The New York Attorney General's Attack on Supplements
The story first broke on February 3rd. New York Attorney General Eric T. Schneiderman announced in a press release that his office had sent letters to four major retailers, GNC, Target, Walmart, and Walgreens, for allegedly selling store brand herbal supplement products in New York that either could not be verified to contain the labeled substance, or which were found to contain ingredients not listed on the labels.1 The letters called for the retailers to "immediately stop the sale of certain popular products, including Echinacea, Ginseng, St. John's Wort, and others." Attorney General Schneiderman requested that the companies "provide detailed information relating to the production, processing and testing of herbal supplements sold at their stores, as well as set forth a thorough explanation of quality control measures in place."
At first glance, the revelations sounded shocking and the call to action reasonable, and, not surprisingly, the press--especially The New York Times--jumped onboard with headline after headline highly critical of the supplement industry. But one couldn't help wondering why this was handled in a press release. Was the motivation more political than public safety? Perhaps Schneiderman is following in the footsteps of Eliot Spitzer, the former New York AG who used a series of dramatic press releases and public intimidation to squeeze high profile defendants as a springboard to the Governor's mansion.2 Then again, how well did that ultimately work out for Elliot Spitzer, who ultimately became the first New York governor to leave office amid scandal in nearly a century?3
But back to the current AG. Specifically, the Schneiderman press release stated, "DNA testing, performed as part of an ongoing investigation by the Attorney General's Office, allegedly shows that, overall, just 21% of the test results from store brand herbal supplements verified DNA from the plants listed on the products' labels--with 79% coming up empty for DNA related to the labeled content or verifying contamination with other plant material."
At first glance, that sounds very damning. But as we will see a little later, that's not necessarily the correct interpretation of the data. To be sure, this article is not an endorsement of any of supplements in question. We are merely examining the validity of AG's use of DNA Barcoding and its applicability for testing supplements. For now, let's just focus on the fact that DNA testing for supplements is actually a brand new field. Only one company is authorized by the FDA to perform it, and they clearly state that DNA testing cannot reliably test products that contain multiple ingredients. But we'll come back to that. Suffice it to say that at this point, everything else that follows in the AG's press release is dubious at best. As a side note, the testing was done for the State of New York by James Schulte at Beckman Coulter Genomics. Schulte's academic specialty is in snake and lizard evolution,4 which means he's not quite the expert in botanical testing that the New York Times claimed.5
Anyway, back to the release, which went on to quote the AG directly, "This investigation makes one thing abundantly clear: the old adage ‘buyer beware' may be especially true for consumers of herbal supplements. The DNA test results seem to confirm long-standing questions about the herbal supplement industry. Mislabeling, contamination, and false advertising are illegal. They also pose unacceptable risks to New York families--especially those with allergies to hidden ingredients. At the end of the day, American corporations must step up to the plate and ensure that their customers are getting what they pay for, especially when it involves promises of good health."
Okay, that pretty much answers the question as to whether or not the press release was about politics of public safety. It's pure politics! If you read it out loud, it sounds remarkably like a campaign speech designed to fire up the masses. However, if you still have any doubts, let's take a look at the next two paragraphs, which again contain no science or facts, but self-serving quotes from two New York political figures.
" ‘As the sponsor of a measure that would require labeling that states whether a product has been evaluated by the FDA or not, and legislation to establish a dietary supplements safety committee, I fully support the Attorney General's efforts in this area,' said New York State Senator Ken LaValle. ‘I will continue to fight for legislation that will provide adequate labeling information for the public.' "
" ‘Since 2005, I have sponsored legislation to create a dietary food supplements safety committee,' said New York State Assembly member Felix Ortiz. ‘This bill was crafted for the very same reasons the Attorney General is now targeting retailers selling generic supplements that may or may not contain the substances contained on the labels. I support the Attorney General's efforts and I will continue to push for the passage of my bill (A3548) to help reduce this kind of consumer fraud. We need adequate standards and better enforcement over these dietary supplements so consumers will feel confident knowing what they are buying.' "
And in case you did not yet understand how marvelous the AG is, the release used the next paragraph to let you know in no uncertain terms.
