Lion's Mane vs alpha-GPC vs citicoline vs bacopa for cognition

Updated 2026-05-20 · Reviewed by SupplementScore editors · No sponsorships · No affiliate links

Four supplements dominate the "nootropic" shelf in 2026: a culinary mushroom with a neurotrophic story, two synthetic choline precursors with crossover into Eastern European and Japanese prescription pharmacology, and an Ayurvedic herb that takes three months to do whatever it's going to do. They work on different parts of cognition, on different timelines, with different evidence weights. Choosing well requires being honest about what you're actually trying to fix — focus today, memory over weeks, or possible neuroprotection over years.

TL;DR — which supplement for which cognitive goal

The framework: what each of these is actually doing

"Cognition" is not one thing. The trial endpoints in this literature fall into roughly four buckets, and each supplement targets a different mix: attention (sustained focus, processing speed, reaction time — measured acutely), working memory (holding and manipulating information — measured acutely or subchronically), declarative memory (encoding and delayed recall — measured over weeks), and neurotrophic/structural endpoints (nerve growth factor expression, hippocampal volume, white-matter integrity — typically inferred from clinical conditions like MCI or post-stroke). A product that improves working memory in 90 minutes will not necessarily slow age-related cognitive decline, and vice versa.

The four supplements here group naturally into two mechanism families: the cholinergic precursors (alpha-GPC and citicoline) act fast by supplying choline for acetylcholine synthesis and phospholipid membrane repair; the herbals/fungals (bacopa, Lion's Mane) act slow, with effects emerging over 8–12 weeks via neurotrophic, antioxidant, or neuromodulatory pathways. Each family also has a niche clinical-population evidence base distinct from its healthy-adult marketing claims.

Citicoline (CDP-choline)

Citicoline — pharmacologically cytidine 5′-diphosphocholine — is a naturally occurring intermediate in phosphatidylcholine biosynthesis. Orally, it hydrolyses to cytidine and choline; the cytidine is converted to uridine, which appears to be a meaningful active metabolite for neuronal membrane and dopamine-system effects. Citicoline is an approved drug in Japan, Italy, and several Eastern European countries for stroke and TBI recovery (Somazina brand). In the US it is sold as a supplement, often under the trade name Cognizin.

The cleanest acute-attention data in healthy adults comes from the McGlade group: a 28-day RCT in adolescent females found 250 mg or 500 mg Cognizin produced modest but statistically significant improvements in attention and motor speed on a computerised vigilance task, with the 250 mg dose performing similarly to the 500 mg [1]. A follow-up MRS study in the same group showed increased brain phosphocreatine and ATP — consistent with citicoline's neuronal-membrane mechanism [2]. In healthy older adults with age-related memory complaints, 500 mg/day citicoline for 12 weeks produced small but reliable improvements in episodic memory vs placebo [3]. The post-stroke and MCI literature is larger and broadly positive, though heterogeneous; the international IRIS-I and Citicholinage trials provide the bulk of the modern data.

Practical use: 250 mg one to two times daily, swallowed whole or with food, effect on attention often perceptible within the first hour. Side-effect profile is excellent — occasional mild headache, GI discomfort, or transient insomnia at higher doses; no consistent serious adverse events in the trial literature. Drug interactions are minimal in healthy adults. The premium "Cognizin" form is the citicoline used in most US RCTs; generic citicoline at matched doses is reasonable when budget matters but trial-replication assurance is weaker.

Cost-per-effective-dose: $0.25–$0.90 per 250–500 mg serving.

Alpha-GPC (L-alpha-glycerylphosphorylcholine)

Alpha-GPC is a choline-containing phospholipid that crosses the blood-brain barrier and donates choline for acetylcholine synthesis. It is a prescription drug in Russia, Italy, and several Eastern European countries (brand name Gliatilin or Delecit) for cognitive impairment after stroke or in Alzheimer's, and an over-the-counter supplement elsewhere. The Alzheimer's-population data — most prominently the De Jesus Moreno Moreno 2003 RCT of 261 mild-to-moderate AD patients given 1,200 mg/day alpha-GPC for 180 days — show modest improvements vs placebo on standardised cognitive scales, comparable in magnitude to but not exceeding cholinesterase inhibitors [4]. Whether those results extrapolate to healthy adults at lower doses is the open question.

The sports-performance niche has the cleanest small-RCT data in non-clinical adults: 600 mg alpha-GPC taken 45–60 minutes pre-exercise has shown modest peak-power-output benefits in resistance-trained men in two small trials, plausibly via growth-hormone release and acetylcholine availability at the neuromuscular junction [5]. The acute-cognition evidence in healthy young adults is weaker — most "focus" claims rely on cross-over to the older-adult/MCI literature.

