How to Read a Herb Study or News Headline

The one-line version. A headline such as "herb X lowers Y" rarely tells you which plant, which part, which product, which people, or how big the effect was. NCCIH notes that the news is often a main source of health information, and while many reports are reliable, some are missing important information or are confusing, conflicting or misleading. [1] This page turns NCCIH's guidance into questions you can ask of any herb story.

Start with the type of study

NCCIH describes several kinds of research. The kind tells you how much weight a result can carry.

Questions to ask of the study

• How many people? Small samples may give inconclusive results and are more likely to produce results due to chance. [4]
• Was there a comparison group? In a controlled trial, results from the group given the treatment are compared with a control group given standard care or an inactive substance. [5]
• Was it blinded? NCCIH says to check whether the trial was "blinded" or "masked" so that neither participants nor investigators knew who received which treatment. [6]
• Statistically significant or clinically significant? A difference can be real yet so small that it is not clinically significant in terms of usefulness for patients. [7]
• Who paid for it? NCCIH suggests asking whether the sponsor or investigators had a financial or reputational stake in the outcome. [7]
• Is it the first report? One study rarely proves anything, and a single study rarely provides a final, definitive answer. [3][7]
• Did people like you take part? NCCIH asks whether participants were similar to you in ways that may matter, such as age, race or sex. [3]
• Did it last long enough? Natural products may take time before they show benefits, and some side effects may take months or years to show up. [3]

Which plant and which product?

A herb result applies to the plant, part and preparation that was tested. NCCIH says supplements bought in stores or online may differ in important ways from products tested in studies. [8] The St. John's wort entry in NIH's LactMed database adds that test results on one product may not apply to others. [9] See botanical names and plant parts on labels and fresh, dried or extract.

A worked example: ginkgo and memory

A news item might say a plant "protects the brain." NCCIH's ginkgo page reports a large study in which more than 3,000 people aged 75 or older were randomly assigned to a ginkgo supplement or a placebo for a median of 6 years, and rates of dementia did not differ. [10] NCCIH also notes studies of echinacea did not find benefit against the common cold. [11] A promising early finding does not guarantee a later one.

What to do with a story

1. Find the original study if you can. NCCIH notes it can be hard to know whether a news story is accurate unless you read the original sources. [12]
2. Notice who is quoted. A report is more likely to be right if it includes experts not connected to the study. [12]
3. Check the herb in official sources. How to check a herb for interactions lists them, and our evidence and safety table links every profile.
4. Before trying anything, talk with a health care provider or pharmacist. NCCIH says this no matter how promising an approach sounds. [3]

For how herb trials are built, see how herb clinical trials are designed. For the labels this site uses (established, promising but limited, mixed, insufficient evidence, preclinical, traditional use only) see traditional use vs research evidence.

Limits

Authors write their own conclusions, so others may not completely agree with their explanation of the results. [13] The sources we read do not give a numerical cut-off for a "big enough" study, so this page gives none. Proper evidence not found, in the NCCIH pages read, for a single score that rates study quality.