‘It makes you not want to eat’: Perceptions of anti‐obesity public health campaigns in individuals diagnosed with an eating disorder

Abstract

Objective

Recent research has highlighted the potentially deleterious effects of anti‐obesity public health campaigns on vulnerable audiences such as those at risk of, or diagnosed with, eating disorders (ED). The qualitative experiences of campaigns in this population group have been little explored.

Methods

Twelve interviews were conducted with participants either currently or previously diagnosed with an ED using the Interpretative Phenomenological Analysis approach. Participants were shown a series of real‐world public health campaigns and, using a semi‐structured interview guide, asked about their responses to the images used.

Results

Several themes emerged from the data, including the effect of campaigns on participants, perceived effect on others, and general campaign attributes. It was evident that participants felt that anti‐obesity campaigns had the potential to confirm ED thoughts and behaviours, and could act as potential triggers. Participants also felt that they could impact vulnerable groups (e.g., young audiences) and stigmatise those living in larger bodies.

Conclusion

Campaigns addressing overweight and obesity should consider reframing these conditions and care must be taken to avoid causing potential harm to vulnerable audiences.

Highlights

• Individuals who have experienced eating disorders (EDs) felt that anti‐obesity public health campaigns could reconfirm thoughts and behaviours pertinent to the ED.

• Participants were particularly concerned over how campaigns may be received by vulnerable audiences such as young people, or those with an already‐heightened level of awareness and concern about their weight.

• Participants recognised the importance of addressing overweight and obesity as public health issues, however suggest future efforts be more sensitive to the potential for adverse effects.

Abbreviations

AN

anorexia nervosa

APA

american psychiatric association

BED

binge eating disorder

BN

bulimia nervosa

COVID‐19

coronavirus disease 2019

ED

eating disorder(s)

IPA

interpretative phenomenological analysis

NEDC

national eating disorders collaboration

OSFED

other specified feeding or eating disorders

1. INTRODUCTION

The public health response to the global prevalence of overweight and obesity has been vast. Advertising and media campaigns distributed on a whole‐of‐population level are one approach to curbing the growth of these conditions. There are many examples of this in the public health landscape (e.g., Croker et al., ²⁰¹²; Grunseit et al., ²⁰¹⁵; James et al., ²⁰²⁰; Morley et al., ²⁰¹⁶; O'Hara, Grunseit, et al., ²⁰¹⁶; Talbot & Branley‐Bell, ²⁰²¹), with campaigns often targeting weight loss through modification of diet and increased physical activity (Walls et al., 2011; Watson, ²⁰¹¹). This is despite the growing body of evidence indicating that weight‐loss efforts are futile in the long‐term, with weight regain probable (Blomain et al., 2013; MacLean et al., ²⁰¹⁵; Mann et al., ²⁰⁰⁷; Melby et al., ²⁰¹⁷; Tomiyama et al., ²⁰¹³; Tsai & Wadden, ²⁰⁰⁵).

Recent research has suggested that such campaigns distributed on a population‐based scale have the potential for negative effects in vulnerable people, such as those at risk of, or diagnosed with, an eating disorder (ED) (Bristow et al., 2020, 2021). These studies have shown that when exposed to anti‐obesity campaigns (those which promote the prevention of overweight and obesity), risk factors for EDs (e.g., body dissatisfaction, desire to engage in unhealthy behaviours for weight control) may be increased in some audiences. In particular, those demonstrating high ED symptomology appeared to be more affected (Bristow et al., 2021). EDs, including anorexia nervosa (AN), bulimia nervosa (BN), binge eating disorder (BED), and other specified feed or EDs (OSFED; American Psychiatric Association [APA], ²⁰¹³), are complex psychiatric conditions often characterised by restrictive and/or abnormal weight control behaviours (Peckmezian et al., 2017). Dieting and excessive exercise are also well‐established risk factors for the development of EDs (Peckmezian et al., 2017; Watson et al., ²⁰¹⁰). Therefore, it is not implausible that campaigns promoting these behaviours could have deleterious effects. Given the serious burden of these conditions on individuals (for example, the standardised mortality rate from suicide due to AN is the highest of any psychiatric condition; Watson et al., ²⁰¹⁰), the potential for harm in this population group must be ascertained.

