Seven-Year Follow-up of Offspring of Women Taking Prenatal DHA in an RCT

Jacqueline F. Gould, BSoSc, BHSc, PhD; Karli Treyvaud, BSc, DPsych; Lisa N. Yelland, BMa, CompSc, PhD; et al.
March 21, 2017

The sale of prenatal supplements with docosahexaenoic acid (DHA) continues to increase, despite little evidence of benefit to offspring neurodevelopment.1 We randomized pregnant women to receive 800 mg of DHA daily or a placebo during the last half of pregnancy and found no group differences in cognitive, language, and motor development at 18 months of age, although secondary analyses revealed less cognitive delay but lower language scores in the DHA group.2 At 4 years of age there was no benefit of DHA supplementation in general intelligence, language, and executive functioning, and a possible negative effect on parent-rated behavior and executive functioning.3 This follow-up was designed to evaluate the effect of prenatal DHA on intelligence quotient (IQ) at 7 years, the earliest age at which adult performance can be indicated.

Methods

Methodology for the trial2 and the 7-year follow-up4 have been published. Written informed consent was obtained and approval granted by the local institutional ethics review boards. Children selected for neurodevelopmental assessment who were alive or withdrawn were invited to attend an appointment at age 7 years (corrected age for children born preterm). Assessments were administered by trained psychologists blinded to group allocation (June 26, 2013, to September 12, 2015; for a description of the assessments, see the eAppendix in the Supplement). The primary outcome was full-scale IQ from the Wechsler Abbreviated Scale of Intelligence, Second Edition (mean, 100 [SD, 15]; delayed performance, full-scale IQ score <85). Language, academic abilities, and core components of executive functioning (memory, inhibition, and mental flexibility) were assessed as secondary outcomes (see Table for details).4 Parents completed standardized questionnaires about their child’s behavior and executive functioning and provided information on children’s DHA intake and neurodevelopmental diagnoses.

Analyses were performed using SAS (SAS Institute), version 9.3, and Stata: Release 13 (StataCorp) on an intention-to-treat basis. All families consenting to the follow-up were included in analyses using multiple imputation to handle missing data. Continuous data were analyzed using linear regressions and binary data using log binomial regression. A 2-sided P value less than .05 was considered significant. No adjustment for multiple comparisons was done. Therefore, secondary outcomes should be interpreted with caution. Analyses were adjusted for center, parity, child’s sex, mother’s secondary and further education, and smoking status at baseline.

Results

Of those eligible, 543 children (85.1%) participated in the 7-year follow-up (DHA group: 73.8%, 259 of 351 invited; control group: 75.7%, 284 of 375 invited) compared with 96% at 18 months and 89% at 4 years (eFigure in the Supplement). Baseline data did not significantly differ between participants and nonparticipants at 7 years. Mean IQ of the DHA and control groups did not differ (98.31 for the DHA group vs 97.32 for the control group; adjusted mean difference [AMD], 1.30 [95% CI, −0.47 to 3.08], P = .15) (Table). Performance on direct measures of language, academic functioning, and executive functioning did not significantly differ between groups, with the exception of slightly higher perceptual reasoning scores in the DHA group. Parents in the DHA group reported more behavior problems (total difficulties score, 9.71 for the DHA group vs 8.63 for the control group; AMD, 1.09 [95% CI, 0.18 to 2.00], P = .02) and executive dysfunction (Global Executive composite, 54.89 for the DHA group vs 52.54 for the control group; AMD, 2.38 [95% CI, 0.67 to 4.08], P = .01) in their children. Diagnosis of neurodevelopmental disorders and child DHA intake did not significantly differ between groups.

Discussion

This randomized clinical trial provides strong evidence for the lack of benefit of prenatal DHA supplementation on IQ at 7 years and cognition at 18 months2 and 4 years,3 despite higher numbers of preterm children in the control group. Direct assessments consistently demonstrated no significant differences in language,2,3 academic abilities, or executive functioning.3 Although perceptual reasoning was slightly higher in the DHA group, parent-reported behavioral problems and executive dysfunction were worse with prenatal DHA supplementation. Differences found in secondary outcomes may be chance findings due to the high number of comparisons made. The small but consistent negative effects of prenatal DHA on behavior and executive functioning at 7 and 4 years3 may reflect true effects, although effect sizes were small and neurodevelopmental diagnoses did not differ between groups. Differences are unlikely due to methodological issues as follow-up rates and variables after randomization were balanced between the groups.