(P01-008-23) Relative Validity of Nutrition Assessment Tools for Estimating Infant Carotenoid Intake

Objectives: Carotenoids may support infant neurodevelopment, and biospecimen concentrations may provide biomarkers of fruit, vegetable, and carotenoid intake. However, a lack of validated infant diet assessment tools hinders the study carotenoids in infant nutrition. The study purpose is to compare the validity of infant carotenoid intake assessment methods, relative to a reference method of a 7-day food diary (7dFD) analyzed with NDSR and adjusted for actual milk carotenoid concentrations.

Methods: This is a secondary analysis of data collected from healthy, 4-mo infants and their caregivers over time, who were recruited from Houston, TX. Repeated diet assessments were collected at 4- (n=21), 6- (n=12), and 8- (n=9) mo. Intake was assessed with either FFQ, 24 hour recalls (24HR), or 7dFD analyzed with 1 of 3 nutrient databases, and adjusted to account for milk and formula carotenoid composition. The effect of fewer food diary record days (5, 3, and 1) on estimate correlation with 7 food diary days was tested. Spearman’s bivariate correlation test examined the associations between the test and reference method.

Results: Across age groups, the carotenoid intake estimated from 7dFD methods analyzed with different combinations of nutrient databases and milk carotenoid adjustments, were most strongly correlated with the reference, followed by 24HR analyzed with ASA24. At 4 mo, adjustment for actual milk carotenoids, independent of assessment tool and database, conferred the greatest correlations with the reference (rho=0.62-0.97). At 6 and 8 mo, the 7dFD analyzed with NDSR, regardless of milk carotenoid adjustment, was most strongly correlated with the reference (rho=0.90-1.00). A minimum of 3 food diary days resulted in high correlation (rho=0.79 to 1) with 7dFD estimates for all ages.

Conclusions: The validity of carotenoid intake estimates obtained from different methods varied by age/feeding stage, with adjustment for actual milk carotenoids being influential in exclusively milk fed infants, and nutrient database and assessment tool being more important in infants receiving complementary foods. Valid tools for infant carotenoid assessment can support future research on carotenoid bioactivity and use of carotenoid measures as biomarkers of dietary intake.

Funding Sources: USDA/ARS, Texas Children’s Hospital/Baylor College of Medicine Pediatric Pilot Award