#DiabetesOnAPlate: the everyday deployment and contestation of diabetes stigma in an online setting

ABSTRACT

People living with Type 1 and Type 2 diabetes (T1DM &T2DM) report that media – including journalism, health promotion, and popular culture – are a primary source of stigmatic representations of diabetes, and that this compromises their physical and mental health. This view of diabetes as stigmatised is not shared by many health professionals and nor the wider public. We used an existing representation of diabetes on Instagram, #DiabetesOnAPlate, to examine how stigmatic representations of diabetes are (re)produced, discussed, and contested. Our analysis found that, consistent with the notion of diabetes as stigmatised, use of the hashtag #DiabetesOnAPlate on Instagram is associated with public health discourses of indulgent eating and with posts that signal recognition of this as a moral transgression. A subsequent on-line survey found that participants who did not have diabetes did not recognise a prototypical #DiabetesOnAPlate post as stigmatic. Whilst some with T1DM and T2DM did perceive and contest the stigma, T1DM participants did so for their group (but not for T2DM); and there was evidence that some T2DM participants internalised the stigma. These findings support concerns about the everyday, divisive nature of stigmatic representations of diabetes on social network sites, which on the one hand reflect public health discourse and on the other may compromise health policy objectives and exacerbate health inequalities.

KEYWORDS:

• Diabetes
• stigma
• public health policy
• health messaging
• social media
• Instagram
• thematic analysis

Introduction

People living with diabetes in the UK report feeling judged based on stereotypes associated with weight stigma, such as being ‘fat’, ‘lazy’ and ‘eat(ing) too much’ (Teixeira & Budd, 2010); and they experience blame for their condition including from friends and family, health professionals, and most particularly, the media. In common with wider literature on the deleterious effects of living with a stigmatised health condition (Hatzenbuehler et al., 2013; Major & O’Brien, 2005; Panza et al., 2018), perceived stigmatisation of diabetes is associated with internalisation of blame and shame, avoidance of social activities, concealment of the condition, and poor medical adherence and self-care (Himmelstein & Puhl, 2021; Kato et al., 2017; Shiu et al., 2003). These findings attest to the importance for public health policy of understanding community-level representations of diabetes and how these may (re)produce stigmatising public health discourses.

Below we review some of the complexities involved in regarding diabetes as a stigmatised condition (or set of conditions) and how the notion of diabetes as stigmatised is contested by health professionals and the wider community. We then consider social media (specifically Instagram) as a modern context for the observation and production of shared social representations. Finally, we report two studies examining stigmatic representations of diabetes on Instagram and how those representations are understood and contested by those living with diabetes as well as members of the wider public.

Definitional complexity of diabetes: good diabetes and bad diabetes

Within the medical profession, understandings (and contestations) of the aetiology of diabetes are framed by the historical differentiation made between two categories or types of diabetes (American Diabetes Association, 2018; Himsworth, 1936). Type 1 diabetes mellitus (T1DM) is described as unpreventable and ‘blameless’ (Greiner et al., 2001), whilst Type 2 diabetes mellitus (T2DM) is described as resulting from poor health choices (Porte & Kahn, 2001) rendering it the legitimate target of behaviour change messaging (Bombak et al., 2020). This delineation has in recent years been challenged by research showing that misclassification rates are high (Carmody et al., 2016); and that both forms are affected by genetic, environmental, and socio-economic factors (Raphael et al., 2012; Wang & Beydoun, 2007). However, despite definitional complexity, T2DM continues to be conflated with fatness (Bombak et al., 2020; McNaughton, 2013), and some have even suggested that Type 2 be renamed ‘diabesity’ (Astrup & Finer, 2000).

Diabetes as stigmatized condition

The notion that irresponsible behaviour causes diabetes dates to the 13^th century, with medieval writers asserting that the condition was due to ‘too much meddling with women’ (O’Donnell, 2015, p. 1105). Diabetes continued to be viewed as a preventable condition (caused by over-indulgence of sexual or dietary appetites) up to the 1850s, when the psychosocial distress associated with intellectual professions became the primary explanation. During the 1920s, the explanation reverted again to the personal indulgence thesis. According to (O’Donnell, 2015), these historical shifts were not reflective of any new forms of scientific evidence, but rather shaped by the condition’s transition from a disease of the rich to a disease of the poor.

