Prolonged fasting: read the evidence beyond the headline
A close reading of multi-day fasting studies, body composition, inflammatory signals, supervision and what happens when eating resumes.
Fasting Free educational synthesis. Independent clinical review of this site content has not been completed.
THE QUESTION BEHIND THE HEADLINE
What do human studies actually establish about going several days with little or no food?
They document substantial short-term changes in weight, metabolism and biological signals. They provide much less certainty about lasting health benefits, the best duration, or safety outside the studied setting. The protocol, participants, measurements and clinical support must remain attached to every claim.
Prolonged fasting deserves its own evidence review. An overnight gap, a daily eating window, several days without food and a restricted-food program ask different biological and practical questions. A finding about one cannot simply be carried across to the others because all use the word fasting. Start by naming the actual intervention, then the outcome being claimed.
This guide brings together small physiological experiments, a narrative review, clinic safety records and clinical guidance. Those sources have different jobs. A protein study can reveal a response worth investigating; a clinical record can describe observed problems; guidance can explain why risk assessment matters. None becomes a substitute for the others. Read the source cards alongside the claims, particularly the comparison, follow-up, study access and disclosures. The aim is to make scientific claims understandable, not to provide a multi-day fasting or refeeding program.
BEFORE COMPARING RESULTS
Get the terms right.
- Water-only fasting
No energy-containing food is supplied during the fasting period. In these papers, screening, monitoring and the return to food form part of the study context; the label alone does not describe those safeguards.
- Buchinger or modified fasting
A program allowing small amounts of food or energy-containing drinks. The large Buchinger cohort also involved clinical care and other activities, so its findings should not be relabeled as water-only results.
Read the evidence5- Fasting-mimicking diet
A food-containing dietary intervention designed to reproduce some fasting-related responses. These trials were excluded from the prolonged-water-fasting review. A separate Wei trial is linked on that review’s source card to make the distinction explicit.
Read the evidence3- Lean mass, fat-free mass and muscle
These terms are not interchangeable measurements. Non-fat body components include more than contractile muscle, and body-composition estimates need to be interpreted alongside hydration, function and the measurement method.
CLAIM BY CLAIM
What can the research support?
Do multi-day fasts have proven advantages over other approaches?
Short-term changes are documented. A clear long-term advantage over a well-supported alternative has not been established by the studies assembled here.
First identify the benchmark. A before-and-after observation asks what changed during a program. A randomized comparison would ask what the program adds relative to another approach. Without that comparison, changes in food intake, clinical care, activity, medication management and the passage of time remain possible contributors. The distinction is especially important for an intensive clinic program that changes several features of daily life together.
The narrative review found a small, varied literature and limited follow-up. That makes it useful as a map of questions, but not as a final ranking of treatments. Ask whether the outcome is a number measured while food is withheld, a sustained result after eating resumes, or a health event occurring years later. Evidence for the first does not supply evidence for the other two. A statistically detectable change also needs a meaningful clinical interpretation.
Keep this boundary in view
The available designs do not justify promising prevention of diabetes, cardiovascular disease or premature death from a specific multi-day schedule.
Saved only in this browser, after you choose Save.
Does the seven-day protein study reveal when health benefits begin?
No. A time pattern in measured proteins is a research finding, not a validated countdown to a clinical benefit.
The 2024 experiment repeatedly sampled blood during a seven-day water-only fast. Its broad proteomic response emerged after several days, but the study did not assign different fasting lengths to test which produced better health. Detecting a group-level difference also depends on the measurement platform, sample size, sampling schedule and analysis. A day at which a pattern becomes detectable is not automatically the day a beneficial process switches on in every person.
The researchers also used external genetic data to explore possible disease implications. That creates hypotheses about biological pathways; it does not show that the volunteers experienced fewer diseases. Ask what evidence connects each step: fasting to a measured signal, the signal to a relevant tissue response, and that response to an outcome people value. If the paper measures only part of this chain, the headline should stop at that part rather than turn a mechanism into a promise.
Keep this boundary in view
A small uncontrolled proteomic study cannot establish an ideal fasting duration, a universal autophagy threshold or a human longevity benefit.
Saved only in this browser, after you choose Save.
How should we interpret rapid weight loss and changes in lean mass?
Weight loss, fat loss, lean-mass estimates and physical performance are different outcomes. A change in one does not determine the others.
