Citicoline (CDP-Choline)

Overview

Citicoline (CDP-choline) is the molecule every cell uses one step before building phosphatidylcholine, the main fat in brain cell membranes, and it also carries a choline payload for the neurotransmitter acetylcholine. Swallowed as a capsule it is split in the gut into choline and uridine, which the brain reassembles. Trials at 250 to 500 mg a day improved attention in healthy adults and memory in older adults with age-related memory complaints, mostly in manufacturer-funded studies of four to twelve weeks.

What is Citicoline (CDP-Choline)?

Key takeaways

• Swallowed citicoline never reaches the brain intact: the gut wall and liver split it into choline and cytidine, the cytidine shows up in human blood as uridine, and brain cells rebuild CDP-choline from the parts (Wurtman 2000, Biochemical Pharmacology).
• In 60 healthy women aged 40 to 60, 28 days of 250 mg or 500 mg of citicoline cut errors on a 14-minute computerized attention test compared with placebo, and the 250 mg group did at least as well as the 500 mg group (McGlade 2012, Food and Nutrition Sciences).
• In 100 adults aged 50 to 85 with age-related memory complaints, 500 mg a day for 12 weeks improved episodic memory and a four-test memory composite more than placebo; 99 of 100 finished and blood chemistry stayed normal (Nakazaki 2021, Journal of Nutrition).
• Every healthy-adult trial on this page was funded by or co-authored with Kyowa Hakko, the company that makes Cognizin, and none has been independently replicated at supplement doses.
• The large safety datasets come from prescription use: 2,298 patients on 2,000 mg a day for six weeks had the same adverse-event rate as placebo (Dávalos 2012, Lancet), and 4,191 patients on 500 to 4,000 mg a day had a side-effect rate of 0.73 percent (Cho 2009).
• Two interactions appear on the Spanish prescribing information: citicoline potentiates levodopa, and it is not to be combined with meclofenoxate (centrophenoxine); plain choline has no known drug interactions (NIH ODS).

Citicoline is cytidine 5'-diphosphocholine, usually shortened to CDP-choline: a cytidine (the same cytidine found in RNA) bonded through two phosphate groups to a choline. That makes it a nucleotide, the same chemical family as ATP, and not a vitamin, an amino acid, or a plant extract. Every cell in the body makes it all day, because it is the step just before phosphatidylcholine in the Kennedy pathway, the assembly line worked out in Eugene Kennedy's laboratory in the 1950s that builds the most abundant phospholipid in cell membranes. In the brain, membranes are the whole product. Every synapse, every receptor, and every mitochondrion sits inside one.

That double payload, choline and cytidine together, is what separates citicoline from a plain choline supplement. Choline bitartrate or alpha-GPC gives the body choline and nothing else. Citicoline gives it choline plus cytidine, and the cytidine half is the one that feeds the rate-limiting enzyme of the Kennedy pathway. The trade is that citicoline is a weak choline source by weight: 500 mg carries about 107 mg of choline, roughly half of what 500 mg of choline bitartrate or alpha-GPC carries. If the goal were choline alone, citicoline would be the expensive way to get it. The argument for citicoline, made most clearly by Richard Wurtman's group at MIT, is that the cytidine half does something choline cannot.

Here is the fact that reframes most of the marketing. Citicoline does not survive digestion. Oral absorption is close to complete (less than 3 percent of a labeled dose turns up in urine and feces, and oral bioavailability is about the same as an injection, according to the Spanish prescribing information), but the molecule is hydrolyzed in the intestinal wall and liver into choline and cytidine before it reaches the bloodstream. In 12 adults given 500, 2,000, or 4,000 mg, plasma choline stayed elevated for 5, 8, and 10 hours, cytidine was never reliably detectable, and plasma uridine rose 70 to 90 percent after 500 mg and 100 to 120 percent after 2,000 mg, with no further rise at 4,000 mg (Wurtman 2000, Biochemical Pharmacology). Humans, unlike rats, convert the cytidine to uridine. What reaches a human brain is choline and uridine, and neurons rebuild CDP-choline from those on site. Paweł Grieb's 2014 review in CNS Drugs adds the honest caveat that nobody has fully excluded a role for the phosphorylated fragments, and that the mechanism of action is still not adequately described. Either way, "which brand reaches the brain better" is the wrong question. Nothing reaches the brain as citicoline.

