Supplements and Herbal Remedies

TL;DR. Supplements are regulated as food, not as medicine, which means nobody has to prove they work or, in most jurisdictions, that the bottle contains what the label says before it goes on sale. Independent testing repeatedly finds products with the wrong dose, the wrong species, or undeclared pharmaceutical ingredients. A short list of supplements has genuinely good evidence for specific people; almost everything else is a transfer of money. And "natural" does not mean safe: herbal products are the second most common cause of drug-induced liver injury in the US after antibiotics, and St John's wort is one of the most interaction-prone substances anyone can buy.

1. The regulatory gap

In the US, the Dietary Supplement Health and Education Act of 1994 (DSHEA) classifies supplements as food. Manufacturers do not need FDA approval before selling, do not need to prove efficacy, and do not need to prove safety in advance. The FDA must demonstrate a product is unsafe after it is on the market to act against it.

In the EU and UK, food supplements are regulated as food with permitted vitamin and mineral lists and authorised health claims, which is somewhat tighter. Traditional herbal medicines can be registered (THR scheme in the UK) on the basis of traditional use plus quality and safety, but not efficacy.

What this produces:

• Content problems. Independent testing programmes have repeatedly found products containing a fraction of the labelled dose, several times the labelled dose, or the wrong species entirely. A widely reported 2013 DNA barcoding study of North American herbal products found substitution or contamination in a majority of samples tested.
• Undeclared drugs. The FDA has issued hundreds of warnings about supplements spiked with actual pharmaceuticals: sildenafil analogues in "natural" erectile products, sibutramine (a withdrawn weight-loss drug) in slimming products, and corticosteroids in arthritis and eczema remedies. Athletes have failed drug tests from contaminated supplements.
• Heavy metals. Ayurvedic products in particular have repeatedly been found to contain lead, mercury, and arsenic, sometimes deliberately as part of rasa shastra preparations.

What to look for if you use supplements: third-party certification marks (USP Verified, NSF Certified for Sport, Informed Sport), a named standardised extract, the plant species and part used, and a reputable manufacturer. Informed Sport certification matters specifically for athletes, because strict liability rules mean a contaminated supplement is still your ban.

2. What actually has evidence

The short list, from Chapter 15 and elsewhere:

Everything not on this list should be assumed to lack good evidence until you check, and the default answer for most supplements in well-fed people is that trials have not shown benefit.

3. What has been shown to cause harm

This is the part that gets least attention.

Herbal and dietary supplements account for roughly 20 percent of drug-induced liver injury cases in US registries, second only to antibiotics, and the proportion has risen over time.

4. St John's wort: the interaction problem

Hypericum perforatum has reasonable evidence for mild to moderate depression, comparable to SSRIs in several trials. It is also one of the most interaction-prone substances available without prescription, and it is sold in supermarkets.

It is a potent inducer of CYP3A4 and of P-glycoprotein, which means it accelerates the clearance of a very large number of drugs:

If you take any prescription medicine, do not start St John's wort without asking a pharmacist. This is the clearest single reason to tell clinicians about herbal products.

5. Other interactions worth knowing

6. Categories worth a specific note

Multivitamins. In well-fed populations, trials have consistently failed to show benefit for cardiovascular disease or cancer, and the USPSTF recommends against beta-carotene and vitamin E specifically (Chapter 15). They are unlikely to harm at standard doses and are mostly a transfer of money. They are genuinely useful in documented deficiency, malabsorption, after bariatric surgery, in restricted diets, and in some older adults.

Protein powders. Food, conveniently packaged, and there is nothing wrong with them (Chapter 12). Independent testing has repeatedly found heavy metal contamination in some plant-based protein powders, and most people meeting their needs from food gain nothing.

Collagen. Widely marketed for skin and joints. Collagen is a protein and is digested into amino acids like any other; it does not travel to your skin intact. There is a modest evidence base for hydrolysed collagen peptides and skin elasticity and joint pain, and the mechanism proposed is signalling by specific dipeptides rather than direct supply. Treat it as an expensive and low-quality protein source with some interesting preliminary data.