A scan’s lean-tissue estimate should not be rewritten as an equivalent amount of muscle destroyed. Water and other non-fat components complicate that translation. Conversely, calling all of the change water would also go beyond the evidence. In the proteomics study, much of the measured lean-mass decrease reversed during the short period after eating resumed. That makes the timing of the scan essential to interpreting the result, while leaving longer-term tissue and function questions open.
The separate performance study illustrates why function needs its own measurements: leg-strength tests and aerobic exercise capacity did not respond alike. Preserving one strength measurement does not prove that demanding training, recovery or endurance are unaffected. Nor does an acute reduction in a scan category establish a permanent disability. Look for the actual test, its timing, who completed it and whether recovery was assessed. Different questions need different outcomes rather than one broad claim that fasting either protects or destroys muscle.
Keep this boundary in view
These small studies do not establish the effects of repeated multi-day fasts on muscle, frailty, injury or long-term training adaptation.
Saved only in this browser, after you choose Save.
Does prolonged fasting reliably reduce inflammation or improve brain health?
The response is more complicated than a general anti-inflammatory claim. A newer human study found increased inflammatory and platelet-related signals.
The 2025 study examined a selected clinic sample before fasting, at its end and after supervised refeeding. Its results challenge a simple claim that every fasting-related change must be favorable. Yet the opposite shortcut would also be unjustified: an elevated platelet-related marker is not a measured rate of blood clots. The clinical implications need direct study, with adequate comparison groups and follow-up. A marker can raise a serious research question without answering the size of an individual’s risk.
The paper also discusses molecular pathways related to muscle, bone and amyloid. Those terms can sound like direct measurements of stronger tissue or a healthier brain, but pathway analysis is a different kind of evidence. Lower circulating concentrations of selected amyloid proteins do not demonstrate removal of brain plaques or prevention of dementia. Read what was measured before interpreting the name of the pathway. Also check the denominator: the clinical sample and the subset used for untargeted proteomics were not the same size.
Keep this boundary in view
The single-arm design, small samples and short observation period support cautious biological interpretation, not proven benefit or proven long-term harm.
Saved only in this browser, after you choose Save.
Why can’t a supervised safety study establish that fasting at home is safe?
Selection, clinical monitoring and the ability to intervene are part of what was studied. Removing them changes the situation.
Clinic studies do more than observe a timer. They screen participants, review health and treatment circumstances, monitor changes and make decisions about stopping or responding to problems. A retrospective safety analysis also depends on which visits qualify and what the records capture. Missing laboratory information, excluded visits and unrecorded events limit the completeness of the picture. A result from that selected setting does not estimate the risk of an unmonitored experiment in a different population.
Read adverse-event language precisely. Severity describes how intense an event is; seriousness is a distinct classification that can include hospitalization. Events recorded during a program are not automatically caused by it, especially when participants already have relevant conditions. Equally, “most events were mild” does not mean every visit was uneventful. Keep the number of people separate from the number of visits, and inspect important events rather than relying on an overall reassuring phrase. The linked safety card retains the original denominator and flags a published percentage discrepancy.
Keep this boundary in view
These records cannot prove that symptoms are harmless adaptations, that a commercial program is risk-free, or that a home protocol is equivalent.
Saved only in this browser, after you choose Save.
Why is the period after fasting part of the evidence?
It determines both how nutrition is resumed safely and whether an observed change lasts outside the fasting state.
NICE guidance treats refeeding risk as a clinical assessment involving the history of intake, nutritional status, electrolyte information and other circumstances. It is written for nutrition support, not to certify a fasting challenge. A reader cannot turn one criterion into a boundary below which every fast is safe. Nor does copying a clinic’s food sequence reproduce its assessment and monitoring. The useful lesson is that the return to eating can require professional planning, especially when nutrition has been inadequate.
The research question is separate: what happens to weight, function and laboratory measurements after ordinary eating resumes? Some papers follow participants only through the immediate transition; others have sparse later observations. Food choices and other care during follow-up can influence the result, so those details belong beside any durability claim. Ask whether the reported benefit survives beyond the acute state, how many people were reassessed and whether an alternative approach was compared over the same period. Neither a transient improvement nor a transient worsening can substitute for that follow-up.
Keep this boundary in view
No self-directed refeeding, electrolyte or medication protocol can be derived from this guide. Clinical safety and long-term effectiveness remain different questions.
