Choline Forms: Alpha-GPC, CDP-Choline, and Bitartrate — Why the Label Number Lies

Pick up three bottles of "choline" off the same shelf. The first is a 500 mg capsule of choline bitartrate. The second is a 300 mg capsule of alpha-GPC. The third is a 250 mg capsule of CDP-choline. The front of all three bottles shows a brain icon and the word "choline." The supplement facts panel on the back lists the milligrams of the salt the capsule actually contains, not the milligrams of free choline that salt yields, not whether that choline crosses the blood-brain barrier in any meaningful quantity, and not whether the carrier molecule attached to the choline has pharmacological activity of its own.

The three bottles are not interchangeable. They are not even close to interchangeable. They deliver different amounts of free choline per capsule, different blood-brain barrier crossing profiles, different downstream metabolites, different effects on acetylcholine synthesis, and different evidence bases for the cognitive endpoints the bottles are implicitly marketed against. The consumer who buys the cheapest of the three because "choline is choline" is making a category error the labels are doing little to prevent.

This post is the choline-form math: the choline-yield percentage for each form, the bioavailability picture, the human-trial dose ranges the published literature actually used, the situations where each form is appropriate, and the label patterns that distinguish a usable cognitive-choline product from a constipation-and-fishy-breath wallet drain.

The three forms in the aisle (and a fourth)

Choline bitartrate. Choline cation bound to tartaric acid as a salt. The cheapest and most common form in mass-market "choline" supplements, in multivitamins that list "choline," and in some pre-workout blends that name-check choline on the label. Choline bitartrate is about 41% choline by weight — a 500 mg capsule of choline bitartrate delivers roughly 205 mg of free choline.

Alpha-GPC (L-alpha-glycerylphosphorylcholine). Choline esterified to glycerophosphate. A more bioavailable and more brain-targeted form. Alpha-GPC is about 40% choline by weight — a 300 mg capsule of alpha-GPC delivers about 120 mg of free choline. The glycerophosphate carrier is itself an intermediate in phosphatidylcholine synthesis, so the entire molecule is doing useful biological work, not just the choline cation.

CDP-choline (citicoline, cytidine 5'-diphosphocholine). Choline esterified to cytidine diphosphate. A precursor to phosphatidylcholine via the Kennedy pathway. CDP-choline is about 18% choline by weight — a 250 mg capsule of CDP-choline delivers about 45 mg of free choline. The remainder of the molecule, the cytidine fragment, is metabolized to uridine and contributes independent effects on membrane phospholipid synthesis and dopaminergic signaling.

Phosphatidylcholine (lecithin and lecithin-derived phosphatidylcholine). Choline esterified to a phosphate-glycerol-diacyl backbone. The form that choline takes inside cell membranes. Soy lecithin and sunflower lecithin both contain phosphatidylcholine. Lecithin as sold is typically 10-30% phosphatidylcholine, and pure phosphatidylcholine extracts are roughly 13% choline by weight. Useful as a dietary choline source and as an emulsifier; not the form that the cognitive-supplement trials typically use, and not what most "choline" capsules contain.

The label-relevant math, condensed:

| Form | Choline yield by weight | 500 mg capsule yields | |---|---|---| | Choline bitartrate | ~41% | ~205 mg free choline | | Alpha-GPC | ~40% | ~200 mg free choline | | CDP-choline | ~18% | ~90 mg free choline | | Phosphatidylcholine | ~13% | ~65 mg free choline |

But that table is not the whole story, because the form determines where the choline ends up — and that matters more than the milligram count on the front of the bottle.

The blood-brain barrier problem

Choline crosses the blood-brain barrier via a saturable transporter system. Plasma choline concentrations after an oral choline bitartrate dose rise, but the brain-extraction step is rate-limited. The increment in brain choline from a 500 mg dose of choline bitartrate is modest, and the increment in central acetylcholine synthesis from that brain-choline rise is smaller still. The downstream effect on cognitive performance is the question the marketing implies — and the data supporting that link, for choline bitartrate specifically, is weak.

Alpha-GPC is the form with the cleanest cognitive trial signal. Trials in healthy young adults at doses of 300-600 mg/day, in older adults at 1,200 mg/day, and in athletes at 600 mg/day pre-exercise have demonstrated effects on reaction time, attention, growth-hormone release, and power output. The mechanism is partly the more efficient brain uptake of the glycerophosphocholine fragment, partly the contribution of glycerophosphate as a direct phosphatidylcholine substrate, and partly the better-characterized clearance of the molecule into the central nervous system. The strongest case for "I want a choline product for cognitive purposes" is alpha-GPC, dosed at 300-600 mg/day, taken in the morning or pre-task.

CDP-choline has its own trial signal, distinct from alpha-GPC. The cytidine fragment metabolizes to uridine, which independently increases membrane phospholipid synthesis (phosphatidylcholine, phosphatidylethanolamine) and modulates dopaminergic signaling. Trials on CDP-choline at 250-1,000 mg/day have shown effects on attention, processing speed, and recovery from ischemic neurological injury. The cognitive-enhancement case for CDP-choline rests on the dual mechanism (choline + uridine), and the dose range that produces measurable effects starts at 250 mg/day and runs as high as 2,000 mg/day in some clinical neurological indications.

