Physical activity has many physical and psychological benefits for young people living with type 1 Diabetes (T1D) [1]. In addition to the cardiovascular benefits, weight loss, and physical conditioning benefits, physical activity is associated with positive mental well-being, improved self-esteem, and quality of life [2-4]. However, the challenges experienced by young people living with T1D like fear of hypoglycemia can prevent them from being physically active [5-7]. Maintaining stable blood glucose levels (BGLs) before, during, and after exercise is affected by many factors and may not be predictable on repeated exercise occasions [8]. In addition, young people with T1D show significant decline in treatment adherence as they move through adolescence [9].

Several interventions have been developed to help young people with T1D effectively manage their physical activity [10]. However, few randomized controlled trials have explored the efficacy of these interventions [11]. Furthermore, most T1D interventions are educational, despite evidence that educational components alone, while necessary, are not sufficient in supporting adolescents when they are physically active [12]. Adding behavioral elements to an intervention, such as monitoring, goal setting, and linking medication taking with established routines, may enhance outcomes [12]. The use of mobile health (mHealth) technologies has become more common practice for diabetes self-management in middle- and high-income countries [13,14] enabling small to moderate improvements in glycemic stability, life satisfaction, and concern about diabetes and mental health [15-19].

Apps to manage diabetes have become widespread and diverse in their functionality and purpose [18]. The recent consensus report from the joint European Association for the Study of Diabetes and the American Diabetes Association Diabetes Technology Working Group highlights the potential value of digital apps for diabetes self-management [20]. An exercise advisor app for those living with TID has recently been designed [21], and there are an increasing number of apps that give guidance to people living with T1D during exercise [22].

Apps that target specific age groups are needed to help in the management of physical activity and T1D [23]. Adolescents living with T1D have specific health and developmental needs related to the many physical, psychological, and neurodevelopmental changes they are undergoing. They are not only becoming focused on peer relationships but are also developing a need for autonomy and have unique challenges adhering to T1D management regimes [24]. Therefore, apps for this age group need to be specifically tailored.

Bant (Centre for Global eHealth Innovation from University Health Network Toronto), an app designed for young people with T1D collects information on physical activity with no specific design aspect for exercise [12]. While the app T1D Exercise focuses on exercise management strategies, it is only designed for adults [22]. Previous formative research by our team provided anecdotal evidence that young people living with T1D would use an app, if suited to their needs, to help them manage their BGLs before, during, and after physical activity [25]. Until recently, there were no apps that specifically support diabetes self-management and provide individualized information around exercise in young people living with T1D. Diactive-1 App has been recently developed by a team of researchers from University of Turin, Italy. This app which is not commercially available is being tested to explore the effect of this app intervention on insulin dose requirements in children and adolescents with T1D [26].

User-centered design (UCD) of mHealth apps recognizes that all innovation starts with a deep and comprehensive understanding of the end user, such as their current behaviors, motivations, their feelings and emotions, intentions, what technology they depend on, who is important to them, and their self-image [27]. In UCD, the needs, abilities, and desires of the end user drive the design at each stage of the process [28]. Design thinking is a UCD process that encourages iterative exploration of solutions, continual refinement of the problem, and increased understanding of user needs [29]. This approach in app development may ensure the necessary relevance and robustness of the app. This is important given that with high drop-out rates of mHealth app use [30], more effort is needed to incorporate the needs, experiences, and expectations of end users in app development [31,32]. A recent integrative review questioned whether mHealth was indeed being used to address the target audiences’ needs [33].

UCD and co-design can be used together to maximize app relevance. A co-design approach to designing and developing mHealth apps allows people who are the end users of the app, providers who are the health care professionals, and researchers to contribute their collaborative insights [34]. Co-design (also referred to as participatory design) is a key research methodology that enables the perspectives and preferences of the target end user population to influence subsequent development of the mHealth app. Ideally, it involves a process of shared decision-making, which is characterized by health care professionals and end users collaborating to make decisions about the mHealth app, with a balanced focus on both hard-clinical evidence as well as the end user’s priorities and values [35]. This suggests the necessity of engaging both end users and health care professionals in co-design to design and develop the technology they will use together [36]. Although the number of available mHealth apps is increasing rapidly, there is a need to create a rigorous design thinking process for their development [29] and guidance that ensures the sustainability of the apps [37,38].

