For gyms
Reverfit for gymsReverfit for boxesPricingFree
Intermittent fasting
what the evidence shows
The short answer: intermittent fasting leads to about the same weight loss as a normal diet with the same calories, because it works by making you eat less. It improves blood sugar and cholesterol to the extent that you lose weight, no more. With strength training, muscle seems to be maintained. Autophagy, living longer or preventing cancer have not been shown in people. And people who take insulin or sulfonylureas, pregnant women and anyone who has had an eating disorder should not do it without a doctor.
Every figure links to its study and says how strong the evidence is. Looking for what happens at 16, 24 or 36 hours? It is in the fasting clock.
What it is: the most studied forms
- Eating within a window (16:8, 14:10). You eat within a few hours a day and fast for the rest. In studies, windows of 8 to 15 hours (American Diabetes Association, 2026).
- 5:2. Five normal days and two days a week with about 500–600 kcal.
- Alternate-day fasting. One normal day and another with about 500–600 kcal, or without eating.
The fasting clock: what has been measured in people
This is what has been measured in people on a water-only fast, counting from the last meal. These are small studies, almost always with healthy volunteers, young lean men. One is not: Service 2005 studied patients admitted for a 72-hour diagnostic fast. The figures vary quite a lot from one person to another and from one study to another.
- Blood sugar is still normal, about 85–100 mg/dL (Klein 1993; McDougal 2018).
- No ketosis yet: ketones around 0.1 mmol/L (McDougal 2018).
- The liver already makes about half the glucose you use, between 47% and 55%; the rest comes from its glycogen (Landau 1996; Petersen 1996).
- Ketones barely move between 12 and 18 hours (Service 2005).
- More fat starts to come out of storage. The biggest increase comes between 18 and 24 hours: a good part of everything that rises over three days happens then (Klein 1993).
- This is where a 16:8 ends: the switch to burning mainly fat is only just starting.
- Very mild ketosis: about 0.3–0.4 mmol/L; blood sugar, still normal (Browning 2012, 9 women and 9 men).
- Insulin has already completed 70% of its drop (Klein 1993).
- The liver already makes two thirds of the glucose, 64–67% (Rothman 1991; Landau 1996).
- Growth hormone starts to rise (Salgin 2012).
- Ketones rise a lot, but with huge differences: in 10 young men, 3.1 mmol/L (Møller 2009); other studies give quite a lot less even at 48 hours.
- The liver already makes 82% of the glucose; its glycogen runs down until close to 40 hours (Rothman 1991).
- Resting energy expenditure does not drop: it rises 6% (Webber 1994, 29 people).
- More protein starts to be used: at 40 hours urea in the urine goes up and more amino acids come out of muscle (Nørrelund 2001, 8 people).
- Clear ketosis: a median of 1.2 mmol/L in women and 1.9 in men; blood sugar drops to 73–79 mg/dL (Browning 2012).
- The liver makes almost all the glucose, 93% (Landau 1996).
- Growth hormone increases fivefold on the second day, but IGF-1 does not change: it is there to save glucose and protein, not to build muscle (Hartman 1992, 9 men).
- Blood sugar drops 25% from 12 hours (Klein 1993). With a continuous monitor there were periods below 70 mg/dL from the first day: 5 of 6 men had them on the first day, 4 on the second and 5 on the third (McDougal 2018).
- Ketones around 2–2.5 mmol/L (McDougal 2018); glucagon almost triples (Vendelbo 2014).
- Resting energy expenditure: the studies do not agree. One sees it rise 14% (Zauner 2000), another the same as at the start (Webber 1994) and another drop 8% (Nair 1987).
- Muscle keeps releasing amino acids and its building signal (mTOR) falls by half (Vendelbo 2014).
And autophagy at 16, 24 or 36 hours?
Autophagy exists and is proven: it is how the cell recycles its old parts, and its discovery won the 2016 Nobel Prize in Medicine. What does not exist is a figure measured in people of the kind «at X hours of fasting autophagy starts or ramps up». The tables by hour that circulate do not come from studies in humans.
- Measured in people · small studies With muscle biopsies, the results are unclear. At 72 hours one marker rises (LC3B-II), but so does another (p62), and the authors themselves cannot tell whether there is more autophagy or whether it is blocked (Vendelbo 2014, 8 men). With biopsies at 2, 12, 24 and 36 hours, the authors conclude that it changes little in muscle (Dethlefsen 2018). In women with obesity, 24 hours of fasting did not activate it in muscle (Chaudhary 2022). At 72 hours some signals that set it in motion do increase (Bak 2016).