Citicoline has two lives. In Spain it has been a prescription drug since June 1978 (Somazina, made by Ferrer, dosed at 500 to 2,000 mg a day, ATC code N06BX06) for neurological indications that the evidence section covers, and it is prescribed in several other European and Asian countries. In the United States it is a dietary ingredient. Kyowa Hakko, the Japanese fermentation company that makes Cognizin, self-affirmed GRAS (generally recognized as safe) status for it in 2009, which is a company determination rather than an FDA review, and says it now sits in more than 200 supplements and drinks. The inner-salt form, which is the free citicoline molecule rather than its sodium salt, is the version declared a food ingredient in the major markets, while the prescription products use citicoline sodium (Synoradzki and Grieb 2019, Nutrients). No health claim for citicoline has been authorized in the EU, and none has been authorized by the FDA. Our full review lays out the case for the fermentation-made branded form in detail; this page is the shorter and more skeptical layer on top of it.

Most people are short on choline, which is a separate question from whether they need citicoline. The Adequate Intake is 550 mg a day for men and 425 mg for women, and NHANES data put average intake at 402 mg and 278 mg (NIH ODS). A citicoline capsule is a poor way to close that gap. Two eggs supply nearly three times the choline of a 500 mg citicoline dose. The reason to take citicoline is the cytidine, the membrane and energy signals in the scanner studies, and the attention and memory trials graded below, not the choline count. Where it sits among the other ingredients people stack for focus is covered in the mood, mind, and memory supplement overview.

How it works

Citicoline has two jobs in a neuron, and both start with the same split. Once the choline and the uridine from a dose arrive, the Kennedy pathway can run: choline kinase phosphorylates choline to phosphocholine, then CTP:phosphocholine cytidylyltransferase (the rate-limiting enzyme, which needs cytidine triphosphate, CTP) joins it to cytidine to remake CDP-choline, and a final enzyme hands the phosphocholine head group to a diacylglycerol to produce phosphatidylcholine. Uridine is the part that matters here, because neurons convert it to UTP and then CTP, and CTP availability is what limits the slow step. Supply both substrates and, the proposal goes, membrane synthesis runs faster.

Two small scanner studies are the human evidence. Phosphorus magnetic resonance spectroscopy (31P-MRS) measures phosphorus-containing molecules in living brain tissue without a needle. In healthy older adults given 500 mg a day for six weeks, brain phosphodiesters, the breakdown products that track membrane turnover, rose 7.3 percent, glycerophosphoethanolamine rose 11.6 percent, and the increase correlated with improvement on a verbal learning test; a second six weeks added nothing further (Babb 2002, Psychopharmacology). In 16 adults averaging 47 years old who took 500 or 2,000 mg of Cognizin for six weeks, the anterior cingulate cortex showed 7 percent more phosphocreatine, 14 percent more beta-nucleoside triphosphate (mostly ATP in brain), and a 32 percent higher ratio of phosphocreatine to inorganic phosphate, with the lower dose producing the larger change (Silveri 2008, NMR in Biomedicine). That study had no placebo group, 16 people, and Kyowa co-authors. Real signal, thin evidence.

The second job is acetylcholine, the neurotransmitter that attention and memory circuits run on. Choline acetyltransferase combines choline with acetyl-CoA to make it, and the free choline from a citicoline dose feeds that reaction directly. Wurtman's older hypothesis, built on animal work, is that when neurons fire hard and free choline runs short, they strip choline out of their own membrane phosphatidylcholine to keep making acetylcholine, a process he called autocannibalism; supplying choline is meant to spare the membrane. Choline is also, on its own, a selective agonist at the alpha-7 nicotinic acetylcholine receptor, a receptor involved in filtering sensory noise. That mechanism is how Verner Knott's group explained one of the more useful findings on this page: in 24 healthy men, a single 500 or 1,000 mg dose of CDP-choline improved processing speed, working memory, verbal learning, and executive function in the people who started below average, did nothing in the average performers, and made the high performers slightly worse (Knott 2015, Pharmacology Biochemistry and Behavior). If you are already sharp, the receptor may already be busy.

The third mechanism is dopamine, and it has the least human data. In animals, citicoline increases dopamine release and the density of dopamine receptors in the striatum (reviewed in Secades and Gareri 2022). The authors of the attention trials proposed dopamine as the reason their participants made fewer errors, and said plainly that they had not measured it (McGlade 2012). Call it proposed. What would prove it is a human PET or microdialysis study that nobody has run.

This is a different route to cellular energy than coenzyme Q10, which works inside the mitochondrial electron transport chain, or NMN, which raises the NAD+ those mitochondria burn. Citicoline's energy signal shows up as phosphocreatine on a scan, and nobody has yet shown that the scan result is something a person feels. What people in the trials noticed was nothing. The measured differences took 28 days to appear on an attention test and 12 weeks on memory tests. The mechanism predicts three outcome families, attention, memory, and brain energy, and the evidence below grades each one on what the trials actually measured.

