Alpha-GPC
Controlled trials found short-term improvements in cognitive symptom scores with Alpha-GPC in Alzheimer disease and amnestic MCI. It is also registered as a medicine in several countries. Healthy nootropic and sports evidence is small and inconsistent, while the observational stroke association and long-term high-dose safety remain unresolved.
Choline alfoscerate / L-alpha-glycerylphosphorylcholine
- The 261-person Alzheimer trial found a six-point separation in ADAS-Cog change over 180 days, but used a one-tailed primary test and was sponsor-run.
- The 100-person MCI trial reported a 1.37-point difference in ADAS-Cog improvement; its efficacy analysis included only 74 participants.
- A 20-man acute healthy crossover improved Stroop change but not N-back, Flanker, mood, physical performance or growth hormone.
- Strength and power findings come from small studies and conflict with a 48-person trial that found no placebo advantage for isometric strength.
- A Korean claims cohort linked prescriptions with stroke; observational data cannot prove causation.
- No healthy dementia-prevention, muscle-gain or human-longevity benefit was found.
What is Alpha-GPC, and where is it used as a medicine?
Alpha-GPC is L-alpha-glycerylphosphorylcholine, also called choline alfoscerate. It supplies choline and glycerophosphate. Italy, Poland and Russia list nationally authorized or registered medicines, while the US GSRS record establishes identity rather than FDA approval. Oral supplements and registered injectable medicines are not interchangeable.
| Record | What it establishes | What it does not establish |
|---|---|---|
| EMA national list | Named Italian and Polish choline-alfoscerate medicines | FDA approval or healthy-user efficacy |
| Russian state register | Current oral and injectable products | Equivalence to US supplements |
| FDA GSRS | Ingredient identity | An approved indication, dose or label |
FDA GSRS · alpha-GPC identity
Government substance database
U.S. Food and Drug Administration. Global Substance Registration System, choline alfoscerate, UNII 60M22SGW66.
The record identifies choline alfoscerate/alpha-GPC as C8H20NO6P, molecular weight 257.22 g/mol, CAS 28319-77-9.
- Participants / model
- Not applicable
- Treatment
- Not applicable
- Follow-up
- Not applicable
- Study design
- Primary source record
A UNII is an identity code, not FDA drug approval.
Read the original sourceEMA list · Italy and Poland national products
EU regulatory product list
European Medicines Agency. List of nationally authorised medicinal products: active substance choline alfoscerate. EMA/174242/2023, PSUSA/00010599/202208. April 14, 2023.
The EMA list identifies nationally authorized choline alfoscerate products in Italy and Poland, including oral and injectable forms.
- Participants / model
- Patients covered by national product labels in Italy and Poland
- Treatment
- Choline alfoscerate medicinal products
- Follow-up
- Current list at the publication date
- Study design
- EU pharmacovigilance/national authorization list
This is not a centralized EMA approval and does not make every supplement or indication equivalent; local product labels govern.
Read the original sourceRussian registered products · oral and injectable forms
Russian government medicines/price register
Ministry of Health of the Russian Federation. State Register of Medicines / public price-limit records for choline alfoscerate and Cereton.
The official register surface lists current choline alfoscerate and Cereton registration records and oral or intravenous/intramuscular formulations in Russia.
- Participants / model
- Patients under Russian product-specific labels
- Treatment
- Registered choline alfoscerate products
- Follow-up
- Living register
- Study design
- Government medicine-register record
The register proves product registration, not a single universal indication or interchangeability across oral/injectable brands. Product-specific Russian instructions govern use.
Read the original sourceHow strong is the Alzheimer disease result?
The largest Alpha-GPC-alone trial randomized 261 adults with mild-to-moderate probable Alzheimer disease to 400 mg three times daily or placebo for 180 days; 229 completed. ADAS-Cog changed by -3.20 points with Alpha-GPC and +2.90 with placebo. Several secondary scales also favored treatment. This supports short-term symptomatic benefit; the trial did not measure disease modification or prevention.
| Design | Cognitive result | Material limits |
|---|---|---|
| 261 randomized; 132 active, 129 placebo; 229 completed | ADAS-Cog -3.20 vs +2.90 points | Prespecified one-tailed primary test; no biomarker, post-withdrawal or long-term progression outcome |
| Adverse-event comparison | Treatment-related events 11/132 vs 3/129, p=0.030 | Mainly constipation and nervousness; not a long-term safety study |
| Sponsorship | Italfarmaco supported design, conduct, analysis and publication | Independent replication remains valuable |
Alzheimer full paper · 261 randomized, 229 completed
Randomized double-blind placebo-controlled trial
De Jesus 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. Clinical Therapeutics. 2003;25(1):178-193. doi:10.1016/S0149-2918(03)90023-3. PMID 12637119.
Mean ADAS-Cog changed from 35.52 to 32.32 with choline alfoscerate and from 36.74 to 39.64 with placebo at day 180, equivalent to -3.20 versus +2.90 points. Several secondary scales also favored treatment.
- Participants / model
- 261 adults aged 60-80 with mild-to-moderate probable Alzheimer's dementia: 132 active and 129 placebo; 229 completed
- Treatment
- Oral choline alfoscerate 400 mg three times daily or placebo
- Follow-up
- 180 days
- Study design
- Multicenter randomized double-blind placebo-controlled trial; prespecified primary analysis used a one-tailed alpha of 0.05
- Funding
- Italfarmaco supported study design, conduct, analysis, and publication
Treatment-related adverse events occurred in 11/132 active participants (10 constipation, 5 nervousness events) and 3/129 placebo participants; none withdrew for these events. The trial was old, sponsor-run, used a one-tailed test, and assessed symptomatic scales rather than disease modification.
Read the original sourceWhat did the newer amnestic-MCI trial show?
