Generally Recognized as Safe does not mean the FDA reviewed it. When researchers analysed 451 GRAS safety determinations that companies had voluntarily filed with the FDA between 1997 and 2012, not one had been made by an independent panel selected by a third party — every single assessment came from an employee of the manufacturer, a consulting firm the manufacturer hired, or a panel those parties chose (JAMA Internal Medicine, 2013).
That is the sort of thing a supplement label will tell you, once you know which words carry legal weight and which are decoration. What follows is a line-by-line read of a Supplement Facts panel: what it actually proves, what it cannot, and the six checks that take about a minute.
What does Generally Recognized as Safe actually mean?
It means a manufacturer concluded that its own ingredient was safe, and was not required to tell anyone. Under US law, GRAS substances are exempt from pre-market approval, and the FDA runs only a voluntary notification programme — which means the determinations nobody ever hears about are the ones that were never filed.
The 2013 analysis in JAMA Internal Medicine looked at the 451 notifications that were filed. Of those safety assessments, 22.4% were made by an employee of the additive manufacturer, 13.3% by an employee of a consulting firm the manufacturer selected, and 64.3% by an expert panel chosen by the manufacturer or its consultants. A standing panel selected by an independent third party made none of them. One individual sat on 128 of the 290 panels (PubMed 23925593). Its limits matter and belong in the same sentence as its finding: it covers only voluntary filings, it stops at 2012, and its authors were at the Pew Charitable Trusts, an organisation that campaigns on food-chemical policy.
A 2023 follow-up in Environmental Health examined 403 GRAS notices filed between 2015 and 2020, after the FDA had published best-practice guidance for managing conflicts of interest on these panels. It found no evidence that panels had adhered to that guidance, and reported that seven individuals occupied almost half of all available panel seats (PubMed 37670318). That paper is a qualitative assessment rather than a measured trial, and it too can only see the voluntary filings.
None of this makes GRAS ingredients dangerous. It means the phrase describes who decided, not who checked.
Which part of the label is actually regulated?
The Supplement Facts panel, and very little else. Serving size, the named dietary ingredients, their amounts and the % Daily Value where one exists are all required; the marketing copy on the front of the box is not held to the same standard.
Two habits pay off. First, read the serving size before any number, because every figure on the panel hangs off it — an impressive dose listed per two gummies is a different product from the same dose in one. Second, read the boring rows. Minerals do not travel alone, so mineral supplements often carry sodium along with them; the Daily Value used on US labels is 2,300 mg, and anything contributing meaningfully to that total should be declared on the panel rather than left for you to infer.
What is a proprietary blend, and why does it hide the dose?
A proprietary blend lists the ingredients without telling you how much of each one is present. The Dietary Supplement Health and Education Act of 1994 requires that the total weight of the blend and the names of everything inside it be declared — but not the individual amounts.
A 2023 perspective in The Journal of Nutrition, written by authors at the NIH Office of Dietary Supplements and the FDA together with one from an industry trade association, states the consequence plainly: from label information alone, the amount of a dietary ingredient inside a proprietary blend is not available for calculating intake or for determining doses in clinical trials (PubMed 37004873). If researchers cannot recover the dose from the label, neither can you — and neither can you check it against the dose used in whatever study the brand is gesturing at. Being a commentary rather than original research, it describes the rule; it does not measure how often blends mislead.
Does the label guarantee what is actually in the jar?
No. A label states an intention; only testing confirms it. The clearest published demonstration is a 2017 analysis in the Journal of Clinical Sleep Medicine, which measured 31 commercial melatonin supplements and found the melatonin content ranged from 83% below to 478% above the labelled amount. More than 71% missed their own label claim by more than 10%, one product varied by as much as 465% between lots, and eight contained serotonin, which appeared on no label (PubMed 27855744).
Its limits: 31 products from one market, one laboratory, one moment in time, and melatonin specifically — a compound with known stability problems. It is not evidence that every category is that erratic. It is evidence that a printed number is a claim rather than a measurement, which is why manufacturing standards and independent testing are the things worth looking for.
Should you worry about heavy metals?
Worth checking, and worth not panicking about — this is where the honest evidence cuts against the scare. A 2021 risk assessment in the International Journal of Environmental Research and Public Health tested 26 Ginkgo biloba products sold in Mexico City and found lead in every single one, arsenic in 54% and cadmium in 81%.
