Alpha-GPC (L-α-glycerylphosphorylcholine, also called choline alfoscerate) is a small, water-soluble molecule that the body uses as a source of choline. Choline is the building block of acetylcholine, the neurotransmitter that carries signals from nerves to muscle and plays a role in memory and attention. In some countries alpha-GPC is a prescription drug for cognitive decline; in the US it is sold as a dietary supplement, often marketed for focus and for power output in the gym.
The most important fact is the gap between those two uses. The drug evidence comes mostly from older patients with dementia, and much of it from one research group. The sports evidence comes from a handful of small, short, often industry-supplied trials with mixed results. A very large Korean cohort study has also linked alpha-GPC prescriptions to a higher long-term stroke risk, which has not been confirmed or refuted by trials.
This page summarizes what clinical trials, pharmacokinetic studies and regulators report. It is for education only and is not medical advice.
| Property | Detail |
|---|---|
| Class | Choline-containing phospholipid metabolite; cholinergic precursor |
| Other names | Choline alfoscerate, choline alphoscerate, L-α-GPC; Gliatilin (reference drug product in trials) |
| Formula | C₈H₂₀NO₆P, 257.22 g/mol |
| Route | Oral capsule or tablet; an intramuscular form was used in early studies |
| Studied doses | 1,200 mg/day (400 mg three times daily) in dementia trials; 250–1,000 mg in exercise studies |
| Half-life | ≈5–7 h for the rise in plasma choline after a 1,200 mg oral dose |
| US status | Not an FDA-approved drug; sold as a supplement. FDA had “no questions” about a GRAS notice for use in foods at 10–100 mg |
| Medicine elsewhere | First marketing authorization in 1985; prescription status differs by country |
| WADA status | Not on the 2026 Prohibited List |
What it is
Alpha-GPC is glycerol phosphate joined to choline. PubChem lists choline alfoscerate as C₈H₂₀NO₆P with a molecular weight of 257.22 g/mol (PubChem). It occurs naturally in the body as an intermediate in the breakdown of phosphatidylcholine, the main phospholipid in cell membranes. Foods also contain it: glycerophosphocholine is one of the water-soluble forms of dietary choline, alongside free choline and phosphocholine (Wiedeman et al., Nutrients 2018).
The rationale for taking it is simple. Alpha-GPC is hydrolyzed to free choline in the gut lining, and plasma choline rises after a dose (Lee et al., JAMA Netw Open 2021). The extra choline is meant to raise acetylcholine production in the brain and at the nerve-muscle junction. Whether a bigger choline supply actually translates into more acetylcholine signaling in healthy people is the open question behind both the cognitive and the sports claims.
As a medicine, choline alfoscerate received its first marketing authorization in 1985. Countries still disagree on whether it should be a prescription drug and on how well it works (Sagaro et al., J Alzheimers Dis 2023). Six of the eight studies in a 2023 meta-analysis were run in Italy. In South Korea it has been widely prescribed under national health insurance for dementia, the after-effects of cerebrovascular disease, mood and behavior changes and late-life depression (Lee et al., JAMA Netw Open 2021).
Pharmacokinetics
The studies below measured plasma free choline, the product of alpha-GPC breakdown, rather than alpha-GPC itself.
A randomized bioequivalence trial enrolled 48 healthy Korean men. Each took a single 1,200 mg dose as a tablet or as the reference soft capsule, in random order (Min et al., Drug Des Devel Ther 2019):
| Parameter (baseline-corrected plasma choline) | Tablet | Reference soft capsule |
|---|---|---|
| Peak concentration (Cmax) | 0.37 ± 0.16 µg/mL | 0.38 ± 0.11 µg/mL |
| Time to peak (Tmax) | 3.51 ± 2.57 h | 3.85 ± 3.19 h |
| Half-life | 6.82 ± 6.80 h | 5.25 ± 3.80 h |
| Total exposure (AUC to infinity) | 3.43 ± 2.17 µg·h/mL | 3.31 ± 1.80 µg·h/mL |
The very wide standard deviations show how much the choline response varies from person to person.
Other studies fill in the picture:
- Intramuscular dosing. In 12 healthy volunteers, a 1,000 mg intramuscular injection raised plasma choline within 15–30 minutes, and levels returned to near baseline by 6 hours. The rise was considerably larger than after the same dose of citicoline, another choline donor (Gatti et al., 1992).
- Single oral dose in young men. After 1,000 mg by mouth, plasma free choline was significantly higher at 60 and 120 minutes (Kawamura et al., Nutrition 2012).
- Dose dependence. In a 7-day trial, serum free choline rose by 59–132% vs placebo across the 250 mg and 500 mg groups (Marcus et al., JISSN 2017).
