Herb Research by the Numbers: What a PubMed Snapshot Shows
A PubMed snapshot of how many studies mention 26 popular herbs, what the counts can and cannot tell you, and why a bigger number does not mean a herb works.
Content last updated . This page has not been medically reviewed.
Safety first. Natural products can cause side effects and interact with medicines. Read herb safety and interactions before you use anything on this page.
A count of studies is not a verdict. The table below shows how much has been published about each herb in PubMed, the US National Library of Medicine's index of biomedical articles. A big number means a lot has been written. It does not mean the herb works, is safe, or has been tested well. These counts are not evidence of benefit.
The snapshot
[[snapshot:herbs]]The counts come from this site's own snapshot, taken on 2026-10-22 through the NCBI E-utilities (the public search service for PubMed). [1] The last column of the table opens the same search so you can repeat it. Counts change as new papers are indexed.
How the sources define the terms
- Title/abstract search. The snapshot looks for the herb's names in the title or abstract of each record, using the PubMed field tag [tiab] (Title/Abstract). [2][1] For example, the turmeric search also includes the word curcumin. [1]
- Randomized controlled trial (RCT) tag. The National Library of Medicine's definition is a work reporting a clinical trial with at least one test treatment and one control treatment, in which treatments are selected by a random process. [3]
- Clinical trial tag. A work reporting results of a clinical study in which participants are assigned to receive one or more interventions so researchers can evaluate outcomes. [3]
- Who assigns the tags. PubMed says publication type data may be supplied by the publisher or assigned during MEDLINE indexing. [2] A record's tag therefore reflects labeling, not a quality check.
- All records. Every record that mentions the herb, including animal and laboratory work. PubMed itself offers separate filters for human or animal studies, which shows both are in the index. [2]
Counts next to what NCCIH concluded
The table pairs five rows of the snapshot with what the US National Center for Complementary and Integrative Health (NCCIH) says about the same herb. The point is that the count and the conclusion do not move together. The herbs are listed in the order shown here for readability, not as a ranking.| Herb (scientific name) | RCT-tagged records | All records | What NCCIH says |
|---|---|---|---|
| Elderberry (Sambucus nigra) | 18 | 1,372 | Some preliminary research suggests it may relieve symptoms of flu, colds or other upper respiratory infections. [1][4][5] |
| Feverfew (Tanacetum parthenium) | 9 | 442 | A 2020 review of studies for migraine found them small and using different sources and levels of feverfew, with inconsistent findings. [1][6][5] |
| Saw palmetto (Serenoa repens) | 85 | 669 | For urinary symptoms of prostate enlargement, NCCIH says we know enough to conclude it is probably not helpful. [1][7][5] |
| Ginkgo (Ginkgo biloba) | 337 | 5,886 | Many studies, but ginkgo has not been shown to be effective for any condition studied. [1][8][5] |
| Turmeric (Curcuma longa) and curcumin | 597 | 31,898 | NCCIH says we do not know enough to definitively conclude turmeric or curcumin is beneficial for any health purpose. [1][9][5] |
What to take from it: a herb with few trials (elderberry, feverfew) can still have a cautious "preliminary" or "mixed" label, and herbs with hundreds of trials (ginkgo, saw palmetto) can have clear negative or unproven conclusions. Each herb's own page has the detail: elderberry, ginkgo, saw palmetto and turmeric.
What the numbers can mislead about
- Direction. A trial may find no effect. The count includes those. [1]
- Quality and size. One trial of 20 people and one of 2,000 each count as one.
- Which product. The search words match the herb broadly, so records can involve extracts, mixtures with other ingredients or isolated compounds. A result for one preparation does not carry over to another.
- Humans or not. "All records" includes animal and lab studies, which show less about people. See human, animal and laboratory research.
- Safety. The snapshot does not count reports of harm, interactions or contamination. Proper evidence not found in this snapshot for safety: it was not built to measure it. See herb safety and interactions.
- Coverage. Only 26 herbs are in the snapshot. A herb missing from it may well have a lot of research too.
- Repeatability. We re-ran two rows on 2026-10-23 (elderberry and feverfew) and got the same counts as the snapshot: 1,372 and 442 total records, with 18 and 9 RCT-tagged records. [10]
Frequently asked questions
Does a higher number of studies mean a herb works better?
No. The count only shows how many records were indexed. NCCIH, for example, says ginkgo has not been shown effective for any condition studied despite many studies. [8]
What does "RCT-tagged" mean?
It means the record carries the Randomized Controlled Trial publication type, defined as a clinical trial with test and control treatments assigned by a random process. [3] The tag may come from the publisher or from indexing. [2]
Why are the totals so different from the trial counts?
The total includes every record that mentions the herb, and the search also catches laboratory and animal work. [1][2] In the snapshot, RCT-tagged records are a minority of the total for every herb.
Can I repeat the search myself?
Yes. The last column of the snapshot table links to a PubMed search for each herb. Counts may differ later as more papers are indexed. [1]
Where do I find what the research actually concluded?
On each herb's page, and in agency summaries such as NCCIH. For example, NCCIH says turmeric research is not yet enough to conclude benefit for any health purpose. [9] See herbs by wellness goal.
More details: evidence, limits and cautions
The details behind this page: cautions, limits of the evidence and where sources disagree. Open any note to read it.
Limits and uncertainty
Safety and who should ask a clinician
References
- AlternativeMed.com PubMed evidence snapshot for herbs, counted 2026-10-22 through NCBI E-utilities (site data file pubmedcountsherbs.json). Search wording for each herb is shown in the table's PubMed links.
- National Library of Medicine. PubMed User Guide. Read 2026-10-23.
- National Library of Medicine (NLM). MeSH Publication Types. Read 2026-10-23.
- National Center for Complementary and Integrative Health (NCCIH). Elderberry: Usefulness and Safety. Read 2026-10-23.
- GBIF Secretariat. GBIF Backbone Taxonomy. Names checked 2026-10-23 through the GBIF species match service (api.gbif.org).
- National Center for Complementary and Integrative Health (NCCIH). Feverfew: Usefulness and Safety. Read 2026-10-23.
- National Center for Complementary and Integrative Health (NCCIH). Saw Palmetto: Usefulness and Safety. Read 2026-10-23.
- National Center for Complementary and Integrative Health (NCCIH). Ginkgo: Usefulness and Safety. Read 2026-10-23.
- National Center for Complementary and Integrative Health (NCCIH). Turmeric: Usefulness and Safety. Read 2026-10-23.
- NCBI E-utilities esearch on PubMed, run by the writer on 2026-10-23 to re-check two rows of the snapshot (elderberry, feverfew). E-utilities.