Why there are metals in plants at all

Plants take up what is in the soil. Lead, cadmium, arsenic and mercury occur naturally in the earth's crust at varying concentrations, and they are also deposited by industry, mining, traffic and historical pesticide use.

A review covering nearly two decades of published reports on 88 medicinal plant species concluded that heavy metal accumulation in medicinal plants "is mainly caused by extraction of soluble metals from contaminated soil, sediments and air." Across those reports, a maximum permissible level was exceeded for lead in 21 species, cadmium in 44, and mercury in 10.

Two things follow from that.

This is not primarily a story about bad actors. A root grown in soil with elevated cadmium will contain cadmium regardless of how ethical the farmer is. That is why the question is not "does this product contain any?" but how much, and whether anyone measured.

Roots concentrate more than leaves or fruit. That is awkward, because a great many popular botanicals are roots: turmeric, ginger, ashwagandha, ginseng, valerian, burdock.

There is also a second, separate pathway. Some traditional preparations (notably in rasa shastra, a branch of Ayurvedic practice) intentionally combine herbs with metals and minerals. That is a deliberate formulation choice, not environmental contamination, and it should never be blurred together with soil uptake. The two have different causes and very different magnitudes.

What FDA does and does not do here

This is where the expectations gap is widest.

FDA's "Closer to Zero" initiative does not cover supplements. The program targets arsenic, lead, cadmium and mercury in foods commonly eaten by babies and young children (infant cereals, purees, juices), prioritized because, in FDA's words, children's "smaller body sizes and metabolism make them more vulnerable to the harmful effects of these contaminants." Dietary supplements are not within its scope.

FDA has set lead action levels only for specific food categories: processed foods for babies and young children, juice, and candy likely to be frequently consumed by small children. No FDA action level for lead in dietary supplements was located.

FDA has published Interim Reference Levels for lead, derived from the CDC blood lead reference value with a tenfold safety factor:

• Children: 2.2 micrograms per day
• Females of childbearing age: 8.8 micrograms per day

These are important numbers, and they are widely misused. They are total dietary exposure benchmarks (everything you eat and drink in a day), not supplement limits and not label limits. A supplement does not "pass" or "meet" an IRL.

FDA also states plainly that "there is no known safe level of exposure to lead."

Now the Proposition 65 arithmetic

California's Proposition 65 requires a warning when a product exposes someone to a listed chemical above a defined threshold. Lead has been listed for reproductive toxicity since 1987.

The threshold (the Maximum Allowable Dose Level) for lead is 0.5 micrograms per day.

Put that next to the federal numbers:

The Prop 65 trigger is roughly a quarter of FDA's benchmark for children and about one-eighteenth of the benchmark for women of childbearing age.

That is the explanation for the warnings you keep seeing. The threshold is a daily figure, so a botanical product carrying ordinary environmental lead can take a person past 0.5 micrograms a day while sitting well below every federal figure in the table. The warning is a California disclosure requirement operating at a deliberately conservative threshold.

So a Prop 65 warning is not, by itself, evidence that a product is unusually contaminated. It is also not nothing: it means the manufacturer either measured a level above the threshold or chose to warn rather than test. You cannot tell which from the label, and that ambiguity is a fair criticism of how the system works in practice.

Is there any standard for supplements?

Not a federal one. There is a voluntary one.

The United States Pharmacopeia publishes a chapter on elemental contaminants in dietary supplements, which sets limits and requires analysis by inductively coupled plasma techniques. Published work citing that chapter identifies a permitted daily exposure for lead of 5 micrograms per day.

Two caveats. Published descriptions of this chapter differ in how they frame the limits (some as daily exposure figures, others as concentration limits), and the chapter itself is behind a paywall, so figures for arsenic, cadmium and mercury are not quoted here. And these limits are voluntary: they bind a company only if it chooses to comply or pursues a certification built on them. They are not FDA-enforced.

That last point is the reason third-party certification matters more in this category than in most. A certification program that screens for lead, mercury, arsenic and cadmium is applying a standard that would otherwise be optional.

What the survey data shows

The best-documented work concerns ayurvedic products, and it is worth reporting precisely because it is frequently cited loosely.

A 2004 study published in JAMA analyzed 70 South-Asian-manufactured herbal products bought in Boston-area stores. Fourteen of the 70 (20 percent, with a 95 percent confidence interval of 11 to 31 percent) contained lead, mercury and/or arsenic.