The 2024 Korean study randomized 100 adults with amnestic MCI to Alpha-GPC or placebo for 12 weeks, but the efficacy analysis included 74: 42 active and 32 placebo. ADAS-Cog improved 2.34 points versus 0.97, a 1.37-point separation with p=0.048. This suggests a small symptomatic effect in MCI; the trial did not test prevention in cognitively normal adults.
| Flow | Result | Limit |
|---|---|---|
| 100 randomized; 74 in efficacy analysis | ADAS-Cog improvement 2.34 vs 0.97 points; p=0.048 | 26% absent from the reported efficacy denominator |
| Safety | No serious adverse events or AE-related withdrawals reported | Short and too small for rare events |
| Interpretation | Symptomatic score signal in diagnosed amnestic MCI | No dementia incidence, disease-modification or healthy-prevention outcome |
MCI trial · 100 randomized, 74 efficacy analysis
Randomized controlled trial
Jeon J, Lee SY, Lee S, Han C, Park GD, Kim SJ, Chang JG, Kim WJ. Efficacy and safety of choline alphoscerate for amnestic mild cognitive impairment: a randomized double-blind placebo-controlled trial. BMC Geriatrics. 2024;24:774. doi:10.1186/s12877-024-05366-7. PMID 39300341.
One hundred adults with amnestic MCI were randomized for 12 weeks, but the reported efficacy analysis included 74 (42 active, 32 placebo). ADAS-Cog improved 2.34 points in the active group versus 0.97 with placebo; between-group p=0.048. No serious adverse events or AE-related withdrawals were reported.
- Participants / model
- 100 randomized adults with amnestic MCI; 74 included in primary efficacy analysis
- Treatment
- Oral alpha-GPC 600 mg/day or placebo
- Follow-up
- Twelve weeks
- Study design
- Multicenter randomized double-blind placebo-controlled trial
- Funding
- Company-funded/product study
Twenty-six percent were absent from the reported efficacy analysis, the effect was small/borderline, and the study was short and product-funded.
Read the original sourceDo the Russian relative cohorts prove dementia prevention?
No. In the three-year report, all 30 treated first-degree relatives already had MCI or objective deficits, while 24 of 32 untreated relatives were cognitively normal at baseline. The five-year abstract compared 56 treated with 66 untreated relatives and reported improvement and conversion patterns, but remained nonrandomized and omits full baseline, attrition, funding, and safety data. Baseline imbalance and selection can create the apparent prevention effect.
| Cohort | Reported result | Why causality is weak |
|---|---|---|
| Three years: 30 treated, 32 untreated | Treated MMSE +1.6 and MoCA +2.6; controls declined | No randomization or placebo; baseline cognition strongly imbalanced |
| Five years: 56 treated, 66 untreated | Abstract reports 96.8% marked/moderate CGI-I improvement and control conversion to MCI | Baseline, attrition, funding and safety details are not reported in the abstract; overlap with the earlier program is unclear |
Russian full paper · 30 treated, 32 untreated
Prospective nonrandomized comparative study
Selezneva ND, Kolykhalov IV, Gavrilova SI. Comparative prospective multidisciplinary study of the effectiveness of choline alfoscerate in preventing progression of cognitive deficit in relatives of patients with Alzheimer's disease. Psychiatry. 2020;18(1):6-15. doi:10.30629/2618-6667-2020-18-1-6-15.
At three years, mean MMSE increased 1.6 points and MoCA 2.6 points in the treated group, while untreated means fell 1.2 and 1.1 points. CGI-I rated 17/30 as moderately or markedly improved, 12 minimally improved, and one unchanged.
- Participants / model
- 62 first-degree relatives of people with Alzheimer disease: 30 treated, all with MCI or objective deficits; 32 untreated, 24 cognitively normal and 8 with deficits at baseline
- Treatment
- Three three-month courses of oral Cereton/choline alfoscerate 1200 mg/day, approximately every 1.5 years
- Follow-up
- Three years
- Study design
- Prospective nonrandomized, unblinded treated-versus-untreated cohort with repeated psychometric testing
The groups were grossly imbalanced in baseline cognition, with no placebo or blinded allocation. Many outcomes were tested. UKU monitoring reported no adverse events, but that zero-event report comes from 30 treated participants.
Read the original sourceRussian 2026 abstract · 56 treated, 66 untreated
Prospective nonrandomized comparative study report
Selezneva ND, Roshchina IF, Gavrilova SI. Efficacy of choline alfoscerate treatment for subjective cognitive impairment in first-degree relatives of patients with Alzheimer's disease: Results of a comparative 5-year study. S.S. Korsakov Journal of Neurology and Psychiatry. 2026;126(4-2):92-99. doi:10.17116/jnevro202612604292. PMID 42054337.
The abstract reports improvement after each treatment course and says 96.8% of treated participants had marked or moderate CGI-I improvement. Untreated participants showed worsening from month 39 and reported SCI-to-MCI conversion of 16.7% by month 60.
- Participants / model
- 122 first-degree relatives with subjective cognitive impairment: 56 treated and 66 untreated
- Treatment
- Four three-month courses of oral choline alfoscerate 400 mg three times daily, spaced about 1.5 years apart
- Follow-up
- Five years
- Study design
- Prospective nonrandomized treated-versus-untreated comparison with nine visits and ten cognitive scales/tests
No randomization, blinding, or placebo is described in the abstract. Multiple repeated outcomes and an untreated comparator make causality uncertain; independence from the earlier cohort program is not established.
Read the original sourceDoes Alpha-GPC improve focus, strength, power or recovery in healthy people?
