Simonette R Mallard, Lisa A Houghton, Suzanne Filteau, Molly Chisenga, Joshua Siame, Lackson Kasonka, Anne Mullen, Rosalind S Gibson

In the monitoring of infant and young child feeding, dietary diversity is used as an indicator of micronutrient adequacy; however, their relation may have weakened with the increasing use of fortified complementary foods. The objectives were to assess the relation between dietary diversity and micronutrient adequacy in an urban infant population with a high consumption of fortified foods and to investigate whether dietary diversity and micronutrient adequacy were independently associated with subsequent growth. We used longitudinal data on 811 infants in the Chilenje Infant Growth, Nutrition, and Infection Study conducted in Lusaka, Zambia. The relation between mean micronutrient adequacies and dietary diversity scores derived from 24-h diet recalls at 6 mo of age was investigated with the use of Spearman rank correlation. Multiple linear regression was used to assess the association between micronutrient adequacy, dietary diversity, and subsequent growth to 18 mo of age. Overall mean micronutrient density adequacy (MMDA) and MMDA of "problem micronutrients," defined as those micronutrients (calcium, iron, zinc) with mean density adequacies less than half of estimated needs, were correlated with dietary diversity scores (ρ = 0.36 and 0.30, respectively, both P < 0.0001). Consumption of "sentinel foods" (iron rich, fortified, animal source, dairy) showed better correlation with MMDA than with dietary diversity (ρ = 0.58-0.69, all P < 0.0001). In fully adjusted analyses, MMDA calcium, iron, zinc, and dietary diversity, but not overall MMDA, were associated with linear growth to 18 mo (both P ≤ 0.028). Micronutrient adequacy in infants consuming fortified foods may be more accurately assessed using locally specific sentinel food indicators rather than dietary diversity scores. Nonetheless, dietary diversity has a positive effect on subsequent linear growth apart from that of micronutrient adequacy, warranting its continued monitoring and further investigation into the mechanisms underlying this finding. This trial was registered at www.controlled-trials.com as ISRCTN37460449.