Alpha-GPC vs Citicoline: Which Choline Source Has the Better Human Evidence?

What the two compounds actually are

Both are choline-carrying molecules marketed as ways to raise acetylcholine, the neurotransmitter most closely tied to attention and memory formation. They differ in what they carry alongside the choline.

Alpha-GPC (L-alpha-glycerylphosphorylcholine, also called choline alphoscerate) is choline attached to a glycerophosphate backbone. In several European countries it is a prescription medicine rather than a supplement, which is why so much of its trial evidence sits in clinical populations.

Citicoline (cytidine 5'-diphosphocholine, or CDP-choline) is choline joined to cytidine. Once absorbed it separates into choline and cytidine, and the cytidine becomes uridine — a nucleoside used in building cell membranes. That second component is citicoline's genuine point of difference, and it is why the two products are not interchangeable even at matched choline doses.

The choline arithmetic, done in public

Every brand blog repeats the figures “40 percent” and “18 percent” without showing where they come from. They come from molecular weights, and you can check them in one line. Choline weighs about 104 g/mol. Alpha-GPC weighs about 257 g/mol, so 104 ÷ 257 is close to 40 percent. Citicoline weighs about 488 g/mol as the free acid, so 104 ÷ 488 is closer to 21 percent, and a little lower for the sodium salt used in most products. The widely quoted 18 percent is in the right neighbourhood but conservative; the honest range is 18 to 21 percent.

The practical consequence: 600 mg of alpha-GPC supplies roughly 240 mg of choline, and matching that from citicoline would take near 1,200 mg. What none of this establishes is that more choline means more benefit. No trial has compared matched choline doses from the two sources, so the arithmetic is a fact about chemistry, not a finding about cognition.

The head-to-head evidence table

This comparison is assembled from the actual studies rather than marketing copy. Read the final row first, because it governs everything above it.

Compound	Choline by weight	Studied human dose	Best human trial (n, duration, population)	Outcome	Evidence strength
Alpha-GPC (choline alphoscerate)	~40%	600 mg/day in the MCI trial; ~1,200 mg/day across pooled clinical studies; 315–630 mg as a single acute dose	Jeon et al. 2024: 100 adults with amnestic mild cognitive impairment, 600 mg/day, 12 weeks, multicentre randomised double-blind placebo-controlled	ADAS-cog score improved by 2.34 points more than placebo; well tolerated, no serious adverse events	Moderate, but in a clinically impaired population — not healthy adults
Alpha-GPC — acute, healthy adults	~40%	315 mg or 630 mg, single dose	Kerksick 2024 (Nutrients): 20 resistance-trained men, mean age 31, randomised double-blind crossover	Stroop total score improved versus placebo at both doses; no difference on the Flanker or N-Back tests	Weak — very small, one positive test out of three, acute only
Citicoline (CDP-choline)	~18–21%	500 mg/day	Nakazaki et al. 2021 (J Nutr): 100 healthy adults aged 50–85 with age-associated memory impairment, 500 mg/day, 12 weeks, randomised double-blind placebo-controlled	Improved composite memory and episodic memory versus placebo; safety measures unremarkable	Moderate, and the closest thing to a healthy-adult trial either compound has — but industry funded
Alpha-GPC — pooled review	—	~1,200 mg/day, sometimes with donepezil	Sagaro, Traini & Amenta 2023 (J Alzheimers Dis): 7 RCTs plus 1 cohort study, adults with cognitive impairment linked to cerebrovascular injury	Improvements in cognitive, behavioural and functional scores versus comparator	Moderate for that clinical population; explicitly not healthy adults
Head-to-head: no direct randomised comparison of alpha-GPC and citicoline exists. Every table like this one, including this one, is built from separate trials in different populations at different doses using different cognitive tests. That is the single most important row here, and it is missing from every comparison article we measured.

Two caveats before anyone quotes the table. The Nakazaki citicoline trial was funded by a citicoline manufacturer and several authors were affiliated with it — that does not invalidate the result, but it belongs in the record. And the two alpha-GPC entries measure different things: a 12-week clinical endpoint in impaired adults, and a single-dose colour-word test in twenty gym-goers. Presenting them as one body of evidence overstates both.

Two different cholinergic strategies: supply versus preservation

Here is the part the comparison articles never reach, and it reframes the question. Alpha-GPC and citicoline are both precursor strategies: supply more raw material and hope the brain builds more acetylcholine. That is one of two ways to influence a cholinergic system, and the more indirect one.

The other strategy is preservation. Acetylcholine released into a synapse is broken down within milliseconds by an enzyme called acetylcholinesterase. An acetylcholinesterase inhibitor slows that breakdown, so the acetylcholine already present stays available for longer. Huperzine A, the alkaloid found in clubmoss (Huperzia serrata), works this way as a reversible inhibitor of that enzyme, which is why it is dosed in micrograms while choline precursors are dosed in hundreds of milligrams — the microgram dose figures are set out here. The two approaches do not operate at the same scale or on the same step.

