Potential impact of the adult-targeted LiveLighter “Sugary Drinks” campaign on adolescent consumption: Findings from a national cross-sectional school survey
Abstract
Issue addressed
To examine the potential impact on trends in reported adolescent sugar-sweetened beverage (SSB) consumption following the adult-targeted Western Australian (WA) LiveLighter “Sugary Drinks” campaign, which is a television-led healthy lifestyle promotion and education mass media campaign that aired in multiple bursts between July 2013 and December 2018.
Methods
Data were from a repeated cross-sectional survey of Australian secondary school students aged 12-17 years conducted nation-wide in 2009-2010 (n = 13 790 from 238 schools), 2012-2013 (n = 10 309 from 196 schools) and 2018 (n = 9102 from 104 schools).
Results
Significant declines in high SSB consumption (≥4 cups/week) were observed between 2009-2010 and 2012-2013 among students in both WA (27.1% vs 20.3%; odds ratio [OR] = 0.65, 95% confidence interval [CI] = 0.50-0.84, P = .001) and all other states and territories combined (30.6% vs 21.4%; OR = 0.60, 95% CI = 0.53-0.68, P < .001). Compared to 2012-2013, the proportion of students in WA who reported high SSB consumption was significantly lower in 2018 (20.3% vs 9.6%; OR = 0.37, 95% CI = 0.21-0.65, P = .001), whereas this significant decrease was not observed in all other states and territories combined (21.4% vs 17.8%; OR = 0.81, 95% CI = 0.67-0.97, P = .024). Interaction testing indicated that the recent decline observed in WA was more pronounced among female than male students.
Conclusions
Mass media campaigns targeted towards the Australian adult population may have favourable effects among adolescents with the steeper decline in high SSB consumption observed in WA between 2012-2013 and 2018 coinciding with the repeated airing of the LiveLighter “Sugary Drinks” campaign in this state.
So what?
Government investment in mass media campaigns may hasten declines in adolescents' SSB consumption.
1 INTRODUCTION
Dietary habits formed during adolescence shape the health and wellbeing of individuals throughout life.1 However, many adolescents do not consume a nutritious, healthy diet2 and consequently one in four Australians aged 2-17 years are considered to be above a healthy weight.3 High consumption of discretionary food and drinks (ie, energy-dense and nutrient-poor products) is known to play a role in the development of overweight and obesity,2, 3 particularly sugar-sweetened beverages (SSBs).2, 4 Discretionary food and drinks made up 41% of total energy intake among Australians aged 14-18 years in 2018, and SSBs alone made up approximately a third of daily intake of added and free sugars.2 The World Health Organization has identified the reduction of SSB consumption as a priority in the prevention of childhood overweight and obesity.5
Numerous strategies can be employed to reduce population consumption of SSBs, such as the introduction of tax on these products6 and restrictions on the way that the ultra-processed food industry can market their products to children.7 Mass media campaigns can also change population-level behaviours given their ability to reach large proportions of the community through multiple channels.8 In 2012, the Western Australian (WA) Government launched the LiveLighter mass media healthy lifestyle promotion and education campaign that illustrates the health consequences of carrying extra weight using graphic imagery and encourages the adoption of simple changes towards leading a healthier lifestyle. In July 2013, the “Sugary Drinks” phase of the campaign commenced, which had a specific focus on reducing consumption of SSBs. Population-based evaluation of the “Sugary Drinks” campaign among the target audience of adults, using cross-sectional and controlled cohort designs, has provided evidence of reduced SSB consumption associated with the campaign.9, 10 However, little is known about the impact of the campaign among adolescents. In tobacco control, mass media campaigns targeted towards the general adult population have been shown to have favourable effects on reducing likelihood of smoking11, 12 and increasing negative attitudes towards tobacco use13, 14 among adolescents. To our knowledge, this effect has not previously been reported in the field of diet-related behaviour change.
The aim of the current study was to explore the potential impact of the WA LiveLighter “Sugary Drinks” campaign by comparing differences in adolescents' SSB consumption over time in WA compared to all other Australian states and territories combined using data collected as part of the National Secondary Students' Diet and Activity (NaSSDA) repeated cross-sectional survey.
