NutraReview Hub method
How we compare supplement categories
This page documents the criteria, weighting, and sourcing rules behind every category comparison on the site. It exists so you can audit the method rather than take our word for it.
Four criteria. Same weight. No brand scores.
Method version September 2026 · Applies to all categories
Decision aid
Scope
What a comparison page is for
A comparison page exists to help you choose a category, not a specific bottle. The reader we write for is a shopper in the United States standing in an aisle or scrolling a product listing, trying to understand what the label in front of them actually means.
The comparison shows how categories differ in research state, label transparency, dose clarity, and format fit. It also names the specific label lines that change the meaning of a product inside the category — the stated amount per serving, the form of the ingredient, and whether that amount is disclosed at all. A page that only says one group is better than another does not help you read the next label you meet. Ours is built to travel with you.
It is a decision aid, and it is not a ranking. We do not crown a winner, we do not score brands against each other, and we do not let popularity move a category up or down. If a category has mixed evidence but unusually clear labels, the page says both things plainly so you can weigh them yourself.
The aisle test: if the page does not help you read the next label you pick up, it needs a rewrite.
A comparison answers four questions
- Evidence What do human studies actually support at label doses?
- Label Are individual amounts visible, or hidden inside blends?
- Dose Is a studied amount practical in a normal serving?
- Format Does the usual capsule, softgel, or gummy suit the ingredient?
The four criteria
We use the same four criteria on every category in our supplement comparison methodology. The wording on each comparison page is plain on purpose: you should be able to check our work with the bottle in your hand.
Research state
How much human evidence supports the primary ingredients. We ask whether findings repeat across studies, whether study doses appear on real labels, and whether the preparations studied resemble what is sold. Strong, mixed, and thin are defined in plain language on each page.
Equal share · one of four Read first
Label transparency
How much a typical product discloses. The key line is whether individual doses are visible instead of folded into a proprietary blend. Extract type, strain where relevant, elemental weight for minerals, and a batch number all count.
Equal share · one of four Label line
Dose clarity
Whether a meaningful dose is practical in the standard serving. One capsule at a studied amount reads differently than six capsules to reach the same amount. We compare the labeled dose against amounts used in human studies.
Equal share · one of four Serving check
Format fit
Whether the common format suits the ingredient. Fat-soluble compounds often sit in oil-based softgels for a reason. Ingredients sensitive to heat or light often avoid gummies. When the standard format fits, that counts in the category favor.
Equal share · one of four Delivery match
No criterion outweighs another. A category cannot make up for thin evidence with attractive packaging, and a well-studied ingredient does not excuse a label that hides its amounts.
How research state is assigned
Research state is assigned after reading the primary human literature on the ingredients that define the category. We look for consistency across studies, whether the dose in the studies is achievable in a typical product, and whether the preparations used in studies resemble what is sold.
-
Strong
Multiple human trials with reasonably consistent findings at doses that appear on labels. The preparation studied looks like the preparation sold.
-
Mixed
Findings depend on preparation, dose, or the outcome measured. Some trials show an effect, others do not, and the difference can be traced to how the ingredient was given.
-
Thin
The literature is small, short, or inconsistent. Early results may be interesting, but there is not yet enough human work to describe the category with confidence.
Animal and cell work can suggest where to look next, but it does not move a category out of thin. Only human data moves the needle, and the dose in that data has to be one you could plausibly find on a shelf.
How transparency is scored
Transparency is about the label, not the brand. We check a spread of typical products and ask the same questions every time.
- Are individual ingredient amounts stated in milligrams or micrograms?
- Is the extract type, strain, or standardization named?
- For minerals, is elemental weight shown alongside the compound?
- Is a batch or lot number printed for traceability?
A category where most products hide doses behind blends scores lower than one where most disclose. We do not score by reputation, because reputation is not a label line. If disclosure is generally good, that shows up. If it is generally poor, that shows up too.
Dose clarity and format fit
Dose clarity asks whether a meaningful dose is achievable without taking an impractical number of servings. Format fit asks whether the delivery method suits the ingredient inside it.
A category where the usual dose is one capsule and the ingredient is active at a small amount reads very differently than a category where the label needs six capsules to reach a typical research amount. We compare the labeled dose against amounts used in human studies, not against a brand promise. Format limits are part of that math, because only so much powder fits in a capsule and only so much oil fits in a softgel.
For format, the match matters more than the novelty. Fat-soluble compounds are usually carried in oil-based softgels, and a dry tablet may not behave the same way. Ingredients that degrade with heat or light often do better in capsules than in gummies that spend months in a warm warehouse. When the standard format fits, that is a point in the category favor. When it does not, you deserve to know before you buy.