Evidence at a glance

What the evidence doesn't support yet

"Clinically proven to improve memory" is the claim most citicoline labels lean on, and it is ahead of the data for a healthy buyer. The one healthy-population memory trial recruited people who already had age-related memory complaints, found its gains on secondary outcomes, and was designed and run by the manufacturer (Nakazaki 2021). The pooled memory evidence in the Cochrane review comes from older adults with chronic cerebrovascular cognitive impairment in trials mostly under three months (Fioravanti 2005). Nobody has shown citicoline improves memory in a healthy 35-year-old. The single-dose data suggest it may do nothing for people already performing well (Knott 2015).

"Cognizin is proven superior to generic citicoline" is the second claim, and it is ours as much as anyone's. No head-to-head trial exists. The largest efficacy and safety datasets, ICTUS and the Korean surveillance study, used citicoline sodium. Our full review makes the fermentation and stability case; the trial record does not settle it.

Neuroprotection is the third. Decades of animal data showed citicoline shrinking brain lesions, and the largest human test found nothing: ICTUS was stopped for futility (Dávalos 2012), and Grieb's review calls the mechanism not adequately described.

Two smaller gaps. Mood: the only randomized trial used 200 mg a day as an add-on to an antidepressant in 50 patients (Roohi-Azizi 2017), and nothing exists in healthy adults. Citicoline versus alpha-GPC: a 2025 meta-analysis of three trials in 358 people with dementia found alpha-GPC scored better on a global geriatric scale with no difference on memory tests (Sagaro and Amenta 2025), and there is no comparison in healthy adults at all.

