Who Is (Held) Responsible for Diabetes and Depression? A Qualitative Interview Study Exploring Attributions and Reactions to Social Network Responsibility Frames

Responsibility frames and attributions of responsibility are closely linked to health-related stigma and social support intentions, which present relevant influencing factors for health outcomes. According to social-ecological models, health responsibility can potentially be attributed to at least three levels: 1) the individual, 2) the social network, and 3) society. So far, little is known about responsibility attributions to the social network. This qualitative interview study explores how N = 22 persons with and without lived experience with diabetes and depression react to a media frame attributing responsibility to the social network level, aiming to understand how framing and responsibility attributions are linked to health-related stigma and endorsement of social support. Results demonstrate that specifically type 2 diabetes is linked to individually controllable attributions and behavioural stigma, and individuals living with type 2 diabetes are expected to manage their condition without considerable social support. In contrast, depression is seen as less controllable, less manageable, and dependent on social and professional support. For both diabetes and depression, frames attributing responsibility to the social network may stimulate social support but also carry certain risks. These results offer implications for health news reporting, and perspectives for further research on health-related responsibility framing.


Responsibility Frames in the Media
Framing scholars such as Iyengar (1991) and Major and Jankowski (2020) argue that people's responsibility attributions regarding social problems, including health issues, emerge from news media. The responsibility frame is a generic type of emphasis frame which "presents an issue or problem in such a way as to attribute responsibility for its cause or solution to either the government or to an individual or group" (Semetko & Valkenburg, 2000, p. 96). A systematic review found that across various health issues and countries, the media attribute health responsibility predominantly to individuals rather than to society (Temmann et al., 2021). Studies about depression coverage found both individual and societal responsibility framing (Zhang et al., , 2016, while diabetes coverage is more individualised, and social determinants are often neglected (Gounder & Ameer, 2018;Stefanik-Sidener, 2013). A content analysis comparing German newspaper reporting of diabetes and depression also revealed that individual behaviours (healthy diet, losing weight, exercise) were frequently mentioned as prevention and treatment options for diabetes (Reifegerste et al., 2021b). Prevention options were almost twice as prevalent in the German reporting about diabetes as compared to the coverage of depression (p. 33), indicating that diabetes is portrayed as more easily preventable and controllable than depression.

Reception of Responsibility Frames
Experimental studies in the health context indicate that responsibility frames can affect attributions (Major, 2009;Shen et al., 2012;Sun et al., 2016), stigma (Frederick et al., 2016;Vyncke & van Gorp, 2018), emotions (Major, 2009(Major, , 2011, and behavioural intentions (Jin et al., 2018;Sun et al., 2016). Still, there are not enough effect studies understand how responsibility frames are perceived by the audience and how they interact with the subjective experiences and responsibility attributions of the recipients. Few qualitative studies provide insights into public views about health responsibility (Lundell, 2013a) and reactions to narratives about social determinants of health (Lundell et al., 2013b), but do not explicitly investigate subjective perspectives of recipients on responsibility frames. Studies consistently suggest that the predominance of individual-level attributions and media frames contributes to health-related stigma (Barry et al., 2013;Gollust et al., 2010;Frederick et al., 2016), while social and societal level frames facilitate social and policy support (Sun et al., 2016;Coleman et al., 2011;Authors, 2022, under review). However, few links have been made between responsibility frames, attributions, stigma, and social support from a recipient perspective.
Thus, despite a rich research landscape on responsibility framing of health issues, there can be two major challenges identified in the research: First, the recipient perspective on the interaction of responsibility frames, attributions, social support, and stigma has been overlooked so far. Importantly, the recipients' subjective engagement with media responsibility frames remains practically unexplored.
Second, based on social-ecological models, Reifegerste et al. (2021a) propose that responsibility frames can attribute responsibility for health issues to at least three levels: 1) the individual, 2) the social network, and 3) society. Still, most studies only contrast individual versus societal responsibility frames or confound frames at the level of the social network with the individual or societal level (e.g., Lawrence, 2004;. This is surprising as the social network level can also be responsible for health behaviours and health outcomes (Holt-Lunstad & Uchino, 2015;Sallis et al., 2008). These desiderata are underpinned by the results of a systematic review (Temmann et al., 2021), which found a predominance of content analyses (as compared to reception studies), and a conceptual gap at the level of the social network.

