The Standard Yardstick for Kava Potency Testing

Two laboratories, one root, two different numbers

A kava potency number is only as good as the method behind it, and before 2018 there was no agreed answer to what that method had to achieve. Two laboratories could analyse the same lot of Piper methysticum root and report total kavalactone contents differing by a third, and neither was doing anything a customer could point to as wrong — because nothing defined what right looked like.

The consequence was not fraud so much as incommensurability. A 12% kavalactone claim from one laboratory and an 8% claim from another could both be honest and still describe the same root, because the extraction solvent, the calibration standards, the treatment of minor lactones and the moisture basis all differed. Buyers had no way to compare offers. Regulators had no way to audit a label. Researchers had no way to place one study's numbers beside another's. And the flavokavains — the chalcones at the centre of the hepatotoxicity argument that closed European kava markets after 2002 — were frequently not measured at all; where they were, they were measured at levels nobody had agreed a method needed to reach.

AOAC SMPR 2018.005 closed that gap from the specification end rather than the method end. It names ten compounds, pins each to an IUPAC name, a CAS number, a UNII code, an InChI key and a PubChem CID, and then states the analytical range, limit of quantitation, recovery, repeatability and reproducibility that any technique must deliver across six named product matrices. It prescribes no column, no mobile phase and no detector, and it sets no limit on what a kava may contain. It states what a result has to be worth, and leaves the chemistry to whoever can meet it.

The distinction most often lost — The figures in the standard — 5 to 750 mg/g for kavalactones, 0.1 to 25 mg/g for flavokavains, 90 to 110% recovery — are demands on the instrument and the procedure. They are not thresholds a material passes or fails. Where a compositional threshold exists it comes from a quality framework, a label claim or a regulator, and the SMPR is what makes such a threshold testable rather than rhetorical.

What an SMPR is

AOAC INTERNATIONAL addresses this class of problem with a Standard Method Performance Requirement: a consensus specification, developed through a stakeholder panel, that defines what any candidate method must achieve for a stated analyte in a stated matrix. Methods are then developed, validated and adopted against that target, rather than a single method being written into the standard and everything else excluded.

SMPR 2018.005 is the kava instance. It was approved by the AOAC Stakeholder Panel on Dietary Supplements on 16 March 2018 and published in the Journal of AOAC INTERNATIONAL later that year, with twenty-four listed authors drawn from analytical laboratories, standards bodies, federal research institutes, ingredient suppliers and the kava trade.

Identity before quantity

Table 1 of the standard is the part most often skipped and the part that does the most work. Each of the ten analytes is carried with its IUPAC name, CAS registry number, UNII code, InChI key and PubChem CID. Trade and common names for kava constituents are inconsistent across the literature and across supplier certificates — kavain, kawain and (R)-(+)-kavain appear interchangeably, and the racemic form is a different substance with a different registry number. Anchoring each compound to a registry identifier means a certificate of analysis and a research paper can be compared without an argument about which compound was measured.

Table 1. The ten analytes named in AOAC SMPR 2018.005, with CAS registry numbers and PubChem identifiers. Entries 1 and 2 are the two kavain forms, listed separately in the standard, so the six major kavalactones are carried across seven rows.

• 1 — Compound: Kavain ((R)-(+)-kavain) · CAS No.: 500-64-1 · PubChem CID: 5281565
• 2 — Compound: d,l-Kavain · CAS No.: 3155-48-4 · PubChem CID: 5369129
• 3 — Compound: Dihydrokavain · CAS No.: 587-63-3 · PubChem CID: 10220256
• 4 — Compound: Methysticin · CAS No.: 495-85-2 · PubChem CID: 5281567
• 5 — Compound: Dihydromethysticin · CAS No.: 19902-91-1 · PubChem CID: 88308
• 6 — Compound: Yangonin · CAS No.: 500-62-9 · PubChem CID: 5281575
• 7 — Compound: Desmethoxyyangonin · CAS No.: 15345-89-8 · PubChem CID: 52736218
• 8 — Compound: Flavokavain A · CAS No.: 37951-13-6 · PubChem CID: 5355469
• 9 — Compound: Flavokavain B · CAS No.: 1775-97-9 · PubChem CID: 5356121
• 10 — Compound: Flavokavain C · CAS No.: 37308-75-1 · PubChem CID: 6293081

UNII codes and InChI keys are also given in the standard and are omitted here for space.