The ethical considerations of large‐scale health communications are not new and have been addressed previously (Bayer & Fairchild, 2016; Cho & Salmon, ²⁰⁰⁶; Guttman & Salmon, ²⁰⁰⁴). Bayer and Fairchild (2016) argue that public health campaigns have the potential to shame and stigmatise certain groups, citing the HIV/AIDS epidemic as an example of the harm that can ensue when campaigns are not developed with caution. Guttman and Salmon (2004) have also drawn attention to the potential for campaigns to reach audiences for which they are not intended. One of the guiding principles of health promotion and public health practice is to do no harm (Guttman, 2017), which must be considered in the context of anti‐obesity campaigns whose impact on EDs had not been explored until recently.

The experiential perspectives of this population group at high risk of harm have not thoroughly been explored. It is important to understand how this group may be affected by anti‐obesity public health campaigns, to enable more effective, safer messaging. As such, this study aims to explore how individuals with experience of an ED perceive real‐world anti‐obesity campaigns using a qualitative approach.

2. METHODS

Interpretive Phenomenological Analysis (IPA) was the primary qualitative methodology used in this study. This approach seeks to understand the ways in which individuals make meaning of and experience phenomena and has been frequently used in the fields of health and psychology (Brocki & Wearden, 2006; Spiers & Smith, ²⁰¹⁹). Therefore, this methodology was deemed the most appropriate for exploring how individuals diagnosed with EDs experience public health anti‐obesity campaigns. This study was approved by the Monash University Human Research Ethics Committee (#24029). All participants provided written informed consent.

2.1. Participants

2.1.1. Inclusion criteria

One of the core components of IPA methodology is the implementation of purposive sampling (Spiers & Smith, 2019). Due to the intricate nature of the IPA approach, small sample sizes are recommended (5–16 participants; Smith, ²⁰¹¹; Spiers & Smith, ²⁰¹⁹). This study purposively recruited Australians 15 years and over with a current or previous diagnosis of an ED. Current or previous diagnosis was chosen for inclusion because recovery from EDs is non‐linear with long‐lasting effects (Bowlby et al., 2015; Steinhausen, ²⁰⁰⁹; Wetzler et al., ²⁰²⁰). Participants currently being treated as in‐patients were deemed high risk and excluded due to the sensitive nature of the interview content. Additionally, 15 years was selected as the minimum age due to concerns that younger audiences might be too impressionable regarding the image material. Interviews were limited to Australia, as originally it was intended that these be conducted face‐to‐face. However, due to the global COVID‐19 pandemic and physical distancing restrictions, all interviews were conducted using the teleconference technology platform Zoom©.

2.1.2. Recruitment

Recruitment was through advertisements on the Internet and social media, and by word of mouth through author networks. The study was advertised as comprising interviews exploring how public health anti‐obesity campaigns affect persons with an ED. Participants received a gift voucher as a token of appreciation for their time.

2.1.3. Participant characteristics

Twelve (n = 12) participants completed interviews between May and August 2020. All were female, aged between 19 and 42 (27.3 ± 7.0). Diagnosis varied between AN (n = 7), BN (n = 3), OSFED (n = 1), and both AN and BN (n = 1). Participants were a mix of previously diagnosed and recovered (n = 3), currently diagnosed (n = 6) or in between (n = 3).

The all‐female study sample was not intentional. Rather, it reflected the population group that responded to the study advertisements. This is possibly due to the higher proportion of females who experience EDs compared to males, despite prevalence increasing in males in recent years (Peckmezian et al., 2017). Due to perceived stigma, males are also less likely to seek help for an ED and often remain undiagnosed (Griffiths et al., 2015). It is this perceived stigma, as well as lower rates of diagnosis that may result in greater hesitancy in sharing their experiences.

