Validity and reliability of a semi-quantitative food frequency questionnaire for assessing dietary vitamin D and calcium intakes in Iranian childbearing age women.

Abstract

This study aimed to examine the validity and reliability of a semi-quantitative food frequency questionnaire (FFQ) designed to estimate dietary vitamin D and Calcium (Ca) intakes in a sample of Iranian childbearing age women. An 87-item FFQ was developed and 84 healthy childbearing age women completed it. FFQ was validated by comparing its results with the 24-h dietary recall (24-h recall) and serum 25-hydroxyvitamin D (S-25(OH)D) as the references methods. The FFQ was completed for the second time after 4 weeks to assess the reliability of the questionnaire. Data were analyzed using spearman's correlation, cross-classification analysis, Bland-Altman plots, the weighted κ, intraclass classification, and the method of triads. Spearman's correlations between vitamin D and Ca intakes by the FFQ and 24-h recalls and between vitamin D intakes and S-25(OH)D were significant (r: 0.706, r: 0.959, and r: 0.682, respectively, all, P < 0.001). Cross-classification for vitamin D and Ca between two dietary methods and for vitamin D intake of FFQ and S-25(OH)D demonstrated that most of the quartiles were categorized into the same or adjacent quartiles. The Bland Altman plots for these nutrients also demonstrated good agreement. For vitamin D, the validity coefficients (VCs) calculated by the method of the triad for three methods were in the range of 0.808-0.843. The weighted κ for agreement of vitamin D and Ca by FFQ1 and FFQ2 were 0.18 and 0.638, respectively. The findings indicated that the developed FFQ has acceptable validity for estimating vitamin D and Ca. Its reliability for Ca was stronger than vitamin D.