Documented, Reported and Theoretical Risks: How to Read Safety Statements
How to read a safety statement about herbs and supplements: what documented, reported, preliminary and theoretical risks mean, with examples from NIH sources.
Content last updated . This page has not been medically reviewed.
Neither a firm warning nor a weak one tells you what to do about your own medicines. Your pharmacist or prescriber can. See also How interactions work.
The four labels we use
These four labels are our own plain-language grouping, applied to what each source says. The sources use their own words, which we quote or paraphrase in the last column.| Label | What it means here | Example, with the source's wording |
|---|---|---|
| Documented | The source says the interaction has been shown | NCCIH says clinically significant interactions between St. John's wort and medicines including cyclosporine, oral contraceptives, warfarin (coumadin), digoxin and benzodiazepines "have been documented". [1] |
| Reported | Case reports or other reports, not a controlled test | NCCIH says interactions between chamomile and some drugs metabolized by the liver, and warfarin, "have been reported". [1] |
| Preliminary | Early, small or limited studies | NCCIH says "preliminary studies suggest" chamomile with birth control pills might decrease the pills' effects. [1] It also cites "some preliminary evidence" that ashwagandha may increase the effects of some benzodiazepines. [2] |
| Theoretical | A reason to suspect it, with no human data in that source | NCCIH says there are "theoretical reasons to suspect" chamomile might interact with other drugs, such as sedatives. [1] |
Some findings do not fit neatly. NCCIH describes mixed results for Asian ginseng and warfarin, and conflicting evidence about cranberry and warfarin. [1] We call those mixed.
What each label does not tell you
It is a measure of evidence, not of danger. A documented interaction can be mild for one person and serious for another. A theoretical one is not proof of harm, and it is not proof of safety either.No reports is not the same as safe. NCCIH says well-designed clinical studies of herbal supplement and drug interactions are limited and sometimes inconclusive, and that it is a daunting challenge to identify all interactions of clinical concern. [1]
Animals and cells are not people. Findings from animals or cell tests are called preclinical. They can raise a question, but NCCIH points out that many herb-drug concerns rest on this kind of indirect evidence. [1]
A blood level is not a health outcome. NCCIH describes a study funded by the center in which metformin levels fell about 25 percent in healthy adults given goldenseal extract plus metformin, and notes the drop was enough to potentially hinder glucose control in people with type 2 diabetes. [1] That study measured drug levels in healthy adults, not blood sugar in people with diabetes.
The product and the amount matter. The green tea finding NCCIH cites is for green tea "at high doses" and the drug nadolol. [1] A result for one extract, plant part or amount does not automatically carry over to a tea, a food or another product.
When even a small chance matters
NCCIH's "Know the Science" module says interactions are especially important in two situations: when you take a medicine with a narrow therapeutic range, and when you are going to have surgery. [3] For medicines in that group, "even a little too low or too high" can cause serious reactions. [3] Examples listed are carbamazepine, cyclosporine, digoxin, phenytoin and warfarin. [3] For these, ask your prescriber or pharmacist before adding anything. See Supplements before surgery.Questions to ask of any safety statement
- Which product was studied: the same plant part, extract or form you are looking at?
- Was it people, animals or cells?
- Was the finding a drug level in the blood or something that happened to patients?
- Does the source name a specific medicine, or just say "may interact with many medicines"?
- Who wrote it, and what do they say they could not find?
For a longer look, see Interactions by drug class and St. John's wort interactions.
Never stop a medicine because of a warning
Do not stop, skip or change a prescribed medicine because of something you read. The NHS (England) says not to stop any prescribed medicine without first talking to your doctor. [4] Ask the prescriber or pharmacist.Frequently asked questions
What is the difference between a documented and a theoretical interaction?
A documented interaction is one a source says has been shown, for example the St. John's wort interactions NCCIH lists. [1] A theoretical one is a reason to suspect an interaction without human data in that source, such as chamomile with sedatives. [1]
If there are no reports of an interaction, is the combination safe?
Not necessarily. NCCIH says well-designed studies of herb-supplement and drug interactions are limited and sometimes inconclusive. [1]
Why do some sources say "hypothetical"?
NCCIH uses that word for concerns inferred from animal studies, cell tests or other indirect means rather than from rigorous research in people. [1]
Which medicines need the most care?
Medicines with a narrow therapeutic range, where a little too much or too little can cause serious reactions. NCCIH lists carbamazepine, cyclosporine, digoxin, phenytoin and warfarin. [3]
Who can tell me if my own medicines are affected?
A pharmacist or the prescriber. NCCIH suggests bringing a written list, or the bottles, of everything you take, including supplements. [5]
References
- NCCIH. Herb-Drug Interactions (Clinical Digest). National Institutes of Health, US, December 2024. Read 2026-10-23.
- NCCIH. Know the Science: How Medications and Supplements Can Interact, "Some Supplements May Increase the Effects". National Institutes of Health, US. Read 2026-10-23.
- NCCIH. Know the Science: Some Medication-Supplement Interactions Can Be Serious. National Institutes of Health, US. Read 2026-10-23.
- NHS (England). Herbal medicines and complementary therapies. Page last reviewed 1 May 2026. Read 2026-10-23.
- NCCIH. Know the Science: Talk With Your Health Care Providers. National Institutes of Health, US. Read 2026-10-23.