Fasting windows help some people eat less, while long fasts add real risks

Key takeaways

• In a 12-month trial of 139 adults with obesity, eating only between 8 a.m. and 4 p.m. on top of a calorie limit led to weight loss that was not significantly different from the calorie limit alone.
• In the largest study of supervised long fasts, 1,422 people at a German clinic lost weight and blood pressure, but their uric acid rose by nearly half, and a company tied to the clinic paid for the study.
• The danger can peak when a long fast ends: refeeding syndrome, a sudden shift of minerals as food returns, can trigger heart rhythm problems and breathing failure.
• Every fasting plan in these studies included water; dry fasting removes that safeguard, and dehydration can damage the kidneys and become life-threatening.
• Skip extended fasts if you take insulin or a sulfonylurea, are pregnant or breastfeeding, are underweight, are under 18, or have had an eating disorder; the National Alliance for Eating Disorders runs a free helpline at 1-866-662-1235.

At the Buchinger Wilhelmi clinic in Überlingen, Germany, guests fast for up to three weeks. Each day brings about 250 calories: a glass of juice at noon, a bowl of vegetable soup at night, perhaps a spoonful of honey, and about three liters of water or herbal tea⁴.

Nurses weigh each guest every morning. Doctors examine them two or three times a week⁴.

In 2019, researchers tied to the clinic published data on 1,422 of these fasters⁴. People lost weight. Blood pressure fell. And 93% said they didn’t feel hungry.

The fine print tells a fuller story. Uric acid, the waste product behind gout, rose by nearly half. A few people developed low sodium or low blood sugar, one had a gout attack, and two went to the hospital⁴. And a company run by the study’s first author, who also sits on the clinic’s board, paid for the research⁴.

That is fasting science in miniature. A daily eating window is a reasonable tool for some people, while longer and drier fasts bring risks that grow with each day.

An eating window works about as well as counting calories

Time-restricted eating means eating all your food within a set number of hours each day, such as 8 or 10, and fasting the rest.

In 2022, researchers in China published a year-long test in the New England Journal of Medicine¹. They randomly assigned 139 adults with obesity to one of two plans for a year. A randomized trial assigns people by chance, so the groups start out alike.

Both groups had the same calorie limit: 1,500 to 1,800 a day for men and 1,200 to 1,500 for women¹. One group also had to eat only between 8 a.m. and 4 p.m.

After 12 months, the eating-window group had lost about 8.0 kg (17.6 pounds) and the calorie-only group about 6.3 kg (13.9 pounds)¹. The gap was not statistically significant, which means chance could explain it. Side effects were similar¹.

Not everyone reads that as a tie. In a 2024 commentary, David Most argued that the eating-window group did better on most measures, and that “not significant” does not prove “no difference”². The window may help a little, but it is not magic.

A second trial shows what happens when the window is the only change. In the TREAT trial, run by the University of California, San Francisco, adults with overweight or obesity either ate only between noon and 8 p.m. or ate three meals a day for 12 weeks, with no calorie advice. The analysis covered 116 of them³.

The window group lost just 0.94 kg, about 2 pounds, no better than the three-meal group³. Among the 50 people tested in person, lean mass, which includes muscle, made up about 65% of what the window group lost, far above the usual 20% to 30%³. The authors flagged that as a concern for people at risk of muscle loss³.

The lesson: an eating window helps if it makes eating less easier for you. It does not seem to work through some special power of the clock.

Inside the biggest study of long, supervised fasts

The Buchinger study followed adults aged 18 to 99 who fasted for about 5, 10, 15 or 20 days⁴. Weight loss ranged from 3.2 kg (about 7 pounds) for five-day fasts to 8.6 kg (about 19 pounds) for 20-day fasts⁴.

Average blood pressure fell from 131.6/83.7 to 120.7/77.9⁴.

