Assessing the Coverage of Biofortified Foods: Development and Testing of Methods and Indicators in Musanze, Rwanda.

ABSTRACT

BACKGROUND: Biofortification of staple crops has the potential to increase nutrient intakes and improve health outcomes. Despite program data on the number of farming households reached with and growing biofortified crops, information on the coverage of biofortified foods in the general population is often lacking. Such information is needed to ascertain potential for impact and identify bottlenecks to parts of the impact pathway. OBJECTIVES: We aimed to develop and test methods and indicators for assessing household coverage of biofortified foods. METHODS: To assess biofortification programs, 5 indicators of population-wide household coverage were developed, building on approaches previously used to assess large-scale food fortification programs. These were 1) consumption of the food; 2) awareness of the biofortified food; 3) availability of the biofortified food; 4) consumption of the biofortified food (ever); and 5) consumption of the biofortified food (current). To ensure that the indicators are applicable to different settings they were tested in a cross-sectional household-based cluster survey in rural and peri-urban areas in Musanze District, Rwanda where planting materials for iron-biofortified beans (IBs) and orange-fleshed sweet potatoes (OFSPs) were delivered. RESULTS: Among the 242 households surveyed, consumption of beans and sweet potatoes was 99.2% and 96.3%, respectively. Awareness of IBs or OFSPs was 65.7% and 48.8%, and availability was 23.6% and 10.7%, respectively. Overall, 15.3% and 10.7% of households reported ever consuming IBs and OFSPs, and 10.4% and 2.1% of households were currently consuming these foods, respectively. The major bottlenecks to coverage of biofortified foods were awareness and availability. CONCLUSIONS: These methods and indicators fill a gap in the availability of tools to assess coverage of biofortified foods, and the results of the survey highlight their utility for identifying bottlenecks. Further testing is warranted to confirm the generalizability of the coverage indicators and inform their operationalization when deployed in different settings.