We see in this brief historical account that diabetes has long borne the hallmarks of a stigmatised health condition entailing negative social judgement and blame (Weiss et al., 2006); and that it now takes the familiar form of personal indulgence (Schabert et al., 2013). Importantly, whilst both people with T1DM and people with T2DM report stigmatization, Liu and colleagues’ (Liu et al., 2017) survey of 12,000 people in the US found that those with T1DM were significantly more likely to perceive stigma and to experience it in the form of misunderstanding. This tallies with Browne and colleagues’ (Browne et al., 2013, 2014) observation that what was galling for T1DM participants in their research was not stigmatic representation and treatment per se, but rather that it had been misapplied to them because of mis-categorization.

Theorising differences in perceptions and responses to stigma

We bring the social identity perspective comprising social identity theory (Tajfel & Turner, 1979) and self-categorisation theory (Turner et al., 1987) to our understanding of stigmatisation as an inter- and intra-group process located in structural relations of unequal power and status. This perspective eschews more individualist and apolitical accounts that inform much psychological (Major & O’Brien, 2005) and sociological research (Tyler & Slater, 2018) on stigma. From this perspective, stigmatisation entails defining and categorising a group of people as outgroup on dimensions that promote positive ingroup distinctiveness (Oakes et al., 1994). Thus, in line with critical sociological theorising, stigma is mobilised to (re)produce social difference and inequalities (Parker & Aggleton, 2003) and draws on dimensions of comparison (e.g. competence and morality) that are defined in terms that are normatively valued by the group (Doise, 1978; Leach et al., 2007). For instance, in Western societies, discipline and self-control are normatively recognised and available core values associated with Protestantism and individualism (Joffe & Staerklé, 2007; Scambler, 2018).

Research informed by the social identity perspective provides valuable insights into how and why the notion of diabetes as a stigmatised condition may (or may not) be recognised and contested by majority and minority group members. First, there are experiential and motivational reasons for people with diabetes perceiving the frequency and impact of prejudice and discrimination differently to majority group members. Most obviously, members of devalued (sub)groups attend more to meta-perceptions (how they are viewed by powerful others) and are hyper-vigilant to threats of being the target of stigma or prejudice (Blackwood et al., 2015; Shelton et al., 2005). Majority group members by contrast, engage in less perspective taking (Lammers et al., 2008), are less exposed and attentive to stigma, and are motivated to view their group as moral (Leach et al., 2007).

Second, membership of a low status or devalued group has negative consequences for self-esteem and there are various strategies for dealing with this. Typically, the most favoured is ‘passing’; this involves dis-identifying from the group and where possible, concealing an identity experienced as shameful. We see this strategy at play where people avoid social situations that would entail revealing their condition (Shiu et al., 2003). Another strategy is sub-categorisation; this also entails an active distancing from the devalued identity and is in evidence in (Browne et al., 2013, 2014) research wherein those with T1DM contrast their blamelessness and moral rectitude with people with Type 2. Importantly, the above strategies atomise people and undermine any potential to achieve a positive diabetes identity (Walker & Litchman, 2021) and pursue either a social creativity strategy through changing value attached to the identity (e.g. see Saguy & Ward, 2011 for discussion of strategies associated with the Fat Acceptance movement); or a social change strategy, which in this instance would be one that sought to challenge and change social structural factors, which contribute to diabetes.

Diabetes and social media

People with diabetes report media – including journalism, health promotion, and popular culture – to be the largest source of stigma (Browne et al., 2013, 2014). Research on diabetes-related discourse in news articles (Gollust & Lantz, 2009) and twitter (Karami et al., 2018) bears this out. Because of the strong Diabetes Online Community (DOC) there has been much research interest in how and why people with diabetes use social media; this typically focuses on engagement related to the medical condition (e.g. social support: Oser et al., 2020). Although one recent study (Basinger et al., 2020) examined talk about stigma in 22 online diabetes forums, this research was unable to distinguish between T1DM and T2DM posters. What has not been considered is the ways in which, within mainstream online contexts involving the wider public (i.e. not sites for the diabetes community), stigmatic representations of diabetes are (re)produced, discussed, and even contested.

