Validity and reproducibility of a food frequency questionnaire as a measure of recent dietary intake in young adults.

Abstract

This research assessed the relative validity and reproducibility of the Dietary Questionnaire for Epidemiological Studies (DQESV2) over one month in young adults, given the lack of concise and convenient instruments for assessing recent dietary intake in this population. Participants were recruited from a large Australian university (N = 102; 35% male; age 18-34 years; body mass index 16-37 kg/m(2)). Five one-day weighed food records (WFR) were administered over one month followed by the DQESV2. Estimates for nutrients (energy, protein, total fat, saturated fat, carbohydrate, sugars, dietary fibre, and alcohol) and fruit and vegetable servings were compared between methods using correlation coefficients, 95% limits of agreement, and quintile classifications. One week later, a second DQESV2 was completed by n = 77 of the participants to assess reproducibility using intra-class correlations (ICC) and weighted kappa. Comparing methods, all nutrients and fruit and vegetable servings showed significant positive correlations (P<0.05) except protein intake in males; over 60% of participants were within one quintile classification except total fat and dietary fibre intakes in males (55% and 56%, respectively); and differences in nutrient and food intakes between methods were all within +/-20% of the mean WFR values except alcohol intake in females. Between first and second administrations of the DQESV2 all ICC coefficients were positive (P<0.01) and weighted kappa coefficients ranged from 0.54 for fruit servings (including fruit juice) in males to 0.91 for protein intake in females. Over a one month period, the DQESV2 demonstrated good reproducibility for the studied nutrients and for fruit and vegetable servings and provided a valid measure of the studied nutrients, except alcohol in females, and of fruit servings (including fruit juice) in both genders, at the group level in this young adult population.