Saved only in this browser, after you choose Save.
COMPARE THE METHODS
Different designs answer different questions.
A controlled feeding trial, a training study, and clinical guidance are different kinds of evidence. Start with the design before comparing their conclusions.
| Source and design | People and duration | What was compared |
|---|---|---|
| 1. Seven days of fasting: a changing blood proteomeSmall, uncontrolled human intervention with exploratory proteogenomic analyses | Apparently healthy adults; predominantly White European participants12 participants: 5 women and 7 men7-day water-only fast; measurements through 3 days after eating resumed | Repeated measurements against each participant’s baseline; no randomized or nonfasting control group. |
| 2. Inflammation and platelet signals during prolonged fastingSingle-arm clinical study with exploratory molecular analyses and external cohort comparisons | Selected clinic volunteers; mean age 52.2 years and mean BMI 28.820 participants; untargeted proteomics in 15; targeted clinical and urinary measures in 20Mean 9.8 days of water-only fasting, followed by mean 5.3 days of supervised refeeding | Within-person changes; no nonfasting control. External comparisons reused earlier water-only and modified-fasting cohorts. |
| 3. What a small prolonged-fasting literature can tell usNarrative review of human water-only and Buchinger fasting studies | Adults in varied clinical and research settings8 studies; only 2 measured changes in fat and lean massFasting interventions of approximately 5–20 days; limited later follow-up | Mixed designs, mostly without a control group; fasting-mimicking diets and intermittent fasting were excluded. |
| 4. Adverse events in a supervised water-only fasting programRetrospective observational chart review; no randomized comparison | Adults selected for a residential fasting program with clinical supervision768 qualifying visits from 652 patientsFasts ranged from 2–41 days; median 7 days; analysis included supervised refeeding | Recorded events during qualifying visits, without a control group. |
| 5. A large clinic cohort studying modified Buchinger fastingProspective observational clinic cohort | Self-selected adults attending a specialist inpatient clinic1,422 participants selected from 3,929 admissions4–21 days; duration groups were chosen rather than randomly assigned | Before-and-after measurements during modified fasting with small amounts of food, clinical care and other lifestyle activities. |
| 6. Seven-day fasting: strength and endurance respond differentlySmall uncontrolled physiological intervention | Young healthy adults; no prescription-drug or tobacco use15 initially included; 13 completed, including 7 men and 6 women7-day water-only fast; performance tested on day 6 and body composition on day 7 | Each participant’s baseline; free-living activity with daily laboratory visits and professional monitoring. |
| 7. NICE guidance on assessment and refeeding riskClinical guidance for nutrition support and risk assessment | Adults who are malnourished or at risk and need nutrition-support assessment | Clinical assessment guidance, not a trial comparing fasting schedules. |
On a narrow screen, scroll the comparison to see every column.
Compare three core papers in the source libraryTHE PAPERS, WITH A READING PLAN
Read the originals.
Each numbered source connects to the claims above. Open its findings, methods, and disclosures, or go directly to the publication. Access notes describe what we could review.
View all references
These references credit the original publications. The explanations and reading notes in this guide are by Fasting Free.
Systemic proteome adaptions to 7-day complete caloric restriction in humans
Pietzner et al. · Nature Metabolism
A detailed example of discovering biological responses without directly measuring disease prevention.
Read for: Follow the distinction between observed proteins and exploratory disease projections.
Final metadata and institutional accepted manuscript, including methods, reviewed.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Small, uncontrolled human intervention with exploratory proteogenomic analyses. Apparently healthy adults; predominantly White European participants; 12 participants: 5 women and 7 men.
- Duration
- 7-day water-only fast; measurements through 3 days after eating resumed
- Comparison
- Repeated measurements against each participant’s baseline; no randomized or nonfasting control group.
- What was measured
- Plasma proteins, clinical chemistry, weight and DXA body composition; disease implications modeled using external genetic evidence.
- Finding
- Many measured proteins changed, with a broad pattern emerging after several days. Mean weight fell 5.7 kg; much of the measured lean-mass decrease reversed during the brief refeeding follow-up.
- Limits
- Molecular changes do not establish a benefit threshold, disease prevention, or longer life. The small selected sample, lack of a control group, and short follow-up limit causal and clinical conclusions.
- Funding and disclosures
- The authors declared no competing interests.