Clinical evidence

• Citicoline and Memory Function in Healthy Older Adults: A Randomized, Double-Blind, Placebo-Controlled Clinical Trial — 100 healthy men and women aged 50 to 85 with age-associated memory impairment (memory complaints with otherwise normal cognition), 12 weeks, 500 mg/day Cognizin citicoline versus placebo (49 versus 51). Measured computerized memory tests (Cambridge Brain Sciences) plus adverse events, weight, blood pressure, hematology, and a metabolic panel. Episodic memory on the paired-associate test improved more with citicoline (0.15 versus 0.06, P = 0.0025) and composite memory across four tests improved 3.78 versus 0.72 points (P = 0.0052); 99 of 100 completed with no safety signal. Limitation: the significant results were secondary outcomes, and the trial was designed, funded, and partly authored by Kyowa Hakko, the maker of Cognizin.
• Improved Attentional Performance Following Citicoline Administration in Healthy Adult Women — 60 healthy women aged 40 to 60 in Salt Lake City, 28 days, one daily capsule of 250 mg or 500 mg Cognizin citicoline or placebo (20 per group). Measured the Conners Continuous Performance Test II, a 14-minute computerized test of sustained attention and response inhibition. The 250 mg group made fewer omission (P = 0.04) and commission (P = 0.03) errors than placebo; the 500 mg group made fewer commission errors (P = 0.03) and trended on omissions (P = 0.07). Limitation: 20 per arm, women only, one-tailed statistics, a journal not indexed in PubMed, and three of eight authors employed by Kyowa Hakko.
• The Effect of Citicoline Supplementation on Motor Speed and Attention in Adolescent Males — 75 healthy adolescent males, 28 days, 250 or 500 mg/day Cognizin citicoline (n = 51) versus placebo (n = 24). Measured the Ruff 2 and 7 selective attention test, a finger-tap test, and the Conners CPT-II. Citicoline improved attention (P = 0.02) and psychomotor speed (P = 0.03) versus placebo, and a higher weight-adjusted dose predicted better attention accuracy, better signal detection, and lower impulsivity. Limitation: unequal randomization, two doses pooled, published online in 2015, and co-authored by Kyowa Hakko staff.
• Improvements in concentration, working memory and sustained attention following consumption of a natural citicoline-caffeine beverage — 60 healthy adults, a single session, a commercial beverage containing citicoline and caffeine versus placebo. Measured EEG event-related potentials and tests of attention, concentration, and reaction time. The beverage group had faster maze learning and continuous-performance reaction times, fewer go/no-go errors, better information-processing accuracy, and higher P450 amplitudes, read as greater sustained attention. Limitation: citicoline and caffeine were given together, so the citicoline contribution cannot be separated, and the beverage maker supported the study.
• Neurocognitive effects of acute choline supplementation in low, medium and high performer healthy volunteers — 24 healthy men, three sessions, single doses of 500 mg or 1,000 mg CDP-choline or placebo in random order. Measured the CogState battery. Relative to placebo, CDP-choline improved processing speed, working memory, verbal learning, verbal memory, and executive function in participants with low baseline performance, had no effect in medium performers, and reduced performance in high performers; effects appeared at both doses. Limitation: 24 men, acute dosing only, and stratification into performance tiers was done after the fact.
• Chronic citicoline increases phosphodiesters in the brains of healthy older subjects: an in vivo phosphorus magnetic resonance spectroscopy study — A small group of healthy older adults, 12 weeks: all took 500 mg/day oral citicoline for six weeks, then were randomized to citicoline or placebo for a second six weeks, with 31P-MRS brain scans at baseline, 6, and 12 weeks. Phosphodiesters rose 7.3 percent (P = 0.008) and glycerophosphoethanolamine 11.6 percent at six weeks, and the rise correlated with improvement on the California Verbal Learning Test; continuing for six more weeks added nothing. Limitation: small sample, the first six weeks were open-label, and the memory correlation was exploratory.
• Effect of oral CDP-choline on plasma choline and uridine levels in humans — 12 mildly hypertensive but otherwise normal fasting adults, four treatment days in random order: 500, 2,000, or 4,000 mg oral citicoline or placebo, with 11 plasma samples over 12 hours. Measured plasma choline, cytidine, and uridine. Choline rose dose-dependently and stayed elevated for 5, 8, and 10 hours; uridine rose 70 to 90 percent after 500 mg and 100 to 120 percent after 2,000 mg, with no further rise at 4,000 mg; cytidine was not reliably detectable. Limitation: 12 people, single doses, and plasma levels rather than brain levels.
• Citicoline Combination Therapy for Major Depressive Disorder: A Randomized, Double-Blind, Placebo-Controlled Trial — 50 adults with major depressive disorder already taking citalopram, 6 weeks, 100 mg citicoline twice daily or placebo added on. Measured the Hamilton Depression Rating Scale at weeks 2, 4, and 6. The citicoline group improved more at every time point (P = 0.030, 0.032, 0.021) and had a higher remission rate (P = 0.045). Limitation: 25 per arm, a single center, a clinical population on an antidepressant, and no data in healthy adults.
• Citicoline in the treatment of Parkinson's disease — 85 patients with Parkinson's disease on levodopa, 4 weeks, randomized to their usual levodopa dose (mean 381 mg) plus 1,200 mg/day citicoline or half their usual levodopa (mean 196 mg) plus citicoline. Measured the Webster Rating Scale, pegboard, drawing, writing, walking, and mood tests. No significant between-group differences; more patients improved in the half-dose group. Limitation: single-blind, no placebo arm, and four weeks; it is cited here for the levodopa interaction, not as an efficacy claim.
• Citicoline in the treatment of acute ischaemic stroke: an international, randomised, multicentre, placebo-controlled study (ICTUS trial) — 2,298 patients with moderate-to-severe acute ischemic stroke at 59 centers in Spain, Portugal, and Germany, 6 weeks, 1,000 mg citicoline every 12 hours intravenously for 3 days then 500 mg orally twice daily, versus placebo. Measured global recovery at 90 days (NIHSS, modified Rankin, Barthel). Recovery was the same in both groups (odds ratio 1.03) and the trial was stopped for futility; adverse events and safety variables did not differ. Limitation: a disease population at four times a supplement dose; included for the safety data and as the largest negative efficacy trial.
• Cytidinediphosphocholine (CDP-choline) for cognitive and behavioural disturbances associated with chronic cerebral disorders in the elderly — 14 double-blind, placebo-controlled randomized trials in older adults ranging from subjective memory complaints to mild-to-moderate vascular cognitive impairment or dementia, mostly 20 to 90 days long, one lasting 12 months, at heterogeneous doses. Pooled outcomes for attention, memory, behaviour, global impression, and tolerability. No evidence of benefit on attention; evidence of benefit on memory and behaviour, stronger on global clinical impression; well tolerated. Limitation: short, heterogeneous trials with no intention-to-treat data, in populations that are not healthy adults.
• Efficacy and safety of oral citicoline in acute ischemic stroke: drug surveillance study in 4,191 cases — 4,191 Korean patients with acute ischemic stroke, at least 6 weeks, oral citicoline 500 to 4,000 mg/day, with 125 continuing beyond 12 weeks. Measured stroke and function scales and all adverse reactions. Function scores improved over six weeks; 37 side effects occurred in 31 patients (0.73 percent), most often nervous-system (8 of 37) and gastrointestinal (5 of 37) symptoms. Limitation: no control group, so the efficacy result is uninterpretable; it is the largest tolerability dataset for oral citicoline.
• Association of L-α Glycerylphosphorylcholine With Subsequent Stroke Risk After 10 Years — 12,008,977 South Koreans aged 50 and older without prior stroke or Alzheimer disease, followed from 2009 to 2018 after alpha-GPC prescription use in 2006 to 2008 (108,877 users). Measured incident total, ischemic, and hemorrhagic stroke. After matching on all covariates, alpha-GPC users had a 43 percent higher risk of total stroke (hazard ratio 1.43), rising with longer use. Limitation: observational prescription data in people given the drug for a reason, so confounding by indication is unresolved; it concerns alpha-GPC, not citicoline, and is here because the two are routinely confused.
• No Acute Effects of Choline Bitartrate Food Supplements on Memory in Healthy, Young, Human Adults — 94 healthy young adults across three double-blind, placebo-controlled crossover experiments (28, 26, and 40 participants), single doses of 2.0 to 2.5 g choline bitartrate, tested about 60 minutes later. Measured visuospatial working memory, declarative picture memory, and verbal working memory. Choline did not improve any memory measure, and Bayesian analysis strongly supported the null. Limitation: acute dosing only; it tests choline bitartrate, included as the comparison buyers most often confuse with citicoline.
• Acute Alpha-Glycerylphosphorylcholine Supplementation Enhances Cognitive Performance in Healthy Men — 20 resistance-trained men averaging 31 years, three sessions, 630 mg or 315 mg alpha-GPC or placebo 60 minutes before testing. Measured Stroop, N-Back, and Flanker tests, mood scales, hemodynamics, jump and bench-throw power, and growth hormone. Stroop total score improved more than placebo at both doses (effect sizes 0.61 and 0.48) and completion time at the higher dose; N-Back, Flanker, power, and hormones did not differ. Limitation: 20 men, a single acute dose, a sponsor-supplied ingredient, and retrospective trial registration; included as the healthy-adult alpha-GPC comparison.
• Citicoline enhances frontal lobe bioenergetics as measured by phosphorus magnetic resonance spectroscopy — 16 healthy men and women averaging 47 years, 6 weeks, 500 mg or 2,000 mg/day Cognizin citicoline, with 31P-MRS at 4 tesla before and after. Measured phosphorus metabolites in the anterior cingulate cortex and a parieto-occipital comparison region. Phosphocreatine rose 7 percent, beta-nucleoside triphosphates (mostly ATP) 14 percent, and the phosphocreatine to inorganic phosphate ratio 32 percent in the anterior cingulate, with membrane phospholipid changes, and the lower dose tended to produce the larger change. Limitation: no placebo group, 16 people, and Kyowa co-authors.

