(P17-019-20) Comparison of Four Methods to Estimate the Prevalence of Dietary Supplement Use Among U.S. Children

Objectives: Little is known whether the prevalence of dietary supplement (DS) use among children varies by assessment method as it does among adults. We assessed the prevalence of use of different DS product types among U.S. children using four different combinations of assessment methods constructed from two instruments (i.e., frequency-based questionnaire and 24-hour dietary recall (24HR)).

Methods: DS use information is collected in the National Health and Nutrition Examination Survey (NHANES) via a 30-day frequency questionnaire in conjunction with an in-home inventory (Dietary Supplement and Prescription Medicine, DSMQ) and during up to two 24HR. NHANES 2013-2016 data were used to estimate the prevalence of use of any DS and selected types of DS products among 8,369 children (≤18y) using four methods constructed from DSMQ and 24HR: 1) DSMQ; 2) 24HR day 1; 3) 24HR day 1 and day 2; and 4) DSMQ and ≥1 24HR.

Results: About one-third (34.4%) of children took at least one DS in the previous 30 days according to the DSMQ. This decreased to 21.0% when data from only the 24HR day 1 were used, increased to 26.2% when data from both the 24HR day 1 and day 2 were used, and increased further to 37.3% when data from the DSMQ and ≥1 24HR were combined. Similar patterns were found when stratified by age (< 1y, 1-3y, 4-8y, 9-13y, and 14-18y). Prevalence of use of specific product types also differed by method of assessment, but the percentages of those using specific products among DS users were similar across methods. The most commonly consumed products were multivitamin-minerals (73.6% to 75.5%, depending on the method). Many of these products contained over 100% of the Daily Value for some vitamins and minerals.

Conclusions: Prevalence estimates of DS use differed by method among U.S. children, which may be related to the length of the reference period of assessment. Accurately capturing DS users is critical because nutrient intakes from DS can contribute substantially total nutrient intakes. Our findings suggest that using data from both DSMQ and 24HR captures more DS users compared to a single instrument alone.

Funding Sources: This work was supported through the National Institutes of Health.