Emerging apps do provide exercise support. However, none of the currently available apps exclusively provide tailored advice on exercise for young people with T1D. Hence, the aim of the study was to apply a co-design process and design thinking principles to build a novel app for young people living with T1D that delivers individualized diabetes management advice to support physical activity. The objectives of the study were to (1) detail the application of a co-design process and design thinking principles to inform the design and development of the app and (2) identify the content and functionality that young people living with T1D need in the app to meet their physical activity goals.

Adolescents and young adults aged 13-25 years with a diagnosis of T1D for more than 6 months and competent in speaking English were eligible to participate. Multiple recruitment strategies were used to provide broad representation. Eligible participants were identified through the Western Australian Children’s Diabetes Database, a state-wide, population-based, longitudinal diabetes registry. Participants were provided with participation information sheets via email, or when approached by researchers at diabetes clinics at Perth Children’s Hospital. The studies were also advertised on local diabetes organization websites and social media. Recruitment of focus group participants continued until data saturation was attained. In the subsequent phases, recruitment concluded when all recruitment methods were exhausted.

Participants provided consent in accordance with the Child and Adolescent Health Service Human Research Ethics Committee, registered with the National Health and Medical Research Council’s Australian Health Ethics Committee. Parental consents were also obtained for participants younger than 18 years. Ethical approval of this project (RGS0000000743) from the Child and Adolescent Health Service Human Research Ethics Committee was valid from January 4, 2018, to January 4, 2021, subject to compliance with conditions of Ethics Approval for a Research Project.

To identify the challenges faced by young people living with T1D when physically active, 6 focus groups were conducted with a total of 14 adolescents and 7 young adults with T1D. Four focus groups were also conducted with 14 parents of the adolescents. Thematic analysis of focus group data suggested many challenges experienced by young people living with T1D when physically active and the results are reported elsewhere [25]. Briefly, these included, managing food and equipment, managing glucose levels with changing environment conditions, and psychosocial challenges. Parents also experienced their own challenges when their adolescents were physically active, including supporting them to be autonomous. These challenges were revisited in later phases of the design process to inform the presentation and support for the physical activity guidelines.

Clinicians and researchers were involved in this stage (n=6). Analysis of the discussion data were used to define the following:

Workshop duration was approximately 90 minutes for each. In total, 22 young adults aged 18-25 years (mean age 20.4 years) with a diagnosis of T1D participated in the first participatory design workshop, and 17 adolescents aged 13-17 years (mean age 14.5 years) with a diagnosis of T1D participated in the second workshop.

The research team reviewed the individual and group responses as well as the facilitator notes from the participatory design workshops and developed a list of desirable app features that participants indicated they needed to overcome the challenges experienced when physically active (Table 1). The team then selected the features rated the highest by the participants to incorporate into the first version of the app.

In the software development process, an exercise advisory algorithm with 240 pathways was created based on the decision tree from the published international consensus guidelines [44-46]. The algorithm expanded from a skeletal exercise activity mapping tree to separate decision trees for 4 different types of exercise: mild, moderate, mixed-intermittent intensity, and resistance exercise of up to 60 minutes in duration. Personalized insulin dose and carbohydrate advice is generated by the decision tree algorithm based on the end user’s weight, type, intensity and duration of physical activity, the duration since the last insulin bolus, and the participant’s glucose levels at the start of activity. In addition, more information on hypoglycemia treatment, pre- and postexercise insulin and carbohydrate advice, and an educational food guide which highlights the importance of low and high–glycemic index foods in the context of exercise was developed to be incorporated into the app.