Dosage & forms

• Cognizin citicoline (branded inner salt, Kyowa Hakko, fermentation-made):
• Generic citicoline sodium:
• Alpha-GPC (L-alpha-glycerylphosphorylcholine, choline alphoscerate):
• Choline bitartrate:
• Phosphatidylcholine and lecithin:

Safety & interactions

How much to take

The research dose for healthy adults is 250 to 500 mg a day, and most products carry exactly that. The attention trial in women found 250 mg worked at least as well as 500 mg over 28 days (McGlade 2012); the adolescent trial pooled 250 and 500 mg and found the higher weight-adjusted dose predicted better accuracy (McGlade 2019); the memory trial in older adults used 500 mg for 12 weeks (Nakazaki 2021). So a 250 mg serving matches the attention trials, and a 500 mg serving matches every healthy-adult trial on this page. Below 250 mg, no trial has looked. Above 500 mg, the scanner studies give a surprising answer: the 500 mg dose produced a larger brain signal than 2,000 mg (Silveri 2008), and plasma uridine stopped rising between 2,000 and 4,000 mg (Wurtman 2000). More is not better here. A 1,000 mg capsule is buying a prescription dose for a supplement question.

Take it once a day; a split dose is not needed. Morning fits the attention data, and it does not matter whether you take it with food, since the gut wall releases the choline and uridine either way (Somazina prescribing information). The one trial that stacked it with caffeine in a beverage found better sustained attention than placebo but could not separate the two (Bruce 2014). Keep it away from nothing in particular, with the two prescription-drug exceptions in the table below.

Expect no felt effect. In the trials the attention difference appeared at 28 days, the brain phosphorus markers at six weeks, and the memory difference at 12 weeks. Same-day effects exist in one dataset only: people who started below average on the test (Knott 2015). Twelve weeks is the longest any healthy-adult trial has run, so that is the honest ceiling on what "long term" means for the evidence. No tolerable upper intake level has been set for citicoline. The one that applies to its choline content, 3,500 mg of choline a day for adults (NIH ODS), is unreachable from citicoline: 500 mg supplies about 107 mg of choline, a thirtieth of the limit.

Safety profile

The safety record is the strongest part of the citicoline file, and most of it comes from prescription use at doses two to eight times a supplement serving. In the ICTUS trial, 1,148 patients received 2,000 mg a day for six weeks, intravenous for three days and oral after that. Their adverse-event rate did not differ from the 1,150 on placebo (Dávalos 2012, Lancet). A Korean surveillance study followed 4,191 patients on 500 to 4,000 mg a day for at least six weeks and recorded 37 side effects in 31 patients, a rate of 0.73 percent; the most common were nervous-system symptoms (8 of 37) and digestive symptoms (5 of 37) (Cho 2009). The Cochrane review of 14 placebo-controlled trials in older adults concluded the drug was well tolerated (Fioravanti 2005). At supplement doses, the 100-person memory trial tracked adverse events, body weight, blood pressure, hematology, and a metabolic panel for 12 weeks and found nothing of concern, with 99 of 100 completing (Nakazaki 2021).

