École Nationale Supérieure de
Techniques Avancées

Abstract

I did my internship at INRAE (National Research Institute for Agriculture, Food and Environment), within the multidisciplinary research unit SAD-APT. Supervised by Pierre Levasseur, a development economist specialized in nutrition and health issues in Mexico, my subject was the study of the socio-economic determinants of obesity and overweight in sub-Saharan Africa. My internship was carried out in two stages: the elaboration of a literature review and then the completion of an original econometric analysis on the link between socio-economic status and body mass for women aged between 18 and 49 years old. There is a global trend towards obesity. The sub-Saharan sub-continent, long limited to the issues of undernutrition, is beginning to face the "double burden of malnutrition", i.e. the simultaneous presence of undernutrition and overnutrition at different levels (country, municipality, household). The nutrition transition theorizes the shift in food consumption and energy expenditure, coinciding with economic, demographic and epidemiological changes. Thus, some authors point out that when a country reaches a specific level of economic development, the number of people with excess body mass increases. This increase in overweight and obesity has direct economic consequences, in terms of public health, but also indirect ones, such as a decrease in productivity. In addition, excess body mass can be determined by economic variables such as socio-economic status. However, this link is not fully stipulated, as demonstrated by the ambiguity that exists in the literature. My mission has centered around this general vision. Thus, I first conducted a review of the existing literature about the relationship between socio-economic status and body mass in sub-Saharan Africa. The main conclusions of this review were that a shift is taking place, given the increasing number of articles showing a non-significant or non-linear relationship between the two variables of interest, particularly among lower middle-income countries in sub-Saharan Africa. Nevertheless, there are many methodological, empirical and quantitative limitations that remain in the literature. Then, I got familiar with the Demographic and Health Survey database, which provides nationally representative demographic, health and nutrition survey data from many countries and over several survey waves. After creating a meta-base of anthropometric and socio-economic data of non-pregnant and non-lactating women aged 18-49 years from 36 sub-Saharan African countries between 1990 and 2019, I was able to test the hypotheses submitted at the end of the literature review. Thus, I implemented two regression models (one by ordinary 53 least squares method and one by instrumental variable method), to draw and then infer a causal relationship between socio-economic status and body mass. Using education as an indicator of socio-economic status, I obtained an inverted U-shaped relationship, highlighting the emergence of a middle class more likely to be overweight. Using wealth as an indicator of socio-economic status, the relationship stood more ambiguous, possibly taking a N-shaped form. The work carried out has technical limitations, such as not considering the heterogeneity of countries in the entire sub-Saharan subregion or the difficulty finding a suitable instrumental variable, but also temporal limitations. For example, I did not have time to deepen the relationship for specific categories, such as women living in urban areas or women from lower middle-income countries. Thus, the study of this topic could be deepened by future research. Nevertheless, this internship gave me a comprehensive view of the research work, for which I developed a particular appetence.