Huperzine A

What it is

Huperzine A comes from Huperzia serrata, a club moss known in traditional Chinese medicine as Qian Ceng Ta. The plant contains several alkaloids; huperzine A is by far the best studied. Chinese researchers developed it as a drug, and in China it is used in Alzheimer’s patients.

In the West, huperzine A is known mainly as a nootropic. In capsules it is marketed as a memory and focus booster, and also as an ingredient in pre-workout powders. A substance with two lives, then: a medicine in Asia, a supplement in online retail.

How it is supposed to work

Acetylcholine is the messenger substance for attention and memory. In Alzheimer’s disease, it is precisely the nerve cells that produce it that are lost. The approved Alzheimer’s medicines donepezil, rivastigmine and galantamine therefore target its breakdown: they inhibit acetylcholinesterase so that the available messenger acts for longer. Huperzine A does the same, reversibly and very selectively. It largely leaves the related butyrylcholinesterase alone.

This is not just laboratory knowledge. In a study of 12 healthy older people, the activity of the enzyme in red blood cells fell by 30 to 40 percent at the lowest level and by over 50 percent at the highest, while butyrylcholinesterase remained unchanged. After 48 hours, enzyme activity was still about 10 percent below baseline. Huperzine A is absorbed quickly, the blood level peaks after just under an hour, and the half-life is around 12 hours.

In addition, there are findings from cell and animal models: huperzine A dampens excessive signaling at the NMDA receptor and protects nerve cells against glutamate, oxidative stress and beta-amyloid. According to a review by the Chinese developer group, it crosses the blood-brain barrier better than donepezil or rivastigmine. Whether these protective effects play a role in humans is open.

What comes next

Research is not finished. In China, a double-blind phase II/III trial with a planned 720 Alzheimer’s patients has been recruiting since August 29, 2025, testing an extended-release tablet against placebo and against donepezil; according to the trial registry, it is due to be completed in August 2028. An extended-release form is also being tested in an early study in focal epileptic seizures, and in a study of 123 patients after brain surgery, the memory test declined less in the first 96 hours under huperzine A than without it. That study, however, was not blinded, and huperzine A was given by injection.

What the studies show

Multicenter trial in China (2002)

Double-blind, placebo-controlled trial in 15 centers with 202 patients with mild to moderate Alzheimer’s dementia over 12 weeks. On the ADAS-Cog, patients under huperzine A improved by 4.6 points, and on the Mini-Mental State Examination by 2.7 points. A clear improvement of at least 4 ADAS-Cog points was achieved by 56.1 percent under huperzine A and 12.5 percent under placebo. Mild, transient side effects occurred in 3 percent.

US phase 2 of the ADCS (2011)

The Alzheimer’s Disease Cooperative Study tested huperzine A, funded by the US National Institute on Aging, in 210 patients divided between placebo and two dose levels, over at least 16 weeks; 177 completed the study. The primary endpoint, the ADAS-Cog at week 16 under the lower dose, was not reached. Under the higher dose, patients improved by 2.27 points at week 11, while they worsened by 0.29 points under placebo. At week 16, the difference of 1.92 versus 0.34 points was just above the significance threshold (p = 0.07). Activities of daily living and clinical global impression did not change in any group.

First double-blind study (1995)

Multicenter, placebo-controlled study with 103 Alzheimer’s patients, 50 under huperzine A and 53 under placebo, over 8 weeks. Memory, cognition and behavior improved in 58 percent under huperzine A and in 36 percent under placebo. No serious side effects were observed.

Where the data stop

The positive summaries rest on a narrow foundation. Most studies come from China, are small and have a high risk of bias; even the Cochrane review considered only one study sufficiently large and well conducted and therefore made no recommendation. An overview of 6 systematic reviews came to the same conclusion in 2021. The methodologically most careful study, from the US, missed its primary goal; the effect of the higher dose was no longer significant after 16 weeks, and nothing measurable changed in the patients’ daily lives.

Beyond Alzheimer’s disease, the data become even thinner. For mild cognitive impairment, Cochrane found not a single suitable study in 2012, and for vascular dementia only one, with 14 participants. For healthy adults hoping for sharper focus there is no robust study; the only work in cognitively healthy people dates from 1999 and examined 34 pairs of school students. Long-term data over many months are lacking, as are studies on chronic toxicity.

Status, approval and legal

In Germany, huperzine A is not an approved medicine, and food control authorities object to it in food supplements: German authorities reported it in the European Rapid Alert System RASFF in 2020 as an unauthorized substance and in 2022 as an unauthorized novel ingredient. EU-wide, the status has not been uniformly clarified. The European Commission’s Novel Food Catalogue has no entry; many member states treat huperzine A as an unauthorized novel food, some, such as Belgium, France and Romania, do not. A working group of the European food authorities chaired by the BVL (German Federal Office of Consumer Protection and Food Safety) provisionally classified it in 2024 as presumably not novel, but at the same time lists it among substances with a potential for harm. The Dutch RIVM has advised against consumption since 2024. In the US it is sold as a dietary supplement; the legal situation there is considered unclear. Huperzine A is not on the 2026 Prohibited List of the World Anti-Doping Agency.

Safety

In the studies, side effects were mostly mild and typical of cholinesterase inhibitors: nausea, vomiting, diarrhea, dizziness, sweating, blurred vision, insomnia and a slowed pulse. The RIVM assessed the data in 2024 and could not derive a safe intake level, because studies on long-term toxicity are lacking and the animal data contain indications of harm to the unborn child. The margins between the amounts in commercial products and the lowest harmful dose in animal experiments ranged from 26 to 429; 1,500 is considered sufficient. Two cases after overdose were reported to the Dutch Poisons Information Center, with nausea, tremor, speech disturbances and visual disturbances. Not during pregnancy and breastfeeding, not together with other cholinesterase inhibitors or with anticholinergics such as scopolamine; anyone taking prescription medicines should speak to their doctor first. Declining memory should be medically evaluated.

Sources

• Li et al., Cochrane Database of Systematic Reviews 2008 – huperzine A for Alzheimer’s disease
• Yang et al., PLoS One 2013 – meta-analysis of 20 randomized trials
• Rafii et al., Neurology 2011 – phase 2 trial of the Alzheimer's Disease Cooperative Study
• Zhang et al., Zhonghua Yi Xue Za Zhi 2002 – multicenter double-blind trial with 202 patients
• Haigh et al., Chemico-Biological Interactions 2008 – enzyme inhibition in the blood of healthy older people
• Yue et al., Cochrane Database of Systematic Reviews 2012 – huperzine A for mild cognitive impairment
• RIVM report 2024-0028 – risk assessment of preparations containing Huperzia serrata
• BVL 2024 – first report of the HoA Working Group on Food Supplements
• Vanhee et al., Journal of Xenobiotics 2025 – market surveillance of illegal nootropics in Europe
• ClinicalTrials.gov – phase II/III trial with huperzine A extended-release tablets in Alzheimer’s disease