A highly polymorphic microsatellite within intron 5 of the porcine 54/56 kDa vacuolar H(+)-ATPase subunit gene (V-ATPase)

OBJECTIVE: To assess the validity of a short calcium food frequency questionnaire (FFQ) for use in young children. DESIGN: Calcium intake from an estimated 4 d diet record (4DDR) was compared with the calcium intake from a 35 item FFQ specifically designed to assess habitual calcium intake and previously validated for adult women. SUBJECTS: Forty-one girls and 26 boys aged 3-6 y recruited by advertisement for studies of nutrition and bone health. RESULTS: Mean (s.d.) calcium intakes were 798 mg (271) and 942 mg (419) for the 4DDR and FFQ respectively, (r = 0.52). Mean difference (s.d. of difference) in calcium intake between the two methods was 144 mg (355), showing that the FFQ may estimate calcium intakes 565 mg below to 854 mg above diet record values. 84% of subjects when classified by the 4DDR fell into the same or adjacent quartiles when classified by the FFQ. Only two subjects were classified in extreme quartiles for the two methods. The FFQ correctly identified 68% of children with recorded intakes less than 800 mg. CONCLUSIONS: The short calcium FFQ tended to overestimate actual calcium intakes in young children, and would not be appropriate for determining calcium intake of individuals. However, the FFQ demonstrated good ability to classify subjects into extremes of calcium intake. Moreover, the predictive value of the FFQ in identifying children with intakes below the current recommended intake of 800 mg was reasonably high (79%).