Selected nutritional biomarkers predict diet quality.

To examine associations of biomarkers of nutrient intake with overall diet quality. A convenience sample of 102 healthy postmenopausal women in Seattle, Washington (USA). Participants attended a study centre where they provided fasting blood specimens and completed a 122-item validated food-frequency questionnaire (FFQ). Data from the FFQ were used to calculate Diet Quality Index (DQI) scores and categorise women as having diets of excellent, good, fair or poor quality. The blood specimens were analysed for nine phospholipid fatty acids (as percentage of total) and serum concentrations of vitamin C, alpha-tocopherol, gamma-tocopherol, vitamin B12, folate and six carotenoids. Multivariate linear regression was used to model associations of the nutrient biomarkers with DQI scores. Compared with women with poor-quality diets, women with excellent diets, as measured by the DQI, had higher plasma concentrations of vitamin C (P for trend=0.01), alpha-tocopherol (P for trend=0.02) and beta-cryptoxanthin (P for trend=0.03). Women with excellent diets also had lower proportions of plasma phospholipid fatty acids of two potentially atherogenic fatty acids: stearic acid (P for trend=0.01) and behenic acid (P for trend=0.03). A group of six biomarkers explained a moderate proportion of the total variability in DQI scores (36%). These objective measures of dietary intake support the use of the DQI as a useful tool to measure dietary patterns.