Informing mHealth and Web-Based Eating Disorder Interventions: Combining Lived Experience Perspectives With Design Thinking Approaches

Background App-based interventions designed to prevent and treat eating disorders have considerable potential to overcome known barriers to treatment seeking. Existing apps have shown efficacy in terms of symptom reduction; however, uptake and retention issues are common. To ensure that apps meet the needs and preferences of those for whom they were designed, it is critical to understand the lived experience of potential users and involve them in the process of design, development, and delivery. However, few app-based interventions are pretested on and co-designed with end users before randomized controlled trials. Objective To address the issue, this study used a highly novel design thinking approach to provide the context and a lived experience perspective of the end user, thus allowing for a deeper level of understanding. Methods In total, 7 young women (mean age 25.83, SD 5.34, range 21-33 years) who self-identified as having a history of body image issues or eating disorders were recruited. Participants were interviewed about their lived experience of body image and eating disorders and reported their needs and preferences for app-based eating disorder interventions. Traditional (thematic analysis) and novel (empathy mapping; visually depicting and empathizing with the user’s personal experience) analyses were performed, providing a lived experience perspective of eating disorders and identifying the needs and preferences of this population in relation to app-based interventions for eating disorders. Key challenges and opportunities for app-based eating disorder interventions were also identified. Results Findings highlighted the importance of understanding and identifying problematic eating disorder symptoms for the user, helpful practices for recovery that identify personal values and goals, the role of social support in facilitating hope, and aspects of usability to promote continued engagement and recovery. Conclusions Practical guidance and recommendations are described for those developing app-based eating disorder interventions. These findings have the potential to inform practices to enhance participant uptake and retention in the context of app-based interventions for this population.


Background
Eating disorders are serious, complex, and potentially fatal conditions. Up to 18.6% of women and 6.5% of men experience a diagnosed eating disorder in their lifetime, a figure that has increased over the past 18 years [1], particularly throughout the COVID-19 pandemic [2]. Subclinical or prodromal and clinically diagnosed eating disorders are associated with high levels of medical and psychological comorbidities and poor quality of life and have one of the highest mortality rates among all mental disorders [3][4][5]. Alongside considerable health costs for individuals and their families, eating disorders also substantially impede productivity, with the cost of disease burden for eating disorders each year estimated at Aus $69.7 billion (US $43.6 billion) in 2012 [6] and approximately US $64.7 billion in 2021 [7].
Early response and treatment are recommended to ensure the best outcomes for individuals [8]; however, there are system-wide and individual-level issues that prevent detection, recognition, diagnosis, and access to treatment for eating disorders [9,10]. Alongside the stigma associated with eating disorders, symptomatic individuals may not realize the severity of their symptoms or may be afraid of change, hence obstructing help seeking [9,11]. Depending on geographic location, health systems vary widely in relation to pathways, costs, and access to diagnosis and treatment [11,12]. Combined, these factors mean that as few as 25% of individuals with eating disorder seek help [13], one of the lowest treatment-seeking rates across mental disorders [14].
Self-help psychoeducational and treatment resources, delivered via the web through digital health platforms (ie, accessed via an app or computer), represent an opportunity for cost-effective access to intervention in a convenient and private manner. These approaches have demonstrated promise in the field of psychiatry (eg, depression and anxiety) [15] and specifically within eating disorders [16,17]. Studies of web-based self-help interventions for eating disorders report moderate acceptability and accessibility [18], with some evidence suggesting that they may be at least as effective as in-person interventions [19,20]. Despite the potential and realized advantages of digital health interventions, limitations exist. Individuals may be ambivalent to change [9] or express concerns regarding data privacy and quality of content within eating disorder apps [16]. For these reasons, mental health and eating disorder apps typically experience low engagement and high attrition rates [15,19].
Although these concerns are present in digital health more broadly, this study will focus specifically on mobile health. Given that app development as a whole is resource-intensive, researchers are encouraged to use user engagement practices to ensure that these investments translate to outcomes [21], particularly given the high levels of attrition in many digital health interventions. Focusing and prioritizing efforts to determine user needs and meeting them in app development will ensure that this investment of time and money will result in increased uptake and use. Although studies have started to explore eating disorder users' preferences [16], reviews have concluded that studies involving end users in the design of such apps is lacking [19] and the few examples that exist do not adhere to best practices (eg, not using succinct and clear communication to end users [22]).
Incorporating voices of lived experience through participatory research and co-design is recognized as being critical in the intervention design process [23]. Design thinking is a novel approach for creative problem-solving, which involves empathizing with future end users, defining the problem, ideating and testing early concepts, and iterating until the product meets accessibility and usability needs [23][24][25]. This study focused on empathizing with potential end users of eating disorder apps and identifying possible challenges and opportunities within this population.
Empathizing with potential future users is the first phase within all design thinking frameworks. This is typically done by interviewing individuals with lived experience of the condition or topic that is being addressed. The purpose is to better understand the nature of their experiences, implications of the problem, and their attitudes toward possible solutions. Following the interview, the researcher independently develops an empathy map for each user to visually depict the user's personal experience, including what they see, hear, say, do, think, and feel. Then, these are used to elicit deeper values, needs, and motivational drivers that may inspire more innovative solutions [23,26]. Although similar in structure to more traditional qualitative analysis, empathy maps are advantageous in that they provide empathy regarding what the individual sees, feels, and so on, thus allowing the researcher to contextualize participant data relative to the individual's lived experience, rather than coming from the perspective of the researcher. In addition, the empathy mapping approach can be used to identify both challenges faced by the population of interest on a particular issue (ie, pain points) and possible opportunities, including how to overcome these obstacles (ie, gain points). Therefore, empathy maps allow for better understanding of the context, problems, and needs of the population, while also providing opportunities for addressing the topic or issue.
Design thinking approaches, such as empathy interviews and empathy maps, may be especially helpful when considering individuals with significant body image and eating concerns, including eating disorders, who often have a fear of change or losing control and have ambivalent attitudes toward change [9]. Empathy maps may contribute to knowledge gain regarding users' wants and needs and inform the marketing and dissemination of these apps to address barriers and increase uptake, retention, engagement, and efficacy. Design thinking approaches have been used to inform technology-enabled eating disorder services [27][28][29] and in the design of a body image and eating disorder prevention program for adolescent boys [30]. However, no study has yet conducted an in-depth examination of the lived experience of individuals with eating disorders to inform app-based intervention development.