Neither strategy is automatically superior; both are mechanistic arguments with modest human evidence attached. But the distinction matters when reading a label, because a formula built around preservation will not contain a choline precursor and is not supposed to. We cover what the preservation route does and does not demonstrate in our review of the huperzine A, ginseng and caffeine evidence, and this site's own formula — four named actives with the amounts printed on the panel — is built on that side of the split.

What the alpha-GPC trials actually found

The strongest alpha-GPC result is the 2024 multicentre randomised trial by Jeon and colleagues in BMC Geriatrics. One hundred participants with amnestic mild cognitive impairment took 600 mg a day of choline alphoscerate or placebo for 12 weeks, and the primary endpoint — change on the Alzheimer's Disease Assessment Scale cognitive subscale — favoured alpha-GPC by 2.34 points, with no serious adverse events. It is a genuine positive result, and it needs reading for what it is: a study in people with a diagnosed impairment, a different population from a healthy adult buying a supplement for everyday focus.

The 2023 systematic review by Sagaro, Traini and Amenta pooled seven randomised trials and one cohort study, at roughly 1,200 mg a day, in adults with cognitive impairment linked to cerebrovascular injury, reporting improvements in cognition, behaviour and function. The authors were explicit that healthy adults were not part of the evidence base.

The only alpha-GPC data in healthy people is thin. The 2024 crossover study by Kerksick in Nutrients gave 20 resistance-trained men single doses of 315 mg or 630 mg and found improved Stroop scores but no difference on the Flanker or N-Back tests. One positive result out of three measures in twenty people is a hypothesis, not a finding.

What the citicoline trials actually found

Citicoline's most useful trial here is Nakazaki and colleagues in the Journal of Nutrition in 2021. One hundred adults aged 50 to 85 with age-associated memory impairment took 500 mg a day of citicoline or placebo for 12 weeks, with computerised memory testing at baseline and week 12. Ninety-nine completed. The citicoline group improved more on episodic memory and on a composite memory score, and safety measures were unremarkable. This is the closest thing either compound has to a controlled trial in non-clinical older adults. The funding source — a citicoline manufacturer — is a real limitation.

Citicoline also has a substantial literature in stroke recovery and vascular cognitive impairment, and that work is the source of most of the confident language you will see online. It should not be carried across to a healthy reader: a trial of an intervention after a stroke tells you nothing about that intervention in someone who has not had one.

Safety, TMAO and the alpha-GPC stroke cohort

Both compounds were well tolerated in the trials above, but two safety topics deserve space. The first is TMAO. Gut bacteria convert dietary choline into trimethylamine, which the liver converts to trimethylamine N-oxide, a metabolite associated in observational research with cardiovascular risk. That concern attaches to high choline intakes generally, and applies more to alpha-GPC than citicoline simply because alpha-GPC carries more choline per milligram.

The second is a specific finding. In 2021, Lee and colleagues published a population-based cohort study in JAMA Network Open using South Korean national health insurance records covering more than 12 million people aged 50 and over. Individuals prescribed alpha-GPC had a higher rate of subsequent stroke over ten years than non-users, with an adjusted hazard ratio of 1.43 (95% CI 1.41–1.46), and longer prescription duration was associated with higher risk. This must be read carefully. It is observational, so it cannot establish cause. The authors themselves noted that alpha-GPC users were older and had more comorbidities, which may reflect underlying vascular disease already present before the prescription. It is a signal that warrants attention and further study, not a proven harm — and if you are considering alpha-GPC, it is a reason to talk to a clinician rather than to decide alone.

On total intake: the tolerable upper intake level for choline in adults is 3.5 g a day, set primarily on low blood pressure at high intakes, with fishy body odour, sweating and gastrointestinal effects as supporting evidence. That odour comes from trimethylamine excretion. Typical supplement doses sit well below the limit, but stacking several choline products is a way to approach it without noticing.

What the evidence does not show

No head-to-head trial exists. Every ranking of these two compounds — including this article's title — is inference across incomparable studies. Anyone telling you one is better is expressing a preference.

Almost none of the evidence is in healthy young adults. The good alpha-GPC trial was in mild cognitive impairment; the good citicoline trial in adults aged 50 to 85 with memory complaints; the pooled alpha-GPC review in cerebrovascular patients. The one healthy-adult study was twenty men and one positive measure out of three. If you are 35 with brain fog, none of this was studied on you.

Neither compound has a large outcome trial. No long-duration study in thousands of people shows a durable cognitive benefit for either. The evidence base is short trials with modest sample sizes, which is normal for supplements and exactly why claims should stay modest.

Most people are not choline deficient. Eggs, liver, fish, soybeans and cruciferous vegetables all supply choline, and one large egg provides a meaningful share of the adult adequate intake. Before adding a precursor, ask whether the diet already covers it. Related evidence on membrane phospholipids appears in our review of phosphatidylserine and memory, which shows the same pattern: plausible mechanism, small trials, strongest signal in older adults.

Frequently asked questions