Reassessing the Health Star Rating: New study highlights need for ultra-processed food adjustments

By Hugo Francisco de SouzaReviewed by Benedette Cuffari, M.Sc.Aug 21 2024

In a recent study published in Nutrition & Dietetics, researchers reassess the Australian and New Zealand Health Star Rating (HSR) given the increasing prevalence of ultra-processed foods.

Study: Modifying the Health Star Rating nutrient profiling algorithm to account for ultra‐processing. Image Credit: sasirin pamai / Shutterstock.com

The health impact of ultra-processed foods

Chronic non-communicable diseases account for almost 75% of all human mortality, with diet representing the foremost modifiable risk factor at 26% to combat these ailments. The rapid and unprecedented influx of industrially manufactured ultra-processed foods has significantly contributed to diet-related mortality, with recent reports estimating that these foods comprise 30-60% of calorific intakes.

Ultra-processed foods are produced using multiple chemical and physical processes that alter the nutritional profiles of their constituent ingredients. These food products often contain significantly increased concentrations of added sugars, salts, and chemical additives such as monosodium glutamate (MSG), which are rarely used in homemade foods.

Alarmingly, a growing body of research suggests that the physiological demerits of ultra-processed foods may extend beyond their ingredients and processing techniques. These findings have prompted the public health policymakers of at least seven nations to recommend limited intakes of these harmful foods.

The HSR system is a nutrition labeling system used in Australia and New Zealand to help consumers make informed choices on the healthiness of packaged food products. The rating ranges from 0.5, which reflects the least healthy food, to five stars.

About the study

The present study aims to update the HSR algorithm to account for the health demerits of ultra-processed food. Data for the study were obtained from the 2022 FoodSwitch dataset comprising 35,645 barcoded food products that represented over 90% of barcoded packed foods in Australia. Vitamins, supplements, alcohol, herbs and spices, and infant formula not covered by the HSR system were excluded from the analysis.

The NOVA classification approach was used to identify cosmetic additives or industrial food substances in products’ ingredient lists as proxies for ultra-processed foods. All anti-caking or glazing agents, color, emulsifiers, flavor enhancers, sweeteners such as dextrose, fructose, and maltitol, protein powders, and food isolates, including lactose, triglycerides, and wheat gluten, were categorized as ultra-processed.

The researchers tested four modifications, which were categorized into inclusion, capping, and hybrid approaches. The inclusion approach, which comprised modification 1 (m-HSR), added five negative HSR points for ultra-processed foods irrespective of their NOVA classification.

The capping approach included m-HSR 2 and restricted ultra-processed foods to a maximum HSR of three. The hybrid approach included both m-HSR 3 and m-HSR 4 and was only applied when the product had 10 or more baseline points.

These modifications were then compared to the original HSR system. Subsequently, the alignment of the modified HSR system to the NOVA classification system and the Australian Dietary Guidelines was calculated.

Study findings

After excluding products not included in the original HSR system, 25,486 food products were analyzed, 64% of which were categorized as ultra-processed foods. About 96% of all included confectionery products were considered ultra-processed.

The m-HSR 1 changed the median HSR from three to two, whereas m-HSR 3 changed the rating from three to 2.5. No statistically significant changes were observed in median HSRs across m-HSR 2 and 4.

Of the six major categories included in the NOVA classification system, median HSRs of ‘egg and egg products’ and ‘sugar and honey products’ remained unchanged. All other categories experienced declines in the modified HSR algorithm as compared to the original rating scheme.

Alignment calculations revealed that all m-HSRs increased alignment with the NOVA classification system from 61% to 81-88%. In contrast, none of the m-HSRs increased alignment with the Australian Dietary Guidelines classifications, with m-HSR 1 and 2 instead reducing alignment from 77% to 72% and 69%, respectively.

Conclusions

Given the alarmingly high prevalence of ultra-processed foods in packaged foods, it is crucial to modify the HSR system to help consumers make better-informed choices. The study findings also highlight the need for Australia to update its Dietary Guidelines to account for these relatively harmful food products, as modifying the HSR without updating these guidelines may lack coherency and confuse consumers about what foods would benefit their health.