Choline
What it is
Choline is an essential nutrient that the body can synthesise endogenously to a limited extent, but not in quantities sufficient to meet requirements during normal metabolic function. It was formally recognised as essential by the US Institute of Medicine in 1998, making it one of the more recently classified essential nutrients, which partly explains why it remains underrepresented in public health guidance and absent from many standard prenatal supplements in the UK and Europe.
Structurally, choline is a water-soluble quaternary amine. It serves as a precursor to several physiologically important molecules: acetylcholine (a neurotransmitter involved in memory, muscle control, and autonomic function), phosphatidylcholine (a major structural component of cell membranes and the dominant phospholipid in neural tissue), sphingomyelin (another membrane phospholipid concentrated in the nervous system), and betaine (a methyl donor involved in homocysteine metabolism and one-carbon metabolic pathways shared with folate). Through these roles, choline participates in membrane integrity, neurotransmission, lipid transport, and epigenetic regulation.
The richest dietary sources are eggs (approximately 115 mg per large egg, primarily as phosphatidylcholine in the yolk), liver and organ meats, beef, fish, and dairy products. Choline content in plant foods is considerably lower, with soy, kidney beans, and cruciferous vegetables providing meaningful but smaller amounts. This distribution means that people following plant-based or egg-free diets are at elevated risk of inadequate intake, and the shift toward plant-based eating in many Western populations may be widening the gap between habitual intake and recommended levels.
Endogenous synthesis occurs primarily via the phosphatidylethanolamine N-methyltransferase (PEMT) pathway in the liver, which is upregulated by oestrogen. This is clinically relevant: premenopausal women can compensate to some extent for low dietary intake through enhanced endogenous synthesis, but postmenopausal women lose much of this capacity and are more dependent on dietary or supplemental sources.
What the evidence shows
The evidence base for choline is most developed in pregnancy, where the combination of observational data, depletion-repletion studies, and a small number of RCTs supports ensuring adequate intake during pregnancy, with preliminary RCT evidence for early neurocognitive outcomes in infants and observational evidence linking lower maternal choline status with higher neural tube defect risk. These trials should not be read as independent confirmations of a clinically meaningful and replicated developmental benefit; they are small, mostly short-term studies using surrogate or early outcome measures that point in the same direction. The primary rating of Emerging reflects that consistent directional signal alongside the limitations of trial size and outcome maturity.
Outside pregnancy, the evidence thins considerably. The importance of choline adequacy for liver function is well established from depletion-repletion experiments, particularly in postmenopausal women. Evidence for cognitive benefit in healthy replete adults is observational and inconsistent, with no adequately powered RCTs of choline bitartrate for cognitive outcomes in this population.
Five questions
Does low status cause harm?
In the context of pregnancy, lower maternal choline intake or circulating concentration is associated with increased neural tube defect risk. A systematic review and meta-analysis by Obeid and colleagues (2022, Journal of Nutrition) pooled five case-control studies including 1,131 mothers of infants with neural tube defects and 4,439 controls, finding a pooled odds ratio of 1.36 (95% CI: 1.11 to 1.67) for low maternal choline status and neural tube defect risk. The prediction intervals are wide (0.78 to 2.36), indicating heterogeneity across studies, and this evidence should be treated as associative. It supports the importance of adequate intake but does not prove that supplementation reduces neural tube defect incidence.
In non-pregnant adults, controlled choline depletion studies demonstrate that inadequate intake causes liver dysfunction (elevated liver enzymes and hepatic steatosis) and muscle damage in a proportion of adults, with postmenopausal women and men showing greater susceptibility than premenopausal women, consistent with the oestrogen-PEMT hypothesis. Whether inadequate intake from typical dietary patterns produces clinically meaningful harm in otherwise healthy replete adults is not established.
Does supplementation prevent disease?
In pregnancy, small supplementation trials suggest favourable effects on selected early neurocognitive measures when intake is increased above the adequate intake level. The systematic review by Obeid and colleagues (2022) identified four RCTs in healthy pregnant women, all of which reported favourable directional findings, although outcomes, follow-up timing, and precision varied across studies. The most frequently cited is Caudill et al. (2018, FASEB Journal), in which third-trimester women were randomised to 480 mg or 930 mg of choline per day for 12 weeks; infants born to mothers in the higher-dose group showed faster information processing speed at four, seven, ten, and thirteen months of age. The evaluable sample was small (26 mother-infant pairs), and information processing speed is a surrogate developmental marker rather than a confirmed clinical endpoint. The available evidence does not support the claim that supplementation at doses above the current adequate intake meaningfully prevents neurodevelopmental disorders in the offspring of women already meeting baseline recommendations.
