72-Hour Fast

Overview

How it works

What 3 days actually reaches

The tracer study in 6 healthy men gives the shape of the whole fast. Blood sugar fell from 100.5 to 74.6 mg/dL (5.58 to 4.14 mmol/L), about 25%. The rate sugar entered the blood fell from 11.0 to 8.3 micromol/kg/min. Insulin fell from 64.6 to 30.1 pmol/L, roughly half, and 70% of that drop happened in the first 24 hours. Fat release roughly doubled. 60% of that rise came between 12 and 24 hours. The authors concluded that early fat release follows the drop in insulin, not blood sugar. Ketones: in 34 adults tested every 6 hours, the typical change was 0% from 12 to 18 hours, a 333% rise from 18 to 36, 210% from 36 to 54, and 167% from 54 to 72. Using the standard hospital cut-off, 74% of people with a normal test result went above 2.7 mmol/L before 72 hours. The brain measurement is the most direct proof of the fuel switch anyone has in a living person. Ketones at the back of the brain, measured by scanner, went from 0.05 mmol/L when not fasting to 0.60 after the second day and 0.98 after the third. Brain lactate rose from 0.69 to 1.47 mmol/L. Blood and brain levels tracked closely. Autophagy (the cell's clean-up process) is where this fast length earns its place, and also where it needs the most care. In 8 healthy men at 72 hours, a key muscle-building switch (mTOR) dropped about 50%, and related signals (ULK1, 4EBP1 and rpS6) dropped too. One clean-up marker (LC3B-II) rose about 30%. But another marker (p62) also rose about 10% in the same samples. Clean-up breaks p62 down, so if clean-up were really speeding up, p62 should fall. The authors say plainly that this makes it hard to read. They cannot rule out that clean-up bags were piling up because clean-up was blocked, not boosted. In a separate 72-hour water-only study, white blood cells showed the clean-up machinery turned up at the gene and protein level. There was no measure of how fast clean-up ran, and no comparison group. So here is the evidence in people. The upstream signal moves the same way every time. A downstream marker moves in a confusing way. And nobody has measured how fast clean-up runs at this length. That is enough for this page to list autophagy as something this fast engages, but not enough to say by how much. No plan here claims mitophagy (clean-up of the cell's power plants), because no study in people has measured it on any fast.

What studies in people found

The most detailed tracer study in the field, a 6-hourly ketone study in 34 adults, direct brain scans, muscle samples with forearm tests at exactly 72 hours, a white blood cell gene and protein study, and the cancer study that used this length as its longest fast.

• Qian 2021, 72-hour water-only fasting, with white blood cells sorted and studied. 57 people gave blood for tests, and the detailed gene and protein data passed quality checks in 4 of them. The clean-up machinery was turned up at the gene and protein level. There was no measure of how fast clean-up ran and no comparison group.
• Service 2005 tested ketones every 6 hours through the standard 72-hour hospital test fast in 34 adults whose test came back normal, plus 21 people with a confirmed insulin-making tumour (insulinoma). The typical changes were 0%, 333%, 210% and 167% across the four stretches. 74% of people with a normal test went above 2.7 mmol/L before 72 hours.
• Dorff 2016, 20 patients in groups fasting 24, 48 or 72 hours around platinum chemotherapy. The 72-hour group fasted 48 hours before and 24 after. Side effects from fasting were all mild to moderate. DNA damage in white blood cells was lower at 48 hours or more, but not clearly so, and the drop in low white cell counts was not clear either. IGF-1 fell 30%, 33% and 8% in the three groups after the first fast. No group ate normally for comparison.
• Bak 2016, 9 lean and 9 people with obesity compared at 12 and 72 hours, with muscle and belly fat samples. The muscle-building switch (mTOR) signalling fell, and a clean-up starter protein (ULK1) rose. The fat samples measured fat release, not clean-up.
• Vendelbo 2014, 8 healthy men tested after an overnight fast and after 72 hours, with forearm tracer tests and muscle samples. The mTOR switch dropped about 50%, and ULK1, 4EBP1 and rpS6 signals all clearly dropped. LC3B-II rose about 30%, and p62 rose about 10%. The authors say the p62 rise makes it hard to tell how fast clean-up was running.
• Klein 1993, 6 healthy men sampled at 12, 18, 24, 30, 42, 54 and 72 hours, using tracers. Blood sugar fell from 5.58 to 4.14 mmol/L. The rate sugar entered the blood fell from 11.0 to 8.3 micromol/kg/min. Insulin fell from 64.6 to 30.1 pmol/L, with 70% of the drop in the first day. Fat release roughly doubled.
• Pan 2000 measured ketones at the back of the brain in healthy adults by scanner: 0.05 mmol/L when not fasting, 0.60 after two days, and 0.98 after three. Brain lactate rose from 0.69 to 1.47 mmol/L. Blood and brain levels tracked closely.