Objectives
Given the opportunities that design thinking may provide [31], this exploratory study aimed to use a novel empathy mapping process to provide insights into end users' perspectives of app-based interventions for eating disorders, including identifying key challenges and opportunities of app-based interventions among this population.

Recruitment
The study forms part of a large project (Web-Based Interventions to Reduce Eating Disorders [WIRED]), which involves the development and evaluation of an eating disorder app. Participants were recruited via posts on the WIRED project's social media accounts, highlighting the value of sharing lived experience, focus on involving end user experiences in app development, and nature and aims of the project. In some instances, these posts were also reshared by partner organizations (eg, the Butterfly Foundation). Young adults were asked to express their interest in participating in the study if they were aged 18 to 35 years, lived in Australia, and had experienced body image issues or eating disorders. Recruitment focused on young adults because (1) eating disorders are prevalent in this age group and (2) this group is likely to be the highest users of an app. Therefore, lived experience was self-identified and not reliant on past or current formal diagnoses. Individuals who have not received an eating disorder diagnosis but nonetheless experience significant symptoms are at high risk of not receiving appropriate treatment and are therefore a key group that the to-be-developed app aims to engage. In an adjacent study, young adults were invited to participate in a quantitative survey, in which they were asked to provide demographic details and respond to questions focused on their needs and preferences for an eating disorder app (results presented elsewhere; refer to study by Daugelat et al, unpublished data, May 2020). However, participants who were interviewed for this paper did not necessarily complete that questionnaire.

