Short food frequency questionnaire can discriminate inadequate and adequate calcium intake in Croatian postmenopausal women

Valid and cost-effective instruments are needed for measuring dietary calcium intake in clinical or epidemiologic studies as well as for education. A food frequency questionnaire (FFQ) is the dietary assessment tool most often used for such purposes; but FFQs are population-, age-, and sex-specific, so each FFQ must be tested for its validity in the population of concern. The aim of this study was 2-fold: to develop and validate a short FFQ for assessing calcium intake in postmenopausal Croatian women. The development of the short calcium FFQ was based on a previous FFQ, after conducting a pilot study in 100 women and including in the short FFQ only those food items that cumulatively accounted for 95% of the total calcium intake by the previous FFQ. The short FFQ (27 items) was then tested in 333 women and compared against a single multipass 24-hour recall. The Pearson and Spearman correlation coefficients between the 2 methods were 0.70 and 0.64, respectively. The weighted κ was 0.43. Cross-classification showed that 55.3% of subjects belonged to the same third. Predictive value of the FFQ at a cutoff of 1200 mg calcium per day was 94.4%. In conclusion, the FFQ can discriminate low or inadequate (<500 mg calcium) and high or adequate (>1000 mg calcium) intakes; but the Bland-Altman method showed that the FFQ is not appropriate for the assessment of individual calcium intake.