The working range, and what it is not

A compliant method must quantify kavalactones from 5 to 750 mg/g — 0.5% to 75% w/w — a span wide enough to cover dried root at one end and a concentrated dry extract at the other. Flavokavains run from 0.1 to 25 mg/g, with quantitation required down to 0.1 mg/g, or 100 ppm.

These are demands on the instrument and the procedure, not permitted content bands. A root that assays above or below them is not out of specification; it is simply outside the range across which the method was required to prove itself. The limit-of-quantitation requirements are ceilings rather than targets: a method must quantify kavalactones at 5 mg/g, and laboratories characterising minor lactones in plant material routinely need to work well below that. The standard permits it. It does not require it.

Both classes are reported as individual constituents — A method that returns only a summed kavalactone figure does not satisfy the standard, which matters more than it first appears. Chemotype — the rank order and relative proportion of the six lactones — is precisely the information a summed number destroys, and it is the information that distinguishes one cultivar's effect profile from another's.

The performance bar

Recovery of 90–110% and repeatability of ≤7.5% are conventional for a well-behaved chromatographic assay of a major constituent, and that is the appropriate benchmark for kavalactones in root and extract. The flavokavain tiering is the more considered decision. Between 0.1 and 5 mg/g the panel allowed repeatability up to 15% and reproducibility up to 20%, acknowledging that trace chalcone determination in a botanical matrix is harder than major-constituent work and that an unrealistically tight limit would simply have gone unmet. Above 5 mg/g the flavokavains are held to the same 7.5% and 10% as the kavalactones.

Table 2. Method performance requirements, consolidated from Tables 3 and 4 of AOAC SMPR 2018.005 and reported as individual constituents.

• Analytical range, mg/g — Kavalactones: 5–750 · Flavokavains 0.1–5 mg/g: 0.1–5 · Flavokavains >5–25 mg/g: >5–25
• Limit of quantitation, mg/g — Kavalactones: ≤5 · Flavokavains 0.1–5 mg/g: ≤0.1 · Flavokavains >5–25 mg/g: ≤0.1
• Recovery, % — Kavalactones: 90–110 · Flavokavains 0.1–5 mg/g: 90–110 · Flavokavains >5–25 mg/g: 90–110
• Repeatability, RSDr % — Kavalactones: ≤7.5 · Flavokavains 0.1–5 mg/g: ≤15 · Flavokavains >5–25 mg/g: ≤7.5
• Reproducibility, RSDR % — Kavalactones: ≤10 · Flavokavains 0.1–5 mg/g: ≤20 · Flavokavains >5–25 mg/g: ≤10

Every value is a requirement on the analytical method; none is a limit on the composition of the material.

Quality control, reference materials and validation

Section 6 of the standard requires blank check samples, check standards at both the lowest and the midrange point of the analytical range, and a control sample in every run. These are requirements on each analytical sequence — not descriptions of a validation exercise performed once and cited thereafter.

The standard also lists commercial sources for the ten analytical standards — PhytoLab, Extrasynthese/Alkemist, Sigma, AvaChem Scientific, AK Scientific, ACC Corp. and Cerilliant — and for botanical reference materials from USP, the American Herbal Pharmacopoeia and Botanical Liaisons. Naming sources in the standard itself is unusual and practical: a laboratory setting up a kava assay for the first time does not have to discover which suppliers can furnish a flavokavain C standard.

On validation, every target analyte and every claimed matrix must be represented in the evaluation. Data may be pooled across matrices to derive the overall analytical range, LOQ, recovery and precision figures, but a submitter cannot validate in dry extract and claim dried plant material. Six matrices are in scope: dried plant material, liquid extracts including tinctures, soft extracts, dry extracts, tablets, and capsules including softgels. The standard is explicitly technique-agnostic — any analytical technique meeting the performance requirements is acceptable — and it states no maximum time-to-determination.

What the standard does not do

SMPR 2018.005 is a constituent-quantitation specification and should be read as one. Four boundaries are worth stating explicitly, because each has been blurred in the trade's use of the document.

