Regulation and adulteration

What it is

What is actually in the bottle, and who checked.

Most people assume that a product on a shelf has been assessed for whether it works and for whether it contains what the label says. For herbal and supplement products in most jurisdictions, neither assumption holds, and the gap between what people assume and what is true is the subject of this article.

Where it came from

Two decisions, on opposite sides of the world, made the modern situation.

The United States, 1994. The Dietary Supplement Health and Education Act placed supplements outside the drug approval system. A manufacturer does not need to demonstrate efficacy or safety before sale; the burden falls on the FDA to prove a product unsafe after it is on the market. The ephedra ban took the agency roughly a decade.

Australia, 1989 onward. The Therapeutic Goods Act created a two-tier system that is better and is still widely misunderstood. Registered medicines (AUST R) are individually assessed for quality, safety and efficacy. Listed medicines (AUST L) — which is what almost every herbal product on a supermarket shelf is — may only use pre-approved lower-risk ingredients and must be manufactured to a licensed standard, but are not assessed for efficacy before sale. The sponsor certifies that they hold evidence; the TGA does not check it up front.

An AUST L number therefore means "the ingredients are permitted and the factory is licensed". It is regularly read as "the government has confirmed this works", and it does not mean that.

How it is used

By manufacturers, entirely legally. The framework is public and the marketing is written to sit just inside it.

What we can and cannot say

We can report the contamination findings, which are the most alarming part.

Robert Saper and colleagues, publishing in JAMA in 2004 and again in 2008, bought Ayurvedic products from shops around Boston and later over the internet, and tested them. Around one in five contained detectable lead, mercury or arsenic, some at levels that would exceed acceptable daily intake many times over in a single dose. The second study found the internet- purchased products no better than the shop-bought ones.

Some of this is contamination. Some of it is not: rasa shastra is a branch of Ayurvedic practice that adds purified metals deliberately, on the doctrine that specific processing renders them therapeutic. Products in that tradition had the highest metal concentrations in the sample. That is a genuine doctrinal position within a living tradition, and it is also a toxicological fact about heavy metals, and both things are true.

We can say undeclared pharmaceuticals are a recurring finding. Products marketed for weight loss, erectile dysfunction, sports performance and pain relief have repeatedly been found to contain actual prescription drugs — sibutramine, sildenafil analogues, corticosteroids, NSAIDs — unlisted on the label. The FDA maintains a running list of hundreds of such products. This is not adulteration in the accidental sense. It is a product that works because it is a drug, sold as though it were a herb, to people who are not told they are taking a drug and may be taking something else that interacts with it.

We can say species substitution is routine. DNA barcoding studies of retail herbal products have repeatedly found that a substantial proportion contain species other than the one on the label, or contain fillers — rice, wheat, soy — instead of much of the stated ingredient. A widely reported 2015 New York Attorney General investigation of major retailers found exactly this, prompting a substantial industry response about barcoding methodology, some of which was fair. The underlying finding has been replicated in other markets.

Substitution is not merely a matter of value for money. The Belgian aristolochia cases were a substitution, and the consequence was permanent kidney failure.

We can say the wild-harvest problem is real and separate. Demand for slow-growing wild plants outstrips supply. American ginseng, goldenseal, several Aloe species and a long list of others are under CITES protection or on threatened lists. A market this size cannot be supplied from wild populations, and the pressure to substitute or to over-harvest follows directly from the economics.

And we can say what is actually checkable by a reader, since the rest of this is structural:

• An AUST L number is not an efficacy assessment. An AUST R number is.
• Buying from a local pharmacy gets you into a regulated supply chain, which imported online purchases sit outside entirely.
• Products making narrow, modest claims are more likely to be operating honestly than ones making broad ones, because the broad claims are the ones the framework does not permit for assessed products.
• Standardised extracts state their active content. Most preparations do not.

We cannot tell you which products are sound. What can be said is that the regulatory reassurance most people believe they are getting does not exist in the form they think, that the testing literature keeps finding things that should not be there, and that this is one of the few places on this site where the practical advice is simply to buy locally and tell your pharmacist.

Further reading

• Saper et al., JAMA (2004) and (2008), on heavy metals in Ayurvedic preparations.
• The TGA's own explanation of listed versus registered medicines — short and clearer than most commentary about it.
• The FDA's tainted products database, which is searchable and grimly educational.