Contamination: Heavy Metals, Bacteria and Substitution

Most pages in this series are about supplements that contain what their labels say and hurt people anyway. This page is about the bottle that does not match its label: lead where the label says turmeric, a mould where it says probiotic, a toxic plant where a harmless one was ordered. None of these can be seen, smelt or tasted, none announces itself with an immediate reaction, and all of them fall hardest on the people with the least margin — young children, pregnant women, premature infants and anyone whose immune system is suppressed. Every figure below comes from a peer-reviewed study or a government record, and each one is linked so you can read it yourself.

Table of Contents

  1. 1. Three Ways a Bottle Can Differ from Its Label
  2. 2. Heavy Metals in Ayurvedic Products: The Two Saper Surveys and Rasa Shastra
  3. 3. Toxic Elements Across Natural Health Products: The Genuis Comparison
  4. 4. Lead in Turmeric: Lead Chromate and the Yellow Trade
  5. 5. Protein Powders, Cocoa and Kelp: What the Measurements Actually Show
  6. 6. What Lead Does at Low Doses, and Why Children and Pregnancy Come First
  7. 7. Bacteria, Mould and Manufacturing: What “cGMP” and a Warning Letter Mean
  8. 8. The Balance of Nature Record: 2019 Warning Letters to 2023 Consent Decrees
  9. 9. Contaminated Probiotics and Raw Botanical Powders: The Cases on Record
  10. 10. Substitution: The Wrong Plant in the Capsule
  11. 11. When Substitution Causes Harm: Aristolochia in Belgium and Foxglove in “Plantain”
  12. 12. How to Reduce Your Risk
  13. 13. What to Do If You Are Worried
  14. 14. The Honest Bottom Line
  15. Research Papers
  16. Connections
  17. Featured Videos

1. Three Ways a Bottle Can Differ from Its Label

Three words are used almost interchangeably in news stories, and they describe three different failures. Keeping them apart tells you where to look and who is responsible.

The regulatory backdrop matters for all three. The FDA states plainly that it “does not have the authority to approve dietary supplements before they are marketed.” The manufacturer is required to follow a set of manufacturing rules (Section 7), and the FDA looks for problems afterwards — through inspections, through testing after complaints, and through outbreak investigations. That is why the evidence on this page comes from university surveys, public-health case reports and enforcement records rather than from any pre-market test. Nobody checked the bottle before it was sold; someone measured it afterwards.

Back to Table of Contents

2. Heavy Metals in Ayurvedic Products: The Two Saper Surveys and Rasa Shastra

The best-known numbers in this field come from two surveys led by Robert Saper of Harvard Medical School and then Boston University, both published in JAMA.

2004, Boston stores. The team searched out every store within 20 miles of Boston City Hall selling Ayurvedic herbal medicine products made in South Asia, bought every unique product they found — 70 in all — and measured lead, mercury and arsenic by X-ray fluorescence. Fourteen of the 70 (20%) contained at least one of the three metals: lead in 13 products (median 40 micrograms per gram, range 5 to 37,000), mercury in 6 (median 20,225 µg/g, range 28 to 104,000) and arsenic in 6 (median 430 µg/g, range 37 to 8,130). The authors’ conclusion was direct: “If taken as recommended by the manufacturers, each of these 14 could result in heavy metal intakes above published regulatory standards.”

2008, the internet. Four years later the same group tested whether the problem was a matter of where a product was made. From 25 websites selling Ayurvedic medicines they randomly selected 230 products, received 193, and found detectable lead, mercury or arsenic in 20.7% of them. The share was the same whether the product was made in the United States (21.7%) or in India (19.5%). Two details from that paper are worth remembering when you read a label. First, 75% of the metal-containing products claimed Good Manufacturing Practices on their labelling, and 95% were sold by US websites — the claim printed on the label had no relationship to the result in the laboratory. Second, every one of the metal-containing products exceeded at least one standard for acceptable daily intake of a toxic metal.

The 2008 study also separated a specific tradition. Rasa shastra is the branch of Ayurveda that deliberately combines herbs with metals, minerals and gems. In those products the metal is the ingredient, not an accident. Rasa shastra medicines were more than twice as likely to contain metals as the others (40.6% versus 17.1%), and their median mercury concentration was 20,800 µg/g against 34.5 µg/g — roughly 600 times higher. The practical rule that follows is simple: treat a rasa shastra product as containing lead or mercury unless a laboratory report for that lot says otherwise. The 17.1% figure for the non-rasa-shastra products is the other half of the lesson: avoiding the tradition that adds metals on purpose does not, by itself, make a product safe.

These surveys are not abstract. The CDC has published the case records. During 2000–2003, twelve adults in five states were reported to the CDC with lead poisoning associated with Ayurvedic medications. During 2011–2012, the New York City health department investigated six cases of lead poisoning in pregnant women, all foreign-born, all found through the lead-risk screening that New York state requires at prenatal visits. Their blood lead levels ran from 16 to 64 micrograms per decilitre (Section 6 explains what those numbers mean). Ten oral Ayurvedic medications made in India were involved; lead made up as much as 2.4% of some of them by weight, and several also contained mercury or arsenic. The department ordered a local business to stop selling them. The site’s pages on lead, mercury and arsenic cover what each does in the body.

Back to Table of Contents

3. Toxic Elements Across Natural Health Products: The Genuis Comparison

A 2012 study from the University of Alberta asked a sharper question than “are supplements contaminated?” It tested 121 natural health products — Ayurvedic, traditional Chinese, and various marine-source products — alongside 49 routinely prescribed pharmaceutical preparations, using the same laboratory methods for both. Toxic element contamination turned up in many products of both kinds, which is expected: trace metals are everywhere. The difference was in the amounts. Levels above established regulatory limits were found “only in a small percentage of the NHPs tested and none of the drugs tested.” Some of the natural products exceeded the preferred daily limits for mercury, cadmium, lead, arsenic or aluminium, and those manufactured in China generally carried more mercury and aluminium.