2 METHODS
2.1 LiveLighter “Sugary Drinks” campaign
The LiveLighter “Sugary Drinks” campaign was launched in WA in July 2013 (after the completion of the 2012-2013 NaSSDA survey round) and focuses on the contribution of excessive sugary drink consumption to the development of “toxic fat” and ultimately disease.10 Targeted towards adults aged 25-64 years, the campaign comprised a television advertisement (30- and 15-s versions) depicting a man entering a convenience store and buying a sugary drink. As he goes outside and begins drinking, the visuals go inside his body to show fat around his organs, with a voiceover explaining that “The sugar in any sugary drink is sugar your body doesn't need.” A total of seven bursts of the WA LiveLighter “Sugary Drinks” campaign aired in the intervening period between the 2012-2013 and 2018 NaSSDA survey years, including a campaign burst comprising 6 weeks of television advertising that achieved 145 average weekly Target Audience Rating Points (TARPs) in the first half of 2018 prior to the commencement of NaSSDA data collection in 2018. TARPs are a standard measure of television media exposure indicating the reach (percentage of the target audience exposed) and frequency (number of times exposed) of the campaign burst,15 with research in tobacco control indicating that campaign exposures of between 100 and 200 TARPs per week are optimal.15 The regular stream of paid mass media campaign activity addressing SSB consumption over an extended time period in WA has not been replicated in other Australian states and territories.
2.2 Study design and sample
Nationally representative cross-sectional samples of Australian adolescents completed web-based questionnaires as part of the NaSSDA survey conducted in 2009-2010, 2012-2013 and 2018. Using a two-stage probability design, random selection of schools stratified by education sector (government, Catholic and independent) was followed by random selection of classes. Students were eligible to participate if they were in year levels 8-11 (ages 12-17 years) and had received active written parental consent (or passive/opt-out in selected jurisdictions in 2018 where this was acceptable to the relevant education authority). The surveys were administered by an independent data collection agency, with written assent provided by all participating students.
2.3 Measures
To assess consumption of SSB, students were asked “How much soft drinks (like Coke, lemonade), cordials or sports drinks (like Gatorade) do you usually drink? (1 cup = 250 ml, one can of soft drink = 1 ½ cups). Do not include diet soft drinks, diet cordials, or diet sports drinks.” Response options included: “I don't drink soft drinks, cordials or sports drinks”; “less than one cup a month”; “about 1-3 cups a month”; “about 1-3 cups a week”; “about 4-6 cups a week”; “about 1-2 cups a day”; “about 3-4 cups a day”; and “5 cups or more a day.” Students who reported consuming four or more cups per week (equivalent to at least 1 L) were classified as high SSB consumers based on limits set in previous studies using NaSSDA survey data.16-19
Students self-reported their sex, year level and residential postcode, which was used to compute a measure of socio-economic area according to the Socio-Economic Index for Areas (SEIFA) Index of Relative Socio-economic Disadvantage.20-22 Using the national deciles to create quintiles, students were categorised into low (first and second quintiles), mid (third and fourth quintiles) and high (fifth quintile) socio-economic area groups. Residential postcode was also used to classify student geographic location (metropolitan or regional/remote) according to the Australian Statistical Geography Standard Remoteness Structure.23, 24
2.4 Statistical analysis
Data were analysed using Stata/MP 16.1 (StataCorp) and restricted to male and female students aged between 12 and 17 years who had completed at least one-third of the questionnaire. Each sample was weighted by state, education sector, year level and sex to bring it in line with the Australian student population.25-27 Information on study sample sizes, response rates and demographic characteristics of participants for each survey year are summarised in Table 1. All analyses accounted for clustering of students within each surveyed school. Unadjusted prevalence estimates are reported. Multivariable logistic regression with an interaction term was used to determine whether change over time in the prevalence of high SSB consumption varied by state (WA vs all other states and territories combined) after controlling for sex, year level, socio-economic area, geographic location and education sector. Further interaction testing was conducted to assess whether trends in high SSB consumption within each state/territory category differed based on students' sex, given that consumption levels are typically higher among adolescent males.16, 28-30 A significance level of P < .20 was accepted for interaction tests,31 while P < .01 was accepted for all other analyses except those using the smaller WA sub-sample (P < .05).