Animal toxicology matches. The oral LD50 is 27 g/kg in mice and 18.5 g/kg in rats, dogs given 1.5 g/kg a day by mouth for six months showed no abnormalities, and rabbits given 800 mg/kg during organogenesis showed no maternal or fetal toxicity apart from a slight delay in skull ossification in 10 percent of treated fetuses (Secades and Gareri 2022; Somazina prescribing information). The intravenous LD50 of citicoline is about 44 times that of choline chloride, which is part of why the cytidine-bound form was developed in the first place (Secades and Gareri 2022).

Side effects by frequency

Common (above placebo): none established in any trial. Uncommon: digestive upset and headache, the two categories that led the surveillance study at under one percent combined (Cho 2009). Very rare (fewer than 1 in 10,000, from individual reports on the Spanish prescribing information): hallucinations, headache, vertigo, high or low blood pressure, shortness of breath, nausea, vomiting, occasional diarrhea, flushing, hives, rash, purpura, chills, and swelling. There is no known overdose syndrome; the prescribing information states that no intoxication is expected even when therapeutic doses have been accidentally exceeded.

Drug interactions

Who should ask a clinician first

Anyone taking levodopa or another dopamine-replacement drug, before the first dose. Pregnant and nursing women: the Spanish prescribing information says there are insufficient data in pregnancy and the drug should not be used unless clearly necessary, and every trial on this page excluded them. Anyone under 18, since the one adolescent trial ran 28 days in healthy boys and nothing longer exists (McGlade 2019). Anyone whose clinician has described them as having excessive parasympathetic (vagal) tone, which the Spanish label lists as a contraindication. People on cholinesterase inhibitors or multiple anticholinergic drugs, for the reasons in the table. Anyone with a history of hallucinations, which sit on the very-rare list.

Signs of taking too much are the signs of choline excess, fishy body odor, sweating, salivation, low blood pressure, and digestive upset (NIH ODS), and they take grams of choline, not milligrams. If you take levodopa, this is the one supplement on your shelf that can change how your prescription behaves, so the conversation happens before the first capsule.

Buying guide

The dose: 250 to 500 mg, stated as milligrams of citicoline

Match the number on the label to the trials. The attention trials used 250 and 500 mg and the memory trial used 500 mg, so a product anywhere in that window is inside the research. A product that lists "citicoline sodium 500 mg" is giving you about 479 mg of citicoline, which is fine, as long as you know that is the math. A product at 100 mg is below anything tested in healthy adults. A product at 1,000 mg or more is paying for a prescription dose that produced a smaller brain signal than 500 mg in the only studies that compared them. Combination formulas typically land at 250 mg, which is the low end of the studied range and the dose that worked for attention; if memory in later life is the goal, 500 mg is the dose with the trial.

The form: citicoline, named, with the salt stated

The label should say "citicoline" or "CDP-choline," and ideally whether it is the inner salt or citicoline sodium. A branded form such as Cognizin tells you the material is the fermentation-made inner salt used in the attention and memory trials, with a published specification behind it. Generic citicoline sodium is the same molecule as a sodium salt and the form behind the largest safety datasets. Either is a legitimate choice. What is not a substitute is "choline" on the label with no further detail: choline bitartrate is a different compound with a null trial record for memory in healthy adults, and alpha-GPC is a different compound with its own open questions, both covered in the forms table above. "Choline complex" or "cholinergic blend" with no milligrams is a proprietary blend, and proprietary blends hide doses.

Proof of testing

Citicoline is a white, water-soluble powder that pulls moisture from the air, and a wet, clumped raw material is the most common quality failure. Ask any brand, including us, for a certificate of analysis for the lot in your bottle: a lab report showing identity, potency against the label claim, and contaminant screens. Our products are made in our own cGMP facility in Mt. Angel, Oregon; every batch is tested in our in-house lab for identity, potency, and purity; and certificates of analysis are published by lot on the lab-testing page. How to read a supplement certificate of analysis explains which lines matter, and what we found when we tested 44 supplements bought on Amazon explains why we keep asking. Demand the same document from any brand.

Fillers and the capsule

A citicoline capsule needs almost nothing besides citicoline. Look for a short "other ingredients" line: a vegetable cellulose capsule, rice extract or rice concentrate as a flow agent, silica. Magnesium stearate is harmless and common. A long list of colors, sweeteners, or "natural flavors" in a capsule signals a formula built around the marketing rather than the dose. Powders and drinks are fine for the compound itself, since it is water-soluble and stable in solution, but check that the citicoline milligrams are listed per serving rather than per container.