The problems were uncommon but real:

• Uric acid rose from about 338 to 495 µmol/L, above the normal range, and peaked in the 15-day group⁴. One man who had been treated for high uric acid before the fast had a gout attack⁴.
• Low sodium showed up in a few people, with the lowest reading at 127 mmol/L. All cases were mild and were treated with salt⁴. The paper counts three cases in one table and six in its text.
• Heart rhythm problems were recorded in 0.21% of fasters, or three people, and low blood sugar in 0.14%, or two⁴.
• Hospital stays: A 75-year-old man had a heart attack on day 9. A 67-year-old woman spent a day in the hospital with vomiting, dizziness and diarrhea⁴.
• Blood thinners: In people taking anticoagulants, a clotting test called the INR rose, and the authors advised close monitoring⁴.

The study had no comparison group, clinic staff collected the data, and the authors wrote that the results may not apply to other places or to people fasting without medical supervision⁴.

The research was paid for by Amplius GmbH. The first author sits on the clinic’s board and, as head of Amplius, leads the clinic’s research. Other authors worked for or consulted for the clinic⁴.

That does not make the data false. It does mean the skepticism you would bring to a drug company’s study belongs here too.

A second safety study points the same way. Researchers reviewed charts from 768 water-only fasting stays at a residential medical facility between 2006 and 2011⁵. Most side effects, about 4,490 events or 75%, were mild and typical of fasting⁵.

Still, 212 stays, more than a quarter, included at least one grade 3 event on a five-step scale where grade 5 means death⁵. One stay reached grade 4, and two had serious adverse events⁵. All of this happened with doctors on hand.

A fasting-mimicking diet was tested by the people who sell it

In 2017, researchers at the University of Southern California published a trial of a fasting-mimicking diet in Science Translational Medicine⁶⁷.

They randomly assigned 100 generally healthy adults to their normal diet or to a five-day meal kit once a month for three months⁶. The kit supplied about 750 to 1,100 calories a day⁷.

Seventy-one people finished three cycles⁶. Compared with the control period, they lost about 2.6 kg (about 6 pounds), and systolic blood pressure, the top number, fell by about 4.5 points⁶. IGF-1, a hormone that affects metabolism and that the team ties to cancer risk, also fell⁶⁷. People who started with higher risk markers improved the most⁷.

No serious side effects were reported. Fatigue, weakness and headaches were the most common complaints⁶.

Now the financial ties. The meal kits came from L-Nutra, a company founded by lead researcher Valter Longo⁷. Longo and another author held equity in it, the university licensed related inventions to it and may earn royalties, and patents were filed⁶. The paper says the two shareholders did not collect or analyze the data⁶.

The trial also measured risk markers, not disease. Longo said larger trials were needed to confirm any effect on preventing or treating illness⁷.

The riskiest part of a long fast can be breaking it

When food returns after a long stretch of starvation, sudden shifts in fluids and minerals can turn dangerous. Doctors call this refeeding syndrome⁸.

Eating carbohydrates releases insulin, which drives phosphate and potassium into cells. Low phosphate is the hallmark, and magnesium and thiamine (vitamin B1) can also run short⁸.

The results can include dangerous heart rhythms, a weakened heart muscle, breathing failure, kidney failure and cardiac arrest⁸. A shortage of thiamine can cause Wernicke-Korsakoff syndrome, a serious brain disorder⁸.

Risk rises with low body weight, unplanned weight loss, long periods of little or no food, heavy drinking, and low mineral levels before eating resumes⁸.

The American Society for Parenteral and Enteral Nutrition advises starting high-risk patients at 10 to 20 calories per kilogram of body weight on the first day, building up over several days, and giving thiamine before feeding begins⁸. Blood minerals are checked every 12 hours for the first three days⁸.

Fasting programs also restart food slowly. At Buchinger, food came back over about four days, rising from 800 to 1,600 calories a day⁴. The water-fasting facility built in a refeeding period half as long as the fast itself⁵.

Picture someone ending a week-long water fast at home with a huge dinner. That is exactly what supervised programs are designed to avoid.

Dry fasting removes the one safeguard every study kept

Dry fasting means no food and no water. Every plan in the studies above kept the water: Buchinger fasters were asked to drink about three liters a day, and the other safety study looked only at water fasts⁴⁵.

Dehydration happens when you lose more fluid than you take in. It can become severe enough to threaten your life, according to MedlinePlus, the National Library of Medicine’s consumer health site⁹.