Healthy evidence is small and inconsistent. A sponsor-funded 20-man acute crossover improved Stroop change after 315 and 630 mg, while N-back, Flanker, mood ratings, physical performance and growth hormone were null. Exercise studies report isolated force, Wingate or biomarker signals in samples of 7 to 13, but a 48-person study found no Alpha-GPC advantage over placebo for isometric strength or psychomotor vigilance.
| Study | Positive result | Nulls or design limit |
|---|---|---|
| 2024 crossover: 20 trained men | Stroop change improved at 315 and 630 mg | N-back, Flanker, ratings, physical tests and GH null; retrospective registration, no pre-dose cognitive baseline, sponsor-controlled data |
| 2021 study: 39 healthy adults | Night-time self-rated motivation signal | Overall motivation only a trend; anxiety null; single-blind and retrospectively registered |
| 2008 conference crossover: 7 | Post-exercise GH and bench peak-force signal | Peak power, rate of force, metabolic rate, RER, heart rate and blood pressure null |
| 2015 study: 13 | Lower-body isometric-force change favored 600 mg/day | Upper-body force null; tiny product-funded sample |
| 2017 study: 48 | Plasma choline rose; 500 mg lowered TSH | Upper/lower isometric strength and vigilance null; jump result was 250 vs 500 mg, not placebo |
| 2022 crossover: 12 women with overweight/obesity | Wingate peak/mean power and selected recovery measures favored 1000 mg | Fatigue index and diastolic-pressure result null; one acute small population |
20-man crossover · Stroop selective, other tests null
Randomized double-blind crossover trial
PMID 39683633; PMCID PMC11644786.
Stroop total-score change was greater after 630 mg and 315 mg than placebo. Flanker, N-back, mood ratings, physical performance and growth hormone were null.
- Participants / model
- 20 resistance-trained healthy men
- Treatment
- Single 315 mg, 630 mg or placebo exposure
- Follow-up
- Three acute crossover periods
- Study design
- Randomized double-blind three-period crossover
- Funding
- NNB Nutrition
Retrospectively registered, no a priori sample calculation, no clean pre-dose cognitive baseline, sponsor-controlled data, paid-adviser author and unblinded CRO analysis.
Read the original source39-person study · night-time self-rating only
Randomized single-blind placebo-controlled trial
PMID 34207484; PMCID PMC8235064.
A night-time self-rated motivation comparison favored 400 mg/day. Overall motivation only trended and anxiety was null.
- Participants / model
- 39 healthy adults
- Treatment
- Oral Alpha-GPC 400 mg/day versus placebo
- Follow-up
- Two weeks
- Study design
- Randomized single-blind placebo-controlled trial
Small, repeatedly sampled bespoke self-rating; registration followed recruitment.
Read the original sourceSeven-man conference crossover · force and GH signals
Randomized crossover conference report
PMCID PMC3313098.
One 600 mg exposure increased post-exercise growth hormone and reported 14% higher bench-press peak force.
- Participants / model
- Seven healthy men
- Treatment
- Single oral Alpha-GPC 600 mg versus placebo
- Follow-up
- Acute crossover
- Study design
- Randomized crossover conference report
- Funding
- Chemi Nutra
Peak power, rate of force development, metabolic rate, respiratory exchange ratio, heart rate and blood pressure were null. Conference abstract only.
Read the original sourceEight-man crossover · biomarkers only
Randomized double-blind crossover study
PMID 22673596.
A 1000 mg exposure produced transient increases in growth hormone, free fatty acids, acetoacetate and 3-hydroxybutyrate.
- Participants / model
- Eight healthy men
- Treatment
- Single oral Alpha-GPC 1000 mg versus placebo
- Follow-up
- Acute
- Study design
- Randomized double-blind crossover
No strength, hypertrophy, fat-loss, recovery or clinical outcome was measured.
Read the original sourceThirteen-person study · lower-body signal, upper-body null
Randomized double-blind placebo-controlled trial
PMID 26582972; PMCID PMC4650143.
Lower-body isometric-force change was +98.8 N versus -39.0 N, reported p=0.044.
- Participants / model
- 13 healthy adults
- Treatment
- Oral Alpha-GPC 600 mg/day versus placebo
- Follow-up
- Six days
- Study design
- Randomized double-blind placebo-controlled trial
- Funding
- Chemi Nutra
Upper-body force was null; tiny product-funded study and reported interaction statistics are not fully transparent.
Read the original source48-man study · isometric strength null
Randomized placebo- and caffeine-controlled trial
PMID 29042830; PMCID PMC5629791.
Plasma choline rose; the 500 mg arm had lower TSH than all groups.
- Participants / model
- 48 healthy men
- Treatment
- Alpha-GPC 250 mg, 500 mg, caffeine or placebo
- Follow-up
- Seven days
- Study design
- Randomized controlled trial
- Funding
- Chemi Nutra supplied products
Upper- and lower-body isometric strength and psychomotor vigilance were null. The jump result was 250 mg versus 500 mg, not a demonstrated placebo effect.
Read the original sourceTwelve-woman crossover · selected Wingate/recovery signals
Randomized placebo-controlled crossover
PMCID PMC9572742.
Peak and mean Wingate power plus selected HRV and systolic-pressure recovery measures favored Alpha-GPC.
- Participants / model
- 12 women with overweight or obesity
- Treatment
- Single oral Alpha-GPC 1000 mg versus placebo
- Follow-up
- Acute crossover
- Study design
- Randomized placebo-controlled crossover
Fatigue index and diastolic-pressure comparison were null; tiny single-population acute experiment.
Read the original sourceDoes a choline rise prove brain delivery or benefit?
Alpha-GPC can raise plasma free choline, which may feed acetylcholine and phospholipid metabolism. That curve does not measure intact Alpha-GPC in brain tissue. A 12-man crossover found free-choline elevation after 1200 mg, and a six-man salt comparison found rapid TMAO increases after water-soluble choline forms. Neither defines a parent-compound half-life or proves cognitive benefit.
| Study | Finding | Boundary |
|---|---|---|
| 12 healthy men, crossover | Plasma free choline rose from 0.75 hours and peaked around 2 hours after 1200 mg | Free choline, not intact Alpha-GPC PK or brain residence |
| Korean tablet/capsule study | Mean choline Tmax 3.51 vs 3.85 hours | Bioequivalence marker, not parent half-life |
| Six-man equal-choline crossover | Water-soluble choline forms rapidly raised TMAO | Tiny acute study; mechanism and vascular consequence unresolved |
| Four-week supplement comparison | No significant between-product TMAO difference | Also too small to settle chronic vascular risk |
EFSA 2026 · narrow ingredient and intake boundary
Regulatory food-safety opinion
European Food Safety Authority. EFSA Journal. 2026. DOI 10.2903/j.efsa.2026.10008.