" ‘This study undertaken by Attorney General Schneiderman's office is a well-controlled, scientifically-based documentation of the outrageous degree of adulteration in the herbal supplement industry,' said Arthur P. Grollman, M.D., Professor of Pharmacological Sciences at Stony Brook University. ‘I applaud the New York Attorney General for taking the additional step of seeking to remove these products from the marketplace as they can cause serious harm to consumers unaware of the actual ingredients in the pills and capsules they ingest. Hopefully, this action can prompt other states to follow New York's example and lead to the reform of federal laws that, in their current form, are doing little to protect the public.' "
For the most part, you can read the AG's press release yourself for more specifics. But there is one more section that I would like to quote from so that we get it exactly right because it goes to the heart of issue.
"Using DNA barcoding technology to examine the contents of herbal supplements, the Attorney General's investigation is focused on what appears to be the practice of substituting contaminants and fillers in the place of authentic product…The testing revealed that all of the retailers were selling a large percentage of supplements for which modern DNA barcode technology could not detect the labeled botanical substance."
"An expert in DNA barcoding technology, Dr. James A. Schulte II of Clarkson University in Potsdam, N.Y., was hired by the Attorney General's office to perform the testing. DNA barcodes are short genetic markers in an organism's DNA and are used to identify it as belonging to a particular species. Barcodes provide an unbiased, reproducible method of species identification. Barcodes can be used to determine the exact plant species being tested."
"The DNA tests were performed on three to four samples of each of the six herbal supplements purchased from the New York stores. Each sample was tested with five distinct sequence runs, meaning each sample was tested five times. Three hundred and ninety tests involving 78 samples were performed overall."
And now let's take a look at the key result for each manufacturer, because once you understand what it's saying, you'll be more than a little angry.
- GNC: Of 120 DNA tests run on 24 bottles of the herbal products purchased, DNA matched label identification 22% of the time.
- Target: Of 90 DNA tests run on 18 bottles of the herbal products purchased, DNA matched label identification 41% of the time.
- Walgreens: Of the 90 DNA test run on 18 bottles of herbal products purchased, DNA matched label representation 18% of the time.
- Walmart: Of the 90 DNA test run on 18 bottles of herbal products purchased, DNA matched label representation 4% of the time.
This is what the AG was referring to when he said earlier, "21% of the test results from store brand herbal supplements verified DNA from the plants listed on the products' labels--with 79% coming up empty." But as I said earlier, that's not necessarily the correct interpretation of the data. Take a look at the Target numbers for a moment. 41% of the time they ran the test they got one result. 59% of the time they got the opposite result. This is not much better than a coin flip. It's a long, long way away from the trillion to one numbers they use when talking about the reliability of DNA testing in murder trials. Can you imagine a jury convicting someone on evidence if the person giving the testimony said, "Well, it's pretty much a 60/40 call? What gives? What kind of DNA testing are we talking about that produces "now you see it, now you don't" results?
Do you remember a few years ago when everybody was selling nanotechnology supplements? Heck, they were even selling nano water for $528 a quart.6 In truth, all nano means is "very small," but it was used to sell everything under the sun until it became meaningless. We're now beginning to see the same thing with the appellation DNA.
When most people think of DNA, they think of murder trials and CSI TV shows. And indeed, the use of DNA evidence in criminal investigations has grown in recent years. DNA testing has helped law enforcement identify criminals and solve difficult crimes. And vice versa, DNA evidence has helped prove that many convicted people are actually innocent. And as used in criminal trials, it's quite accurate, although probably not as accurate as most people think. Whereas DNA experts in criminal trials often cite numbers such as one in 113 billion or even one in a trillion in terms of reliability, it turns out that the real numbers are far, far less accurate. Arizona state crime lab analyst Kathryn Troyer has actually identified dozens of men--unrelated and of different races--producing "matching" DNA.7
Okay, so it's not as accurate as promised. Maybe the odds of making a mistake are not one in 113 billion, but we're still talking about one in tens of thousands or hundreds of thousands, right? Maybe it's not good enough to execute someone, but it's certainly good enough to identify a bad product, right?