Regulatory status
- United States. No choline alfoscerate product appears in FDA’s drug approval database (Drugs@FDA). It is sold as a supplement ingredient. In 2012 FDA closed GRAS notice 419 with a “no questions” letter. The notice covered its use as a choline source in beverages, bars, cereals, yogurt and snacks at 10–100 mg (FDA, GRN 419). It is not a controlled substance.
- Other countries. It is a prescription or reimbursed medicine in several countries, with differing approval status (Sagaro et al., 2023).
- Sport. Alpha-GPC, choline and related choline donors do not appear on the WADA 2026 Prohibited List.
What the research shows
Alzheimer disease and dementia
The key placebo-controlled trial enrolled 261 patients aged 60–80 with mild to moderate Alzheimer disease in Mexico. They took 400 mg three times daily (1,200 mg/day) or placebo for 180 days (De Jesus Moreno Moreno, Clin Ther 2003). The ADAS-Cog scale scores cognitive errors, so a lower score is better:
| Measure | Choline alfoscerate (n = 132) | Placebo (n = 129) |
|---|---|---|
| ADAS-Cog change at day 90 | −2.42 points | +0.36 points |
| ADAS-Cog change at day 180 | −3.20 points | +2.90 points |
Differences between groups were also significant for the MMSE, the Global Deterioration Scale and the Clinical Global Impression.
The ASCOMALVA trial tested whether adding choline alfoscerate to donepezil (a standard Alzheimer drug) helps patients with Alzheimer disease plus vascular brain damage on imaging. It planned to enroll 210 patients. Interim reports covered 91 patients at 12 months and 113 at 24 months (Amenta et al., J Neurol Sci 2012; Amenta et al., J Alzheimers Dis 2014). The donepezil-plus-placebo group slowly worsened, while the combination group declined more slowly on most measures, though not on basic activities of daily living at 12 months.
Pooled evidence
A 2023 meta-analysis included 7 randomized trials and 1 cohort study run between 1993 and 2022, six of them in Italy (Sagaro et al., J Alzheimers Dis 2023):
| Comparison | Outcome | Studies | Mean difference (95% CI) |
|---|---|---|---|
| Donepezil + alpha-GPC vs donepezil | Cognition | 4 RCTs | 1.72 (0.20–3.25) |
| Donepezil + alpha-GPC vs donepezil | Function | 3 RCTs | 0.79 (0.34–1.23) |
| Donepezil + alpha-GPC vs donepezil | Behavior | 4 RCTs | −7.61 (−10.31 to −4.91) |
| Alpha-GPC vs placebo or other drugs | Cognition | Pooled | 3.50 (0.36–6.63) |
Two cautions apply. The donepezil-combination analyses include the ASCOMALVA trial, and the meta-analysis was written by the university group that ran it. An earlier review by members of that group counted 13 published trials with 4,054 patients, but three of them, covering 2,484 stroke and TIA patients, were uncontrolled (Parnetti et al., Mech Ageing Dev 2001).
Mild cognitive impairment
Results in milder impairment are less clear:
- Type 2 diabetes with mild impairment. A randomized trial in 36 patients (MMSE 25–28) compared 1,200 mg/day with placebo. The MMSE difference was +1.4 points at 6 months (p = 0.059, not significant) and +1.7 points at 12 months (p under 0.001) (Sohn et al., Diabetes Obes Metab 2025).
- Progression to dementia. An observational study of 3,062 propensity-matched pairs from three Korean hospital databases found no association between choline alfoscerate use and progression from mild cognitive impairment to dementia (HR 0.93, 95% CI 0.59–1.26) or to Alzheimer dementia (HR 1.05, 95% CI 0.51–1.59) (Pyun et al., Dement Neurocogn Disord 2024).
Strength and power in healthy adults
This is where most supplement marketing focuses, and the evidence is thin. The trials are small, last a week or less, and several used product supplied by the manufacturer.
| Study | Design | Dose | Main result |
|---|---|---|---|
| Bellar et al., JISSN 2015 | 13 college-aged men, crossover, 6 days | 600 mg/day | Mid-thigh pull peak force +98.8 N vs −39.0 N on placebo (p = 0.044); upper-body force not significant (p = 0.127) |
| Marcus et al., JISSN 2017 | 48 college-aged men, parallel groups, 7 days | 250 or 500 mg/day (vs caffeine 200 mg or placebo) | No differences in isometric strength or vigilance; jump velocity and power improved most at 250 mg |
| Kerksick, Nutrients 2024 | 20 resistance-trained men, crossover, single dose | 315 or 630 mg | No differences in jump, bench throw or growth hormone |
The isometric mid-thigh pull result in the first trial did not replicate in the larger second trial from the same lab.