For liver disease prevention in adults, the evidence does not support supplementation as a preventive strategy in people with normal dietary choline intake. The depletion-repletion literature establishes the importance of adequacy rather than providing a rationale for supplementation above adequate intake.
Does it affect biomarkers?
Yes. Supplemental choline reliably raises plasma free choline and betaine levels. It also raises TMAO, a gut-derived metabolite discussed in the safety section. In pregnancy, higher choline intake attenuates the decline in choline-derived plasma biomarkers that occurs during the third trimester when fetal demand is highest. In the context of liver health, choline repletion in deficient individuals reverses elevations in liver enzymes and reduces hepatic fat accumulation on imaging. These are biomarker effects; their translation to clinical benefit depends on whether the biomarker change lies on a causal pathway, which is reasonably established for liver function in deficiency and plausible but not proven for fetal outcomes.
Does it help clinical populations?
The clearest practical application is pregnancy adequacy. Surveys reviewed in recent literature suggest that many pregnant and lactating women fall below adequate intake recommendations, although estimates vary by country, dietary pattern, and assessment method. A 2025 clinical practice review by Derbyshire and colleagues (Nutrients, DOI: 10.3390/nu17091558) concluded that the majority of pregnant and lactating women in most surveyed populations do not meet adequate intake levels, with mean reported intakes ranging from approximately 233 to 383 mg per day against an adequate intake of 450 mg per day for pregnancy and 550 mg for lactation.
In patients with non-alcoholic fatty liver disease, observational data suggest that low choline intake is associated with worse fibrosis severity, particularly in postmenopausal women. A cross-sectional analysis of 664 biopsy-proven NAFLD patients by Guerrerio and colleagues (2012, American Journal of Clinical Nutrition) found that postmenopausal women with deficient choline intake had significantly worse liver fibrosis after adjustment for confounders (p=0.002), while no significant association was found in children, men, or premenopausal women. A small RCT of phosphatidylcholine supplementation (2400 mg per day for 12 weeks, n=79) showed improvements in oxidative stress markers and liver enzymes in NAFLD patients, but the study was small and single-blinded, and the overall evidence is insufficient to recommend supplementation as a clinical NAFLD treatment. Correcting inadequate intake is a more defensible framing than treating established disease.
Does it benefit healthy individuals?
In healthy replete adults, the available evidence has not shown benefit from supplemental choline. Observational data on dietary choline and cognitive function in older adults are inconsistent. Cross-sectional analyses using NHANES data have found no association between total choline intake and cognitive scores in adults aged 60 and over. A 22-year prospective cohort study from the China Health and Nutrition Survey (Huang et al., 2024, Nutrients) reported an association between higher dietary choline intake and preserved cognitive function in 1,887 adults aged 55 to 79, but cohort data cannot establish causation and dietary assessment tools at this scale carry substantial measurement error. No adequately powered RCT of choline bitartrate for cognitive outcomes in cognitively healthy adults has been published. The cognitive evidence for specific forms (alpha-GPC and citicoline) is meaningfully different and is covered in those separate entries.
Individual variation
Pregnancy and lactation represent the highest-risk period for inadequate intake. Requirements increase during pregnancy and lactation, and surveys consistently find that most pregnant women in Western countries do not reach recommended levels through diet alone, partly because many prenatal supplements do not include choline and partly because the richest sources (eggs, liver) are sometimes avoided during pregnancy.
Postmenopausal women have reduced endogenous synthesis capacity relative to premenopausal women due to declining oestrogen and its effect on PEMT activity, increasing dietary dependency. Observational data suggest this population shows greater susceptibility to low-choline-associated liver changes than premenopausal women, men, or children.
People following plant-based or egg-free diets are at substantially elevated risk of inadequate intake, given the concentration of choline in animal foods. This group is frequently not captured in guidance around choline adequacy, and the risk is likely to increase as plant-based eating becomes more prevalent.
Genetic variation in the PEMT gene affects endogenous synthesis capacity and likely modifies individual choline requirements. Common PEMT variants that reduce enzyme activity increase susceptibility to deficiency symptoms on low-choline diets. Routine genetic testing for this purpose is not currently validated for clinical supplementation decisions, but awareness of this variation is relevant when counselling individuals who follow low-animal-food diets.
Sex differences are relevant beyond the pregnancy context. In controlled depletion studies, men and postmenopausal women are significantly more likely to develop organ dysfunction on low-choline diets than premenopausal women, consistent with the oestrogen-PEMT model.
