How a real testing lab tests herbal extracts

The uproar caused by New York State Attorney General Eric Schneiderman in February is based upon testing methods that are neither scientifically valid nor fit for purpose, the two cardinal rules upon which accurate testing is founded. While DNA barcode testing has a place in the pantheon of testing methods, it can't give accurate results on extracts.

Elan Sudberg, CEO

March 16, 2015

7 Min Read
How a real testing lab tests herbal extracts

New York State Attorney General Eric Schneiderman sent cease and desist letters to Target, GNC, Walgreens and Walmart on Feb 3, accused them of selling fraudulent and potentially dangerous supplements, and demanded they be removed from shelves in New York. The tests this action was based upon found the products contained little to none of the herbs claimed on the label for seven house-brand products: echinacea, St. John’s wort, ginkgo, valerian, ginseng, saw palmetto and garlic. But they did contain other things not listed on the label, such as rice, asparagus, wild carrot and even garlic in the non-garlic product.

Schneiderman based his legal actions upon testing he said was definitive and supported by more than 70 studies proving the test method’s ability to identify the ingredient being tested. The same day, The New York Times health and science reporter Anahad O’Connor had a significant article in the Times—front cover, above the fold—ensuring every other news outlet in the country would cover this story as well. The Times, which has never hidden its disapproval of the dietary supplements industry, went on to print five additional articles and two editorials on dietary supplements over the next two weeks, including a call for Congress to take action for supplements to be regulated like drugs.

Testing methods brought into question

All of this uproar is based upon testing methods that are neither scientifically valid nor fit for purpose, the two cardinal rules upon which accurate testing is founded. While DNA barcode testing has a place in the pantheon of testing methods when performed by trained experts who understand the kinds of substances they are working with, it just can’t give accurate results on extracts. This is because the commercial extraction process leaves no actual plant tissue, degrading DNA while leaving the actives present. It also doesn’t quantify, so the rice and other substances were likely present in trace amounts via excipients. And anyone with experience in testing would immediately retest to rule out cross contamination when testing a garlic product, then finding garlic in other products in the lab.

This resulting massive uproar among consumers, health care providers, the dietary supplements industry and retailers has been unprecedented in our industry. As CEO of one of the testing labs focusing on natural products, I think about testing all the time. Suddenly, it’s all everyone else in this industry is talking about.

Bets testing methods

The World Health Organization (WHO) and other government and nonprofit organizations, including the U.S. Pharmacopeia (USP) and American Herbal Pharmacopeia (AHP), have published extensively on which methods are best for ensuring the quality of botanical materials, which includes identity testing. Appropriate testing methods fit for purpose are readily available (and have since been provided in great detail to the New York Attorney General by AHP).

For the record, the usual tests a competent lab would perform for identity testing of finished products would include a purposeful combination of high-performance liquid chromatography (HPLC) fingerprint analysis or high-performance thin layer chromatography (HPTLC). While they do have some limitations, these methods have stood the test of time and are accepted by government agencies such as FDA, Australia’s TGA and Health Canada, among others. These tests work well in the hands of specifically trained technicians who are committed to accurate results.

DNA is not on that list. Botanical extracts contain a chemical profile or fingerprint, and depending on what solvents are used, that profile or fingerprint can change drastically. Every morning, I make an extract using hot water and the ground seeds of Coffea arabica. I call my extract “espresso." If I were making a botanical extract for a dietary supplement, I'd then dry this liquid extract into a powder. Because water was used as the solvent, it pulls out from the ground seeds the chemicals that “jive" with water. In science terms, water pulls out polar phytochemicals from the seeds and leaves behind others. That’s what we drink in our coffee, a solution of water, pigments and polar compounds, mostly.