Participants
Young Australian women (N=7) who identified as having lived experience of an eating disorder or significant body image concerns provided informed consent to participate in this study. Given the exploratory nature of this study, the sample was not intended to be representative; therefore, a small sample size was deemed acceptable [32]. Furthermore, this sample size is aligned with the number of participants typically engaged in empathy interviews for user experience studies [33]. Of the 7 women, 1 (14%) woman chose not to provide demographic details; therefore, these are reported for the remaining participants (6/7, 86%). This 86% (6/7) of the participants was aged between 21 and 33 (mean 25.83, SD 5.34) years. Of the 6 participants, 5 (83%) lived in Melbourne, Victoria, and 1 (17%) lived in regional New South Wales. All except 1 participant had received a formal eating disorder diagnosis from a health professional (5/6, 83%) and had engaged in treatment, either previously (5/6, 83%) or currently (2/6, 33%).
This study adopted a combined methodology, which incorporated a design thinking approach with traditional qualitative interview research. The semistructured interview schedule is provided in Multimedia Appendix 1. The interviewer (female, Masters of Health Psychology student, aged 22 years, and with recent experience of a subclinical eating disorder) initially asked participants to share their experience of disordered eating and significant body image concerns and how it affected them personally. Following this, questions focused on the web-based environment and the development and use of digital resources such as apps. The aim of the interviews was two-fold: to hear the story and body image or eating disorder journey of the participant and to ask participants what they would have needed or wanted from an app-based intervention around that time.

Ethics Approval
The Victoria University Human Research Ethics Committee approved all the procedures for conducting this research project (HRE 20-010). Participants provided active consent before completing a web-based survey to collect their demographic and contact details.

Procedure
Interviews were conducted through the web via Zoom and were transcribed using web-based transcription software. Participants were provided a Aus $50 (US $30) gift card to thank them for their time and for sharing their lived experience.

Data Analysis Strategy
Empathy mapping is the first stage in the design thinking process [23]. Using the interview transcripts, an empathy map was created for each participant to empathize with the individual by viewing the world through their eyes and visually articulating what is known about them and their experience through four key questions: what did they see, what did they say and do, what did they hear, and what did they think and feel. This process provides context with which one can reflect on the deeper needs of individuals, as expressed by the individuals, rather than coming from the perspective of the researcher. Given that the timeline for app development on the project was short, empathy maps were initially drafted by the interviewer and then extended to provide more detail by the lead author (female, aged 29 years, and with no lived experience of an eating disorder). The last author (female, aged 39 years, and with lived experience of a subclinical eating disorder), who is experienced in both design thinking and eating disorder research, ensured congruence between these stages.
Next, to ensure a lived experience perspective, an empathic lens was used to interpret the interview data, and inductive thematic analysis was performed [34]. This process meant that the lead author used the empathy maps to inform the interpretation of the data, to ensure that the lived experience of the participants was front of mind during the analyses, rather than their own perspective. These analyses were focused on the data that were relevant to intervention and recovery, specifically app-based interventions. An iterative, reflexive process was used, meaning that ideas were formed and developed over time. The lead author identified the initial themes [34]. Then, these were shared with the authors and multiple members of the research team (who were in the early stages of content development for the app). Following discussion and further development, themes were finalized by the lead author.
Alongside each theme, a single pain and gain point was identified to encapsulate key considerations regarding the challenges faced (ie, pain points) and how apps may overcome these issues (ie, gain points). Although pain and gain points traditionally form part of the empathy mapping process, given that the aims of the paper were focused on app-based interventions, interpretation of the pain and gain points was focused specifically on app-based interventions.

Empathy Maps
An example of an empathy map from a participant is shown in Figure 1. As described previously, the aim of this process was for the lead author to ensure that the lived experience perspective was considered during the thematic analysis. Furthermore, given that the purpose of the study was focused on app-based interventions rather than describing the lived experience (which is presented in detail elsewhere [35,36]), an overview of the common elements across the empathy maps is provided in Multimedia Appendix 2.

Overview
Following the empathy map process and attending to this lived experience perspective, 4 themes were identified relating to insights into the needs and preferences for app-based eating disorder interventions among this population. Following each theme, a pain and gain point is described to summarize the personal challenges faced by users and opportunities to overcome these and meet their needs through app-based interventions. The findings are also presented in Table 1. Pseudonyms are used to maintain participants' anonymity.

