Prolonged fasting explained: what seven days without food does to weight, muscle and the body's proteins

Twelve healthy adults in Norway drank only water for seven days while researchers took blood from them daily. The study that resulted, led from Queen Mary University of London and published in Nature Metabolism in 2024, is the most detailed account so far of what a week without food does inside the body. It returned to wide attention in the first week of October 2026, when its sponsoring university’s summary was republished as a lead science story.

Multi-day fasting has a growing following, promoted for weight loss and for claimed benefits to metabolism and aging. This explainer sets out what the study found, what other studies show about weight, muscle and lasting effects, and what the risks are.

What prolonged fasting is

A review of human trials from the University of Illinois Chicago gives a working definition. Prolonged fasting is characterized by consciously eating little to no food or caloric beverages for several days to weeks. Some regimens allow water alone. Others, used in European fasting clinics, allow a few hundred calories a day as juice or broth.

It should not be confused with intermittent fasting, which compresses eating into part of the day or skips food on set days. In a review in the New England Journal of Medicine, researchers wrote that eating in a 6-hour period and fasting for 18 hours can trigger a metabolic switch from glucose-based to ketone-based energy. Ketones are fuel molecules the liver makes from fat when carbohydrate runs short. A prolonged fast pushes that switch much further and holds it for days.

The practice is old. Fasting has been used in various forms to help manage conditions including epilepsy and rheumatoid arthritis. What it does beyond changing the body’s fuel has been far less studied. “We still have a very limited understanding of the systemic adaptions in humans to extreme caloric restriction of different durations,” the Queen Mary team wrote.

What the seven-day fasting study found

The researchers recruited 12 volunteers (5 women, 7 men), all healthy. For seven days they consumed water only. Participants were closely monitored each day. Blood was sampled before the fast, every day during it and after eating resumed, and about 3,000 different proteins were measured in each sample.

The scale result was large. A 7-day water-only fast leads to an average weight loss of 5.7 kg, the paper reports, about 12.5 pounds.

The protein measurements were the new part. Proteins in blood come from every organ, so tracking thousands of them gives a readout of how the whole body is responding. For the first two days little changed beyond the expected shift from burning glucose to burning fat. Then the picture altered. The team found nine distinct patterns of response, with systemic changes evident only after 3 days of complete calorie restriction. More than 1,000 proteins eventually rose or fell, and the response is remarkably conserved across volunteers, meaning the 12 people reacted in much the same way.

Some changes had nothing obvious to do with fuel. Proteins that form the scaffolding around cells, including one specific to the brain, shifted markedly.

Claudia Langenberg, who directs the Queen Mary institute that ran the analysis, put the timing at the center of the interpretation. “Our results provide evidence for the health benefits of fasting beyond weight loss, but these were only visible after three days of total caloric restriction - later than we previously thought,” Langenberg said.

The phrase “health benefits” needs unpacking. The team did not measure health. They used genetic databases to predict what a change in each protein might mean for disease, and found candidates pointing in both helpful and harmful directions. The paper itself is careful on this. Those predicted effects, the authors wrote, should be considered exploratory. They added: “our results should not be interpreted as fasting being a sole solution to the disease examples shown”.

The university’s stated hope is for drugs. If the protein changes that matter can be identified, they might be triggered without anyone going a week without food.

How much of the weight lost in a prolonged fast is fat

The 5.7 kg was mostly not fat.

Outcome in the seven-day water-only fast	Average change
Body weight	5.7 kg lost
Lean mass	3.6 kg lost
Fat mass	1.6 kg lost
Lean mass three days after eating resumed	0.69 kg below the start
Body weight three days after eating resumed	3.1 kg below the start

Lean mass covers tissues that are not body fat, such as muscle and other water-rich tissues. It made up nearly two-thirds of the loss, and the scan-based figures for lean and fat tissue do not sum exactly to the change measured on the scale. The rebound was fast too. The loss of lean mass was almost completely reversed during the three days following the end of fasting, while the fat loss persisted. The university’s summary put it the same way: most of the lost lean mass had returned, while the reduction in fat mass remained.

That pattern suggests much of the “lean” loss is water and stored carbohydrate, which refill quickly, more than muscle fiber.