Price per day for the real dose

Price the milligrams, not the bottle. Multiply the citicoline per serving by the servings per container, divide the price by the result, and compare brands on cents per 100 mg. Then ask what else is in the capsule, because a combination formula is priced for everything in it.

Brain Advantage carries 250 mg of Cognizin citicoline per two-capsule serving, alongside 320 mg of Synapsa bacopa extract (standardized to 55 percent bacosides), 500 mg of lion's mane extract, 60 mg of ginkgo extract, 30 mg of L-theanine, and vitamin C, niacinamide, folate, magnesium, and zinc; 30 servings for $61.99, about $2.07 a day, in a vegetable cellulose capsule with rice extract, silica, and rice concentrate as the only other ingredients. It is caffeine-free. The 250 mg of citicoline is the dose that improved attention scores in the 28-day trial and half the dose the 12-week memory trial used; the formula supports attention, memory, and mental stamina as part of a daily routine.* The bacopa half of the formula has its own trial record, covered in our Synapsa bacopa review. Price, serving count, and label checked on the live product page 2026-10-14.

Red flags

• "Choline" with no compound named, or citicoline inside a proprietary blend with no milligram amount.
• "Clinically proven" with no study named. The trials on this page are the ones a label can honestly point to, and all of them are small.
• 1,000 mg or more per serving sold as "clinical strength" with no reason given.
• A label that does not distinguish citicoline from choline bitartrate, or prices them the same.
• Any brand that will not produce a lot-specific certificate of analysis on request.

The bottom line

If you are over 50, notice your memory is not what it was, and want the one dose with a placebo-controlled trial behind it, 500 mg of citicoline a day for 12 weeks is that dose, and the trial was clean on safety. If you want a focus ingredient that has moved an attention test in healthy adults, 250 to 500 mg for a month is where the evidence is, with the caveat that the manufacturer paid for the trials. If you are already sharp and under 40, the single-dose data say you may get nothing, and your money is better spent on sleep. If you take levodopa, do not buy it until your prescriber says so. If what you actually have is stress rather than a focus problem, ashwagandha has the larger trial record for that. And if you are taking it for the choline, eat two eggs instead.

Frequently Asked Questions

What is citicoline and how is it different from choline or alpha-GPC?

Citicoline is choline bonded to cytidine through two phosphates, the molecule every cell makes one step before phosphatidylcholine. Choline bitartrate and alpha-GPC supply choline only; citicoline supplies choline plus cytidine, which reaches the human brain as uridine and feeds the rate-limiting step of membrane synthesis (Wurtman 2000). Per milligram it is the weaker choline source, about 107 mg of choline per 500 mg versus about 200 mg from the same weight of alpha-GPC or choline bitartrate. The trial records differ too: citicoline has 28-day and 12-week placebo-controlled trials in healthy adults, alpha-GPC has single-dose data, and choline bitartrate has a null memory trial (Lippelt 2016). Nothing reaches the brain as citicoline; the parts do.

What did the human trials of citicoline measure?

Tests, not feelings. The healthy-adult trials measured a 14-minute computerized attention test (errors of omission and commission) in women aged 40 to 60 (McGlade 2012), attention, finger-tap speed, and impulsivity in adolescent males (McGlade 2019), and computerized memory tests, mainly paired-associate episodic memory, in adults aged 50 to 85 with age-related memory complaints (Nakazaki 2021). Two scanner studies measured brain phosphorus compounds: phosphocreatine and ATP in the frontal lobe (Silveri 2008) and membrane phosphodiesters in older adults (Babb 2002). A single-dose crossover measured processing speed and working memory on the CogState battery (Knott 2015). The prescription trials measured recovery and function scales in patient populations, which is a different question.

What dose was used in the attention and memory trials?

250 mg and 500 mg for attention, 500 mg for memory. The women's attention trial gave 250 or 500 mg once daily for 28 days and both beat placebo, with 250 mg doing at least as well (McGlade 2012). The adolescent trial used 250 or 500 mg for 28 days (McGlade 2019). The memory trial in older adults used 500 mg for 12 weeks (Nakazaki 2021). The scanner studies compared 500 mg with 2,000 mg and found the smaller dose produced the larger brain signal (Silveri 2008), and plasma uridine plateaued between 2,000 and 4,000 mg (Wurtman 2000). A 250 mg serving matches the attention data; 500 mg matches everything. Nothing above 500 mg has shown more in healthy adults.

Is Cognizin different from generic citicoline?

