Methodology

This document describes exactly how Longevity Lab Report grades ingredients and stacks, how it handles expert positions, and how the record stays current. It is written to be read by anyone, including a sponsor's legal team. If any practice described here changes, this document is updated first and the change is noted in the version history at the bottom.

1Purpose and scope

Longevity Lab Report is an independent record of what the evidence says about longevity ingredients and the stacks built from them. It grades ingredients (compounds such as creatine or magnesium), never brands, companies, or products. It reports what studies and experts say. It does not provide medical advice, diagnose, treat, or recommend that any individual take anything.

The product has three layers: (1) the weekly pulse of what changed in the evidence, (2) debates showing where experts agree and disagree, and (3) interactive tools, the stack builder, the worth-it leaderboard, and stack comparisons. The ingredient database is the library underneath these layers.

2How we grade individual ingredients

Every ingredient receives two grades: evidence (strength of the research record) and worth it (evidence per dollar). Both follow the rubric below. Plus and minus refine within a band. A grade is never a single study; it is the weight of the record.

Worth-it grading. Evidence points (A = 4.0 down to D = 1.0) are divided by the midpoint of the typical monthly cost of a common form. The result ranks the public worth-it leaderboard, recomputed monthly and stamped with the month. Cost is the typical price of a common form, never a brand recommendation.

Grade states. Every grade is an editorial assessment derived through this rubric with sources on record. Grades are reviewed on a rolling basis and updated when new evidence lands; the changelog on each supplement page shows what changed and why. Nothing is presented as stronger than the underlying evidence supports.

3How we source expert positions

• Direct quotes only. Expert positions are published as direct quotes with a link to the source page and its date. We never paraphrase experts from memory.
• Primary sources preferred. The expert's own site, newsletter, podcast transcript, or published article outranks any secondary summary. A secondary source is used only when it carries a direct, checkable quote.
• Verification before publication. No quote is published until it has been confirmed on the source page. Unsourced slots render as “position not yet verified,” never as filler.
• Stance labels. Each verified position is labeled supportive, skeptical, or nuanced, based on the quoted content. The label describes the position, not the person.
• Disagreement is shown, not smoothed. When experts disagree, debate pages show the split and the consensus meter reflects it. We do not average away real disagreement.

4How we grade and compare stacks

A stack is scored on four dimensions, each 0 to 100, computed only from published data on this site. The overall score is the mean of the available dimensions, converted to a letter grade (A: 85+, B: 70+, C: 55+, D: 40+, F: below 40).

• Evidence depth (weight: equal). The mean of the ingredient evidence points in the stack, scaled to 100. A stack of A-graded ingredients scores near 100; a stack of D-graded research compounds scores near the floor. Formula: (mean evidence points / 4) × 100.
• Interaction safety (weight: equal). Starts at 100 and subtracts 25 for each established interaction between items in the stack, floored at 0. “No established interaction data” is neutral: it means no well-established interaction, not a guarantee of safety, and it is stated that way on every page.
• Goal alignment (weight: equal, when goals are stated). The share of items matching at least one of the user's stated goals, as a percentage. Omitted (not zeroed) when no goals are given.
• Cost efficiency (weight: equal). Total typical monthly cost of the stack against a $30/month reasonableness line: 100 at $30 or below, minus 2 points per dollar above, floored at 0.

What scores well. A high-scoring stack is small, built from well-evidenced ingredients, matched to stated goals, free of established interactions, and inexpensive. A low-scoring stack is large, built from thin-evidence or research compounds, mismatched to goals, or expensive per unit of evidence. The score describes the stack's evidence profile, not its suitability for any person.

Limitations. The score cannot account for individual health, medications, genetics, diet, or training. It does not see interactions beyond the established list. It is a comparison aid, not a prescription. Two stacks with the same score can differ in ways the rubric does not capture, which is why every score is shown with its full breakdown, never as a bare number.

5Peptides and research compounds

Peptides such as BPC-157, TB-500, GHK-Cu, and CJC-1295 are covered deliberately, because they are widely discussed and thinly covered by rigorous sources. The policy is strict:

• Regulatory status is stated up front: not approved for human use.
• Human trial status is stated plainly. For most of these compounds, no published human randomized trials exist, and the pages say so.
• A compound with no human data cannot earn better than a D for human efficacy, regardless of popularity or animal data.
• Products sold online are described as what they are: unregulated research chemicals of unknown purity.
• Safety notes flag WADA bans and the absence of human safety data.
• The stack builder excludes research peptides from beginner recommendations entirely and flags them for advanced users.

6How we find and fold in new data

The record is living. The weekly research process runs as follows:

• Monday, scan. New studies from PubMed and major journals for tracked ingredients; the tracked expert list for new positions (episodes, newsletters, videos); typical prices for common forms.
• Tuesday, draft. For each finding: what changed, the proposed grade or position update, the rationale, and the changelog note. No grade changes without a recorded rationale.
• Wednesday, review and publish. Routine updates publish directly. Material updates (a grade moves, experts newly disagree, safety flags) are flagged for owner sign-off; silence after 48 hours counts as approval except for medical-safety changes, which always wait.
• Thursday, newsletter. The weekly brief goes out: grade changes, new studies, new expert positions, one deep dive.

Changelog policy. Every ingredient and debate page carries a public changelog. Every grade change records the date, the old and new grade, and the reason. The record is versioned; nothing is silently rewritten.

Corrections policy. Errors are corrected publicly with a changelog entry stating what was wrong and what changed. We do not stealth-edit grades, quotes, or scores.

7Independence: sponsors and objectivity

• We grade ingredients, never brands. No sponsor's product can be the subject of a grade, because grades attach to compounds.
• Sponsors buy audience access, never grades. Placements are labeled. The people who manage sponsor relationships do not set grades.
• Disclosure. If a sponsor operates in a graded category, the relationship is disclosed on the relevant pages.
• No pay-to-rank. Leaderboard positions, stack scores, and debate meters cannot be purchased, and no placement fee influences them.

8Medical boundary

This site reports what studies and experts say. It does not tell any individual what to take, and nothing here is medical advice. Supplement decisions interact with medications, conditions, pregnancy, and individual risk in ways no database fully captures. Talk to your doctor before changing anything about your health. The stack builder is labeled explore-and-learn on every screen where it appears.

9Version history

• v1.0, October 2026. Initial public methodology. Covers ingredient grading, expert sourcing, stack scoring (evidence depth, interaction safety, goal alignment, cost efficiency), peptide policy, weekly research process, changelog and corrections policy, sponsor independence, and medical boundary.

Questions about this methodology, including from sponsors and their legal teams, are welcome. The record is public and so is the rubric. Browse the ingredients.