Additionally, in relation to participant stages of diagnosis, some participants struggled to identify with an endpoint and felt their experience to be fluid: for these participants, the ED and its related thoughts and behaviours come and go. This is consistent with recent research in which participants described recovery as non‐linear and ‘fuzzy’, with no clear start or end points (LaMarre & Rice, 2021). Hence, it was difficult to provide an analysis that distinguished clearly between these categories. Instead, the analysis aimed to highlight participant use of tense to indicate the stage to which they felt they belonged (i.e. past and present self).

2.2. Materials

Participants were exposed to four different public health campaign images (one at a time) during each interview. These campaigns were selected based on their use in a previous related study. An overview of the campaigns selected is presented in Table 1. Campaign images are provided in Supplementary Materials.

TABLE 1.

Description of campaign images

Image	Country	Visual description	Text
1	UK	2 packets of cigarettes, one open and one closed—open packet is filled with French fries	“Guess what is the biggest preventable cause of cancer after smoking”
2	US	A female drinking from a bottle of soft drink with liquid fat pouring out of the bottle onto her mouth and body	“Don't drink yourself fat choose water instead”
			“Rethink your drink”
3	Australia	An overweight male balloon character	“Swap it don't stop it”
		Measuring tape used as an exclamation mark
4	Australia	A male clasping his stomach in his hands positioned above an image of visceral fat surrounding an internal organ	“Grabbable gut outside means toxic fat inside” “Reduce your risk of heart disease, type 2 diabetes and cancer by eating less and moving more every day”
		Heart Foundation Logo
		Cancer Council Logo

2.2.1. Interview development

The process of developing the interview structure was guided by the hermeneutics of IPA with its endeavour to make sense of other's sense making (Smith, 2017; Smith & Osborne, ²⁰¹⁵). A semi‐structured interview schedule enabled participants to explore and elaborate on concepts in their own way (Spiers & Smith, 2019). Interview questions used to guide the interviews were drafted and refined by authors. Follow up prompts were included to aid further discussion if needed. The interview was separated into two parts. In Part A, questions were asked in relation to each campaign image, in Part B, questions related to the campaigns overall. The full interview schedule is presented in Table 2.

TABLE 2.

Semi‐structured interview schedule

Schedule part A

Potential prompts

Have you seen this campaign/similar campaigns? Do you notice these in your day‐to‐day life? What is your initial response to this campaign in particular?1 What do you like/dislike about this campaign?2 What is your understanding of this campaign and what it aims to achieve?3 In relation to your health condition, how do you think it affects your understanding and/or interpretation? How do you think others would respond to these campaigns? How do you think your understandings and/or perceptions of the campaign may differ from others? Do you have any suggestions of how this campaign may be improved?4

1For example, what type of feelings, emotions, and thoughts does this bring up for you? 2For example, you may choose to explain how it makes you feel, or how you think it may make others feel 3For example, how do you receive, understand and interpret the messages conveyed? Adjectives may be helpful here 4According to how you have previously described its impact?

Schedule part B	Potential prompts
What are your overall thoughts and feelings about these images and campaigns as a whole? How do you perceive these, in relation to your condition?5 Can you explain whether or not you believe there are differences between these, or they are mostly the same? Do you have any other thoughts, insights or comments you would like to add?	5For example, how do they make you feel, from the perspective of your diagnosis with an eating disorder?

Note: ^1,2,3,4,5Superscript number corresponds to potential prompt.

In line with IPA recommendations, initial questions were general in nature to ease participants into the conversation, with more sensitive prompts specific to their condition arising later in the interview once rapport had been established (Spiers & Smith, 2019).

2.2.2. Analysis

Interview recordings were transcribed verbatim. IPA stipulates that all transcriptions be completed before the analysis process can begin (Smith, 2011). After transcription, interviews were read multiple times to ensure data immersion. Interviews were analysed individually at first, followed by overall thematic analysis to identify recurring themes, convergences, and divergences (Smith, 2017; Spiers & Smith, ²⁰¹⁹). More specifically, each interview was analysed by arranging the transcriptions into tables, with a column for coding themes on the right of the text, and a column for exploratory commentary on the left (IPA methodology; Spiers & Smith, ²⁰¹⁹). Transcripts were then exported into the qualitative software NVivo®, and the process was repeated to allow any further concepts to arise and to modify previous theme classifications. Sample transcripts and analyses were provided to a second researcher for checking and additional input, and to ensure researcher congruence. To ensure validity, participants were provided with a copy of their transcripts and offered the opportunity to further clarify or amend as they felt necessary. Results of this analysis are presented below thematically, with the inclusion of quotes to illustrate and support findings.

