Skip to main content

An analysis of changing dietary trends and the implications for global health.

Le, T. H., 2021. An analysis of changing dietary trends and the implications for global health. Doctoral Thesis (Doctoral). Bournemouth University.

Full text available as:

LE, Thai Hong_Ph.D._2021.pdf
Available under License Creative Commons Attribution Non-commercial.



Worldwide, obesity has reached epidemic proportions and has almost tripled between 1975 and 2016. Acknowledging that weight gain is a complex and multifactorial condition involving changes in both dietary and physical activity patterns, this research focuses on the dietary origin of obesity. Given the dearth of empirical literature on food consumption at the global level, the aim of this research is to explore dietary trends and dietary types around the world linked with the indications that they imply for obesity and global health. To this aim, annual food availability data collated by the Food and Agriculture Organisation of the United Nations (FAO) covering the period from 1961 to 2013 for 118 countries are scrutinised by econometric convergence tests, clustering techniques and spatial analysis. Results indicate that countries with lower levels of initial calories tend to exhibit higher growth rates of calorie consumption. However, this process is not homogeneous across countries. Low-income countries have converged at the fastest pace and the convergence rate reduces as income rises. In addition, the dietary convergence is conditioned by a range of structural indicators including agroecological, demographic and socio-economic variables. Evidence suggests that economic factors have become a more important determinant of the dietary convergence since the Millennium. Applying innovative fuzzy clustering algorithms which allow multiple diets to coexist within a single country, several dietary trends and dietary types are detected. While the identified clusters are all associated with relentlessly increasing calories and deteriorating dietary healthiness over the past half a century, the most calorific cluster has shown signs of stabilising calorie consumption. A notable contribution of this research is the examination of spatial patterns of global food consumption using both traditional and non-traditional measures of spatial proximity. Differing from the earlier literature emphasising the role of geographical closeness, this research utilises an economic indicator for proximity and finds that countries with similar income levels tend to have similar diets. Spatial convergence analysis reveals a convergence process that is about three times as fast as the non-spatial model; thus, ignoring the spatial relationship leads to biased results. Incorporating the spatial dimension in cluster analysis also affects the clustering results dramatically. Cluster profiling shows that only the segment of more educated and health-aware populations exhibits the behavioural changes towards better diets, hence underlining the importance of improved education and access to knowledge. The finding that dietary evolutions are ‘spatially’ dependent provides a basis for the development of group-specific interventions that target populations at risk of worsening diets. While these policy measures are often place-based, this research lays foundations for the implementation of coherent food policies beyond geographical boundaries. Overall, this research highlights that healthier diets are possible, but we need to act now. As we are living longer but not necessarily healthier, current attempts to improve diets are obviously inadequate and existing efforts need to be redoubled. This is an urgent message for policymakers considering the sobering fact that no country has been able to significantly reverse the rise in obesity.

Item Type:Thesis (Doctoral)
Additional Information:If you feel that this work infringes your copyright then please contact the BURO Manager.
Uncontrolled Keywords:dietary trends; diets; obesity; cluster analysis; convergence analysis; food consumption; Food Balance Sheet
Group:Bournemouth University Business School
ID Code:36106
Deposited By: Symplectic RT2
Deposited On:13 Oct 2021 11:14
Last Modified:14 Mar 2022 14:29


Downloads per month over past year

More statistics for this item...
Repository Staff Only -