In keeping with Goffman (1963), we conceptualise stigma as located in the interaction between “the normal” (those who do not have diabetes) and “the stigmatized” (those living with diabetes). Accordingly, our focus is on how diabetes is represented within mainstream Social Networking Sites (SNS). SNS are a form of social media that allows users to form their own ‘profile’, connect with a network of other user-profiles, and view content generated by this network (Boyd & Ellison, 2007). The very nature of a user profile is relevant to identity. Indeed, previous research has concluded that SNS can be considered a mirror of offline social interaction, with a significant body of research linking SNS directly to theories of stigma as ‘spoiled’ identity (e.g. Belk, 2013).

The current research

The central concern of this research is how is diabetes represented and understood in everyday discourse in a contemporary social media context. For this purpose, we chose to focus the research on an existing hashtag “#DiabetesOnAPlate”, which has been included in over 6,000 posts on Instagram. Our reasons for doing so are three-fold. First, by linking diabetes and food, #DiabetesOnAPlate arguably constitutes a stigmatic representation and so allows us to look at the everyday and mundane mobilisation of stigma. Second, Instagram is the world’s leading photo sharing application (Kallas, 2018) with lifestyle (especially food), among the most popular photo categories. Third, Instagram is emblematic of the centrality of the visual in identity performance on-line; and the accompanying explanatory text (including hashtags, descriptions or comments), often forms the basis for meaning construction and contestation among audience members (Rose, 2012).

Study 1: everyday representations of diabetes on instagram

Our question in Study 1 was what is the nature of images associated with #DiabetesOnAPlate posts and how is diabetes represented in the accompanying text? Specifically, we are interested in whether and how these posts (images and text) support or contest stigmatic representations of diabetes?

Method

#DiabetesOnAPlate was chosen after preliminary searches of the hashtag #diabetes. Through Instagram’s API we identified and downloaded 881 publicly available posts that included the #DiabetesOnAPlate hashtag, between 1 July 2016 and 1 July 2017. A 12-month period was used to avoid the time of year affecting the type of posts collected (Webb et al., 2017). After removal of re-posts, posts by organisations (rather than individuals), and posts that could not be analysed (e.g. accompanying text was not written in English), 530 posts were retained for analysis. The posts were in a public domain and so consent was not considered necessary (McKee & Porter, 2009). We were mindful, however, that even where individuals have agreed to a web service providers’ End User Licence Agreements, they may not wish to be identifiable. Accordingly, faces were covered and usernames were removed in order to ensure anonymity and confidentiality. Preliminary tests using our data confirmed that the Instagram posts were not traceable. Given this and the concern that paraphrasing might alter a post’s initial message (Pitts & Smith, 2007), comments were quoted without adaptation. This study was approved by the University of Bath Ethics committee (PREC:17–203).

We conducted content analysis (Bell, 2001) for the images and reflexive thematic analysis for the text (Braun & Clarke, 2019). First, one researcher coded images and text drawn from a random sample of 100 posts and created two coding guides (one for images and one for text). Two researchers used the guides to code the same 50 posts. Consistent with content analysis, intercoder reliability was calculated for the images using Cohen’s Kappa (K = .82 ; Cohen, 1960). Following discussion among all authors of discrepancies and additional features, the coding guides were amended, and the full data set comprising 100 posts was coded.

The content of images was coded using binary judgments, to indicate presence or absence of a feature (see Figure 1 for example images). For instance, for each image we asked the question, is food that appears high in fat content present or absent?

Figure 1. Example Instagram images accompanying #DiabetesOnAPlate.

Reflexive thematic analysis (Braun & Clarke, 2019) of the text entailed a collaborative process; the authors refined and collated the codes based on patterns of shared meaning, returning to the transcripts and coding in an iterative fashion. The authors worked toward consensus on each theme while paying particular attention to “negative cases” or data that might challenge the themes (Mays & Pope, 2000).