Growth hormone
The claim that alpha-GPC boosts growth hormone rests on one small study. In 8 young men, a single 1,000 mg dose raised plasma growth hormone at 60 minutes and raised free fatty acids and ketones at 120 minutes, while placebo did not (Kawamura et al., Nutrition 2012). The later trial in 20 trained men found no growth hormone difference after exercise at 315 or 630 mg (Kerksick, 2024). A brief rise in growth hormone has not been shown to change muscle mass or strength.
Cognition in healthy adults
In the same 2024 crossover trial, both doses improved scores on the Stroop test of attention vs placebo:
- 630 mg: effect size d = 0.61 (p = 0.013);
- 315 mg: d = 0.48 (p = 0.046).
The Flanker and N-Back tests showed no significant differences. The trial was registered retrospectively and funded by an alpha-GPC manufacturer, for which the author is a paid advisor (Kerksick, 2024). In the 2017 trial, 250 or 500 mg/day for a week did not change psychomotor vigilance (Marcus et al., 2017).
Side effects and risks
Short-term tolerability
The controlled trials above report it as generally well tolerated. In the 48-man bioequivalence study, no serious adverse events occurred after a 1,200 mg dose (Min et al., 2019). One unexpected finding was a significant drop in thyroid-stimulating hormone (TSH) in the 500 mg/day group of the 7-day exercise trial (Marcus et al., 2017). It has not been studied further.
Stroke risk in a large cohort
The largest safety signal comes from South Korea’s national health insurance database. Researchers followed 12,008,977 adults aged 50 and over without prior stroke, TIA or dementia drug use. They compared 108,877 people prescribed alpha-GPC in 2006–2008 with non-users through January 2018 (Lee et al., JAMA Netw Open 2021):
| Outcome | Total cohort, aHR (95% CI) | Matched cohort, aHR (95% CI) |
|---|---|---|
| Total stroke | 1.46 (1.43–1.48) | 1.43 (1.41–1.46) |
| Ischemic stroke | 1.36 (1.33–1.39) | 1.34 (1.31–1.37) |
| Hemorrhagic stroke | 1.36 (1.28–1.44) | 1.37 (1.29–1.46) |
Risk also rose with longer prescriptions. Compared with under 2 months of use, total stroke risk was 1.13 times higher for 2–6 months of use and 1.36 times higher for over 12 months. The association remained after extra adjustment for smoking, alcohol, BMI, blood pressure, diabetes and cholesterol, and after excluding strokes in the first 4 years.
This is an observational study and cannot prove cause:
- users were older and sicker than non-users;
- the drug was often prescribed for early vascular or cognitive symptoms, which may themselves signal higher stroke risk;
- no trial has tested alpha-GPC’s effect on stroke.
A 2026 correction fixed swapped ICD-10 code labels in the Methods and an author-contribution entry (JAMA Netw Open 2026).
TMAO and atherosclerosis
The proposed mechanism is trimethylamine N-oxide (TMAO), a compound that gut bacteria and the liver make from choline and that is linked to atherosclerosis. Animal and human data so far point in different directions:
- In mice prone to high cholesterol (Apoe-deficient), dietary alpha-GPC raised TMAO, shifted gut bacteria and worsened atherosclerosis (Wang et al., Int J Mol Sci 2021). Mouse results do not show the same happens in people.
- In 12 healthy men, a single dose of GPC raised plasma choline but did not significantly raise plasma TMAO (Tanaka-Kanegae et al., Biosci Biotechnol Biochem 2024). Long-term human data are lacking.
Total choline intake
Each dose of alpha-GPC adds to total choline intake. The US Institute of Medicine set the adult tolerable upper intake level for choline at 3.5 g/day, based on low blood pressure seen at much higher doses. Adequate intakes are 550 mg/day for men and 425 mg/day for women (Wiedeman et al., 2018).
Groups with open questions
The evidence does not establish clear contraindications. The unresolved questions above are most relevant to:
- older adults and people with existing stroke or cardiovascular risk factors, the population in which the stroke association was seen;
- people with thyroid disease, given the single TSH finding;
- people taking dementia drugs such as donepezil, since evidence on the combination comes mainly from the trials described above.
Key takeaways
- A choline donor. Alpha-GPC is split into choline after absorption. A 1,200 mg dose raises plasma choline for several hours, with a choline half-life of ≈5–7 h.
- Dementia evidence is modest and concentrated. One 261-patient trial and the ASCOMALVA program showed benefits on cognitive scales. Much of the pooled evidence comes from one research group.
- Sports evidence is weak. Three small trials lasting a week or less gave inconsistent strength and power results. The growth hormone claim rests on one 8-person study and was not reproduced.
- A stroke signal exists. A 12-million-person Korean cohort linked prescriptions to a 1.3–1.5-fold higher stroke risk, rising with duration. It is observational and may reflect who gets prescribed the drug.
- Supplement in the US, drug elsewhere. It has no FDA drug approval, is GRAS for low-dose food use, and is not on the WADA 2026 Prohibited List.
