24-hour history is more closely associated with vitamin A status and provides a better estimate of dietary vitamin A intake of deficient Indonesian preschool children than a food frequency method

Objectives To determine if the Simplified Dietary Assessment to Identify Groups at Risk of Inadequate Intake of Vitamin A developed by the International Vitamin A Consultative Group (IVACG) correctly classified a group of vitamin A-deficient children as being at risk, and to see if a food frequency questionnaire (FFQ) or 24-hour history (24HH) yielded estimated dietary vitamin A intakes most closely associated with vitamin A status.

Design Forty-seven foods were identified as contributing most of the vitamin A to the diet of the study population. For each food, usual portion sizes were determined during a pilot study. Intake was calculated from data collected by FFQ and 24HH. Four modifications of the 24HH analysis were made to determine if this method of analysis could be simplified.

Subjects/setting Subjects were 265 Indonesian children with or at high risk of developing xerophthalmia.

Results Mean and median intakes of vitamin A based on the 24HH analysis were 50% and 27% of the US Recommended Dietary Allowance, respectively, which accurately identified the study sample as being at risk. Dietary intake based on the 24HH was significantly associated with serum retinol concentration (P=.01, trend test). Eliciting portion sizes during the 24HH was not necessary once the usual portion sizes consumed by the population were estimated in the pilot study. Mean and median intakes of vitamin A based on the FFQ were 150% and 118% of the Recommended Dietary Allowance, respectively, which suggests that the FFQ overestimated intake. Intake based on the FFQ was not correlated with serum retinol concentration.

Conclusions Our findings do not support the IVACG recommendation that the FFQ be regarded as more reliable than the 24HH when the 2 methods produce different conclusions, nor the recommendation of some users of the method that the 24HH be dropped from the assessment method. J Am Diet Assoc. 2000;100:1501-1507.