Theme 1: Recognizing Symptoms and Attitudes Toward Accessing Help
Findings revealed that none of the participants who had been diagnosed with an eating disorder (0/7, 0%) identified with their diagnosis at the time they were experiencing symptoms. Participants did not think they were "sick enough" to warrant or deserve treatment and, on reflection, had incomplete knowledge of eating disorders and their associated symptoms. For example, media portrayals of eating disorders seemed to form the basis of their understanding of such conditions, where symptoms were perceived as more severe than their own. Furthermore, some participants appeared to normalize these behaviors, often downplaying their symptoms and severity: Other participants also commented that they did not want to take steps to address their eating problems. This stemmed either from fear of making changes and the daunting nature of beginning a recovery journey or from concerns about losing the connection to self and identity that their eating disorder had provided-making them feel special, important, and powerful, often for the first time in their lives:

Pain and Gain 1: From Not Wanting Help to Acknowledgment of a Problem
It appears that this denial of a problem or ambivalence to change may act as a significant barrier to app-based eating disorder interventions (pain 1). This suggests the importance of early intervention and positioning of the app for young people early in the emergence of disordered eating attitudes and behaviors, before symptoms become very severe or the eating disorder becomes part of the individuals' identity. The gain point that emerges from this theme is that young adults may need to be enlightened and educated about the extent of their symptoms to demonstrate that they may be in need of early intervention. For example, questionnaires may be used in marketing, screening, and tailoring of app content to highlight and reinforce the relevance and potential of such a resource to them (gain 1).

Theme 2: Effective Content That Reflects Key Processes in Recovery
Participants discussed a range of approaches that were helpful in their recovery journey and recommended that these could be incorporated into an early intervention app. Of the 7 participants, 6 (86%) reported that using reflective questioning to identify unhelpful thoughts or behaviors was critical for their recovery. This technique provided participants with deeper understanding of the connections between their thoughts and behaviors and allowed them to identify that their eating behavior was disordered. This practice also helped participants to see how these attitudes and behaviors may have been affecting their lives and how they may start to address it: In addition to identifying problematic thoughts and behaviors, participants also emphasized the need to replace these with more positive and helpful thoughts and behaviors. Providing ideas for how to think differently about the self and body was suggested as particularly valuable. It was clear that many participants (5/7, 71%) judged themselves based solely on their appearance, typically resulting in negative self-evaluations. In contrast, during and after recovery, participants had started to take a more holistic view of themselves and their bodies, which allowed them to feel more comfortable. Having a broader understanding and experience of their bodies led them to a range of helpful perspectives including aspects of positive body image, body neutrality, and enhancing self-worth: These quotes emphasize the value of focusing on alternative views of the self and one's appearance that does not focus solely on esthetics. By doing so, participants were able to identify and value aspects of themselves that were distinct from their appearance, and this provided them with opportunities to enjoy or be proud of their body, a feeling that was new to them. Other elements that participants found to be helpful in their journey of recovery were identifying other critical aspects of their identity beyond the eating disorder, incorporating a values-based approach, and focusing on the future and potential goals:

Pain and Gain 2: Taking Control of Personal Values and Goals
In this study, young adults with body image and eating concerns frequently used food or exercise as a means to exert control, regulate emotions, or manage stress and anxiety, often when they felt that a situation or aspect of their life was out of their control. The challenge for early interventionists and app developers is that, if these behaviors become regular coping mechanisms, users may fear losing the control that they desire, which may result in a subjective evaluation of change as risky and reluctance to engage in intervention content through an app (pain 2). A way to address this is to leverage personal motivations by identifying values and goals that are specific to the user. By concentrating on future potential achievements and outcomes and targeting realistic goals, users may be less frightened and more motivated to initiate their recovery journey, as they know the possible benefits that are personally meaningful to them (gain 2).

Theme 3: Social Connection to Facilitate Hope and Support
Participants discussed a desire for an app to provide a feeling of connection. This was especially pertinent as all participants (7/7, 100%) had experienced some level of loneliness and isolation in the experience of their eating disorder. A number of participants (4/7, 57%) suggested that an app can allow the user to communicate with others, which can facilitate feelings of social support. Although some participants suggested formats within an app including chats, forums, or meetups, most participant (3/4, 75%) also acknowledged that this may be unhelpful or even dangerous if used inappropriately, and thus strict moderation will be needed: Although existing support, such as professionals or family members, may provide some element of care, it was presumed that they will not understand the issues experienced by participants. Consequently, their opinions seemed less relevant or helpful to participants. In contrast, participants wanted to talk to and hear from others who had first-hand experience.
There was some indication that individuals who had been through the same experience will have an overall different and deeper level of understanding and empathy. Participants felt that they will just "get it." A number of positives of this lived experience perspective were identified by participants, including sharing useful strategies of how individuals dealt with certain situations and providing hope and encouragement for recovery and the future: It appears that this shared lived experience may be an especially powerful aspect of recovery. The idea of hearing from others who had experienced the same concerns also facilitated some level of hope for participants. Participants felt that being able to hear from these people will provide some sense of trust in recovery-that others have done it and that there is light at the end of the tunnel. Having this alternative route and point of view can keep users motivated, knowing that, although it is hard, it is both possible and worthwhile. Seeing a role model may encourage users to achieve this for themselves.