What this does not tell you: The tracer study had 6 healthy young men and the muscle sample study had 8. The 34-adult ketone study was patients having a hospital test, not healthy people choosing to fast, though their tests came back normal. The white blood cell study had no comparison group and no measure of clean-up speed, and its detailed data passed quality checks in only 4 people. The cancer study had 20 patients and no group eating normally. No study has tested a 72-hour fast in healthy people against a comparison group for any real health outcome.

Reading the research record

What happens to muscle

Doing it

By day three, symptoms stop warning you

Glucose monitors on 12 healthy adults through a 7-day water fast showed time spent below 70 mg/dL rising from 3.0% at the start to 66.0% by day 5. The daily low fell from 76 to 50 mg/dL. No one reported symptoms of low blood sugar. In a healthy person, that is the body adapting to ketones. In anyone on a blood-sugar-lowering drug, it is a lost warning system.

You are losing salt faster than you take it in

Salt in the urine starts rising after about 48 hours and then goes above what you take in. It peaks at day 4 at 109 mmol a day on a fixed intake of 51 mmol a day. Across week one, people lost 203 mEq more than they took in, on average. The salt-holding hormone aldosterone cannot make up for it: given on day 4, it had only about 40% of its normal effect. The one most serious problem in the biggest supervised water-fasting group was low blood salt.

Uric acid rises, and water will not fix it

Uric acid in the blood roughly doubles by day 7, and the rise tracks how deep into ketosis you are. Ketones and uric acid compete for the same exit route in the kidneys, and uric acid clearance drops more than kidney clearance overall. So drinking more water does not fix it. In 1,422 supervised people there was one gout attack, in a man already on allopurinol who had frequent attacks before.

Potassium is where community habits carry real risk

Some plans passed around online use 2,000 mg of potassium a day or more, in a drink. Strong potassium taken quickly can cause heart rhythm problems and gut damage. The risk is much higher with weaker kidneys, or on an ACE inhibitor, ARB or potassium-saving water pill. The community's own guides say to avoid tablets above 99 mg and never to take it all at once.

Training

No study has tested weight training during a 72-hour fast. Community practice is to keep walking, go easier, and not try for personal bests. People describe lifting on days two and three as clearly weaker. What has been measured on longer supervised fasts is that light activity was kept up, and strength in weight-bearing muscles rose across a 10-day low-calorie fast.

Plan the refeed before you start

The most common advice in fasting communities is about behaviour, not the body. Buy and prepare your first food before the fast ends, so you are not deciding while hungry. What people say goes wrong after a long fast is a big or carb-heavy first meal, and the overeating that follows.

Sources

1. Study in people2025
   Effects of seven days' fasting on physical performance and metabolic adaptation during exercise in humans

   Nature Communications

   13 healthy young adults, 6 women and 7 men, finished a seven-day water-only fast, and 2 more dropped out. They lost 5.8 kg, including 4.6 kg of lean mass (8.0%) and only 1.4 kg of fat, and muscle sugar stores halved. Yet top strength did not change, while peak fitness fell 13% and endurance dropped. The authors say it is not clear whether the lean mass lost includes actual muscle fibre.

   13 healthy young adults, 6 women and 7 men, completed a seven day water only fast (15 enrolled, 2 discontinued). Body weight fell 5.8 +/- 0.3 kg. DXA lean mass fell 4.6 +/- 0.3 kg, an 8.0% decrease, with 1.7 +/- 0.5 kg of it from the extremities, a 6% reduction across arms and legs. Fat mass fell 1.4 +/- 0.1 kg. Muscle glycogen halved. Maximal isometric and isokinetic strength were UNCHANGED, while peak oxygen uptake fell 13% and endurance capacity dropped. Electron transport chain protein expression was unchanged, PDK4 rose 13-fold, and AMPK activity was unaffected, which the authors say challenges its proposed role in muscle protein degradation. The authors state that whether fasting protein degradation includes contractile proteins is uncertain.