Although, the testing methodology contracted by the New York AG's office is called DNA Barcoding, it's not the same thing as the DNA testing used in criminal trials. It involves a much smaller sampling of the DNA, and therefore much less accuracy. The technique was first proposed in 2003 by a researcher from the University of Guelph in Ontario, Canada.8 DNA Barcoding uses a very short genetic sequence from a standard part of the genome in much the same way that a supermarket scanner distinguishes products using the black stripes of the Universal Product Code (UPC)--thus the use of the word Barcoding. Until now, it has been used exclusively for determining the identity of animal specimens. The gene region that is used as the standard barcode for almost all animal groups is a 648 base-pair region in the mitochondrial cytochrome c oxidase 1 gene (CO1 or COI). And for animals, it seems to work well. COI has proven highly effective in identifying birds, butterflies, fish, flies and many other animal groups. Unfortunately, COI is not an effective barcode region in plants because it evolves too slowly, so researchers have had to look for alternatives. Two gene regions in the chloroplast, matK and rbcL, have been identified as the barcode regions to use for plant identification, but the accuracy of results based on those regions is still open to question. And, as we will see, totally falls apart when you are looking at formulas that contain extracts and have multiple ingredients.
How does DNA Barcoding work?
Essentially, the specimen in question is analyzed and searched for the appropriate matK and rbcL sections of DNA code. Then, like a barcode, that code is run against an existing database of all previously identified species looking for a match. If you get a match in your database, you've identified your species.
So far so good, but:
How does the identification process work when the original specimen has been "processed" or "extracted"? As it turns out, not so well. Nandukumara Sarma, PhD, director of dietary supplements at U.S. Pharmacopeia, says "Use of solvents and/or heat during extraction process can degrade--and sometimes destroy--DNA in extracted material (but not the constituents associated with the benefits). DNA may not be extracted in sufficient amounts or may be eliminated by further purification or refinement processes. Thus, it's possible that when target compounds are extracted from its [sic] plant source to make an herbal ingredient, there's no longer any detectable DNA present in the final product. In this case, it is possible that the right compounds are present in an extract but there would be no DNA present to confirm it. For this reason, the presence or absence of DNA in an extract should not be used to confirm the identity of an ingredient in the form of a plant extract."9
- How does the process differentiate when something is a contaminant and when something is a trace amount of an inert ingredient that has been added to the formula as part of the processing? Again, not so well. Inert ingredients such as wheat, rice, even wild carrots (all identified as contaminants in the AG's press release) are actually all perfectly legal in dietary supplements when present in trace amounts. There is no indication that the testing performed for the AG calculated the quantity of these ingredients, which is extremely difficult to do using DNA Barcoding. All the testing determined that there was some DNA from these ingredients present--no matter how minute that quantity was.
In other words, whereas, DNA Barcoding may have some efficacy in identifying actual plant specimens (and that's still open to question), it has only marginal efficacy when it comes to analyzing herbal formulations--and none if those herbs have been process or extracted.
There is only one lab authorized to do DNA testing for botanicals: AuthenTechnologies, based in Northern California. In September 2014, the FDA awarded AuthenTechnologies an exclusive contract for up to three years to develop a collection of validated reference DNA sequences and reference materials for a variety of botanical species. If you're looking for an expert in using DNA Barcoding for botanicals, it's AuthenTechnologies--not James, the snake and lizard evolutionist. But as AuthenTechnologies stated in a special report they published in response to the New York AG's calls for DNA testing of herbal formulations10--and remember, AuthenTechnologies is the only lab authorized by the FDA to develop DNA Barcoding for commercial use.
"Very little research has been conducted on the use of DNA barcoding for materials other than those that are fresh or living, especially on botanical extracts in dietary supplements where the DNA can be removed or degraded. Therefore the use of DNA barcoding for finished dietary supplements is largely misunderstood and misapplied--even by those who claim to perform it."
"While DNA sequencing can be a useful tool for authentication of raw herbs (if the appropriate genes are used), its application for finished herbal/botanical dietary supplements is limited due to the generally low quality of DNA in those products…The failure to appreciate the effects of extraction on botanical material indicates the general lack of understanding by scientists performing DNA testing on finished products and the importance of the distinction between raw and processed materials, especially extracts."
"The NY AG made public pronouncements with potentially far-reaching legal and market implications that a majority of the dietary supplement products were devoid of the labeled ingredients and contained a myriad of unlabeled ones. However, these conclusions are not valid for a number of reasons."
What's the better alternative?
To be clear, the U.S. Pharmacopeial Convention does not currently recommend DNA Barcoding in any of its standards.11
Daniel Fabricant, PhD, and CEO of the Natural Products Association, says that chemical analysis or chromatography should be used instead of DNA barcoding. "Once you get an extract, there essentially won't be any DNA left. All of the products were extracts so we wouldn't expect there to be any DNA." The use of DNA barcoding for testing supplements containing herbal extracts is "like using a screwdriver to hammer a nail and then saying the lumber was bad. It's over-reaching and nonsensical."12
The American Botanical Society (ABS) concurs. "Microscopy and validated chemical test methods, like those found in official pharmacopeias for these seven herbs, should have been conducted to confirm the DNA findings. …We respectfully must question whether an appropriate level of scientific rigor has been applied in this case."
