Bacopa

Bacopa monnieri is a creeping perennial with small, oblong leaves and small white to purple flowers (StatPearls). Its soft stems spread across mud and shallow water, and the Merck Manual describes it as a herb that commonly grows in wetlands in many areas of the world; StatPearls gives its native range as India, Indochina, Australia, and Sri Lanka. Other common names include water hyssop, thyme-leaved gratiola, Indian pennywort, and herb of grace. Older references, StatPearls among them, place the plant in the figwort family, Scrophulariaceae; NCBI Taxonomy (taxon 263974) now files it under Plantaginaceae, the plantain family, in the order Lamiales. The leaves are the part used medicinally, according to StatPearls, although some extracts, such as the one tested by Calabrese and colleagues, are made from the whole plant.

The name brahmi is the first trap for buyers. StatPearls explains that brahmi derives from Brahma, the creator in Hindu tradition, and that the term has been used for Bacopa monnieri, for Centella asiatica (gotu kola), or for a combination of the two. Gotu kola belongs to an entirely different family, the carrot family Apiaceae (NCBI Taxonomy 48106), and has different chemistry. Both plants have a reputation in Ayurveda for supporting the mind, and some products combine them, but the memory trials described on this page used Bacopa monnieri extracts. Their results cannot be transferred to a product made from gotu kola, and a label that says only “brahmi” does not tell you which plant is inside. Look for the Latin name.

StatPearls lists the plant’s constituents as triterpenoid saponins (bacoside A, bacoside B, and bacopasaponins), alkaloids such as brahmine and herpestine, flavonoids including luteolin and apigenin, and sterols such as stigmasterol and beta-sitosterol. It attributes most of the herb’s pharmacology to the saponins, collectively called bacosides. Bacoside A3 (PubChem CID 91827005, C47H76O18) is one of these triterpenoid saponins. Commercial extracts are standardised to bacoside content, but the standards differ: StatPearls notes that extracts are usually standardised to 24–55% bacosides, while the trials summarised by Pase and colleagues used extracts standardised to 10–20% bacopa glycosides. Branded extracts tested in trials include KeenMind (also coded CDRI 08), BacoMind, and Bacognize, and results from one cannot be assumed for another. Because bacopa grows in water and wet soils it also takes up metals from its surroundings, a quality issue covered under side effects.

Bacopa’s reputation comes from Ayurveda, the classical medicine of India. StatPearls notes that its use is documented in the Charaka Samhita, one of the founding texts of Ayurveda, as a treatment for mental conditions, and that Ayurvedic medicine classes it as a medhya rasayana: a group of herbs believed to improve mental health, memory, and intellect and to promote rejuvenation and longevity. Its traditional indications, according to StatPearls, were memory improvement, insomnia, epilepsy, and anxiety, and it has been described as a calming cognitive enhancer. The same Sanskrit-derived name, brahmi, has long been applied to gotu kola as well, so older texts and modern labels do not always distinguish the two plants.

Laboratory research in India and elsewhere followed the traditional claims. StatPearls summarises animal studies reporting better learning and memory in rodents given bacopa extracts or purified bacosides, reversal of drug-induced memory impairment in mice, and possible benefits in animal models of Alzheimer’s disease, epilepsy, Parkinson’s disease, and stroke, along with protection against gastric ulcers. These studies explain why bacopa attracted pharmaceutical interest, but animal models of memory and neurodegeneration have a poor record of predicting benefit in people, and none of these findings is evidence of a treatment.

Controlled human trials began around the turn of the century. Stough and colleagues in Victoria, Australia, published a 12-week placebo-controlled trial of the KeenMind extract in healthy adults in 2001, and Roodenrys and colleagues followed in 2002. Trials in older adults in the United States and Australia appeared in 2008 and 2010, systematic reviews in 2012 and 2014, and a review of Alzheimer’s disease trials in 2022. Bacopa is now a staple of memory supplements in the West, where it is sold as a dietary supplement rather than a medicine. StatPearls notes that the FDA’s position is that bacopa products are not approved for any medical purpose, and that in 2019 the agency warned supplement manufacturers about making therapeutic claims for products containing it.

Nobody knows for certain how bacopa might affect memory, and nearly all mechanistic work comes from cell cultures and animals. StatPearls attributes the cognitive effects mainly to the bacosides and lists several proposed mechanisms: changes in acetylcholine release, muscarinic receptor binding, and the activity of choline acetyltransferase, the enzyme that makes acetylcholine; inhibition of acetylcholinesterase, the enzyme that breaks it down; reduced beta-amyloid; increased cerebral blood flow; effects on monoamine neurotransmitters; antioxidant protection of brain tissue; modulation of the stress-hormone axis; and support for GABA signalling, which may explain its mild sedative and anti-anxiety reputation. The cholinergic effects are the most relevant for safety, because they are the basis of the interaction and contraindication warnings below.

Timing is the practical lesson. The trials that found benefits ran for 12 weeks, and Stough and colleagues (Psychopharmacology 2001), who tested people at 5 and 12 weeks, reported maximal effects after 12 weeks. Kongkeaw’s meta-analysis included only trials of at least 12 weeks for that reason. One small crossover study of 17 volunteers reported some effects on attention tasks and the stress hormone cortisol within hours of a single dose (StatPearls), but that has not been replicated in larger samples. Roodenrys and colleagues (Neuropsychopharmacology 2002) found that bacopa did not speed up learning but reduced how quickly newly learned material was forgotten, suggesting any effect is on retention rather than on alertness.

