Abstract:
The nutrition transition and global convergence on a high fat, low complex carbohydrate diet have led to corresponding increases in obesity and associated chronic diseases, now recognised as global public health problems. Population food and nutrition monitoring plays a critical role in understanding and addressing these challenges, yet current monitoring methods, such as national nutrition surveys, are not practical on a continuous (e.g. annual) basis. Supermarket sales data could fill a gap by providing detailed, timely and inexpensive food and nutrition information helpful in developing policies and predicting future trends. This dissertation aims to examine the feasibility and applicability of supermarket sales data in population food and nutrition monitoring; and includes two components: (1) a literature review of published research that has used supermarket sales data for various food and nutrition monitoring purposes, and (2) a research proposal involving use of supermarket sales data to evaluate ethnic and socioeconomic differences in food purchasing patterns in New Zealand and the multilevel determinants of any such differences. A systematic review of the literature was undertaken up to April 2006. Forty seven studies were identified, of which four were feasibility studies, 19 assessed population food purchasing patterns, and 24 evaluated nutrition interventions. Four feasibility studies showed promising results for using sales data (scanned data, commercially available datasets and grocery receipts) to assess population food purchasing patterns. Six main purposes of using supermarket sales data observed in this review were to assess the impact of policy changes on population food purchases; to examine the effect of food pricing on healthier food choices; to compare food and nutrient purchase patterns between population groups; to validate food consumption data; to identify factors related to food purchasing behaviours; and to evaluate nutrition interventions. Based on the findings of the reviews, it appears that supermarket sales data have great potential for supplementing population food and nutrition monitoring. Further research is recommended to address some methodological limitations such as restricted data coverage and issues related to capturing, coding and analysing of sales data. The findings of this review have been accepted for publication in an international journal and presented at an international conference. Based on the review, a proposal was written for research that uses electronic supermarket sales data to evaluate ethnic and socioeconomic differences in food purchase patterns and the multilevel determinants of any such differences. Such information is clearly important in New Zealand where the burden of suboptimal nutrition in terms of reduced quality of life, increased morbidity, premature mortality and associated health care costs is substantial and increasing, with a disproportionate burden falling on Maori, Pacific and low socioeconomic group. Since most New Zealand household food purchases are made at supermarkets, the proposed study will inform the development of effective policies, programmes and services to improve nutrition of these priority groups in a more focused and efficient manner.