2.2.3. Positionality statement

Intrinsic to IPA is the process of interpretation, meaning the analysis of the data provided is subjective and may differ in presentation between researcher. The lead author conceptualised the study, designed the interview schedule, conducted interviews and performed data analysis and the final write up. This study was part of a larger research project undertaken for their Ph.D.’s interest in the field stems from personal lived experience. This enabled them a certain level of sensitivity in approaching the subject and a shared understanding of the participants' experience. At the same time, it is important to recognise their past experience likely influenced their interpretation of the data. To address this, researcher triangulation was performed in collaboration with throughout to ensure questions were not leading or presumptuous and allowed participants to share their own perspectives freely and without boundaries. During interviews, was constantly mindful to not project their own experiences, but simply listen to participants and prompt where necessary based off their interpretations only. During analysis, read and re‐read transcripts multiple times in an attempt to gain multiple perspectives and remove their own biases as much as possible.

3. RESULTS

Analysis of the interviews established three main themes, within which further subordinate themes are presented. Participants are referred to using a pseudonym to protect confidentiality. Participant diagnosis is presented in brackets. An overview of themes and subordinate themes is presented in Table 3. Standard ellipses (…) indicates omitted text (text not relevant); [sic] indicates quote transcribed as is despite grammatical errors, and en dash (−) is used to denote a short pause. Quotes may be in response to any of the four campaign images, or the campaigns overall. Where a specific image being referenced is important for context, it is denoted in brackets.

TABLE 3.

Overview of major domains and themes

Domain	Themes
Effect on present or past self	Confirmation and justification of the eating disorder
	Campaigns as potential ‘triggers’
	Reinforcing fears of food and weight gain
Effect on others	Potential to impact vulnerable populations
	Stigmatisation of persons in larger bodies
Campaign attributes	Recognition of necessity
	Negative associations
	Recommendations for improvement

3.1. Effect on present or past self

One of the most prominent themes that emerged throughout participant interviews was the effect of campaign exposure on the self. The self emerged as the present self; how this affected participants currently, or the past self; how this would have affected them when affected by their ED. This distinction between present and past self was dependent on where the participant was in their ED journey and is evident in the use of tense throughout interviews.

3.1.1. Confirmation and justification of the eating disorder

That campaigns and their underlying messages confirmed and justified some of the thoughts pertinent to the participants' ED was consistently expressed. That is, participants felt that campaigns often fed into the beliefs they already held about food and weight. Sarah (OSFED) describes this as a reiteration, referring to her past self.

From an eating disorder perspective, I would say that it just reiterates—it just kind of reconfirms what I was thinking at the time that obesity is unhealthy and it needs to be avoided and for me that was—like it needs to be avoided at all costs.

(Sarah, in response to Image 1)

Another participant, May (AN), when reflecting on how campaigns would have made her feel before she embarked on her recovery journey, took on the persona of her ED, explaining how campaigns would have forced her to ‘stay’. This highlights the difficulty participants may face during their recovery attempts when presented with messaging that reaffirms beliefs associated with the ED.

I think, maybe sort of just being like yeah see? Like you don’t want to look like that so stay with me and do what I’m telling you because you know that’s, that’s what you’re going to end up like if you stop following my rules and you stop restricting you’ll just blow up like a balloon and you’ll have toxic fat, literally that’s like, you know, complete confirmation that I should stay in my miserable existence with my eating disorder.

(May, in response to Image 4)

3.1.2. Campaigns as potential triggers

Not only was it evident that campaigns could reaffirm ED thoughts, but also that they could also act as potential triggers (set off an ED thought or behaviour). This is evident in the following quote from Kate (AN, BN), who describes the triggering of an ED behaviour (purging) in response to Image 4.

