Attitudes towards mental health, mental health research and digital interventions by young adults with type 1 diabetes: A qualitative analysis

Abstract Background Young people with type 1 diabetes are at increased risk of mental disorders. Whereas treatment need is high, difficulty recruiting young people with type 1 diabetes into psychosocial studies complicates development, testing and dissemination of these interventions. Objective Interviews with young adults with type 1 diabetes were conducted to examine attitudes towards mental health and mental health research, including barriers and motivators to participation in mental health studies and preferred sources of mental health support. The interviews were audio‐taped, transcribed and evaluated via thematic analysis. Setting and participants Young adults with type 1 diabetes were recruited via social media channels of 3 advocacy organizations. A total of 31 young adults (26 females and 5 males) with an average age of 22 years were interviewed between October 2015 and January 2016. Results Participants were largely unaware of their increased vulnerability to common mental health problems and knew little about mental health research. Major barriers to participation included perceived stigma and lifestyle issues and low levels of trust in researchers. Opportunities to connect with peers and help others were described as key motivators. Psychological distress was considered normal within the context of diabetes. A need for some level of human contact in receiving psychosocial support was expressed. Discussion and conclusion Findings provide valuable insights into the complex dynamics of engaging young adults with type 1 diabetes in mental health studies. Interviewees provided practical suggestions to assist investigation and delivery of psychosocial interventions for this vulnerable group.


| INTRODUCTION
Prevalence of mental disorders is greatest in adolescence and early adulthood, 1 and adolescents and young adults with type 1 diabetes are an especially vulnerable group. 2,3 Daily self-care of type 1 diabetes involves complex activities that are often highly intrusive into daily life including monitoring of blood glucose levels and hyperglycaemic and hypoglycaemic symptoms, assessing carbohydrate intake with insulin treatment (multiple injections or infusion from a pump); and meal-tomeal adjustment of insulin dose depending on diet and physical activity. How well young people cope with the emotional and behavioural challenges of type 1 diabetes, while at the same time negotiating the normative developmental transition to adulthood, has important implications for long-term physical and mental health, quality of life and life expectancy. 4,5 Unfortunately, data suggest that many young adults with type 1 diabetes do not seek face-to-face support for their emotional health needs. 2,6 Despite a clear need, research into development, evaluation and dissemination of novel mental health interventions for young adults with type 1 diabetes is limited, possibly because of difficulties recruiting this group for psychosocial studies. 7 Young adults, generally, are less likely to volunteer for psychosocial research for various pragmatic and lifestyle reasons including scheduling conflicts (due to work, education and other responsibilities), lack of interest and perceived personal irrelevance of the research focus and treatments. 7,8 Recruitment of young adult men is complicated further by high mental health stigma, inadequate knowledge of mental health signs and symptoms, and privacy and confidentiality concerns. [9][10][11][12] For researchers targeting young people with type 1 diabetes, low disease prevalence in the 15-to 24-year-age group (approximately 0.2%) is an additional recruitment barrier. 6 Direct referral to psychosocial studies by health practitioners can be a valuable and cost-effective method for recruiting people with medical conditions. 13 However, low rates of attendance at diabetes clinics and medical appointments by young people with type 1 diabetes 14 mean that even this targeted recruitment approach is likely to yield only a small group of eligible candidates.
Comfort, confidence, proficiency and engagement with digital technologies are frequently offered as reasons for using social media channels and Internet outreach to access young adult candidates for research studies. [15][16][17][18][19] At the same time, evidence suggests that recruitment of young adults with medical conditions may still benefit from adjunctive local strategies encouraging direct referral via healthcare providers in specialist programs, clinics and services targeting young people. 15 We adopted both recruitment methods in a recent controlled study of a broadly available, mobile phone and web-based self-help program for reducing mental health symptoms and improving positive well-being in young people (aged 16-25 years) with type 1 diabetes. Specifically, we liaised with stakeholder and advocacy groups to develop trial messaging for dissemination via a range of social media outlets, including Facebook, Twitter and appropriate websites. At the same time, young people attending type 1 diabetes "transition" services (ie, supporting transition from paediatric to adult health services) were approached directly by clinic staff during routine follow-up appointments. Despite an exhaustive and laborious 12-month recruitment campaign involving more than 40 different social media and Internet options and ongoing direct referral, we were unable to enrol sufficient young people into our study.
Understanding how best to increase recruitment into clinical research trials is gaining increased scientific interest, [20][21][22] with research suggesting that effective strategies are unlikely to be universal, but to vary as a function of the target group and study methodology. 23 At the same time, identifying the factors impacting research participation decision making seems critical for designing interventions aimed at optimizing trial recruitment. 20 For this reason, and in the light of our recruitment experience, we decided to examine the factors influencing decisions made by young adults with type 1 diabetes in regard to their mental health and participation in mental health research.
Specifically, the study used qualitative methods to obtain an indepth understanding of: • the attitudes and feelings of young adults with type 1 diabetes towards mental health and mental health research and • the enthusiasm of this cohort for digital mental health interventions.
Our aim was to gain vital information to inform the design, promotion and recruitment methods of future studies of psychosocial interventions for young adults with type 1 diabetes. Discussion and conclusion: Findings provide valuable insights into the complex dynamics of engaging young adults with type 1 diabetes in mental health studies.
Interviewees provided practical suggestions to assist investigation and delivery of psychosocial interventions for this vulnerable group.