2. Study in people1993
   Progressive alterations in lipid and glucose metabolism during short-term fasting in young adult men

   American Journal of Physiology

   6 healthy men were tested at points from 12 to 72 hours of fasting. Blood sugar fell from 5.58 to 4.14 mmol/L (100.5 to 74.6 mg/dL), and insulin fell from 64.6 to 30.1 pmol/L, with 70% of that drop in the first 24 hours. Fat release from body stores rose sharply. That early fat release tracked falling insulin, not blood sugar.

   6 healthy men sampled at 12, 18, 24, 30, 42, 54 and 72 hours with stable isotope tracers and indirect calorimetry. Glucose 5.58 ± 0.08 to 4.14 ± 0.10 mmol/L (100.5 to 74.6 mg/dL); glucose rate of appearance 11.0 ± 0.4 to 8.3 ± 0.3 micromol/kg/min; insulin 64.6 ± 12.9 to 30.1 ± 7.9 pmol/L (p < 0.001) with 70% of the decline inside the first 24 hours; glycerol 2.08 ± 0.22 to 4.36 ± 0.36 and palmitic acid 1.63 ± 0.20 to 3.26 ± 0.40 micromol/kg/min. Early lipid mobilisation tracked the fall in insulin, not glucose kinetics.

3. Study in people2005
   Increasing serum betahydroxybutyrate concentrations during the 72-hour fast: evidence against hyperinsulinemic hypoglycemia

   Journal of Clinical Endocrinology and Metabolism

   34 adults did the standard 72-hour hospital fast used to test for an insulin-making tumour and tested negative, alongside 21 whose tumour was confirmed by surgery. In the negative group, ketones checked every 6 hours did not move from 12 to 18 hours. They then rose 333% from 18 to 36 hours, 210% from 36 to 54 and 167% from 54 to 72. 74% passed 2.7 mmol/L before 72 hours.

   Beta-hydroxybutyrate measured every 6 hours through the standard 72 hour diagnostic fast in 34 adults whose fasts were negative for insulinoma, alongside 21 with surgically confirmed insulinoma. Median changes 0% from 12 to 18 hours, +333% from 18 to 36, +210% from 36 to 54 and +167% from 54 to 72. 74% of those with a negative fast exceeded 2.7 mmol/L before the 72 hour mark.

4. Study in people2000
   Human brain beta-hydroxybutyrate and lactate increase in fasting-induced ketosis

   Journal of Cerebral Blood Flow and Metabolism

   Scans measured ketones directly in the brains of healthy adults. Brain ketones went from 0.05 mmol/L when fed to 0.60 after the second day of fasting and 0.98 after the third. Brain lactate also rose, from 0.69 to 1.47 mmol/L. Brain levels closely tracked blood levels.

   Occipital beta-hydroxybutyrate measured directly by 4-tesla magnetic resonance spectroscopy in healthy adults: 0.05 ± 0.05 mmol/L non-fasted, 0.60 ± 0.26 after the second day, 0.98 ± 0.16 after the third. Brain lactate 0.69 ± 0.17 to 1.47 ± 0.22 mmol/L. Plasma and brain correlated at r = 0.86 with a brain-to-plasma slope of 0.26.

5. Study in people2014
   Fasting increases human skeletal muscle net phenylalanine release and this is associated with decreased mTOR signaling

   PLoS One

   8 healthy men were tested after an overnight fast and after a 72-hour fast, with muscle samples. A key muscle-building switch (mTOR) was about 50% less active, and related signals fell too. Two cell clean-up (autophagy) markers rose, by about 30% and 10%. But the authors say those markers are hard to read, and the clean-up process may have been blocked rather than switched on.

   8 healthy men studied postabsorptive and after a 72 hour fast, with forearm amino acid tracer kinetics and muscle biopsies. mTOR Ser2448 phosphorylation fell about 50%, with ULK1 Ser757, 4EBP1 Thr46 and rpS6 Ser235/236 all significantly reduced. LC3B-II rose about 30% and p62 rose about 10%. Because p62 is itself degraded by autophagy, the authors state this makes autophagic flux interpretation problematic and that they cannot exclude autophagosome accumulation and autophagy inhibition rather than induction.

6. Study in people2016
   Differential regulation of lipid and protein metabolism in obese vs. lean subjects before and after a 72-h fast

   American Journal of Physiology, Endocrinology and Metabolism

   9 lean people and 9 people with obesity were compared at 12 and 72 hours of fasting, with muscle and fat samples. A muscle-building switch (mTOR) dialled down, and a clean-up starter protein (ULK1) went up. The fat tissue tests measured fat release, not cell clean-up.

