Validity and Reproducibility of Malaysian Food Frequency Questionnaire for Dietary Intake Related to Colorectal Cancer

Abstract

This study aims to report on the validity and reproducibility of a 142-food item Food Frequency Questionnaire (FFQ) for dietary factors related to colorectal cancer among Malaysians. Population aged 30 to 70 years from two cities of Peninsular Malaysia were recruited through voluntary participation. A semi-quantitative FFQ was modified from an established FFQ used in the national survey. It includes specific questions to measure the consumption of food sources related to colorectal cancer development. FFQ was administered two times in two weeks to evaluate reproducibility (FFQ1 and FFQ2). Then the validity was assessed by comparing FFQ against the 3-day Food Record method (FR). A total of 100 respondents (mean age 50.6 years) provided data for both validity and reproducibility. The FFQ had significantly higher estimates of most nutrients and food groups’ intake than the FR. The Spearman correlation showed moderate agreement between FFQ and FR while moderate to strong correlation between FFQs. The limit of agreement between both methods using Bland Altman plot was acceptable or both validity and reproducibility. The classification into the same and adjacent quartiles was between 62 to 75% for validity and 77 to 89% for reproducibility assessment. Overall, the validity was satisfactory and reproducibility of the FFQ was good for estimating absolute nutrient and food group intakes. Hence, the FFQ could be used as a valid tool for assessing dietary intake among Malaysians to study dietary
factors related to colorectal cancer risk.