Understand, I'm not saying not to test. Customers should absolutely be assured their supplement contains what it promises on the label. I’m just saying:
- The test the NYAG used doesn’t accomplish that. It adds unnecessary costs that you will ultimately have to pay for—purely for political grandstanding.
- There is testing in place already that is used as standard procedure in all cGMP compliant products that does accurately guarantee that promised bioactives are indeed present.
- The same procedures also test for contaminants.
- Just make sure you opt for supplements that are cGMP compliant.
For more on cGMP, check out: Can You Trust Your Supplements?
On March 30th, AG Eric Schneiderman's office issues another press release.13 To quote:
"Attorney General Eric T. Schneiderman today announced a landmark agreement with Pennsylvania-based retail giant GNC to implement new standards in authenticating herbal supplements, ensuring their purity, and educating consumers about their chemical content. Under today's agreement, GNC will perform DNA barcoding on the "active" plant ingredients used in its products; implement testing for contamination with allergens, both before and after production; and post prominent signage advising consumers of the processed, chemical nature of extracts."
"Within 18 months, GNC will begin utilizing DNA barcoding to confirm the authenticity of all plants used as sources for its herbal supplements products prior to processing."
Incidentally, the reason for the 18 month delay is that there is currently no actual database of verified DNA from authenticated specimens of all the appropriate herbs to use as the reference for running DNA Barcoding tests. Also, as previously mentioned, there is currently only one lab authorized to run commercial testing using DNA Barcoding. The backlog would be monumental if GNC's competitors felt forced to immediately implement the same testing for their products. In fact, it would be fundamentally impossible for the market to instantly be able to handle such a dramatic shift in testing protocol--as silly as that shift might be.
The Reality behind the Agreement
The agreement itself is worth reading,14 but if you don't have the time, here are the key points and what they mean:
- GNC fully complied with all Current Good Manufacturing Practices and tested all supplements during production and with an independent laboratory. Translation: GNC didn’t do anything wrong.
- NYAG and GNC agree that DNA Barcoding “holds great promise” for further authentication—in the “supply chain,” but not for finished goods. Translation: We messed up big time when we accused GNC, Target, Walmart, and Walgreen's of not having any active ingredients in their supplements. DNA Barcoding doesn’t work now for extracted herbal products--but may someday in the future work for assessing ingredients in the "supply chain."
- But this is a false hope. Even for ingredient testing as part of the manufacturing process, DNA Barcoding is useless, and potentially misleading, when it comes to supplement testing. Let me give you an example. Beet powder is an ingredient sometimes used in super food formulations. High quality beet powder has the same bright red color that you see in beets themselves. And why wouldn’t it? It’s made by drying fresh beets and grinding them into a powder. Sometimes, however, ingredient suppliers will ship you a beet powder that is pale orange/yellow. Why the difference in color? This is actually a powder made from the pulp of beets that have been juiced. Everything good and useful has already been extracted from it. At this point, it’s pretty much nothing but fiber. It has zero value as a supplement. However, and this is the true elephant in the room when it comes to DNA Barcode testing, it will still register as pure, 100% American beet. In other words, DNA Barcoding will give you a big thumbs up for what is essentially a useless ingredient—simply because it contains beet DNA as designated in a reference table.
- GNC has up to 18 months to implement source material standards that utilize DNA Barcoding to confirm the authenticity of the "active" herbal/ botanical ingredients prior to extraction. Translation: Again, Barcoding does not work for finished herbal products. If there is any value, it's in testing raw source materials. But since DNA Barcoding is not ready for prime time even for that purpose, let’s push it off 18 months and see if it's ready in a year and a half.
- GNC will implement allergen testing throughout its production process. Translation: Allergens need to be identified, but that’s already covered by Federal law. Oh, and as for Barcoding, it may tell you that the DNA of an allergen is present, but not how much of the whole that allergen represents. When you harvest a crop of some herb, you may get a couple of weeds in that batch. With Barcoding, the DNA of that weed will show up even if it represents as little as 0.001 percent of the total batch. At 0.001 percent, there will be zero allergenic impact; but as far as Barcoding is concerned, it will look exactly the same as the herb itself at 99.999 percent. Barcoding can't differentiate amounts. That makes it useless for identifying allergens.