   9 lean and 9 obese subjects compared at 12 and 72 hours with muscle and subcutaneous adipose sampling. mTOR signalling fell and ULK1 protein and messenger RNA rose. The adipose measures were lipolytic rather than autophagy markers.

7. Study in people2021
   Innate immune remodeling by short-term intensive fasting

   Aging Cell

   People did a 72-hour water-only fast, and their white blood cells were studied. 57 gave samples for blood tests, but only 4 passed quality checks for the deepest tests. The cells made more of the cell clean-up (autophagy) machinery. There was no comparison group and no test of how fast the machinery ran, only how much of it there was.

   72 hour water-only fasting with sorted CD45-positive leukocytes, 57 subjects for biochemistry with multi-omics quality control passing in 4. Transcriptomic and proteomic upregulation of autophagy machinery. There was no flux assay and no control group, so this measures the expression of the machinery rather than the rate at which it runs.

8. Animal study2014
   Prolonged fasting reduces IGF-1/PKA to promote hematopoietic-stem-cell-based regeneration and reverse immunosuppression

   Cell Stem Cell

   In mice. This is where the claim that a 72-hour fast rebuilds the immune system comes from, and the stem cell renewal and immune recovery after chemo were all seen in mice. The paper calls its human part early data on protecting white blood cells from chemo in fasting patients. No study has shown that a 72-hour fast rebuilds blood stem cells in people.

   The origin of the claim that 72 hours regenerates the immune system. The hematopoietic stem cell regeneration, the reversal of chemotherapy-induced immunosuppression and the reversal of age-related myeloid bias were all measured in mice. The paper's own abstract describes the human component as preliminary data on the protection of lymphocytes from chemotoxicity in fasting patients. No human study has demonstrated hematopoietic stem cell regeneration after a 72 hour fast.

9. Study in people2017
   Metabolic effects of fasting on human and mouse blood in vivo

   Autophagy

   Blood from 4 to 5 people per group was studied after up to 4 days of fasting. White blood cells showed no rise in cell clean-up (autophagy) in the body, only after they were grown in a dish with a blocking drug, and then only in one cell type. In mice, 48 hours of fasting costs about 20% of body weight. People lose less than 2% over 4 days.

   Human and mouse blood run in the same study, with 4 to 5 subjects per condition. No enhanced autophagy-associated lipidation was detectable in vivo in human leukocytes after up to 4 days of fasting; flux appeared only after ex vivo culture with leupeptin, and among human cell types only in neutrophils. The paper also quantifies the dose mismatch: 48 hours of fasting causes approximately 20% weight loss in mice and less than 2% in humans over 4 days.

10. Study in people2016
    Safety and feasibility of fasting in combination with platinum-based chemotherapy

    BMC Cancer

    20 patients fasted 24, 48 or 72 hours around platinum chemo, the longest being 48 hours before and 24 after. Fasting was doable, and its side effects were all mild to moderate. Less DNA damage in white blood cells and fewer severe drops in white cells were not clear results, and IGF-1 fell 30%, 33% and 8% in the three groups. Nobody ate normally for comparison.

    20 patients in cohorts fasting 24, 48 or 72 hours around platinum-based chemotherapy, the longest arm being 48 hours before plus 24 after. Feasibility was met and fasting-related toxicities were all grade 2 or below. Leukocyte DNA damage by COMET assay was reduced at 48 hours or more at p = 0.08, not significant, and the trend to less grade 3/4 neutropenia was p = 0.17. IGF-1 fell 30%, 33% and 8% in the 24, 48 and 72 hour cohorts after the first fast. No control arm ate normally.

11. Study in people2026
    Marked Increases in Continuous Glucose Monitor-Detected Hypoglycemia During a Seven-Day Water-Only Fast in Healthy Men and Women

    Journal of Diabetes Science and Technology

    12 healthy adults, 7 of them men, average age 29.7, wore glucose monitors through a 7-day water-only fast. Time with blood sugar under 70 mg/dL rose from 3.0% to 66.0% by day 5, and the daily low fell from 76 to 50 mg/dL. Nobody felt any low blood sugar symptoms. So you cannot rely on how you feel to spot low blood sugar during a long fast.