A follow-up published in JAMA in 2008 examined 193 products purchased over the internet. The overall prevalence of metal-containing products was 20.7 percent. Two findings from that study deserve to be stated directly:

US manufacture was not protective. US-made products showed 21.7 percent prevalence versus 19.5 percent for Indian-made, a difference that was not statistically significant.

A GMP claim was not protective either. Among the metal-containing products, 75 percent carried a Good Manufacturing Practices claim on the label.

Products in the rasa shastra tradition, where metals are added intentionally, showed substantially higher prevalence (40.6 percent versus 17.1 percent) and much higher median concentrations. This is, again, a different phenomenon from soil uptake.

One important limitation. Those purchases were made in 2003 through 2005. No comparable US-market survey of botanical supplements from the last fifteen years was located during this research. Anyone citing that 20 percent figure as a current US prevalence (including this article, if it were to do so) would be extrapolating from twenty-year-old sampling. The findings are solid for what they measured, and what they measured was a market two decades ago.

How testing works

The method. The standard technique is inductively coupled plasma mass spectrometry, usually written ICP-MS. It is what FDA itself uses: FDA's Elemental Analysis Manual includes a method for determining arsenic, cadmium, chromium, lead, mercury and other elements in food by ICP-MS following microwave-assisted digestion, and a separate method combining liquid chromatography with ICP-MS for methylmercury.

It is sensitive, it is well established, and it is not expensive enough to be a real barrier for a company that wants to do it.

Speciation matters, and is often skipped. Arsenic is the clearest case: inorganic arsenic and organic arsenic compounds have very different toxicity, and a total arsenic number does not distinguish them. A test reporting only total arsenic on a product like a seaweed-derived ingredient can look alarming for reasons that may not be meaningful. Asking whether arsenic was speciated is a sophisticated question that gets a revealing answer.

Certificates of analysis: what to ask for and how to read one

A certificate of analysis (a COA) is a testing record for a specific lot. It is the single most useful document a supplement company can give you, and also the most frequently misunderstood.

The federal manufacturing regulation itself tells you what a meaningful COA contains, with one important limit: a supplier's certificate can never substitute for the manufacturer's own identity test on a dietary ingredient. Where a manufacturer does rely on a supplier's certificate for other components, the regulation requires that the certificate "includes a description of the test or examination method(s) used, limits of the test or examinations, and actual results of the tests or examinations," that the supplier first be qualified by confirming its test results, that the qualification be documented, that the supplier be periodically re-confirmed, and that quality control personnel review and approve it.

Those requirements map directly onto the questions worth asking:

Which lot does it cover? A COA is lot-specific. One for a different lot tells you about a different batch of material.

Who ran the test? A supplier's own COA and an independent laboratory's report are different kinds of evidence. Both have a place; they are not interchangeable.

What method and what detection limits? "Non-detect" means nothing without a detection limit. Non-detect at a sensitive limit is reassuring; non-detect at a crude one may only mean the test was not sensitive enough to see anything.

Is it for the finished product or the raw ingredient? A clean raw material can be contaminated during manufacturing, and a clean ingredient COA does not cover the capsule.

How current is it? The regulation contemplates periodic re-confirmation. A COA from several years ago describes material from several years ago.

A company that emails a lot-specific finished-product COA with methods and detection limits stated has a real quality system. A company that sends a generic certificate with no lot number, or declines entirely, has told you something too.

What to do, practically

Ask for the COA for your lot. The lot number is on the bottle. The contact information is on the label because federal law requires it. This is the highest-value action available to you and hardly anyone takes it.

Favor third-party certification for botanicals specifically. Programs that screen for lead, mercury, arsenic and cadmium are doing the thing no federal rule requires. This matters more for roots and for traditional preparations than for, say, a synthesized B vitamin.

Take the rasa shastra distinction seriously. Traditional preparations that intentionally include metals are a different category with different risk, and they warrant a conversation with a clinician rather than a label check.

Think about total intake, not per-product. The federal benchmarks are for everything you eat in a day. Someone taking eight botanical supplements is in a different position from someone taking one, and no single label accounts for that.

Be most careful for children and during pregnancy. The federal reference levels are lower for children and for women of childbearing age for good reason. Supplements for children, and use during pregnancy or breastfeeding, are worth discussing with a pediatrician or an obstetric provider rather than deciding from packaging.

Do not read a Prop 65 warning as a verdict, but do not read its absence as clearance either. A product without the warning may have been tested and found low, or may simply not be sold in California.