Objectives and Research Questions
Even though attributions, social support, and stigma in the context of depression and diabetes have each been relatively well researched so far, there is a missing link to the concept of responsibility frames, which are a crucial source of health-related responsibility attributions (Major & Jankowski, 2020). A recipient-centred perspective can help to link these concepts, and better understand responsibility attributions and endorsement of social support in the context of depression and diabetes. Based on the research gaps outlined above, the following four research questions guided the study: RQ1: How is responsibility for diabetes and depression attributed?
RQ2: How is health-related stigma regarding depression and diabetes linked to specific responsibility attributions?
RQ3: How are responsibility attributions related to the endorsement of social support in the context of depression and diabetes?
RQ4: How do recipients make sense of media frames attributing responsibility for depression or diabetes to the social network?

Sampling and Participants
To ensure an equal number of participants for both diabetes-centred and depression-centred interviews, a purposive quota sampling method (Campbell et al., 2020) was used to recruit participants. Studies on framing effects have shown that for frames to be processed by recipients, there must be a minimum level of knowledge and attitudinal content (i.e., cognitive schemas) that can be activated and made applicable through media frames (Matthes, 2007;Price & Tewksbury, 1997;Scheufele, 2004). Thus, to make sure that participants already had pre-existing schemas of diabetes and depression, they were required to be at least interested in either of these health issues to be eligible. Also, individual experiences with health issues may have an impact on cognitive schemas, responsibility attributions, stigma beliefs, and reactions to media frames (Corrigan et al., 2003;Rose et al., 2019;Valkenburg & Peter, 2013;Vyncke & van Gorp, 2018). This suggests that people living with a disease such as diabetes or depression are a particularly relevant audience for responsibility frames. In addition, the interviews are intended to shed light on how media recipients with no direct relationship to diabetes or depression react to responsibility frames at the level of the social network. Hence, participants were recruited from two relevant groups: 1) people who have lived experience with either diabetes or depression 1 , and 2) people without lived experiences but with at least an interest in one of these health issues. In the depression group (n = 11), the interview guide and responsibility frame focused on depression, and diabetes was discussed later. In the diabetes (n = 11) group, it was the other way around: the interview guide and stimulus focused on diabetes, and depression was discussed later in the interview. Overall, N = 22 German-speaking adults aged 19-77 years (M = 39.45, SD = 20.87, Mdn = 31) took part in the study.

Procedure and Interview Guide
Upon receiving ethics approval 2 and informed consent, the semi-structured interviews were conducted between December 2020 and March 2021. Due to pandemic restrictions, all interviews were carried out via phone or video calls. A qualitative approach was selected because the literature review revealed that more data is needed to understand responsibility frames from a recipient perspective, and to better capture subjective experiences with sensitive issues like health and illness (Low, 2012). A semi-structured interview guide provided the necessary openness for the interviewees to set their own priorities while allowing the interviewers to keep the conversation focused (Brinkmann, 2020).
Responsibility frames attributing responsibility for depression or diabetes to the social network were used as a stimulus. The frames were incorporated into a constructed news article (see supplementary material 3 ). To increase the similarity with real health news articles and thus external validity, an individual exemplar living with either diabetes or depression (episodic framing; Major, 2018) was embedded, since exemplars and episodic framing are often employed by journalists to illustrate health problems (Hinnant et al., 2013;Major & Jankowski, 2020). Furthermore, health issues like diabetes or depression are often sensitive and complex, making it challenging for interviewees to verbalise their thoughts. Therefore, the responsibility frame in the form of the constructed article also served as a "stimulation of personal recollections that are not always easy to represent" (Hopf, 2004, p. 206).

Data Analysis
The interviews were transcribed verbatim and analysed with MAXQDA 2020 following Kuckartz (2018), using the framework of a content-structuring content analysis (Mayring, 2014). Categories were both developed deductively from the research on responsibility framing, attribution theory and health-related stigma, and also inductively from the material (see Table 3).
Furthermore, inspired by the constant comparison method (Boeije, 2002), attributions and stigma beliefs regarding depression and diabetes were compared along three different dimensions: 1) frame reception, i.e., before and after reading the article within a single interview, 2) lived experience, i.e., between persons living with diabetes or depression, and those who do not live with either diabetes or depression, and 3) health issue, i.e., between diabetes and depression.