Choline bitartrate produces measurable plasma-choline elevations and contributes to dietary choline-adequacy. It is appropriate as a baseline-dietary-choline supplement (for vegans, for pregnant women, for choline-deficient diets), and it is cheaper per gram of free choline than the cognitive forms. It is not the form to reach for if the goal is cognitive performance. The trial evidence supporting choline bitartrate for cognition specifically is sparse compared to the alpha-GPC and CDP-choline literature, and the brain-extraction step is the rate-limiting bottleneck the cheaper form does not address.

The label trick is that the bottle saying "choline 500 mg" is technically not wrong — there is 500 mg of choline bitartrate (or 500 mg of alpha-GPC, or 500 mg of CDP-choline) in the capsule. What the label does not tell the consumer is that 500 mg of bitartrate yields ~205 mg of free choline that does an okay job of raising plasma choline and a mediocre job of raising brain choline, while 500 mg of alpha-GPC yields ~200 mg of free choline that crosses the blood-brain barrier reliably and converts to phosphatidylcholine inside the brain. The two are not the same product even when the milligrams match.

The trial doses the strongest evidence is built on

A quick survey:

Alpha-GPC, cognitive endpoints (healthy adults). 300-600 mg/day, single or split dose. Reaction time and attention improve at the 300-600 mg/day range in trials lasting 2-6 weeks. The higher end of the dose range (600 mg) shows up in pre-exercise trials on power output and growth-hormone release.

Alpha-GPC, cognitive endpoints (older adults with mild cognitive impairment). 1,200 mg/day, split into 2-3 doses. The classic Italian clinical-dementia literature from the 1990s and 2000s used this dose and showed measurable improvements on standardized cognitive scales over 90-180 days.

CDP-choline, healthy young adults. 250-500 mg/day, single dose. Effects on attention and processing speed in 4-6 week trials.

CDP-choline, post-stroke and post-traumatic brain injury (clinical). 500-2,000 mg/day, sometimes IV in acute hospital settings, oral 500-1,000 mg/day in outpatient recovery trials.

Choline bitartrate, dietary-adequacy supplementation. 500-1,000 mg/day. The reference dietary intake for choline is 425-550 mg/day for adults (RDA-equivalent); the bitartrate form is appropriate at this dose for filling a dietary gap. The cognitive-endpoint trials at this dose-range are weaker than the alpha-GPC and CDP-choline literature.

Phosphatidylcholine (lecithin-derived). 1.5-3 g/day in studies on liver and on membrane-phospholipid endpoints. Not a typical cognitive-enhancement vehicle.

If the bottle says "choline 500 mg, take one capsule daily," and the form is choline bitartrate, the consumer is dosing at the floor of dietary-adequacy supplementation and below the threshold of any cognitive-endpoint trial. If the form is alpha-GPC, the consumer is squarely in the cognitive-trial dose range. If the form is CDP-choline, the consumer is at the upper end of the healthy-adult cognitive-trial range. Three "choline 500 mg" bottles can produce three completely different effects, none of which the front-of-bottle marketing accurately distinguishes.

Side effects, fishy-breath, and the TMAO question

Choline metabolism produces trimethylamine (TMA) as a downstream metabolite, particularly at higher doses and particularly in individuals with high gut-flora trimethylamine-lyase activity. TMA has a fishy odor and produces the characteristic "fishy-breath" side effect of high-dose choline bitartrate supplementation — a real and well-documented effect, dose-dependent and form-dependent. The bitartrate form produces it most reliably, followed by free choline; alpha-GPC and CDP-choline produce it at lower rates because the slower release and the alternative metabolic fates of the carrier fragments shift the metabolic flux away from TMA.

The downstream metabolite of TMA, after hepatic oxidation, is trimethylamine-N-oxide (TMAO). Elevated plasma TMAO has been associated with cardiovascular risk in some observational studies, though the causal mechanism and the dose-response in supplement-using populations remain debated. The relevant point for label-reading: anyone using high-dose choline (above ~1,000 mg/day of any form, sustained) should be aware of the TMAO question and should consider lower doses, alternative forms (alpha-GPC and CDP-choline produce less TMA), or food-based choline (egg yolks, organ meats, lecithin) as primary sources.

Other dose-dependent side effects of high-dose choline supplementation include GI distress, hypotension at very high doses (>3 g/day), sweating, and salivation — all of which are mediated by the parasympathetic effects of acetylcholine elevation and which are dose-related across forms.

The label patterns to scan

Pattern 1: "Choline (as choline bitartrate) 500 mg — take 1 capsule daily." The default mass-market form. Useful as a dietary-adequacy supplement. Below the cognitive-endpoint dose range for the form. If the bottle is pitched on the front as a "brain support" or "focus" product, the form does not match the marketing claim well. Look at the supplement-facts panel and check that the form is what is implied.