The complication: a 2021 observational study in over 12 million Korean adults reported an association between cumulative alpha-GPC supplementation and incident ischaemic stroke at 10-year follow-up [6]. The association was robust to several adjustments but is observational, vulnerable to confounding by indication (people taking alpha-GPC likely had subclinical vascular or cognitive complaints to start), and is the only large-scale signal of its kind. It is not a basis for panic but is a basis for caution in adults with vascular risk factors, especially at high doses (≥1,200 mg/day) used chronically. The supplement-industry response to this paper has been mostly to ignore it; the prudent response is to weigh it.

Practical use: 300–600 mg pre-workout for the sports niche, 400–1,200 mg/day in divided doses for the cognitive-decline indications, ideally under physician supervision. Side effects include headache, GI upset, insomnia, and the cholinergic "wired" feeling at higher doses. Avoid in pregnancy and in adults with significant cerebrovascular disease unless prescribed.

Cost-per-effective-dose: $0.30–$1.20 per 300–600 mg serving.

Bacopa monnieri

Bacopa monnieri is a creeping herb used in Ayurveda for over 1,000 years as a memory tonic. The active constituents are the bacosides — saponins concentrated in the leaf — which appear to act through multiple mechanisms: cholinergic enhancement, BDNF upregulation, antioxidant activity in the hippocampus, and modulation of serotonergic and GABAergic tone. Modern trials use extracts standardised to 20–55% bacosides, most commonly the proprietary CDRI-08 or BacoMind formulations.

The trial literature on bacopa is more substantial than the wellness-blog framing implies. Three meta-analyses of RCTs in healthy adults have converged on a small-to-moderate effect on delayed recall, learning rate, and information-processing speed, with effect sizes that emerge between week 8 and week 12 of supplementation [7][8][9]. The Calabrese 2008 90-day RCT (Bacopa 300 mg vs placebo, n=54 older adults) is the cleanest single trial — significant improvements in delayed recall on the AVLT and on the Stroop attention task at trial end, no effect at the interim time-points [10]. Stough 2008 found similar results on the AVLT in younger adults over 12 weeks.

The two important practical implications: this is a slow supplement (the trial-validated time course is months, not days — you should not expect to feel anything in week one), and the trial-validated dose is the equivalent of 300–450 mg of a standardised extract at ~55% bacosides, ideally taken with food because the saponins are fat-soluble and on an empty stomach commonly cause GI cramping, nausea, or diarrhoea (the most frequent reason for trial dropout). Bacopa is also a mild sedative for some users — better taken in the evening than the morning if it produces drowsiness.

Cost-per-effective-dose: $0.20–$0.80 per 300–450 mg serving. The cheapest of the four when the protocol is followed.

Lion's Mane (Hericium erinaceus)

Lion's Mane is an edible mushroom whose fruiting body and mycelium contain two distinctive constituent classes — hericenones (in the fruiting body) and erinacines (in the mycelium) — that stimulate nerve growth factor (NGF) synthesis in cultured cells and in animal models. The biological hypothesis is genuinely interesting: NGF supports cholinergic-neuron survival and plasticity in the basal forebrain, and pharmacological NGF stimulation has been a long-standing (and largely unsuccessful) drug-development target for Alzheimer's and peripheral neuropathy.

The human clinical evidence is more modest than the mechanistic story. The most-cited RCT is Mori 2009, a 16-week trial of 250 mg Lion's Mane (fruiting-body) 3×/day in 30 Japanese adults aged 50–80 with mild cognitive impairment; the supplement group improved on the revised Hasegawa Dementia Scale vs placebo, with effects diminishing 4 weeks after stopping [11]. A 2019 Mori follow-up in older adults with mild cognitive complaints replicated the direction-of-effect on a different scale at a similar 12-week timeline [12]. Beyond MCI, small trials have looked at depression and anxiety in menopausal women, post-stroke recovery, and peripheral nerve regeneration after surgery, each with positive-direction signals but small sample sizes and few independent replications.

The healthy-young-adult cognitive-enhancement use case — the dominant marketing pitch in the US — has very little dedicated trial weight. A short-duration (28-day) study in healthy young adults reported improvements in a sentence-completion task and speed-of-processing measures at 1.8 g/day Lion's Mane, but the sample (n=41) and effect magnitude were small. Most claims of "neurogenesis in healthy adults" generalise from cell-culture or rodent work in a way the human data have not yet supported.