Browse the ingredient library
Servings
One to two servings should carry the studied amount. More than that, and clarity drops.
Carriers
Oil carriers suit fat-soluble ingredients. Dry tablets do not always match that behavior.
Stability
Heat and light sensitivity points toward capsules over gummies for long storage.
If the trace fails, the claim is cut.
Every line on a comparison page must trace back to a label, a regulation, or published work. Marketing alone never carries a claim.
Sourcing rules
We use regulatory documents, peer-reviewed studies, and official label text as primary sources. Review articles and established reference databases are secondary. Brand marketing and news coverage are treated as leads and are never cited as proof.
A claim must trace back to a primary or secondary source before it appears. When the trace fails, the claim is cut — even if it would have made the page more interesting. That is why a comparison page can take longer to publish than a short article, and it is the rule we follow regardless of pressure to move faster.
- Primary: regulations, human trials, and the exact label wording.
- Secondary: review papers and reference databases for context.
- Leads only: brand pages and press stories point us where to look, then step aside.
What we do not include
These exclusions are deliberate. They mean some of our pages look less definitive than a competitor shopping list. That is a trade we accept because it produces a page that can still be trusted next year.
- No affiliate rankings, paid placements, or brand scores.
- No best overall recommendation for a category.
- No user reviews, before-and-after photos, or brand-supplied study summaries.
- No popularity signal — sales rank never moves a review criterion.
Our conflict of interest policy is simple: if money could move the answer, the thing that carries money stays off the page. For questions about funding or corrections, write to the editorial desk.
How disagreements between editors are handled
When two editors disagree about a research state or a transparency score, the page does not move forward until both have read the same sources.
Often the disagreement is about scope: one editor is thinking of a specific outcome, the other of the general literature. Resolving that usually means narrowing the language on the page rather than picking a side — saying exactly which outcome the evidence supports, and where it stays quiet. If the disagreement cannot be resolved, we say so on the page and describe the range of views instead of smoothing it over.
That note stays visible until new work settles the question. Readers should see where judgment was close, not just where it was comfortable.
Review cycle and updates
Comparisons are reviewed on a rolling basis. A review is triggered by a change in labeling rules, a meaningful new trial, or a reader correction that holds up after we check it.
Ingredient entries carry dates. Category comparisons are re-checked when their underlying ingredient entries change. Updates are logged on the Corrections and Update Log page with a short note describing what moved. There are no silent edits, and no page is marked as reviewed unless it has actually been read.
Read the update logLimits of the method
The method has limits and we name them up front, on the same page that asks for your trust.
-
No lab bench of our own
We do not test products ourselves. We rely on published work and third-party certificates, and published work changes as new trials land.
-
Typical product, not every product
We do not have data on every bottle on the market. Some comparisons describe the typical product in a category rather than every formula inside it.
-
Categories, not bottles
Our read is about categories. A careful product in a mixed category can still be a sound choice, and a careless product in a strong category still exists.
Supplements are intended to support general wellness alongside food and daily habits. They are not medicines, and nothing on this site diagnoses, treats, or prevents disease. If you take prescription medicines or manage a health condition, talk with a qualified professional before changing what you take.
How to use the method as a reader
You can apply the same four review criteria yourself without opening any of our pages. Two out of four is usually enough to set a product aside. The routine below makes that check repeatable.
-
1
Look up the ingredient
Search the primary ingredient in our research library or a peer-reviewed summary. Ask whether human studies exist at all, and at what amount.
-
2
Check whether the dose is practical
Divide the studied amount by the amount per serving. If the label needs four or six servings to get there, dose clarity fails.
-
3
Read the label for disclosed amounts
Look for individual milligrams, extract type, and elemental weight. A blend total without breakdowns is a reason to pause.
-
4
Confirm the format matches the ingredient
Oil-based softgels for fat-soluble compounds, capsules for sensitive ingredients. If the format fights the chemistry, keep looking.
Why do categories share the same four checks?
A shared method lets you compare across categories without relearning the rules. The same lens reads a mineral, a botanical, and a vitamin, so differences you see come from evidence and labels — not from shifting goalposts.
Do you test products in a lab?
No. We read labels, regulations, and published studies, and we explain how to interpret third-party test results. For lab testing itself, see our guide to certificates and what a passing result actually covers.
What if a label looks good but the evidence is thin?
Then the page says exactly that: clear label, limited human evidence. A transparent product in a thin category is easier to evaluate, but clarity does not create evidence that is not there yet.
Put the method to work
Open a category comparison with these four checks in mind, skim the update log for what changed recently, or send a correction if a label line looks off. That loop is how the pages stay honest.
NutraReview Hub · 1420 Madison Avenue, Suite 310, New York, NY 10029 · September 2026