K E Y W O R D S
digital interventions, psychosocial support, type 1 diabetes 2 | METHODS

| Participants
In-depth interviews and focus group discussions were conducted with young Australian adults (18-30 years) with type 1 diabetes between October 2015 and January 2016. Participants responded by email to advertisements placed on Facebook and Twitter, announcements on websites of local diabetes advocacy agencies and/or were referred by staff of transition services in NSW.
Potential candidates were contacted by phone for screening by one of 2 researchers who are clinically trained (VV, JC). Candidates were eligible to take part in the study if they obtained depressive symptoms scores in the mild-to-moderately severe range (ie scores 5-19 inclusive) on the Patient Health Questionnaire-9 (PHQ-9; 24 ), and/or reported diabetes distress in the moderate range (ie scores greater than or equal to 3) on the Diabetes Distress Scale-2 (DDS-2; 25 ). Return of a signed consent form was also a criterion for eligibility. Potential candidates who scored in the severe range on the PHQ-9 were recommended to seek face-to-face support and provided advice and information about self-help resources and crisis mental health services.
A "sampling to saturation" recruitment method was used in which data collection continued until no new themes emerged from the interviews. Participants received a $40.00 gift voucher as appreciation for their time and contribution. The study was approved by the UNSW Sydney's (The University of New South Wales) Human Research Ethics Committee (HC15519).

| Analysis
Data analysis was conducted primarily by one of the authors (CM) and was guided by Braun and Clarke's 26 thematic analysis framework. This involved working sequentially through each of the transcripts, making notes about potential themes, key observations and highlighting each participant's key attitude, preference concerns and experiences. The transcripts were then reviewed a second time to ensure that nothing had been missed, and initial codes were discussed with authors JP and JC. NVivo 11 qualitative research software (QSR International Pty. Ltd., Victoria, Australia) was then used to facilitate analysis, and codes were then further grouped and organized. The organized themes and subthemes were then reviewed and agreed by CM, JP and JC.

| Participants
As shown in Table 1, thirty-one (31) young adults with type 1 diabetes were recruited for this study. The sample was predominantly female (n = 26) with an average age of 22 years (SD = 3 years) and an average age of diabetes onset of 13 years (SD = 6.5 years). Sixty-five per cent (n = 20) of participants screened positive for diabetes distress, and the average score on the PHQ-9 was in the "mild" range (x = 7.9; SD = 4.1).
Only 5 participants had previously taken part in mental health research.
Two focus groups were conducted, each comprising 4 participants.
The authors (JC and VV) each lead one focus group, with the other author observing to ensure consistency of approach. The remaining 23 participants were interviewed by one of the authors (either JC or VV). Table 2 summarizes the major themes and subthemes that were derived from participant responses regarding attitudes towards mental health. Almost all participants were unaware of the relevance of emotional well-being for health outcomes and longevity in people with type 1 diabetes. Many described diabetes management as overly preoccupied with "numbers" (ie controlling blood glucose levels), with little attention to helping young adults "live" well with the disease. Some participants described raising emotional issues in the context of routine diabetes care, however, felt that their concerns had been either minimized or dismissed by their treating health professionals. Even participants with no previous mental health diagnosis reported struggling from time to time with a broad range of emotional issues with varying levels of severity, including shame, guilt, fear, low motivation, loneliness, helplessness and hopelessness. Overwhelmingly, these difficulties were considered a normal response or "side-effect" of living with a complex and demanding illness. Participants felt that it was inappropriate to "overpathologize" these experiences by applying diagnostic labels to them.