| Survey year | ||||||
|---|---|---|---|---|---|---|
| 2009-2010 | 2012-2013 | 2018 | ||||
| Western Australia | All other states/territories | Western Australia | All other states/territories | Western Australia | All other states/territories | |
| Number of students | 1501 | 12 289 | 1406 | 8903 | 609 | 8493 |
| Number of schools | 30 | 208 | 26 | 170 | 9 | 95 |
| Student response rate | 51.5% | 54.4% | 53.8% | 52.9% | 62.1% | 67.3% |
| School response rate | 35.6% | 39.4% | 21.7% | 20.4% | 6.5% | 8.4% |
| Sex | ||||||
| Male | 51.3% | 50.6% | 51.5% | 51.5% | 51.2% | 51.9% |
| Female | 48.7% | 49.4% | 48.5% | 48.5% | 48.8% | 48.1% |
| Socio-economic areaa | ||||||
| Low (1%-40%) | 29.1% | 38.6% | 28.5% | 40.1% | 6.4% | 29.4% |
| Mid (41%-80%) | 36.8% | 39.6% | 43.8% | 35.7% | 56.9% | 46.7% |
| High (81%-100%) | 34.1% | 21.7% | 27.8% | 24.2% | 36.7% | 23.9% |
| Year level | ||||||
| 8 | 25.0% | 25.6% | 27.9% | 25.6% | 26.3% | 26.4% |
| 9 | 25.5% | 26.2% | 27.9% | 25.4% | 25.4% | 26.3% |
| 10 | 25.5% | 25.7% | 27.1% | 25.6% | 24.7% | 25.5% |
| 11 | 23.9% | 22.6% | 17.1% | 23.4% | 23.6% | 21.8% |
| Geographic locationb | ||||||
| Metropolitan | 64.2% | 59.2% | 59.5% | 65.3% | 96.2% | 66.4% |
| Regional/remote | 35.8% | 40.8% | 40.5% | 34.7% | 3.8% | 33.6% |
- Note: Percentages are based on weighted data and may not sum to 100% due to rounding.
- a Socio-economic area is missing where students did not report a valid home postcode or no SEIFA Index of Relative Socio-economic Disadvantage value was available for their home postcode (2009-2010: n = 18 for WA and n = 85 for all other states/territories; 2012-2013: n = 2 for WA and n = 12 for all other states/territories).
- b Geographic location is missing where students did not report a valid home postcode or no remoteness area code was available for their home postcode (2009-2010: n = 1 for WA and n = 69 for all other states/territories; 2012-2013: n = 5 for all other states/territories).
3 RESULTS
In 2018, 1 in 10 WA students reported high SSB consumption (≥4 cups/week) compared to 18% of students from all other states and territories. Interaction tests suggested that change over time in the prevalence of high SSB consumption differed between WA and all other states and territories (P = .073). Among participants residing in WA, the proportion of students who reported high SSB consumption decreased significantly from 2009-2010 to 2012-2013 (27% vs 20%; odds ratio [OR] = 0.65, 95% confidence interval [CI] = 0.50-0.84, P = .001) and 2012-2013 to 2018 (20% vs 10%; OR = 0.37, 95% CI = 0.21-0.65, P = .001; Table 2). While the prevalence of high SSB consumption in all other states and territories also decreased to a significant extent from 2009-2010 to 2012-2013 (31% vs 21%; OR = 0.60, 95% CI = 0.53-0.68, P < .001), there was no statistically significant change observed between 2012-2013 and 2018 (21% vs 18%; OR = 0.81, 95% CI = 0.67-0.97, P = .024).
| Prevalence (%)a | 2012-2013 (ref 2009-2010) | 2018 (ref 2012-2013) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 2009-2010 | 2012-2013 | 2018 | Adj. ORb | 95% CI | P value | Adj. ORb | 95% CI | P value | |
| Western Australia | |||||||||
| Males | 34.6 | 26.1 | 17.0 | 0.65 | 0.47-0.91 | .013 | 0.49 | 0.26-0.90 | .023 |
| Females | 19.4 | 13.7 | 1.9 | 0.67 | 0.49-0.90 | .010 | 0.11 | 0.05-0.26 | <.001 |
| Total | 27.1 | 20.3 | 9.6 | 0.65 | 0.50-0.84 | .001 | 0.37 | 0.21-0.65 | .001 |
| All other states and territories | |||||||||
| Males | 38.4 | 27.9 | 23.0 | 0.61 | 0.53-0.70 | <.001 | 0.80 | 0.66-0.97 | .024 |
| Females | 22.6 | 14.6 | 12.1 | 0.58 | 0.48-0.72 | <.001 | 0.82 | 0.61-1.10 | .179 |
| Total | 30.6 | 21.4 | 17.8 | 0.60 | 0.53-0.68 | <.001 | 0.81 | 0.67-0.97 | .024 |
- a Unadjusted prevalence estimates.
- b Odds ratios adjusted for sex, year level, socio-economic area, geographic location and education sector.
Among students residing in all other states and territories excluding WA, the change in high SSB consumption between 2012-2013 and 2018 was nonsignificant for both males (28% vs 23%; OR = 0.80, 95% CI = 0.66-0.97, P = .024) and females (15% vs 12%; OR = 0.82, 95% CI = 0.61-1.10, P = .179). However, among students residing in WA, the statistically significant decrease in the proportion of students who reported high SSB consumption during this time period was observed to a greater extent among females (14% vs 2%; OR = 0.11, 95% CI = 0.05-0.26, P < .001) compared to males (26% vs 17%; OR = 0.49, 95% CI = 0.26-0.90, P = .023).