Sources
- National Center for Biotechnology Information. Choline alfoscerate, PubChem CID 657272. PubChem.
- US Food and Drug Administration. GRAS Notice No. 419: L-alpha-glycerylphosphorylcholine (ChemiNutra). Closed November 20, 2012. FDA; Drugs@FDA.
- World Anti-Doping Agency. The 2026 Prohibited List. WADA.
- Wiedeman AM, Barr SI, Green TJ, et al. Dietary Choline Intake: Current State of Knowledge Across the Life Cycle. Nutrients. 2018;10(10):1513. PubMed.
- Min MH, Park JH, Hur JH, et al. Formulation and bioequivalence studies of choline alfoscerate tablet comparing with soft gelatin capsule in healthy male volunteers. Drug Des Devel Ther. 2019;13:1049–1058. PubMed.
- Gatti G, Barzaghi N, Acuto G, et al. A comparative study of free plasma choline levels following intramuscular administration of L-alpha-glycerylphosphorylcholine and citicoline in normal volunteers. Int J Clin Pharmacol Ther Toxicol. 1992;30(9):331–335. PubMed.
- Kawamura T, Okubo T, Sato K, et al. Glycerophosphocholine enhances growth hormone secretion and fat oxidation in young adults. Nutrition. 2012;28(11-12):1122–1126. PubMed.
- De Jesus Moreno Moreno M. Cognitive improvement in mild to moderate Alzheimer’s dementia after treatment with the acetylcholine precursor choline alfoscerate: a multicenter, double-blind, randomized, placebo-controlled trial. Clin Ther. 2003;25(1):178–193. PubMed.
- Amenta F, Carotenuto A, Fasanaro AM, et al. The ASCOMALVA trial: association between the cholinesterase inhibitor donepezil and the cholinergic precursor choline alphoscerate in Alzheimer’s disease with cerebrovascular injury: interim results. J Neurol Sci. 2012;322(1-2):96–101. PubMed.
- Amenta F, Carotenuto A, Fasanaro AM, et al. The ASCOMALVA (Association between the Cholinesterase Inhibitor Donepezil and the Cholinergic Precursor Choline Alphoscerate in Alzheimer’s Disease) trial: interim results after two years of treatment. J Alzheimers Dis. 2014;42(Suppl 3):S281–S288. PubMed.
- Sagaro GG, Traini E, Amenta F. Activity of Choline Alphoscerate on Adult-Onset Cognitive Dysfunctions: A Systematic Review and Meta-Analysis. J Alzheimers Dis. 2023;92(1):59–70. PubMed.
- Parnetti L, Amenta F, Gallai V. Choline alphoscerate in cognitive decline and in acute cerebrovascular disease: an analysis of published clinical data. Mech Ageing Dev. 2001;122(16):2041–2055. PubMed.
- Sohn M, Park YH, Lim S. Effects of choline alfoscerate on cognitive function and quality of life in type 2 diabetes: A double-blind, randomized, placebo-controlled trial. Diabetes Obes Metab. 2025;27(3):1350–1358. PubMed.
- Pyun JM, Lee I, Lee K, et al. Effect of Choline Alfoscerate on the Progression From Mild Cognitive Impairment to Dementia: Distributed Network Analysis of a Multicenter Korean Database Using a Common Data Model. Dement Neurocogn Disord. 2024;23(4):202–211. PubMed.
- Bellar D, LeBlanc NR, Campbell B. The effect of 6 days of alpha glycerylphosphorylcholine on isometric strength. J Int Soc Sports Nutr. 2015;12:42. PubMed.
- Marcus L, Soileau J, Judge LW, Bellar D. Evaluation of the effects of two doses of alpha glycerylphosphorylcholine on physical and psychomotor performance. J Int Soc Sports Nutr. 2017;14:39. PubMed.
- Kerksick CM. Acute Alpha-Glycerylphosphorylcholine Supplementation Enhances Cognitive Performance in Healthy Men. Nutrients. 2024;16(23):4240. PubMed.
- Lee G, Choi S, Chang J, et al. Association of L-α Glycerylphosphorylcholine With Subsequent Stroke Risk After 10 Years. JAMA Netw Open. 2021;4(11):e2136008. PubMed. Correction: Errors in Methods and Article Information. JAMA Netw Open. 2026;9(9):e2639482. PubMed.
- Wang Z, Hazen J, Jia X, et al. The Nutritional Supplement L-Alpha Glycerylphosphorylcholine Promotes Atherosclerosis. Int J Mol Sci. 2021;22(24):13477. PubMed.
- Tanaka-Kanegae R, Kimura H, Hamada K. Pharmacokinetics of soy-derived lysophosphatidylcholine compared with that of glycerophosphocholine: a randomized controlled trial. Biosci Biotechnol Biochem. 2024;88(6):648–655. PubMed.