• It sets no limits on composition. Nothing in the standard says how much kavalactone a root should carry, how little flavokavain B a noble cultivar should show, or what a lot must contain in order to be sold. Those questions belong to quality standards, to label claims and to regulators.
• It does not address sampling. How a lot is drawn, which plant part is analysed, and whether a result is reported as-is or on a dry-weight basis are all outside its scope — and those choices move a kavalactone figure further than any difference between two compliant methods.
• It does not distinguish noble from tudei kava. Chemotype and cultivar authentication are ratio and marker questions. The standard quantifies the constituents from which such ratios are computed, but takes no position on where the boundaries between cultivar classes fall.
• It does not certify a laboratory. Meeting the SMPR is a property of a method as validated, not a standing attestation about a facility. The useful question to put to a laboratory is whether its method meets SMPR 2018.005 for the specific analytes and matrix being reported.

What it means in practice

What follows is the Association's reading rather than the text of the standard, and it is addressed to three different readers.

For laboratories and producers

A potency claim becomes auditable. A laboratory asked to support a label claim can be asked a specific question — does your method meet AOAC SMPR 2018.005 for the analytes and matrix you are reporting? — and the answer is yes or no rather than a description of an in-house procedure. Producers commissioning analysis should ask it at the point of engagement rather than after a dispute.

Individual-constituent reporting should be specified on the purchase order. The standard requires it, but a laboratory not working to the SMPR may still report a single total. A certificate showing all six lactones separately is worth materially more than one showing a sum, both for cultivar management and for any downstream chemotype claim.

For standards bodies and certification programmes

A compositional limit without a method specification behind it is unenforceable. The 0.4% w/w total flavokavain figure in the Kava Coalition's 2025 quality framework, carried into the American Kava Quality Standard, comes from that framework and not from AOAC; the SMPR takes no position on whether 0.4% is the right number, or whether any number belongs there. What it does is make the figure testable.

It also says how much scatter is permitted at the decision point — A threshold of 0.4% w/w is 4 mg/g, comfortably inside the range a compliant method must quantify — but in the lower flavokavain tier the permitted repeatability is 15% and the permitted reproducibility 20%. Two competent laboratories may therefore return materially different results on the same material at exactly the concentration where a certification decision is made. The practical responses are to report flavokavain B separately, to require duplicate determination near the limit, and to name the method alongside the number whenever a lot is certified.

For rapid and field methods

The standard places no constraint on throughput. With no maximum time-to-determination and no prescribed technique, a rapid or non-destructive method competes on equal terms provided it meets the same recovery and precision requirements as a chromatographic assay. Near-infrared spectroscopy is the obvious candidate for kava: it is fast enough to screen material at the point it changes hands rather than only where it is shipped for laboratory analysis.

The relationship between such methods and the SMPR is complementary rather than competitive. A spectroscopic method is calibrated and checked against reference chemistry, and the SMPR is what makes that reference chemistry comparable between laboratories. The stronger the reference layer, the more a rapid method can be trusted to carry a number into the field.

Conclusions

AOAC SMPR 2018.005 fixed what a kava potency number has to mean. It names ten analytes by registry identity, requires them to be reported individually, and sets recovery and precision requirements that any method — chromatographic or otherwise — must meet across six product matrices, with quality control run in every sequence rather than described once.

It is a measurement specification and nothing more. It sets no minimum or maximum content, addresses no sampling question, and draws no line between cultivar classes. Read as a composition standard it will mislead; read as what it is, it is the document that makes every other kava standard enforceable.

For the American kava industry the practical consequence is simple. Ask whether the method meets the SMPR for the analytes and matrix in question, insist on individual constituents, and treat any compositional threshold as meaningful only in company with the method used to test it.

Disclosures and conflicts of interest

The author is one of the twenty-four listed authors of AOAC SMPR 2018.005, and this review is published by an organisation he co-founded. Readers should weigh it accordingly: this document is an account of the standard's contents and significance rather than an independent verdict on it.

Tyler Blythe co-founded the American Kava Association in 2014 as an industry organisation to bring standards and laboratory testing to the North American kava market — four years before this standard was approved, and on the same premise. He has worked in kava distribution and analysis since 2006, and is co-founder of Root of Happiness, a kava brand founded in 2012 on lab-tested material at a point when routine analysis was the exception rather than the rule in the American trade. He is also co-founder of KavaLytics, a near-infrared quality-assessment method for kava, and therefore has a commercial interest in the rapid-method position set out above.

The Association is member-funded and its members are commercially engaged in the kava trade. No payment was received from AOAC INTERNATIONAL or from any laboratory or supplier named in this review.