4 DISCUSSION
Overall, the results from the NaSSDA survey indicate that there have been steeper declines in high SSB consumption among adolescents residing in WA compared to all other states and territories combined. Specifically, the proportion of WA students reporting high SSB consumption (≥4 cups/week) halved between 2012-2013 and 2018, coinciding with the repeated airing of the LiveLighter “Sugary Drinks” campaign in this state, while the rate of decline in high SSB consumption across the rest of Australia appeared to slow during this same period.
While adolescents are not the target audience of the WA LiveLighter campaign, the current study elucidates a potential effect of the campaign on SSB consumption among this population group. This replicates previous findings of spill-over effects among adolescents of adult-targeted tobacco control mass media campaigns.11-14 Various hypothesised effect mechanisms may explain this pattern of results. First, although the media buy for the campaign specifically targets adults aged 25 years and over, adolescents engage with many of the media channels utilised by the campaign, including commercial television, out-of-home and digital media, and may be exposed to campaign messages. Second, parents make up a substantial proportion of the WA LiveLighter campaign target audience. This may have direct effects on adolescent SSB consumption by encouraging parents to limit the amount of SSBs they purchase or provide for their children.32 Furthermore, indirect effects may exist, for example through influencing social norms related to attitudes and dietary intentions and behaviours.33, 34
The decline in high SSB consumption in WA was driven to a larger extent by female adolescents, while the decline observed among male adolescents was less pronounced, which reflects previous studies showing variation in SSB consumption by sex.16, 28-30 This may be related to the tendency for sugary drink advertising to employ themes of masculinity, particularly for sports and energy drinks.35 Alternatively, this could be explained by a greater uptake of artificially sweetened diet drinks displacing the consumption of SSBs, which tends to be more common among females.36
In interpreting the findings of the current study, caution should be applied due to the cross-sectional nature of the study design, which inhibits the ability to make causal inferences. The differential declines in high SSB consumption by state/territory may have been related to external unidentified factors. These may include concurrent public health campaigns or changes to school food environments, including as a result of the introduction of the WA Healthy Food and Drink Policy for school canteens in 2014,37 which tends to impose more stringent guidelines comparative to policies in other Australian jurisdictions.38 The study is also limited by the use of a self-report measure of SSB consumption and is hence subject to social desirability bias or recall bias. This may have resulted in an underestimation of the proportion of students classified as high SSB consumers39; however, this potential bias is unlikely to differ across survey year or state/territory. Future research would benefit from objective measures of SSB consumption. Finally, the representativeness of the 2018 sample was compromised by a lower school response rate comparative to previous survey years, although this applied equally to WA and other jurisdictions, so does not explain the differential change observed in WA. Those residing in low socio-economic and regional/remote areas were under-represented in the WA sample in 2018, but not in the sample of participants residing in all other Australian states and territories combined. While this may have resulted in an underrepresentation of frequent SSB consumers in WA in 2018 given that prevalence of high SSB consumption tends to be greater in lower socio-economic areas,40 all analyses controlled for socio-economic area and geographic location to account for potential biases. Nevertheless, we acknowledge residual confounding may be present and we encourage future research to further investigate the impact upon adolescents of population-wide campaigns primarily aimed at the adult population.
5 CONCLUSIONS
The current study suggests that mass media campaigns targeted towards the general Australian adult population may have favourable effects among adolescents regarding their own consumption of SSBs. Long-term substantial government investment in a LiveLighter “Sugary Drinks” style of campaign at a national level may hasten declines in Australian adolescents' SSB consumption.
AUTHOR CONTRIBUTIONS
All authors have contributed to the present paper through involvement in the conception and design of the study or in analysis and interpretation of the data. All authors were involved in writing and revising the paper and have approved the final article.
ACKNOWLEDGEMENTS
The authors thank the school principals, teachers and students who participated in the study, as well as Roy Morgan Research (2018) and I-view (2009-2010 and 2012-2013) who were responsible for fieldwork coordination. The National Secondary Students' Diet and Activity (NaSSDA) survey is principally funded by state and territory Cancer Councils through Cancer Council Australia, with additional funding support received from the National Heart Foundation of Australia for the 2009-2010 and 2012-2013 survey rounds, the State and Territory Government Health Departments for the 2012-2013 survey round, and from the South Australian Health and Medical Research Institute for the 2018 survey round. The funders were not involved in the study design, data collection, data analysis, data interpretation, or writing of the paper.
CONFLICT OF INTEREST
All authors are currently employed by an organisation involved in public health research and advocacy.
ETHICS STATEMENT
The NaSSDA study was approved by the Human Research Ethics Committee of Cancer Council Victoria (HREC 1117) and conducted in accordance with the Declaration of Helsinki.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