It makes you not want to eat, do you know what I mean, it would make you maybe want to throw it up. Maybe you just saw this having a meal and you’re just like, oh, no, like I can’t even think of all that fat inside of me, I need to get it out…

3.1.3. Reinforcing fears of food and weight gain

The word fear appeared frequently in the interviews, particularly in relation to fear of food and fear of gaining weight. Consider the following quote from Kate, who describes Image 1 as positioning food as an enemy. Of interest is the use of we and people like me, indicating a belief that fearful feelings may be universal to those with EDs.

It’s uncomfortable. It just pushes the idea that a lot of people like me feel that food is the enemy, like it’s this dangerous thing that can hurt you, and obviously, that’s not true, but that’s how we feel and this kind of provokes that unhealthy idea that we deal with.

(Kate, in response to Image 1)

Also responding to Image 1, another participant (Ella, AN) reflected on how the campaign would have induced fear when she was in the midst of her ED, and acknowledged this fear still lingers for her at present. Once again, the use of tense acknowledges the varying levels at which participants' felt they may be affected, depending on their stage of recovery.

I think perhaps a few years ago it would have been fear inducing for me, now I think I'm able to like look at it and kind of separate myself I guess from the word obese, but I think it still plays into, like at the moment it still plays into the fear of putting on weight and the fear of losing control and becoming obese or overweight.

(Ella)

3.2. Effect on others

As well as discussing the effect that the campaigns had on themselves, participants often addressed how they thought the campaigns could affect others, in particular vulnerable populations and persons living in larger bodies. Participants were also concerned about how campaigns would be received by young people, who may be particularly susceptible due to being more impressionable than older age groups.

3.2.1. Potential to impact vulnerable populations

The concern for the effects of campaign exposure on others appeared, at the least, equal to the concern participants held for themselves. Consider Sarah, who was consistent in her belief that campaigns distributed on a population‐based scale have the potential to negatively impact people who may be vulnerable to that message. While Sarah's response was during exposure to Image two in particular, she generalised this to any messaging that suggested restriction in some form.

…I think that any advertising—anything that says you can’t have a certain food or drink or whatever is detrimental to the population who are trying to restrict that thing.

(Sarah)

That campaigns could potentially impact vulnerable populations was reinforced by Olivia (AN), while acknowledging that there are people who would likely not be affected at all. Unlike Sarah, Olivia's response was specific to Image 1.

Yeah, I guess the average person probably wouldn’t be too affected because they’ll, they’d see the word obesity and go oh well that’s not necessarily me. If they’re not obese or something like that or if they’re normal weight they’d go oh yeah, that campaign doesn’t apply to me and kind of ignore it and continue with the day. But I guess the concern is that people who are vulnerable to eating disorders or in their early stages can use it as fuel I guess to develop that even further…

(Olivia)

Also referring to Image 1, Kate described a cumulative effect, where ongoing exposure to messaging could eventuate in an ED, particularly in younger audiences.

It’s like a slippery slope. It starts promoting that idea in their heads, especially for younger people that are more susceptible, it kind of puts that idea in their minds that like, hi, like you should associate this as being bad, and that just can kind of like over time fester and then become an eating disorder.

(Kate)

3.2.2. Stigmatisation of persons in larger bodies

In considering the ways that campaigns could affect others, participants spoke of how they felt the campaigns were inherently stigmatising towards persons in larger bodies. This often involved feelings of shame and guilt, as well as blame directed towards the individual for their size and weight. One such example is the following excerpt from Olivia, who took issue with the use of a slim model and the paradox that this represented. The irony of using a slim model to imply weight gain was also commented on by numerous participants.

I think the biggest issue for me with this particular image is the fact that the person in the picture is already quite thin and seems quite healthy—if you’re struggling with body image and might be even a little bit larger than this person, but not necessarily overweight or at risk of anything yeah, I think it would be quite difficult to see—using the model that they did means that a large portion of the population are I guess bigger than that person. And the fact that they’re saying that her drinking the soft drink is bad makes that even more monumental for someone who is larger than her. The fact that she’s drinking herself fat by drinking the soda, like what would that look like to someone who is of a larger weight but not necessarily again at risk of the things that they’re trying to prevent.

