Abstract
Background
To address concerns that healthy weight and lifestyle campaigns (HWLCs) could have unintended psychological or behavioral consequences, this study tested reactions to TV advertisements from two HWLCs: LiveLighter® (employs graphic health effects messaging) and Swap It Don’t Stop It (employs animation and light-hearted messaging).
Methods
An online between-subjects experiment tested reactions to one of five advertisements: “Toxic Fat”; “Sugary Drinks” (both from LiveLighter); “Become a Swapper”; “How to Swap It” (both from Swap It); and “HSBC Bank” (control) among 2,208 adults from Western Australia and Victoria, Australia. Responses assessed were cognitive and emotional reactions, behavioral intentions, internalized weight bias, antifat attitudes, self-esteem and body dissatisfaction.
Results
The HWLC advertisements prompted favorable cognitive reactions and weak to neutral positive and negative emotional responses. HWLC advertisements promoted stronger intentions to engage in adaptive lifestyle behaviors compared with the control advertisement. Intention to engage in maladaptive behaviors (e.g., skipping meals) was low overall and did not differ by condition. Compared with the control condition, participants who saw LiveLighter “Sugary Drinks” showed weaker disagreement with antifat attitudes; however, mean antifat attitude scores were still at the low end of the scale. HWLC advertisements did not promote internalized weight bias, reduced self-esteem or body dissatisfaction compared with the control advertisement.
Conclusions
Overall, HWLC advertisements performed favorably compared with the control advertisement and showed no clear evidence of adverse impacts. Findings suggest HWLC advertisements may not promote negative psychological and behavioral consequences and can continue to be used in obesity prevention efforts.
Excess consumption of energy-dense, nutrient-poor foods and beverages and insufficient physical activity are important in the development and maintenance of obesity and associated noncommunicable diseases (Hruby et al., 2016; Millar et al., 2014; Sui et al., 2017). As most Australians’ (>80%) physical activity levels and diets fall short of national guidelines, and 67% of adults and 25% of children aged 2 to 17 years are overweight or obese (Australian Bureau of Statistics, 2018b; Jongenelis et al., 2018; Mihrshahi et al., 2019; National Health and Medical Research Council, 2013), experts have called for urgent investment in obesity prevention programs (Lung et al., 2020).
Comprehensive approaches are required to address obesity (Hawkes et al., 2015; Swinburn et al., 2019), with interventions encouraging healthier dietary practices and increased activity considered important elements of these approaches. The World Health Organization (WHO, 2013) recommends implementing evidence-based mass media campaigns to inform and motivate the public about the benefits of healthy eating and physical activity, and facilitate these behaviors. Well-designed campaigns of sufficient duration and intensity can produce positive changes in health-related behaviors across large populations (Wakefield et al., 2010). There is also a need for regulations to reduce the public’s exposure to competing advertising for energy-dense, nutrient-poor food products, which vastly overshadows advertising for healthy weight and lifestyle campaigns (HWLCs) (Kornfield et al., 2015).
A number of nations, including Australia, have implemented HWLCs employing mass media advertising as a centerpiece (Goodman et al., 2021; Kite, Grunseit, et al., 2018; Kornfield et al., 2015). According to the theory of planned behavior, HWLCs should initially influence attitudes and beliefs then intentions and behavior (Ajzen, 1991; Ajzen & Fishbein, 1980). A worldwide systematic review found adult-targeted HWLCs can impact intermediate outcomes (nutrition content knowledge, positive attitudes to weight gain prevention, self-efficacy beliefs, healthy eating intentions); however, evidence on behavioral impacts is limited (Kite, Grunseit, et al., 2018). Another systematic review of quantitative evaluations of HWLCs focusing on healthy eating concluded that those pairing “stop” (nonrecommended) and “go” (recommended) behaviors are the most successful in achieving behavior change (Abril & Dempsey, 2019).
In Australia, two campaigns highlighting “stop” behaviors (e.g., consuming energy-dense food) and modeling alternative “go” behaviors (e.g., choosing water instead of a sugary drink) have aired. The Australian Government’s Swap It campaign (2008–2011) used light-hearted animated messaging to encourage small health-related lifestyle changes, with evaluation showing modest campaign awareness and limited effect on behaviors (O’Hara et al., 2016). The Western Australian Government-funded LiveLighter campaign (2012-ongoing) presents graphic anatomical images of visceral fat to illustrate negative health effects of obesity as well as modeling recommended behaviors (Morley et al., 2016). Population-based surveys have found favorable changes in beliefs, intentions, and behaviors associated with the campaign (Morley et al., 2016, 2018, 2019).
