Use of a food frequency questionnaire to assess the relationship between dietary habits and cardiovascular risk factors in the NESCAV study: validation with biomarkers

Background In the interregional cross-sectional study NESCAV (Nutrition, Environment and Cardiovascular Health), a modified semi-quantitative Food frequency questionnaire (FFQ) was used to assess dietary habits of the “Greater Region” (Luxembourg, Wallonia in Belgium and Lorraine in France) and to explore the relationship between diet and cardiovascular risk factors (CVRF). Validation of the FFQ is essential since incorrect information may lead to erroneous associations between dietary factors and diseases. The aim of the study was to evaluate the validity of the FFQ used in the NESCAV study by comparing the estimated intakes of several micro- and macro-nutrients with the corresponding nutritional biomarkers. Methods Relative validity was assessed in a sample of 984 subjects (429 men and 455 women). Comparisons between FFQ-estimates and their corresponding biomarkers were performed by means of correlation analysis and cross classification into quintiles using both crude and energy-adjusted FFQ-estimates. Correlations were also adjusted for potential confounders. Results For micro-nutrients, significant correlations were found for vitamin B9, D and β-carotene in both men and women. Iodine intake, vitamin B12, E and sodium were also related to their biomarkers but only in women. Energy-adjustment led to an increase in all correlations listed above. However, after excluding supplement users, correlations for vitamin D were no longer significant. For macro-nutrients, there was almost no association when crude data were used. However, correlations became significant after adjusting for energy and confounders. In both men and women, saturated fatty acids (SFA), dietary carbohydrate and total fiber were correlated with their biomarkers and so was dietary cholesterol with plasma cholesterol but only in men. Similar findings were obtained with cross-classification analyses. Conclusions The present study demonstrates the validity of the FFQ as valuable tool to assess micro- and macro-nutrients in population-based studies. Key message Our modified FFQ is a reasonable tool to assess dietary patterns in people living in the Greater region, although protein and Vitamin D intakes should be interpreted with caution.