How Aviado rates the evidence.
Aviado comprehensively ingests research and regulatory evidence every week, including PubMed, ClinicalTrials.gov, and the NIH and FDA databases that govern what a product contains and how it was made. From that record we publish three distinct ratings: the evidence behind a conclusion, the research case for an ingredient, and the quality of a product on the shelf. This page is the standard those ratings are held to.
Three ratings. Three different questions.
A strong study, a well-characterized ingredient, and a well-made product are different facts. Aviado rates them separately, so each has to stand on its own. The ratings are a reading of the published science and the regulatory record, in service of a decision.
The human evidence behind one conclusion in the Research Library. The same ingredient can be Strong for one outcome and Preliminary for another. The tier belongs to the conclusion.
The research case for an ingredient, from 0 to 100, across the strength of the human studies, the form's bioavailability, the safety record, and independent replication.
The quality and transparency of a branded product: manufacturing, testing, labeling, certifications, and the FDA record. A quality score for what you can buy.
In the Aviado product, an effect on a health measure carries its own evidence status, kept independent of both the article tier and the ProofMark.
What Aviado ingests every week
The scientific literature and the regulatory record both move constantly. Aviado ingests them on a weekly cadence, across the sources a serious reading of this category actually requires.
From the National Library of Medicine, PubMed: more than 40 million biomedical citations, among them more than 670,000 randomized controlled trials, with new and revised records added continuously. Aviado takes that literature in every week, including trials, systematic reviews, and observational studies across vitamins, minerals, botanicals, and the outcomes they are studied for.
From ClinicalTrials.gov, studies at registration, often long before a paper exists. Following a question from protocol to publication is part of reading the evidence, rather than arriving only when a journal has something flattering to print.
From the National Institutes of Health, the Dietary Supplement Label Database: the country's structured record of what labels say is in the bottle.
From the Food and Drug Administration, the manufacturing and safety record: facility registration, inspections, warning letters, and recalls. Those records inform ProofMark, the rating of a product. The literature informs the rating of a claim.
Where a paper is available in full, Aviado reads the full text. Every claim we publish is anchored to a citable source, with a PubMed identifier or a DOI a reader can open.
The 4-tier evidence rubric
Each published article carries one of these labels. The label is about the human evidence for the conclusion that article is drawing. The same ingredient can be Strong in one article and Preliminary in another.
- Strong evidence
- The conclusion rests on replicated controlled human trials, or on a systematic review of those trials. The studies agree on the direction of the effect.
- Moderate evidence
- The conclusion rests on one controlled human trial, plus observational studies that point the same way.
- Preliminary
- Early human data, mixed results, or a single trial. Published so the state of the evidence is visible. Not a firm conclusion.
- Insufficient
- Animal studies, a proposed mechanism, or anecdote. This is not a finding of no effect. There is not enough human evidence to judge.
The tier is a judgment on the human evidence for that conclusion. Concordant randomized trials, or a systematic review of them, support Strong. One controlled trial with observational studies in agreement supports Moderate. Early, mixed, or single-trial human data remains Preliminary. Animal evidence, a proposed mechanism, and anecdote remain Insufficient: the question is open, and Aviado will say so.
What counts as a source
Aviado cites the primary literature. A PubMed identifier links to the National Library of Medicine record. A DOI is given whenever the paper has one. The reader should be able to leave our page and stand in front of the study itself.
A preprint does not carry an evidence tier. A retracted paper is withdrawn from the page. A study is cited for a claim only when that study actually speaks to the claim. Popular coverage, brand materials, and mechanism diagrams are context. They are not evidence.
When the literature disagrees
Aviado does not resolve a conflict by citing the most favorable study. Where human trials disagree, the conclusion stays Preliminary. A single positive trial does not outrank a body of null results. Animal data and a plausible mechanism do not promote a claim into Strong or Moderate.
Disagreement is reported. It is a finding about the state of the evidence, and the tier is how that finding is labeled.
Evidence Score and ProofMark
An ingredient, scored from 0 to 100. Four judgments make the score: the human evidence, the bioavailability of the form under review, safety in the populations actually studied, and whether independent investigators have reproduced the effect. The score characterizes the ingredient. It says nothing about a brand.
A specific product, scored for quality and transparency. Manufacturing practice, third-party testing, ingredient form and dose, label clarity, certifications, and the FDA record of warning letters and recalls. A certification counts when it can be confirmed with the body that issued it. A sentence on a label is recorded as a claim by the brand.
The four Evidence Score axes
- Evidence
- Strength and consistency of the human studies behind the ingredient.
- Bioavailability
- How well the reviewed form is absorbed and delivered.
- Safety
- Tolerability in the populations those studies actually enrolled.
- Replication
- Whether independent studies find the same effect.
The two scores answer to different records, and they are kept apart on purpose. A high ProofMark is a statement about the product in front of you. A high Evidence Score is a statement about the research behind the ingredient. The full product-quality standard is set out in Inside ProofMark.
Scientific oversight
Working scientists oversee the standard and review the record against it. A review asks a precise question: does this study truly support this claim, and does the published article represent the studies it cites, including the tier, the size of the effect, the dose, and who was studied.
Where a reviewer and the draft disagree, the scientific judgment prevails and the published page is corrected. Scientists who review a specific article are credited by name. The scientific advisory board advises on the standard itself. A seat on the board is an oversight role, and a byline is earned article by article.
Who stands behind it
The Research Library is published by Aviado Health, Inc. Evidence reviews carry the Aviado Research byline. Explainers and this standard carry Aviado Editorial. Named scientific credit appears on an article when that scientist has reviewed it.
Aviado sells no supplements and takes no commission on any product it scores. The evidence tier answers to the literature. ProofMark answers to the product's manufacturing, testing, and regulatory record. Independence is the point of keeping them separate.
Currency and corrections
The underlying evidence is ingested every week. Each published page carries the date it was last revised. When a citation is wrong, a tier overstates the studies beneath it, or a product record has changed, we correct the page and advance that date.
Send the page address and the sentence to support@aviado.com.
Standard current as of 23 September 2026