Results

Our content analysis of images found that most (96.8%) included food; the foods depicted typically appeared to be high in sugar (92.9%) and/or fat (92.2%) and served in large portions (51.7%); and the context typically involved ‘eating out’ (e.g. at a restaurant: 52.2%). Thirty-two images featured people; of these, most appeared happy (76.3%) and did not appear overweight (86.1%). Finally, there were a small number of images (n = 10) depicting diabetes related medical equipment and health foods and portion sizes; these posts came solely from people who identified as having T1DM (See Supplementary materials for complete analysis of images in each coded category).

Our thematic analysis of text identified three key themes: (1) Positivity and lack of direct (enacted) stigmatisation towards people with diabetes; (2) Use of cultural frames relating to unhealthy eating and mitigation strategies; and (3) Counter-representations showing the ‘reality’ of living with diabetes. The posts associated with themes 1 and 2 were distinct from the posts associated with Theme 3; the latter accompanied the subset of images posted by people identifying with T1DM.

Theme 1: positivity and lack of direct (enacted) stigmatisation towards diabetics

Saturday well spent #baking #choccake #brucebogtrotter #chocoverload #diabetesonaplate #imafatty”

The food depicted in images was typically described in very positive and even rapturous terms (e.g. “Flymeoverthemoon” and “#dessertperfection”) and there was often a humorous and self-deprecating element (e.g. “Experimenting with properties of Nutella” and “put on weight by just looking at it”). Where obesity or weight gain was mentioned, it was always self-referential and not obviously directed at people with diabetes (e.g. “#imafatty”). Indeed, other than the hashtag #DiabetesOnAPlate, most posters made no reference to diabetes or diabetics (see Theme 3 for the only exceptions). Thus, the overall impression is one of innocence; people joyfully and playfully indulging in food.

Theme 2: the use of cultural frames of unhealthy eating and mitigation strategies

Easter Oreo brownies, let’s just say I found my sweet-tooth-tolerance level #OMG #diabetesonaplate #sugaroverload #toosweet #mustruntomorrow

However, something else was also present in the text that suggested posters were keenly aware of the available cultural representations of the images they were posting. Specifically, posters typically drew on recognizable frames for talking about ‘unhealthy’ foods in terms of poor nutritional value (e.g. many referred to the high sugar and fat content in the images posted); excessive portion sizes (e.g. inability to finish and food ‘overload’); and negative health outcomes (e.g. “major sugar coma” and “#thatsagooddeath”). Moreover, there was an additional nod to cultural frames around what constitutes unhealthy eating with posters drawing attention to their awareness of having transgressed acceptable norms of eating (e.g. referring to a ‘cheat’ and diets broken: “#cheatmeal”). These transgressions of eating norms were presented as aberrations; posters linked their transgressions to special occasions (e.g. weekend, holiday, or celebration); and spoke of intentions to take compensatory action (e.g. through running or dieting: “#fastingitofftomorrow”). Thus, there was evidence of a shared representation of what constitutes ‘unhealthy’ eating and through strategies of mitigating one’s ‘over’ consumption, an implicit recognition of the potential for negative social judgements.

Theme 3: counter-representations showing the ‘reality’ of living with diabetes

Check out this giant salad: lettuce, black beans, tomato, mango, with roasted portobello mushroom, corn, red cabbage, red onion, summer squash/garlic/jalapeño. #vegan #vegetarian #type1diabetes #type1 #t1d #diabetes #diabetesonaplate #eatyourveggies #wfpb

Finally, a small sub-set of posts (images and text) by people who identified themselves in their posts with T1DM, explicitly spoke to the ‘realities’ associated with living with diabetes, such as health management (e.g. images of needles accompanied by text such as“#glucosemonitoring”); and normatively virtuous healthy eating choices, such as vegan and vegetarian diets. Some posts also challenged the use of diabetes as the butt of a joke and drew attention to the seriousness of the condition through comparisons with cancer and the use of common metaphors for living with chronic illness (e.g. describing it in terms of a “daily battle”). As seen in the above example where #type1diabetes is used, these posts sometimes signalled that the poster identified as Type 1 and in so doing set themselves apart from judgement associated with T2DM.

