Choline in pregnancy: what the 2026 Colorado trial found, and what it does not settle

Choline is a nutrient the body uses to build cell membranes and the messenger chemical acetylcholine. The liver makes some, but not enough, so the rest has to come from food such as eggs, meat, fish, dairy, soya and beans. In a trial published online on 19 October 2026 in the American Journal of Obstetrics and Gynecology, University of Colorado researchers gave 206 pregnant women either phosphatidylcholine, the main form of choline in food, or a placebo from 16 weeks of pregnancy until birth. Children of mothers who took the supplement had fewer social withdrawal problems and scored about 8 points higher on an IQ test at age 4. The result has been widely shared since it appeared in October as a reason to add choline to a prenatal routine.

Short answer: this is the largest randomised trial of prenatal choline so far, and it deserves attention. It is not enough on its own to justify ten large capsules a day. One of its two main outcomes, attention problems, showed no significant difference. The IQ result was a secondary outcome measured in 103 children, a third of the women were lost before the 4-year assessments, and parents rated their children's behaviour. An earlier 140-woman trial at a lower choline dose found no difference in infant development. Choline from food is worth a look in pregnancy. A high-dose supplement is a decision to make with a midwife, GP or obstetrician, not on the strength of one study.

What the Colorado trial did

Women aged 19 to 44 were enrolled at 12 to 15 weeks of pregnancy at University of Colorado prenatal clinics and partner community hospitals between 2017 and 2021. After a one-week placebo run-in to check they could take the capsules, they were randomised at 16 weeks to five 900 mg capsules of phosphatidylcholine twice a day (9,000 mg a day) or identical placebos until delivery. Phosphatidylcholine is about 14% choline by weight. The authors state that the 9,000 mg daily dose supplied 1,285 mg of choline on top of the diet. Parents rated children's behaviour from 18 months to 4 years using a standard checklist, and children took a preschool IQ test (WPPSI-IV) at ages 3 and 4.

• Social withdrawal (a main outcome): 4% of children in the supplement group had elevated withdrawal problems, against 24% in the placebo group, based on parent ratings of 142 children.
• Attention problems (the other main outcome): no significant difference.
• IQ at age 4 (a secondary outcome): average Full Scale IQ was 106.0 in the supplement group (54 children) and 98.2 in the placebo group (49 children), a difference of 7.78 points (95% confidence interval 1.88 to 13.7).
• Exploratory findings: 1 of 71 children in the supplement group and 10 of 71 in the placebo group had checklist scores in the range that calls for an autism assessment, with language difficulties. The analysis was exploratory, the confidence interval was very wide, and a checklist cannot diagnose autism.
• Safety: the authors report that nausea, vomiting, dizziness and high blood pressure were not more common with the supplement, and that no serious side effect in the mothers or children was attributed to it. Birth measurements were similar, apart from greater length in boys whose mothers took the supplement.

The work was supported by the US National Institutes of Health, the Institute for Children's Mental Health and The Anschutz Foundation. The authors conclude that prenatal phosphatidylcholine provides safe, effective support of fetal brain development, and that a comparison of doses across trials points to 6,300 mg of phosphatidylcholine, or about 900 mg of choline, a day, starting as early as possible in pregnancy. That conclusion goes further than a single trial can carry.

Why the result needs caution

The authors list the main limitations themselves: no significant effect on attention problems, parent-rated social behaviour, and about one-third of women lost between enrolment and the 4-year assessments. Losing that many families can bias a result if the ones who stayed differ between groups. The IQ difference was a secondary outcome, and the autism-range and high-IQ comparisons were small exploratory counts with wide confidence intervals. During the COVID-19 pandemic, 20 women, 10 in each group, also received open-label phosphatidylcholine while staying in their assigned group. When they were left out of the analysis, the IQ difference was similar, but its confidence interval included zero.

