Background: Food-based clinical trials are vital to advance the scientific evidence for the impact of food on health. These trials reqUire stringent dietary assessment to substantiate effects. We are evaluating the use of a self-administered computerised dietary assessment (DietAdvice) in a current food based weight loss trial. Objective: This cross sectional study aims to compare data from DietAdvice with diet history (DH) and food record (FR) dietary assessments measured at baseline. Materials and Methods: Baseline data for n=71 overweight (23-60 years, BMI 25-37 kg/m2) participants was utilised. Macronutrient data for matched dietary assessments from n=32 participants was obtained for the DH assessment while only n=30 matched FR data sets were available. Pearson's correlations and Goldberg cut-off limits were calculated to determine relationships and levels of underreporting between assessment methods respectively. Results: DietAdvice provided significantly higher reported energy intake (kJ) compared to both the DH and the FR (P < 0.01). There were relatively high correlations (r2 =0.740 and 0.596, respectively) between data from the methods. In the DH 35% (n=25/71) of participants underreported their energy intake whereas only 19% (n=13/69) underreported in the FR and only 16% (n=5/32) while using DietAdvice. Significance: This study suggests that further evaluations of the DietAdvice program in a clinical trial is warranted, particularly in determining the efficiency of food based interventions.. Biomarker validation of data may be of value as DietAdvice consistently prOVided larger reported intakes compared to the DH and FR and the traditional dietary assessment methods displayed high levels of underreporting of energy intake. The results indicate that the DietAdvice and the FR are more comparable than the DH though replacement of the traditional FR is not yet warranted.
Available at: http://works.bepress.com/l_tapsell/52/