EFSA found no safety concern for the specified ingredient under a proposed maximum 203.7 mg/day use, corresponding to 82.5 mg choline/day.
- Participants / model
- Proposed adult food-supplement users
- Treatment
- Specified soy-phospholipid-derived Alpha-GPC ingredient
- Follow-up
- Regulatory assessment
- Study design
- EFSA novel-food safety opinion
The conclusion does not cover arbitrary products, indefinite higher trial doses or vascular outcomes.
Read the original sourceKorean crossover · plasma choline, not parent alpha-GPC
Human bioequivalence study
Min MH, Park JH, Hur JH, Shin HC, Cho Y, Kim DD. Formulation and bioequivalence studies of choline alfoscerate tablet comparing with soft gelatin capsule in healthy male volunteers. Drug Design, Development and Therapy. 2019;13:1049-1058. doi:10.2147/DDDT.S193424. PMID 31040642.
A randomized single-dose two-period crossover compared tablet and soft capsule in healthy Korean men. Plasma choline, not intact parent alpha-GPC, had mean Tmax 3.51 and 3.85 hours for the formulations and met bioequivalence criteria.
- Participants / model
- Healthy Korean men in a randomized crossover bioequivalence study
- Treatment
- Single oral 1200 mg choline alfoscerate tablet or soft capsule
- Follow-up
- Blood sampling through 12 hours after each period
- Study design
- Randomized single-dose two-period crossover
- Funding
- Whanin Pharmaceutical; multiple authors were employees
The assay measured endogenous/baseline-corrected plasma choline rather than intact parent alpha-GPC, so those Tmax values cannot populate parent-drug PK.
Read the original sourceSix-man crossover · water-soluble forms raised TMAO
Randomized crossover metabolism study
PMCID PMC8783899.
Equal-choline doses of water-soluble choline forms, including Alpha-GPC, rapidly increased plasma TMAO; egg phosphatidylcholine did not.
- Participants / model
- Six healthy men
- Treatment
- Equal-choline exposures from different choline forms
- Follow-up
- Acute crossover
- Study design
- Randomized crossover metabolism study
Tiny acute metabolism study; it does not establish a vascular event mechanism.
Read the original sourceSmall four-week study · no between-product TMAO difference
Randomized supplement comparison
DOI 10.2174/2665978602666210212115014.
No significant between-product TMAO difference was reported after four weeks.
- Participants / model
- Healthy adults
- Treatment
- Different choline supplements including Alpha-GPC
- Follow-up
- Four weeks
- Study design
- Randomized comparison
Small and short; it does not settle chronic vascular risk.
Read the original sourceFDA GSRS · alpha-GPC identity
Government substance database
U.S. Food and Drug Administration. Global Substance Registration System, choline alfoscerate, UNII 60M22SGW66.
The record identifies choline alfoscerate/alpha-GPC as C8H20NO6P, molecular weight 257.22 g/mol, CAS 28319-77-9.
- Participants / model
- Not applicable
- Treatment
- Not applicable
- Follow-up
- Not applicable
- Study design
- Primary source record
A UNII is an identity code, not FDA drug approval.
Read the original sourceWhat are the stroke association and safety limits?
A Korean claims study of 12,008,977 adults aged at least 50 linked Alpha-GPC prescriptions with total stroke, ischemic stroke and hemorrhagic stroke after matching. Indication, frailty, prodromal cognitive disease and unmeasured risk can confound that association, so it does not prove causation. Short randomized trials cannot dismiss or confirm a long-term vascular effect.
| Source | Result | Interpretation |
|---|---|---|
| Korean claims cohort | Total stroke aHR 1.43, 95% CI 1.41 to 1.46; ischemic 1.34; hemorrhagic 1.37 | Large observational association, not randomized causality |
| 2026 EFSA opinion | No safety concern for a specified soy-derived ingredient at 203.7 mg/day | Does not cover indefinite 600 to 1200 mg/day use or vascular outcomes |
| 2026 kidney-cancer cohort | aHR 0.95, 95% CI 0.84 to 1.08 | Endpoint-specific null; does not answer stroke or all-cancer safety |
| Alzheimer RCT | Treatment-related events more frequent, mainly constipation and nervousness | Controlled short-term signal, not chronic safety |
Korean cohort · 12,008,977 adults
Retrospective population cohort
Lee G, Choi S, Chang J, Choi D, Son JS, Kim K, Kim SM, Jeong S, Park SM. Association of L-alpha Glycerylphosphorylcholine With Subsequent Stroke Risk After 10 Years. JAMA Network Open. 2021;4(11):e2136008. doi:10.1001/jamanetworkopen.2021.36008. PMID 34817582.
In a South Korean claims cohort of 12,008,977 adults aged 50 or older, 108,877 alpha-GPC users had higher subsequent total stroke risk after matching (adjusted HR 1.43, 95% CI 1.41-1.46), with duration-response patterns.
- Participants / model
- 12,008,977 South Korean adults aged 50 or older; 108,877 exposed
- Treatment
- Prescription alpha-GPC exposure in claims data
- Follow-up
- Ten-year retrospective follow-up
- Study design
- Retrospective national cohort with propensity matching
The observational association cannot establish causality. Confounding by indication, disease severity, and unmeasured factors remains possible.
Read the original sourceEFSA 2026 · narrow ingredient and intake boundary
Regulatory food-safety opinion
European Food Safety Authority. EFSA Journal. 2026. DOI 10.2903/j.efsa.2026.10008.
EFSA found no safety concern for the specified ingredient under a proposed maximum 203.7 mg/day use, corresponding to 82.5 mg choline/day.