The comparison is the useful part. Prescription drugs are made under pharmaceutical manufacturing rules that require every incoming ingredient to be tested and every batch to meet a written specification; none of the 49 exceeded a limit. The chemistry of a herb is no harder to control than the chemistry of a tablet. What differs is the rule book and how hard it is enforced — a point the CDC investigators of a supplement outbreak made in almost the same words (Section 9).

Back to Table of Contents

4. Lead in Turmeric: Lead Chromate and the Yellow Trade

Turmeric is the clearest example of contamination that is neither natural nor accidental. Lead chromate is a bright yellow industrial pigment, and it has been added to dried turmeric root to make dull, poor-quality roots look vivid and sell for more.

A Stanford-led team documented the practice across Bangladesh in 2019. They visited the nine major turmeric-producing districts, interviewed 152 people along the supply chain and collected 524 samples of turmeric, pigments, dust and soil. They found evidence of lead-chromate adulteration in seven of the nine districts. Turmeric powder reached 1,152 µg/g of lead in the Dhaka and Munshiganj districts and 690 µg/g in the producing districts; the yellow pigments themselves were 2 to 10% lead by weight, and soil at polishing mills held up to 4,257 µg/g. Wholesalers told the researchers the practice began more than 30 years ago and continues, driven by consumer preference for brightly coloured curries; farmers said merchants could sell otherwise poor roots at a profit by asking polishers to add pigment. The polishers doing it did not know the pigment was a neurotoxin.

This is not a distant problem. New York City’s health department tested 1,496 spice samples of more than 50 spices from 41 countries between 2008 and 2017, collected during investigations of lead-poisoned children and adults and surveys of local stores. More than half had detectable lead and more than 30% had over 2 parts per million. Spices bought abroad were significantly worse than spices bought in the United States, with the highest concentrations in spices purchased in Georgia, Bangladesh, Pakistan, Nepal and Morocco. A 2017 commentary in Public Health Reports summarised the US evidence: turmeric with excessive lead is available in US grocery stores, and childhood lead-poisoning cases attributable to it have occurred here. The index case was a 12-month-old boy seen at Boston Children’s Hospital in 2010 with a blood lead level of 28 µg/dL; the Massachusetts health department traced the exposure to daily consumption of several lead-contaminated spices, including turmeric. Between 2010 and 2014, five more childhood cases attributable to culinary spices were documented by health departments in Arizona, California, Colorado, Connecticut and New York. The Boston Children’s group that treated the first child also measured 86 imported spices and 71 cultural powders: 22 of the spices and 46 of the powders contained more than 1 µg/g of lead, and three sindoor products (a ceremonial powder) were more than 47% lead.

The same mechanism surfaced in a food rather than a spice jar in 2023. The FDA’s lead page records the October 2023 recall of WanaBana apple cinnamon fruit puree pouches “due to elevated lead levels linked to cinnamon in those products,” and the agency’s follow-up work is titled for what it found: elevated lead and chromium. Lead chromate, again, in a product marketed to toddlers.

For supplement buyers the translation is this. A capsule labelled “turmeric root powder” is ground turmeric in a different package, and it is taken every day, which is exactly the exposure pattern that raises blood lead. Standardised curcumin extracts are a different product that these studies did not test; whether they carry lead depends on the extraction and on whether anyone measured it. In either case, the only thing that answers the question is a certificate of analysis for that lot showing a lead result — not the word “pure” on the front.

Back to Table of Contents

5. Protein Powders, Cocoa and Kelp: What the Measurements Actually Show

Three product categories generate recurring headlines about heavy metals, and the evidence behind each is of a different quality. This page tells you which is which.

Protein powders. The numbers usually quoted come from consumer-group testing, not from peer-reviewed studies: a Consumer Reports analysis of 15 products, which found that three servings a day would exceed the daily limits proposed by the US Pharmacopeia, and a later advocacy-group screen of 133 products that reported “elevated” metals in 40%. Neither was published in a journal with its methods reviewed. The one peer-reviewed analysis of those measurements, published in 2020 by scientists at a consulting firm (Cardno ChemRisk), took the reported arsenic, cadmium, mercury and lead concentrations and ran them through the EPA’s standard risk models at one and three servings a day. It concluded that the hazard index stayed below 1 — the threshold of concern — for every product; that the metals driving the result were arsenic and cadmium, not lead; that the highest values, approaching the threshold, were in “mass gain” products and the lowest in whey; and that modelled adult blood lead stayed below 5 µg/dL. An honest reading is that the metals in protein powders are real, they come mostly from plant ingredients (plants take up cadmium and arsenic from soil), and the amounts measured so far have not been shown to harm an adult at ordinary use. It is not evidence that a given tub is clean, and the words “clean” or “lab-tested” on the front are marketing until the company shows you the lot’s certificate of analysis.

Cocoa and dark chocolate. Here there is a proper multi-year dataset. Researchers at George Washington University and ConsumerLab tested 72 cocoa-containing products bought in the United States in 2014, 2016, 2019 and 2022. Measured against California’s Proposition 65 maximum allowable dose levels — 0.5 µg/day for lead and 4.1 µg/day for cadmium, thresholds the authors themselves call conservative — 43% of products exceeded the lead level per serving and 35% the cadmium level; none exceeded the arsenic level. Against the FDA’s interim reference level for lead, 97% of products fell below. The median product was under even the conservative limits, meaning the problem is outliers rather than chocolate in general; and, notably, products labelled “organic” were significantly more likely to be high in both cadmium and lead. FDA scientists have explained the sources in a 2020 perspective: cocoa picks up cadmium from soil, with the higher levels seen in some cocoa from parts of Latin America, and international mitigation work is under way. A cocoa or cacao “superfood” powder taken daily by the spoonful is the version of this that a supplement buyer should think about, because a spoonful is several servings.

