Dietary diversity is a good predictor of the micronutrient density of the diet of 6- to 23-month-old children in Madagascar.

This study was conducted in the context of a multicountry validation of indicators of diet quality and had the following objectives: 1) to determine how well dietary diversity scores (DDS) predict diet quality of children aged 6-23 mo in urban Madagascar; and 2) to assess whether the prediction was improved by changing the food groups included and by imposing a minimum amount restriction. Correlation and regression were used to describe the relationship between 4 diversity scores (2 based on 8 and 7 food groups, the latter excluding fats and oils, and 2 that imposed a 10-g minimum restriction on food groups) and the mean micronutrient density adequacy (MMDA) of the diet. MMDA, the dietary quality score used, was calculated as the mean individual micronutrient density adequacy for 9 or 10 "problem" nutrients (depending on age and breast-feeding status), each capped at 100%. We used sensitivity and specificity analysis to determine how well DDS predicted MMDA below or above selected cut-offs. All scores were positively correlated with MMDA. When the fats and oils group was omitted, correlations were 10-16% higher for breast-fed children and 19-28% higher for non-breast-fed children. Correlations were only slightly improved with the 10-g minimum. With the 7-food group score, a score of <or=2 best predicted low dietary quality (MMDA <50%), with 64% sensitivity, 82% specificity, and 22% misclassification. Imposing a 10-g minimum increased misclassification (30%). These results support the growing evidence of the usefulness of dietary diversity to predict dietary quality, and among infants and young children more specifically.