- Participants / model
- Proposed adult food-supplement users
- Treatment
- Specified soy-phospholipid-derived Alpha-GPC ingredient
- Follow-up
- Regulatory assessment
- Study design
- EFSA novel-food safety opinion
The conclusion does not cover arbitrary products, indefinite higher trial doses or vascular outcomes.
Read the original source2026 Korean cohort · kidney cancer null
Nationwide retrospective cohort
PMID 42361543.
Among 81,970 users and 409,801 matched non-users, kidney-cancer incidence was not associated with Alpha-GPC, adjusted HR 0.95, 95% CI 0.84 to 1.08.
- Participants / model
- Korean prescription users and matched non-users
- Treatment
- Alpha-GPC prescription exposure
- Follow-up
- Longitudinal claims follow-up
- Study design
- Retrospective matched cohort
Endpoint-specific observational null does not adjudicate stroke, other cancers or all-cause safety.
Read the original sourceAlzheimer full paper · 261 randomized, 229 completed
Randomized double-blind placebo-controlled trial
De Jesus 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. Clinical Therapeutics. 2003;25(1):178-193. doi:10.1016/S0149-2918(03)90023-3. PMID 12637119.
Mean ADAS-Cog changed from 35.52 to 32.32 with choline alfoscerate and from 36.74 to 39.64 with placebo at day 180, equivalent to -3.20 versus +2.90 points. Several secondary scales also favored treatment.
- Participants / model
- 261 adults aged 60-80 with mild-to-moderate probable Alzheimer's dementia: 132 active and 129 placebo; 229 completed
- Treatment
- Oral choline alfoscerate 400 mg three times daily or placebo
- Follow-up
- 180 days
- Study design
- Multicenter randomized double-blind placebo-controlled trial; prespecified primary analysis used a one-tailed alpha of 0.05
- Funding
- Italfarmaco supported study design, conduct, analysis, and publication
Treatment-related adverse events occurred in 11/132 active participants (10 constipation, 5 nervousness events) and 3/129 placebo participants; none withdrew for these events. The trial was old, sponsor-run, used a one-tailed test, and assessed symptomatic scales rather than disease modification.
Read the original sourceSix-man crossover · water-soluble forms raised TMAO
Randomized crossover metabolism study
PMCID PMC8783899.
Equal-choline doses of water-soluble choline forms, including Alpha-GPC, rapidly increased plasma TMAO; egg phosphatidylcholine did not.
- Participants / model
- Six healthy men
- Treatment
- Equal-choline exposures from different choline forms
- Follow-up
- Acute crossover
- Study design
- Randomized crossover metabolism study
Tiny acute metabolism study; it does not establish a vascular event mechanism.
Read the original sourceWhat does B tier apply to?
B tier applies to short-term symptomatic cognition evidence in Alzheimer disease and amnestic MCI. It does not certify healthy focus, sports performance, growth-hormone benefit, prevention or longevity. The healthy studies are small and selective; Russian prevention reports are nonrandomized; and the observational stroke signal plus long-term higher-dose safety remain unresolved.
| Claim | Judgment |
|---|---|
| Symptomatic Alzheimer/MCI cognition | Substantive but short-term, sponsor-linked evidence |
| Healthy focus or memory | One selective acute Stroop signal amid null tests; low confidence |
| Strength, power or recovery | Small inconsistent acute/short-term studies; no training-adaptation evidence |
| Dementia prevention and longevity | No randomized healthy prevention, mortality or lifespan evidence |
| Long-term safety | Unresolved; observational stroke signal requires causal follow-up |
Alzheimer full paper · 261 randomized, 229 completed
Randomized double-blind placebo-controlled trial
De Jesus 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. Clinical Therapeutics. 2003;25(1):178-193. doi:10.1016/S0149-2918(03)90023-3. PMID 12637119.
Mean ADAS-Cog changed from 35.52 to 32.32 with choline alfoscerate and from 36.74 to 39.64 with placebo at day 180, equivalent to -3.20 versus +2.90 points. Several secondary scales also favored treatment.
- Participants / model
- 261 adults aged 60-80 with mild-to-moderate probable Alzheimer's dementia: 132 active and 129 placebo; 229 completed
- Treatment
- Oral choline alfoscerate 400 mg three times daily or placebo
- Follow-up
- 180 days
- Study design
- Multicenter randomized double-blind placebo-controlled trial; prespecified primary analysis used a one-tailed alpha of 0.05
- Funding
- Italfarmaco supported study design, conduct, analysis, and publication
Treatment-related adverse events occurred in 11/132 active participants (10 constipation, 5 nervousness events) and 3/129 placebo participants; none withdrew for these events. The trial was old, sponsor-run, used a one-tailed test, and assessed symptomatic scales rather than disease modification.
Read the original sourceMCI trial · 100 randomized, 74 efficacy analysis
Randomized controlled trial
Jeon J, Lee SY, Lee S, Han C, Park GD, Kim SJ, Chang JG, Kim WJ. Efficacy and safety of choline alphoscerate for amnestic mild cognitive impairment: a randomized double-blind placebo-controlled trial. BMC Geriatrics. 2024;24:774. doi:10.1186/s12877-024-05366-7. PMID 39300341.
One hundred adults with amnestic MCI were randomized for 12 weeks, but the reported efficacy analysis included 74 (42 active, 32 placebo). ADAS-Cog improved 2.34 points in the active group versus 0.97 with placebo; between-group p=0.048. No serious adverse events or AE-related withdrawals were reported.
- Participants / model
- 100 randomized adults with amnestic MCI; 74 included in primary efficacy analysis
- Treatment
- Oral alpha-GPC 600 mg/day or placebo
- Follow-up
- Twelve weeks
- Study design
- Multicenter randomized double-blind placebo-controlled trial
- Funding
- Company-funded/product study
Twenty-six percent were absent from the reported efficacy analysis, the effect was small/borderline, and the study was short and product-funded.
Read the original source20-man crossover · Stroop selective, other tests null
Randomized double-blind crossover trial
PMID 39683633; PMCID PMC11644786.
