Invisible heterogeneity in crop zinc concentration and child zinc intake in rural Uganda

Micronutrient deficiencies affect more than two billion individuals worldwide, with dire outcomes for human health and productivity. Yet estimates of individual-level nutrient intake or population-wide deficiency rates often rely on food composition tables (FCTs) that do not capture the true nutrient content of consumed foods. This is justified by the assumption that food nutrient heterogeneity, while present, is too minor to significantly impact dietary intake. Using a unique dataset from rural Uganda with food recall data for children under five as well as the zinc concentration of household-sampled and market-sampled crops, I document massive heterogeneity in the zinc concentration of staple crops. I additionally find that crops sampled from markets are lower in zinc concentration than crops sampled at households. Zinc inadequacy shifts significantly according to assumptions about the nutrient content of foodsinadequacy appears to double when zinc intake is estimated using market-sampled crops rather than household-sampled crops. The HarvestPlus FCT for Uganda does fairly well estimating dietary zinc intake as a whole, even though it often fails to capture median zinc concentration for a specific crop. However, in a setting with less dietary diversity, or in urban settings where children rely heavily on market-purchased crops, the HarvestPlus FCT might perform worse. Additionally, low-zinc market-purchased crops play a key role in shaping dietary zinc intake, which raises questions about the efficacy of biofortification for tackling low nutrient intake in poor, rural areas of Uganda.