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

Cross-sectional validation study within the NESCAV project; 922 adults (452 men, 470 women) aged 18–69 years; general population from the Wallonia region; FFQ estimates compared with nutritional biomarkers; analyses performed separately for men and women; three correlation models (crude, energy-adjusted model I, energy-adjusted model II); cross-classification into quintiles; adjustments for age, BMI, current smoking and specific micro-nutrient supplement use; sensitivity analyses excluding supplement users; analyses performed with SAS.

FFQ showed acceptable validity for ranking intakes of fruits/vegetables and several micronutrients. Energy adjustment generally improved correlations. Vitamin B9 had the strongest and most consistent correlations with biomarkers (serum folates and erythrocyte folates). Vitamin D correlations were modest and became non-significant after adjusting for supplement use. Vitamin B12 and Vitamin E correlations were mainly observed in women. Beta-carotene around 0.20; iodine around 0.10–0.23; iron and sodium showed weaker associations. Cross-classification indicated substantial within-quintile agreement for several nutrients (notably Vitamin B9). The FFQ is suitable for ranking nutrient intakes in the NESCAV study, but protein and vitamin D estimates should be interpreted with caution.

No recovery biomarkers; cross-sectional design; potential misreporting and supplement-use effects; 62 participants excluded due to incomplete supplement data; some nutrients show weak or inconsistent validity; results may be influenced by supplement use and seasonality.