- All product pages on the GNC consumer website for GNC Herbal Plus brand products that are listed as extracts will include in the product description section prominent links to the explanation of the differences between whole herbs and extracts using the wording in footnote 4, and a description of GNC's extraction methods. Translation: As we’ve said repeatedly throughout this report, DNA Barcoding doesn’t work for herbal extracts. By the way, the AG tried to spin this like the announcement is some kind of warning. It's not.
- This agreement constitutes an assurance of discontinuance for purposes of Executive Law § 63(15) that will discontinue NY A G's investigation of GNC Herbal Plus brand products for authenticity, purity, and related labeling issues. Translation: GNC's allegedly compromised products are now allowed back on store shelves because DNA Barcode testing produces bogus results when testing processed herbs. It's also an admission that chemical analysis and/or chromatography, the tests currently used by all cGMP compliant herbal manufacturers, actually are more accurate in assessing the presence of herbal bioactives.
Everything you need to really know is contained in that last provision that GNC is now allowed to once again sell all of the products in question--immediately. Why is this important? Because it can mean only one of three things.
- The AG's office doesn't care a rat's behind about consumers and is allowing product that they truly believe to be tainted with allergens and that contains no active ingredients to once again be sold to the public.
- The AG's office now understands that they were absolutely wrong in their original assessment and is doing their best to polish the turd they made into something they can spin as high-grade, organic fertilizer.
- Both 1 and 2. They really never cared about consumers in the first place; the whole deal was purely political grandstanding; they were caught with their pants down; and they are now going to make consumers pay for valueless additional testing to cover their behinds.
I'll let you decide which. Unfortunately, no matter how you look at it, this is a bad deal for consumers. It doesn't help them at all, provides no reliable new information, improves safety not one iota; but it does add another layer of cost to the price of supplements that will be passed on to those same consumers for the foreseeable future. It sets a horrible precedent.
I understand why GNC caved and came to an agreement so quickly. Their stock dropped 7% in the days immediately following the AG's first announcement attacking GNC. And for a company the size of GNC, that represents a whole lot of money. Unjustly or not, they felt their feet were being held to the fire and so they "settled" ASAP. Incidentally, their stock rose 5% immediately after the announcement of their agreement. Their PR department will now spin the disaster as a benefit--claiming that GNC is the only company that goes the extra step and is the only company to follows the NYAG standards. They will have the "nanotechnology edge." For similar reasons, it is likely that Walmart, Target, and Walgreen's will settle along similar lines. But again, it sets a horrible precedent.
How much more interesting things would have been if the Gang of Four or their shareholders who actually lost the money in question, instead of capitulating, had filed suit against the State of New York for their false accusations--claiming hundreds of millions in damages. Perhaps, then, The New York Times might not have been quite so smug in their April 1st editorial in which they said:
"Now that the perils of lax oversight have become inescapably clear, Congress ought to grant the F.D.A. enhanced powers to regulate the industry and remove products that are fraudulently marketed or potentially dangerous."15
The AG and the Times love to cite all of the studies behind DNA Barcoding as justification for the AG’s actions. What a crock! Not one of those studies supports the use of DNA Barcoding for processed supplements or herbal extracts—not one. And as we’ve already seen, the only lab authorized by the FDA for testing herbal ingredients categorically states it doesn’t work for herbs that have been processed. The bottom line is that the AG backtracked from his accusations, nothing was fraudulently marketed, and nothing was actually "removed" from the market. Res ipsa loquitur--the thing speaks for itself! The only thing that is "inescapably clear" is that once again we get to see politics trump common sense, good science, and anyone who wishes to pursue natural health alternatives. This was a total mess-up by the NYAG, replete with libelous accusations and cover-your-behind backpedaling--all repackaged as a great consumer victory that will, in truth, provide no benefit to the consumer, but merely add to the cost of your supplements.
Again, this report is not an endorsement of the supplements in question or of the quality of the herbs used in them or even if the ingredients are included in meaningful amounts--all of which separate an effective formula from one that doesn't do anything. Instead, the purpose of this article is make clear that DNA Barcoding is not suitable for the task at hand and that GNC's agreement to nevertheless incorporate it into their manufacturing process will quite likely to create a ripple effect in which Walgreen's, Walmart, and Target also agree--so they don't look like they're lagging behind. And then, of course, every other company will be forced to follow suit simply as a result of market pressure. In the end, you're going to end up paying more for your supplements down the road, with no benefit to you--all so Eric Schneiderman can remain a viable candidate as the next governor of New York. You should be angry—very angry.
