Refeed Day, Diet Break or Calorie Cycling: Which One, and When?

A refeed day is one day at about maintenance calories inside a calorie deficit. A diet break is one to two weeks at maintenance. Calorie cycling moves calories between days while the weekly total stays put. Of the three diet-break trials below, only one found people lost more, and it ran 14 weeks longer. Refeed days and calorie cycling rest on small studies.

What is the difference between a refeed day, a diet break and calorie cycling?

The three differ mainly in length, and in how much testing each has had. Campbell and colleagues, summarizing earlier work, describe a refeed as a brief stretch of eating above dieting level, mostly from carbohydrate. The target is maintenance or a little over. It lasts 1 to 3 days and repeats weekly. A 2019 review of intermittent dieting uses “refeed” and “diet break” for the same idea, a spell of higher intake between spells of eating less. Length is the practical difference. The review adds that in adults with a higher BMI, undoing part of the slowdown may take at least 7 to 14 days at maintenance.

Calorie cycling is the odd one out. Here it means planned high and low days that add up to the same weekly total. You never leave the deficit. You move calories around inside it.

	Refeed day	Diet break	Calorie cycling
Length	1–3 days, every week	1–2 weeks, every few weeks	Set days, every week
Calories	Maintenance or a little over	Maintenance	Higher on some days, lower on others
Weekly deficit	Shrinks unless the other days drop	Zero during the break	Unchanged
Closest trial	27 trained adults, 7 weeks	12 trials pooled, 881 people	74 adults, 6 weeks, different design
Result	Same weight and fat lost; lean-mass gap disputed	No difference in fat lost, slower slowdown	No trial isolates it

Look at what the table leaves out. Nothing in it says to eat well past maintenance to restart your metabolism, and I’d be wary of any page that does. Campbell’s refeed days averaged 19 kcal above maintenance, and none of the trials below tested a big overshoot.

Does a refeed day speed up fat loss?

There’s no measured evidence that it does. The idea runs like this. Dieting pushes down leptin, a hormone tied to appetite and energy use. A day of extra carbohydrate pushes it back up.

The best-known measurement comes from 10 lean women who ate 3 days of extra carbohydrate. Their leptin rose 28% and their 24-hour energy expenditure rose 7%. Three days of extra fat changed neither. The authors found no link between the leptin change and the energy change, so leptin may not be the cause. They weren’t dieting. The test ran 3 days, not 1. Here is my own illustration, not a figure from the paper: on a day that costs 2,000 kcal, 7% is 140 kcal. A refeed that adds 500 kcal would outrun that rise by 360 kcal.

Campbell’s is the one refeed trial I found in lean, trained people. It enrolled 27 men and women. They lifted four days a week for 7 weeks and ate 1.8 g of protein per kg of body weight. One group ate 25% below maintenance every day. The other ate 35% below on five weekdays and at maintenance on the weekend, which also averages 25% across the week (5 × 35 ÷ 7). Weight lost was close: 3.2 kg (7 lb) against 3.6 kg (8 lb). Fat lost was 2.8 kg against 2.3 kg.

Lean mass is where the paper drew its conclusion, and where it was challenged. The refeed group lost 0.4 kg of fat-free mass, the continuous group 1.3 kg. Resting metabolic rate fell 38 kcal a day in one group and 78 kcal in the other, a gap of 40. By my arithmetic, that is 8% of a 500 kcal deficit. The authors wrote that the refeed preserved both. Other researchers reran the shared data and found no significant gap between groups in fat-free mass or resting rate. Only dry fat-free mass, which takes water out of the measure, differed: 0.2 kg lost against 1.9 kg. Campbell’s reply called that reanalysis valid and defensible.

Twenty-seven people for seven weeks is a pilot, and I’d read it that way. The groups began at different body weights, 76.4 kg and 83.1 kg, and seven people’s resting-rate readings came from a second machine. Nobody measured what happened inside the refeed days. The 2019 review noted that 24 to 48 hour refeeds are common in athletes. It said the literature doesn’t suggest such short bouts fully undo the body’s adaptation. Whether they undo part of it was unknown. Trials that measured how large that adaptation is, and how much of it comes back, are covered in starvation mode.