Pain and Gain 3: Moving From Loneliness to a Sense of Belonging
Although intense feelings of loneliness meant that participants desired the ability to communicate with others, regarding an app, this would be challenging to implement in such a way that it would not cause harm (pain 3). The opportunity from this is that, if app developers can provide helpful content and facilitate a feeling of connection and belonging through an app, this may help users to turn off their inner voice and move away from disordered patterns of thoughts and behaviors. Ultimately, the content, approach, and language used in an app can promote a feeling of belonging and an element of safe social connection, while providing evidence-based and accurate information. This may help users to begin to feel more motivated to engage in the content and develop a sense of hope that they can and will recover (gain 3).

Theme 4: Usability and Language Are Key
The usability of the app appeared to be very important for participants, including aspects such as content readability and app accessibility. Participants expressed a desire for an app that is intuitive and easy to use. Relatedly, the look and feel of the app are critical factors that likely affect uptake and retention over time. A participant described their experience of using a similar mental health app: Participants also had the expectation that an eating disorder app should be developed by experts and professionals to ensure its credibility, reliability, and effectiveness and that this would make them more inclined to spend time on an app. This could be achieved by the use of research, connections to universities, and having health professionals directly refer people to the app. The language used within apps also seems to play an important role here, specifically ensuring that users both trust and understand the information being presented and do not feel overwhelmed by it. Achieving this professional yet informal balance will likely encourage continued use and engagement: In comparison with face-to-face support, a key advantage that an app provides is the flexibility and freedom for the user to engage at any time. Participants described how an app should have the ability to be used fluidly to suit the needs of the user: So it's not so much of a structured application or process within the app. I think that would be really appealing for people who were worried about having to commit to something and knowing that they can always, not opt out, but they can always just come back into it. [Alex] Participants also indicated that it was important for apps to be personalized or tailored to users' needs. Participants perceived that an app of this nature will require hard work and effort. They described fear of commitment to a long program and concerns regarding the mental capability needed to work through the app content. The emphasis on flexibility suggests that an app will be used more sporadically, rather than, for example, scheduling a certain time to engage each week. The expressed preference for fluidity within an app may also indicate that some users may have difficulty in maintaining motivations or that they only feel moderately committed to working through an app. Aspects that may help to maintain motivation and retention are personalization and relevance for the user within the app and providing some features throughout the app that allow for participant input, including the ability to write ideas, comments, or responses directly into the app and check-in questions for the user: It appears that users may find continued use of an app to be challenging, including maintaining motivation over a prolonged period. Given these concerns, participants welcomed and seemed open to the idea of reminders or check-in notifications used within the app to prompt use. This may also make users more accountable for their engagement.

Pain and Gain 4: Addressing Fear of Failure Through Encouragement and Compassion
Participants perceived that an app will require significant effort, meaning that some will prefer to use it more sporadically (pain 4). From the empathy mapping process, it appears that this sample typically demonstrated some perfectionistic and competitive qualities and tended to be high achievers, meaning that they often strived to be the best. Therefore, it is likely that they fear feelings of failure when using the app and may be extremely self-critical. In contrast, participants craved and thrived on acknowledgment and praise for their efforts, which some felt that they did not receive during their illness. These types of prompts and rewards can be easily incorporated into apps to foster consistent use and combat lapses, thus ensuring progressive recovery. Given these perfectionistic traits, it may also be important to model a self-compassionate approach and remind users that recovery is neither linear nor straightforward (gain 4). If end users are reminded that setbacks are normal and encouraged to return after periods of less engagement, they may be less likely to drop out or lose interest in an app.

