(P08-007-25) Utility of Two Food Frequency Questionnaires to Assess Usual Intake of Docosahexaenoic Acid (DHA) and Eicosapentaenoic Acid (EPA) Among Toddlers

Objectives: High-quality food frequency questionnaires (FFQ) to estimate usual intake of specific long-chain fatty acids among very young children are lacking. Our objective was to evaluate 2 food frequency questionnaires for assessing DHA and EPA intake among toddlers.

Methods: In a clinical trial of omega-3 fatty acid supplementation of 377 children ages 12-20 months, usual intake of DHA and EPA were assessed at baseline via caregiver completion of a 7-item FFQ that listed high DHA and EPA foods (Kuratko, 2013) and the 111-item Harvard semi-quantitative FFQ (2007) designed for adults (Willett, 1985). Red blood cell (RBC) fatty acids were measured in 205 children by gas chromatography. Pearson correlation coefficients estimated the strength of correlation among estimates of DHA, EPA, and DHA+EPA intake between the FFQs and with RBC fatty acids. Participant characteristics were examined as potential modifiers of the correlations.

Results: Among 372 children with data from both FFQs, estimates of DHA, EPA, and DHA+EPA intake were moderately correlated among FFQs (r ranged 0.62-0.69, all p< .0001). Those correlations were stronger for children of mothers with greater than a high school education and from households with income >$35,000/yr, compared to children of other households. Estimates of DHA and DHA+EPA from both FFQs were weakly correlated with RBC DHA and DHA+EPA (r ranged 0.14-0.15, p ranged 0.03-0.04). Those correlations did not differ by household income for the 7-item FFQ, but for the Willett only children from households with income >$35,000/yr showed correlation. Both FFQs generated estimates of EPA intake that were not correlated with RBC EPA (r ranged 0.07-0.11, p ranged 0.13-0.31).

Conclusions: A 7-item FFQ generated estimates of fatty acid intake similar to those generated by the longer Willett FFQ. Estimates from neither FFQ were strongly correlated with RBC fatty acids. Families with low incomes and caregivers with less education may have had difficulty with the lengthy Willett FFQ, as one possible explanation for differences by participant characteristics.

Funding Sources: National Institutes of Health, Health Resources and Services Administration, March of Dimes, Allen Foundation, Cures Within Reach, and internal support