RQ1: Attributions of Responsibility for Diabetes and Depression
Overall, the interviews revealed that all participants had multifaceted attributions: None of them attributed responsibility to only one influence level, but always discussed multiple causes and treatments for both health issues after probing. Some interviewees explicitly mentioned that different causes potentially interact with each other, but this multicausality was discussed more in relation to depression than diabetes. Comparing cases with and without lived experience revealed that lived experiences were not necessarily linked to specific attributions. Instead, attributions were structured very individually, and informed by participants' subjective experiences and contexts. Margit, for example, has extensive contact with older people living with diabetes due to her job as a geriatric nurse: "I have a lot of older people at work. They are restricted. They are restricted in their movement, and now they don't go out so much [due to the pandemic]. Well, and then they are also restricted in their diet, because they can no longer eat everything." (Margit, 68, geriatric nurse) Consequently, she perceived age as a major uncontrollable cause for type 2 diabetes. Anke, whose husband lives with type 1 diabetes, emphasised the behavioural causes of type 2 diabetes to distinguish them from the uncontrollable causes of her husband's type 1 diabetes: "With type 2, it's the diet and lifestyle. If I eat too much fat, too many sweets, too unhealthy, and of course if I gain weight and don't exercise. That is the influence for type 2. With type 1… for my husband, […] they said that it was triggered by a spread flu. But it can also be inherited." (Anke, 56, relative type 1 diabetes) Attributions of Responsibility for Diabetes. By comparing attributions before and after reception of the responsibility frames, and especially attributions between diabetes and depression, clear differences could be identified. See Table 4 for an overview of the full range of diabetes attributions.
Diabetes was primarily seen as an avoidable lifestyle disease. All spontaneous causal attributions referred to individually controllable behaviours, predominantly diet and (lack of) physical activity ("When you eat too many sweets […] and eat too many fatty and unhealthy things, that promotes diabetes." Anke, 56, relative type 1 diabetes). After probing, most participants acknowledged that while "unhealthy" behaviours contribute to diabetes, its causes were not always fully controllable ("Well, the doctors like to say that this is also due to exercise, but an athlete can also get it." Renate, 74, living with type 2 diabetes). Non-controllable causes such as genetics and age were also recognised for type 2 diabetes: "Of course, you could say it's because of diet or body weight. But that's not always the case, I think it's mostly an inherited disease, and the person can't do much about it. It's already in the body." (Galina, 65, interested in diabetes) Only respondents with direct or indirect experience and more knowledge of diabetes (Mila, Anke, Noah, Aylin) explicitly distinguished between the different diabetes types. In their attributions, they especially emphasised that the onset of type 1 diabetes is neither behavioural nor controllable, and treatments are more complex compared to type 2 diabetes. Like Anke, Mila also expressed the concern that people might see type 1 diabetes as being as self-inflicted as type 2 diabetes due to a lack of awareness:   Causes at the level of society were only discussed upon questioning, especially regarding the prompted examples (poverty, food industry). The opinions of the interviewees on societal diabetes causes diverged considerably. While part of the interviewees in the diabetes subsample recognised these as relevant causes of type 2 diabetes, others objected and said that it was still possible for people to prevent it through healthy behaviours.
"I can understand the point about poverty to some extent. Because poverty often leads to an unhealthier lifestyle.
[…] But basically, everyone can choose what they eat.
[…] I can understand the idea to a certain extent, but I can't really support it." (Aylin, 22, living with type 1 diabetes) Table 5 for all depression-specific attributions. Depression was perceived as an unpredictable mental illness with multiple, sometimes undeterminable causes, a condition which could potentially affect anyone, and which had consequences for both individuals and their social network.