Kelp and seaweed. A 2007 case report from the University of California, Davis, describes a 54-year-old woman with two years of worsening hair loss, memory loss, rash, fatigue, nausea and vomiting who was taking daily kelp supplements. Her urine arsenic was 83.6 µg per gram of creatinine (normal is under 50); her kelp contained 8.5 mg/kg of arsenic; within weeks of stopping, her symptoms resolved and arsenic was no longer detectable in her blood or urine. The authors then bought nine kelp supplements at local health-food stores: eight of the nine contained arsenic above the FDA tolerance level of 0.5 to 2 ppm that applies to certain foods, and none carried any warning. Seaweed arsenic needs one qualification. A UK Food Standards Agency study of 31 retail seaweed samples found total arsenic in all of them (18 to 124 mg/kg), but the inorganic form — the form that causes cancer — only in hijiki, at 67 to 96 mg/kg; the other varieties were below detection for inorganic arsenic. The agency advised consumers not to eat hijiki. Most seaweed arsenic is organic and far less toxic, so a high total-arsenic number on a kelp product is not by itself alarming; a product that has never been tested for the inorganic form is a different matter.

Back to Table of Contents

6. What Lead Does at Low Doses, and Why Children and Pregnancy Come First

Lead is the metal that turns up most often in the studies above, and the reason it matters at the small doses a capsule delivers is that its effects have no floor. The CDC’s position is stated in one sentence: “No safe blood lead level in children has been identified.” The FDA’s food-safety page says the same thing from the other direction: “Because there is no known safe level of exposure to lead, the FDA monitors and regulates levels of lead in foods.”

The evidence for that position is the pooled analysis published by Bruce Lanphear and colleagues in 2005, which followed 1,333 children from seven population-based cohort studies from infancy to age 5–10. Raising a child’s concurrent blood lead from 2.4 to 30 µg/dL was associated with a loss of 6.9 IQ points — and 3.9 of those points were lost between 2.4 and 10 µg/dL. The curve is steepest at the lowest exposures. That is why the CDC lowered its blood lead reference value for children from 5.0 to 3.5 µg/dL on October 28, 2021; the value is the level at which the agency recommends action, not a level that is safe. Lead does its damage during active brain development, and the FDA notes that the very young are especially vulnerable “because of their smaller body sizes and rapid metabolism and growth.” In pregnancy, the CDC’s investigators wrote, fetal exposure “can adversely affect neurodevelopment, decrease fetal growth, and increase the risk for premature birth and miscarriage” — which is why the six New York cases in Section 2 were caught by prenatal screening.

Adults are not exempt. Lanphear’s 2018 follow-up of 14,289 US adults in the NHANES III survey, followed for a median of 19 years, found that moving from a blood lead of 1.0 to 6.7 µg/dL (the 10th to the 90th percentile of the population) was associated with a 37% higher risk of death from any cause, a 70% higher risk of cardiovascular death and roughly double the risk of death from ischaemic heart disease. The authors estimated that low-level lead exposure accounts for about 412,000 deaths a year in the United States, mostly cardiovascular. Chronic lead exposure in adults is also associated, in the FDA’s summary, with kidney dysfunction and hypertension. These are population figures, not a prediction for one person; the point is that lead from a daily supplement adds to a background exposure whose effects are already measurable, and that a “small” amount taken every day for years is not small.

Mercury and arsenic have their own profiles and their own pages on this site: mercury and arsenic, with the whole group at Toxic Minerals.

Back to Table of Contents

7. Bacteria, Mould and Manufacturing: What “cGMP” and a Warning Letter Mean

Metals are usually a raw-material problem. Bacteria and mould are usually a manufacturing problem, and so is a product with 200 times the labelled dose of a mineral (Section 9). Understanding two terms lets you read an enforcement record for yourself.

cGMP. “Current Good Manufacturing Practice” for dietary supplements is a federal regulation, Title 21 of the Code of Federal Regulations, Part 111, which the FDA published as a final rule on June 25, 2007. It does not require the FDA to test anything before sale. It requires the company to write down what its product is supposed to be and to prove, batch by batch, that it is. The sections that matter most for this page: §111.70 requires written specifications for the identity, purity, strength and composition of every ingredient and every finished batch, and “limits on those types of contamination that may adulterate” the product; §111.75 requires tests or examinations to confirm those specifications are met, including confirming the identity of each incoming ingredient; §111.365 sets out the precautions that must be taken to prevent contamination during manufacturing; §111.205 requires a master manufacturing record for every formulation; and §§111.553–111.570 require written procedures for handling, investigating and keeping records of product complaints. Subpart B even requires procedures “for preventing microbial contamination from sick or infected personnel.” The full text is on the eCFR site, linked under Regulatory Records.

A warning letter. The FDA describes it this way: “When FDA identifies what it believes are significant violation(s) of federal requirement(s), FDA notifies the appropriate party. This notification is often in the form of a Warning Letter. The Warning Letter identifies the concern(s), such as poor manufacturing practices, problems with claims for what a product can do, or incorrect directions for use.” The company is given a chance to respond with a plan for correction; the FDA then checks, usually by a follow-up inspection, and may issue a close-out letter once corrections are verified. A warning letter is therefore a public record of what an inspector found on a given date, written in regulatory language, with the company’s response noted. It is not a finding that anyone was harmed, and its absence is not a finding that a company is compliant — the FDA inspects a fraction of facilities in any year. All letters are searchable on fda.gov, and products found to contain hidden drugs are listed separately in the Health Fraud Product Database.

