Evidence is the product.

1. Source hierarchy

LifeMeter prefers primary research, systematic reviews, official statistics, clinical guidelines, trial registries, regulatory documents, and complete source datasets. Secondary reporting may supply context or identify a source, but it does not replace the underlying evidence when that evidence is available.

• Claims with material health implications should link to the supporting paper, dataset, or official document.
• Dates, population, intervention, comparator, endpoint, follow-up, and important limitations should be stated when they alter interpretation.
• Preprints, conference abstracts, company releases, animal studies, and mechanistic work must be identified as such.

2. Claims and uncertainty

Measured findings, interpretation, inference, and projection are separate categories. Association is not described as causation. Biomarker change is not automatically treated as clinical benefit. A result in animals or cells is not presented as evidence of human lifespan extension.

Absolute effects, denominators, uncertainty intervals, study size, and conflicts of interest are included when available and decision-relevant. When sources disagree, the disagreement is described rather than collapsed into false certainty.

3. Calculators, models, and graphics

Calculators identify their intended population, source vintage, important assumptions, and known blind spots. Outputs are estimates or scenarios, not diagnoses. Visuals must preserve the meaning of the underlying data; truncated axes, mixed denominators, and incomparable vintages require explicit treatment.

User-entered health and financial values are not used as editorial evidence. Aggregate product events may be used to improve presentation, but private inputs and calculated results are excluded from analytics.

4. Medical and safety boundaries

LifeMeter does not provide individualized diagnosis, prescribing, treatment, emergency guidance, or a clinician–patient relationship. Content involving drugs, supplements, devices, laboratory tests, or experimental interventions must distinguish regulatory status, evidence quality, known risk, and unknown risk. Readers should consult qualified professionals before acting on information that could materially affect health.

5. Commercial independence

Advertising, sponsorship, affiliate compensation, and report sales do not determine editorial conclusions, rankings, scores, or methodology. Commercial material must be visibly disclosed near the relevant recommendation or placement. Sponsors receive no access to calculator inputs, subscriber identities, or editorial veto. Unsupported disease, reversal, treatment, or lifespan claims are not accepted.

6. Automation and AI-assisted work

Software may assist with discovery, transcription, drafting, coding, formatting, visualization, and quality checks. Generated text or output is never treated as a source. Citations must resolve to real material, calculations must be reproducible, and publication standards remain the same regardless of the tools used.

7. Updates and corrections

Material factual or computational errors are corrected promptly after verification. A substantive revision should identify the changed evidence, data, method, or conclusion and retain the original publication context where practical. Typographical and formatting repairs may be made without a correction note when meaning is unchanged.