Pattern 2: "Alpha-GPC 300 mg — take 1-2 capsules daily." Legitimately dosed for the cognitive use case. Free-choline yield per capsule is about 120 mg; two capsules per day delivers ~240 mg of free choline through the bioavailable form, comfortably inside the 300-600 mg/day cognitive-trial range when reported as the alpha-GPC dose. The bottle is honest about form, honest about dose, and matches the trial literature.

Pattern 3: "Citicoline (CDP-choline) 250 mg — take 1-2 capsules daily." The Cognizin trademarked CDP-choline ingredient is the most studied form for the cognitive use case at this dose range. The bottle delivers 250-500 mg of CDP-choline per day, which matches the healthy-adult trial dose. Free-choline yield per 250 mg capsule is about 45 mg, but the cytidine/uridine fragment is doing pharmacological work independent of the choline mass, so the milligram-of-free-choline math undersells the value of the form.

Pattern 4: "Choline 200 mg" (form unspecified in the front panel, supplement-facts panel shows "choline bitartrate"). The deceptive pattern. Front of the bottle lists "choline" without form. Back of the bottle reveals the form is bitartrate, often at a sub-dietary-adequacy dose. The bottle is functionally a placeholder ingredient in a multivitamin or a multi-ingredient blend.

Pattern 5: "Choline complex — alpha-GPC + CDP-choline + choline bitartrate" (proprietary blend). The hide-the-dose pattern. The blend's total mass is stated, but the breakdown by form is hidden behind a proprietary-blend wall. Per the proprietary-blends post, this is the format where underdosing of the expensive forms (alpha-GPC, CDP-choline) is most likely. A bottle that uses this pattern is almost always loading most of the blend with the cheap bitartrate and putting a sprinkle of the expensive forms in for the label.

The pattern Scythene considers the floor for a cognitive-choline product worth selling: a single, named form (alpha-GPC or CDP-choline), dosed at a per-serving milligram that matches the trial literature, with no proprietary-blend obfuscation.

What about timing and stacking

Timing. Alpha-GPC has a relatively short plasma half-life (about 4-5 hours) and the cognitive effects peak 30-90 minutes after dosing. Morning dosing or pre-cognitive-task dosing is appropriate. CDP-choline has a longer half-life because the cytidine/uridine arm continues to contribute downstream; once-daily morning dosing works fine for cognitive use.

With food or without. Both alpha-GPC and CDP-choline are absorbed reasonably well in either state. Bitartrate is slightly better tolerated with food because the parasympathetic side-effect risk (salivation, sweating, GI distress) drops when the dose is paired with a meal.

Stacking with caffeine. Alpha-GPC and caffeine are commonly stacked in pre-workout and cognitive products. The stack is rational — caffeine's acetylcholine-receptor effects and alpha-GPC's acetylcholine-precursor effects are complementary — though the additive effect on subjective stimulation can produce overstimulation at higher doses of both.

Stacking with racetams. Both alpha-GPC and CDP-choline are commonly stacked with piracetam and the other -racetams in nootropic protocols. The choline arm is necessary to prevent the headache that racetams produce when central acetylcholine demand outstrips choline availability. The trial evidence for this specific stack is sparser than the marketing implies; the rationale is mechanistic rather than RCT-backed.

Drug interactions. Choline at high doses can interact with anticholinergic medications, with cholinergic dementia medications (donepezil, rivastigmine, galantamine), and with some antipsychotics. Anyone on prescription medications affecting cholinergic signaling should not stack high-dose choline without consulting the prescribing physician.

The bottom line

Three bottles, all labeled "choline 500 mg," can deliver three completely different products: a dietary-adequacy supplement, a cognitively-active brain-penetrant precursor, or a dual-mechanism cytidine-uridine product. The front-of-bottle "choline" claim does not distinguish them. The supplement-facts panel does, if the consumer knows which form to look for.

For dietary-adequacy and general health: choline bitartrate at 500-1,000 mg/day is appropriate and inexpensive.

For cognitive performance, attention, reaction time, and pre-task focus: alpha-GPC at 300-600 mg/day is the strongest-evidence form. The bottle that names alpha-GPC specifically on the supplement-facts panel and doses at this range is doing what the label implies.

For attention and processing speed with the dual mechanism: CDP-choline (citicoline) at 250-500 mg/day is the strongest-evidence form. The Cognizin trademarked ingredient is the most-trialed specific source.

Any bottle that lists "choline" without specifying the form on the supplement-facts panel, that hides the form inside a proprietary blend, or that doses below the form-appropriate threshold is doing label work, not pharmacology work.

The label tells you everything if you know to read past the front-of-bottle word "choline" and look at the form, the dose, and the implied use case. Most of the cognitive-choline products on the shelf are dosed correctly for the form they contain. Most of the multi-ingredient "brain support" blends are not. Read the panel, check the form, check the dose against the trial range, and the choice gets a lot less ambiguous.