When a company does not correct what a warning letter describes, the next step is a court order. The Department of Justice files a complaint on the FDA’s behalf, and the case usually ends in a consent decree of permanent injunction: an order the company consents to that stops it operating until the FDA is satisfied. The next section walks through one such record in full, because it is the one most often mentioned in the videos that prompted this series.

Back to Table of Contents

8. The Balance of Nature Record: 2019 Warning Letters to 2023 Consent Decrees

Balance of Nature is a widely advertised brand of fruit and vegetable capsules. The public record about it consists of two FDA warning letters and an FDA press release announcing court orders. This section reports what those documents say and nothing more; in particular, none of them describes any illness or injury, and neither does this page.

The 2019 inspections. The FDA inspected the St. George, Utah facility of Premium Production LLC, the contract manufacturer, on February 6–8, 2019, and the distributor Evig LLC (doing business as Balance of Nature) on February 4–8, 2019. The warning letter to Premium Production, dated July 29, 2019, states that the inspection “revealed serious violations” of Part 111 and that the products were therefore adulterated. The specifics are the sections in Section 7: “you have no in-process, raw material, or finished product specifications” (§111.70); “you have no written procedures for the responsibilities of the quality control operations” (§111.103); “you do not have written master manufacturing records for any of the dietary supplements you manufacture” (§111.205); and batch records that omitted the equipment used and the identity of each component (§111.260). The company’s written response said it would establish these; the FDA replied that it could not evaluate the response because no documentation was provided.

The warning letter to Evig, dated August 20, 2019, covers two things. On labelling, the FDA reviewed the company’s website and YouTube channel and concluded that the claims made there — the letter quotes videos titled “How to Help Overcome Relapsing MS” and “No More Inhalers for Her Asthma,” and product-label statements about diabetes, arthritis and cholesterol — established the products as unapproved new drugs and misbranded drugs. On manufacturing, the letter states that Evig, as a distributor using a contract manufacturer, had “not established and implemented a system of production and process controls” (§111.55), had no written quality-control procedures, and had no written procedures for product complaints (§§111.553 and 111.560). The letter is explicit that contracting out manufacturing does not contract out responsibility for the product.

The 2023 consent decrees. On November 16, 2023, the FDA announced that the U.S. District Court for the District of Utah, Central Division, had entered two consent decrees of permanent injunction: against Evig LLC and its CEO, Douglas Lex Howard, and against Premium Production LLC and its manager, Ryan Petersen. The press release states that Balance of Nature products carried “labeling that rendered them unapproved new drugs and misbranded drugs,” with claims that they “could be used to diagnose, cure, mitigate, treat, or prevent diseases such as cancer, heart disease, cirrhosis, diabetes, asthma, and COVID-19”; that Evig “violated current good manufacturing practice (CGMP) requirements, which rendered its products adulterated dietary supplements”; and that Premium Production manufactured the products “in violation of multiple CGMP requirements, such as failing to establish ingredient and finished products specifications for identity, purity, strength, and composition.” It records that after the 2019 warning letters, “FDA inspections showed that the firms failed to address the deviations,” which is why the agency pursued injunctions. The FDA’s Acting Associate Commissioner for Regulatory Affairs is quoted: “the public cannot have confidence that their products are what they purport to be.”

The decrees prohibit both firms from manufacturing or distributing products until they comply with CGMP and labelling regulations; both must hire CGMP experts, submit documents demonstrating compliance and receive the FDA’s approval to resume operations, and Evig must also hire a labelling expert. The Department of Justice filed the complaints on the FDA’s behalf. That is the record. What it shows is not a poisoning; it is a company that, on the FDA’s account, had no written specification for what its product contained and no procedure for handling a complaint about it — which is exactly the situation in which contamination, if it occurred, would go undetected.

Back to Table of Contents

9. Contaminated Probiotics and Raw Botanical Powders: The Cases on Record

Microbial contamination of a supplement rarely makes a healthy adult seriously ill. The documented harm is concentrated in three groups: premature infants, people whose immune systems are suppressed, and people who ingest an agricultural product that was never cooked.

A contaminated probiotic, Connecticut, 2014. The CDC’s MMWR reported the case of an infant born at 29 weeks and 1,400 grams who had received a probiotic powder, ABC Dophilus Powder made for Solgar, for seven days beginning on the first day of life. About a week after birth the infant developed what looked like necrotizing enterocolitis; at surgery the entire gastrointestinal tract was found to be without blood supply, and the infant died. Tissue examination showed an invasive fungal infection, gastrointestinal mucormycosis, caused by the mould Rhizopus oryzae. The CDC’s health advisory of November 25, 2014 records that unopened bottles of the same lot (074 024 01R1) were contaminated with the same organism, and that Solgar recalled three lots on November 14, 2014. The product was sold as a dietary supplement and, as the CDC’s advisory noted, was not regulated as a drug; the CDC advised clinicians who had used it in infants to review those patients.