Stroop total-score change was greater after 630 mg and 315 mg than placebo. Flanker, N-back, mood ratings, physical performance and growth hormone were null.
- Participants / model
- 20 resistance-trained healthy men
- Treatment
- Single 315 mg, 630 mg or placebo exposure
- Follow-up
- Three acute crossover periods
- Study design
- Randomized double-blind three-period crossover
- Funding
- NNB Nutrition
Retrospectively registered, no a priori sample calculation, no clean pre-dose cognitive baseline, sponsor-controlled data, paid-adviser author and unblinded CRO analysis.
Read the original sourceKorean cohort · 12,008,977 adults
Retrospective population cohort
Lee G, Choi S, Chang J, Choi D, Son JS, Kim K, Kim SM, Jeong S, Park SM. Association of L-alpha Glycerylphosphorylcholine With Subsequent Stroke Risk After 10 Years. JAMA Network Open. 2021;4(11):e2136008. doi:10.1001/jamanetworkopen.2021.36008. PMID 34817582.
In a South Korean claims cohort of 12,008,977 adults aged 50 or older, 108,877 alpha-GPC users had higher subsequent total stroke risk after matching (adjusted HR 1.43, 95% CI 1.41-1.46), with duration-response patterns.
- Participants / model
- 12,008,977 South Korean adults aged 50 or older; 108,877 exposed
- Treatment
- Prescription alpha-GPC exposure in claims data
- Follow-up
- Ten-year retrospective follow-up
- Study design
- Retrospective national cohort with propensity matching
The observational association cannot establish causality. Confounding by indication, disease severity, and unmeasured factors remains possible.
Read the original sourceEFSA 2026 · narrow ingredient and intake boundary
Regulatory food-safety opinion
European Food Safety Authority. EFSA Journal. 2026. DOI 10.2903/j.efsa.2026.10008.
EFSA found no safety concern for the specified ingredient under a proposed maximum 203.7 mg/day use, corresponding to 82.5 mg choline/day.
- Participants / model
- Proposed adult food-supplement users
- Treatment
- Specified soy-phospholipid-derived Alpha-GPC ingredient
- Follow-up
- Regulatory assessment
- Study design
- EFSA novel-food safety opinion
The conclusion does not cover arbitrary products, indefinite higher trial doses or vascular outcomes.
Read the original sourceStudies and sources
FDA GSRS · alpha-GPC identity
Government substance database
U.S. Food and Drug Administration. Global Substance Registration System, choline alfoscerate, UNII 60M22SGW66.
The record identifies choline alfoscerate/alpha-GPC as C8H20NO6P, molecular weight 257.22 g/mol, CAS 28319-77-9.
- Participants / model
- Not applicable
- Treatment
- Not applicable
- Follow-up
- Not applicable
- Study design
- Primary source record
A UNII is an identity code, not FDA drug approval.
Read the original sourceEMA list · Italy and Poland national products
EU regulatory product list
European Medicines Agency. List of nationally authorised medicinal products: active substance choline alfoscerate. EMA/174242/2023, PSUSA/00010599/202208. April 14, 2023.
The EMA list identifies nationally authorized choline alfoscerate products in Italy and Poland, including oral and injectable forms.
- Participants / model
- Patients covered by national product labels in Italy and Poland
- Treatment
- Choline alfoscerate medicinal products
- Follow-up
- Current list at the publication date
- Study design
- EU pharmacovigilance/national authorization list
This is not a centralized EMA approval and does not make every supplement or indication equivalent; local product labels govern.
Read the original sourceRussian registered products · oral and injectable forms
Russian government medicines/price register
Ministry of Health of the Russian Federation. State Register of Medicines / public price-limit records for choline alfoscerate and Cereton.
The official register surface lists current choline alfoscerate and Cereton registration records and oral or intravenous/intramuscular formulations in Russia.
- Participants / model
- Patients under Russian product-specific labels
- Treatment
- Registered choline alfoscerate products
- Follow-up
- Living register
- Study design
- Government medicine-register record
The register proves product registration, not a single universal indication or interchangeability across oral/injectable brands. Product-specific Russian instructions govern use.
Read the original sourceAlzheimer full paper · 261 randomized, 229 completed
Randomized double-blind placebo-controlled trial
De Jesus 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. Clinical Therapeutics. 2003;25(1):178-193. doi:10.1016/S0149-2918(03)90023-3. PMID 12637119.
Mean ADAS-Cog changed from 35.52 to 32.32 with choline alfoscerate and from 36.74 to 39.64 with placebo at day 180, equivalent to -3.20 versus +2.90 points. Several secondary scales also favored treatment.
- Participants / model
- 261 adults aged 60-80 with mild-to-moderate probable Alzheimer's dementia: 132 active and 129 placebo; 229 completed
- Treatment
- Oral choline alfoscerate 400 mg three times daily or placebo
- Follow-up
- 180 days
- Study design
- Multicenter randomized double-blind placebo-controlled trial; prespecified primary analysis used a one-tailed alpha of 0.05
- Funding
- Italfarmaco supported study design, conduct, analysis, and publication
Treatment-related adverse events occurred in 11/132 active participants (10 constipation, 5 nervousness events) and 3/129 placebo participants; none withdrew for these events. The trial was old, sponsor-run, used a one-tailed test, and assessed symptomatic scales rather than disease modification.
Read the original sourceMCI trial · 100 randomized, 74 efficacy analysis
Randomized controlled trial
Jeon J, Lee SY, Lee S, Han C, Park GD, Kim SJ, Chang JG, Kim WJ. Efficacy and safety of choline alphoscerate for amnestic mild cognitive impairment: a randomized double-blind placebo-controlled trial. BMC Geriatrics. 2024;24:774. doi:10.1186/s12877-024-05366-7. PMID 39300341.
One hundred adults with amnestic MCI were randomized for 12 weeks, but the reported efficacy analysis included 74 (42 active, 32 placebo). ADAS-Cog improved 2.34 points in the active group versus 0.97 with placebo; between-group p=0.048. No serious adverse events or AE-related withdrawals were reported.
