How to read a supplement COA: 7 red flags that look fine on paper
A supplement certificate of analysis can be genuine, signed and in spec, and still hide filler or missing tests. Seven red flags to check before you order.
Verilyth Team 6 min read
In short
- A supplement COA can be genuine and in spec and still not prove what your label says: check the method, the lab and the lot, not just the numbers.
- The most common red flag is a “polysaccharide” figure from the phenol-sulfuric acid method, which also counts starch and fillers.
- Results should be numbers with units and limits, from a lab whose ISO/IEC 17025 scope covers that test, for the exact lot you are buying.
- In the US, 21 CFR 111.75 only lets you rely on a supplier’s COA after you qualify the supplier and confirm its results.
A certificate of analysis is a claim, not a result. It says a lot of material was tested, by someone, by some method, and met a specification someone wrote. Every part of that sentence can be true and the document can still tell you almost nothing about what you are about to buy.
None of the patterns below require forgery. Most COAs we read are not fake. They are just answering a different question from the one the brand is asking.
US brands have a specific reason to care. Under 21 CFR 111.75, a manufacturer may rely on a supplier’s COA for some specifications only after qualifying the supplier, which includes confirming the COA’s results through its own testing or examination1. A COA you have not checked is not a qualified COA.
1. The right number from the wrong method
The most common pattern in botanical and mushroom extracts is a correct result from a method that cannot measure the claimed compound.
Lion’s mane and other mushroom extracts are often sold on a “polysaccharide” percentage. When that figure comes from the phenol-sulfuric acid method, it is a colorimetric reading that responds to sugars and related substances in general2. Starch from grain, added maltodextrin and fungal β-glucan all contribute. McCleary and Draga showed that an enzymatic approach, measuring total glucan and subtracting α-glucan, separates fungal β-glucan from starch, and that products grown on grain were dominated by starch3.
So a COA that reports “Polysaccharides 30%, UV” can be perfectly accurate and still not support a β-glucan claim.
What to check: every assay line names a method, and the method can measure what the label will say.
2. A method that is named but not specific
Some COAs name a method that sounds rigorous but is not specific to the analyte. “HPLC” on its own is a technique, not a method. An HPLC purity figure without the chromatogram, the reference standard and some evidence that the method separates the analyte from related compounds is a number you cannot test.
What to check: ask for the method summary and the chromatogram for the lot. If the supplier cannot provide them, ask why.
3. Missing parameters
A COA is only as complete as the specification behind it. If the specification does not include heavy metals, microbiology or pesticide residues, the COA will not report them, and it will still say “conforms”.
For dietary supplements, US Pharmacopeia general chapters set out widely used approaches for elemental contaminants4, botanical pesticide residues5 and microbiological quality6. You do not have to adopt those chapters, but your specification should cover the same ground for your markets.
What to check: compare the COA against your own specification, not the supplier’s. List what is absent.
4. “Conforms” instead of a value
“Conforms”, “complies” and “pass” are not results. They tell you the lab, or the factory, compared a value with a limit. They do not tell you the value, the limit or the limit of quantification.
This matters most for contaminants. A lead result of “conforms” against an unstated limit gives you nothing to compare between lots or suppliers, and nothing to compare against the limits in the markets you sell into.
What to check: numeric results, with units, for every contaminant line, and the limit of quantification where the result is below it.
5. A real lab, outside its scope
Accreditation to ISO/IEC 17025 shows a laboratory’s competence for the specific tests listed on its scope of accreditation7. It is not a general badge. A lab can be accredited for heavy metals by ICP-MS and not for β-glucan, or for microbiology and not for pesticide residues.
Accreditation bodies that are signatories to the ILAC mutual recognition arrangement recognise each other’s accreditations8. In China, the national accreditation body is CNAS, and its directory lets you look up an accredited lab’s scope9.
The pattern to watch for: a COA on a reputable lab’s letterhead, for a test that lab is not accredited to perform. Or a COA from the factory’s own QC lab, which may be competent but is not independent.
What to check: look up the lab’s accreditation number and confirm the scope covers each test on the COA.
