Reproducibility and relative validity of a semi-quantitative food frequency questionnaire designed for assessing the intake of dioxin-like contaminants.

The aim of this study was to assess the reproducibility and relative validity of a food frequency questionnaire (FFQ) designed to estimate the intake of dioxin-like contaminants.In total, 180 adults (50-65 years) completed the FFQ twice and 157 of them kept four times a 2-day estimated dietary record (EDR) spread over a year.Spearman rank correlation coefficients varied between 0.58 and 0.67 for reproducibility, versus between 0.28 and 0.46 for validity for the different food groups under study. Exact agreement within quartiles was between 34% and 39%, only between 3% and 9% was grossly misclassified. The weighed kappa statistics indicated fair agreement between the two methods (0.29 for the total diet). The correction factors used had a relevant influence on the absolute estimated intake.The FFQ presented here is a valuable tool for ranking individuals in the study population on the basis of estimated intake of dioxin-like contaminants. However, absolute intakes should be estimated without correction factors and interpreted with caution.