To test this raw material, one must know what to look for. One easy and time-tested way is to compare it to a reference sample of verified quality. That reference sample must be treated/prepared the same way as the test sample (hot water extract) in order to accurately compare the two. This is best done via HPTLC because HPTLC is all about comparative chemical fingerprinting. This is not unlike a DNA analysis that compares a known to an unknown however, in the case with most botanical extracts, little or no DNA is present.

While HPTLC is more than adequate, the other technique used to identify a botanical extract is HPLC; however, this is far more complicated and expensive to run and requires a handful of expensive phytochemical standards or reference materials to compare and contrast a phytochemical fingerprint. One thing to know, though, is that HPLC is designed to quantify and not to qualify. Methods are readily available to quantify the Gingerols in samples of ginger; however, the presence of Gingerols does not mean you have ginger root, rather, you have something that has Gingerols in it. In this way, it is not uncommon that HPLC is improperly used to assess the identity of a test sample. Saw dust spiked with Gingerols might pass. The folks who use HPLC fingerprinting to identify a botanical do so because they have access to HPLCs and not HPTLC equipment.

HPTLC can successfully confirm the presence of specific botanical ingredients in a finished product. There are limitations with all methods, and one that seems to stump HPTLC is when “fairy dusting" is employed. Unfortunately, the issue of fairy dusting cannot be solved in the lab— the formulators need to stop listening to the marketing department and stop tossing every “hot new ingredient" in a product without it being a meaningful addition to a formula and in a therapeutic dosage.

In the lab, researchers must inform finished product brands that sometimes, we simply can’t indicate the presence of a particular botanical ingredient. In other words, we fail the test sample. That said, a good product's label claim should be 100-percent verifiable by good science, and if it’s not, then they might need to reverse engineer so that it can be verifiable with today's limitations.

To ensure accurate label claims—test, test, test, test and test again. At each change of hands in the chain of custody, an identity test should be employed. Many products out there route through at least three different companies on its way to store shelves. What matters most is what is in the bottle, not the shipping container.

When I was getting my degree in chemistry, I could never give the professor a lab report with the words “trust me." I needed to show the data. There was no alternative. The same goes in court. The judge will need to see proof that X killed Y before sentencing X. This is what strikes me the most in the New York Attorney case. This is an Attorney General making a misstep of not fully disclosing the data to prove his claims, as we are all mandated to do. In addition, before I were to broadcast results of this magnitude, I'd surely get at least three labs to reproduce the data. Best lab practices are all about full method disclosure and full transparency. It is not always necessary to use two labs to reproduce each other's work, especially if transparency is followed.

Transparency is key to regaining the trust of consumers and the medical community. The age-old industry practice of failing to reveal the identity of testing labs and methods used needs to end now. Consumers are demanding, and deserve, to know that companies are taking care to give them what they think they are getting when they buy the products our industry sells. And that requires using valid test methods, fit for purpose.

Élan M. Sudberg is CEO of Alkemist Labs, a contract testing laboratory specializing in plant authentication, botanical ingredient identification and quantitative analytical services to the food and beverage, nutraceutical and cosmeceutical industries. He holds a degree in chemistry from California State University Long Beach, and has authored numerous journal articles on phytochemistry and analytical techniques for the natural products and nutraceutical industry. He is a board member of the American Herbal Products Association (AHPA), as well as AHPA’s Education and Research on Botanicals Foundation, and former chair of the Hemp and Medical Marijuana committee.

About the Author(s)

Elan Sudberg

CEO, Alkemist Labs

Élan M. Sudberg is CEO of Alkemist Labs, a passionately committed contract testing laboratory specializing in plant and fungal identity, potency and purity testing for the food, beverage, nutraceutical and psychedelic industries. His favorite part of his job is catching cheaters, and he is known for pushing the industry to continually raise the bar on quality and transparency. Élan holds a degree in chemistry from California State University Long Beach, and he is on the American Herbal Products Association (AHPA) board of trustees.

Subscribe and receive the latest insights on the health and nutrition industry.
Join 37,000+ members. Yes, it's completely free.

You May Also Like