Salmonella in kratom, 2018. Kratom is a dried, powdered leaf — an agricultural product, not a sterile one — and in 2018 it caused a multistate outbreak. The FDA and CDC investigators’ published account counts 199 people in 41 states, with 54 hospitalisations and no deaths, and product traceback to imports from several countries, most often Indonesia. Of 76 kratom samples the FDA collected from retailers and distributors, 42 (55%) were positive for Salmonella, in many different strains; the investigators concluded the plant was “widely contaminated across multiple sites from which it was grown, harvested, and packaged.” When one company, Triangle Pharmanaturals, refused the FDA’s repeated requests to recall its products voluntarily, the agency issued the first mandatory food recall order in its history, in April 2018, after two of the firm’s products bought in Oregon and four more samples collected by the FDA tested positive. (Kratom’s own pharmacology is covered in Sedatives, Phenibut and Kratom.)

A manufacturing error, 2008. Not every manufacturing failure is microbial. The CDC investigated 201 cases in 10 states of acute selenium poisoning from a liquid supplement that contained 200 times the labelled concentration of selenium; the median estimated intake was 41,749 µg a day against a recommended 55. Diarrhoea, fatigue, hair loss, joint pain and discoloured or lost fingernails were the common symptoms, and some persisted beyond 90 days. The investigators’ final sentence belongs on this page: “Had the manufacturers been held to standards used in the pharmaceutical industry, it may have been prevented.” See Selenium Toxicity and Vitamin and Mineral Overdose.

Probiotics that are not contaminated. A probiotic does not have to be contaminated to cause an infection in the wrong patient. A 2005 report in Pediatrics described two children who developed bloodstream infection and sepsis from Lactobacillus; DNA fingerprinting showed the organism in their blood was indistinguishable from the probiotic strain they had been given. The authors were careful: infections from probiotic lactobacilli are extremely rare and the report “should not discourage the appropriate use” of probiotics, but these agents “can cause invasive disease in certain populations.” A 2006 safety review in the American Journal of Clinical Nutrition reached the same balance — an excellent overall safety record, with caution “particularly [in] neonates born prematurely or with immune deficiency,” and a reminder that effects are strain-specific and cannot be generalised from one product to another.

The common thread is that a capsule is not cooked. A spice goes into a hot pan; a powdered leaf or a live culture goes straight into the gut of whoever swallows it. For most people that is fine. For a premature infant, a person on chemotherapy or high-dose steroids, a transplant recipient or someone with a central venous line, it is a decision that should be made by the clinician who knows their immune status, with a specific strain, not from a shelf.

Back to Table of Contents

10. Substitution: The Wrong Plant in the Capsule

How often is the plant on the label actually in the bottle? The most complete answer is a 2019 analysis in Frontiers in Pharmacology that pooled every published DNA-based authentication study it could find: 5,957 commercial herbal products sold in 37 countries on six continents. Twenty-seven percent were adulterated — their contents did not match their labelled species. The rate varied by region: 33% for North America, 29% for the United States specifically, 47% for Europe, 23% for Asia. The adulterated products, in the author’s words, “contain undeclared contaminant, substitute, and filler species, or none of the labeled species, which all may be accidental or intentional, economically-motivated and fraudulent.” (An earlier and much-quoted 2013 DNA-barcoding paper on North American herbal products has been retracted and is deliberately not cited here.)

DNA authentication works the way a paternity test does: a short, species-specific stretch of DNA is read from the powder and matched against a reference library. It is powerful on whole-plant material — dried leaf, root, powder — and it has one honest limitation: a purified extract may contain little or no intact DNA, so a “not detected” result on an extract is not by itself proof of fraud. The 27% figure comes from products that could be tested.

The economics are not mysterious. A herb that is expensive, slow-growing, wild-harvested or in short supply is the one that gets replaced with a cheap relative that looks the same when dried and ground, or padded with a filler that adds weight. A capsule of brown powder gives no clue, and a company that never tests the identity of what it buys — the §111.75 test that the warning letters in Section 8 found missing — will pass the substitution straight through to the customer. This is also how a product can be both substituted and adulterated: a 2022 analysis in JAMA Network Open of 30 immune-support supplements bought on Amazon found that 17 had inaccurate labels; 13 were missing an ingredient the label listed, 9 contained substances the label did not declare, and none of the 30 carried a third-party certification seal.

Back to Table of Contents

11. When Substitution Causes Harm: Aristolochia in Belgium and Foxglove in “Plantain”

Most substitution cheats the buyer without hurting them. Two documented episodes show what happens when the substitute is itself a poison.

Brussels, 1990–1993. In 1993 a group of Brussels nephrologists reported in The Lancet nine women under 50 with rapidly progressive scarring of the kidneys — two index cases and seven more admitted for dialysis in 1991 and 1992 — who had all followed the same slimming regimen at the same clinic. The clinic had run for 15 years without trouble; in May 1990 it had added two Chinese herbs to its regimen, labelled Stephania tetrandra and Magnolia officinalis. The patients’ serum creatinine had doubled within about three months. Chemical analysis showed that the capsules had an alkaloid profile different from the plants they were supposed to contain. The following year aristolochic acid was identified in the herbs, and the mechanism became clear: the Stephania had been replaced, through what the later NEJM paper calls a manufacturing error, by Aristolochia fangchi — a plant that is both nephrotoxic and carcinogenic. The two roots have similar traditional names and look alike, and the substitution of one for the other is a recognised, recurring problem in the herb trade. The Belgian cohort grew to more than 100 patients.

The cancer followed the kidney failure. Because aristolochic acid binds to DNA, the Brussels team offered their patients with end-stage disease prophylactic removal of the native kidneys and ureters. Among the 39 who agreed, 18 (46%) already had urothelial carcinoma, and every tissue sample analysed contained aristolochic-acid DNA adducts; a cumulative dose above 200 grams of the herb carried the higher risk. The US National Toxicology Program lists aristolochic acids as “known to be human carcinogens” in its Report on Carcinogens, first listed in 2011, citing precisely this evidence. What began as one slimming clinic’s supply problem is now recognised as a worldwide disease — aristolochic acid nephropathy — occurring wherever traditional formulas substitute an Aristolochia species for another herb, and, in the Balkans, where wheat fields were contaminated with Aristolochia clematitis seeds. The Brussels group noted in 2008 that despite FDA warnings, botanicals containing aristolochic acid were still being sold on the internet.