Some authors have raised concerns that exposure to HWLCs, especially those that explicitly mention or depict obesity and/or use fear appeals with graphic imagery, may have negative psychological and behavioral consequences (Couch et al., 2018; Lupton, 2014; Pausé, 2017; Varshney, 2020). It has been argued that such HWLCs may stigmatize individuals with obesity and reinforce societal weight-based stigma (Carter et al., 2011; Johnstone & Grant, 2019; Puhl, Luedicke, & Peterson, 2013; Rubino et al., 2020), since exposure to weight-stigmatizing media content has been found to increase weight-biased attitudes and endorsement of negative weight-related stereotypes (Pearl et al., 2012). It has also been suggested that HWLCs focusing on obesity may increase internalized weight bias and reduce motivation to eat well, exercise and lose weight (Puhl et al., 2020; Vartanian et al., 2018). Internalized weight bias is associated with low self-esteem, eating disorder psychopathology, depressive symptoms, and body dissatisfaction (Pearl & Puhl, 2018; Wu & Berry, 2018).
HWLC messaging may have varying effects on people of different weight status. In U.S. studies, exposure to media content rated as stigmatizing lowered self-efficacy to engage in health behaviors for those with obesity and increased the magnitude of obesity-reinforcing behaviors (Puhl, Luedicke, & Peterson, 2013; Major et al., 2014). By contrast, for healthy-weight individuals, such communications increased self-efficacy in maintaining healthy weight and promoted greater intention to engage in healthy behaviors (Major et al., 2014; Young et al., 2016). Qualitative research with Australian adults suggested those furthest from achieving health recommendations for weight, diet and activity were motivated by “what” and “how” advertisements (ads) involving gentle persuasion and helpful hints, whereas those closer to meeting recommendations were motivated by “why” ads featuring more graphic and emotive content and new information (Dixon et al., 2016). However, an experiment assessing Australian adults’ cognitive and emotional reactions to HWLC ads found the LiveLighter “Toxic Fat” ad (which emphasized negative health consequences of overweight) outperformed ads employing a less graphic approach to encourage behavior change, with overweight/obese adults showing stronger perceptions of the effectiveness of the ads than healthy-weight adults (Dixon et al., 2015).
Concerns around potential stigmatizing effects of HWLCs highlight a need to monitor unintended impacts (Carter et al., 2011). Evaluation surveys have found no increase in endorsement of stereotypes of overweight individuals (a measure of social stigma) in response to LiveLighter (Morley et al., 2016, 2018, 2019). Although these findings are encouraging, the measures assessed have been necessarily parsimonious to allow monitoring of other outcomes. Focused studies using a wider range of validated measures to assess unintended impacts are critical to avoid counterproductive effects of HWLCs.
The aim of this study is to test potential intended and unintended reactions to ads from two HWLCs employing different content and executional styles: LiveLighter and Swap It. The following research questions are posed:
Method
Design and Stimuli
An online between-subjects experiment tested reactions to five ad conditions (four HWLC ads and one control ad). Two of the HWLC ads tested were from the LiveLighter campaign, which graphically portrays the negative health effects of obesity and demonstrates simple lifestyle changes to encourage healthier eating and increased activity. Two were from the Swap It campaign, which portrays obesity in a light-hearted way using animated figures to model recommended behaviors. The specific ads tested (depicted in Figure 1) were: (a) “Toxic Fat” (LiveLighter), (b) “Sugary Drinks” (LiveLighter), (c) “Become a Swapper” (Swap It), (d) “How to Swap It” (Swap It), and (e) HSBC Bank (control ad, unrelated to HWLC). The HWLC ads selected for testing were chosen because they exemplified the messaging of the respective campaigns.
Description of the advertisements tested accompanied by stills of key scenes.
