Water fasting and extended fasting

• Lose weight quickly.
• Lower blood pressure, in some cases enough to stop medication under a doctor.
• Improve blood sugar, insulin and blood lipids.
• Switch the body into ketosis and trigger autophagy and cell renewal, which proponents link to longevity and disease prevention.
• Reset eating habits before moving to a new diet.

A water fast is a controlled switch from eating to living on stored fuel. The sequence is well mapped, mostly from older studies of people with obesity who fasted for weeks under medical care.

• First day. The liver releases stored glycogen and starts making new glucose, mostly from amino acids taken from protein [3].
• Days 2 to 3. Fat breakdown speeds up and the liver turns fatty acids into ketone bodies. Insulin and the active thyroid hormone T3 fall, while glucagon and reverse T3 rise [3].
• After about 3 days. A 2024 study that measured about 3,000 blood proteins every day of a 7-day water fast found that body-wide changes only appeared after roughly 3 days, and more than 1,000 proteins changed in a consistent way across volunteers [7].
• Weeks. In people who fasted for 5 to 6 weeks, ketone bodies replaced glucose as the brain's main fuel [18]. That shift is what lets the body spare protein during long fasts [19].

Weight comes off fast at first: about 0.9 kg a day in the first week, slowing to 0.3 kg a day by the third week. Most of the early loss is sodium and water, not fat [3]. In the 7-day study, volunteers lost 5.7 kg on average [7].

In mice, repeated multi-day fasts lower the growth signal IGF-1 and trigger blood stem cell renewal [14]. That work drives many of the regeneration claims you will see online, but it has not been shown in people.

Weight loss on a water fast is real and large on the scale. What the scale hides is where the weight comes from.

A 2026 meta-analysis of 49 prolonged fasting studies found falls in body weight, BMI, fat mass and waist size, but the drops in fat-free mass and total body water were bigger than the drop in fat [5]. A review of human trials by Krista Varady's group put it plainly: in fasts of 5 to 20 days, roughly two thirds of the weight lost was lean mass and one third was fat [1]. That is the reverse of a typical slow diet.

Some of that "lean" loss is water and glycogen that come back when you eat. In a careful 10-day study of 16 men on a 200 to 250 kcal modified fast, lean soft tissue made up about 60% of the 5.9 kg lost; extracellular water explained 44% of it, glycogen and its water 14%, and active lean tissue 42%. Strength held up, and leg strength even improved, but lean tissue was still 3.2% below baseline 3 months later [6].

Does it stay off? Follow-up is short. In a randomized pilot in type 2 diabetes, a 7-day modified fast left people 3.5 kg lighter at 4 months, against 2.0 kg with usual care [8]. Six weeks after a supervised water fast and plant-based refeed, weight was still below baseline [13]. Across trials, though, metabolic gains were gone 3 to 4 months later even when weight stayed down [1].

For lasting weight change, the evidence favors patterns you can keep, such as intermittent fasting or moderate calorie restriction.

Blood pressure is the most consistent finding. In 174 consecutive patients with hypertension who fasted on water for about 10 to 11 days, almost 90% ended below 140/90 mm Hg. The average fall was 37/13 mm Hg, and 60/17 mm Hg in people with the most severe hypertension [2]. A newer single-arm trial of 29 people saw median blood pressure fall below 130/80 mm Hg, with drugs stopped, and the effect held for at least six weeks [9]. Reviews agree that blood pressure falls consistently with prolonged fasting [1].

The catch: these are uncontrolled studies from fasting clinics, and the authors include the clinics' owners [2] [9]. The one randomized trial of a week-long fast used a 300 kcal liquid fast, not water only, and found lower blood pressure at 4 months [8].

Cholesterol is less clear. A 14-day modified fast lowered LDL cholesterol by about 28 mg/dL (0.72 mmol/L)0.72 mmol/L (about 28 mg/dL) and triglycerides by 0.35 mmol/L, but HDL also fell, and ApoB, lipoprotein(a) and hs-CRP did not change [10]. After a median 17-day water fast, LDL and hsCRP were lower, yet triglycerides and HOMA-IR were higher after refeeding [11]; six weeks later those two had returned to baseline [13]. A single 24-hour water fast briefly raised total and HDL cholesterol [12].

What this means for a blood test

If you get labs during or just after a fast, expect unusual values: higher ketones, shifted lipids and, after refeeding, sometimes higher insulin resistance. The fasting you do before a routine blood test, usually overnight with water allowed, is a different thing from a water fast.

The best safety data come from clinics that run fasts every day, and they look reassuring for selected, monitored patients.

• A chart review of 768 water-only fasts of at least 2 days found that 75% of adverse events were mild. 212 visits had a grade 3 event, 1 visit had a grade 4 event and 2 had serious adverse events [20].
• In 1,422 people on 4 to 21-day Buchinger fasts at about 200 to 250 kcal a day, adverse effects were reported by under 1% [21].

Both studies come from the clinics themselves, and their authors disclose financial ties to them [20] [21]. Both screened out people at high risk.

The older literature is harsher. When doctors used weeks-long starvation to treat severe obesity in the 1960s, they reported Wernicke's encephalopathy, a brain injury from thiamine deficiency [22], and deaths [23]. A Glasgow series of 18 patients on 60-day fasts called the procedure generally safe but warned of breakdown in electrolyte balance and the need for close blood monitoring [24]. Reviews list gout and uric acid kidney stones, low blood pressure on standing and cardiac arrhythmias as known complications [3].

Refeeding syndrome

The most dangerous moment can be the first meals. After more than 5 to 10 days of little or no food, a sudden load of carbohydrate drives phosphate, potassium and magnesium into cells. Blood levels can crash, causing heart rhythm problems, heart failure and confusion. Clinical guidance flags people as high risk with a BMI under 16, more than 15% weight loss in 3 to 6 months, more than 10 days of little intake or low electrolytes before feeding, and recommends starting at about 10 kcal per kg a day with vitamins given first [4].