(Olivia, in response to Image 2)

Similar sentiments were expressed by Hannah (AN), who described Image 2 as demonising. This further contributed to the general consensus that on the whole, campaigns shamed and blamed those in larger bodies for their size.

… I mean it’s very—it demonises those people doesn’t it? It really says, says to them, you’re not as valued because of your weight. You’re fat, you probably, it kind of generalises that people have unhealthy habits, yeah I think that’s probably—it certainly would be quite offensive I could imagine to those people.

(Hannah)

Additionally, in considering blame, Ella expressed concern that campaigns failed to recognise the external factors contributing to an individual's weight status.

I think, I don’t really like health campaigns particularly like this because as I said, like a few years ago this would have been something that like I would have seen the word obesity and kind of been very scared and associated it with myself, and I just don’t think that this kind of, these kind of campaigns are quite blamatory and they don’t, like they blame the person quite a lot and don’t really focus on the socio‐economic factors of obesity, they just kind of focus on obese people eat chips, obese people eat a lot.

(Ella, in response to Image 1)

3.3. Campaign attributes

The previous themes demonstrated that participants spoke candidly about the ways that they experience campaigns, and perceive others to. However, the way they referred directly to specific aspects of the campaigns also stood out. These descriptions were less related to experience, but provided important insight into interpretation and the factors they deemed to be noteworthy.

3.3.1. Recognition of necessity

Though participants often criticised campaigns, they also recognised that overweight and obesity are highly prevalent conditions whose health and economic burdens need to be addressed. They also seemed to grapple with the concept that messages may have a positive overall effect on a large proportion of the population. Olivia expressed with underlying tones of empathy.

I think you can’t really blame the person who came up with the strategy or the people who promoted it and spent money on disseminating it. I think that generally the intention is good and I think the net effect in terms of you know, weight loss, creating awareness amongst people who might not be aware is good but I don’t think it takes into consideration the impact on the individuals who already consider these things to I guess a normal level.

(Olivia, in response to image 4)

The dual themes of overweight/obesity and ED prevention were also evident for May. This further demonstrates the contention that many participants held between understanding that weight status has the potential to influence health outcomes, but also that the messaging surrounding this can be harmful.

… I grapple with it all the time to be honest because obesity is a massive, massive issue—it is a massive drain on the public health system but then as are eating disorders—and they’re continuing to rise, umm and the age of onset is getting younger and you know there are all these frightening statistics around that as well so it’s like neither of them can be ignored—it’s that fine balance of yes we need anti‐obesity campaigning—absolutely we do, but how can we do it in a way that’s sensitive to those who are susceptible to eating disorders.

(May)

3.3.2. Negative associations

A persistent theme for participants was the way the campaigns often created negative and moral associations for both food and body size. In regards to Image 1, Olivia spoke of this in reference to likening food to cigarettes. This was mostly consistent across images, with the exception of Image 3 which compared to the others, was often thought to encourage a healthier message.

I guess the biggest issue for me is just the link of the particular food with obesity and thus, cancer and smoking and that kind of negative approach. Like as someone who’s trying to reintroduce these kinds of foods that are quite difficult for me.

(Olivia, in response to Image 1)

Another concern was the way that campaigns intrinsically labelled fatness as a bad and negative thing. This was particularly consistent for Image 2.

Probably the word fat, and the fact that they’re just targeting that word and placing so much emphasis on it and adding so much meaning to it—a lot of the reading I’ve been doing is about you know the power of the messages we’re given and that thin is good fat is bad, whereas every body is right in its own way and there’s perfectly healthy people in larger bodies, and this particular Image sort of demonises fat people as if it’s the worst thing that can happen to you and I think that’s fairly irresponsible to be honest.

(Hannah, in response to Image 2)

Anna (BN) also spoke of campaigns placing moral value on—in this case a beverage—and that she found this type of narrative to be unhelpful in the overall context of health and wellbeing.

