Most products on the focus shelf that say "choline" contain choline bitartrate, which is inexpensive and which has not performed well in healthy adults. The two forms with a real research record behind them are alpha-GPC and citicoline, and they are genuinely different molecules with genuinely different evidence. Nobody has run them head to head. This is what is known about each separately.
What each one actually is
Alpha-GPC is L-alpha-glycerylphosphorylcholine: a choline group attached to glycerophosphate. It occurs naturally in the body and in food, concentrated in milk and organ meats. Around 40% of its molecular weight is choline, which is high — 300 mg of alpha-GPC carries roughly 120 mg of choline.
Citicoline is cytidine-5'-diphosphocholine, usually written CDP-choline. It is a larger molecule: a choline group joined to cytidine through two phosphates. Only about a fifth of its weight is choline, so 500 mg of citicoline carries roughly 100 mg. The rest is not filler — the cytidine portion is metabolised to uridine in humans, and the argument for citicoline is partly that this second component does something of its own. That argument is mechanistically reasonable and, as the next section shows, has not been settled in people.
What has been measured in healthy adults
This is where the two diverge, and where most comparisons go wrong by quoting whichever literature suits them.
Alpha-GPC's healthy-adult record is short-term and mostly performance-flavoured. A 2024 crossover trial in Nutrients gave 315 mg and 630 mg to 20 resistance-trained men of mean age 31 and found better Stroop scores against placebo at both amounts, with no difference on the Flanker or N-Back tasks. A 2017 trial in the Journal of the International Society of Sports Nutrition gave 250 mg or 500 mg daily for seven days to 48 healthy college-aged men and found no difference in isometric strength or psychomotor vigilance, with an improvement in countermovement-jump peak power. A 2023 crossover in the same journal used 300 mg for seven days in 30 trained cyclists of mean age 43, but gave it alongside two other actives, so alpha-GPC cannot be separated out. Small samples, short durations, mostly men, mostly athletic.
Citicoline has one substantially larger and longer healthy-adult trial. Nakazaki and colleagues, Journal of Nutrition, 2021: 100 healthy men and women aged 50 to 85 with age-associated memory impairment, randomised to 500 mg a day of a branded citicoline or placebo for 12 weeks. Episodic memory improved more in the citicoline group than in the placebo group (p = 0.0025), as did a composite memory score (p = 0.0052). Both were secondary outcomes rather than the primary one, which matters — but 100 participants over 12 weeks is a better-powered study than anything in the alpha-GPC healthy-adult literature.
The counterweight is worth stating just as plainly. In 2024 the European Food Safety Authority assessed a proposed memory claim for citicoline. The panel identified three human intervention studies in healthy people, noted that only one — 500 mg a day for 12 weeks — showed a benefit, that the effect was not seen at 1 g a day for three months, and concluded that a cause-and-effect relationship has not been established. No European regulator has accepted a comparable claim for alpha-GPC either.
The safety picture is not symmetrical
This section counts in citicoline's favour and it would be dishonest of us to bury it.
In 2021 JAMA Network Open published a South Korean national cohort study: 12,008,977 people aged 50 and over, followed for ten years, comparing 108,877 who had been prescribed alpha-GPC with matched non-users. The adjusted hazard ratio for total stroke was 1.46 (95% CI 1.43–1.48) before matching and 1.43 afterwards. Three things belong alongside that number. Korean prescriptions run at roughly 1,200 mg a day, far above anything used in healthy-adult research, in an older and clinically diagnosed population. The dose-response the authors measured was by length of prescription rather than by milligrams, so the study does not show that a smaller amount is safer. And a 2025 review in Nutrition Reviews concluded that the risk "awaits necessary validation".
We found no comparable population-scale signal for citicoline. That absence is not proof of safety — nobody has run the equivalent study — but as things stand, one of these two ingredients has a large published cohort raising a question and the other does not.
Cost, with the arithmetic shown
Prices below were read from each brand's own shopping page in August 2026, normalised to a day at the amount used in the relevant healthy-adult trial: 300 mg for alpha-GPC, 500 mg for citicoline.
| Product | Form | Per capsule | Price / count | Cost per trial-amount day |
|---|---|---|---|---|
| Nutricost Alpha GPC | Alpha-GPC | 300 mg | $27.97 / 120 | $0.23 |
| Nootropics Depot Alpha GPC | Alpha-GPC | 150 mg | $34.99 / 180 | $0.39 |
| Swanson Alpha-GPC | Alpha-GPC | 300 mg | $31.39 / 60 | $0.52 |
| Jarrow Formulas Citicoline | Citicoline (CDP-choline) | 250 mg | $33.99 / 60 | $1.13 |
| Swanson Cognizin Citicoline | Citicoline (branded) | 500 mg | $58.49 / 60 | $0.97 |
| Weal Focus | Alpha-GPC | 150 mg × 2 | $39.99 / 60 | $1.33, with four other actives |
Amounts from the supplement facts panels via the NIH Dietary Supplement Label Database; prices from each brand's own site, August 2026. Weal Focus is listed for completeness and is not a like-for-like comparison — it carries four additional actives.
At the amounts used in research, citicoline runs roughly two to four times the daily cost of alpha-GPC. Part of that premium is real: the citicoline trial used a specific branded, patented material, and buying that same material is buying the thing that was tested. Part of it is simply that citicoline is a more complex molecule to manufacture.
Who each one suits
Citicoline suits someone over fifty who wants the best-evidenced option for memory over months rather than hours, and who is willing to pay roughly a dollar a day for it. It has the larger and longer healthy-adult trial, the branded material from that trial is buyable, and we found no population-scale safety signal against it. The honest caveats are that Europe's regulator reviewed the same literature and declined to accept the claim, and that the positive outcomes in that trial were secondary rather than primary.
Alpha-GPC suits someone who wants more choline per milligram at a lower daily cost, and whose interest is nearer-term attention performance than a twelve-week memory outcome. Its healthy-adult data is thinner and shorter, and the Korean cohort is a genuine reason for anyone over fifty, anyone with cardiovascular history, and anyone taking anticoagulants to talk to a physician before starting — and a reason for anybody to think about how long they intend to take it continuously, since duration was the axis that cohort actually measured.
Neither suits someone shopping on the word "choline" alone. Most bottles carrying that word contain choline bitartrate, which is a third form, and cheaper than both of these for reasons that show up in the research.
What we would like to see, and do not have
A twelve-week trial in healthy adults comparing 300 mg of alpha-GPC against 500 mg of citicoline against placebo, on the same cognitive battery, would settle most of this page. It does not exist. Until it does, anyone claiming one of these forms is superior to the other is making an argument from mechanism, not from measurement, and should say so.
What we make, and why we chose what we chose
Weal Focus uses alpha-GPC at 300 mg, alongside lion's mane 400 mg at 100% fruiting body with beta-glucans of at least 25%, L-theanine 200 mg, zinc 15 mg and B12 500 mcg, in two size-00 capsules. Every amount is printed and there are no proprietary blends.
We chose alpha-GPC on choline density and cost per day, and we set it at 300 mg rather than higher because 300 mg sits inside the range used in healthy-adult research while the larger figures come from people with diagnosed cognitive impairment. A reader who weighs the same evidence and prefers citicoline has reached a defensible conclusion from the same page, and three of the products in the table above will sell it to them.
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.