Foxglove in “plantain,” United States, 1997. Plantain (Plantago, the roadside herb, not the banana) is a common ingredient in supplements sold for digestion. In 1997 two women were hospitalised with signs of digitalis toxicity — the heart-rhythm poisoning caused by foxglove — days after starting a supplement regimen sold for “internal cleansing.” FDA scientists, using chemical methods, traced the cause to a batch of plantain raw material that was contaminated with Digitalis lanata, the foxglove species from which the heart drug digoxin is extracted. They published the investigation in the New England Journal of Medicine in 1998. Nobody in that supply chain intended to sell foxglove; the leaves were harvested, dried and ground together, and no one tested the identity of the material before it went into capsules. The Digitalis case is the reason identity testing of every incoming lot is written into Part 111.

Back to Table of Contents

12. How to Reduce Your Risk

Back to Table of Contents

13. What to Do If You Are Worried

If you or a child have been taking a product in one of the risk categories above — an Ayurvedic or traditional formula, a daily spice supplement, an untested kelp product — ask for a blood lead level. It is a standard, inexpensive blood test, it is the same test used for prenatal screening, and it settles the question in a day. Chronic lead exposure in adults can be silent or show as fatigue, anaemia, high blood pressure or reduced kidney function. If arsenic is the concern, a urine arsenic test is the equivalent; the kelp patient in Section 5 had hair loss, memory loss, rash, fatigue and nausea that resolved within weeks of stopping the supplement. Mercury has its own tests. The Lab Tests section explains what each measures and how to read the result. Bring the product to the appointment; the doctor cannot look up a tested lot, but the manufacturer and the health department can.

If someone vulnerable develops a fever or becomes unwell after starting a probiotic or a raw botanical powder — a premature infant, a person on chemotherapy or immunosuppressants, anyone with a central line — that is an urgent medical visit, and the clinician needs to know the product and the strain so the laboratory can look for it.

Report it. Adverse events from supplements are reported to the FDA through MedWatch (link under Regulatory Records); since 2007 manufacturers have been required to report serious ones too, but the reports from consumers and clinicians are what connected the cases above. For a suspected acute poisoning, call Poison Control at 1-800-222-1222, any hour, and have the bottle in your hand.

Back to Table of Contents

14. The Honest Bottom Line

Contamination is not an exotic risk and it is not a universal one. In the largest surveys, one in five Ayurvedic products contained a toxic metal and one in four herbal products worldwide contained the wrong plant; most products in most categories were what they claimed to be. The failures cluster where you would expect: traditions that add metals on purpose, spices from supply chains with a pigment problem, raw agricultural powders that are never cooked, and companies that — on the FDA’s own record — never wrote down what their product was supposed to contain. Metals matter most for children and in pregnancy, because lead has no safe dose and its steepest damage is at the lowest levels; microbes matter most for the premature and the immunosuppressed; substitution matters most when the substitute is Aristolochia or foxglove. None of this can be detected by looking at the bottle. All of it can be detected by a laboratory, which is why the single most useful question a buyer can ask a supplement company is not “is it natural?” but “show me the test.”