… it’s weighing in on this, there’s a good type of food and there’s a bad type of food, its moralising the drink she’s got in her hands, and that kind of thinking is really harmful, particularly people with eating disorders, there is no such thing as a good food or a bad food… if you take away the moral value of food, because there shouldn’t be any, it would make conversations around health and wellbeing so much easier.

(Anna, in response to Image 2)

3.3.3. Recommendations for improvement

Finally, when participants were asked how they thought campaigns could be improved, there was consistent support for messages to encourage and highlight the benefits of healthy behaviours as opposed to shaming undesirable ones. Only some participants expressed desire for campaigns to be abolished completely, with many recognising the importance of acknowledging that higher weight status may influence other health outcomes. Participants also frequently commented on their preference for Image 3 because it was less stigmatising and did not promote complete restriction.

…I do like this idea because, you know, we should be encouraging people to make small changes in their life, not giant, crazy ones they can’t sustain… I think it’s a far more measured way to promote, you know, balance.

(Charlotte, BN, in response to Image 3)

4. DISCUSSION

This study sought to explore the experiences of participants currently or previously diagnosed with an ED when exposed to real–life public health anti‐obesity campaigns. This was in light of recent findings demonstrating a negative influence of campaigns on ED related outcomes (Bristow et al., 2020, 2021). In particular, the study enabled a thorough exploration of why and how campaigns may contribute to ED thoughts and behaviours in a high‐risk population.

The first study theme to be addressed is how participants experienced campaign exposure. Division of the self into two (me as my present self and me as my past self) reflected that participants were either in the throes of their ED or well into recovery. Whether reflecting on their present or past selves, it was clear that participants felt that the campaigns further justified thoughts and feelings associated with their ED. Described as either reiteration, confirmation, or justification, the message was consistent in the way participants described feeling as though the campaigns encouraged their disordered thoughts. Similarly, the participants described the campaigns' ability to trigger ED symptoms. Importantly, this concerned setting off actual behaviours as well as reinforcing thoughts. For example, the phrase ‘it would make you maybe want to throw it up’ demonstrates the impact that these types of campaign can have on both unhealthy thoughts and actual behaviours in extremely vulnerable persons. The final subordinate theme in the overarching effect on self theme was reinforcing fears of food and weight gain. This theme also highlights how campaigns can exacerbate beliefs typical of ED populations. Given that an intense fear of food and gaining weight is a key criterion for the diagnosis of EDs (APA, ²⁰¹³), statements such as ‘it really makes me scared of getting obese’ are cause for concern when considering exposure in this population group. Note also that 18% of respondents in a recent retrospective cohort study of 2901 patients being treated for EDs attributed the onset of their ED to public health messages (Mensinger et al., 2021). Together, these findings highlight the potential for unintended harm in this population group.

Further to the way that participants experienced the campaigns were their perceptions on how the campaigns could affect others. Most notable was the campaigns' potential influence on vulnerable populations, in particular younger audiences or persons susceptible to developing an ED. Numerous participants commented that campaign effects are cumulative. In other words, while brief exposure to one campaign may not cause significant damage, repeated exposure to dieting and weight loss messages could fuel the development of an ED. The idea of a dose‐response relationship between media messages on thinness and increased ED symptomology has been explored (Andersen & DiDomenico, 1992; Vaughan & Fouts, ²⁰⁰³). However, just as participants acknowledged vulnerable populations, they also recognised that for many people it may be ‘water off a duck's back’—[verbatim]. The ethical considerations of public health campaigns on a population‐based scale are still subject to debate (Guttman, 2017; Guttman & Salmon, ²⁰⁰⁴; Puhl & Heuer, ²⁰¹⁰), and the question of whether potential harm to vulnerable audiences is offset by potential benefits to the rest of the population remains complex (Bayer & Fairchild, 2016; Cho & Salmon, ²⁰⁰⁶). Although these findings may not answer the question entirely, they do provide insights from the perspective of those most at risk into the potential harms that can result from exposure in vulnerable groups.