Attributions of Responsibility for Depression. See
Most participants in the depression sample referred to the aspect of multicausality unprompted. Spontaneous causal attributions referred mainly to non-controllable factors in the social environment, i.e., traumatic life events, childhood experiences, an unsupportive social network, bullying at the workplace or online.
Solutions for depression on the part of the social network were also diverse, especially after the reception of the article, and included increased sensitivity in dealing with affected persons (e.g., suicidal thoughts), involvement in therapy, and various forms of emotional and tangible social support (e.g., going for a walk together, talking, making calls). On the societal level, causes for depression were less differentiated but still acknowledged; they mostly referred to pressure to perform due to modern meritocracy, which was prompted by the interview questions.
In addition, bullying and hate on social media were discussed as possible factors that might lead to a rise in depression at the population level: "Way back in school days, you just avoided [the bullies] or had a few people lying in wait for you, but that was it. Today, some idiot can inform the whole world about something and cause a shitstorm." (Gerhard, 64, living with type 2 diabetes). At the same time, social media was also seen as an opportunity to de-stigmatise depression, especially in younger generations where depression was perceived to be less stigmatised in general: "I think we know more about it, are better informed -through social media, through celebrities and stars who can talk openly about it. And for us this taboo has also gone away a bit, but at our age we are confident enough to talk about it, much more than perhaps our parents or our grandparents." (Antonia, 20, relative depression) If respondents attributed causal responsibility for depression to individuals, they mostly discussed causes they thought to be relatively stable, i.e., personality traits like neuroticism or introversion, or low resilience. Regarding treatment, all interviewees in the depression group expressed in one way or another that people with depression have a responsibility to communicate openly with their environment, to seek help and to take measures to improve their everyday life (e.g., improve coping with stress, exercise, diet). However, it was consistently pointed out that people were not expected to cope with their depression alone, and that help seeking and behavioural solutions were often complicated or rendered impossible by the symptoms of depression (e.g., lack of motivation: "You could start working on your

RQ2: Issue-Specific Stigma
In terms of diabetes stigma, the interviews revealed two seemingly paradoxical findings: On the one hand, diabetes was assessed as less stigmatised as compared to depression "because it's such a physical illness" (Lea, 20, relative type 2 diabetes). At the same time, many of the same respondents (both with and without lived experience) made implicitly stigmatising statements about diabetes, relating diabetes to unhealthy habits, negligence, and high body weight. Interviewees frequently created a link between type 2 diabetes and specific "unhealthy" foods (pizza, fast food, sugary drinks, cake), which was not prompted by any of the interview questions.
"If a person is very fat and eats a lot of pizza, these things that are just not healthy, it is obvious that the body can react with diabetes at some point. Because obesity is often accompanied by a lack of exercise […]. Especially with obesity or overweight, people are more likely to get diabetes." (Renate, 74, living with type 2 diabetes) Although the interviews did not reveal explicit instances of discrimination against people with diabetes, the underlying assumptions reflect the (social and internalised) stigma associated with type 2 diabetes, and due to a lack of distinction in the public discourse, also to type 1 diabetes. The fact that the persons with type 2 diabetes in the sample also made these stigmatising statements indicates that they have already internalised the assumption that diabetes is self-inflicted.
Depression, in contrast, was perceived as less manageable, less tangible, and more difficult to overcome. Most interviewees shared the opinion that depression, like a physical condition, required medical treatment and should be recognised as a 'real' illness like any other, e.g., Mila (19, interested in depression): "[Depression] is simply a disease of the psyche, just as there are diseases of the body." Throughout the interviews, the recovery perspective was seen as much more pessimistic for depression than for (type 2) diabetes. For example, when asked how she would react if a friend would be diagnosed with depression vs. diabetes, Lea expressed concern that depression could "never really be cured".
This finding reflects two different, issue-specific manifestations of health-related stigma. While type 2 diabetes is perceived as a straightforward, mostly behavioural issue with simple, behavioural solutions, depression is seen as a less tangible illness with more uncontrollable causes and complex, multi-level treatments.