- Participants / model
- 100 randomized adults with amnestic MCI; 74 included in primary efficacy analysis
- Treatment
- Oral alpha-GPC 600 mg/day or placebo
- Follow-up
- Twelve weeks
- Study design
- Multicenter randomized double-blind placebo-controlled trial
- Funding
- Company-funded/product study
Twenty-six percent were absent from the reported efficacy analysis, the effect was small/borderline, and the study was short and product-funded.
Read the original sourceRussian full paper · 30 treated, 32 untreated
Prospective nonrandomized comparative study
Selezneva ND, Kolykhalov IV, Gavrilova SI. Comparative prospective multidisciplinary study of the effectiveness of choline alfoscerate in preventing progression of cognitive deficit in relatives of patients with Alzheimer's disease. Psychiatry. 2020;18(1):6-15. doi:10.30629/2618-6667-2020-18-1-6-15.
At three years, mean MMSE increased 1.6 points and MoCA 2.6 points in the treated group, while untreated means fell 1.2 and 1.1 points. CGI-I rated 17/30 as moderately or markedly improved, 12 minimally improved, and one unchanged.
- Participants / model
- 62 first-degree relatives of people with Alzheimer disease: 30 treated, all with MCI or objective deficits; 32 untreated, 24 cognitively normal and 8 with deficits at baseline
- Treatment
- Three three-month courses of oral Cereton/choline alfoscerate 1200 mg/day, approximately every 1.5 years
- Follow-up
- Three years
- Study design
- Prospective nonrandomized, unblinded treated-versus-untreated cohort with repeated psychometric testing
The groups were grossly imbalanced in baseline cognition, with no placebo or blinded allocation. Many outcomes were tested. UKU monitoring reported no adverse events, but that zero-event report comes from 30 treated participants.
Read the original sourceKorean cohort · 12,008,977 adults
Retrospective population cohort
Lee G, Choi S, Chang J, Choi D, Son JS, Kim K, Kim SM, Jeong S, Park SM. Association of L-alpha Glycerylphosphorylcholine With Subsequent Stroke Risk After 10 Years. JAMA Network Open. 2021;4(11):e2136008. doi:10.1001/jamanetworkopen.2021.36008. PMID 34817582.
In a South Korean claims cohort of 12,008,977 adults aged 50 or older, 108,877 alpha-GPC users had higher subsequent total stroke risk after matching (adjusted HR 1.43, 95% CI 1.41-1.46), with duration-response patterns.
- Participants / model
- 12,008,977 South Korean adults aged 50 or older; 108,877 exposed
- Treatment
- Prescription alpha-GPC exposure in claims data
- Follow-up
- Ten-year retrospective follow-up
- Study design
- Retrospective national cohort with propensity matching
The observational association cannot establish causality. Confounding by indication, disease severity, and unmeasured factors remains possible.
Read the original sourceKorean crossover · plasma choline, not parent alpha-GPC
Human bioequivalence study
Min MH, Park JH, Hur JH, Shin HC, Cho Y, Kim DD. Formulation and bioequivalence studies of choline alfoscerate tablet comparing with soft gelatin capsule in healthy male volunteers. Drug Design, Development and Therapy. 2019;13:1049-1058. doi:10.2147/DDDT.S193424. PMID 31040642.
A randomized single-dose two-period crossover compared tablet and soft capsule in healthy Korean men. Plasma choline, not intact parent alpha-GPC, had mean Tmax 3.51 and 3.85 hours for the formulations and met bioequivalence criteria.
- Participants / model
- Healthy Korean men in a randomized crossover bioequivalence study
- Treatment
- Single oral 1200 mg choline alfoscerate tablet or soft capsule
- Follow-up
- Blood sampling through 12 hours after each period
- Study design
- Randomized single-dose two-period crossover
- Funding
- Whanin Pharmaceutical; multiple authors were employees
The assay measured endogenous/baseline-corrected plasma choline rather than intact parent alpha-GPC, so those Tmax values cannot populate parent-drug PK.
Read the original sourceRussian 2026 abstract · 56 treated, 66 untreated
Prospective nonrandomized comparative study report
Selezneva ND, Roshchina IF, Gavrilova SI. Efficacy of choline alfoscerate treatment for subjective cognitive impairment in first-degree relatives of patients with Alzheimer's disease: Results of a comparative 5-year study. S.S. Korsakov Journal of Neurology and Psychiatry. 2026;126(4-2):92-99. doi:10.17116/jnevro202612604292. PMID 42054337.
The abstract reports improvement after each treatment course and says 96.8% of treated participants had marked or moderate CGI-I improvement. Untreated participants showed worsening from month 39 and reported SCI-to-MCI conversion of 16.7% by month 60.
- Participants / model
- 122 first-degree relatives with subjective cognitive impairment: 56 treated and 66 untreated
- Treatment
- Four three-month courses of oral choline alfoscerate 400 mg three times daily, spaced about 1.5 years apart
- Follow-up
- Five years
- Study design
- Prospective nonrandomized treated-versus-untreated comparison with nine visits and ten cognitive scales/tests
No randomization, blinding, or placebo is described in the abstract. Multiple repeated outcomes and an untreated comparator make causality uncertain; independence from the earlier cohort program is not established.
Read the original source20-man crossover · Stroop selective, other tests null
Randomized double-blind crossover trial
PMID 39683633; PMCID PMC11644786.
Stroop total-score change was greater after 630 mg and 315 mg than placebo. Flanker, N-back, mood ratings, physical performance and growth hormone were null.
- Participants / model
- 20 resistance-trained healthy men
- Treatment
- Single 315 mg, 630 mg or placebo exposure
- Follow-up
- Three acute crossover periods
- Study design
- Randomized double-blind three-period crossover
- Funding
- NNB Nutrition
Retrospectively registered, no a priori sample calculation, no clean pre-dose cognitive baseline, sponsor-controlled data, paid-adviser author and unblinded CRO analysis.
