Burn
Get Burn
Burn
Get Burn
HOME / EVIDENCE POLICY
Last updated: 22 August 2026
Our Evidence Policy explains how we evaluate sources, how we decide what claims the evidence supports, and how we communicate uncertainty.
Burn publishes information about fasting, nutrition, metabolism, weight management, health, fitness, and related topics.
These subjects often involve a mixture of well-established evidence, emerging research, individual variation, and areas where scientific understanding is still developing.
The strength of a claim should reflect the strength of the evidence behind it.
We do not treat all studies, sources, or forms of evidence as equivalent.
A finding from a small preliminary study should not be presented with the same certainty as a conclusion supported by multiple high-quality studies.
Where evidence is limited, mixed, indirect, or uncertain, our language should reflect that.
For scientific and health-related claims, we generally prioritise sources such as:
Where available, we prefer original research over summaries of that research.
Secondary sources can be useful for context, but they should not normally be the primary evidence for significant scientific or health claims.
Different types of evidence can answer different questions, but some provide stronger support for cause-and-effect claims than others.
As a general guide, Burn considers evidence in roughly the following order.
These examine findings across multiple studies and can provide a broader view of the available evidence.
Their quality still depends on the studies included, the methods used, and whether the underlying research is sufficiently comparable.
We do not assume that every meta-analysis automatically provides definitive evidence.
Well-designed randomised controlled trials can provide strong evidence about whether an intervention causes a particular outcome.
We consider factors including:
One trial rarely settles a broad scientific question by itself.
Observational research can identify associations between behaviors and outcomes.
However, an association does not necessarily mean that one factor caused the other.
When relying on observational evidence, Burn should avoid presenting correlation as causation.
Laboratory research can help explain how biological processes may work.
This type of evidence can be useful when discussing metabolism, hormones, energy use, or the physiological response to fasting.
Mechanistic evidence alone does not necessarily demonstrate that a particular intervention will produce a meaningful real-world health outcome.
Animal, cellular, and laboratory research can provide useful early evidence and help researchers understand biological mechanisms.
However, findings in animals or isolated cells do not automatically apply to humans.
When Burn discusses this type of research, it should be clearly identified as such.
Expert interpretation can provide useful context, particularly where evidence is incomplete or complex.
Expert opinion does not replace high-quality empirical evidence.
Where reasonable experts disagree, we should not present one interpretation as universally settled.
Personal experiences, testimonials, individual case reports, and social media accounts can be useful for understanding how people describe their experiences.
They cannot establish that a treatment, fasting method, diet, or behavior causes a particular outcome.
Burn does not use anecdotal evidence as the primary basis for scientific claims.
Where Burn makes claims about effects in humans, we prioritise human evidence.
Animal or laboratory research may be discussed where relevant, but it should not be used to imply that a benefit has been demonstrated in people when it has not.
For example, research showing a biological effect in mice would not justify stating that the same effect has been proven in humans.
The type of study is only one part of evaluating evidence.
We also consider the quality of the research itself.
This may include:
A large study is not automatically a good study, and a peer-reviewed publication is not automatically free from limitations.
Small studies can produce useful early findings, but they are generally less reliable for making broad conclusions.
Smaller samples can make results more sensitive to chance, individual variation, and unusual participant characteristics.
When a finding comes from a small study, we should avoid implying that it applies equally to the wider population.
Short-term studies cannot always tell us what happens over months or years.
This is particularly important when discussing:
A short-term effect should not automatically be described as a long-term benefit.
A result can be statistically significant without necessarily being meaningful in everyday life.
We consider both:
Where possible, Burn should avoid focusing solely on whether a study reached statistical significance.
Health statistics can appear much more dramatic when expressed only as relative changes.
Where relevant and available, we aim to provide sufficient context for readers to understand the actual size of an effect.
For example, a large relative percentage change may represent a small absolute difference.
We should not deliberately select the most dramatic numerical framing.
Many nutrition and health studies are observational.
These studies can identify patterns, but they cannot always determine why those patterns exist.
If people who regularly practice a particular behavior have different health outcomes, other factors may contribute to that difference.
Where research demonstrates an association rather than causation, Burn should use language such as:
We should avoid stronger causal language unless the evidence supports it.
"Intermittent fasting" describes a broad category of eating patterns rather than a single intervention.
Research may study different approaches, including:
Results from one type of fasting should not automatically be applied to every other fasting method.
We consider the specific fasting protocol studied when interpreting research.
When assessing evidence about fasting and weight loss, we consider the comparison being made.
For example, a study may compare intermittent fasting with:
These comparisons answer different questions.
If fasting produces similar weight loss to another calorie-reduction strategy, we should not describe the study as proving that fasting is superior.
