 16 |
| Calendar weeks | 16 | 30 |
| Weight loss | 9.1 ± 2.9 kg | 14.1 ± 5.6 kg |
| Fat mass loss | 8.0 ± 4.2 kg | 12.3 ± 4.8 kg |
| Weight change during the maintenance blocks | — | 0.0 ± 0.3 kg |
| Loss per calendar week (our arithmetic) | 0.57 kg | 0.47 kg |
Source: Byrne et al. 20181; the last row is 9.1 ÷ 16 and 14.1 ÷ 30. The authors also report that, after adjusting for changes in body composition, the reduction in resting energy expenditure was smaller in the intermittent group, and they conclude that interrupting restriction with periods at energy balance "may reduce compensatory metabolic responses". Unadjusted, the drop in resting expenditure did not differ between groups (−502 vs −624 kJ a day, P = 0.48); adjusted for fat and fat-free mass it was −360 vs −749 kJ a day (P = 0.03), and fat-free mass loss was similar1.
Two honest readings sit side by side. The intermittent group lost more weight and more fat per week *of dieting*, held its weight through the breaks, and showed a smaller measured drop in resting expenditure. But per week *on the calendar* the continuous group lost faster, and anyone with a date in mind should notice that the break nearly doubled the length of the study. This is one trial, in men with obesity, at a restriction near the calculator's cap, under supervision. It does not tell us what a break does for a lighter person at the moderate 0.5% rate, and it says nothing about refeeds or carb cycling, which it did not test.
Refeed trials in trained people
A refeed was tested directly in 2020. Twenty-seven resistance-trained men and women were put on the same weekly deficit for 7 weeks with four supervised lifting sessions a week and 1.8 g/kg of protein2. One group cut 25% every day. The other cut 35% for five days, then ate at maintenance for two consecutive days, the extra calories from carbohydrate only2. That is the "two refeeds, budget held" row above, in a lab.
Fat mass fell by about the same in both groups (2.8 vs 2.3 kg on average, no difference). Fat-free mass (everything that is not fat: muscle, water, organs, bone) fell 0.4 kg in the refeed group, not significant, against 1.3 kg in the continuous group, which was2. Resting metabolic rate fell 38 vs 78 kcal a day; the refeed group's drop was not significant, the continuous group's was2.
Change in fat-free mass over 7 weeks of dieting, trained lifters · kg
The authors read the kept resting rate as a side effect of the kept fat-free mass, and note glycogen and insulin were not measured, so the mechanism is a guess2. The refeed group also achieved a slightly larger restriction (26% vs 21%)2.
Diet breaks in lean, trained people have not shown the same edge. In the 12-week ICECAP trial, 61 resistance-trained adults dieted continuously or in three-week blocks separated by a week at energy balance; fat loss, fat-free mass, strength and resting expenditure came out the same, while the break group reported lower hunger and desire to eat3. In 38 resistance-trained women, a week at maintenance after every two weeks of a 25% deficit gave no difference in body composition or resting metabolic rate4. One eating-behaviour score moved apart: disinhibition (a questionnaire measure of losing control of eating) rose with continuous dieting, from 4.91 to 6.17, and fell with breaks, from 6.80 to 6.05, a significant group-by-time difference with no single group's change significant on its own4. The cost: eight weeks instead of six4.
Pooled, the picture is similar. A 2025 meta-analysis of 12 randomised trials (881 participants) found body mass, fat mass, body-fat percentage and waist improved with both approaches, with no significant difference between dieting with breaks and dieting straight through6. Resting metabolic rate fell significantly only with continuous dieting, and the drop was smaller with breaks, more clearly in people with overweight or obesity than in resistance-trained people6.
Randomised trials of pausing a diet, oldest first
- 2018MATADOR · men with obesityTwo weeks on, two off: more fat lost, but over 30 weeks instead of 16
- 2020Trained lifters · 7 weeksTwo-day refeed each week: same fat lost, less fat-free mass lost
- 2021ICECAP · trained adults · 12 weeksWeek-long breaks: same fat and fat-free mass; lower hunger
- 2023Trained women · 6 weeksWeek-long breaks: same body composition and resting rate; disinhibition fell
- 2026Women with obesity · 12 weeksRefeeds plus breaks: same fat lost; fat-free mass trend not significant
For carb cycling we found no human trial to cite, and we will not borrow one from a help centre. What the arithmetic supports is modest: it leaves the week's deficit intact when the budget is held, and pauses of any kind can make a diet easier to keep — more food around training, a day that looks like everyone else's at the weekend. The trials above hint at that (lower hunger, less disinhibition), but it is adherence, not extra fat loss.