The participants' belief that campaigns are stigmatising is consistent with findings from both our systematic literature review and other studies in the field highlighting the stigma and bias experienced by persons in larger bodies (Puhl, 2013; Puhl et al., ²⁰¹³). This is especially concerning due to the potential for internalisation of weight stigmatisation to further lead to engagement in disordered eating behaviours (Carels et al., 2010; Durso et al., ²⁰¹²; O’Hara, Tahboub‐Schulte & Thomas, ²⁰¹⁶). Most similar to the present study is that by Lewis et al. (2010) who reported participants feeling blamed and stigmatised by public health anti‐obesity messages in a qualitative study of adults classified as ‘obese’. In that study, participants said campaigns positioned them as ‘failures’ and lazy, and that being overweight was simply a choice (Lewis et al., 2010). These views were in line with those expressed in the present study, and further contribute to the ongoing concern of stigmatisation experienced by persons living in larger bodies. It also adds to the discussion of the use of shame and blame as tactics to force behavioural change, while neglecting the environmental, societal, and biological influences on a person's weight status (Guttman & Salmon, 2004; Puhl & Heuer, ^{2009, 2010}; Thille et al., ²⁰¹⁷; Vartanian & Smyth, ²⁰¹³).

The concluding theme presented participants' reflections on the campaigns themselves and provided perspective into the need for campaigns and recommendations for future improvement. Analysis of the results showed that participants saw both positives and negatives in the campaigns. They recognised overweight and obesity as a public health issue worth addressing but were critical of the enforcement of weight loss on the overall population. Perhaps the greatest concern they had with the campaigns was the overtly negative association of food and overweight/obesity. They objected most to the ongoing narrative of a moral dimension to the food/weight nexus. Hence, recommendations for improvement centred on promoting the positive benefits of desired health behaviours rather than shaming behaviours perceived as negative. These recommendations are consistent with those from the NEDC, which suggests that healthy messages should promote the adoption of healthy behaviours for overall physical health and emotional wellbeing rather than promoting the elimination of foods and making moral judgements on dietary choices (Watson, 2011).

By nature of the IPA approach and qualitative research in general, the findings of this study are not generalisable to the population. Though the in‐depth exploration of participant experience offers valuable insight into the issues, qualitative research cannot, and does not aim to, establish generalisable findings. Additionally, the authors acknowledge that detailed demographic information (e.g., sexual orientation, racial and ethnic data) was not collected and that this limits the understanding of these factors on individuals' experiences. Future research will include demographic/sociodemographic data collection to explore this further. Finally, due to the COVID‐19 pandemic, all interviews were held via videoconferencing technology. Despite efforts to make the participants feel as comfortable as possible, the absence of in‐person contact may have resulted in less rapport between the researchers and participants.

The findings of this study contribute to growing concern over anti‐obesity public health campaigns and their impact on at‐risk audiences. The detailed and thorough participant reflections show that these messages not only reinforce ED‐related thoughts and feelings but may also encourage unhealthy weight control behaviours. This exploration also enabled the identification of certain campaign characteristics that were contributing to the participants' uneasiness. These were primarily the stigmatising and blaming subtext of the campaigns, the contribution to unhelpful weight and health narratives, and the demonisation of food and weight. The benefit of this finding is that it points directly to actionable areas of improvement. It enables government and health authorities to reflect on their current practices and create sustainable change. As it is likely that campaigns addressing overweight and obesity will continue to be developed, the present study adds to the growing body of literature indicating that substantial change and care are needed to minimise the negative and potentially detrimental impacts on vulnerable audiences.

CONFLICT OF INTEREST

The authors report there are no competing interests to declare.

Supporting information

Supporting Information S1

Bristow, C. , Simmonds, J. , Allen, K.‐A. , & McLean, L. (2023). ‘It makes you not want to eat’: Perceptions of anti‐obesity public health campaigns in individuals diagnosed with an eating disorder. European Eating Disorders Review, 31(1), 166–177. 10.1002/erv.2950

DATA AVAILABILITY STATEMENT

Data from participant interviews is not publicly available to protect anonymity.