DHA and arachidonic acid linked to modest attention gains in select preterm children

A six-month course of docosahexaenoic acid (DHA) and arachidonic acid (AA) supplementation in toddlerhood was associated with modestly better attention and inhibitory control in some children born preterm when they were assessed at 8.5 to 10.5 years of age. The long-term follow-up of the Omega Tots randomized placebo-controlled trial found no difference between the supplemented and placebo groups in the likelihood of attention-deficit/hyperactivity disorder (ADHD) or autism spectrum disorder (ASD) diagnoses documented in medical records. The original Omega Tots trial enrolled children born before 35 weeks of gestation when they were 10 to 16 months old, corrected for prematurity.

The children were recruited at Nationwide Children’s Hospital in Columbus, Ohio, between 2012 and 2017. In the fully blinded, parallel-group trial, the intervention group received daily dissolvable powder containing 200 mg of DHA from algal oil and 200 mg of AA from fungal oil for 180 days. The placebo group received 400 mg of corn oil daily for the same period.

The supplement was administered during toddlerhood, when children may begin replacing breast milk or formula with solid foods containing less DHA. Researchers conducted the long-term follow-up from 2021 to 2024, when participants had a median age of 9.2 years. At least one of the examined outcomes was available for all 376 children eligible for follow-up, although data completeness varied by measure.

Caregivers of 317 children completed the Child Behavior Checklist, and medical records were available for 356 children. An exploratory analysis examined the 252 children who had results for all outcomes included in the analysis. Mixed linear models compared attention outcomes while accounting for clustering among siblings, twins and triplets, while generalized estimating equations with a log-binomial model compared the likelihood of recorded diagnoses.

The clearest group-level difference appeared on the National Institutes of Health Toolbox Flanker Inhibitory Control and Attention Test, a computerized task in which children focus on a target while ignoring competing visual cues. Children assigned to DHA plus AA scored 3.02 points higher than placebo recipients on the age-corrected Flanker scale, which ranges from 0 to 100 and gives higher scores for better attention and inhibitory control. The 95% confidence interval was 0.01 to 6.04.

The difference was therefore modest and close to the conventional threshold for statistical significance. No significant group difference was found in Flanker response-time variability or in overall caregiver- and teacher-reported attention-problem scores. Exploratory analyses suggested that the Flanker difference was larger among children with lower dietary omega-3 intake at the beginning of the trial.

For children below the median intake of DHA plus eicosapentaenoic acid, the supplemented group scored 4.88 points higher than the placebo group, with a 95% confidence interval of 0.53 to 9.23 and an effect size of 0.38. Among children above the median intake, the estimated difference was 1.06 points, with a 95% confidence interval of minus 3.24 to 5.37. In the analysis restricted to children with complete outcome data, the overall Flanker difference decreased to 2.84 points and was no longer statistically significant, although the estimated effect size remained similar.

Birthweight also appeared to influence the teacher-reported attention results. Among children with a birthweight above 1,500 g, those receiving DHA plus AA had 2.97 points fewer teacher-rated inattention symptoms than placebo recipients. The 95% confidence interval ranged from minus 5.38 to minus 0.55, and the effect size was minus 0.42.

No treatment effect was detected among children born weighing less than 1,500 g; the estimated difference was 0.44 points, with a 95% confidence interval of minus 2.50 to 3.39. These were exploratory subgroup findings. Interaction tests used a screening threshold of p below 0.20 before the researchers examined stratified estimates, making the subgroup results hypothesis-generating rather than definitive.

Medical records documented an ASD diagnosis in 25 of 356 children, or 7.0%, and an ADHD diagnosis in 85 children, or 23.9%. The likelihood of either diagnosis did not differ significantly between the DHA-plus-AA and placebo groups. For ASD, the estimated risk ratio was 0.79 with a 95% confidence interval of 0.37 to 1.68; for ADHD, it was 1.03 with a 95% confidence interval of 0.70 to 1.51.

Birthweight was screened as a possible moderator of ADHD risk, but neither the above-1,500-g nor below-1,500-g stratum showed a significant difference between treatment groups. The complete-outcome analysis produced similarly nonsignificant results for diagnosis outcomes. The study’s design strengthened the comparison between groups.

Random assignment, full blinding and placebo control reduced the likelihood that observed differences reflected expectations about the treatment, while the follow-up combined caregiver ratings, teacher ratings, a performance-based attention test and medical-record diagnoses. Nevertheless, some outcomes were missing, and families with lower household incomes and maternal education levels were less likely to participate in follow-up, although these characteristics did not differ between treatment groups. Other limitations constrain what the results can establish.

The medical records did not reveal how individual ADHD or ASD diagnoses had been made, and diagnostic coding or the absence of specific testing could have introduced error. The analysis also did not directly assess core autistic traits, making subtle changes in autism-related features difficult to detect. In a separate exploratory comparison that was not a test of supplementation, children with an ADHD diagnosis had significantly poorer caregiver-reported attention, teacher-reported attention and Flanker standard scores than children without an ADHD diagnosis; their Flanker response-time variability did not differ significantly, with p equal to 0.06.

Earlier Omega Tots follow-up data had also reported adverse findings in some subgroups, including poorer language outcomes among the lowest-birthweight children and reduced effortful control in several income strata. The long-term result is therefore a qualified signal rather than a general endorsement: six months of DHA plus AA in toddlerhood was associated with small advantages on selected attention measures in some preterm children, but it did not reduce the likelihood of ADHD or ASD diagnoses by age 8.5 to 10.5 years.

Subject of Research: Long-term attention and neurodevelopmental outcomes after DHA and arachidonic acid supplementation in children born preterm

Article Title: Fatty Acid Supplementation in Toddlers Born Preterm and Attention and Autism Outcomes at 8–10 Years: Omega Tots Randomized Controlled Trial Long-Term Follow-Up

Article References: Robinette, L. M., Taylor, H. G., Boone, K. M., Rogers, L. K., Rausch, J., Bartlett, C. W., Klebanoff, M. A., & Keim, S. A. (2026). Fatty Acid Supplementation in Toddlers Born Preterm and Attention and Autism Outcomes at 8–10 Years: Omega Tots Randomized Controlled Trial Long-Term Follow-Up. Journal of Autism and Developmental Disorders. https://doi.org/10.1007/s10803-026-07566-z

Image Credits: AI Generated

DOI: 10.1007/s10803-026-07566-z

Keywords: Omega Tots trial, Preterm children, DHA supplementation, Arachidonic acid, Attention, Inhibitory control, ADHD, Autism spectrum disorder, Randomized controlled trial, Long-term follow-up, Fatty, Acid