Why do plant and animal DAO products list such different potencies?

Because the figures come from different assays measuring different things. Porcine products typically publish tens of thousands of units per capsule and plant products publish millions. That hundredfold gap is a measurement artefact, not a statement that one is a hundred times stronger.

So can I compare two products by their labelled units?

Only within the same product type, and even then with caution. Comparing a porcine figure against a plant figure is comparing two different scales, which is why we render them in separate tables and would not merge them.

Is there good evidence these work?

The evidence is thin and the most-cited assay of commercial products has a serious conflict problem. Independent replication is what this category needs and we could not find it.

What was wrong with that assay?

It measured activity across commercial products and found a range from 0% to 244% of label claim. All five authors worked for the manufacturer of the two products that passed, and the three measuring zero were competitors. The range is informative about variability. The direction is not something we would report as established.

What would you want before trusting a number on one of these labels?

An assay run by someone with nothing to sell, publishing its method, testing products it did not make. That is a low bar and as far as we can find it has not been cleared.

DAO supplements

By B. Gray Randolph · Updated August 5, 2026

This category has a specific and unusually clean problem, and understanding it takes about a minute.

The hundredfold gap

Porcine-derived products typically publish potency in the tens of thousands of units per capsule. Plant-derived products publish figures in the millions. That is not a statement that one is a hundred times stronger. The numbers come from different assays measuring different things, so the gap is a measurement artefact and the two scales are not comparable.

Almost every comparison you will find online ranks these products in one table by that number. Doing so puts plant products at the top by a wide margin for a reason that has nothing to do with what they do in a person.

Why we render two tables and will not merge them

A merged ranking would be actively misleading, so the component that displays this does not offer the option. That is a deliberate constraint rather than an oversight: the way to stop a misleading comparison being made is to make it unavailable rather than to caption it.

The evidence problem

Alemany-Fornes M et al., bioRxiv 2023 Preprint, not peer reviewedWe have not yet confirmed a direct link to this document, so we are not giving you one. The citation above is enough to find it, and we would rather say this than send you to a journal front page.All five authors work for the manufacturer of the two products that passed. The three products measuring zero activity were competitors. Treat the direction of this result as marketing until replicated independently.

Read the caveat attached to that citation, because the caveat is the finding. Measured activity across commercial products ranged from 0% to 244% of label claim. Every author was employed by the manufacturer of the two products that passed, and the three products measuring zero activity were competitors.

What is not settled here

We report that range because a spread from zero to two and a half times label is genuinely informative about how variable this category is. We do not report the direction as established, because a manufacturer testing its own products against competitors and finding its own products superior is the least surprising result in science. What this category needs is an assay by someone with nothing to sell, and as far as we can find it does not have one.

What this page will not do

No ranking, no dosing, no recommendation. Ranking products whose potency figures are not comparable across types, in a category whose main published assay has an unresolved conflict, would imply a basis for comparison that does not exist.

If you are considering one, the useful question for a pharmacist or clinician is not which brand but whether anything you already take interacts, and what you would expect to observe if it were working. A supplement with no expected observable is a supplement nobody can tell you to stop.

Common questions

Why do plant and animal DAO products list such different potencies?: Because the figures come from different assays measuring different things. Porcine products typically publish tens of thousands of units per capsule and plant products publish millions. That hundredfold gap is a measurement artefact, not a statement that one is a hundred times stronger.
So can I compare two products by their labelled units?: Only within the same product type, and even then with caution. Comparing a porcine figure against a plant figure is comparing two different scales, which is why we render them in separate tables and would not merge them.
Is there good evidence these work?: The evidence is thin and the most-cited assay of commercial products has a serious conflict problem. Independent replication is what this category needs and we could not find it.
What was wrong with that assay?: It measured activity across commercial products and found a range from 0% to 244% of label claim. All five authors worked for the manufacturer of the two products that passed, and the three measuring zero were competitors. The range is informative about variability. The direction is not something we would report as established.
What would you want before trusting a number on one of these labels?: An assay run by someone with nothing to sell, publishing its method, testing products it did not make. That is a low bar and as far as we can find it has not been cleared.