Poster Topical Area: Global Nutrition
Location: Hall D
Poster Board Number: 603
We sought to track changes in diets in Bangladesh over 25 years and identify potential causes of changing diets using household consumption expenditure surveys.
We identified dietary patterns using principal component analysis on data from seven cross-sectional rounds of the Bangladesh Household [Income and] Expenditure Surveys. We used linear probability models to assess associations between adherence to dietary patterns, socioeconomic characteristics of households, and agricultural production at the household and regional level. For dietary patterns that increased/decreased over time, Blinder-Oaxaca decomposition was used to assess factors associated with these changes.
We identified seven dietary patterns: modern, traditional, festival, winter, summer, monotonous, and spices. All diets were present in all survey years, with the modern, winter, summer, and monotonous dietary patterns increasing over time, while the traditional decreased (Figure 1). In the pooled sample, many factors were associated with adherence to dietary patterns, with expenditure, presence of child[ren], and smaller households positively associated with every dietary pattern, while other associated factors varied by pattern. Decomposition results were not significant for the summer dietary pattern, while observed factors predicted an increase in adherence to the traditional diet when a decrease was observed (Table 1). For the remaining dietary patterns, observed factors explained from 12% (modern) to 48% (no pattern) of the observed change in adherence. Of observed factors, changes in agriculture and expenditure were the largest drivers of dietary change.
Dietary patterns can be identified from household consumption data and are similar to those obtained from individual data in number and amount of variance explained. In Bangladesh, dietary patterns have changed gradually over time, and these changes were linked to both economic development and the agricultural system.
Director - Research, Learning, and Evaluation
University of Heidelberg
Dhaka, Dhaka, Bangladesh