Hippuric acid in 24-hour urine collections is a potential biomarker for fruit and vegetable consumption in healthy children and adolescents

An objective noninvasive biomarker for fruit and vegetable (FV) consumption would help to more reliably characterize the relationship between FV intake and health status in observational studies. Because increases in urinary hippuric acid (HA) were observed after consumption of several FV varieties, we aimed to investigate whether 24-h urinary HA may represent a potential biomarker for FV consumption in children and adolescents. The association of FV and juice (FVJ) intake calculated from 3-d weighed dietary records with 24-h urinary HA excretion was analyzed in 240 healthy children and adolescents and compared with associations of the established biomarkers urinary nitrogen (uN) and urinary potassium (uK) with protein and potassium intake, respectively. Spearman correlation coefficients (r) and cross-classifications were calculated for all diet-biomarker associations. Potential confounders for the HA-FVJ association were examined in linear regression models. In children, correlations of HA with FVJ (r = 0.62), uN with protein (r = 0.64), and potassium intake with uK (r = 0.65) were comparable. In adolescents, the HA-FVJ association was weaker (r = 0.41) compared with the biomarkers uN (r = 0.60) and uK (r = 0.58) (all P 85% for all analyzed comparisons except for a 75% classification agreement between HA and FVJ in adolescents. Unadjusted and adjusted linear regression models indicated significant (P < 0.0001) HA-FVJ associations in both age groups. FVJ explained more of the variability in HA excretion in children (R(2) = 0.38) than in adolescents (R(2) = 0.22). Our findings in children showing HA-FVJ associations comparable to those for well-established biomarkers with their respective dietary intakes suggest that HA may represent a useful biomarker for FVJ.