Participants
The sample comprised 2,208 adults aged 18 to 64 years residing in Western Australia (WA) or Victoria (Australian states where LiveLighter and Swap It had previously aired). An ISO-accredited web panel provider, PureProfile, accessed the sample and administered the survey. PureProfile’s panel of Australians with diverse geographic and socioeconomic profiles has been established via multiple recruitment strategies, including internet and radio advertising, referrals, and publicity. Respondents receive small financial incentives for survey participation. Quotas were applied to achieve a sample with even gender and state (WA, Victoria) split; 80%/20% metropolitan/regional; 40%/40%/20% low/mid/high socioeconomic status (SES) residential area; and 35%/65% aged 18 to 25/26 to 64 years. Due to difficulties recruiting younger adults and those residing in low SES areas, these groups were under-represented in the sample. Sample composition did not differ significantly by condition (see Table 1).
Sample Characteristics by Advertisement Condition.
Notes. Percentages are rounded so may not sum to 100.
Missing values (n = 14) treated listwise. b Significant difference between conditions.
Measures
Participant Characteristics
Participants reported their gender, age, residential postcode, height, weight and highest level of education. Postcode determined metropolitan versus regional residential location and area-based SES (Australian Bureau of Statistics, 2018a). Height and weight were used to calculate body mass index (BMI; kg/m2) and weight category (underweight/healthy weight [BMI ≤ 24.9] vs. overweight/obese [BMI ≥ 25.0]). Prior ad exposure was measured by asking respondents, “Have you seen this ad before today?” (response options: yes/no).
Cognitive Responses
Cognitive responses to the ads were assessed by asking respondents to rate the extent to which they agreed with a series of statements about the ad they just watched (as per Brennan et al., 2014; Dixon et al., 2015; Noar et al., 2018; Wakefield et al., 2017), see Table 2). Perceived argument strength was assessed by averaging participants’ scores in response to five statements (Cronbach’s α = 0.90). Perceived personal effectiveness was assessed by averaging participants’ scores in response to 12 statements (Cronbach’s α = 0.95). Ease of understanding and psychological reactance were assessed using individual item scores. All ratings were made on a scale of 1 (strongly disagree) to 5 (strongly agree).
Items Assessing Cognitive Reactions to the Ads.
Note. Response scale for all items ranged from 1 “strongly disagree” to 5 “strongly agree”.
Emotional Responses
Participants indicated from 1 (strongly disagree) to 5 (strongly agree) the extent to which they experienced 11 positive and 19 negative emotions after viewing their ad. Emotions were presented in random order. Exploratory factor analysis (EFA) using maximum likelihood extraction with Promax rotation to allow for correlated variables was conducted to inform the development of emotion categories. Three categories were identified and computed: “anxious/distressed” (comprising the emotions of afraid, alarmed, anxious, ashamed, disgusted, distressed, fearful, guilty, jittery, nervous, sad, scared, uncomfortable, and upset; Cronbach’s α = 0.96), “responsive/engaged” (comprising alert, amused, attentive, determined, enthusiastic, excited, hopeful, inspired, interested, proud, and strong; Cronbach’s α = 0.91), and “oppositional/resistant” (comprising angry, bored, confused, hostile, and irritated; Cronbach’s α = 0.78).
Behavioral Intentions
Participants indicated their likelihood of engaging in 14 behaviors over the next week from 1 (very unlikely) to 5 (very likely), with a don’t know option. EFA with Promax rotation identified two categories of behavior: adaptive behaviors (e.g., eat more fruit and vegetables) and maladaptive behaviors (e.g., go without food for a day or more in order to lose weight), with Cronbach’s α = 0.84 and 0.85, respectively. Participants with 100% don’t know responses (<1% of cases for each computed category) were treated as missing for analysis purposes.
Psychological Measures
Internalized weight bias was assessed using the 11-item Modified Weight Bias Internalization Scale (Pearl & Puhl, 2014). Statements such as “I am less attractive than most other people because of my weight” were rated from 1 (strongly disagree) to 7 (strongly agree). Items were reverse scored where necessary and total scores were averaged to create an overall measure (Cronbach’s α = 0.93).
Antifat attitudes were assessed using the 13-item Anti-fat Attitudes Questionnaire (Crandall, 1994). Statements such as “Fat people make me somewhat uncomfortable” were rated from 0 (very strongly disagree) to 9 (very strongly agree), with mean scores computed (Cronbach’s α = 0.87).
Self-esteem was assessed using the 10-item Rosenberg Self-esteem Scale (Rosenberg, 1965). Statements such as “On the whole, I am satisfied with myself” were rated from 1 (strongly agree) to 4 (strongly disagree). Items were reverse scored where necessary and total scores were averaged (Cronbach’s α = 0.90).