RQ3: Endorsement of Social Support
The different responsibility beliefs about depression and diabetes have consequences for people's endorsement of social network responsibility and social support. Because attributions for type 2 diabetes are so individualised, most respondents -even some living with type 2 diabetes -could not imagine that the social network could play any role in the development of type 2 diabetes. If they did, it was associated with specific pre-conceptions; Anke, for example, agrees to some degree to social causes of type 2 diabetes, but reduces it to poor families who Comparing different support behaviours between diabetes and depression, the analysis revealed that not only the endorsement of, but also the forms of support differed considerably between the two health issues. Diabetes support was mostly limited to tangible support like being considerate of special dietary needs ("Inviting people to beer and sweets and coffee or fast food all the time, that should certainly be avoided. […] I would say that should be enough support, the omission of negative things." Gerhard, 64, living with type 2 diabetes), or in some cases even social control like restricting exposure to sugary foods: "I would maybe say: 'Pull yourself together and stop for a bit, that's unhealthy and it will only make your diabetes worse!' But that's all I can do." (Christa,77,living with diabetes,unspecified) Respondents often revealed they would be less considerate towards someone with diabetes, and more careful in their choice of words towards a person with depression, again illustrating the different stigmatising beliefs related to both health issues.
"I think the subject of depression is much more sensitive and that people are much more careful about it. So, with the third piece of cake, I would say [to someone with diabetes]: 'Is that possible? Can you inject so much that you can eat it?' I would simply say that without having a guilty conscience. In the case of depression... because you know that people are so emotionally on the edge and very sensitive and so completely thrown out of their mental balance, I would ask this question much more cautiously." (Sabrina,39,interested in depression) Overall, social support for depression has more variety than diabetes support and considers the emotional vulnerability of affected individuals -partly to the point that people are afraid to say something 'wrong', because their illness is perceived to make them emotionally unstable and over-sensitive.
While social support for depression was seen as highly relevant in all interviews, participants also spoke from experience that depressive symptoms might hinder the affected person from accepting support: "You can offer it [support], but whether the person accepts it is something else entirely. And that is not necessarily the responsibility of the others." (Leonie, 20, recovered from depression).
A more specific case is that of two respondents who have lost a loved one to suicide because of depression (Marion and Jennifer). Both discussed the difficulty of recovering from depression, particularly from their experiences as close relatives. They underlined that the symptoms of depression posed obstacles for individuals to seek help and accept social support, mitigating both individual and social network responsibility. Consequently, Jennifer stated that relatives of persons with depression should set boundaries, and that she would always suggest professional support first: "Because if I say something wrong and the psychotherapist says something completely different, it confuses the person even more." (Jennifer, 35, relative depression) This statement also ties in with the previous finding that people are often scared to do or say something 'wrong' to individuals with depression, because their illness is perceived to make them emotionally unstable and over-sensitive.

RQ4: Interaction with the Social Network-level Responsibility Frame
During and after exposure to the social network frame, the interviewees regularly referenced their pre-existing responsibility beliefs. The social network frame did not alter the recipients' previous attributions but made causes and solutions at the social network level more accessible or solidified previous attributions.
The fictional person with depression or diabetes in the article triggered both recollections of own experiences or memories, and elicited intentions of social support from participants when being asked how they would react if the fictional person was their friend.
"I would tell her: 'There is a way out. Type 2 is curable. And we're going to tackle it now.' And then I would say, 'Come on. Then let's go for a walk once a day in the evening for an hour. It's a good time to talk. Or we can cook at the weekend. Or we can go to the gym.' There are so many things. I would give her so much support so that she doesn't have to go through it alone." (Anke, 56, relative type 1 diabetes) Some participants even theorised that media coverage including both individual exemplars and social responsibility may stimulate empathy and instruct relatives how to provide social support.
"With relatives, it could make some of them just wake up and think about how they can help these people better. The closest relatives and friends will help the affected person anyway, but an article like this sharpens the senses to see what more I can do." (Galina,65,interested) After exposure to the social network frame, Leonie (20), who is recovered from depression, criticised the individual framing of treatment responsibility she often notices in the media coverage about mental illness. She echoed the opinion of most other interviewees that depression cannot be treated individually, but only with social or professional support: "There were a few [articles] where it was just like 'And she did it on her own […]', and you're like 'You sure?' (laughs). Of course, there's people who can do it on their own. But sometimes I think it's kind of wrong to send messages like 'Hey, you can do it on your own.' Some people just can't do it all on their own." The person portrayed in the article elicited unequivocally positive reactions like empathy and support intentions, which was also related to the perception that she took responsibility into her own hands by openly seeking and accepting help from others: "First of all, I think it's really good that she is reflective and that she has actually realised that I can't get any further at a certain point. I really need help and I also want to work." (Heike,48,psychotherapist) In some interviews, however, the combination of social network responsibility frame and the exemplar triggered unexpected negative reactions. Finja thought the positive portrayal of depression recovery was oversimplified and too optimistic: "But I also don't know if she'll end up being 100% clear yet, because it just sounds too good to be true. It's somehow too superficial. It doesn't go into enough depth to convince a person who has no contact with [depression] that it's something serious." (Finja, 18, relative depression) Kim (19, interested in depression) felt the fictional person seemed "really insecure", "lacking willpower", and "helpless". Despite the agreement to social causes of both depression and type 2 diabetes, interviewees did not agree with a too biased portrayal of social network responsibility, criticising that it might absolve individuals of their own responsibility and motivation to recover: "I think that external influences are often exaggerated, and people are not reminded enough that they can also do something themselves and that they are responsible." (Gerhard, 64, living with type 2 diabetes) Especially in the context of diabetes, where individualised attributions were predominant, the social network responsibility frame disrupted expectations of individual health responsibility and the exemplar was perceived as being "too dependent" (Lea, 20, relative type 2 diabetes) and avoiding responsibility.