Practical use: 1–3 g/day of a fruiting-body extract standardised for hericenones, taken consistently for at least 8–12 weeks before assessing effect. Side effects are minimal — occasional GI upset, rare contact dermatitis or respiratory sensitisation in regular handlers. Quality is a meaningful concern: many "Lion's Mane" supplements are predominantly mycelium grown on grain, which dilutes the active constituents with starch. Look for fruiting-body or dual-extract labels and verify with a third-party tester.

Cost-per-effective-dose: $0.50–$2.50 per 1–3 g serving (wide range driven by sourcing and extract quality).

Head-to-head matrix

Which should you pick — decision tree

Evidence quality call-out

Citicoline: Tier 2 in healthy adults; Tier 1 in stroke and MCI populations. Mixed funder mix including independent and academic [1][2][3].

Alpha-GPC: Tier 2 for MCI/Alzheimer's (prescription-grade indication); Tier 3 for healthy-adult cognitive endpoints; Tier 3 for sports performance. The Korean stroke signal warrants weighing [4][5][6].

Bacopa monnieri: Tier 2 for delayed recall and learning rate at 8–12 weeks. Multiple meta-analyses; consistent direction; modest effect sizes [7][8][9][10].

Lion's Mane: Tier 3 in healthy adults; Tier 2 in mild cognitive impairment. Mechanism well-characterised in vitro; translation to human cognitive endpoints supportive but not yet replicated at scale [11][12].

Funder mix: alpha-GPC and citicoline trials have a meaningful share of industry sponsorship (Cognizin, Gliatilin) alongside independent and government-funded post-stroke work. Bacopa trials are predominantly academic with some industry (CDRI-08, BacoMind manufacturers). Lion's Mane trials are weighted toward Japanese academic groups; the dominant Mori 2009 trial received support from a domestic supplement supplier [11].

Common misconceptions

"Lion's Mane grows new brain cells in healthy adults." NGF stimulation in cell culture and rodents is real; clinical demonstration of meaningful structural brain change in healthy young adults is not yet on the books. The honest framing is "modest effects in mild cognitive impairment, biological plausibility for more, evidence not yet there."

"Bacopa works in a few days." The trial-validated time course is 8–12 weeks. Subjective reports of acute effect on day 3 are not what the controlled literature shows; if you stop bacopa in week 4, you have not given it a fair trial [7][8][10].

"Alpha-GPC is just better-absorbed choline." The pharmacokinetics of alpha-GPC differ from dietary choline in clinically meaningful ways (faster BBB transit, more acetylcholine signalling), but the supplement-industry framing as "premium choline" understates both the prescription-drug status in some jurisdictions and the recent vascular safety question [6]. It is not equivalent to a choline-rich diet.

"Stacking all four is better than any one." No trial has compared a multi-product stack to any single component on cognitive endpoints. Stacking multiplies cost and obscures attribution; pick one, run a 12-week trial, then decide.

"A nootropic will fix my brain fog." "Brain fog" is most often a downstream symptom of sleep deprivation, insufficient cardiovascular fitness, perimenopause, undiagnosed hypothyroidism, depression, ADHD, post-viral syndrome, or a medication side effect. None of the four supplements on this page is a sensible first-line response without ruling out the higher-yield causes.

Who should not take any of these

Pregnant or breastfeeding women have limited safety data on all four and should avoid them in the absence of a specific medical recommendation. People taking cholinesterase inhibitors (donepezil, rivastigmine, galantamine) for dementia should not add alpha-GPC or citicoline without specialist input — the cholinergic stacking is unstudied. People on SSRIs, SNRIs, or other serotonergic agents should be cautious with bacopa, which has weak serotonergic modulation. People with mushroom allergies should avoid Lion's Mane. People on warfarin or other anticoagulants should clear bacopa with their prescriber (modest interaction signal). Anyone preparing for surgery should stop all four supplements 2 weeks ahead.

What we'd actually buy

For a healthy adult under 60 wanting to test a single cognitive supplement: third-party-tested citicoline (Cognizin or generic), 250 mg with breakfast, for 4 weeks of focused self-observation on attention and reaction time. If nothing perceivable, stop. For an adult 60+ adding a cognition-and-memory hedge: citicoline 500 mg/day plus a Lion's Mane fruiting-body extract 1–2 g/day, both for at least 12 weeks. For someone specifically targeting verbal memory and willing to commit to 12 weeks: bacopa 300 mg of standardised extract with the evening meal. For pre-workout cholinergic support: alpha-GPC 300–600 mg before the session — with the caveat that this is the indication where evidence is thinnest in healthy adults and where the recent safety signal is most relevant.

None of these recommendations are sponsored. Verified-tested brand options are listed on each supplement's individual page (look for the "Verified brands" panel).