6. The wrong lot, or the wrong material
A COA describes one lot. If the lot number is missing, or does not match the shipment, the document describes something else. The same applies when the COA describes the raw material and you are buying the extract, or describes the extract before a carrier was added.
Dates are a quick check. A test date before the manufacture date, a report issued before the sample was received, or an expiry that does not fit any stability data are all signs that the document was assembled rather than produced.
What to check: lot number, product description and dates against the commercial invoice, the packing list and the spec sheet.
7. Recycled and edited documents
Recycled COAs are a pattern rather than a single feature. The same template, the same fonts and the same results across lots tested months apart. Real analytical results vary from lot to lot. Identical results to the decimal across three lots should prompt a question, not reassurance.
Edited documents are harder to spot from one file. Ask for COAs from three consecutive lots, and where a third-party lab is named, contact the lab and ask it to confirm the report number.
What to check: request three lots. Compare them side by side. Verify one report directly with the issuing lab.
What a useful COA looks like
A COA you can rely on, after qualifying the supplier, has a few features in common:
- one lot, identified by a lot number that matches your documents
- a specification that covers identity, assay, contaminants and microbiology for your markets
- a named method for every line, specific to what it measures
- numeric results with units and limits
- a laboratory whose accreditation scope covers each test
- dates that are consistent with each other and with the shipment
None of this requires a lab of your own. It requires reading the document as a set of claims and checking each one against something outside the document.
If you want a structured version of these checks, the COA red-flag checklist lists twelve of them on one page. If you would rather someone else read the document, send it for a free 24-hour flag check and we reply with the top 2 red flags, or book a full COA audit.
Footnotes
-
US Code of Federal Regulations, 21 CFR 111.75, What must you do to determine whether specifications are met? https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-111/subpart-E/section-111.75 ↩
-
DuBois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. Colorimetric method for determination of sugars and related substances. Analytical Chemistry. 1956;28(3):350–356. https://doi.org/10.1021/ac60111a017 ↩
-
McCleary BV, Draga A. Measurement of β-glucan in mushrooms and mycelial products. Journal of AOAC International. 2016;99(2):364–373. https://pubmed.ncbi.nlm.nih.gov/26957216/ ↩
-
United States Pharmacopeia. General Chapter 2232, Elemental contaminants in dietary supplements. ↩
-
United States Pharmacopeia. General Chapter 561, Articles of botanical origin. ↩
-
United States Pharmacopeia. General Chapters 2021 and 2022, Microbial enumeration tests and absence of specified microorganisms, nutritional and dietary supplements. ↩
-
ISO/IEC 17025:2017, General requirements for the competence of testing and calibration laboratories. https://www.iso.org/standard/66912.html ↩
-
International Laboratory Accreditation Cooperation. ILAC MRA and signatories. https://ilac.org/ilac-mra-and-signatories/ ↩
-
China National Accreditation Service for Conformity Assessment (CNAS). https://www.cnas.org.cn/ ↩
Common questions
What is a COA for supplements?
A certificate of analysis (COA) is a document from the supplier or a lab that reports test results for one lot of an ingredient: identity, assay, heavy metals, microbiology and other parameters, against a specification.
Can a real COA still be misleading?
Yes. A COA can be genuine and in spec and still not support your label, for example when the assay uses a method that cannot measure the claimed compound, or when key tests are missing.
Which tests should a supplement ingredient COA include?
At minimum: identity, assay of the labelled compound with a named method, heavy metals (lead, cadmium, arsenic, mercury) as numbers, microbiology including pathogens, and pesticide residues for botanicals.
Is the supplier’s COA enough for FDA compliance?
Not on its own. Under 21 CFR 111.75, US brands can rely on a supplier’s COA only after qualifying the supplier, which includes confirming the COA’s results with their own testing or examination.
How do I know if the lab is accredited for the test?
Ask for the lab’s ISO/IEC 17025 accreditation number and check its scope on the accreditation body’s website. The specific test on the COA should be listed in that scope.
- COA
- methods
- laboratories
- supplier qualification