Bacopa also affects drug-metabolising enzymes in the laboratory. Ramasamy and colleagues (Molecules 2014) found that a standardised extract inhibited the human drug-metabolising enzymes CYP2C19, CYP2C9, CYP1A2, and CYP3A4 in test-tube assays, and that at the concentration expected in the gut after a 300 mg dose it cut CYP3A4, CYP2C9, and CYP2C19 activity to less than 10% of normal. Purified bacosides had little effect, so other constituents are responsible. No human interaction studies confirm this, but it is the basis for caution with medicines cleared by those enzymes. StatPearls suggests that the gut side effects may come from increased acetylcholine activity, irritation by the saponins, or both.

Memory in healthy adults is the main use and the best studied. Pase and colleagues (Journal of Alternative and Complementary Medicine 2012) systematically reviewed six randomised controlled trials in adults without dementia; all ran for 12 weeks and used one of three extracts at 300–450 mg a day. Bacopa improved performance on 9 of 17 tests of free recall, but there was little evidence of benefit in other cognitive domains, many of which had barely been studied. Kongkeaw and colleagues (Journal of Ethnopharmacology 2014) found nine trials with 518 participants that gave standardised extracts for at least 12 weeks. Pooling 437 participants, bacopa shortened completion times on the Trail Making Test part B, a test of attention and mental flexibility, and reduced choice reaction time. The authors judged the trials to be at low risk of bias but concluded that only a large head-to-head trial against an existing medicine, using a standardised preparation, would give definitive answers.

Individual trials show the pattern. Stough and colleagues (Psychopharmacology 2001) found that 300 mg a day of KeenMind improved the speed of visual information processing, learning rate, memory consolidation, and state anxiety compared with placebo. Roodenrys and colleagues (Neuropsychopharmacology 2002), studying 76 adults aged 40 to 65, found better retention of new information but no effect on attention, short-term memory, retrieval of existing knowledge, everyday memory, or anxiety. Calabrese and colleagues (Journal of Alternative and Complementary Medicine 2008) randomised 54 people aged 65 or older without dementia to 300 mg a day or placebo for 12 weeks; 48 finished, and the bacopa group improved on delayed word recall and the Stroop test of attention. Morgan and Stevens (Journal of Alternative and Complementary Medicine 2010) gave 98 healthy Australians over 55 the BacoMind extract, 300 mg a day, or placebo; among the 81 who finished, verbal learning and delayed recall improved, but other memory tests and self-rated memory did not differ significantly.

Not every trial is positive. Sathyanarayanan and colleagues (Psychopharmacology 2013) gave 72 healthy adults aged 35 to 60 in Bangalore 450 mg a day or placebo for 12 weeks and found no significant difference on any cognitive measure, only a trend towards lower anxiety. The Merck Manual, in a 2025 review, cites a six-week trial in 60 Indian medical students given 150 mg twice daily that improved some memory tests but not others, and concludes that despite modest benefits in small trials there are no high-quality studies demonstrating that bacopa improves memory or cognitive function. The trials are small, use different branded extracts and outcome tests, and measure statistical rather than clearly meaningful everyday changes.

Dementia and mild cognitive impairment are where marketing outruns evidence. Basheer and colleagues (Interactive Journal of Medical Research 2022) systematically reviewed bacopa in Alzheimer’s disease and found only five eligible trials, three of which combined bacopa with other herbs. Doses ranged from 125 mg to 500 mg twice daily, all five trials were judged at high risk of bias, and the certainty of evidence was very low. There was no difference between bacopa and placebo or donepezil, a standard dementia drug. Bacopa should not be used in place of a diagnosis or proven treatment for dementia, and nobody with a dementia diagnosis should start it without discussing their other medicines with a doctor.

Anxiety and other uses have thinner evidence. Several memory trials measured anxiety as a secondary outcome: Stough found lower state anxiety, Calabrese found falling combined anxiety and depression scores in the bacopa group, and Sathyanarayanan found only a non-significant trend. StatPearls mentions small trials in children, including an open-label study of attention-deficit hyperactivity disorder, and laboratory or animal work on epilepsy, Parkinson’s disease, stroke, and gastric ulcers. None of these is established. Bacopa is not a treatment for epilepsy, and anyone with a seizure disorder should not change treatment because of it.

Clinical trials used standardised dry extracts: 300 mg a day in the studies by Stough, Calabrese, and Morgan, 450 mg a day in Sathyanarayanan’s trial, and 150 mg twice daily in the trial of medical students described by the Merck Manual, usually for 12 weeks. StatPearls gives typical adult ranges from traditional practice and trials as 300–450 mg a day of extract, usually standardised to 24–55% bacosides; 5–10 g a day of crude dried herb, taken in two or three divided doses; or 10–20 ml a day of a 1:5 tincture, also in divided doses. There is no official European or US monograph setting a dose, and no regulator has approved bacopa for any medical use.

Products vary widely. Bacoside percentages, plant parts, extraction methods, and the branded extracts used in trials (KeenMind or CDRI 08, BacoMind, Bacognize) are not interchangeable, so a trial result applies only to a similar product at a similar dose. Check that the label names Bacopa monnieri rather than just brahmi, which can mean gotu kola. Prefer products that are tested for heavy metals by an independent laboratory: Saper and colleagues (JAMA 2008) found lead, mercury, or arsenic in about one in five Ayurvedic medicines bought online, and 75% of the contaminated products claimed to follow Good Manufacturing Practices, so that claim alone is not reassurance.

Expect any effect to take weeks. The positive trials assessed people after 12 weeks, so a few days of use says little either way, and the benefits seen were modest improvements on memory tests rather than dramatic change. If digestive side effects are troublesome, stop. StatPearls mentions use in children aged 6 to 12 at 225 mg a day for up to six months, based on an open-label study; given the small evidence base, do not give bacopa to a child without advice from their doctor, particularly a child taking medicines for attention problems or seizures.