Body dissatisfaction was assessed via the Eating Disorder Inventory—Body Dissatisfaction Scale (Garner, 1991), which presents 10 body parts (e.g., thighs, face, shoulders) that participants rate individually on a scale of 1 (very dissatisfied) to 5 (very satisfied). Scores on this scale were averaged (Cronbach’s α = 0.90).
Procedure
The study was approved by Curtin University’s Human Research Ethics Committee (HRE2018-0597). Respondents provided informed consent prior to participation. After answering questions assessing their personal characteristics, participants were required to view their randomly assigned ad at least twice before they could proceed with the survey. Participants indicated whether they had seen the ad before, then cognitive and emotional reactions were assessed, followed by behavioral intentions. Finally, internalized weight bias, antifat attitudes, self-esteem, and body dissatisfaction were assessed in random order to minimize order effects.
Statistical Analysis
Data were analyzed using Stata version 16.1 (Statacorp, 2020). Descriptive statistics were used to summarize participants’ characteristics and responses by condition. Although demographic characteristics did not differ by condition, prior exposure to the ads did (see Table 1). Accordingly, linear regression analyses tested differences by condition on the response measures, with prior ad exposure included as a covariate (RQ1–RQ3). For outcomes where a significant (p < .05) omnibus test for condition was observed, pairwise differences were assessed applying the Holm–Bonferroni correction for multiple comparisons. Each linear regression model was then rerun to examine whether there was an interaction between condition and participants’ weight category (underweight/healthy weight vs. overweight/obese; RQ4).
Results
Cognitions, Emotions, and Behavioral Intentions (RQ1)
As per Table 3, participants tended to agree the HWLC ads were understandable (means ≥ 4.36), made a strong argument (means ≥ 3.93) and were somewhat personally effective (means ≥ 3.60), with each HWLC ad scoring significantly higher than the control ad on these measures. Participants tended to disagree the HWLC ads were manipulative or exaggerated (means ≤ 2.57). All HWLC ads were rated as significantly less exaggerated compared with the control ad. Both Swap It ads were rated less manipulative than the control ad; there was no difference for ratings of “manipulative” for LiveLighter ads compared with the control ad.
Cognitive and Emotional Responses to the Advertisements by Advertisement Condition.
Notes. Linear regression analyses included prior ad exposure as a covariate. For each outcome where the omnibus test for condition was significant (p < .05), pairwise differences were assessed with a Holm–Bonferroni correction applied.
Significant difference compared with Control condition at p < .05. bSignificant difference compared with How to Swap It condition at p < .05. cSignificant difference compared with Become a Swapper condition at p < .05. dSignificant difference compared with Sugary Drinks condition at p < .05.
Overall, none of the ads prompted particularly strong emotional responses (mean scores ranged from 1.71 to 3.54 on 5-point scales, see Table 3). Compared with the control ad: All the HWLC ads except “How to Swap It” prompted higher anxiety/distress emotions; all the HWLC ads prompted higher responsive/engaged emotions; both Swap It ads prompted lower oppositional/resistant responses.
Mean scores for intention to engage in adaptive behaviors were above the midpoint of the scale, suggesting participants were likely to engage in these in the immediate future. Participants exposed to HWLC ads reported significantly greater intention to engage in adaptive behaviors compared with those exposed to the control ad. Intention to engage in maladaptive behaviors was low and did not differ between the control condition and the HWLC ad conditions (Table 4).
Scores on Behavioral Intentions and Psychological Measures by Advertisement Condition.
Notes. Linear regression analyses included prior ad exposure as a covariate. For each outcome where the omnibus test for condition was significant (p < .05), pairwise differences were assessed with a Holm–Bonferroni correction applied.
Adaptive behaviors comprised: Cut down the amount of high fat and/or sweet foods; drink fewer soft drinks, cordials or sports drinks; eat smaller portion sizes; eat more fruit and vegetables; exercise more often/be more active; use active transport (e.g., walk instead of drive); sit less (e.g., sit less while watching TV). bSignificant difference compared with Control condition at p < .05. cMaladaptive behaviors comprised: Go without food for a day or more in order to lose weight; Drastically cut down the amount of food you would normally eat, for a short period of time to lose weight; Make yourself vomit to lose/maintain weight; Use laxatives to lose/maintain weight; Smoke cigarettes to lose/maintain weight; Take diet pills to lose/maintain weight; Skip meals to lose/maintain weight.
