Mis-match Between the Healthy Food and the Ultra-processed Food Classifications in Australia (OR14-03-19)

Abstract

ObjectivesMany countries produce guides to healthy food choices for their populations. Healthy food choices are typically based on ensuring micronutrient adequacy while limiting components such as saturated fat, sugar and sodium. It has also been proposed that the degree of food processing can be used as the criterion for classification either for advice to the public or for making regulatory decisions around which food should be allowed to be voluntarily fortified or carry health claims. This study examined the concordance between two such systems. MethodsThe food database used to analyze intakes reported in the 2011–12 Australian National Nutrition and Physical Activity Survey has been classified by others to identify a) which foods are recommended or ‘discretionary’ in the Guide to Healthy Eating and b) by degree of processing according to the four levels in the NOVA system. The NOVA classifications were collapsed to compare ‘ultra-processed’ foods to the other categories. ResultsThere were 5645 foods in the database after excluding codes used to generate recipes for fortified foods (e.g., vitamins and minerals) and foods added for the 2013 Indigenous wave of the survey. Of the 4014 recommended foods, 23.5% were classed as ultra-processed whereas 31.2% of the discretionary foods were classified as not ultra-processed. In this dataset, nearly one-quarter of the foods would be recommended to the public by one classification but not by the other. Some notable discrepancies were the classification of recommended foods such as margarine and mass produced bread and buns and breakfast cereals as ultra-processed whereas butter, cream, sugar, honey, and homemade cakes, biscuits and jams are not classed as ultra-processed. It should also be noted that the descriptions in classification systems are often imprecisely worded and so some decisions in the two dataset could be debated. ConclusionsThe survey database serves as a useful starting point to screen possible tools even though it contains averages rather than brand-level information. These two specific classifications do not lead to equivalent advice about which foods to choose. Funding SourcesNone.