Discussion

Three things stood out from our analysis of Instagram posts in Study 1, which were united only by their use of the hashtag #DiabetesOnAPlate, that are consistent with the notion of diabetes as a stigmatised condition. First, is the sense of a normative way of speaking about practices of (un)healthy eating that creates a moralised subject; someone who is responsible and answerable for unhealthy choices and adverse health outcomes and has mitigation strategies at their disposal. Second, is the implicit distancing from ‘spoiled’ identities entailed in the use of self-deprecating humour and management of moral transgression (e.g. through positioning behaviour as aberration and to be followed by acts of moral restitution). Third, consistent with wider social psychological research on prejudice and perspective taking (Lammers et al., 2008), majority group members appeared oblivious to stigma and its potential consequences. More prosaically, our analysis suggests that #DiabetesOnAPlate is associated with shared stigmatic representation; and whilst some employ strategies of disassociation, others (notably those with T1DM) actively contest this representation. However, on this latter point, we cannot be entirely sure of the diabetic status of all those who are posting and so of the reactions and strategies employed by those without diabetes nor those differentiated by T1DM versus T2DM. Study 2 was designed to address this.

Study 2: how those not living with diabetes and those living with T1DM and T2DM respond to #DiabetesOnAPlate

In Study 2 we look at how different groups (e.g. the wider public, and those living with T1DM and T2DM) engage with a stigmatic representation of diabetes. Here, we are interested in whether and how different groups recognise and contest stigma in a prototypical post taken from our #DiabetesOnAPlate data set.

Method

We recruited participants for an online survey on ‘trending hashtags on Instagram’, through a range of approaches: Online advertisements were shared on Instagram and Facebook and on diabetes organizations’ websites (www.diabetes.org.uk, www.diabetesdaily.com and www.diabetes.co.uk). Our sample comprised 67 participants (female = 33; male = 33, prefer not to say = 1) from three groups (No diabetes = 19; Type 1 = 25; and Type 2 = 23). Recruitment stopped once we had sufficient participants in each of the three groups (Hammersley, 2015). Consistent with later onset of T2DM (Health and Social Care Information Centre [HSCIC], 2014), there was a statistically significant age difference between groups, F(2, 64) = 20.09, p < .001; participants with T2DM (52.3 ± 12.3 years) were significantly older than participants with Type 1 (33.0 ± 14.2 years, p < .001) and without diabetes (28.68 ± 2.03, p < .001). Written consent was obtained prior to participation in the study. This study was approved by the University of Bath Ethics committee (PREC: 17–206).

Participants read a short explanation of the task and were then presented with a screenshot of a prototypical #DiabetesOnAPlate post taken from our Instagram dataset; the post depicted a meal with high sugar-content (see Figure 2).

Figure 2. Stimulus presented to participants.

Our reflexive thematic analysis (Braun & Clarke, 2019) of the text proceeded in a similar fashion to Study 1. We first coded all the responses blind to the participants’ group membership and identified patterns of meaning that could constitute clear themes. Once this was done, we organised the data according to group membership and looked at what meaning was shared and what was distinct.

Results

Our reflexive thematic analysis identified four themes: Attendance to the ‘literal’ meanings of the image and hashtags; questioning the poster’s motives; misrepresentation of the link between food and diabetes; and failing to distinguish between T1 and T2 diabetes. For each extract, group (N=no diabetes, T1 = T1 diabetes, T2 = T2 diabetes), gender and age is indicated in parentheses.

Theme 1: ‘literal’ meanings of the image and the hashtags

For many of our respondents, particularly those not living with diabetes, the association between diabetes and food was not questioned. Indeed, there was the suggestion that diabetes should be associated with extreme excess with two participants questioning whether the food represented in the image was sufficiently unhealthy to warrant the #DiabetesOnAPlate hashtag. Moreover, some commented matter-of-factly on the juxtaposition of the appeal of the food against the gravity of the health message.