- Design and scope
- Small, uncontrolled human intervention with exploratory proteogenomic analyses. Apparently healthy adults; predominantly White European participants; 12 participants: 5 women and 7 men.
- Duration
- 7-day water-only fast; measurements through 3 days after eating resumed
- Comparison
- Repeated measurements against each participant’s baseline; no randomized or nonfasting control group.
- What was measured
- Plasma proteins, clinical chemistry, weight and DXA body composition; disease implications modeled using external genetic evidence.
- Finding
- Many measured proteins changed, with a broad pattern emerging after several days. Mean weight fell 5.7 kg; much of the measured lean-mass decrease reversed during the brief refeeding follow-up.
- Limits
- Molecular changes do not establish a benefit threshold, disease prevention, or longer life. The small selected sample, lack of a control group, and short follow-up limit causal and clinical conclusions.
- Funding and disclosures
- The authors declared no competing interests.
Original source: https://pubmed.ncbi.nlm.nih.gov/38429390/
DOI: https://doi.org/10.1038/s42255-024-01008-9
Prolonged fasting promotes systemic inflammation and platelet activation in humans: A medically supervised, water-only fasting and refeeding study
Commissati et al. · Molecular Metabolism
A useful test of the assumption that every fasting-related signal is beneficial.
Read for: Compare assay denominators, the clinical endpoint and the pathway interpretations.
Published full-text sections and institutional journal pre-proof methods reviewed.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Single-arm clinical study with exploratory molecular analyses and external cohort comparisons. Selected clinic volunteers; mean age 52.2 years and mean BMI 28.8; 20 participants; untargeted proteomics in 15; targeted clinical and urinary measures in 20.
- Duration
- Mean 9.8 days of water-only fasting, followed by mean 5.3 days of supervised refeeding
- Comparison
- Within-person changes; no nonfasting control. External comparisons reused earlier water-only and modified-fasting cohorts.
- What was measured
- Inflammatory markers, plasma proteomics, urinary platelet-related markers and metabolic measurements.
- Finding
- Inflammatory and platelet-activation markers increased. Urinary 11-dehydro-TXB2 remained elevated after refeeding. These were biological signals, not observed rates of thrombosis or cardiovascular events.
- Limits
- Small uncontrolled sample, variable durations and clinical medication changes complicate attribution. Inferred muscle, bone and amyloid pathways are not direct demonstrations of preserved tissue or improved brain health.
- Funding and disclosures
- A coauthor founded the fasting clinic; two worked for C2N Diagnostics. Academic/philanthropic support was disclosed; funders reportedly had no role in the findings.
- Design and scope
- Single-arm clinical study with exploratory molecular analyses and external cohort comparisons. Selected clinic volunteers; mean age 52.2 years and mean BMI 28.8; 20 participants; untargeted proteomics in 15; targeted clinical and urinary measures in 20.
- Duration
- Mean 9.8 days of water-only fasting, followed by mean 5.3 days of supervised refeeding
- Comparison
- Within-person changes; no nonfasting control. External comparisons reused earlier water-only and modified-fasting cohorts.
- What was measured
- Inflammatory markers, plasma proteomics, urinary platelet-related markers and metabolic measurements.
- Finding
- Inflammatory and platelet-activation markers increased. Urinary 11-dehydro-TXB2 remained elevated after refeeding. These were biological signals, not observed rates of thrombosis or cardiovascular events.
- Limits
- Small uncontrolled sample, variable durations and clinical medication changes complicate attribution. Inferred muscle, bone and amyloid pathways are not direct demonstrations of preserved tissue or improved brain health.
- Funding and disclosures
- A coauthor founded the fasting clinic; two worked for C2N Diagnostics. Academic/philanthropic support was disclosed; funders reportedly had no role in the findings.
Original source: https://pubmed.ncbi.nlm.nih.gov/40268190/
DOI: https://doi.org/10.1016/j.molmet.2025.102152
Efficacy and safety of prolonged water fasting: a narrative review of human trials
Ezpeleta et al. · Nutrition Reviews
A compact map of the early human literature and its missing comparisons.
Read for: Check selection criteria and the small evidence base behind broad summary statements.
Full text and study-selection methods reviewed.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Narrative review of human water-only and Buchinger fasting studies. Adults in varied clinical and research settings; 8 studies; only 2 measured changes in fat and lean mass.
