ALBANY, N.Y. - DNA barcoding has exposed some infamous cases of food fraud, like cheap catfish sold as pricey grouper and expensive "sheep's milk" cheese that was really made from cow's milk.
But can it tell if a pill touted as an energy-booster contains ginseng or is just a mix of rice powder and pine?
Some scientists say yes, while industry groups and some independent experts say DNA testing alone is inadequate for analyzing botanical products that have gone through a lot of processing from leaf to tablet.
About 65,000 dietary supplements are on the market, consumed by more than 150 million Americans, according to a 2013 Canadian government study. The American Botanical Council estimates U.S. sales of herbal supplements came to $6 billion that year.
The U.S. Food and Drug Administration requires companies to verify their products are safe and properly labeled, but supplements are exempt from the FDA's strict approval process for prescription drugs.
New York Attorney General Eric Schneiderman says lax oversight of supplements can have serious public health consequences, noting a 2013 hepatitis outbreak traced to a tainted diet supplement and the death of a baby at a Connecticut hospital after doctors gave the child a probiotic supplement later found to be contaminated with yeast.
Last week, Schneiderman ordered Wal-Mart, Walgreen's, Target and GNC to stop selling store-brand herbal supplements that DNA tests found questionable.
Schneiderman's action followed a Clarkson University study he commissioned and he also referenced a 2013 study published by University of Guelph researchers.
The Clarkson study tested hundreds of bottles of store-brand herbal supplements sold as treatments for everything from memory loss to prostate trouble, and found 4 out of 5 contained none of the herbs listed on the labels.
The investigation looked into numerous supplements, including echinacea, ginseng, St. John's wort, garlic, ginkgo biloba and saw palmetto, were contaminated with substances including rice, beans, pine, citrus, asparagus, primrose, wheat, houseplant and wild carrot. In many cases, unlisted contaminants were the only plant material found in the product samples.
The retailer with the poorest showing was Walmart, where only 4 percent of the products tested showed DNA from the plants listed on the labels.
At the University of Guelph, researchers used DNA fingerprinting to find that a third of 44 supplements tested contained no trace of the plant on the label.
But the dietary supplement industry takes issue with DNA testing and some consumer advocacy groups say the method by itself is inadequate because it cannot detect the most serious problems with supplements - contamination with heavy metals or chemical adulterants.
"There's no problem with DNA barcoding as a science; however, it should be used appropriately. It has limitations," said Nandakumara Sarma, director of dietary supplements for US Pharmacopeia, which sets quality standards and testing protocols for drugs, vitamins and supplements.
The United Natural Products Alliance, a trade group, says it is sending people out around the country to buy large quantities of the supplements cited by Schneiderman and submit the bottles, unopened, to five or six certified botanical testing labs for analysis.
"They will perform universally accepted methods and procedures to test the products and will independently report their findings, which will be made public," said Loren Israelsen, the group's president. "We feel the most appropriate response to bad science is good science."
The American Botanical Council, a nonprofit research and education organization based in Austin, Texas, was particularly critical of the Guelph study.
"We raised the question if any of these products are extracts, and if so, what other analytical technologies were used to help ensure the validity of the results obtained by DNA testing," said Mark Blumenthal, the council's founder and director.
"DNA testing seldom is able to properly identify chemically complex herbal extracts, because often DNA doesn't get through the extraction process," Blumenthal said.
He said at least some of the products cited by Schneiderman are likely made from extracts, which can be validated by other common lab tests.
Tod Cooperman, president of the consumer-funded ConsumerLab.com, also dismissed the Guelph and attorney general's DNA studies as inappropriate for validating herbal supplements.
"There are definitely problems with herbal supplements but this is not the right method to test those products," Cooperman said.
His lab performs a range of tests on vitamins, herbal products and other dietary supplements to determine quality and purity, and provides the results to consumers who subscribe to his service.
Blumenthal also is working to expose bad products. He launched the Botanical Adulterants Program four years ago to ferret out suppliers who sell adulterated or mislabeled ingredients to manufacturers of herbal supplements. One goal of the program is to identify what laboratory tests are most effective.
Pieter Cohen, a Harvard Medical School researcher whose area of expertise is tracking down dangerous supplements, said the new versions of methamphetamine he has found in diet and sports supplements would never be spotted by DNA testing.
"There is so much wrong with the quality of supplements today that it's a shame the New York attorney general is not using sound science to focus on the most important problems," Cohen said. "The FDA has done hundreds of spot inspections of supplement companies, and they have found that 7 in 10 are not compliant with basic manufacturing practices."
Schneiderman's office stands by the validity of its DNA testing.
"Rather than attacking testing methods that have been validated by more than 70 published papers, the time has come for the herbal supplements industry to put concerns about what is and is not included in its products to rest," Matt Mittenthal, spokesman for the attorney general, said in an email.
The attorney general's office did not respond to a follow-up email asking why conventional, widely accepted testing wasn't used to verify the DNA test results, which would have made the findings more indisputable.
US Pharmacopeia provides independent third-party certification to drug and supplement manufacturers, who can then use the certification to assure consumers their products are genuine.
But John Atwater, director of verification programs for the organization, said less than 1 percent of herbal supplements carry the USP mark.
He noted the program is voluntary. "It's up to manufacturers and consumers to demand it," he said.