Discussion
This qualitative interview study with N = 22 participants contributes to health communication by 1), linking the concepts of responsibility framing, attributions, stigma, and social support, 2), extending the concept of responsibility frames to the level of the social network and 3), including subjective perspectives of recipients.
Overall, the respondents' attributions were differentiated and multifaceted, both for diabetes and for depression. External influences such as the food industry (diabetes) or modern meritocracy (depression) as well as the family (both) were generally acknowledged. Nevertheless, it was also revealed that despite all external influences at the level of the social network and society, ultimately the responsibility always falls back on the individual living with the disease. In this respect, the results reflect the predominance of individual responsibility frames in health news (Temmann et al., 2021).
Regarding stigma, the interviews revealed an interesting paradox: On the one hand, type 2 diabetes was perceived as less publicly stigmatised than depression because of its physical nature, yet many respondents made implicitly stigmatising statements and expressed diabetesspecific stereotypes, specifically the belief that type 2 diabetes is self-inflicted through negligence and an over-consumption of unhealthy foods. Thus, this finding partly contradicts earlier findings that physical illnesses are less stigmatised (Weiner et al., 1988), but rather reflects the highly individualised and morally charged discourse around food choices and body weight as possible risk factors for type 2 diabetes Himmelstein & Puhl, 2021;Stefanik-Sidener, 2013). This stigma was also applied to people living with type 1 diabetes, which is in line with an insight by Schabert et al.: "Blame for self-infliction of the condition may also affect those with type 1 diabetes by association, (particularly as media reports rarely make any attempt to distinguish the two conditions)" (2013, p. 7). This insight can be further supported by the present results, as people living with type 2 diabetes did not feel explicitly stigmatised but expressed more implicit, internalised stigma. At the same time, respondents with direct or indirect lived experience with type 1 diabetes expressed their frustration that shame and stigma are unfairly attributed to them as well, due to insufficient differentiation between diabetes types in the media discourse. To address this problem, media should 1) better differentiate between diabetes types and 2) reduce the behavioural stigma associated with type 2 diabetes and diabetes in general. The more uncontrollable causal attributions for depression in contrast to diabetes reflect low individual responsibility and low behavioural stigma. The interviewees mostly had a high awareness that depression is a social and societal problem, whereas diabetes (especially type 2) was seen as an individual problem. Still, it cannot be inferred that depression is less stigmatised than diabetes based on the interviews. On the contrary, depression was labelled as unpredictable and difficult or even impossible to treat. These attributions and labels bear resemblance to essentialism, i.e., the belief that depression is an unchanging characteristic of a person (Peters et al., 2020). Previous studies have shown that this perception might increase perceived dangerousness, aversion, and a pessimistic view on recovery (Loughman & Haslam, 2018). In this regard, the way media present prevention and treatment options for depression and diabetes (e.g., Reifegerste et al., 2021b) was reflected in the interviews: albeit diabetes was perceived as more individually controllable, it was also seen as better preventable easier to treat than depression.
Interestingly, some interviewees perceived depression to be less stigmatised in younger generations, while studies examining the public stigma towards depression consistently show that depression stigma did not decrease in recent years (Schomerus & Angermeyer, 2017;von dem Knesebeck et al., 2015). This paradox might be partly explained by the predominantly young and female sample in the depression group (see Limitations).
Regarding social support, respondents believed that it could not be effective in managing type 2 (in contrast to type 1) diabetes -although studies demonstrate significant positive effects of social support in this context (Strom & Egede, 2012;Yokobayashi et al., 2017). Depression, on the other hand, was perceived as dependent on social and professional support. This finding corresponds to a medical model of health responsibility, in which people are largely freed of responsibility for both causes and treatments (Brickman et al., 1982). Still, in line with Parson's sick role (Parsons, 1975;Varul, 2010), people with depression are expected to seek and accept medical and social support. Type 2 diabetes corresponds more to a moral model of health responsibility (Brickman et al., 1982;R. C. H. Brown, 2013), in which individuals are held responsible for both the causes ('unhealthy' diet, lack of exercise) and solutions (management of body weight and blood glucose, dietary control, exercise) to their health issue, and are expected to deal with it largely without social support.
The individual exemplar portrayed in the article contributed to the external validity of the study since exemplars and episodic stories are highly prevalent in news coverage about health (L. D. Brown et al., 2018;Hinnant et al., 2013;Iyengar, 1991). The combination of a fictional exemplar and the frame attributing responsibility for diabetes or depression to the social network turned out to be a double-edged sword: On the one hand, it showed potential to educate people about the social influences in the health context and encouraged social support. On the other hand, a one-sided emphasis on social network responsibility in combination with an individual exemplar carries the risk of evoking negative reactions (especially the perception of avoiding or deflecting responsibility), which could potentially be extended to other affected individuals (see representativeness heuristic; Zillmann, 2006, p. S223

