B — systematic reviews with limitations 5 sources · checked 2026-09-04

Alternate-day fasting

What it is

The strictest of the intermittent-fasting family: one day at roughly a quarter of usual energy, the next without restriction, repeated indefinitely — about three and a half fast days a week. It is the pattern with the best-designed long-term trial in the registry, which is why it is here despite being the one fewest people keep.

How it works in practice

Fast days of five or six hundred kilocalories, usually as one meal; feast days of ordinary eating. In the year-long trial the feast days were not free — participants were asked to eat about 125 percent of their needs, so the week averaged out to the same deficit as the comparison group's daily restriction. Real life rarely structures the feast days, which is where the pattern usually diverges from the trial.

What the evidence shows

A one-year randomised trial in adults with a BMI around 34 compared alternate-day fasting, daily calorie restriction to 75 percent of needs, and no intervention. At the end of the year the fasting group had lost about six percent of starting weight and the daily-restriction group about five — no significant difference — and the maintenance phase showed no difference either. The fasting group lost more participants: about 38 percent dropped out against 29 percent with daily restriction, many citing dissatisfaction with the pattern. The umbrella review of fasting meta-analyses places it among the patterns that reduce weight versus usual eating, with the same caveat as always: the comparisons are mostly not against a matched daily deficit, and where they are, the results are similar.

Who it tends to suit

People who genuinely prefer alternating hard and easy days to a steady moderate one; people with flexible, solitary schedules; people who have tried daily restriction and found the everyday counting worse than the hunger.

How it most often fails

By dropout, which the trial measured; by feast days that outgrow the fast days; and by the rhythm colliding with a social life — half of all invitations fall on a fast day.

What to watch

Protein, which half the days cannot carry; the feast days, logged honestly; and, for anyone on diabetes medication, a doctor first. The eating-disorder note applies with full force to a pattern that alternates restriction and licence by design.

In the app

The app plans every day with the same kind of target and does not schedule fast days; alternate-day fasting is not something it verifies or supports, and the diary will simply show the alternation.

Sources

  1. Trepanowski JF et al., „Effect of alternate-day fasting on weight loss, weight maintenance, and cardioprotection“, JAMA Intern Med 2017;177(7):930–938 — randomized clinical trial (1 година, 3 рамена)
    Limitation: едноцентрово, BMI ~34; „феаст“ дните са били структурирани, не свободни
    cited, not reproduced
  2. Patikorn C et al., „Intermittent fasting and obesity-related health outcomes: umbrella review“, JAMA Netw Open 2021;4(12):e2139558 — umbrella review of meta-analyses of RCTs (GRADE)
    Limitation: сравненията са спрямо обичайно хранене или непостоянни контроли; не доказва предимство пред еквивалентен дневен дефицит
    cited, not reproduced
  3. Alberta Health Services, Adult Obesity Care, Protein — official provincial clinical guideline
    Limitation: препоръчва dietitian; не доказва hard three-meal rule
    cited, not reproduced
  4. CDC, Weight Gain During Pregnancy — official public-health guidance
    Limitation: не е weight-loss solver за pregnancy
  5. Наредба № 1 от 22.01.2018 г. за физиологичните норми за хранене на населението, МЗ, ДВ бр. 11/02.02.2018 — national normative act (Bulgaria)
    Limitation: популационни норми за здрави хора, не за отслабване; отменя Наредба № 23/2005

Educational information, not medical advice. It does not replace a consultation with a doctor or a registered dietitian. Not for people under 18, during pregnancy or breastfeeding, or with a history of an eating disorder — talk to a professional first.