Internalized Weight Bias, Antifat Attitudes, Self-Esteem, and Body Dissatisfaction (RQ2)
Mean scores across all conditions were at or below the midpoint of the scale for internalized weight bias, antifat attitudes, and body dissatisfaction following ad exposure (see Table 4). No significant differences between HWLC ad conditions and the control condition were observed for internalized weight bias, self-esteem or body dissatisfaction. Participants exposed to “Sugary Drinks” scored higher on antifat attitudes compared with participants who saw the control ad, although their scores were toward the lower end of this 9-point scale (M = 3.89) indicating a tendency to disagree with the antifat attitudinal statements.
Reactions to the Respective HWLC Ads (RQ3)
“How to Swap It” was rated easier to understand than “Toxic Fat.” ‘‘Become a Swapper’’ was rated significantly lower on argument strength than the other three HWLC ads. All HWLC ads were rated comparably on personal effectiveness. “How to Swap It” was rated less exaggerated than “Toxic Fat” and “Become a Swapper”. Both Swap It ads were rated less manipulative than the respective LiveLighter ads.
Both Swap It ads prompted less anxiety/distress, higher responsive/engaged, and lower oppositional/resistant responses than the two LiveLighter ads. “Toxic Fat” prompted the highest level of anxious/distressed response, although scores were around the midpoint, indicating participants tended to neither agree nor disagree this ad made them feel anxious/distressed. “How to Swap it” received the lowest scores for anxious/distressed and oppositional/resistant emotions and the highest score for responsive/engaged emotions.
No significant differences between HWLC ad conditions were observed for intentions to engage in adaptive behaviors or maladaptive behaviors, internalized weight bias, self-esteem, body dissatisfaction or antifat attitudes.
Responses to HWLC Ads by Weight Status (RQ4)
There were no significant interactions between ad condition and respondent’s weight category for the cognitive measures, behavioral intentions, nor the psychological measures.
For the emotional response measures, significant condition-by-weight category interactions were found for anxious/distressed (F(4, 2,197) = 4.17, p = .002) and oppositional/resistant emotions (F(4, 2,197) = 2.46, p = .044). Post hoc analyses indicated that participants classified as overweight/obese showed higher anxious/distressed responses to “Sugary Drinks” (M = 3.03 cf. M = 2.85, p = .037) and “How to Swap It” (M = 1.84 cf. M = 1.60, p = .001) than participants classified as underweight/healthy weight. However, post hoc analyses found no significant differences by weight category on oppositional/resistant responses for any of the HWLC ads.
Discussion
To address concerns that exposure to HWLCs may have negative psychological and behavioral consequences (Couch et al., 2018; Lupton, 2014; Pausé, 2017), this study systematically tested public reactions to ads from two HWLCs from Australia. Results indicate LiveLighter and Swap It HWLC ads prompted favorable responses in terms of intended outcomes, with scant evidence of adverse impact.
This study examined whether these HWLC ads promoted cognitions, emotions and intentions supportive of healthy lifestyle behavior or maladaptive behavior. Cognitive responses to all four HWLC ads were favorable and typically differed to those for the control ad. The HWLC ads were considered easy to understand and were believed to make a strong argument for being a healthy weight, eating a healthy diet, and exercising regularly. They were also perceived as somewhat personally effective, but not exaggerated or manipulative. Such measures of perceived message effectiveness are predictive of health-related intention and behavior change (Brennan et al., 2014; Davis et al., 2013), suggesting these HWLC ads have the potential to improve lifestyle behaviors. The HWLC ads prompted greater intention to engage in adaptive lifestyle behaviors than the control ad, attesting their potential to promote positive behavior change. Intention to engage in maladaptive behaviors was low overall and did not vary by condition, indicating these HWLC ads did not prompt intentions to engage in behaviors such as skipping meals, deliberately vomiting and/or taking diet pills.
