Using Household Surveys and Other Secondary Data to Assess the Need for and Impacts of Micronutrient Interventions (FS01-03-19)

Abstract

ObjectivesGiven the scarcity of nationally-representative dietary intake data in most low- and middle-income countries, our objective was to compare estimates of micronutrient (MN) intake and predicted MN intervention effectiveness generated using household consumption and expenditure surveys (HCES) and other secondary data vs individual-level 24-hr dietary recalls (24HR), focusing on vitamin A (VA) in children and iron in women of reproductive age (WRA). MethodsWe calculated “apparent” food consumption using HCES data from Cameroon and applied the adult male equivalent method to estimate individual-level food intake. For children, intake was matched with food composition data to calculate VA intake, inadequate intake, and, combined with Demographic and Health Survey (DHS) data, effective coverage (shift from inadequate to adequate intake) of VA interventions. For WRA, we combined DHS data with HCES estimates of fortifiable food intake to predict cases of iron deficiency anemia averted. Results were compared to those generated by modeling data from a national 24HR survey. ResultsMean apparent VA intake among children 6–59 mo from HCES data was similar to 24HR estimates for non-breastfed children (206 vs 215 μg RAE/d) and followed the same sub-national pattern (highest in the South macro-region, followed by the cities, and lowest in the North). Prevalence of inadequate intake was also similar (64% vs 59% nationally). Both data sources predicted effective coverage was highest for VA supplements, micronutrient powders, and oil fortification (Fig. 1).Preventive iron supplements to WRA was predicted to avert a relatively high proportion of anemia based on both the 24HR (24–62% of cases of anemia averted in the cities, 15–26% in the North, and 12–57% in the South) and secondary data (28–60%, 17–37%, and 27–56%). Iron-fortified wheat flour (60 ppm) and bouillon cubes (600 ppm) were predicted to avert few cases of anemia among WRA (< 7% and < 3%, respectively) in all macro regions. ConclusionsWhere dietary intake data are unavailable, secondary data may be useful for identifying effective MN interventions. Challenges include uncertainty about intrahousehold food distribution, MN intake from breastmilk, and combining multiple secondary data sources. Funding SourcesThis study is based on research funded by a grant to UC Davis from the Bill & Melinda Gates Foundation. Supporting Tables, Images and/or Graphs▪