It is saying that the waffle is a gorge treat but bad for you. Friends do this quite a lot but makes you think more about the calorific value of food and its associated health impacts than you would otherwise normally. defo wouldn’t stop me eating that waffle though (NM24)

In the above response, it is implied that such posts could be regarded in terms of normative practices of health messaging that can potentially have a positive effect (although not for this respondent who would eat the waffle anyway). Certainly, some commented that the reference to diabetes was “offputting” (NF22a) and “makes me not want to eat the food” (NF22b). One respondent took a very different view commenting that “It is irresponsible to normalise this kind of diet by plastering it all over instagram and associating it with jokey hashtags” (NF21). Regardless of the position taken, the responses described above suggest serious engagement with the post and it being viewed as a familiar (“Friends do this quite a lot” NM24) and unquestioned cultural framing of diabetes as caused by poor dietary choices. None in the group who were not living with diabetes found the #DiabetesOnAPlate message problematic and none made a distinction between T1DM and T2DM.

Some of our respondents living with T2DM (but not T1DM) responded similarly to those without diabetes; seeing the post as a harmless joke or a legitimate health message: “I think it means that the meal is unhealthy as it is the type of meal which could contribute to becoming diabetic” (T2F78). Where they differed was in seeing the health message as self-relevant: “I get that it’s making fun of the high sugar/calorie content of the food, but as a type 2 diabetic I wish I hadn’t been so flippant about foods that I ate” (T2F39). Indeed, even some of those who expressed dismay at the post did not question that the message was warranted (e.g. “I find it rude, even if actually a plate like that it’s not advisable for a person with diabetes” T2M46).

Theme 2: questioning the posters’ motives

The tenor of some comments from people not living with diabetes suggested that their sense-making about the Instagram post was shaped by their understanding of norms surrounding ‘#FoodPorn’ (Bienenstein et al., 2020). This was evident in expressed confusion about why someone posting a picture of food would accompany it with a more serious health message: “Does the person posting the food have diabetes hence their hashtags? Or are they trying to be funny?” (NF23). It was also evident in more dismissive responses:

I wonder how many attempts they took at getting the ‘right photo’? By the looks of the ‘flat’ whipped cream they must’ve been there a while! JUST EAT IT. (NM22)

The note of contempt in this response is echoed by others: “Some sad fuck seeking approval of their food” (NM22). Thus, for some of our respondents the Instagram post was viewed negatively and even seen as being offensive; but this was wholly unrelated to the (mis)representation of diabetes.

Theme 3: the misrepresentation of the link between food and diabetes

None of our respondents without diabetes raised issues that might signal recognition and concern about how the #DiabetesOnAPlate post (mis)represents diabetes. In stark contrast to this group, almost all respondents in the T1DM group and some in the T2DM group, commented on people’s ignorance and the misinformation about diabetes contained in the post. Some wrote only short comments: e.g. “Not necessarily diabetes, just obesity and rotten teeth” (T1F67); “It shows the posters ignorance” (T1M57). “I think this is just another example of how ignorant people are about diabetes generally” (T13F29). Where respondents did elaborate, the main source of contention shared by both T1DM and T2DM respondents was the misrepresentation of the link between the food in the image and diabetes. In the words of this T2DM respondent, the link is simply inaccurate: “Some of the hashtags imply that eating this will cause diabetes, which of course is incorrect” (T2F41). Below we have a T1 respondent who intimates that such inaccuracies are perhaps not unexpected.

The fact that waffles with banana and syrup isn’t particularly the highest sugar meal you can eat for breakfast, yet he has already made the connection between sugar and diabetes. (T1FX)

In commenting on the inaccuracy of this link, this respondent alludes to the availability of a frame or social representation; “he has already made the connection between sugar and diabetes” (T1FX). The two extracts below from T2DM respondents make explicit their perception of the #DiabetesOnAPlate post reflecting a ‘stereotypical’ view of the link between food and diabetes, particularly for T2DM.