Very low blood pressure from dehydration is one listed cause of acute kidney failure¹⁰. Dehydration also raises uric acid, on top of the rise seen during fasting in the Buchinger study⁴¹¹.

Older adults face extra risk because they may lose their sense of thirst⁹. Get medical care right away if someone is confused, faints, stops urinating, or has a racing heart or fast breathing⁹.

Claims that dry fasting works faster have no support in these studies, and dehydration makes it the most dangerous version of fasting.

Who should not try an extended fast

These warnings come from the studies above and from federal health sources. If any row applies to you, talk with a clinician before any fast.

If this applies to you	Why it matters
You take insulin, a sulfonylurea or a meglitinide for diabetes	Skipped meals can drop blood sugar below 70 mg/dL; severe lows can cause seizures or loss of consciousness12
You take a diuretic (“water pill”)	Diuretics are linked to low sodium and higher uric acid1113
You have gout or high uric acid	Uric acid rose by nearly half during supervised fasts, and one gout attack occurred4
You take a blood thinner	Clotting test results shifted during fasting, and doses may need adjusting4
You are pregnant or breastfeeding	Listed as reasons not to fast at the Buchinger clinic4
You are underweight or have had anorexia	Listed as reasons not to fast; low weight also raises refeeding risk48
You have advanced kidney or liver disease, or dementia	Listed as reasons not to fast4
You are under 18	The studies here enrolled adults only345

If you have ever had an eating disorder, or if food rules tend to take over your thinking, talk with a clinician who knows your history before you try any fast.

The National Alliance for Eating Disorders runs a free, confidential helpline at 1-866-662-1235. Calls go to licensed therapists who specialize in eating disorders, on weekdays from 9 a.m. to 7 p.m. Eastern. If the lines are full, a therapist calls back within a day¹⁴.

It is not a crisis line, so in an emergency, call 911.

A fasting safety routine: before, during and after

This routine is for healthy adults and does not replace a clinician’s advice.

1. Bring your medicine list to a clinician first. Diabetes drugs, blood pressure pills, diuretics and blood thinners may need changes. At Buchinger, doctors adjusted doses during fasts⁴.
2. Start with the mildest version. An 8-hour eating window was the format tested in the year-long trial¹.
3. Keep drinking water. Never dry fast.
4. Go past a day only with supervision. The studies with low complication rates were done under medical supervision, with nurses weighing fasters every morning at Buchinger⁴⁵.
5. Know your stop signs. Stop and get help for shakiness, confusion or a fast or irregular heartbeat, which can signal low blood sugar¹². Stop for dizziness, fainting, dark urine or little urine⁹. Stop for muscle cramps, nausea or headache that could mean low sodium¹³. Stop for severe pain and swelling in one joint, which can signal gout¹⁵.
6. Call 911 for chest pain, fainting, a seizure or confusion.
7. Break the fast slowly. Use small meals and build up over several days⁴⁵.

How strong is the evidence on fasting claims?

This table uses four levels: strong evidence, promising but early, not enough evidence, and evidence says no (or unsafe).

Claim	Evidence level	What the research shows
An eating window helps you lose weight when you also cut calories	Promising but early	Similar to calorie limits alone over a year, with a possible small edge12
An eating window alone causes meaningful weight loss	Evidence says no	About 2 pounds in 12 weeks, no better than three meals, with lean mass loss3
Supervised multi-day fasts lower weight and blood pressure	Promising but early	Large clinic study, but no comparison group and clinic-linked funding4
A fasting-mimicking diet improves risk markers	Promising but early	One trial of 100 people; the company and university hold financial stakes67
Long fasts prevent disease or extend life	Not enough evidence	Studies tracked short-term markers, not disease or lifespan46
Long water fasts at home without supervision are safe	Evidence says no, or unsafe	Even supervised fasts produced low sodium, low blood sugar, rhythm problems and hospital stays45
Dry fasting is a faster, safer cleanse	Evidence says no, or unsafe	Not tested in these studies; dehydration can cause kidney failure910

Fasting is old wisdom. Treat it like strong medicine

Fasting runs through many faiths and healing traditions, and the body has real tools for going without food. In the Buchinger study, blood sugar settled at a steady level while ketones, a fuel the body makes from fat, rose⁴. That built-in ability may help explain why many healthy adults tolerate a short fast.