Back to Table of Contents

Research Papers

Heavy Metals

  1. Saper RB, Kales SN, Paquin J, et al. Heavy metal content of Ayurvedic herbal medicine products. JAMA. 2004;292(23):2868-2873. doi:10.1001/jama.292.23.2868 — 14 of 70 (20%) Boston-area Ayurvedic products contained lead, mercury and/or arsenic; each could exceed regulatory intake standards if taken as directed.
  2. Saper RB, Phillips RS, Sehgal A, et al. Lead, mercury, and arsenic in US- and Indian-manufactured Ayurvedic medicines sold via the Internet. JAMA. 2008;300(8):915-923. doi:10.1001/jama.300.8.915 — 20.7% of 193 products contained a toxic metal; US- and Indian-made alike; rasa shastra 40.6% vs 17.1%; 75% of metal-containing products claimed GMP.
  3. Centers for Disease Control and Prevention. Lead poisoning in pregnant women who used Ayurvedic medications from India — New York City, 2011–2012. MMWR Morb Mortal Wkly Rep. 2012;61(33):641-646. PMID 22914225 — six pregnant women, blood lead 16–64 µg/dL; medications up to 2.4% lead by weight.
  4. Centers for Disease Control and Prevention. Lead poisoning associated with Ayurvedic medications — five states, 2000–2003. MMWR Morb Mortal Wkly Rep. 2004;53(26):582-584. PMID 15241300 — 12 adult cases reported to the CDC.
  5. Genuis SJ, Schwalfenberg G, Siy AK, Rodushkin I. Toxic element contamination of natural health products and pharmaceutical preparations. PLoS One. 2012;7(11):e49676. doi:10.1371/journal.pone.0049676 — 121 natural health products vs 49 prescription drugs; limits exceeded only by a small percentage of the natural products and by none of the drugs.
  6. Forsyth JE, Nurunnahar S, Islam SS, et al. Turmeric means “yellow” in Bengali: lead chromate pigments added to turmeric threaten public health across Bangladesh. Environ Res. 2019;179(Pt A):108722. doi:10.1016/j.envres.2019.108722 — pigment adulteration in 7 of 9 producing districts; turmeric lead up to 1,152 µg/g; practice more than 30 years old.
  7. Cowell W, Ireland T, Vorhees D, Heiger-Bernays W. Ground turmeric as a source of lead exposure in the United States. Public Health Rep. 2017;132(3):289-293. doi:10.1177/0033354917700109 — lead-contaminated turmeric in US grocery stores; the 2010 Boston case and five further US childhood cases 2010–2014.
  8. Lin CG, Schaider LA, Brabander DJ, Woolf AD. Pediatric lead exposure from imported Indian spices and cultural powders. Pediatrics. 2010;125(4):e828-e835. doi:10.1542/peds.2009-1396 — 22 of 86 spices and 46 of 71 cultural powders above 1 µg/g lead; three sindoor products over 47% lead.
  9. Hore P, Alex-Oni K, Sedlar S, Nagin D. A spoonful of lead: a 10-year look at spices as a potential source of lead exposure. J Public Health Manag Pract. 2019;25 Suppl 1:S63-S70. doi:10.1097/PHH.0000000000000876 — 1,496 New York City spice samples; over half with detectable lead, over 30% above 2 ppm; worst from spices bought abroad.
  10. Bandara SB, Towle KM, Monnot AD. A human health risk assessment of heavy metal ingestion among consumers of protein powder supplements. Toxicol Rep. 2020;7:1255-1262. doi:10.1016/j.toxrep.2020.08.001 — consulting-firm risk assessment of consumer-group measurements; hazard index below 1; arsenic and cadmium the drivers; highest in mass-gain products, lowest in whey.
  11. Hands JM, Anderson ML, Cooperman T, Balsky JE, Frame LA. A multi-year heavy metal analysis of 72 dark chocolate and cocoa products in the USA. Front Nutr. 2024;11:1366231. doi:10.3389/fnut.2024.1366231 — 43% exceeded the Prop 65 lead level and 35% the cadmium level per serving; 97% below the FDA interim reference level for lead; “organic” products higher in both.
  12. Abt E, Robin LP. Perspective on cadmium and lead in cocoa and chocolate. J Agric Food Chem. 2020;68(46):13008-13015. doi:10.1021/acs.jafc.9b08295 — FDA scientists on the sources of cadmium (soil, notably in parts of Latin America) and lead in cocoa, and mitigation efforts.
  13. Amster E, Tiwary A, Schenker MB. Case report: potential arsenic toxicosis secondary to herbal kelp supplement. Environ Health Perspect. 2007;115(4):606-608. doi:10.1289/ehp.9495 — symptomatic arsenic exposure from daily kelp that resolved on stopping; 8 of 9 store-bought kelp supplements above the FDA tolerance level for certain foods.
  14. Rose M, Lewis J, Langford N, et al. Arsenic in seaweed — forms, concentration and dietary exposure. Food Chem Toxicol. 2007;45(7):1263-1267. doi:10.1016/j.fct.2007.01.007 — total arsenic in all 31 retail seaweeds; inorganic arsenic (67–96 mg/kg) only in hijiki, which the UK Food Standards Agency advised against eating.
  15. Lanphear BP, Hornung R, Khoury J, et al. Low-level environmental lead exposure and children’s intellectual function: an international pooled analysis. Environ Health Perspect. 2005;113(7):894-899. doi:10.1289/ehp.7688 — 1,333 children in seven cohorts; 6.9 IQ points lost from 2.4 to 30 µg/dL, 3.9 of them below 10 µg/dL; the steepest loss at the lowest levels.
  16. Lanphear BP, Rauch S, Auinger P, Allen RW, Hornung RW. Low-level lead exposure and mortality in US adults: a population-based cohort study. Lancet Public Health. 2018;3(4):e177-e184. doi:10.1016/S2468-2667(18)30025-2 — 14,289 adults; blood lead 1.0 to 6.7 µg/dL associated with 37% higher all-cause and 70% higher cardiovascular mortality; an estimated 412,000 US deaths a year.

Microbial Contamination and Manufacturing

  1. Vallabhaneni S, Walker TA, Lockhart SR, et al. Notes from the field: fatal gastrointestinal mucormycosis in a premature infant associated with a contaminated dietary supplement — Connecticut, 2014. MMWR Morb Mortal Wkly Rep. 2015;64(6):155-156. PMC4584706 — Rhizopus oryzae in a probiotic powder given from the first day of life to a 29-week infant.
  2. Nsubuga J, Baugher J, Dahl E, et al. Multistate outbreak investigation of Salmonella infections linked to kratom: a focus on traceback, laboratory, and regulatory activities. J Food Prot. 2022;85(5):747-754. doi:10.4315/jfp-21-319 — 199 cases in 41 states, 54 hospitalisations; 42 of 76 FDA product samples positive; contamination across multiple growing and packing sites.
  3. MacFarquhar JK, Broussard DL, Melstrom P, et al. Acute selenium toxicity associated with a dietary supplement. Arch Intern Med. 2010;170(3):256-261. doi:10.1001/archinternmed.2009.495 — a liquid supplement with 200 times the labelled selenium; 201 cases in 10 states.
  4. Land MH, Rouster-Stevens K, Woods CR, et al. Lactobacillus sepsis associated with probiotic therapy. Pediatrics. 2005;115(1):178-181. doi:10.1542/peds.2004-2137 — two children with bloodstream infection by a strain indistinguishable from the probiotic they received.
  5. Boyle RJ, Robins-Browne RM, Tang ML. Probiotic use in clinical practice: what are the risks? Am J Clin Nutr. 2006;83(6):1256-1264. doi:10.1093/ajcn/83.6.1256 — excellent overall safety record; caution in premature neonates and immune deficiency; effects are strain-specific.