    12 healthy adults, 7 men, mean age 29.7 and mean BMI 25.0, wearing continuous glucose monitors through a 7 day water-only fast. Time below 70 mg/dL rose from 3.0% ± 7.1% at baseline to 66.0% ± 25.7% by day 5 (P < .001) and minimum daily glucose fell from 76 ± 14 to 50 ± 7 mg/dL (P < .001). No symptoms of hypoglycaemia were reported, which makes symptom-based self-monitoring unreliable during an extended fast.

12. Study in people1981
    Adverse effects of liquid protein fast on the handling of magnesium, calcium and phosphorus

    American Journal of Medicine

    6 people with obesity spent 40 days on a liquid protein fast on a hospital research ward. They lost more minerals in urine than they took in: 58% more calcium, 75% more phosphorus and 500% more magnesium. Yet blood phosphorus and magnesium did not change, and blood calcium fell only 0.5 mg/dL. A normal blood test during a fast does not mean your stores are fine, which is why eating again is the dangerous part.

    6 obese subjects studied for 40 days on a metabolic ward. Cumulative urinary losses exceeded intake by 58% for calcium, 75% for phosphorus and 500% for magnesium, with mean daily balances of -104 mg calcium, -48 mg magnesium and -363 mg phosphorus, while serum phosphorus and magnesium did not change and serum calcium fell only 0.5 mg/dL. A normal serum magnesium or phosphate during a fast does not mean the store is intact, which is why refeeding is the dangerous phase.

13. Study in people2022
    Effects of 10-Day Complete Fasting on Physiological Homeostasis, Nutrition and Health Markers in Male Adults

    Nutrients

    13 men lived in a controlled facility through 10 days of water-only fasting, then 4 days of light eating and 5 days of full recovery. They lost 7.28 kg (9.8%), and by fasting day 6 fat was down 10.7% and lean mass 9.2%. By recovery day 5, fat was down 17.2% and lean mass was back where it started. Blood sugar hit its low on day 4, sodium fell below normal on day 9, and the authors suggest a salt supplement.

    13 male adults in a controlled facility: 3 day baseline, 10 day complete water-only fast, 4 days of calorie restriction, 5 days of full recovery, with DXA at baseline, fasting day 6 and recovery day 5. Body weight fell 7.28 kg (9.8%). At fasting day 6, total fat mass was down 10.7% and lean mass 9.2%; by recovery day 5, fat mass was down 17.2% and lean mass had returned to baseline. Glucose reached a nadir of 3.508 ± 0.397 mmol/L on day 4. Serum sodium fell below the 137 mmol/L lower limit on day 9, and the authors conclude that a sodium chloride supplement should be considered.

What people report

These are things people say about their own fasting. Nobody checked them, so they are not evidence. They are here because they tell you what to expect and what to watch for, which the studies do not. They cannot tell you whether anything works.

• Three days seen as the first real long fast and a line of its own, with day two described as the wall and the most common point to quit.
• Physical hunger mostly fading by day three. This is the community's most confident claim about long fasts.
• Appetite hanging on after physical hunger goes. People describe constant thoughts about food with no stomach feeling, and they separate the two carefully.
• Nail-polish breath (acetone) and a metal taste, strong by day two or three.
• Dizziness when standing up that is worse than at 48 hours. Almost everyone advises standing up in stages.
• Feeling cold all the time, with people adding layers indoors.
• A calm, quiet, emotionally flat focus on day three. A smaller group describe brain fog on day three instead, and nobody offers a reason for the split.
• A racing or pounding heart is one of the few signs where the community's usual advice is to take salts and minerals, and if it keeps going, stop instead of pushing through.

Sources: These are self-reports from public fasting communities, with no comparison group, no count of how many people tried, and no follow-up on people who stopped posting. Hunger has a supervised match pointing the same way. In a group of 1,422 people fasting 4 to 21 days on 200 to 250 calories a day, 93.2% said they had no hunger. Feeling faint and trouble sleeping both show up in the supervised review, at 217 and 257 of 768 visits.