| Potential barriers to and motivators for participating in mental health research
Tables 2 also summarizes the major themes and subthemes that emerged from discussion of the potential barriers and facilitators to participation in mental health research.
By far, the greatest perceived barrier for participants was poor understanding of the research process generally. Specifically, most interviewees admitted to knowing little about the purpose of mental

| Strategies for engaging young adults in mental health research
Common major themes and subthemes that emerged in participant responses to questions about how best to engage with young adults are shown in Table 3.
Almost overwhelmingly, participants emphasized the need for better education for the community generally, people with type 1 diabetes and diabetes health professionals about the link between mental and physical health in type 1 diabetes. They felt that raising awareness would assist with: • "normalizing" the emotional struggles that many young adults with type 1 diabetes face, • facilitating more open and "comfortable" discussion about mental health issues (by young adults and health-care providers), • emphasizing the importance and personal relevance of mental health research; and • making research participation a less stigmatizing and more acceptable option for young adults with type 1 diabetes. Participants generally agreed that mental health research was potentially a "hard sell" to young adults, due largely to stigma-related concerns. Low tolerance was expressed for messaging that made explicit reference to diagnostic labels (eg "depression" and "anxiety").
Instead, they preferred a subtler approach that normalized emotional issues as part of the experience of living with type 1 diabetes.
The following suggestions were offered as examples; "Does diabetes stress you out?," "Feel like a mood tune-up from time to time?".
[ ] it is okay to feel down, it is okay to struggle, everyone occasionally will go through this.

| Sources of mental health support
Young adults nominated 3 preferred sources of psychosocial support, namely peer-supports, online tools and face-to-face counselling (See Table 3). Overwhelming, peer support was discussed as a means of facilitating connection with others and reducing feelings of isolation and loneliness. Both online and face-to-face peer support options were discussed as providing opportunities for young adults to share their diabetes experience without feeling judged, to seek opinion on practical issues and to obtain advice when required. "Chat rooms," in particular, were mentioned as options for ensuring that peer support was provided in a way that ensured participant anonymity and privacy. Online and unguided (ie no human support) were described as providing the benefits of anonymity, privacy and convenience.
However, the extent to which they could be sufficiently supportive and able to assist with "serious" emotional issues was questioned by some participants, who fell that these tools were generally impersonal and incapable of responding empathically and with encouragement to a young person's unique emotional needs. There was greater confidence in digital tools that included human contact in the form of peer support, mentoring or therapist support, for example chat rooms, discussion boards and Skype facilitated therapy.

Major themes Common subthemes
Engaging young adults with type 1 diabetes in mental health research It was further suggested that these tools might have the advantage "breaking the ice" for someone in need of more structured face-toface therapy.

It's helpful to hear a voice behind it [ ] not just see a screen with information on it. That's how I am in terms of
websites.
(Female, age 20, diagnosed at 9 years) There was a unanimous preference for support for serious emotional difficulties to be provided in the face-to-face context.