Principal Findings
Web-based and app-based psychological interventions offer some advantages to delivering early interventions and treatment by overcoming barriers related to cost, time, and location and concerns regarding the lack of anonymity and perceived stigma of treatment seeking. However, difficulties such as low uptake and retention remain. Using a novel design thinking approach, this study aimed to provide insights into the lived experience perspectives of potential end users regarding app-based interventions for eating disorders, including identifying key challenges and opportunities. Several issues arose from these insights. First, people with eating disorder will need to recognize that they have a problem that is sufficiently serious to warrant treatment and want to seek help for them to engage with an app. Second, specific content for the app, including processes that enable users to record their experiences, was suggested to facilitate recovery. Third, there was a strong desire for an app to facilitate social connection and sense of belonging. Finally, app usability was identified as important, including the use of compassionate and supportive language and setting a tone that is both credible and friendly.

Comparison With Previous Studies
The first issue that became clear in reaching end users was that many may not know that they have a problem or want help. Individuals with lived experience reported that they would likely not have engaged with this type of app when their eating disorder was most severe or established. As found in previous studies, participants did not believe that their problem was sufficiently serious to warrant treatment [37,38] or they feared losing control [9]. A key challenge for the implementation of an early intervention app is identifying and engaging with those in need, encouraging them to recognize that need, and informing them that they can benefit from using an eating disorder app.
To address this, when marketing and disseminating this resource, researchers should aim to increase awareness and understanding and challenge inaccurate perceptions of eating disorders through psychoeducation, including reference to the broad range of conditions and symptoms. In support of this, a recent Australian study also found that 87% of participants reported a preference for psychoeducation and screening scales to assess symptoms within app-based interventions [16]. Another critical consideration is the framing and sales pitch of an app. Researchers developing apps need to think carefully about who the app is intended for and how to engage potential users. For example, is an app intended to be used regularly as a form of treatment to replace or complement therapy, or does it aim to provide in-the-moment support that can be used sporadically? [39] Understanding this and marketing an app accordingly will provide clear expectations to the user, likely resulting in higher retention over time.
Participants identified some particularly helpful activities or perspectives that they experienced, which facilitated their recovery, and recommended that these be incorporated into app content, including verbalizing, writing, and visualizing thoughts, attitudes, and behaviors. These are common in practices such as cognitive behavioral therapy and in other mental health apps, which may theoretically translate well into an app context. A recent systematic review of evidence-based treatment elements in eating disorder apps found that cognitive restructuring was included in only 21% of apps, despite 56% of active monthly users using this function when it was included, making it one of the most frequently used features [40]. In addition, consistent with studies on positive body image, participants in this study also reported content that encouraged more holistic views of beauty and worth as being helpful [41]. Furthermore, the fear of losing control that disordered eating behaviors can cause may be overcome by focusing on the future and identifying personal goals and more meaningful achievements than those associated with weight and appearance. Other studies also suggest that combining goal setting with daily reminders is associated with great app use [42]. This appears to be a helpful strategy to disrupt mechanisms that may maintain eating disorder symptoms, thus making users more open and willing to change.
Given the loneliness and isolation reported by all participants (7/7, 100%) at some stage of their eating disorder, it is perhaps unsurprising that there was a strong desire for an app to facilitate social connection and hope. This may also be useful in helping users to overcome their critical inner voice and allow them to hear alternative views or perspectives. Although a key obstacle of app-based interventions is the lack of human contact or reciprocity [43], participants in this study provided some suggestions to combat this, including communication with other app users through a chat or forum. However, given that this may be accompanied by a range of challenges and potential dangers for users, researchers can instead focus on using innovative and creative thinking to identify alternative ways to reduce feelings of loneliness and enhance a sense of belonging through an app, without direct communication. For example, this can be achieved with inclusive and welcoming language and may include providing notifications or messages to users to indicate that other users are also using the app currently (eg, "[number] other users have been online this week/are now online"). It is possible that an e-therapeutic alliance can also be achieved in digital health interventions through user engagement and careful design choices [44]. In addition, using recovery stories from individuals with lived experience may facilitate feelings of hope for users. Future technology may also find a way to facilitate social connection, including chatbots or moderation for a chat room-type function [15], as has been done in other areas of mental health research [45]. Although these features can be incorporated into an app, they can also be offered as adjunct supports for people with eating disorder.
The final theme highlights the key role of app usability in participant uptake and engagement. Given that poor usability has been found to predict nonadherence to web-based interventions [46], this finding is an important reminder for researchers to focus considerable efforts on app accessibility, including the need for continued and rigorous testing with end users and design approaches to identify and address issues. The language used within an app itself is also important. End users may fear failure and be especially self-critical. Therefore, ensuring compassionate language that emphasizes the nonlinear journey of recovery and praise when completing activities and progressing through an app may be helpful. Studies suggest that self-compassion may be an effective intervention target, more so than alternatives such as mindfulness [47]. Relatedly, the language and tone of an app need to strike a careful balance between appearing credible (ie, created and recommended by professionals) and being personable and informal (ie, similar to talking to a friend and using language that is understandable and relatable). In addition, it appears that users want an app that can be used flexibly to fit around themselves and their lives. These findings highlight how the look and feel, functionality, features, and usability of an app are extremely important, thus indicating that the time invested in making an app appealing and intuitive during its development is critical [21].
App features that emphasize personalization were mentioned favorably by participants, including allowing users to enter their own input (eg, writings and audio messages) and track their mood. Consistent with this, qualitative studies exploring the usability of an existing eating disorder app found that participants value being able to monitor their progress and felt that it improved their self-efficacy and motivation to use the app [43]. Although some personalized elements may be easy to implement (eg, notifications, prompts, or branching of content and feedback based on early user inputs), others will be more complex and require expertise to be incorporated into an app (eg, use of chatbots to enable an ongoing, personalized dialogue with users). It is also important to emphasize that users and designers may have different interpretations of personalization. For users, this may be about feeling that the app is responsive to their inputs and not simply providing generic scripts and videos that fail to pick up on information that may make subsequent content unnecessary. In contrast, programmers may think of personalization in terms of more complex design elements, such as chatbots. Given the rapid evolution of technology, it is important for researchers to be up to date with newly emerging programming that may produce advancements for app personalization. However, it is also important to note that this high-level personalization may be beyond the scope and capacity of most academic-led projects and may not always be desired or necessary for establishing an engaging and efficacious product. This reinforces the importance of end user testing, co-design elements, and ensuring that researchers are clear in their interpretation of user feedback.