Expect the scale to jump the next morning. Glycogen is the carbohydrate stored in muscle and liver. It’s held with three to four parts water, so refilling 100 g of it adds 400 to 500 g on the scale. The 1992 paper that reported this studied 11 people and warned that carbohydrate loading exaggerates the regain you see. Judge a refeed by the weekly average, never by the morning after.

Do diet breaks work better than dieting straight through?

Not in most trials. One found people lost more, and it needed 14 more weeks.

Trial	Who	Pattern	What happened
MATADOR, 2018	47 men, BMI 30–45; 36 finished as planned	16 weeks at 67% of maintenance, straight or in 2-week blocks with 2-week maintenance gaps	14.1 kg (31 lb) lost against 9.1 kg (20 lb)
ICECAP, 2021	61 trained adults, 32 women	12 weeks of dieting, straight or in 3-week blocks with 1-week breaks	Fat mass 15.3 kg against 18.0 kg at the end, a gap the authors couldn’t tell from chance
Siedler, 2023	38 trained women	25% below maintenance, 6 weeks straight or 1 week at maintenance after every 2	No difference in body composition or resting rate

MATADOR supplied the main meals and snacks. The men were sedentary, and the diet ran at 67% of what they needed, a 33% cut. The break group dieted in eight 2-week blocks with seven 2-week blocks of maintenance between them. Their weight changed by 0.0 kg on average during the gaps (give or take 0.3 kg). They lost 12.3 kg of fat against 8.0 kg. Their resting energy use, adjusted for body composition, fell 360 kJ a day against 749 kJ. That’s 86 kcal against 179 kcal, a 93 kcal gap. These are the figures for the 36 men who finished as planned. Counting all 51 men who were randomized, the gap was smaller: 10.9 kg (24 lb) lost against 7.4 kg (16 lb).

That’s the result I find hardest to wave away and hardest to use. Read it with the price in view. The break group’s plan lasted 30 weeks, not 16, which is 88% longer. A 33% cut is also far deeper than the 500 kcal daily deficit this site works with. Six months later, 28 of the 36 finishers came back (15 and 13). Both groups had regained weight. The break group still sat 11.1 kg (24 lb) below where they’d started the diet, and the other group 3.0 kg (7 lb) below.

ICECAP stretched 12 weeks of dieting to 15 weeks. Lean mass came out identical at 56.7 kg in both groups, and so did leptin, resting energy use and strength. The break group did report less hunger, less desire to eat and more satisfaction. Siedler’s trial saw the same split in eating behavior, on one of eight measures. A questionnaire rates how easily eating gets out of hand. The continuous group’s score rose from 4.91 to 6.17, while the break group’s slipped from 6.80 to 6.05. Neither change was significant on its own.

For the big picture, a meta-analysis of 12 trials and 881 people found weight and fat lost did not differ between breaks and straight dieting. Resting rate fell less with breaks. The ISSN position stand reports the average gap as 47.31 kcal a day. It was about 73 kcal in trials of adults with a higher BMI and about 11 kcal in resistance-trained groups.

My reading is plain. Breaks rarely change how much you lose. What they can do is make a long diet easier to stay on, and the price is time. The arithmetic is easy to run: a 2-week break after 12 weeks of dieting means 14 weeks for 12 weeks of deficit, 17% longer. Spread over those 14 weeks, a 500 kcal deficit averages 429 kcal a day (500 × 12 ÷ 14).

A break after about 12 weeks is my planning rule. None of the three trials dieted that long before pausing. They used blocks of 2 or 3 weeks, so that number is a rule of thumb and no result backs it. The reasoning is in how long to stay in a calorie deficit.

Does calorie cycling beat eating the same every day?

Nothing shows it does. The scale answers to the weekly total, and cycling leaves that untouched.

The nearest evidence is a review of 11 trials, 8 to 24 weeks long, where “fast” days gave 25% or less of daily energy needs. Weight loss matched continuous dieting: a difference of 0.61 kg, with a confidence interval that included zero. Those fast days sit far below the floors of 1,200 and 1,500 kcal this site keeps. Treat the review as a data point, not a plan.

