Abstract
This thesis:1. develops an operational definition of the concept of a dietary pattern
2. employs this to explain the relationship between household characteristics and the nutritional status of its food consumption
3. identifies potential policy levers to improve national diet.
The study is an attempted synthesis of economic and nutritional analysis. The nutritionist approach is to statistically group patterns of consumption using some form of cluster analysis. The household characteristics of the members of the resulting groups are then established by regression analysis. Such a two step approach fails to acknowledge the constraints faced by households, particularly in relation to income. A poor diet is thus established purely by an assessment of nutritional content, irrespective of whether this is due to low income, a lack of information or cultural factors.
The economic analysis of diet is based upon the Engel curve. The advantage of this approach is that it highlights how income can constrain the dietary outcomes. The limitations are that social and cultural factors are ignored. Augmenting the Engel curve by household characteristics is explored but the results are difficult to interpret and have little policy content.
Therefore this thesis uses Latent Class Cluster Analysis (LCA) to attempt a synthesis of the economic and nutritional approaches. The Engel curve is central but is allowed to vary between groups of consumers where group membership is estimated by household characteristics; household income is not employed in the cluster analysis. Group membership is estimated simultaneously with the parameters of the Engel curve of the group.
The association between diet and health is well recognized in public health especially with the growth of obesity. From a policy perspective the results suggest that households can be influenced either through their preferences or their production of health.
| Date of Award | May 2011 |
|---|---|
| Original language | English |
| Supervisor | Patrick McGregor (Supervisor) & Ciaran O'Neill (Supervisor) |
Cite this
- Standard