Contrary to concerns that public health campaigns focusing on obesity may increase internalized weight bias (Durso & Latner, 2008), the present results are encouraging overall as they indicate that the HWLC ads tested did not increase internalized weight bias nor promote low self-esteem or body dissatisfaction. However, one difference indicated those exposed to “Sugary Drinks” reported weaker disagreement with antifat attitudes compared with the control group. Qualitative research could yield insight into which attributes of the “Sugary Drinks” ad might have prompted this particular unintended response. Population-level surveys have found no evidence of increased endorsement of stereotypes of overweight individuals in response to the LiveLighter campaign. We strongly recommend campaign planners carefully pretest and evaluate HWLCs to monitor potential intended and unintended outcomes.
Both LiveLighter ads were rated higher on argument strength compared with the “Become a Swapper” ad but did not differ to “How to Swap It”. Previous research with U.S. parents found perceived argument strength was predictive of intentions to reduce their own and their children’s sugary drink intake (Vaala et al., 2016). These findings point to the importance of including strong arguments supporting recommended behaviors in HWLC ads. The LiveLighter ads tended to be rated as more manipulative than the Swap It ads, but ratings were low overall.
The extent to which the four HWLC ads elicited anxious/distressed or oppositional/resistant emotional responses was low overall, suggesting they did not cause undue distress. However, LiveLighter ads elicited greater anxious/distressed and oppositional/resistant emotions compared with both Swap It ads, while Swap It ads elicited greater responsive/engaged emotions, likely reflecting the different executional styles of the respective campaigns. Research testing public health advertisements on quitting smoking indicates that ads evoking intense emotional responses are more effective at prompting positive health behavior change (Durkin et al., 2018; Farrelly et al., 2012). Previous experimental research comparing responses to a larger suite of HWLC ads suggests that “graphic, hard-hitting” ads, such as LiveLighter “Toxic Fat” may be more impactful than “softer, upbeat” ads (Dixon et al., 2015). Furthermore, evaluation surveys have found evidence of behavioral change in response to the LiveLighter HWLC (Morley et al., 2016, 2018, 2019) but not the more light-hearted Swap It (O’Hara et al., 2016) or Make Healthy Normal HWLCs (Kite, Gale, et al., 2018). By contrast, experimental research testing U.S. parents’ reactions to public service announcements encouraging reduced sugar-sweetened beverage consumption concluded that campaigns should pursue empowering messages and avoid evoking negative emotions (Vaala et al., 2016). Further research could elucidate how different emotional pathways contribute to behavioral responses to HWLCs employing different messages and executional characteristics.
One difference between participants of different weight status occurred in emotional responses to two of the HWLC ads: participants who were overweight/obese had higher anxious/distressed scores for “Sugary Drinks” and “How to Swap It” compared with participants classified as underweight/healthy weight. These findings are similar to previous research, which found HWLC ads emphasizing negative health consequences of excess weight elicit stronger emotional responses from adults with overweight or obesity (Dixon et al., 2015). Despite this difference in emotional response, there was no evidence the HWLC ads tested differentially affected those with overweight or obesity on any other outcomes. All participants responded similarly to the HWLC ads on cognitive responses, other emotional responses (responsive/engaged; oppositional/resistant), behavioral intentions, and psychological responses, irrespective of weight category. The latter findings contrast with related research, which found that exposure to still images from obesity prevention campaigns (Puhl, Luedicke, & Peterson, 2013) or simulated news articles (Major et al., 2014) classified as potentially stigmatizing inadvertently increased the risk of obesity-reinforcing behaviors among those who were overweight. Our findings, testing responses to whole HWLC ads rather than message excerpts, are reassuring as they suggest these ads were generally equally effective with participants of varying weight status.
Limitations and Strengths
As with previous research (e.g., Puhl, Peterson, & Luedicke, 2013; Puhl, Luedicke, & Peterson, 2013), the self-report measures used could have been prone to social desirability bias. Although the use of an anonymous online survey may have mitigated such bias, future research could assess physiological (e.g., skin conduction, heart rate) and neurocognitive responses that are not subject to such bias. Future studies could also complement quantitative assessment with qualitative methods, since previous research testing reactions to HWLCs have found qualitative research can provide additional insights to those yielded through quantitative methods (Dixon et al., 2016; Kite et al., 2020).
