Complementary feeding and attained linear growth among 6-23-month-old children.

Objective: To examine the association between complementary feeding indicators and attained linear growth at 6-23 months.

Design: Secondary analysis of Phase V Demographic and Health Surveys data (2003-2008). Country-specific ANOVA models were used to estimate effects of three complementary feeding indicators (minimum meal frequency, minimum dietary diversity and minimum adequate diet) on length-for-age, adjusted for covariates and interactions of interest.

Setting: Twenty-one countries (four Asian, twelve African, four from the Americas and one European).

Subjects: Sample sizes ranging from 608 to 13 676.

Results: Less than half the countries met minimum meal frequency and minimum dietary diversity, and only Peru had a majority of the sample receiving a minimum adequate diet. Minimum dietary diversity was the indicator most consistently associated with attained length, having significant positive effect estimates (ranging from 0·16 to 1·40 for length-for-age Z-score) in twelve out of twenty-one countries. Length-for-age declined with age in all countries, and the greatest declines in its Z-score were seen in countries (Niger, -1·9; Mali, -1·6; Democratic Republic of Congo, -1·4; Ethiopia, -1·3) where dietary diversity was persistently low or increased very little with age.

Conclusions: There is growing recognition that poor complementary feeding contributes to the characteristic negative growth trends observed in developing countries and therefore needs focused attention and its own tailored interventions. Dietary diversity has the potential to improve linear growth. Using four food groups to define minimum dietary diversity appears to capture enough information in a simplified, standard format for multi-country comparisons of the quality of complementary diets.