This paper describes the co-design approach and the practical application of the 5-stage design framework in developing an app to support young people living with T1D to be physically active. The process enabled the development of a unique exercise app that provides individualized insulin and carbohydrate advice and builds a record of “what works” for various activities, potentially facilitating a greater level of end user confidence when physically active. It is intended that just prior to exercise, youth living with T1D would answer questions about the exercise they are about to complete, and their diabetes management at that point in time, and would be provided with personalized insulin dose and carbohydrate advice for exercise. As the relationship between exercise and T1D is very complex, it is expected that end users would engage with the app for each exercise session, especially when commencing a new form of exercise.

Young people with T1D identified desirable features of the app and provided suggestions to improve the individual advice given, including linking it to their own continuous glucose monitor (CGM) and Fitbit, and the recording of responses to advice to build a better understanding of the individualized way their body responds to physical activity. They also wanted the app to count carbohydrates and suggest foods to eat before, during, and after exercise and to provide a search function so users can ask a range of questions. Additionally, they identified the desire to personalize the appearance of the app where possible and to be able to listen to music while using the app.

This project and the acT1ve app are unique for 4 reasons: the app focuses solely on exercise management for young people with T1D, a “state of the art” design thinking process was used in the development of the app, individualized advice is provided to end users and a pilot testing phase was included. In a recent study of m-health apps for people with a chronic condition only a small number included a pilot testing phase [31]. There are apps on the market that involve young people at each stage of the participatory design process [47] such as the Bant app for T1D [12]. However, unlike our project, the effects of involvement on acceptability were not examined for this app.

The failure of health information technology interventions in the past has been largely due to poor design that did not meet the requirements of the users of the technology [28]. In this co-design project end-user’s needs drove the app development, including their input at each stage of the process. To strengthen the robustness of the app, each stage of this project was informed by the previous stage and design iterations were undertaken to refine the app to meet the needs of young people with T1D. The continual synthesis of end user’s feedback with health professional’s advice has enabled the development of a potentially usable and useful app, with high end user satisfaction, as confirmed in the pilot trial [42].

The importance of the participatory design process is highly relevant for the adolescent T1D population. The blood glucose response to physical activity can be highly individualized for young people living with T1D [48], hence the need to involve a group of adolescents and young adults with T1D of different ages and gender in the process of developing the app. Involving adolescents with T1D in a participatory design process enhances their empowerment and addresses their need for autonomy. Feelings of empowerment and autonomy are associated with improved psychological outcomes for young people with T1D [49]. Novel approaches as used in the initial workshops, that is, a “scenario” approach—are also suited to young people who particularly like real-world and relevant activities [50].

Several cofacilitators were needed to conduct the participatory design workshops which contributed to different levels of participant engagement. At times, the participatory design workshops were dominated by more vocal participants, often the older participants, and it is possible that less vocal participants were not able to fully express their viewpoints. This is a well-documented limitation of focus groups that can be offset by experienced facilitators [51]. There was also, in one of the participatory design workshops, a discrepancy in the power relations between adults and young people, with a considerable number of adults present. Adult numbers need to be minimal in participatory design processes with young people to encourage a sense of equality among the participants [52].

Designing an application to display complicated information is challenging. The additional barrier of presenting digestible information for the younger audience on a mobile format required omissions to what could be displayed, such as necessary language to be used being unable to fit the character limits of mobile devices. Additional challenges involved in using mobile devices included limiting the number of devices to be tested on and compatibility issues with operating software updates. Due to limited budget and certain logistic issues, certain features which could increase user engagement (ie, nice-to-haves) like integration with real-time CGM, activity tracker, and data sharing options were deprioritized for future modifications.

This paper describes a co-design approach using design thinking processes undertaken to develop an mHealth app to support young people with T1D when they are physically active. The design thinking processes allowed young people with T1D to identify app features that would support them to be physically active, and particularly that enabled the delivery of individualized advice. In addition, the process of app development has been described in detail to help provide guidance to others embarking on a similar project. Use of the design thinking process has the potential to create highly usable mHealth apps that can improve health behaviors in various populations of young people, especially those who need to self-manage chronic conditions such as diabetes while increasing the relevance of the content and empowering those involved in the process.