Testing and status assessment
There is no routine clinical test for choline status in standard NHS or private practice panels. Plasma free choline can be measured in research settings, but it does not reliably reflect total body stores or tissue availability.
In practice, assessment is best approached through dietary analysis: estimating intake from a food frequency questionnaire or dietary recall and comparing against the adequate intake values (425 mg per day for non-pregnant women, 450 mg in pregnancy, 550 mg in lactation, 550 mg for men). Given that the richest sources are a small number of foods (eggs, organ meats, fish, meat, dairy), this is feasible with a brief dietary history focused on those sources.
PEMT genetic testing is available through direct-to-consumer genomic platforms but is not validated for clinical decision-making around choline supplementation. Its utility is currently limited to informing heightened vigilance around dietary intake rather than driving specific supplementation decisions.
Safety
Choline is generally well tolerated at doses up to the tolerable upper limit of 3,500 mg per day for adults. Adverse effects at higher doses are primarily gastrointestinal: nausea, diarrhoea, and a characteristic fishy body odour caused by the excretion of trimethylamine, a bacterial metabolite of choline. The odour effect is dose-dependent and more prominent with free choline supplements (choline bitartrate, choline chloride) than with phosphatidylcholine forms.
The more substantive safety question is the TMAO pathway. Gut bacteria metabolise dietary and supplemental choline to trimethylamine, which is absorbed and converted in the liver to TMAO by flavin-containing monooxygenases. This pathway and its association with atherosclerosis in animal models and cardiovascular risk in humans is well characterised in the broader literature on gut microbiome metabolism of choline and related dietary compounds. Elevated plasma TMAO is consistently associated with cardiovascular disease and mortality risk in observational cohort studies, and meta-analyses confirm that association.
The direct human evidence for the supplement-versus-food distinction comes from Wilcox and colleagues (2021, Journal of the American Heart Association), who found in a randomised trial of healthy adults with normal renal function that choline bitartrate supplements significantly raised fasting TMAO levels while egg consumption at equivalent choline doses did not. A crossover pharmacokinetic study comparing four supplement forms (Traussnigg et al., 2022, European Journal of Clinical Nutrition) confirmed that free choline salt forms generate more TMAO than phospholipid-bound sources. These findings are relevant to form selection when supplementing.
Whether the TMAO elevation from choline supplementation at doses relevant to correcting dietary inadequacy translates to cardiovascular harm in clinical outcome trials has not been established. TMAO observational associations are subject to confounding, and the causal mechanism in humans remains incompletely characterised. The signal warrants monitoring but does not establish that choline supplements cause cardiovascular disease.
People with chronic kidney disease warrant separate consideration because TMAO clearance is renally dependent, and studies conducted in participants with normal renal function cannot be assumed to generalise to this population. TMAO concentrations are disproportionately elevated in kidney disease, and the cardiovascular risk association is stronger in that context.
There is no established clinical interaction between choline supplementation and antiplatelet or anticoagulant drugs. Because the mechanistic signal includes platelet reactivity at higher doses, high-dose free choline supplementation should be approached cautiously in people with high thrombotic risk until outcome data are clearer.
Choline supplementation during pregnancy at doses used in published RCTs (up to 930 mg per day on top of dietary intake) has not raised safety signals in published trials, and the tolerable upper limit is well above these doses.
What can reasonably be concluded
Choline is an essential nutrient that is widely under-consumed, particularly among pregnant women, postmenopausal women, and people following plant-based diets. The available evidence supports ensuring adequate intake during pregnancy, where small trials suggest favourable early neurocognitive outcomes in infants and observational data link lower maternal choline status with higher neural tube defect risk. These findings are directionally consistent and biologically plausible, but the trial base is too small and the outcome data too preliminary to support stronger clinical prevention claims.
Outside pregnancy, the evidence for supplementation in healthy replete adults has not been established. The liver health signal in deficiency populations, particularly postmenopausal women, is meaningful but rests primarily on observational data. For cognitively healthy adults, choline bitartrate has not demonstrated benefit in adequately powered RCTs; the cognitive evidence for enhanced-bioavailability forms is covered in the separate citicoline and alpha-GPC entries.
The TMAO safety question warrants monitoring, particularly for people considering high-dose or long-term supplementation with free choline salts, and for people with pre-existing kidney disease or elevated cardiovascular risk.
Where evidence is limited or outcomes are uncertain, conclusions should be treated as provisional and subject to revision as the evidence base develops.