Limitations
This study used a novel, design thinking approach to provide better understanding of the needs and preferences of potential end users of app-based interventions for eating disorders. These findings can be used to inform the development, framing, and dissemination of eating disorder apps. However, the study has a number of limitations. First, although this study identified several insights into strategies to enhance uptake and retention, the retrospective nature of these data means that suggestions are informed by hindsight and may be different from what participants may have said if interviewed in the early stages of their illness or during their treatment. Second, the present sample was limited to women and may have been moderately biased in that they self-selected to participate, indicating an openness to discussing eating disorders and engaging in research. However, this study was exploratory in nature and the sample was not intended to be representative. Nonetheless, the authors encourage the use of these methodologies among large samples. Finally, even in this small sample, there were some slight inconsistencies in participants' preferences, which highlighted individual differences. For example, some participants placed more emphasis on social support or journal writing within an app than others. Therefore, further studies will be helpful to unpick some of these nuances to better understand individual preferences among users.

Implications
There are a number of important practical and clinical implications of this study. Researchers developing theoretically informed and evidence-based apps must consider the needs of the target group when developing resources. This study may be used to inform app developers or researchers who may not be familiar with the lived experience of an eating disorder and those trying to put themselves back into an early intervention mindset if they have not experienced this recently. In addition, the pain and gain points can be used to directly inform the design and development of such apps. Relatedly, it is critical that researchers explore and identify key drivers of uptake, which may inform users' willingness to change. Consideration and implementation of such practices will likely enhance app uptake, engagement, and retention, with the potential to support individuals who may not otherwise seek or access eating disorder information elsewhere. Importantly, cost evaluations that examine the impact of various design and implementation features should be performed to ensure that they are appropriate and effective.

Conclusions
This study extended previous research by using a novel design thinking approach to understand and empathize with individuals with lived experience of eating disorders to identify their needs and preferences for app-based early interventions. Although a number of challenges are experienced by this population, including feelings of loneliness and ambivalent attitudes toward change, this study has identified some opportunities that may be considered in the design and development of an eating disorder app. These include helping users to develop insights into eating disorders, identifying personal values and goals to encourage behavior change, providing elements of social connection to facilitate hope, and ensuring that the language and framing of an app is appropriate. Researchers and app developers should consider these aspects to develop future early intervention apps for women in ways that enhance app uptake and retention and improve eating disorder symptoms.