| DISCUSSION
Findings from this qualitative study suggest that attitudes held by young adults with type 1 diabetes towards mental health are complex. Comments made by our participants suggest that this group conceptualizes the relationship between emotional struggles and type 1 diabetes in a nuanced way. While many participants reported feeling stressed, lonely, isolated, hopeless, ashamed, guilty and sad from time-to-time, these were discussed as side-effects of living with a chronic and complex disease, in other words "normal," and not as mental health issues per se.
The use of diagnostic labels to reference psychological distress within the context of diabetes, such as depression and anxiety, was met with some resistance by our participants. One possibility is that this represents an "externalizing" (ie beyond my control) of emotional distress, which reflects either a conscious or unconscious effort on the part of young adults to reduce externally imposed stigma and internally constructed self-stigma. Indeed, attribution theory predicts that external causal predictions of negative situations generate a more positive self-outlook than internal attributions. 27,28 At the same time, these findings provide support for the view that emotional distress in diabetes is more appropriately considered a continuous, scalable psychological characteristic (ie diabetes distress) rather than a discrete comorbid clinical condition (eg depression). 29 Decision making regarding mental health research was found to be The responses of participants confirmed the viability of social media and Internet outreach for recruitment of young adults with type 1 diabetes. 19 They also identified a number of advertisement features that might impact ad success including, graphics (eg the use of "youth ambassadors"), language (eg transparent and de-stigmatizing) and placement (eg Facebook and trusted organizations/advocacy groups

| Limitations
Our recruitment methods yielded a group of mildly distressed predominantly female volunteers with type 1 diabetes. Therefore, our findings cannot be assumed to relate to more distressed young adults, those less willing to take part or to males. Young adult males present a recruitment challenge to mental health researchers, [9][10][11][12] and there are likely to be attitudinal and experiential factors specific to young men with type 1 diabetes that were not captured in our data. Future qualitative studies may benefit from sampling young men using methods with demonstrated effectiveness in research with other groups that are difficult to involve (eg low incidence and stigmatized groups).
For example, snowball sampling is especially useful for research on sensitive topics, where some degree of trust is needed to initiate the recruitment process. 33 Another limitation relates to the unfamiliarity of most participants with mental health research. As young adults' reflections are likely to be grounded in day-to-day experience, the applicability of our findings to a more "experienced" group of participants remains to be determined.

| Implications
Based on the findings of this qualitative study, we can offer the following broad suggestions for health professionals and mental health researchers regarding young patients with type 1 diabetes: Health professionals: • Continue to develop and implement strategies aimed at improving recognition, discussion and management of the emotional needs of young adults with type 1 diabetes in routine care.
• Avoid "overpathologizing" the emotional struggles of young adults with type 1 diabetes.
Mental health researchers: • Adopt a participatory research approach that prioritizes involvement of young adults in all stages of the research process, including protocol development, recruitment messaging and activities, and dissemination of research findings. This may include providing opportunities for involvement as co-researchers and co-authorship in scientific manuscripts.
• Online and social media strategies that take advantage of a study's inclusion criteria (eg age, gender and location) and provide a "clickthrough" option for screening and enrolment are a palatable option to young adults.
• As part of a two-pronged recruitment strategy, preface and/or accompany recruitment activities targeting young adults with adjunctive campaigns aimed at educating stakeholder groups, health professionals and other potential sources of direct referral about the study.
• Carefully consider the content of recruitment messages. Wording of recruitment materials should reflect young adults' conceptualization of distress as normative in diabetes. Pilot message content and formats with young adults prior to dissemination. Messaging that includes diagnostic labels (eg depression and anxiety) and/or potentially stigmatizing wording is likely to be perceived as personally irrelevant and potentially pejorative. Inclusion of testimonials from and images of "youth ambassadors" is likely to be helpful.
• Online and print materials should clearly identify and personalize the research team by providing all relevant biographic details (including a photograph, qualifications and expertise) and contact information. If possible, provide an opportunity to meet the research team in person (eg seminar) or via video presentation (eg webinar).
• In addition to meeting the minimum requirements for standard informed consent, study materials should outline clearly the processes by which the unique needs of young adults have been considered in the research protocol, including details of youth involvement and pilot testing.
• There was limited evidence that offering incentives for participation would boost recruitment; however, young adults reported that they would be motivated by opportunities to learn about themselves and to help others. Clearly specifying the potential personal and broader impacts of the research process and study findings.
• Young adults expressed genuine interest in engaging with others with type 1 diabetes. As far as possible, provide opportunities for young adults with diabetes to connect with their peers, either directly (eg focus groups) or indirectly (eg online), throughout the research process.
Future research testing the impact of these recommendations on study recruitment is warranted.

| CONCLUSION
Effective provision of mental health support for young adults with