RESUMEN
OBJECTIVES: The failure of single-nutrient supplementation to prevent disease in intervention studies underlines the necessity to develop a holistic view of food intake. The objectives of this study were to devise a diet quality index (DQI) and identify biomarkers of multidimensional dietary behavior. DESIGN: A nutrition survey was conducted in Mediterranean southern France by means of a food frequency questionnaire. The DQI was based on current dietary recommendations for prevention of diet-related diseases such as cardiovascular disease and some cancers. A second DQI included tobacco use. STATISTICAL ANALYSES: performed Spearman rank correlations, cross-classifications and intraclass correlations were computed between the DQI and biomarkers. RESULTS: Of the 146 subjects, 10 had a healthful diet and 18 had a poor diet. Erythrocyte omega-3 fatty acids-eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)-beta carotene, and vitamin E concentrations were lower and cholesterol concentrations were higher in the poor diet; the difference was significant for EPA and DHA and borderline significant for vitamin E. Significant correlation was found between the DQI and vitamin E (-0.12), EPA (-0.30), and DHA (-0.28), and beta carotene (-0.17) when tobacco use was considered, but not between the DQI and cholesterol. The correlation coefficient reached 0.58 (P0.01) for a composite index based on all biomarkers except cholesterol. CONCLUSIONS: Subjects with a beta carotene levels greater thanl micromol/L, vitamin E greater than 30 micromol/L and EPA greater than 0.65% and DHA greater than 4% of fatty acids in erythrocytes were likely to have a healthful diet. Each biomarker indicated the quality of diet, but correlation was higher with a composite index.