Plasma carotenoid levels in Dutch men and women, and the relation with vegetable and fruit consumption

Objective: Fruit and vegetable intake is inversely associated with cancer risk in many epidemiological studies. Accurate assessment of consumption of these foods is difficult, and biomarkers of intake would overcome several drawbacks of currently used dietary assessment methods. Therefore, we investigated the relation between plasma carotenoids and usual vegetable and fruit intake. Design: Plasma carotenoid concentrations were measured and vegetable, fruit and juice consumption was assessed by a food frequency questionnaire (FFQ) in a random sample of 591 Dutch men and women aged 20-59 y from the MORGEN-project, one of the contributions to the European Prospective Investigation into Cancer and Nutrition (EPIC)-study. Results: In this sample of the general Dutch population, in both genders, relative to the other carotenoids, plasma β-cryptoxanthin was the best indicator for fruit intake, and for the sum of vegetable, fruit and juice intake, while lutein concentrations best reflected intake of vegetables, although quartiles of intake were not consistently separated. Since levels of lycopene were not associated with any of the main food groups examined, associations with total carotenoids improved when excluding lycopene, and monotonously increasing plasma levels were seen for intakes of vegetables, of fruits, and of the sum of vegetables, fruits and juices. Several vegetable types and orange/grapefruit juice were associated with plasma levels of one of the carotenoids. Conclusion: Plasma carotenoids were only crude indicators of vegetable and fruit intake as assessed by a FFQ; β-cryptoxanthin for fruit intake and lutein for vegetable intake. None of the plasma carotenoids could distinguish all four quartiles of vegetables, fruit and/or juice intake. © 2004 Nature Publishing Group. All rights reserved