Likewise, if a study shows a benefit compared with no intervention, that does not necessarily mean fasting is better than other approaches.
Burn may describe estimated stages associated with fasting.
Human metabolism does not move through perfectly defined stages at identical times for every person.
Changes involving insulin, glycogen use, fat oxidation, ketone production, and other metabolic processes occur along a continuum.
The timing can vary according to factors such as:
For this reason, fasting-stage timings in Burn should be treated as educational estimates rather than exact measurements of an individual's metabolic state.
The production of ketones generally increases as carbohydrate availability decreases and fasting continues, but there is no universal hour at which every person "enters ketosis".
Burn should not present an estimated fasting duration as proof that an individual has reached a particular ketone concentration.
Direct measurement would be required to determine an individual's ketone level at a particular time.
Some biological processes commonly discussed in fasting content are difficult to measure directly in humans.
Where much of the evidence comes from animal, laboratory, mechanistic, or indirect human research, Burn should make that limitation clear.
We should not convert uncertain or emerging evidence into precise claims simply because a particular fasting milestone is popular online.
A change in scale weight does not always represent an equivalent change in body fat.
Short-term weight can be influenced by factors including:
Where Burn discusses weight change, we aim to distinguish short-term scale fluctuations from longer-term changes in body composition.
Nutrition research is particularly difficult to interpret because diet is complex and long-term controlled trials can be difficult to conduct.
When evaluating nutrition claims, we consider:
We avoid treating individual foods or nutrients as having guaranteed effects independent of the broader diet and lifestyle.
A result observed in one group does not necessarily apply equally to everyone.
Research may involve specific populations based on:
Where these differences materially affect interpretation, Burn should make them clear.
Scientific studies do not always agree.
When credible studies reach different conclusions, Burn should not simply select whichever result produces the strongest headline.
We consider the wider body of evidence, including:
If the evidence remains genuinely mixed, we should say so.
Interesting new research can be worth discussing before a scientific consensus exists.
When we cover emerging evidence, we should describe it appropriately.
Phrases such as:
may be appropriate where they accurately reflect the state of knowledge.
We do not use "scientifically proven" as a shortcut for findings that are still preliminary.
Scientific evidence usually describes what happened across groups of people.
It cannot guarantee how a particular individual will respond.
Even where an average effect is well supported, there may be substantial variation between participants.
Burn should distinguish population-level evidence from predictions about an individual user.
We do not normally use the following as sole support for important health or scientific claims:
These materials may sometimes help identify a topic or source, but important claims should be traced back to stronger evidence.
Preprints are research papers that have been made publicly available before completing formal peer review.
They can provide useful information about emerging research, but their findings should be treated more cautiously than established peer-reviewed evidence.
Where a material claim relies on a preprint, its status should be clear.
Research funding does not automatically make a study unreliable.
However, financial relationships can be relevant when evaluating evidence.
Where appropriate, we consider:
We evaluate the methods and evidence rather than automatically accepting or rejecting research based solely on funding.
The wording used by Burn should match the confidence justified by the evidence.
STRONGER EVIDENCE
LESS CERTAIN EVIDENCE
We should not strengthen scientific conclusions for marketing purposes.
When publishing numerical claims, we aim to:
Where a statistic has substantial uncertainty, we should communicate that uncertainty where it matters.
Scientific evidence can inform how Burn explains fasting and related health concepts.
It does not mean that Burn itself has clinically demonstrated a particular health outcome unless Burn has directly been studied for that outcome.
We distinguish between:
Evidence about fasting or human physiology
and
Evidence about Burn as a product.
Research showing that a particular behavior has an effect does not automatically demonstrate that using Burn causes that effect.
Claims about Burn should be supported by evidence appropriate to the claim.
For example:
We should not imply clinical validation where it does not exist.
Scientific understanding changes.
New research can:
We may update Burn content when the evidence materially changes.
An article should not remain unchanged simply because its original sources were accurate when it was first published.
If we identify that a claim is no longer adequately supported, has been misinterpreted, or is based on an unreliable source, we aim to correct it.
Depending on the issue, this may involve:
Our goal is to preserve accuracy rather than defend previous wording.
AI tools may assist with research, organisation, drafting, or editing.
AI output is not itself evidence.
Scientific and health claims should be verified against appropriate underlying sources before publication.
An AI system repeating a claim does not make that claim reliable, regardless of how confidently it is expressed.
For Burn, "evidence-based" does not mean that every scientific question has a definitive answer.
It means that we:
That standard applies whether a finding supports a popular idea about fasting or challenges one.
If you believe a Burn article misrepresents research, relies on an unsuitable source, or overlooks important evidence, we want to know.