A study from a German fasting clinic tried to separate the components. Healthy men fasted for 10 days on 200 to 250 calories a day while staying physically active. Fat mass and lean soft tissues (LST) accounted for about 40% and 60% of weight loss, respectively. The researchers then broke the lean part down: LST loss was explained by the reduction in extracellular water, meaning fluid outside the cells (44%), muscle and liver glycogen and associated water (14%), and metabolic active lean tissue (42%). Glycogen is the body’s stored carbohydrate. So a little over half the lean loss was water and glycogen, and the remainder was working tissue.

Function held up in that study. Strength was maintained in non-weight-bearing muscles and increased in weight-bearing muscles. The authors’ reading was that protein loss occurs in early fast but decreases as ketogenesis increases, ketogenesis being the production of ketones.

The independent review from Chicago was less relaxed. Across trials, it found, approximately two-thirds of the weight lost is lean mass, and one-third is fat mass. Its authors wrote that the excessive lean mass loss suggests that prolonged fasting may increase the breakdown of muscle proteins, which is a concern.

Whether the benefits of prolonged fasting last

Short-term changes during a long fast are consistent across studies. The Chicago review found that prolonged fasting for 5-20 days produces potent increases in circulating ketones, and mild to moderate weight loss of 2-10%. Systolic and diastolic blood pressure consistently decreased with prolonged fasting. Systolic and diastolic are the upper and lower numbers of a blood pressure reading. In people with normal blood sugar, fasting glucose and insulin fell.

Two results temper that. The first concerns diabetes: these glucoregulatory factors, meaning the blood-sugar measures, remained unchanged in patients with type 1 or type 2 diabetes.

And the gains did not persist. In the few trials that checked later, 3-4 months after the fast was completed, all metabolic benefits were no longer observed, even when weight loss was maintained. A fast changes the body’s chemistry while it lasts, and on current evidence the chemistry reverts.

The largest descriptive dataset comes from a clinic that offers fasting stays. Over a year, 1422 subjects participated in a fasting program consisting of fasting periods of between 4 and 21 days. They were not on water alone, with a daily caloric intake of 200-250 kcal, the calories shown on food labels. Weight, waist size and blood pressure fell, and most participants reported feeling well. The study had no comparison group and measured people only at the start and end of their stay.

What the risks of prolonged fasting are

Every study described here took place under supervision, and that shapes the safety record.

In the Queen Mary study, participants did not report any severe adverse events during daily interviews. Twelve healthy people is a small sample from which to judge safety.

In the clinic study, adverse effects were reported in less than 1% of the participants. A more detailed accounting comes from a residential fasting center in California, which reviewed its charts for 768 visits involving medically supervised, water-only fasting of two days or more. Symptoms were the rule. The majority of adverse events identified were mild, among them fatigue, nausea and insomnia. The more severe category was not rare: in 212 visits it was a grade 3 event, the worst recorded for that visit, on a scale where grade 3 means severe. One visit involved a grade 4, or life-threatening, event, and two visits involved events that met the formal definition of a serious adverse event. There were no deaths.

The Chicago reviewers listed what trials have recorded: metabolic acidosis, headaches, insomnia, and hunger were observed in some studies. Metabolic acidosis is a buildup of acid in the blood, here from high ketone levels. Their overall judgment was that prolonged fasting appears to be a moderately safe diet therapy, a phrase that stops short of safe.

Both of these safety sources come from organizations that provide fasting commercially, as does the 10-day muscle study. Their patients were assessed beforehand and monitored daily. None of this evidence describes fasting alone at home.

Whether a fast of any length is appropriate for a person with a medical condition, or one who takes regular medication, is assessed by their doctor.

Gaps in the evidence on prolonged fasting

The molecular study is small and its authors say so directly: “the sample size (n=12) and the homogenous composition of our volunteers restricts the generalizability of our results as did a missing control group”. In that notation n is the number of participants. Without a comparison group that ate normally under the same conditions, some changes could reflect the setting more than the fast.

A change in a blood protein is not a change in health. The study measured proteins for ten days in total. It did not follow anyone long enough to see whether disease risk moved.

Its results also do not transfer to shorter regimens. The university’s own summary states that the results do not mean that shorter fasting schedules necessarily produce the same effects. If the broad response begins after three days, an 18-hour daily fast is a different intervention.

The wider literature has gaps of its own. The Chicago review is a narrative review, which summarizes trials without pooling them statistically, and the trials are few and short. The largest datasets are observational and come from fasting providers.

A week without food removes about 5.7 kg, most of it lean tissue and water that returns within days, and the molecular changes that follow day three have been observed in 12 healthy volunteers with no measure yet of whether they improve anyone’s health.