- Duration
- Fasting interventions of approximately 5–20 days; limited later follow-up
- Search boundary
- The reviewed methods do not state an explicit search cutoff. Online publication in June 2023 predates the 2024–2025 primary papers in this guide.
- Comparison
- Mixed designs, mostly without a control group; fasting-mimicking diets and intermittent fasting were excluded.
- What was measured
- Weight and metabolic markers during fasting and, where available, after refeeding.
- Finding
- Short-term weight and some marker changes were reported, with variable lipid results and limited evidence of lasting improvements.
- Limits
- This is not a pooled meta-analysis. Sparse body-composition and follow-up data do not establish a universal muscle-loss fraction or long-term safety; an explicit search cutoff was not supplied.
- Funding and disclosures
- No external funding reported. One author disclosed fees from publishers of diet books.
- Design and scope
- Narrative review of human water-only and Buchinger fasting studies. Adults in varied clinical and research settings; 8 studies; only 2 measured changes in fat and lean mass.
- Duration
- Fasting interventions of approximately 5–20 days; limited later follow-up
- Search boundary
- The reviewed methods do not state an explicit search cutoff. Online publication in June 2023 predates the 2024–2025 primary papers in this guide.
- Comparison
- Mixed designs, mostly without a control group; fasting-mimicking diets and intermittent fasting were excluded.
- What was measured
- Weight and metabolic markers during fasting and, where available, after refeeding.
- Finding
- Short-term weight and some marker changes were reported, with variable lipid results and limited evidence of lasting improvements.
- Limits
- This is not a pooled meta-analysis. Sparse body-composition and follow-up data do not establish a universal muscle-loss fraction or long-term safety; an explicit search cutoff was not supplied.
- Funding and disclosures
- No external funding reported. One author disclosed fees from publishers of diet books.
Original source: https://pubmed.ncbi.nlm.nih.gov/37377031/
DOI: https://doi.org/10.1093/nutrit/nuad081
Is fasting safe? A chart review of adverse events during medically supervised, water-only fasting
Finnell et al. · BMC Complementary and Alternative Medicine
Shows how recorded harms are classified in a supervised setting.
Read for: Distinguish visits from patients, severe from serious events, and observation from causation.
Open-access publisher full text and methods reviewed.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Retrospective observational chart review; no randomized comparison. Adults selected for a residential fasting program with clinical supervision; 768 qualifying visits from 652 patients.
- Duration
- Fasts ranged from 2–41 days; median 7 days; analysis included supervised refeeding
- Comparison
- Recorded events during qualifying visits, without a control group.
- What was measured
- Adverse-event type and severity; hospitalizations.
- Finding
- Most recorded events were mild or moderate; two visits involved hospitalization, for dehydration and hyponatremia. A grade-3 event was the highest grade in 212 visits.
- Limits
- Selected visits, incomplete laboratory data and retrospective records cannot establish safety at home. Report counts: the article’s printed 0.002% for 2 of 768 visits is inconsistent with the arithmetic (approximately 0.26%, calculated).
- Funding and disclosures
- Clinic leadership and paid TrueNorth Foundation consulting relationships were disclosed, alongside a National Health Association grant.
- Design and scope
- Retrospective observational chart review; no randomized comparison. Adults selected for a residential fasting program with clinical supervision; 768 qualifying visits from 652 patients.
- Duration
- Fasts ranged from 2–41 days; median 7 days; analysis included supervised refeeding
- Comparison
- Recorded events during qualifying visits, without a control group.
- What was measured
- Adverse-event type and severity; hospitalizations.
- Finding
- Most recorded events were mild or moderate; two visits involved hospitalization, for dehydration and hyponatremia. A grade-3 event was the highest grade in 212 visits.
- Limits
- Selected visits, incomplete laboratory data and retrospective records cannot establish safety at home. Report counts: the article’s printed 0.002% for 2 of 768 visits is inconsistent with the arithmetic (approximately 0.26%, calculated).
- Funding and disclosures
- Clinic leadership and paid TrueNorth Foundation consulting relationships were disclosed, alongside a National Health Association grant.
Original source: https://link.springer.com/article/10.1186/s12906-018-2136-6
DOI: https://doi.org/10.1186/s12906-018-2136-6
Safety, health improvement and well-being during a 4 to 21-day fasting period in an observational study including 1422 subjects
Wilhelmi de Toledo et al. · PLOS ONE
Demonstrates how a modified-fasting program differs from complete food abstention.
