Poster Topical Area: Global Nutrition

Location: Hall D

Poster Board Number: 603

P12-122 - Dietary patterns and Determinates of changing diets in Bangladesh from 1985 to 2010

Monday, Jun 11
8:00 AM – 3:00 PM

Objectives:

We sought to track changes in diets in Bangladesh over 25 years and identify potential causes of changing diets using household consumption expenditure surveys.


Methods:

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.


Results:

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.


Conclusions:

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.



Figure 1: Proportion of the population residing in households with high adherence to identified dietary patterns by year of survey

Table 1: Rational for periods selected for decomposition analysis and summary results

CoAuthors: Sheela Sinharoy – Emory Univeristy; Ame Stormer – Helen Keller International; Sabine Gabrysch, PD Dr. med., MSc PhD – Institute of Public Health, University of Heidelberg

Jillian L. Waid

Director - Research, Learning, and Evaluation
University of Heidelberg
Dhaka, Dhaka, Bangladesh