Relative validity and reliability of a novel diet quality assessment tool for athletes: The Athlete Diet Index.

Diet quality indices are a practical, cost-effective method to evaluate dietary patterns, yet few have investigated diet quality in athletes. This study describes the relative validity and reliability of the recently developed Athlete Diet Index (ADI). Participants completed the electronic ADI on two occasions, two weeks apart, followed by a four-day estimated food record (4-dFR). Relative validity was evaluated by directly comparing mean scores of the two administrations (mAdm) against scores derived from 4-dFR using Spearman's rank correlation coefficient and Bland-Altman (B-A) plots. Construct validity was investigated by comparing mAdm scores and 4-dFR derived nutrient intakes using Spearman's coefficient and independent t-tests. Test-retest reliability was assessed using paired t-tests, intra-class correlation coefficients (ICC) and B-A plots. Sixty-eight elite athletes (18.8 ± 4.2 years) from an Australian sporting institute completed the ADI on both occasions. Mean score was 84.1 ± 15.2 (range: 42.5-114.0). The ADI had good reliability (ICC=0.80, 95% CI: 0.69, 0.87; p<0.001), and B-A plots (mean 1.9; LOA -17.8, 21.7) showed no indication of systematic bias (y=4.57-0.03*x) (95% CI: -0.2, 0.1; p=0.70). Relative validity was evaluated in 50 athletes who completed all study phases. Comparison of mAdm scores to 4-dFR derived scores was moderate (rs=0.69; p<0.001) with no systematic bias between methods of measurement (y=6.90-0.04*x) (95% CI: -0.3, 0.2; p=0.73). Higher scores were associated with higher absolute nutrient intake consistent with a healthy dietary pattern. The ADI is a reliable tool with moderate validity, demonstrating its potential for application to investigate the diet quality of athletes.