Biochemical Validation of the Older Australian’s Food Frequency Questionnaire Using Carotenoids and Vitamin E

BACKGROUND Validation of a food frequency questionnaire (FFQ) is important, as inaccurate and imprecise information may affect the association between dietary exposure and health outcomes. OBJECTIVE This study assessed the validity of the Older Australian's FFQ against plasma carotenoids and Vitamin E. METHODS A random subsample (n = 150) of 2420 participants in the Hunter Community Study, aged 55-85 years, were included. Correlations between crude and energy-adjusted FFQ estimates of carotenoids, Vitamin E, and fruit and vegetables with corresponding biomarkers were determined. Percentages of participants correctly classified in the same quartile, and in the same ± 1 quartile, by the two methods were calculated. RESULTS Significant correlations (P < 0.05) were observed for α-carotene (r = 0.26-0.28), β-carotene (r = 0.21-0.25), and β-cryptoxanthin (r = 0.21-0.23). Intakes of fruits and vegetables also showed similar correlations with these plasma carotenoids. Lycopene was only significantly correlated with fruit and vegetable intakes (r = 0.19-0.23). Weak correlations were observed for lutein + zeaxanthin (r = 0.12-0.16). For Vitamin E, significant correlation was observed for energy-adjusted FFQ estimate and biomarker (r = 0.20). More than 68% of individuals were correctly classified within the same or adjacent quartile, except for lutein + zeaxanthin. CONCLUSION With the exception of lutein + zeaxanthin, the Older Australian's FFQ provides reasonable rankings for individuals according to their carotenoids, Vitamin E, fruit and vegetable intakes.