- It cannot be measured in blood with a normal test. The first method to do so dates from 2021 and requires the sample to be treated in the lab (Bensalem 2021). The international guidelines for measuring autophagy warn that a single marker is ambiguous and that measuring how much there really is in patients is «practically impossible» (Klionsky 2021).
- Only in animals Why mouse hours do not apply to you. A mouse runs out of liver glycogen in about 12 hours and, when fasting, loses 12% of its weight in those first 12 hours, 24% a day at that rate; after that it drops to 7% a day, up to 30% at 72 hours. In a person the loss is around 2% on the first day and close to 1% after that. Their days of fasting are equivalent to our weeks (Sokolović 2008). In us, liver glycogen lasts until close to 40 hours (Rothman 1991).
Weight: does it take off more than a normal diet?
- Strong evidence · meta-analyses All forms lead to weight loss, and to a similar degree. A meta-analysis of 99 trials with 6,582 people: compared with the classic diet with daily restriction, only alternate-day fasting lost a little more, 1.3 kg, and in studies of 24 weeks or more there were no differences (BMJ 2025). A Cochrane review of 22 trials finds «little or no difference» compared with classic dietary advice (Cochrane 2026).
- Moderate evidence · trials With the same calories, the same. Over a year, eating from 8:00 to 16:00 lost 8.0 kg and the diet alone 6.3 kg, a difference that was not significant (Liu, NEJM 2022). Alternate-day fasting, 6.0% of body weight versus 5.3% (Trepanowski, JAMA 2017). Eating within 8 hours without counting calories, 4.9% versus 5.3% for cutting calories by 25% (Lin, Annals 2023).
- Moderate evidence · trials Without adjusting what you eat, it may not work. A 16:8 for 12 weeks without set calories lost only 0.26 kg more than the control group, with no real difference (Lowe, JAMA 2020).
- Moderate evidence · trials Where it did come out a bit better, the advantage was small: 2.9 kg in a year with a 4:3 within an intensive support program (Catenacci, Annals 2025) and about 2.3 kg with an early window from 7:00 to 15:00 (Jamshed, JAMA 2022).
Blood sugar, cholesterol and blood pressure
- Strong evidence · meta-analyses The same as with a normal diet. In 28 trials, compared with continuous restriction, there were no differences in glucose, insulin, insulin resistance, cholesterol or blood pressure (Schroor, Adv Nutr 2024).
- Moderate evidence · trials Without eating less, it does not improve. With the same calories, changing only the timing did not improve insulin in two weeks (ChronoFast, Sci Transl Med 2025).
- Moderate evidence · trials Watch out for LDL cholesterol with alternate-day fasting: it rose 11.5 mg/dL compared with the daily diet (Trepanowski 2017).
- Moderate evidence · trials Type 2 diabetes. With a medical protocol to adjust insulin, glycated hemoglobin dropped at 3 months (Obermayer, Diabetes Care 2023), but at two years it was no longer different from the other group (two-year follow-up). Compared with a normal diet, it dropped the same (Pavlou, JAMA Netw Open 2023).
Muscle and people who train
- Strong evidence · meta-analyses With strength training, muscle seems to be maintained. In 8 trials with people who train, muscle mass did not change significantly and fat dropped 1.25 kg. They are few, short studies (Nutrition Research 2026).
- Moderate evidence · trials Maintaining yes, gaining is not clear. In trained men, a 16:8 for 8 weeks lowered fat and maintained muscle and strength, although testosterone and IGF-1 dropped (Moro 2016). In another trial, only the group that did not fast gained muscle (+2.3 kg); strength went up in both (Tinsley 2017).
- Moderate evidence · trials Without training, somewhat more muscle loss: in the 16:8 without set calories, lean mass in the arms and legs dropped more than in the control group (Lowe 2020).
Can people stick with it?
- Strong evidence · meta-analyses People drop it just like other diets. In 15 trials with 1,365 people, dropout was no different from that of the comparison groups (Nutrition Journal 2024).
- Moderate evidence · trials Alternate-day fasting is harder: 38% quit, compared with 29% on the daily diet, and many ate more than planned on fasting days (Trepanowski 2017).
What is claimed and NOT proven in people
- Only in animals «It activates autophagy», the cleaning of cells. Almost everything there is comes from animals and cells (2018 review); what has been measured in people, above. In people there is only one exploratory, weak analysis (Journal of Physiology).
