Healthy or Unhealthy? Health Star Ratings and our food choices

What is an unhealthy food? Usually, identifying ‘unhealthy’ foods is straight forward.  Most would agree foods such as potato chips, meat pies, chocolate and cakes should be avoided in a healthy diet. But what about a high-protein energy bar? Or a frozen ready-to-heat meal? This is where identifying healthier options becomes less clear cut.

The Australian Dietary Guidelines discourage the consumption of ‘discretionary’ choices, described as energy-dense foods and drinks, high in saturated fat, added sugars and/or salt.

Alternatively, foods can be identified as ‘unhealthy’ based on the extent and purpose of food processing using the NOVA classification scheme. The term ‘ultra-processed’ has been coined to described foods that should be avoided in a healthy diet. These are products formulated from industrial processes or that contain industrially derived ingredients. 

The logo and slogan of the Health Star Rating System

Only 4% of Australians meet dietary recommendations, and on average Australians consume 35% of daily energy intake from discretionary foods and 42% from ultra-processed foods. This level of intake is likely influenced by the current Australian food marketing environment. For example, over half the packaged food supply is composed of discretionary and ultra-processed foods, and discretionary foods are price promoted twice as often as healthier foods in supermarkets. 

The Health Star Rating

The World Health Organization recommend front-of-pack labelling as part of a suite of policy tools to address poor diets and associated non-communicable diseases. Presenting simple, easy-to-interpret graphical information, front-of-pack labelling allows consumers to quickly evaluate the healthfulness of packaged products at the point of purchase. In 2014, the Australian federal government implemented the Health Star Rating system. The label allocates products a rating from 0.5 to 5 ‘health stars’ based on ‘risk’ nutrients (energy, total sugars, saturated fat, and salt) and ‘beneficial’ components (fibre, protein, and fruit/vegetable/nut/legume content). The nutrient approach informing the Health Star Rating is termed ‘nutrient profiling’, and is commonly used in front-of-pack labelling and other nutrition policies such as restrictions to marketing and the regulation of health claims.

Is Milo a healthy food?

 The Health Star Rating system has caused controversy since its implementation with many products considered ‘junk foods’ able to receive relatively high ratings. A much-publicised example is the 4.5-star rating allocated to Nestle’s Milo. Nestle took advantage of the ‘as prepared’ loophole, calculating the product’s rating based on two teaspoons of the sugary drink base added to a glass of skim milk. This technical oversight in the implementation has now been amended. However other examples of ‘anomalies’, such as choc chip muesli bars and flavoured icy poles, point to more fundamental issues with the underlying nutrient approach.

Nutrient profiling is a simple way to assess the healthfulness of foods, and hence an appealing approach to apply to policy. However, it assumes the healthfulness of a food can be predicted by summing the individual parts. Yet foods are more complex than a handful of positive and negative nutrients, in fact 26,000 distinct bio-active components have been identified in food. The nutrient approach not only ignores these components, but how they may interact. Food synergies both between components in a food and between foods in a diet also affect how a food is metabolised, in addition to the physical food structure in which the components are packaged. For example, the juice from an orange is metabolised differently from the juice contained in a whole orange.

Nutrients, foods, or diets?

Conceptually foods can be classified as ‘healthy’ using indices underpinned by one of three approaches: nutrients, foods or diets. In our recently published research, my colleagues and I evaluated the Health Star Rating system against a food-based index, NOVA, and a diet-based index informed by the Australian Dietary Guidelines. We analysed 4251 newly released food products displaying the Health Star Rating label from June 2014 (the system’s introduction), and June 2019. We were particularly interested in the star ratings being displayed on ultra-processed and discretionary products, the non-recommended foods according to NOVA and the Australian Dietary Guidelines. We found 73% of ultra-processed foods and 53% of discretionary foods were displaying a Health Star Rating of 2.5 or higher (a logical pass mark). Our results clearly show the nutrient-based approach informing the Health Star Rating system is providing a ‘health halo’ to junk foods.

Where to from here?

In July 2020, the Food Ministerial Forum on Food Regulation confirmed amendments to the Health Star Rating Calculator to improve alignment with the Australian Dietary Guidelines. However, minor tweaks to the algorithm are unlikely to change the substantial misalignment observed in our study. We propose a food’s healthfulness should be determined by more than the sum of the nutrients it contains, and incorporate aspects of food processing and diet-based indices, better reflecting nutrition science principles. Through my PhD research I aim to understand how this can be achieved and incorporated in the design of future polices.

Sarah Dickie is a PhD candidate at the Institute for Physical Activity and Nutrition (IPAN) and tutors within public health nutrition units at Deakin University. Sarah’s PhD analyses the politics and science underpinning the use of nutrition classification schemes in policy actions. Her research to date has focused on evaluating the implementation of Australia’s Health Star Rating front-of-pack labelling system.



Join the conversation

back to top