Relative validity and reliability of a quantitative food frequency questionnaire for adults in Guam

Abstract

Background: Guam is a US territory in the western Pacific with a diverse population that includes understudied ethnic groups such as Chamorros and Filipinos. A food frequency questionnaire (FFQ) to estimate dietary intake was needed to facilitate studies of diet and health among adults living in Guam. Objective: To develop and validate an FFQ to assess dietary intake over a 1-year period among adult Guam residents. Design: A three-part study was conducted: 1) an initial cross-sectional study using 24-h recalls to identify a food and beverage list for the FFQ and resulting in a final FFQ containing 142 food and drink items; 2) to test reliability, 56 different individuals completed the FFQ twice; and 3) to test relative validity, self-administered FFQs and up to 2 days of food record data from an additional 109 individuals were collected, and daily nutrient intake from the two methods was compared. Results: The reliability of the FFQ was very good (ρ range = 0.65-0.75), and the relative validity of the FFQ was good for women (median Spearman's correlation [ρ] between instruments of 0.45 across 20 nutrients and an interquartile range [IQR] of 0.42-0.58) and generally adequate for men (median ρ = 0.31, IQR = 0.23-0.55). Validity was also good for Chamorros (median ρ = 0.47, IQR = 0.38-0.53) and generally adequate for Filipinos (median ρ = 0.42, IQR = 0.20-0.62). Correlations after energy adjustment were lower (overall median ρ = 0.20, IQR = 0.14-0.26). Conclusions: The FFQ can be used to rank nutrient intake for adults in Guam and may be helpful in the analysis of relationships between diet and chronic disease in Guam.