The 382-day fast is often cited as proof that long fasts are harmless. That patient was supervised throughout, had blood glucose near 30 mg/dL for months, low magnesium and a spell of high calcium [25]. It shows how far a monitored body can go, not that unsupervised long fasts are safe.

Autophagy, the cell's recycling of damaged parts, ramps up during fasting in animals, and fasting researchers argue it drives many of fasting's benefits [15]. In people, autophagy is hard to measure, and no trial has linked a measured rise in autophagy during a water fast to a health outcome.

The stem cell and immune claims trace largely to one line of mouse research: cycles of prolonged fasting lowered IGF-1, protected blood stem cells and reversed immune suppression from chemotherapy [14]. The 2024 human proteome study found deep, body-wide changes after 3 days, including in brain extracellular matrix proteins, but it could only estimate whether those changes help or harm, and it flagged both kinds [7].

Cancer claims rest on case reports. A woman with low-grade follicular lymphoma had smaller lymph nodes after a 21-day water fast and a plant-food diet [16]. Ten patients who chose to fast before chemotherapy reported fewer side effects, and the authors themselves wrote that this does not establish practice guidance [17]. Low-grade follicular lymphoma can shrink and grow on its own, which is one reason single cases cannot show cause.

What fasting researchers say they know is narrower: fasting reliably lowers weight and blood pressure, and the animal data justify trials [15]. The commercial and clinical push to make fasting easier to tolerate led to the fasting-mimicking diet, which has more randomized human data than water fasting does.

Each of these protocols cuts energy intake. They differ in how hard they are, how long they last and how much evidence backs them.

Krista Varady, who has run many intermittent fasting trials, led the review that found water fasting's metabolic benefits gone within 3 to 4 months [1], and she has said she would suggest intermittent fasting over a water fast for weight loss. A meta-analysis of randomized trials found that intermittent and continuous energy restriction produce similar weight loss [26], and both are far easier to sustain than a week without food.

Modified fasts, such as Buchinger fasting at about 200 to 300 kcal a day, are the version with the most published data, including randomized pilot trials [8] and a 1,422-person cohort [21]. The fasting-mimicking diet goes further, giving about 1,100 kcal on day one and about 720 kcal on days two to five, and has been tested in randomized trials.

Dry fasting, going without water too, is not a stronger form of water fasting. It removes the one thing that keeps a fast from causing dehydration, and fasting researchers, including Valter Longo, have said it is not a safe way to fast.

These are the practices supervised clinics and the clinical literature describe. They are not a substitute for a clinician who knows your health and medications.

1. Check first. People on insulin, sulfonylureas, blood pressure drugs, or diuretics need medication plans, because blood sugar and blood pressure fall during a fast [2] [3].
2. Step up. Start with a 24-hour fast. Clinics often use a short lead-in of light meals; in one series, patients ate only fruits and vegetables for 2 to 3 days before fasting [2].
3. Drink water. Water fasting is not dry fasting. Drink through the day.
4. Rest and stand up slowly. Low blood pressure on standing is a known complication [3]. Avoid driving or hard training if you feel lightheaded.
5. Monitor longer fasts. Beyond a few days, clinics track blood pressure, weight and blood tests for sodium, potassium, magnesium, phosphate, glucose, kidney function and uric acid [24] [3]. Whether you need electrolyte supplements depends on those results.
6. Refeed slowly. Clinics refeed for about half the length of the fast [20]. Start with small, simple meals. People at high risk of refeeding syndrome start at about 10 kcal per kg a day under medical care [4].
7. Plan the after. The benefits faded within 3 to 4 months in trials [1]. The clinics with the best results moved people straight onto a whole-food, plant-based diet [13] [9].

What do days 3 to 5 feel like?

Hunger usually eases once ketosis sets in; in the Buchinger cohort, 93.2% reported no hunger [21]. Headaches, fatigue, insomnia and dizziness are the common complaints [1]. Palpitations, fainting, confusion, chest pain or persistent vomiting are reasons to stop and seek care.

• Breaking a long fast with a large, carbohydrate-heavy meal, which raises the risk of refeeding problems.
• Fasting on insulin, sulfonylureas, diuretics or blood pressure drugs without a clinician adjusting doses.
• Reading the fast scale drop as fat loss; much of it is water, glycogen and lean tissue.
• Repeating long fasts often without tracking muscle mass or eating enough protein between them.
• Going straight back to the old diet, which erases most of the gains.
• Confusing water fasting with dry fasting, which also cuts out water.

• Pregnancy or breastfeeding.
• A current or past eating disorder.
• Underweight (BMI under 18.5), frailty or recent unintended weight loss.
• Type 1 diabetes, or type 2 diabetes treated with insulin or sulfonylureas, unless a clinician supervises and adjusts medication.
• Gout, kidney stones or kidney disease.
• Heart rhythm problems, heart failure, or drugs that affect electrolytes such as diuretics.
• Heavy alcohol use or other reasons for low thiamine.
• Children and teenagers.
• Active cancer treatment, except inside a supervised trial.

• Hunger, headache, fatigue and insomnia, mostly mild.
• Dizziness or fainting on standing from low blood pressure.
• Loss of lean mass, about two thirds of weight lost in 5 to 20-day fasts.
• Raised uric acid, gout flares and uric acid kidney stones.
• Electrolyte disturbances and heart rhythm problems, more likely in long or unsupervised fasts.
• Refeeding syndrome when eating restarts after a long fast.
• Wernicke's encephalopathy from thiamine deficiency, reported in prolonged starvation.