Household-level consumption data can be redistributed for individual-level Optifood diet modeling: analysis from four...

Area of Science:

• Nutritional epidemiology
• Public health nutrition
• Dietary assessment methodology

Background:

• Optifood linear programming (LP) requires detailed dietary intake data for developing population-specific food-based recommendations (FBRs).
• A significant barrier to Optifood use is the need for individual-level dietary data, such as 24-hour recalls (24HRs).
• Household Consumption and Expenditure Surveys (HCESs) offer an alternative data source, but their utility for young children's dietary assessment is debated.

Purpose of the Study:

• To investigate the feasibility of using HCES data as a substitute for 24HR data in Optifood LP analyses.
• To compare Optifood inputs and outputs derived from paired HCES and 24HR data for 12- to 23-month-old children.
• To assess the impact of using HCES data on the identification of nutrient gaps and the development of FBRs.

Main Methods:

• Paired HCES and 24HR data were collected from rural areas in Kenya, Uganda, Guatemala, and Bangladesh for 12- to 23-month-old children.
• HCES food intakes were adjusted for breastfeeding and estimated using adult male equivalents.
• Percentage agreement was used to compare LP inputs (food lists, parameters) and outputs (nutrient gaps, FBRs) derived from both data sources.

Main Results:

• Mean percentage agreement between HCES and 24HR data was 42% for food, 63% for food subgroups, and 80% for food-group model parameters.
• HCES-derived food lists were, on average, 1.3 times longer than those from 24HRs.
• High agreement (77-100%) was observed in identifying nutrient gaps, food sources of nutrients, and developing FBRs, suggesting comparable outcomes.

Conclusions:

• HCES data can be effectively utilized in Optifood analyses for 12- to 23-month-old children, facilitating nutrient gap identification and FBR development.
• Despite challenges in estimating young children's precise intakes from HCES data, the overall results for nutrient gaps and FBRs are comparable to 24HRs.
• Further research is recommended across diverse age groups and geographical locations to validate the generalizability of these findings.