Diabetes on a plate – from this photo it looks like someone is stereotyping how you get diabetes and is everything wrong about the public perception of diabetes. (T2F50)

As a type 2 diabetic I find this image represents the negative stereotypical view portrayed by tabloid journalism often accepted by the general public that type 2 diabetes is self- inflicted. (T2F60)

Others too expressed concern not just about the poster but the wider community the poster was seen to represent and the consequences of encountering this ‘ignorance’ in their daily lives: “I get tired of having to defend myself and explain that my diabetes is not due to what I’ve eaten” (T1F24).

Respondents described a range of emotional responses. Many spoke of ‘hurt’, ‘anger’ and ‘offensiveness’; particularly in relation to the #diabeticcoma hashtag, which was seen as trivialising a life-threatening event. Others, however, insisted that they were not offended and spoke instead of their greater frustration with those who do get angry “I get more annoyed when people get holier than thou about their diseases, I prefer to be able to laugh about stuff”, (T1F33) and the need to educate people “When people say these comments, it’s not a time to get angry, it’s a time to educate-correctly” (T1F28). We end this section with a comment from a T2DM respondent.

You get these sorta comments a lot these days unfortunately and I just smile and joke along. You know they don’t mean anything by it and hopefully understand it takes more than a few high sugary meals to get diabetes. (T2F33)

At first glance, it may seem that this respondent views #DiabetesOnAPlate as ‘just a joke’. She certainly recognises the expectation for her to “joke along”. However, her use of the word “unfortunately”, suggests that it is nonetheless experienced as hurtful.

Thus, whilst there was a shared understanding amongst many of the T1DM and T2DM respondents in our sample that the post entailed a misrepresentation of the food-diabetes link, there was some variability in how they responded or believed one should respond.

Theme 4: failing to distinguish between T1DM and T2DM

Our final theme is drawn solely from the Type 1 respondents, some of whom made clear that what was being misrepresented was specifically the link between food and T1DM. This was evident in the invocation of their Type 1 status when commenting on people’s ignorance and in their more explicit reference to the difference between T1DM and T2DM. For instance, the respondent below is clear that it is legitimate to attribute T2DM to lifestyle but not her own.

Annoyed. I’m not saying it is right or wrong but the hashtag and image is clearly making reference to Type 2 diabetes whereas I have Type 1 diabetes – an autoimmune condition and not connected to what you eat or lifestyle choices. (T1M49)

In some responses there was a more hostile distancing from T2DM than others.

I find any statement that does not distinguish the difference between Type 1 and Type 2 Diabetes offence. (T1M25)

There are two types of diabetes, there’s the fat greedy type and then there’s the type I’ve got … their diabetes could be prevented but ours could not. (T1M33)

In contrast, only one Type 1 respondent showed any sympathy for people living with T2DM:

I feel this image promotes hate in the form of discrimination against all people with Diabetes Type 1 and 2, it also spreads misinformation about Type 1 Diabetes and spreads the stigma that all people with Diabetes are obese. (T1F23)

What is noteworthy here is that although there is concern about the promotion of hate and discrimination against people with T2DM as well as those with T1DM, this concern does not extend to all members of those categories. That is to say, there is an inference that those who are obese might legitimately be the target of stigma and discrimination.

Discussion

Study 2 shed additional light on the apparent asymmetries in how people not living with diabetes and people living with T1DM and T2DM make sense of these posts. Importantly, the stigmatic nature of the #DiabetesOnAPlate was only explicitly recognised by those who have diabetes; people without diabetes appeared blithely unaware that any offence might be taken. Indeed, far from seeing the potential for harm, people without diabetes, the group who Study 1 suggests are the main posters of the #DiabetesOnAPlate hashtag, saw either humour or a potentially helpful health message. This perspective was shared by some with T2DM, but there was an important difference. Where those without diabetes could be playful about violating normative values around healthy eating, this was less clearly possible for those with T2DM; there was a self-relevance and in some cases self-blame and regret. Thus, people with T2 diabetes in the current study appeared to internalise the stigmatic representation of diabetes as self-inflicted (Browne et al., 2013; Corrigan et al., 2009).