Read for: Identify the clinical setting, accompanying activities and nonrandom duration choices.
Open-access publisher full text, methods and registry entry reviewed.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Prospective observational clinic cohort. Self-selected adults attending a specialist inpatient clinic; 1,422 participants selected from 3,929 admissions.
- Duration
- 4–21 days; duration groups were chosen rather than randomly assigned
- Comparison
- Before-and-after measurements during modified fasting with small amounts of food, clinical care and other lifestyle activities.
- What was measured
- Weight, blood pressure, laboratory markers, reported well-being and adverse events.
- Finding
- Several short-term measurements and well-being scores improved. Two participants were hospitalized. This was modified fasting, not a water-only intervention.
- Limits
- No nonfasting comparison; selection, co-interventions and unblinded assessment limit causal claims. This cohort was reused in later analyses and should not be counted repeatedly as independent participants.
- Funding and disclosures
- Funded by Amplius on behalf of the Buchinger clinics; author leadership, employment and consulting ties were disclosed.
- Design and scope
- Prospective observational clinic cohort. Self-selected adults attending a specialist inpatient clinic; 1,422 participants selected from 3,929 admissions.
- Duration
- 4–21 days; duration groups were chosen rather than randomly assigned
- Comparison
- Before-and-after measurements during modified fasting with small amounts of food, clinical care and other lifestyle activities.
- What was measured
- Weight, blood pressure, laboratory markers, reported well-being and adverse events.
- Finding
- Several short-term measurements and well-being scores improved. Two participants were hospitalized. This was modified fasting, not a water-only intervention.
- Limits
- No nonfasting comparison; selection, co-interventions and unblinded assessment limit causal claims. This cohort was reused in later analyses and should not be counted repeatedly as independent participants.
- Funding and disclosures
- Funded by Amplius on behalf of the Buchinger clinics; author leadership, employment and consulting ties were disclosed.
Original source: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0209353
DOI: https://doi.org/10.1371/journal.pone.0209353
Effects of seven days’ fasting on physical performance and metabolic adaptation during exercise in humans
Kolnes et al. · Nature Communications
Places physical tests beside body-composition estimates.
Read for: Keep strength, aerobic capacity and recovery as separate questions.
Full text, methods, outcome tables and disclosures reviewed.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Small uncontrolled physiological intervention. Young healthy adults; no prescription-drug or tobacco use; 15 initially included; 13 completed, including 7 men and 6 women.
- Duration
- 7-day water-only fast; performance tested on day 6 and body composition on day 7
- Comparison
- Each participant’s baseline; free-living activity with daily laboratory visits and professional monitoring.
- What was measured
- DXA body composition, leg strength, exercise capacity, muscle biopsies and metabolism.
- Finding
- Measured leg strength was unchanged, while absolute peak oxygen uptake fell 13% and exercise performance declined. Lean-mass estimates also decreased.
- Limits
- Completer analysis, no control group and no long-term recovery assessment. Preserved strength does not establish absence of protein loss or suitability for training; this was a different cohort from the 2024 proteomics paper.
- Funding and disclosures
- Norwegian School of Sport Sciences funded the clinical work, with additional foundation/MRC support. Authors disclosed Pfizer/Novo Nordisk employment, consulting or shareholding relationships.
- Design and scope
- Small uncontrolled physiological intervention. Young healthy adults; no prescription-drug or tobacco use; 15 initially included; 13 completed, including 7 men and 6 women.
- Duration
- 7-day water-only fast; performance tested on day 6 and body composition on day 7
- Comparison
- Each participant’s baseline; free-living activity with daily laboratory visits and professional monitoring.
- What was measured
- DXA body composition, leg strength, exercise capacity, muscle biopsies and metabolism.
- Finding
- Measured leg strength was unchanged, while absolute peak oxygen uptake fell 13% and exercise performance declined. Lean-mass estimates also decreased.
- Limits
- Completer analysis, no control group and no long-term recovery assessment. Preserved strength does not establish absence of protein loss or suitability for training; this was a different cohort from the 2024 proteomics paper.
- Funding and disclosures
- Norwegian School of Sport Sciences funded the clinical work, with additional foundation/MRC support. Authors disclosed Pfizer/Novo Nordisk employment, consulting or shareholding relationships.