Respecting the body also means respecting its limits. That same study recorded rising uric acid, low sodium and two hospital stays, all under daily medical care⁴.

Treat a long fast the way you would treat a powerful herb. Know the dose, know who should skip it, and have someone qualified watching when you push past the familiar.

Questions readers ask

The bottom line

A daily eating window can help some people eat less, but over a year it worked about as well as plain calorie counting. On its own, without a calorie limit, it did little.

Longer fasts are a different matter. Even under daily medical care, they brought rising uric acid, low sodium, low blood sugar and a few hospital stays, and ending a fast carries its own refeeding risk. Dry fasting removes the one safeguard every study kept. If you take diabetes medicines, are pregnant, underweight or under 18, or have had an eating disorder, skip extended fasts.

Related reading

• Faith shapes how people eat, fast and decide, and your clinician should know
• Who paid for the study matters, even when the science is real
• Glucose sensors catch every spike, and in healthy people most spikes are normal
• The symptoms that mean call 911 now
• What to ask when your doctor recommends a treatment

Sources

1. Liu D, Huang Y, Huang C, et al. “Calorie Restriction with or without Time-Restricted Eating in Weight Loss.” New England Journal of Medicine, 2022. pubmed.ncbi.nlm.nih.gov
2. Most DE. “Evaluating Evidence Regarding the Efficacy of Time-Restricted Eating for Weight Loss.” Frontiers in Nutrition, 2024. pmc.ncbi.nlm.nih.gov
3. Lowe DA, Wu N, Rohdin-Bibby L, et al. “Effects of Time-Restricted Eating on Weight Loss and Other Metabolic Parameters in Women and Men With Overweight and Obesity: The TREAT Randomized Clinical Trial.” JAMA Internal Medicine, 2020. pmc.ncbi.nlm.nih.gov
4. Wilhelmi de Toledo F, Grundler F, Bergouignan A, Drinda S, Michalsen A. “Safety, Health Improvement and Well-Being During a 4 to 21-Day Fasting Period in an Observational Study Including 1422 Subjects.” PLOS ONE, 2019. pmc.ncbi.nlm.nih.gov
5. Finnell JS, Saul BC, Goldhamer AC, Myers TR. “Is Fasting Safe? A Chart Review of Adverse Events During Medically Supervised, Water-Only Fasting.” BMC Complementary and Alternative Medicine, 2018. pubmed.ncbi.nlm.nih.gov
6. Wei M, Brandhorst S, Shelehchi M, et al. “Fasting-Mimicking Diet and Markers/Risk Factors for Aging, Diabetes, Cancer, and Cardiovascular Disease.” Science Translational Medicine, 2017. doi.org
7. Gersema E. “Scientifically Designed Fasting Diet Lowers Risks for Major Diseases.” USC News, University of Southern California, 2017. today.usc.edu
8. Persaud-Sharma D, Saha S, Trippensee AW. “Refeeding Syndrome.” StatPearls, National Library of Medicine Bookshelf, updated 2022. ncbi.nlm.nih.gov
9. MedlinePlus, National Library of Medicine. “Dehydration.” medlineplus.gov
10. MedlinePlus Medical Encyclopedia, National Library of Medicine. “Acute Kidney Failure.” Reviewed 2026. medlineplus.gov
11. MedlinePlus Medical Encyclopedia, National Library of Medicine. “Uric Acid – Blood.” Reviewed 2025. medlineplus.gov
12. National Institute of Diabetes and Digestive and Kidney Diseases. “Low Blood Glucose (Hypoglycemia).” National Institutes of Health. niddk.nih.gov
13. MedlinePlus Medical Encyclopedia, National Library of Medicine. “Low Blood Sodium.” Reviewed 2025. medlineplus.gov
14. National Alliance for Eating Disorders. “Helpline.” allianceforeatingdisorders.com
15. Centers for Disease Control and Prevention. “About Gout.” cdc.gov