Professor Caroline Miller, Dr Jo Dono and Dr Kerry Ettridge (SAHMRI and The University of Adelaide)
Australia has one of the highest rates of sugar-sweetened beverage (SSB) consumption in the world. It is now well known that there are long term health impacts of consuming SSBs, including weight gain, Type 2 Diabetes, cardio-vascular disease and tooth decay.
Sugary drinks are available everywhere, heavily marketed, and consumption is normalised. Consequently, SSBs are not recognised as a problem by those who consume them regularly. Consumers may be increasingly aware that too much sugar is harmful to health. However, our earlier research has shown that people lack awareness of exactly how much sugar is in SSBs, and they have difficulty understanding abstract nutrition information. Australia has the Health Star Rating (HSR) system to help consumers navigate complex nutrition information, but the system is voluntary and has had relatively low uptake by beverage producers.
SSBs have no nutritional value and consumption is driven by industry practices that have little regard for health consequences. The parallel to cigarette consumption is striking. It took a systematic dismantling of industry influence through regulatory channels to reduce smoking rates, using strategies such as banning advertising, tax increases and mandating warning labels on packets. It is likely that a similar approach is needed to curb SSB consumption at the population level.
Our research
We have been investigating the possibility of applying warning labels to drinks high in sugar as a strategy to reduce consumption. We have conducted multiple studies to test and evaluate a range of message options that could be used as warning labels on bottles of SSBs. From these studies we were able to identify the best performing messages to test in an experimental study with real decision-making stakes.
The paper, recently published in Appetite, found that warning labels showing a range of messages reduced young adults’ selection of SSBs and promoted substitution to water. Furthermore, including a HSR as well as a warning label reinforced this effect for the least healthy drinks. We tested five different warning labels (e.g. health effects, sugar content) but we did not find a difference by label type.
The way we tested change in consumer behaviour was to develop an ‘online convenience store’ app that participants could access from a laptop, set-up on a university campus during orientation week. Participants (n=511, average age of 22 years) were informed that we were trialling the app, and that they would be asked to complete a drink selection task three times, each time with a new piece of information added to the drinks. They were also told that they would get to keep one of the drinks that they had selected to ensure that they were making a decision that had real consequences.
The app displayed 10 drinks to choose from that ranged in sugar content: five SSBs, one 100% fruit juice, two artificially sweetened beverages (ASBs) and two waters. Three drink selection tasks were given to participants:
- The first drink choice task had no intervention, meaning participants saw and chose drinks as they currently exist without a warning label or HSR.
- The display for the second drink choice task changed so that a warning label was placed on drinks high in sugar (five SSBs and one 100% fruit juice).
- The warning labels were retained for the third drink choice task, with a HSR added to all drinks.
On the first drink selection task, SSBs were the most popular choice and waters were the least popular choice. As noted above, the addition of warning labels in round two reduced the selection of SSBs and increased the selection of water, with this selection change maintained in round three with the addition of HSR.
There were some other interesting patterns. Participants increased their selection of ASBs in round three when warning labels were added to SSBs, but not ASBs. There were gender differences in drink selections associated with 100% fruit juice. Only females reduced their selection of 100% fruit juice following the addition of a warning label, but both males and females increased their selection when a four-star HSR label was added in round three.
Implications
One implication of these findings is that an intervention is justified: without any additional information, people will select the least healthy choice. Another implication is that both warning labels and HSR can be used in conjunction to advise against SSB consumption. The different systems operate with different objectives but are compatible, provided that the messaging is consistent.
For SSBs, the message was consistent, with the 0.5 HSR reinforcing the warning label message that the drink was an unhealthy choice. For 100% fruit juice, the messages were inconsistent because it attracted a warning label for being over the sugar content threshold but had an HSR of 4, conveying that it is a healthier option relative to lower rated drinks. The algorithm for the HSR has since been revised, but the 100% fruit juice used in this study would still score a 3.5. Further research is needed to test what range of scores may serve as a health endorsement even in the presence of a warning label, and how to increase substitution to water rather than ASBs and 100% fruit juice.
This study is part of a broader program of work that is informing Australia’s food packaging labelling strategies. We are continuing to explore strategies to reduce SSB consumption, and investigate the policy and practical implications of substitution effects to drinks other than water.
Investigator team: Professor Caroline Miller, Dr Jo Dono, Dr Kerry Ettridge, Professor Simone Pettigrew, Professor Gary Wittert, Professor Melanie Wakefield, Professor John Coveney, Professor David Roder, Dr Jane Martin, Dr Aimee Brownbill
Image: Nik Albert/Unsplash