Abstract
In many countries, statistics from household consumption and expenditure surveys are increasingly being used to inform policies and programs. In household surveys, foods are typically reported as they are acquired (the majority are raw). However, the micronutrient content of some foods diminishes during processing and cooking. Using food consumption data from the 2015/16 Kenya Integrated Household Budget Survey, this study analyzes whether mean consumption estimates of dietary energy, macronutrients, and eight micronutrients are equivalent (applying a two-side paired equivalence test) when matching foods: (1) considering the nutrient content in raw foods (as reported in the survey), and (2) considering the nutrient content in foods as typically consumed, thus applying yield and retention factors as needed. Both food matching approaches rendered statistically equivalent mean consumption estimates, at national and county levels, for dietary energy, protein, fats, available carbohydrates, total fiber, calcium and zinc. Non-equivalent means were found for iron, vitamins A, B1, B2, B12, and C. The higher differences between the means were, in percentage change, for vitamin C (47 %), B1 (34 %) and B12 (26 %).