The earlier evidence is mixed. The same Colorado group ran a 100-woman trial that found more infants with a brain-response marker of sensory filtering at 5 weeks of age (76% against 43%), but no difference at 13 weeks; at 40 months, parents of 23 children in the choline group reported fewer attention and social withdrawal problems than parents of 26 children in the placebo group. A separate 2012 trial in North Carolina gave 140 women phosphatidylcholine supplying 750 mg of choline a day (six capsules, about 5,000 mg of phosphatidylcholine) from 18 weeks of pregnancy to 90 days after birth and found no difference in infant memory, language or overall development at 10 and 12 months; those women already ate about 360 mg of choline a day. A 2018 Cornell feeding study of 26 women in the third trimester found faster information processing in infants whose mothers ate 930 mg rather than 480 mg of choline a day; it was funded in part by the Egg Nutrition Center and The Beef Checkoff.

How much choline, and how much is too much

Official intakes are far below the trial dose. The US adequate intake in pregnancy is 450 mg a day, and the European Food Safety Authority set 480 mg a day. The NIH Office of Dietary Supplements says about 90% to 95% of pregnant women in the US eat less than the adequate intake, that prenatal supplements typically contain little if any choline, and that choline supplements usually contain 10 mg to 250 mg. A large hard-boiled egg provides about 147 mg.

The US tolerable upper intake level for choline from food and supplements is 3,500 mg a day for adults, including in pregnancy, and 3,000 mg for those aged 14 to 18. High intakes are linked to a fishy body odour, vomiting, heavy sweating and salivation, low blood pressure and liver toxicity. Choline can raise production of TMAO, a compound linked to cardiovascular risk in adults. In this trial, phosphatidylcholine did not significantly raise TMAO. Beef liver is rich in choline, but NHS advice is to avoid liver during pregnancy because of its vitamin A content.

Our view

• This trial makes prenatal choline one of the more interesting pregnancy nutrition questions, but one trial with heavy loss to follow-up and a null primary outcome does not settle it. A second large trial by a different research group would.
• Food is the low-risk place to start: eggs, fish, meat, dairy, soya and beans all contribute. If you are vegetarian or vegan, choline intake is worth raising with a dietitian.
• Do not start a high-dose phosphatidylcholine regime without talking to your midwife, GP or obstetrician, and keep total choline below the upper limit of 3,500 mg a day for adults, or 3,000 mg a day at ages 14 to 18.
• Choline does not replace folic acid or vitamin D. Those remain part of routine pregnancy supplement advice, and a midwife, GP or obstetrician may advise other nutrients for you.
• The autism-range finding is exploratory. Nothing in this trial shows that choline prevents autism.

South African rules

In South Africa, health supplements are regulated by SAHPRA as complementary medicines under the Medicines and Related Substances Act 101 of 1965, and SAHPRA publishes guidance on the safety and efficacy of health supplements. A label claim that a prenatal product boosts a child's IQ or prevents a developmental condition is not supported by the evidence above. Ask your midwife, doctor or pharmacist before adding a high-dose supplement in pregnancy, and report suspected side effects to SAHPRA.

FAQ

Should I take a choline supplement in pregnancy?

NHS advice does not include a routine choline supplement. In the United States, there is no recommendation that every pregnant woman take a separate choline supplement. The Dietary Guidelines for Americans 2020-2025 say pregnant women should eat choline-containing foods, and that women who eat a vegetarian or vegan diet should ask a clinician whether a supplement is needed. The American College of Obstetricians and Gynecologists says to get 450 mg of choline a day and notes that most prenatal vitamins do not contain it. The Colorado trial supports a closer look, but the dose was ten large capsules a day and the result has not been repeated. Discuss it with your midwife, GP or obstetrician.

Does my prenatal vitamin contain choline?

Check the label. The NIH Office of Dietary Supplements says prenatal supplements typically contain little if any choline, and that choline supplements usually provide 10 mg to 250 mg, far below the 1,285 mg of choline in the trial dose.

Does choline in pregnancy prevent autism?

This trial does not show that choline prevents autism. The autism-range result was an exploratory analysis of checklist scores in a small number of children, the confidence interval was very wide, and a checklist cannot diagnose autism.

Which foods are high in choline?

Eggs, meat, poultry, fish and dairy are the richest everyday sources. Soya beans, kidney beans, quinoa, broccoli and Brussels sprouts also contribute.