Substitution and Label Accuracy

  1. Ichim MC. The DNA-based authentication of commercial herbal products reveals their globally widespread adulteration. Front Pharmacol. 2019;10:1227. doi:10.3389/fphar.2019.01227 — 27% of 5,957 products in 37 countries adulterated; North America 33%, United States 29%.
  2. Crawford C, Avula B, Lindsey AT, et al. Analysis of select dietary supplement products marketed to support or boost the immune system. JAMA Netw Open. 2022;5(8):e2226040. doi:10.1001/jamanetworkopen.2022.26040 — 17 of 30 Amazon products mislabelled; 13 missing a listed ingredient, 9 with undeclared substances; no third-party seals.
  3. Vanherweghem JL, Depierreux M, Tielemans C, et al. Rapidly progressive interstitial renal fibrosis in young women: association with slimming regimen including Chinese herbs. Lancet. 1993;341(8842):387-391. doi:10.1016/0140-6736(93)92984-2 — the original Brussels cluster: nine women, seven on dialysis, after a regimen that added “Stephania tetrandra” in May 1990.
  4. Vanhaelen M, Vanhaelen-Fastre R, But P, Vanherweghem JL. Identification of aristolochic acid in Chinese herbs. Lancet. 1994;343(8890):174. doi:10.1016/S0140-6736(94)90964-4 — the toxin identified in the herbs supplied to the clinic.
  5. Nortier JL, Martinez MC, Schmeiser HH, et al. Urothelial carcinoma associated with the use of a Chinese herb (Aristolochia fangchi). N Engl J Med. 2000;342(23):1686-1692. doi:10.1056/NEJM200006083422301 — Stephania replaced by Aristolochia fangchi through a manufacturing error; urothelial carcinoma in 18 of 39 patients (46%); aristolochic-acid DNA adducts in every sample.
  6. Debelle FD, Vanherweghem JL, Nortier JL. Aristolochic acid nephropathy: a worldwide problem. Kidney Int. 2008;74(2):158-169. doi:10.1038/ki.2008.129 — a Belgian cohort of more than 100 patients; frequent substitution by aristolochic-acid herbs in Asia; contaminated wheat in the Balkans; products still sold online despite FDA warnings.
  7. Chen J, Zhao Q, Si D, et al. Comprehensive profiling of Stephania tetrandra (Fangji) by stepwise DFI and NL-dependent structure annotation algorithm-based UHPLC-Q-TOF-MS and direct authentication by LMJ-HRMS. J Pharm Biomed Anal. 2020;185:113225. doi:10.1016/j.jpba.2020.113225 — S. tetrandra “is often mistakenly substituted and adulterated with the nephrotoxic Aristolochia fangchi” because of similar names and appearance.
  8. Slifman NR, Obermeyer WR, Aloi BK, et al. Contamination of botanical dietary supplements by Digitalis lanata. N Engl J Med. 1998;339(12):806-811. doi:10.1056/NEJM199809173391204 — the FDA investigation of foxglove-contaminated plantain raw material.
  9. Hunter ES, Literman R, Handy SM. Utilizing big data to identify tiny toxic components: Digitalis. Foods. 2021;10(8):1794. doi:10.3390/foods10081794 — FDA scientists’ summary of the 1997 case (two women hospitalised days after starting an “internal cleansing” regimen; source a batch of Plantago) and a DNA method to detect Digitalis in mixtures.

Regulatory Records

  1. 21 CFR Part 111 — Current Good Manufacturing Practice in Manufacturing, Packaging, Labeling, or Holding Operations for Dietary Supplements (full text). ecfr.gov; FDA overview of the rule, published June 25, 2007. fda.gov
  2. FDA. About Warning and Close-Out Letters. fda.gov; the Warning Letters database. fda.gov; Questions and Answers on Dietary Supplements (no pre-market approval). fda.gov
  3. FDA Warning Letter to Premium Production LLC, July 29, 2019 (MARCS-CMS 579705). fda.gov; FDA Warning Letter to Evig LLC dba Balance of Nature, August 20, 2019 (MARCS-CMS 580888). fda.gov
  4. FDA press release, November 16, 2023: Federal Judge Enters Consent Decrees Against Utah-Based Dietary Supplement Distributor and Manufacturer of Balance of Nature Products. fda.gov
  5. FDA, April 2018: FDA Orders Mandatory Recall for Kratom Products Due to Risk of Salmonella (Triangle Pharmanaturals LLC) — the agency’s first mandatory food recall order. fda.gov
  6. CDC Health Alert Network, November 25, 2014: Fatal Gastrointestinal Mucormycosis in an Infant Following Ingestion of Contaminated Dietary Supplement — Connecticut, 2014 (ABC Dophilus Powder, lot 074 024 01R1; recall of November 14, 2014). archive.cdc.gov
  7. CDC. Blood Lead Reference Value — 3.5 µg/dL since October 28, 2021 (previously 5.0); “No safe blood lead level in children has been identified.” cdc.gov
  8. FDA. Lead in Food, Foodwares, and Dietary Supplements — “no known safe level of exposure to lead”; the October 2023 WanaBana cinnamon applesauce recall. fda.gov
  9. National Toxicology Program. Report on Carcinogens, Fifteenth Edition: Aristolochic Acids — “Known to be human carcinogens,” first listed in the Twelfth Report (2011). ntp.niehs.nih.gov
  10. FDA. Health Fraud Product Database (products found to contain hidden ingredients). fda.gov
  11. FDA MedWatch adverse-event reporting. fda.gov; Poison Control, 1-800-222-1222. poison.org

Back to Table of Contents

Connections

Back to Table of Contents