How Our Evidence
Scores Work

Every ranking on Healthy Aging Atlas follows a structured, repeatable evidence process. We grade the science first, then test product fit, safety, transparency, and value.

What Our Evidence Grades Mean

A supplement can be popular, expensive, or heavily marketed and still have weak evidence for a specific goal. Our grades separate human outcome evidence from mechanism, hype, and product availability.

Strong

Repeated human evidence

Multiple relevant human RCTs, systematic reviews, or meta-analyses support the same practical conclusion for the same goal.

Higher confidence pages, but still not disease-treatment claims.

Moderate

Useful but incomplete human evidence

At least one relevant human trial or review supports the claim, but the evidence may be smaller, shorter, mixed, or indirect.

Most supplement pages live here until larger trials or replicated outcomes exist.

Limited

Early or narrow signal

Human evidence exists, but it is sparse, narrow to one population, short-term, or focused on biomarkers instead of outcomes readers feel.

Requires more cautious copy, stronger safety notes, and no inflated ranking language.

Emerging

Mechanism-first evidence

The biological rationale is plausible, but human evidence is not yet strong enough for confident consumer claims.

Allowed only with clear uncertainty and no strong buying recommendation.

Not Recommended

Risk, mismatch, or weak support

Evidence is too weak, safety concerns are material, the product form is a poor match, or claims depend on unapproved drug-like positioning.

May appear in research explainers or safety pages, not as a promotional ranking.

Five Steps, Every Page

01

Clinical Evidence Review

We start with the science. Our team reviews relevant studies indexed in PubMed, focusing first on human clinical trials, systematic reviews, and meta-analyses. Animal studies and in-vitro research may be referenced for context but are never used as the sole basis for health claims or dosage guidance.

  • Systematic search of PubMed for human clinical trials
  • Priority given to randomized controlled trials (RCTs)
  • Meta-analyses and systematic reviews weighted heavily
  • Study quality assessed by sample size, duration, and methodology
02

Product Data Collection

Our automated pipeline collects product data from major retailers daily. This includes pricing, serving sizes, ingredient lists, and availability. No manual data entry means fewer errors and consistently current information.

  • Daily automated scraping from Amazon, iHerb, and brand sites
  • Price-per-serving calculated from actual label data
  • Ingredient forms and dosages extracted and normalized
  • Availability status monitored in real time
03

Third-Party Testing Verification

Products with independent third-party certifications receive priority weighting in our rankings. We verify certifications against the certifying body's public databases when available.

  • NSF International (NSF Certified for Sport)
  • USP Verified mark
  • ConsumerLab.com approval
  • Other recognized certifications noted when present
04

Customer Experience Analysis

We mine thousands of verified customer reviews from Amazon and iHerb to identify real-world experience patterns. This surfaces information that clinical trials cannot capture — taste, mixability, side effects at scale, and long-term satisfaction.

  • Verified purchase reviews only — no incentivized reviews
  • Natural language analysis for recurring themes
  • Side effect mentions flagged and categorized
  • Review volume and recency factored into confidence
05

Medical Review

Before any page is published, a credentialed registered dietitian reviews every health claim, dosage reference, and safety note. Content that does not pass medical review is revised or withheld.

  • Credentialed RD sign-off required before publication
  • Health claims verified against current clinical evidence
  • Dosage ranges cross-referenced with study protocols
  • Safety notes and contraindications confirmed

How Products Are Ranked

Each product receives a composite score based on four weighted criteria. These weights reflect our editorial judgment about what matters most to consumers making evidence-based supplement decisions.

40%

Evidence Quality

Strength of clinical evidence supporting the supplement for the stated health goal. Weighted by study type, sample size, replication, outcome relevance, and safety context.

25%

Value

Price-per-serving at clinically studied doses. We calculate what it actually costs to take the effective amount, not the label's suggested serving.

20%

Purity & Testing

Third-party certifications (NSF, USP, ConsumerLab), heavy metal testing, and ingredient transparency. Products with verified testing score higher.

15%

Transparency

Label accuracy, proprietary blend disclosure, company reputation, and responsiveness to quality inquiries.

A High Score Is Not A Medical Claim

Product scores compare options for a stated supplement goal. They do not mean a product treats, cures, prevents, or diagnoses disease, and they do not replace clinician guidance.

9.0-10

Best candidate for the stated goal

Strong product fit, clear dose/form match, meaningful testing or transparency, and no unresolved safety or claim problem.

8.0-8.9

Strong option with tradeoffs

Clinically plausible and usually worth considering, but may have a price, form, availability, or evidence limitation.

7.0-7.9

Acceptable but not preferred

May be useful for a narrower reader need, but another product usually has better evidence fit, value, or transparency.

<7.0

Usually excluded from ranked picks

Used for internal comparison, watchlists, or cases where a product fails the fit, safety, testing, or value threshold.

What We Don't Do

  • Accept paid placements or sponsored rankings
  • Make disease treatment, cure, or prevention claims
  • Use only animal studies to support health claims
  • Publish pages without medical review
  • Allow affiliate relationships to influence rankings
  • Recommend products we have not independently evaluated
  • Let freshness, popularity, or affiliate availability override weak evidence

How We Stay Current

Daily

Product Data

Prices, availability, and retailer data are refreshed every 24 hours through our automated pipeline.

Monthly

Content Review

Every published page is reviewed at least once per month for accuracy, outdated claims, and new research.

Ongoing

New Research

When significant new clinical trials are published, affected pages are updated within one editorial cycle.

Questions About Our Process?

Learn more about our editorial standards, see live trust rankings, or review how we handle affiliate relationships.

Last updated: June 2026