Abstract
Accurate dietary assessment in adolescent athletes is essential for optimizing athletic performance, supporting growth, and preventing energy deficiency syndromes. However, conventional self-reported methods suffer from substantial recall bias and participant burden. This study evaluated the comparative validity of a digital image-based 24-hour dietary recall mobile application ("NutriSnap-24") against conventional objective biomarkers and traditional paper-based food records in adolescent athletes. A cohort of 85 competitive athletes aged 14–18 years completed a 7-day protocol incorporating the NutriSnap-24 application, standard 3-day food records, doubly labeled water (DLW) measurements for total energy expenditure, 24-hour urinary nitrogen collection for protein intake, and serum carotenoid analyses for fruit and vegetable consumption. NutriSnap-24 demonstrated a significantly stronger correlation with DLW-derived energy expenditure (r = 0.74, p < 0.001) than paper records (r = 0.48, p < 0.001), reducing mean energy underreporting from 21.2% to 8.4%. Protein intake derived from the application showed high agreement with urinary nitrogen estimates (r = 0.78, p < 0.001), while Bland–Altman analysis confirmed narrower limits of agreement for NutriSnap-24 compared to conventional records. Serum carotenoid levels correlated moderately with app-logged fruit and vegetable intake (r = 0.62, p < 0.001). These findings demonstrate that image-assisted mobile dietary recall provides a highly valid, low-burden alternative for dietary monitoring in pediatric sports nutrition and public health settings.