Original source: https://pmc.ncbi.nlm.nih.gov/articles/PMC11695724/
DOI: https://doi.org/10.1038/s41467-024-55418-0
Nutrition support for adults: oral nutrition support, enteral tube feeding and parenteral nutrition (CG32)
National Institute for Health and Care Excellence · NICE
Explains the need for clinical judgment when nutrition is restarted.
Read for: Read the scope of the guidance before treating any criterion as personal clearance.
Official guidance PDF and relevant recommendations reviewed.
Saved only in this browser, after you choose Save.
Findings, methods, and disclosures
- Design and scope
- Clinical guidance for nutrition support and risk assessment. Adults who are malnourished or at risk and need nutrition-support assessment.
- Comparison
- Clinical assessment guidance, not a trial comparing fasting schedules.
- What was measured
- Recognition of malnutrition and refeeding risk, and professionally managed nutrition support.
- Finding
- Risk assessment considers nutritional intake, weight history, body size, electrolytes and other clinical factors. High-risk care requires appropriately trained professionals.
- Limits
- The criteria are not a home-fasting clearance tool, a universal safe-duration threshold or a self-directed refeeding recipe. The guidance is not an endorsement of prolonged fasting.
- Design and scope
- Clinical guidance for nutrition support and risk assessment. Adults who are malnourished or at risk and need nutrition-support assessment.
- Comparison
- Clinical assessment guidance, not a trial comparing fasting schedules.
- What was measured
- Recognition of malnutrition and refeeding risk, and professionally managed nutrition support.
- Finding
- Risk assessment considers nutritional intake, weight history, body size, electrolytes and other clinical factors. High-risk care requires appropriately trained professionals.
- Limits
- The criteria are not a home-fasting clearance tool, a universal safe-duration threshold or a self-directed refeeding recipe. The guidance is not an endorsement of prolonged fasting.
Original source: https://www.nice.org.uk/guidance/cg32/chapter/Recommendations
PUT THE EVIDENCE TO USE
Practice asking the better question.
Hypothetical headline: “Benefits begin on day three”
An imagined article turns the emergence of a protein pattern into advice that everyone must fast for at least three days. Its illustration shows a switch flipping from ineffective to beneficial.
- Did the study compare clinical benefits at different durations?
- Was the plotted result a group average, an individual threshold or a statistical detection boundary?
- What evidence connects the proteins to an actual improvement in health?
The defensible interpretation describes the measured response in the tested volunteers. To justify the headline, researchers would need evidence about the claimed benefit, the relevant comparison and the proposed threshold. A molecular timeline can motivate that work, but it cannot complete it. A clearer article would name the protein findings and the uncertainty, without attaching a minimum fasting target. This reasoning does not deny that timing matters; it identifies the additional evidence required to turn a pattern into a recommendation.
Hypothetical headline: “Most events were mild, so it is safe”
A fictional reader sees a summary of a supervised program and considers copying its duration at home. The summary omits the participant screening, clinical decisions and follow-up arrangements.
- Who was eligible, and which visits were excluded from the analysis?
- Were important problems captured even if the overall majority of events were mild?
- What medical support was available, including when food was reintroduced?
The missing context changes what the claim means. A program can have mostly mild recorded events and still include clinically significant problems. Its results also cannot estimate risk when different people follow a different process without the same support. The appropriate interpretation is about the observed setting, with its selection and recording limitations. An individual clinical conversation would need the person’s health, treatment and nutritional circumstances; a count of days alone cannot supply that information.
WHERE THE EVIDENCE STOPS
What would make the answer more convincing?
- Which outcomes improve when prolonged fasting is compared with an equally supported alternative over months rather than days?
- How much of an acute body-composition change persists after recovery, and how does it relate to physical function?
- Do the inflammatory and platelet-related signals translate into meaningful clinical outcomes, and for whom?
- How much do post-fast food choices, medication changes and accompanying care contribute to reported results?
- What does complete, prospective adverse-event monitoring reveal in populations underrepresented in existing studies?
Questions worth taking into a conversation.
Use these to prepare for an individual discussion with a qualified clinician, sports dietitian, or coach as appropriate to the question.
- Which outcome matters in the claim you are evaluating: a marker, body composition, function or a health event?
- Does the paper study the same protocol and population that the headline describes?
- What would a clinician need to know about health history, medicines, nutrition and eating-disorder concerns before assessing suitability?
- Which finding remains uncertain after the fast ends, and what study would help resolve it?