Experimentally testing participants’ immediate reactions following brief exposure to HWLC ads meant it was only feasible to assess behavioral intentions rather than actual behavior change. Although intentions are theorized to precede behavior change (Ajzen, 1991; Ajzen & Fishbein, 1980), experiments assessing short-term responses to HWLCs (e.g., Puhl, Luedicke, & Peterson, 2013; Puhl, Peterson, & Luedicke, 2013; Vaala et al., 2016; Young et al., 2016) are complemented by research monitoring longer-term behavioral impacts of such campaigns; for example, Morley et al., 2014, 2016, 2018, 2019; O’Hara et al., 2016). Dedicated research following participants over the long term is needed to firmly establish the relationship between exposure to HWLCs and subsequent engagement in adaptive and maladaptive behaviors.
The sample was recruited from an online panel, so may not have been representative of the general population. However, this is less important for experiments examining differences between groups with comparable baseline characteristics. Applying quotas helped ensure the sample was broadly representative. More than half of the participants were overweight/obese (reflective of the Australian population) enabling capture of responses among those potentially more vulnerable to weight-based stigmatization. This study tested reactions to HWLC ads among a broad sample of Australian adults. However, young people, specific cultural subgroups, or adults in other nations could respond differently, warranting future research with different populations. Recently, the methods of the present study were extended to testing reactions to HWLC ads in a sample of U.K. adults, and similar results were found (Horn & Jongenelis, 2022). In the case of Australian adults, rates of obesity are slightly higher among Aboriginal and Torres Strait Islander Australians (Australian Bureau of Statistics, 2020). A limitation of the present study was that there were insufficient numbers of Aboriginal and Torres Strait Islander participants to enable examination of responses to the HWLC ads among this important subgroup. Dedicated studies, such as that by Browne et al. (2019), are necessary to achieve this objective.
A strength of the present study over most previous experiments exploring unintended impacts of HWLCs is the testing of actual TV ads as stimuli. With the exception of Vaala et al. (2016), previous research (e.g., Puhl, Luedicke, & Peterson, 2013; Puhl, Peterson, & Luedicke, 2013) has tested reactions to campaign stills (e.g., isolated images, slogans). Testing reactions to whole ads provided a more naturalistic assessment of how audiences engage with and respond to HWLC ads. While viewing a single ad is not realistic in some instances (e.g., compared with TV, where multiple ads tend to be aired together) it is typical for online videos. A limitation to testing real ads was that we could not control all aspects of content. Although the differing messaging and execution style of ads from the different campaigns were salient, other subtle variations between the ads tested could have contributed to how participants reacted to them. For example, the ads varied in length (45s for Swap It ads vs. 30s for LiveLighter ads), which could have contributed to the Swap It ads generating higher responsive/engaged emotions than the LiveLighter ads. Also, the Swap It ads included messaging about exercise and diet, whereas the LiveLighter ads focused on diet. Nonetheless, ads from both HWLCs performed comparably in promoting adaptive behavioral intentions pertaining to diet and physical activity.
Conclusion
Overall, the four HWLC ads tested in the present study performed favorably compared with the control ad on the assessed cognitive, emotional, behavioral, and psychological outcomes and showed no clear evidence of adverse impacts. Findings suggest that campaigns like LiveLighter and Swap It can continue to be used safely in public health efforts addressing poor diet, inactivity, and obesity.
Footnotes
Acknowledgements
The authors gratefully acknowledge the work of PureProfile in undertaking the data collection for this study.
Author Contributions
M.J., H.D., and B.M. conceptualized the study. All authors interpreted the data analyses, co-authored the manuscript, and approved the contents of the submitted manuscript. M.J. managed the data collection, and M.J. and M.S. conducted the data analysis in consultation with H.D. and B.M.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: This research was conducted while M.J. was employed at Curtin University. Her position at Curtin University was funded by the WA Cancer Prevention Research Unit, which receives part funding from Cancer Council Western Australia. H.D., M.S., and B.M. are employed by a nonprofit organization that conducts public health interventions, research, and advocacy aimed at reducing the population’s cancer risk, including risks pertaining to diet and obesity. The funder (Cancer Council Western Australia) had no involvement in designing the study, collecting, analyzing, and interpreting the data, nor in writing the manuscript.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by Cancer Council Western Australia, which is contracted by the Department of Health Western Australia to conduct the LiveLighter® campaign.
Ethical Standards Disclosure
Curtin University Human Research Ethics Committee (HREC) approved all study procedures (HREC number HRE2018-0597) and all participants provided informed consent.
