Classification of 'healthier' and 'less healthy' supermarket foods by two Australasian nutrient profiling models.

AIM: To determine whether a modified version of the Heart Foundation Tick (MHFT) nutrient profiling model appropriately classifies supermarket foods to endorse its use for identifying 'healthier' products eligible for promotion in a supermarket intervention trial. METHODS: Top-selling products (n=550) were selected from an existing supermarket nutrient composition database. Percentage of products classified as 'healthier' by the MHFT and a modified comparator model (Food Standards Australia New Zealand; MFSANZ) were calculated. Percentage agreement, consistency (kappa statistic), and average nutrient values were assessed overall, and across seven food groups. RESULTS: The MHFT model categorised 16% fewer products as 'healthier' than the MFSANZ model. Agreement and consistency between models were 72% and kappa=0.46 (P=0.00), respectively. For both models, 'healthier' products were on average lower in energy, protein, saturated fat, sugar, and sodium than their 'less healthy' counterparts. CONCLUSION: The MHFT nutrient profiling model categorised regularly purchased supermarket foods similarly to the MFSANZ model, and both appear to distinguish appropriately between 'healthier' and 'less healthy' options. Therefore, both models have the potential to appropriately identify 'healthier' foods for promotion and positively influence food choices.