Read the original source39-person study · night-time self-rating only
Randomized single-blind placebo-controlled trial
PMID 34207484; PMCID PMC8235064.
A night-time self-rated motivation comparison favored 400 mg/day. Overall motivation only trended and anxiety was null.
- Participants / model
- 39 healthy adults
- Treatment
- Oral Alpha-GPC 400 mg/day versus placebo
- Follow-up
- Two weeks
- Study design
- Randomized single-blind placebo-controlled trial
Small, repeatedly sampled bespoke self-rating; registration followed recruitment.
Read the original sourceSeven-man conference crossover · force and GH signals
Randomized crossover conference report
PMCID PMC3313098.
One 600 mg exposure increased post-exercise growth hormone and reported 14% higher bench-press peak force.
- Participants / model
- Seven healthy men
- Treatment
- Single oral Alpha-GPC 600 mg versus placebo
- Follow-up
- Acute crossover
- Study design
- Randomized crossover conference report
- Funding
- Chemi Nutra
Peak power, rate of force development, metabolic rate, respiratory exchange ratio, heart rate and blood pressure were null. Conference abstract only.
Read the original sourceEight-man crossover · biomarkers only
Randomized double-blind crossover study
PMID 22673596.
A 1000 mg exposure produced transient increases in growth hormone, free fatty acids, acetoacetate and 3-hydroxybutyrate.
- Participants / model
- Eight healthy men
- Treatment
- Single oral Alpha-GPC 1000 mg versus placebo
- Follow-up
- Acute
- Study design
- Randomized double-blind crossover
No strength, hypertrophy, fat-loss, recovery or clinical outcome was measured.
Read the original sourceThirteen-person study · lower-body signal, upper-body null
Randomized double-blind placebo-controlled trial
PMID 26582972; PMCID PMC4650143.
Lower-body isometric-force change was +98.8 N versus -39.0 N, reported p=0.044.
- Participants / model
- 13 healthy adults
- Treatment
- Oral Alpha-GPC 600 mg/day versus placebo
- Follow-up
- Six days
- Study design
- Randomized double-blind placebo-controlled trial
- Funding
- Chemi Nutra
Upper-body force was null; tiny product-funded study and reported interaction statistics are not fully transparent.
Read the original source48-man study · isometric strength null
Randomized placebo- and caffeine-controlled trial
PMID 29042830; PMCID PMC5629791.
Plasma choline rose; the 500 mg arm had lower TSH than all groups.
- Participants / model
- 48 healthy men
- Treatment
- Alpha-GPC 250 mg, 500 mg, caffeine or placebo
- Follow-up
- Seven days
- Study design
- Randomized controlled trial
- Funding
- Chemi Nutra supplied products
Upper- and lower-body isometric strength and psychomotor vigilance were null. The jump result was 250 mg versus 500 mg, not a demonstrated placebo effect.
Read the original sourceTwelve-woman crossover · selected Wingate/recovery signals
Randomized placebo-controlled crossover
PMCID PMC9572742.
Peak and mean Wingate power plus selected HRV and systolic-pressure recovery measures favored Alpha-GPC.
- Participants / model
- 12 women with overweight or obesity
- Treatment
- Single oral Alpha-GPC 1000 mg versus placebo
- Follow-up
- Acute crossover
- Study design
- Randomized placebo-controlled crossover
Fatigue index and diastolic-pressure comparison were null; tiny single-population acute experiment.
Read the original sourceEFSA 2026 · narrow ingredient and intake boundary
Regulatory food-safety opinion
European Food Safety Authority. EFSA Journal. 2026. DOI 10.2903/j.efsa.2026.10008.
EFSA found no safety concern for the specified ingredient under a proposed maximum 203.7 mg/day use, corresponding to 82.5 mg choline/day.
- Participants / model
- Proposed adult food-supplement users
- Treatment
- Specified soy-phospholipid-derived Alpha-GPC ingredient
- Follow-up
- Regulatory assessment
- Study design
- EFSA novel-food safety opinion
The conclusion does not cover arbitrary products, indefinite higher trial doses or vascular outcomes.
Read the original sourceSix-man crossover · water-soluble forms raised TMAO
Randomized crossover metabolism study
PMCID PMC8783899.
Equal-choline doses of water-soluble choline forms, including Alpha-GPC, rapidly increased plasma TMAO; egg phosphatidylcholine did not.
- Participants / model
- Six healthy men
- Treatment
- Equal-choline exposures from different choline forms
- Follow-up
- Acute crossover
- Study design
- Randomized crossover metabolism study
Tiny acute metabolism study; it does not establish a vascular event mechanism.
Read the original sourceSmall four-week study · no between-product TMAO difference
Randomized supplement comparison
DOI 10.2174/2665978602666210212115014.
No significant between-product TMAO difference was reported after four weeks.
- Participants / model
- Healthy adults
- Treatment
- Different choline supplements including Alpha-GPC
- Follow-up
- Four weeks
- Study design
- Randomized comparison
Small and short; it does not settle chronic vascular risk.
Read the original source2026 Korean cohort · kidney cancer null
Nationwide retrospective cohort
PMID 42361543.
Among 81,970 users and 409,801 matched non-users, kidney-cancer incidence was not associated with Alpha-GPC, adjusted HR 0.95, 95% CI 0.84 to 1.08.
- Participants / model
- Korean prescription users and matched non-users
- Treatment
- Alpha-GPC prescription exposure
- Follow-up
- Longitudinal claims follow-up
- Study design
- Retrospective matched cohort
Endpoint-specific observational null does not adjudicate stroke, other cancers or all-cause safety.
Read the original sourceWhy is Alpha-GPC in B tier?
Tier B applies to controlled short-term improvements in cognitive symptom scores in Alzheimer disease and amnestic MCI. Healthy cognition has one selective acute Stroop signal amid null tests. Exercise studies are small and inconsistent, and no prevention or longevity outcome exists. Short-term tolerability is documented, while the observational stroke signal and long-term higher-dose safety remain unresolved.