Whilst a small number of people with T2 diabetes in Study 2 took exception to the post and two commented explicitly on the stereotyping of people with diabetes, the responses of this group were more circumspect and muted than the responses of those with Type 1. Only the latter appeared united in condemning both the stigmatic message as well as the messengers. Moreover, further emphasising the absence of solidarity between the two groups, even those T1 individuals who expressed some shared identity with T2, appeared to establish distance from those who they viewed as more legitimately blameworthy.

Conclusion

Taken together, our findings attest to the deeply entrenched, taken-for-granted assumptions in communities about diabetes as an illness that results from overeating unhealthy food and personal moral failings. This taken-for-granted or normative quality in everyday, on-line discourse gives warrant to claims of stigmatisation amongst those living with both T1DM and T2DM. So too does the evidence for how those not living with diabetes and the ‘stigmatised’ recognise the potential judgement entailed in the posts and the need to affirm one’s own morality.

There is a long-held view that stigmatising behaviours that compromise health (e.g. smoking, poor-eating habits) has societal-level benefits through encouraging the collective policing of these behaviours. Indeed, in the 1920s an influential diabetes specialist, Dr Joslin, suggested we should stigmatise the obese person in the same way the “drunkard is looked upon with pity and contempt” (Joslin, 1921, p. 81). There is, however, a body of evidence that perceived stigma and discrimination undermines psychological wellbeing and positive health behaviours (e.g. seeking assistance) and contributes to weight gain (Pascoe & Smart Richman, 2009). This view resonates with health policy initiatives such as the ‘Language Matters’ position statement developed collaboratively in England and beyond by people with diabetes and health professionals (Cooper et al., 2018).

The strategies observed in our data of distancing from those regarded as ‘legitimately’ saddled with diabetes stigma – e.g. through signalling overindulgence as an aberration or attributing to other categories of person – atomise people and compromise recognition and discussion of health messaging as well as collective resistance to messaging that is wrong or unhelpful (Walker & Litchman, 2021). Moreover, the emphasis on individual responsibility in diabetes (Ong et al., 2014), or indeed cultural group-level attributions (Hardin, 2015), draws attention away from social structural inequalities, which influence people’s health directly or through undermining collective agency; and may come at the expense of policy-change, which could address the issue at a societal level (Kelly & Barker, 2016). Health professionals and policy makers’ interventions need to be developed with this wider view of both the direct and indirect effects of health messaging on health outcomes, and how these articulate with societal-level processes that sustain stigma. There are three insights both from our research and the wider literature on prejudice and social change that can inform strategies to disrupt and challenge these processes.

1. Addressing diabetes stigma cannot be disentangled from addressing obesity stigma; indeed, any attempt to do so will only serve to reinforce both.

2. Stigma operates through shaming, silencing, and atomising. Rather than merely asking health professionals to examine and refrain from expressing societal biases a more radical approach is needed. Supporting patients with diabetes requires professionals having the ability to (a) recognise and speak openly about stigma and (b) work alongside patients to understand the operation of stigma and to develop both individual and collective strategies (e.g. through signposting to a wider range of groups – including those that are more political).

3. Health profession bodies have a potentially powerful role to play in publicly challenging stigmatic representations of diabetes (and obesity) and in shifting the public discourse regarding the underlying, societal-level contributions. This entails building on collaborations such as ‘Language Matters’ with an aim to influence not just the profession, but wider society.

In sum, the challenges for public health and health promotion are (a) to recognise the ways in which diabetes is stigmatised in professional as well as everyday discourse, and (b) to develop forms of communication about healthy eating for diabetes prevention, which do not in themselves contribute to stigmatising discourses.

Acknowledgements

Grateful thanks to Jamie Mahoney at Northumbria University for sharing the Instagram data with us.

Data availability statement

The data that support the findings of this study are available from https://doi.org/10.15125/BATH-01059. See Blackwood (2022). “#Diabetesonaplate”. Data for article in Critical Public Health. University of Bath Research Data Archive. https://doi.org/10.15125/BATH-01059.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.