Limitations
The major limitation of this study is rooted in the size and composition of the sample. N = 22 persons with and without lived experience with diabetes and depression were recruited to provide diverse perspectives on the questions at hand. Since two health issues were addressed concurrently in this study, the subsamples (n = 11) are too small to be able to systematically differentiate between the different perspectives and backgrounds, including relatives and experts. In addition, the sample does not include enough respondents with lived experience in the depression subsample (n = 2), and not enough male identified respondents (n = 20 female identified, n = 2 male identified) to make meaningful comparisons between these groups. Furthermore, no causal inferences are possible regarding the effects of specific frame dimensions or lived experience because of the non-standardised, qualitative study design. Mindful of these limitations, this study provides a first exploration of reactions to media frames attributing responsibility to the social network level, linking the concepts of responsibility framing, social support, and stigma in the context of diabetes and depression.

Implications for Research and Health News Coverage
Regarding diabetes of all types, more education should be provided to emphasise the importance of social support while reducing stigma and individual blame. In addition, there needs to be greater awareness of the different causes and pathogenesis between diabetes types, which would ultimately benefit all individuals living with diabetes. In relation to depression, the results underpin claims by previous research (Vyncke & van Gorp, 2018) that more studies are needed to understand the interplay of different responsibility frames, essentialist beliefs, and stigma. This study illustrates that social network responsibility frames might help to overcome the criticised over-emphasis on individual responsibility in media coverage about health, but they should also be employed with caution due to possible unintended reactions towards affected individuals. Since individual responsibility for health is 1), a value deeply rooted in many societies and entwined with role expectations (Sei-Hill Kim et al., 2017;Traina et al., 2019;Varul, 2010;Wikler, 2002) and 2), essential for adherence and self-efficacy in the context of medical treatments (Martos-Méndez, 2015), it cannot be the goal to abandon individual responsibility framing altogether.
Lastly, the interviews showed that both persons with and without lived experience with diabetes and depression attribute responsibility to multiple levels at the same time -including the individual, society, and the social network. Consequently, media reporting on the causes and treatments of health issues should not be overly simplistic, but rather consider socioecological perspectives on health responsibility that go beyond the individual.

Notes
1. This study did not differentiate between nonclinical and clinical depression, as previous research does not indicate this distinction would influence the participants' attributions, stigma beliefs, or reactions to media frames.
2. The study was approved by the board of ethics of the University of Erfurt (record number 20210107).