In plain terms: a diet break trades time for a gentler diet; a refeed trades calories between days; neither manufactures extra loss.
What the studies found
| Study | Who was studied | Key result | Year |
|---|---|---|---|
| Byrne et al. (MATADOR), *Int J Obes*1 | 51 men with obesity; 33% restriction, 2 weeks on / 2 weeks at maintenance vs continuous | More weight and fat lost, over 30 weeks vs 16; adjusted resting expenditure fell less | 2018 |
| Campbell et al., *J Funct Morphol Kinesiol*2 | 27 resistance-trained men and women, 7 weeks; 2-day refeed each week vs continuous | Same fat loss; fat-free mass −0.4 vs −1.3 kg on average | 2020 |
| Peos et al. (ICECAP), *Med Sci Sports Exerc*3 | 61 resistance-trained adults, 12 weeks of restriction; 1-week breaks vs continuous | Same fat and fat-free mass; lower hunger with breaks | 2021 |
| Siedler et al., *J Hum Kinet*4 | 38 resistance-trained women, 6 weeks of 25% restriction; 1-week breaks vs continuous | Same body composition and resting rate; disinhibition fell with breaks, rose without | 2023 |
| Poon et al., *Nutr Rev*6 | Meta-analysis of 12 randomised trials, 881 participants | No body-composition difference; smaller drop in resting metabolic rate with breaks | 2025 |
| Johnson et al., *Nutrients*5 | 30 women with obesity, 12 weeks of 25% restriction; weekly refeed plus two 1-week breaks vs continuous | Fat loss virtually identical (9.21 vs 9.30 kg); fat-free mass trend not significant | 2026 |
Limits worth knowing
- Small trials. Between 27 and 61 people were randomised in each, and dropouts were common: 58 were recruited in the 2020 refeed trial and 27 analysed2.
- Different people. Men with obesity, trained lifters, trained women and women with obesity respond differently; no trial covers everyone.
- Fat-free mass is not muscle. It includes water and glycogen, which a carbohydrate refeed can shift; the 2020 trial tested at least two days after the last refeed and also reported "dry" fat-free mass to reduce that problem2.
- Unequal deficits. Achieved restriction differed between groups in the 2020 and 2026 trials25.
- Training volume. In the 2023 trial the break group did fewer sets per week (54 vs 72) to match total volume4.
- Industry links. The 2020 trial was funded by a sports-nutrition company institute whose advisory board includes the first author; the funder had no role in the design, data or writing2.
- Abstract only for ICECAP and the 2025 meta-analysis; the other four trials were read in full.
Latest research
The newest trial, published in June 2026, randomised 30 women with obesity to 12 weeks of a 25% deficit, either every day or with a maintenance "refeed" day each week plus two week-long diet breaks5. Fat mass fell almost identically (9.30 vs 9.21 kg on average). Fat-free mass rose in both groups, more with the pauses (+2.40 vs +1.22 kg), but the difference was not significant (p = 0.08), and the authors warn the rise may reflect water rather than muscle5. Measured from body change, the continuous group actually achieved the larger deficit (−1,005 vs −690 kcal a day), and the authors concluded the total energy deficit "remains the primary driver of fat loss"5.
The short version
Refeeds, diet breaks and carb cycling are ways of spending a weekly budget, not modes of the calculator. Two maintenance days in a 12,999 kcal week put the other five at 1,692 kcal and change nothing about the deficit. The MATADOR trial found that two-weeks-on, two-weeks-off dieting lost more in total (14.1 vs 9.1 kg) with a smaller measured drop in resting expenditure, but over 30 weeks rather than 161. Use a break when the diet is sliding and judge it on the trend; expect it to cost calendar time, not to save it.
- A weekly two-day refeed kept more fat-free mass in one small trial of trained lifters, with the same fat loss2.
- Week-long breaks gave the same fat loss as dieting straight through in trained adults and women with obesity345.
- Pauses may make a diet easier to hold (lower hunger, less disinhibition in two trials); none made it faster.
Educational overview only — not medical advice. Zyra is a training and nutrition tracking tool; the rates, caps and floors on this page describe how the calculator behaves, not what any individual should do.
