Use of food frequency questionnaire to assess relationships between dietary habits and cardiovascular risk factors in NESCAV study: validation with biomarkers
1 CRP-Santé, Department of Public Health, 1A rue Thomas Edison, L-1445 Strassen, Luxembourg
2 Ecole de Santé Publique, Université de Liège, Liège, Belgium
3 CRP-Santé, CCMS (Competence center in methodology and statistics), 1A rue Thomas Edison, L-1445 Strassen, Luxembourg
Nutrition Journal 2013, 12:143 doi:10.1186/1475-2891-12-143Published: 6 November 2013
Validation of Food frequency questionnaire (FFQ) is particularly important element, as incorrect information may lead to false associations between dietary factors and diseases. The aim of the study was to evaluate the validity of the FFQ used in NESCAV (Nutrition, Environment and Cardiovascular Health) study, by comparing the estimated intakes of fruits and vegetables and of several micro-nutrients with corresponding nutritional biomarkers.
Relative validity was assessed in a sample of 922 subjects (452 men and 470 women). Comparisons between FFQ-estimates and their corresponding biomarkers were performed through correlation and cross classification into quintiles by using both crude and energy-adjusted FFQ-estimates. Correlations adjusted for confounders were also computed. All analyses were performed separately for men and women.
Concerning micro-nutrients, significant correlations were found for vitamin B9, D, E, B12 β-carotene and iodine in both men and women. Energy-adjustment led to an increase of all correlations cited previously. However, after excluding supplement users, correlations for vitamin D were not significant anymore. Concerning fruits and vegetables, all correlations were significant. Vegetables alone and fruits and vegetables correlated better in men (r around 0.2) than in women (r around 0.1). In men, correlation was also better for vegetables alone and fruits and vegetables than fruits alone.
These data demonstrate that this FFQ is a reasonable tool to assess intakes of fruits and vegetables and of several micro-nutrients. We conclude that our FFQ is suitable to be used in NESCAV study, although protein and vitamin D estimates should be interpreted with caution.