And if you want to let Eric T. Schneiderman, the New York AG, know how angry you are about his attack on the supplement industry, here’s his email address: [email protected]. Tell him his position on supplements is not a vote getting position but, in fact, vote losing. Tell all your friends to write him and tell him. This is politics. If enough people tell him his position will cost him future elections, he will change his position in a heartbeat. It's all about numbers. Pass it on.
- 1. "A.G. Schneiderman Asks Major Retailers To Halt Sales Of Certain Herbal Supplements As DNA Tests Fail To Detect Plant Materials Listed On Majority Of Products Tested." Attorney General Eric T. Schneiderman. Feb 3rd 2015. http://ag.ny.gov/press-release/ag-schneiderman-asks-major-retailers-halt-sales-certain-herbal-supplements-dna-tests
- 2. Daniel Gross. "Eliot Spitzer: How New York's attorney general became the most powerful man on Wall Street." Slate. 21 Oct 2007. (Accessed 8 Apr 2015.) http://www.slate.com/articles/news_and_politics/assessment/2004/10/eliot_spitzer.html
- 3. Michael Brynbaum. "Spitzer Resigns, Citing Personal Failings." The New York Times. 12 Mar 2008. (Accessed 8 Apr 2015.) http://www.nytimes.com/2008/03/12/nyregion/12cnd-resign.html
- 4. "Jim Schulte." Clarkson University. (Accessed 10 Apr 2015.) http://www.clarkson.edu/biology/faculty_pages/jim_schulte_profile.html
- 5. The Editorial Board. "Herbal Supplements Without Herbs." The New York Times. 6 Feb 2015. (Accessed 8 Apr 2015.) http://www.nytimes.com/2015/02/07/opinion/herbal-supplements-without-herbs.html
- 6. http://www.amazon.com/New-Vita-Nano-Water/dp/B00971VCLE
- 7. J. Felch, M. Dolan, "DNA: GENES AS EVIDENCE: FBI resists scrutiny." L.A Times, July 20, 2008. (Accessed 5 April 2015.) http://articles.latimes.com/2008/jul/20/local/me-dna20
- 8. "What Is DNA Barcoding?" Barcode of Life. (Accessed 8 Apr 2015.) http://www.barcodeoflife.org/content/about/what-dna-barcoding
- 9. Nandakumara Sarma. "DNA Testing of Herbal Supplements - Does it Work or Doesn't It?" QM USP. Feb 12, 2015. (8 Apr 2015.) http://qualitymatters.usp.org/dna-testing-herbal-supplements-does-it-work-or-doesnt-it
- 10. D Reynaud, B Mishler, J Neal-Kababick, P Brown. "The Capabilities and Limitations of DNA Barcoding of Botanical Dietary Supplements." AuthenTechnologies. March 2015. (Accessed 8 April 2015.) http://www.crnusa.org/NYAG/The-Capabilities-and-Limitations-of-DNA-Testing-FINAL-3-10-2015.pdf
- 11. http://www.usp.org/
- 12. Kathy Holliman. "DNA Barcode Testing of Herbal Supplements Questioned." Food Quality & Safety. April 2015. (Accessed 8 Apr 2015.) http://www.foodquality.com/details/article/7508461/DNA_Barcode_Testing_of_Herbal_Supplements_Questioned.html
- 13. "A.G. Schneiderman Announces Agreement With GNC To Implement Landmark Reforms For Herbal Supplements." Attorney General Eric T. Schneiderman. March 30th 2015. http://www.ag.ny.gov/press-release/ag-schneiderman-announces-agreement-gnc-implement-landmark-reforms-herbal-supplements
- 14. "Agreement between the New York State Office of the Attorney General ("NY AG") and GNC Holdings, Inc. ("GNC")." http://www.ag.ny.gov/pdfs/NYAG-GNC%20AGREEMENT.%20FINAL%20AGREEMENT.%203.28.2015..pdf
- 15. The Editorial Board. "Making Herbal Supplements Safer." The New York Times. 1 Apr 2015. (Accessed 8 Apr 2015.) http://www.nytimes.com/2015/04/01/opinion/making-herbal-supplements-safer.html