Same molecule, different salt, different paper trail. Cognizin is citicoline as the free inner salt, made by Kyowa Hakko's fermentation process, and it is the material in every healthy-adult trial on this page. Generic citicoline is usually citicoline sodium, the salt used in the prescription products, so 500 mg of it carries about 479 mg of citicoline; it is the form behind the largest safety datasets, including 2,298 patients in the ICTUS trial and 4,191 in the Korean surveillance study. Both are split into the same choline and uridine in the gut. No trial has compared them head to head. [Our full review](/research/cognizin-citicoline-neuropharmacology) makes the case for the branded form on purity and stability; the trial record does not settle it, and we say so.

How long before citicoline has an effect?

Weeks, with one exception. The attention difference appeared at the 28-day test (McGlade 2012, 2019), the brain phosphorus markers at six weeks with no further change by twelve (Babb 2002), and the memory difference at 12 weeks (Nakazaki 2021). The exception is a single 500 or 1,000 mg dose, which shifted CogState scores within hours, but only in people who started below average on the tests; average performers got nothing and high performers got slightly worse (Knott 2015). Plasma choline and uridine stay up for 5 to 10 hours after a dose (Wurtman 2000). If you notice nothing in a month, that matches the trials, where nobody noticed anything either; the differences were on the tests.

Is citicoline safe long term?

Safe for as long as anyone has tested, which is the honest limit. Healthy adults have been studied for 12 weeks at 500 mg with normal blood chemistry and 99 of 100 completing (Nakazaki 2021). Patients have taken 2,000 mg a day for six weeks with adverse events equal to placebo (Dávalos 2012) and 500 to 4,000 mg a day for six weeks or more with a 0.73 percent side-effect rate (Cho 2009). The Cochrane review's 14 trials included one lasting 12 months and called the drug well tolerated (Fioravanti 2005). It has been a prescription drug in Spain since 1978. What does not exist is a multi-year trial in healthy adults, so "long term" means the 12-week record plus decades of prescription use.

Does citicoline interact with any medications?

Two interactions are on the Spanish prescribing information for citicoline: it potentiates levodopa, and it is not to be combined with meclofenoxate (centrophenoxine). The levodopa effect has trial support: 85 patients on levodopa cut their dose in half with 1,200 mg of citicoline and lost nothing on their rating scales (Eberhardt 1990). Beyond those, the interactions are mechanistic: anticholinergic drugs such as diphenhydramine and oxybutynin block the receptors citicoline feeds, and cholinesterase inhibitors push in the same direction it does. Caffeine has no known interaction. Choline itself has no clinically relevant drug interactions (NIH ODS). The safety section above has the full table with what to do for each.

Who should not take citicoline?

Anyone on levodopa, until the prescriber says so, because citicoline changes how the drug behaves. Anyone using meclofenoxate (centrophenoxine), because the two are not to be combined. Pregnant and nursing women, because the prescribing information says data are insufficient and every trial excluded them. Anyone under 18 beyond the single 28-day adolescent trial. Anyone whose clinician has described excess parasympathetic or vagal tone, which is a labeled contraindication. People with a history of hallucinations, which sit on the very-rare side-effect list. And, less dramatically, people already performing well on attention tests, for whom the single-dose data showed no benefit and a slight cost (Knott 2015). If your attention and memory are fine, you probably do not need it.

Can citicoline be taken with caffeine?

Yes. No interaction between citicoline and caffeine has been reported, and the one placebo-controlled trial that combined them did so on purpose: 60 healthy adults drank a beverage containing citicoline and caffeine and showed better sustained attention, faster reaction times, fewer go/no-go errors, and larger P450 brain potentials than placebo (Bruce 2014). The catch is that the trial cannot say how much of that was the caffeine. Treat the two as separate decisions. Citicoline is not a stimulant and did not change blood pressure over 12 weeks in the memory trial (Nakazaki 2021); caffeine raises heart rate and blood pressure, so count it on its own terms. A morning dose of citicoline alongside coffee is how most people in the attention trials would have taken it anyway.

What should a citicoline label show?

Four things. The compound, written as citicoline or CDP-choline, not "choline"; the salt, inner salt (Cognizin states this) or citicoline sodium, so you can do the 479 mg math; the milligrams per serving, outside any proprietary blend, between 250 and 500 mg to match the healthy-adult trials; and a lot number you can match to a certificate of analysis showing identity, potency, and contaminant screens. A short "other ingredients" line (capsule, rice extract, silica) is a good sign. "Clinically studied" should name the study. A label that says 1,000 mg "clinical strength" is quoting prescription doses that produced a smaller brain signal than 500 mg (Silveri 2008). The buying section above walks through each item.

Related Mt. Angel Vitamins Products

• Recommended for Citicoline (CDP-Choline)

  Brain Advantage Nootropic For Memory, Focus And Mental Performance

  $61.99

  Shop Now