Identifying areas with vitamin A deficiency: the validity of a semiquantitative food frequency method.

Abstract

The prevalence of vitamin A deficiency has traditionally been assessed through xerophthalmia or biochemical surveys. The cost and complexity of implementing these methods limits the ability of nonresearch organizations to identify vitamin A deficiency. This study examined the validity of a simple, inexpensive food frequency method to identify areas with a high prevalence of vitamin A deficiency. The validity of the method was tested in 15 communities, 5 each from the Philippines, Guatemala, and Tanzania. Serum retinol concentrations of less than 20 micrograms/dL defined vitamin A deficiency. Weighted measures of vitamin A intake six or fewer times per week and unweighted measures of consumption of animal sources of vitamin A four or fewer times per week correctly classified seven of eight communities as having a high prevalence of vitamin A deficiency (i.e., 15% or more preschool-aged children in the community had the deficiency) (sensitivity = 87.5%) and four of seven communities as having a low prevalence (specificity = 57.1%). This method correctly classified the vitamin A deficiency status of 73.3% of the communities but demonstrated a high false-positive rate (42.9%).