Who this is wrong for

• Anyone taking an SGLT2 inhibitor, unless a prescriber has stopped it at least 3 days ahead, or 4 for ertugliflozin. The ketoacidosis (dangerous acid build-up) it causes happens while blood sugar looks normal. A urine ketone strip is a poor test, because it looks for a different ketone. If nausea, vomiting, belly pain, gasping for air or confusion show up, the tests that matter are a blood gas and a blood ketone (beta-hydroxybutyrate) level, not a finger-prick.
• Anyone on insulin, a sulfonylurea or a meglitinide, unless a prescriber changes the dose first. Do not assume a dose cut is enough: 50% cuts in long-acting insulin and 70% cuts in mealtime insulin each still led to a lot of low blood sugar.
• Anyone with type 1 diabetes outside a set program with a continuous glucose monitor and an insulin plan.
• Anyone with primary adrenal insufficiency. A fast calls for more cortisol, and a fixed pill dose cannot rise to meet it.
• Pregnancy and breastfeeding. There is no similar data for fasting over several days in either, and the body needs sugar all the time, not now and then.
• Anyone with a past eating disorder, or who now has episodes of eating they cannot control. One large supervised fasting clinic turns away anyone with severe wasting or anorexia outright.
• Children and teens.
• Anyone with a BMI under 18.5, or who already meets the UK NICE risk list for refeeding problems. A person with a low BMI starts a fast already low on phosphate, potassium, magnesium and thiamine inside the cells.
• Anyone with chronic kidney disease, especially on an ACE inhibitor, ARB or water pill. In 131 fasting kidney patients, a blood marker of kidney strain (creatinine) rose in 60.4% by day 7. People on this family of blood pressure drugs were twice as likely to have the rise. Serious heart problems happened in six fasting patients against one who was not fasting.
• Anyone with known heart disease who is not under supervision. Of the six deaths in the older review of medical fasting, two were in people with severe heart failure before the fast, which at first seemed to get better with fasting. One had a narrowed heart artery found after death.
• Anyone with gout, and anyone taking lithium who will not be drinking freely.
• Anyone with sickle cell trait. It is common, often unknown to the person who has it, and dehydration is the known trigger. A 29-year-old woman with no health problems on record died suddenly while fasting for religious reasons. After death, sickled red cells were found in her lungs, liver and spleen, and a blood test confirmed the trait. That is a single case report, and it counts as one.

Questions

Does a 72 hour fast reset the immune system?

The study that claim comes from measured stem cell regrowth, undoing of chemotherapy damage to the immune system, and undoing of age-related immune changes in mice. Its own summary calls the human part early data on protecting white blood cells from chemotherapy damage in patients who fasted, and those patients were having chemotherapy. The 72 hours was that small study's chosen length, not a proven line. No study in people has shown blood stem cells regrowing after a fast.

Is autophagy actually elevated at 72 hours?

The upstream signal moves the same way every time: the muscle-building switch (mTOR) dropped about 50% in the muscle of 8 healthy men, and a related signal (ULK1) dropped too. The downstream marker is confusing. One clean-up marker (LC3B-II) rose about 30%, but another (p62) rose about 10% in the same samples, and p62 should fall if clean-up speeds up. The authors say they cannot rule out clean-up bags piling up because clean-up was blocked. A separate study found the clean-up machinery turned up in white blood cells at 72 hours, with no measure of clean-up speed and no comparison group. So this fast engages the process, and nobody has measured by how much.

How low does blood sugar go?

About 25% below the 12-hour level: 74.6 mg/dL (4.14 mmol/L) at 72 hours in 6 healthy men. It keeps falling after that, to 3.5 mmol/L on day 4 and 3.2 on day 5. Glucose monitors through a 7-day fast in 12 healthy adults showed a daily low of 50 mg/dL by day 5, with no symptoms of low blood sugar, because the brain is running on ketones. That silence is reassuring in a healthy person and dangerous in anyone on a blood-sugar-lowering drug.

Is a 72 hour fast safe?

The supervised 72-hour fast is a routine hospital test, which is the strongest evidence we have that healthy people handle this length under watch. In a review of 768 supervised water-only fasting visits, typically 7 days long, the worst problem was mild or moderate (grade 2 or lower) in 555 visits and severe (grade 3) in 212. There was one grade 4 problem, a step above severe, and no deaths. The two events formally rated serious were dehydration on fasting day 3 and low blood salt on day 9. Both were in men over 70, and both were caught because a clinician was there.

How much muscle does three days cost?

Nobody has measured body make-up across a 72-hour fast. At 72 hours the forearm muscles leak protein building blocks, mostly because building slows down rather than breakdown speeding up, so the balance is negative. The closest numbers come from 10-day studies. In one, 58% of the scan's lean loss was water outside the cells plus stored sugar and its water. In another, scan lean mass was back to normal after eating again, while fat stayed down.