Testosterone boosters, nootropics, "detox" products, and weight loss supplements are the categories where undeclared pharmaceutical adulteration is most frequently found.

Homeopathy deserves a plain statement. Its preparations are diluted to the point where, beyond about 12C, no molecule of the original substance is likely to remain. Systematic reviews, including the Australian NHMRC's comprehensive 2015 review and the UK House of Commons Science and Technology Committee's 2010 report, have concluded there is no reliable evidence of effectiveness beyond placebo. The direct harm is usually nil; the indirect harm, when it displaces effective treatment for serious illness, is not.

7. Practical guidance

1. Ask what the deficiency is. If you cannot name a plausible deficiency or a specific evidence-based indication, the supplement is unlikely to do anything.
2. Test rather than guess where a test exists: ferritin, vitamin D, B12.
3. Tell every clinician and pharmacist everything you take, including "just herbal" products. Surveys find a large majority of supplement users do not.
4. Buy third-party certified products if you use them, and be especially careful if you are tested for sport.
5. Prefer food where possible. Food supplies nutrients in the forms, ratios, and matrices that absorption evolved with, alongside fibre and thousands of other compounds.
6. Be most sceptical of the categories with the biggest promises: weight loss, sexual function, muscle building, cognition, and detox. These are where adulteration is concentrated.
7. Stop supplements before surgery.
8. "Natural" tells you about origin, not safety (Chapter 87).

8. The bottom line

• Supplements are regulated as food. Nobody has to prove they work, and in the US nobody has to prove they are safe or accurately labelled before sale.
• Independent testing repeatedly finds wrong doses, wrong species, heavy metals, and undeclared pharmaceuticals, particularly in weight loss, sexual function, and muscle-building products.
• The short list with real evidence is B12 for vegans, folic acid before pregnancy, vitamin D in winter, iron for documented deficiency, iodine in pregnancy, algal omega-3 for vegans, creatine, and caffeine.
• Several supplements have caused documented harm: beta-carotene in smokers, high-dose vitamin E, vitamin B6 neuropathy, green tea extract and kava liver failure, and aristolochia kidney failure.
• St John's wort induces CYP3A4 and causes contraceptive failure, transplant rejection, and loss of anticoagulation. Tell your pharmacist.
• Stop supplements affecting bleeding one to two weeks before surgery.

Sources and notes

US regulation follows the Dietary Supplement Health and Education Act 1994; EU and UK regulation follows the Food Supplements Directive and the Traditional Herbal Registration scheme. Species substitution in herbal products follows Newmaster et al., BMC Medicine, 2013, using DNA barcoding. Undeclared pharmaceutical adulteration follows FDA tainted products database analyses by Tucker et al., JAMA Network Open, 2018. Heavy metals in Ayurvedic products follow Saper et al., JAMA, 2004 and 2008. Supplement-associated liver injury proportions follow the US Drug-Induced Liver Injury Network reports and LiverTox. Specific harms follow: ATBC and CARET for beta-carotene; SELECT for vitamin E; Bjelakovic et al.'s Cochrane antioxidant mortality review; vitamin B6 neuropathy case series and the resulting EFSA and national dose restrictions; EFSA's 2018 green tea catechin assessment; kava hepatotoxicity reviews and national bans; aristolochic acid nephropathy from the Belgian slimming clinic cases (Vanherweghem et al., The Lancet, 1993) and Taiwanese urothelial cancer data; ephedra deaths and the 2004 US ban. St John's wort CYP3A4 and P-glycoprotein induction, and documented contraceptive failure and transplant rejection, follow Izzo and Ernst's reviews. Red yeast rice monacolin K identity with lovastatin follows EFSA's 2018 opinion. Creatine evidence follows the ISSN position stand. Homeopathy conclusions follow the Australian NHMRC 2015 review and the UK House of Commons Science and Technology Committee report, 2010. Perioperative supplement cessation follows ASA guidance.

Open questions. How widespread adulteration is across the whole supplement market, as opposed to in targeted enforcement samples, is unknown. Collagen peptide research is early and mechanistically unresolved.

👉 Next: medicines for children.