One small trial tested a pattern it called calorie shifting. Seventy-four women were randomized to either 6 weeks of constant restriction or three cycles of 11 days of restriction and 3 days of self-selected eating. The restricted days cut intake by 44% and 51%, far beyond what I’d suggest. At day 70, weight lost was 5.79% against 3.38%, and the authors called for longer, larger studies.

The one practical case I see for cycling is the weekend. In 48 adults aged 50 to 60, weight rose 0.06 kg on each weekend day before any intervention and didn’t rise on weekdays. They ate more on Saturdays and moved less on Sundays. A planned higher weekend with lower weekdays puts that extra eating inside the weekly total instead of leaving it as a surprise.

I ran one week of numbers. Say maintenance is 2,300 kcal and the plan sits 500 kcal below it, a weekly deficit of 3,500 kcal (500 × 7). The AHA/ACC/TOS guideline describes a deficit of at least 500 kcal a day in lifestyle programs. It’s often sought by prescribing 1,200 to 1,500 kcal for women and 1,500 to 1,800 kcal for men. I treat the low end of each range as a floor. Each pattern below keeps the same weekly total.

Pattern	Maintenance 2,300 kcal	Maintenance 1,800 kcal (a woman’s example)
Steady, 500 under	7 days at 1,800	7 days at 1,300
One refeed day	6 days at 1,717, 1 day at 2,300	6 days at 1,217, 1 day at 1,800
Cycling	5 days at 1,700, 2 days at 2,050	5 days at 1,250, 2 days at 1,425

The refeed day isn’t free. The other six days pay for it, and each drops by 583 kcal (500 × 7 ÷ 6). With a maintenance of 1,800 kcal they land at 1,217 kcal, which is 17 above the women’s floor. The right-hand column is arithmetic for a woman and respects her 1,200 kcal floor. A man at that maintenance has a floor of 1,500 kcal, which a 500 kcal deficit already breaks, so the column isn’t for him. If the days would fall under your floor, shrink the deficit or skip the refeed.

Overshooting costs more than it feels like. Say maintenance is 2,300 kcal, the refeed day is planned at 2,300, and it ends at 3,500. That’s 1,200 kcal over plan, and it erases 34% of the week’s deficit (1,200 ÷ 3,500).

To get your own maintenance number from age, height, weight and activity, use the calorie calculator. It uses the NASEM 2023 equations and never sets a target under 1,200 kcal for women or 1,500 kcal for men.

Which one should you pick, and when should you skip all three?

Match the tool to your actual problem:

• Hunger and tiredness are climbing after about 12 weeks. Take a diet break of 1 to 2 weeks at maintenance, then work your maintenance out again with how to find your maintenance calories, because you weigh less than when you started.
• Weekends keep erasing the week. Plan cycling, with a higher weekend and lower weekdays, and check the weekly total in your log rather than any single day.
• You lift weights and want to protect muscle. The two-day carbohydrate refeed is the one design with a trial behind it, a small one whose muscle result is disputed. Training advice is in calorie deficit and working out.
• The scale has been flat for three weeks. Choose none of the three. Check the log first, and why you’re not losing weight in a calorie deficit shows what usually hides there.

Log a refeed day like any other, because leaving it out turns the week’s total into a guess.

These are tools for adults who are already in a calorie deficit. If you have diabetes or take medicine that depends on how much you eat, ask your doctor before changing your eating pattern. If you’re pregnant or breastfeeding, calorie targets are a conversation with your midwife or doctor. And if swinging between low and high days makes food feel like something to manage, stop cycling and eat at maintenance for a while.

For today, subtract your daily target from your maintenance number and multiply by 7. That weekly deficit is the number every pattern has to respect. Write it down and check it against Sunday’s log.

If counting calories has started to feel hard to stop, or eating feels out of control, talk to someone: ANAD in the US (1-888-375-7767), Beat in England (0808 801 0677), NEDIC in Canada (1-866-633-4220) or the Butterfly Foundation in Australia (1800 33 4673). If you’re in danger right now, call your local emergency number. Other UK numbers and crisis lines are on our crisis support page.