Then it did the arithmetic, and that is the half usually left out of the headline. Hazard quotients for manganese, arsenic and cadmium indicated no human health risk; the pharmaceutical-grade herbal products stayed within daily intake limits for all five metals, and only the dietary supplements and traditional remedies exceeded the limit for lead. The authors concluded that these products were not a health risk (PubMed 34444035). The study covers one botanical, in one city, at one time, so it does not settle other categories.
What it does settle is the reasoning. Detection is not the same as danger. Modern instruments find parts per billion in almost anything grown in soil, so contains lead means very little without a dose and an exposure limit printed next to it. Ask for a current-batch certificate of analysis with numbers on it, not a reassurance.
Which front-of-box phrases mean nothing on their own?
| Phrase | What it actually requires | What it does not mean |
|---|---|---|
| Generally Recognized as Safe (GRAS) | A safety determination by qualified experts; notifying the FDA is voluntary | That the FDA reviewed it or agreed |
| Patented | A novel, non-obvious composition or method | That it works. Patent offices examine novelty, not efficacy |
| Clinically studied | Nothing specific — the phrase is not defined in regulation | That the finished product was studied, or that the study succeeded |
| Proprietary blend | Total blend weight and the names of the ingredients in it | That you can determine the dose of anything inside it |
| Third-party tested | Nothing, unless the tester, the test and the batch are named | That every batch was tested, or what it was tested for |
| Natural | Nothing defined for supplements | Safe, gentle, or better absorbed |
How do you read a label in sixty seconds?
- Serving size first. Every number on the panel is per serving, and servings are chosen by the brand.
- Find the active and its dose. If it sits inside a proprietary blend, you cannot — treat the dose as unknown.
- Compare that dose to the studied dose. A different dose is a different question, and the answer does not carry over.
- Read the boring rows. Sodium, sugar, and anything carrying a % Daily Value.
- Interrogate the word tested. Tested by whom, for what, and on which batch. A seal without those three answers is a graphic.
- Look up the ingredient, not the brand name. A trademarked ingredient name is marketing; the material underneath it is the thing that has, or has not, been studied.
Step six is where most of the value sits. Trademarks travel further than evidence, and the research attached to a category — shilajit, humic acids, a class of botanicals — is not research on the specific branded material in your hand. If you want the mechanics of why form and delivery are separate questions from dose, we covered that in bioavailability and why nutrient form matters, and the specific misuse of the phrase clinically studied in what clinically studied actually means.
What we make, and how it scores on its own test
We make Vioni+, a tamarind fulvic mineral gummy, carrying 200 mg of Ioniplex per gummy. Running this article's own checks across it: the GRAS status of Ioniplex is self-affirmed, not FDA-reviewed. Its extraction patent, US 9,044,417, expired in 2023, so patented process would be the wrong words for us to use. And there is no published human clinical trial of Ioniplex — the single human study exists only as unpublished data inside a patent, where the between-group difference did not reach statistical significance. The class-level research on shilajit, humic acids and other fulvic materials is research on different materials, and we do not treat it as evidence for ours. Sodium is declared on our Supplement Facts panel; read it on the pack in your hand, because the panel is the record and formulations change.
Frequently asked questions
Does GRAS mean the FDA approved the ingredient?
No. GRAS substances are exempt from pre-market approval, and notifying the FDA is voluntary. The phrase records that a safety determination was made, not that the agency reviewed or endorsed it.
Is a patented ingredient a better ingredient?
Not on its own. Patents are granted for novelty and non-obviousness, not for proof of benefit, and they expire. A patent tells you something is distinctive; it tells you nothing about whether it does anything.
Should I avoid proprietary blends entirely?
Not automatically, but treat the dose as unknown, because legally it is. If a brand is pointing at a study, and the blend prevents you from checking whether its product matches that study's dose, the comparison cannot be made.
What is the single most useful check?
Dose against studied dose. Almost every misleading supplement claim survives on the gap between the amount used in the research and the amount in the product.
Do heavy metal findings mean a supplement is unsafe?
Not by themselves. Trace metals are detectable in most things grown in soil, and safety depends on the amount consumed against an established exposure limit. Ask for a current-batch certificate of analysis with figures on it.
Written by the VyVa Editorial Team and reviewed by Dr. Mohammed Hawa, DHA.
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
