Abstract
BACKGROUND: This study aimed to determine the relative validity of dietary quality index scores estimated using the updated version of the Menzies Remote Short-item Dietary Assessment Tool (MRSDAT), a food frequency questionnaire developed for use with remote Indigenous communities in Australia. METHODS: This validation study compares scores derived from the updated MRSDAT with those derived from 24-h recalls. Participants from nine remote communities across three discrete regions of Australia (Central Australia, North Northern Territory and Cape York) participated in June-September 2021, or May-July 2022. Participants were Indigenous children aged six months to five years, pregnant and/or breastfeeding Indigenous women, and overweight adult Indigenous males and females. The agreement between methods was explored using Lin's concordance correlation coefficient (CCC), Bland-Altman plots, weighted Cohen's kappa, paired samples t-tests and Wilcoxon signed-rank tests. Primary comparisons were completed for the whole sample, and primary and secondary comparisons for only participants with ≥ 2 × 24 h recalls. RESULTS: Ninety-nine participants were recruited. The MRSDAT estimated dietary quality index scores 0.31/100 points higher than 24-h recall estimates for participants with ≥ 2 × 24-h recalls, and 0.18/100 points higher for all participants, with consistent results for adults and children (CCC = Rc 0.58, 95% CI 0.39, 0.72). CONCLUSIONS: Updates to the MRSDAT have resulted in dietary quality index scores highly comparable to those derived from 24-h recalls. The MRSDAT has the practical advantages of speed and convenience for research purposes, and the built-in instant dietary feedback feature has application to service delivery settings given the robust results demonstrated in this validation study.