- Only in animals «It extends life». It has been seen in mice (Nature 2024). In people, no trial has measured deaths, heart attacks or strokes (Cochrane 2021).
- «It prevents cancer». There are no trials in people that measure it.
- «It resets your metabolism». Resting energy expenditure did not change in trained men (Moro 2016), and without eating less nothing improves (ChronoFast 2025).
- «It boosts growth hormone and you gain muscle». A two-day fast multiplies it, but that is to save glucose and protein, not to build (1992, 9 men); with a 16:8, IGF-1 even dropped (Moro 2016).
- Weak evidence · observational And the other way round: «eating within 8 hours kills». An observational study associated eating within less than 8 hours with more cardiovascular deaths (AHA conference 2024; published in 2025). It does not prove cause: very few people ate like that, the timing came from two days of survey and the authors acknowledge other factors.
Who should NOT do it, or only with their doctor
- People who take insulin or sulfonylureas for diabetes: episodes of low blood sugar doubled in a trial with 5:2, even when medication was reduced (Corley 2018). The American Diabetes Association asks for medical supervision (ADA 2026).
- Pregnant and breastfeeding women, and people who have or have had anorexia or bulimia (NHS).
- Children, teenagers, older adults and people with heart, kidney or liver disease: check with a specialist first (Spanish Society of Endocrinology and Nutrition).
- Fasts of several days, only with medical supervision: with a continuous monitor there were periods below 70 mg/dL of blood sugar from the first day, in healthy people (McDougal 2018).
- The most frequent side effects in the trials: hunger, constipation, nausea, headache and dizziness, generally mild (BMJ 2025).
What scientific societies say
- Spanish Society for the Study of Obesity (SEEDO): it is equivalent to continuous restriction; the evidence is preliminary and it is not known whether it holds up in the long term (Nutrición Hospitalaria 2024).
- European obesity and dietitian associations (EASO and EFAD): it has no advantage over continuous restriction (Obesity Facts 2023).
If you decide to try it
- What matters is what you eat in total. The timing only helps if it makes you eat less: work out your calories and your protein and fit your meals within the window.
- If you want to keep muscle, do strength training. It is the situation in which the studies see it maintained.
- Choose the form you can stick to. They lead to similar weight loss; the best is the one you can keep up.
- If you take medication, talk to your doctor first.
Frequently asked questions
What happens in the body at 16, 24 and 36 hours of fasting?
Measured in people: at 16–18 hours ketones have barely risen and more fat starts to come out of storage; at 24 hours there is very mild ketosis (0.3–0.4 mmol/L), insulin has completed 70% of its drop and the liver makes two thirds of the glucose; at 36 hours ketones rise a lot, with big differences between people, the liver makes 82% of the glucose and more protein starts to be used.
How many hours of fasting before autophagy starts?
It is not known in people. Autophagy exists, but there is no measurement in humans that says at what hour fasting starts or ramps it up: the tables by hour come from mice, which run through their reserves much faster. With muscle biopsies the results are unclear, and it cannot be measured in blood with a normal test.
Does intermittent fasting make you lose more weight than a normal diet?
No, or very little. With the same calories, trials give similar results: in the largest, a meta-analysis of 99 trials published in the BMJ in 2025, only alternate-day fasting beat the classic diet, by 1.3 kg, and in studies of six months or more there were no differences. It leads to weight loss because it helps you eat less.
Does intermittent fasting activate autophagy and extend life?
It has not been shown in people. Autophagy and longer lifespan have been seen mainly in animals and in cells. In humans there are no trials measuring how long people live, nor heart attacks or cancer.
Do you lose muscle with intermittent fasting?
If you do strength training, muscle seems to be maintained: a 2026 meta-analysis with 8 trials in people who train found no loss. But there is no evidence that it helps you gain it either, and in one trial only the group that did not fast gained muscle mass.
Who should not do intermittent fasting?
Pregnant and breastfeeding women, people who have or have had an eating disorder, under 18s, and people who take insulin or sulfonylureas for diabetes without medical supervision, because it increases the risk of low blood sugar.
Is it dangerous to eat within only 8 hours?
An observational study from 2024, published after peer review in 2025, associated eating within less than 8 hours with more cardiovascular deaths. It does not prove cause: few people ate like that, the timing came from two days of survey and the authors themselves acknowledge other factors. Trials have not found that harm, but they have not measured mortality either.
Keep count, with or without fasting
Reverfit Diet shows you the calories, protein and carbs of what you eat, in whatever meals you have. It has no fasting timer: it counts what you eat, which is what the studies say matters.
Diet
