Determining the accuracy of reporting using repeat automated dietary assessment methods

Automated assessment technologies are becoming increasingly popular in the field of nutrition. In Australia, the DietAdvice website was developed for self-administered reporting of dietary intakes. Numerous studies report on the accuracy of reporting in a face-to-face setting and only few studies have addressed accuracy for computerised reporting. The aim of this study is to determine the accuracy of reporting using the DietAdvice website. Twenty-nine subjects were recruited. Diets for main meals were modelled for a one week period based on food preferences and 70% estimated energy requirements. Subjects were provided with one week of food for main meals, kept a 7-day food record during this period and used DietAdvice to report intake during the 7-days at day one, three, seven, 14, and 28 following the week of provided food intake. Twenty-one subjects completed five visits, three completed four visits and five completed three or less visits. Subjects were aged 19-68 years with an average BMI of 25.9ÃÂñ6.6kg/mÃÂò. Total energy at day one was reported at 6874ÃÂñ3332kJ and 6870ÃÂñ3327kJ at day 28. Despite these small variations, the actual foods reported for each meal varied between visits suggesting a slight decrease in accuracy with increased time between consumption and recall.