The health benefits of performing sufficient physical activity (PA) are well established and include reduced risk of noncommunicable diseases, reduced risk of premature mortality, and better mental health [1-3]. There is a dose-response relationship between PA and health with a 20% to 30% reduction in chronic illness and premature death for those that meet PA guidelines [1-3]. Adolescence (13 to 17 years) is a key time to intervene, as long-term PA likely confers maximum protective benefit [3]. Those who have high levels of PA in adolescence are more likely to be active in adulthood and lead healthier lifestyles [4,5]. There are psychological and social benefits of increased PA participation in adolescence [6]. However, less than 15% of boys and 10% of girls are achieving the UK government recommendation for adolescents of at least 60 min of moderate-to-vigorous physical activity (MVPA) per day [7]. Levels of adolescent PA are similarly low in other developed countries [8]. Without intervention, activity levels decline by around 7% per year throughout adolescence, particularly in girls [9]. Strategies to increase PA in this group are urgently required.

It remains unclear what works best to change adolescent PA behavior [10,11]. A recent review of digital PA interventions for adolescents recommended that education, goal setting and feedback, self-monitoring, and parental support should be incorporated [12].

Despite recommendation that conducting formative work with target users is important to intervention development [13,14], very few PA interventions have involved adolescents in development. Before developing a school-based intervention, Corder et al conducted focus groups with participants aged between 16 and 18 years and identified choice, novelty, mentorship, rewards, competition, and flexibility as key aspects that young people would like to be included [15]. Although co-design or participatory design (PD) of digital health interventions is not necessarily recommended (as no higher effectiveness was found for games developed with PD), user input is beneficial [16]. This is particularly applicable when considering digital PA interventions [16].

Digital interventions are likely to be particularly appealing for adolescents; more than 90% play video games for at least an hour per day [17]. A variety of exergames are available and many have been commercial successes: Wii Fit sold over 22 million copies worldwide in its first 4 years [18]. More recently, an augmented reality exergame run on smartphones, Pokémon Go, has seen over 800 million downloads [19]. A study conducted in Hong Kong (participants N=210; aged 16 to 64 years) found that the use of Pokémon Go was associated with a short-term increase in the players’ daily walking and running distances, particularly in those who were less physically active [20].

Levels of PA and physiological response when exergaming are comparable with field-based PA, and significantly increased when compared with standard gaming [21]. Exergames also enhanced enjoyment, self-efficacy, and motivation for PA [21]. Exergaming interventions can lead to weight loss in overweight adolescents [22,23]. However, past research has generally involved small studies and earlier generation exergames.

Virtual reality (VR) has the potential to enhance the exergaming experience through immersion and presence contributing to the feeling of absorption, flow, and fun [23,24]. New generation VR technologies deliver increasingly realistic experiences at decreasing cost. Some small laboratory-based studies in adults have found that immersive VR exergames resulted in the same or higher intensity of exercise as standard exercise conditions, but with higher ratings of enjoyment and interest [25,26]. One study also found that perceived exertion was lower and self-efficacy was higher during VR cycling compared with standard stationary cycling [27]. An exploratory pre-post study including 9 children and adolescents suggested that an immersive VR game enhanced motivation to be active [28]. Immersion is likely to distract participants from exertion and possibly negative thinking during PA.

We hypothesized that a VR exergaming intervention could increase PA in adolescents. The aim of this study was to interview adolescents about their recommendations for a PA intervention, use of technology, gaming, and interest in VR for PA.

Participants were recruited from 2 secondary schools in London, United Kingdom between January and July 2017. To be eligible, participants had to be aged between 13 and 17 years. Given the need to recruit a high number of girls, one was a girl’s school whereas the other was mixed. Both schools were in areas of high ethnic and socioeconomic diversity. Each school sent information packs home with students from 2 classes. In total, approximately 65 packs were sent to the interested students. Interested students each returned and completed the parental consent and child assent form. We aimed to interview approximately 30 young people based on guidance provided by Fugard and Potts [29] and we interviewed all students who returned the completed consent forms.

Ethical approval was provided by the University College London Ethics committee (Project ID 10213/001) with all participants and their parents/caregivers providing informed written consent.

We used a semistructured interview. We asked the following: “What is the amount of time you spend being active as well as inactive per week in hours?”; “What is your average sitting time per day?”; “Do you know about the recommended guidelines for PA in your age group?”; What might encourage people of your age to become more physically active?”; “What is your current technology use- what do you use and how long per day or week?”; “What interest you about a particular game?”; “Do you know much about VR?”; “What are your experiences and/or opinion on using VR?”; “Do you use any other health-related technology?”; and “How would you help increase PA through use of technology and VR?”. Finally, participants were asked about key features desired in a VR exergame. Interviews were conducted face-to-face in schools between June and July 2017 by 2 researchers (EY and LV) and transcribed verbatim.

Framework Analysis was used, which is a flexible and systematic approach for analysis of semistructured interviews [30]. A total of 2 researchers (EY and LV) independently analyzed 3 transcripts, each developing an initial set of codes. The participant’s responses within the interviews were assigned codes and coded into a framework matrix, along with relevant quotes to make it visually straightforward and easy to track. Then 2 additional researchers (NF and AF) independently interpreted the data to identify common themes among responses. Theme prevalence was not determined quantitatively but instead grouped together into main themes described below. All researchers then met and discussed their interpretations and codes, compared them, and made minor adjustments to create the final framework. The final interpretations are shared below, with illustrative quotes (followed by participant’s gender and age).

A total of 31 participants aged between 13 and 15 years were interviewed. Of them, 58% (18/31) were girls. Participants identified their ethnicity as white British (n=7), white other (n=5), multiple ethnicities (n=5), Asian (n=5), Indian (n=2), Bangladeshi (n=2), Pakistani (n=1), Caribbean (n=1), African (n=1), and black (n=2). Interviews lasted for 40 to 70 min (mean 55 min). The themes were apparent and straightforward with no disagreements between the researchers. The main themes and their subthemes are described in Table 1.

Our findings suggest strong support for the potential of VR to promote PA in adolescents. However, there were a number of factors relevant to researchers developing any digital PA intervention discussed below.

Participants raised the importance of parental support, in line with previous research on digital health interventions for adolescents and studies exploring determinants of adolescent PA [12]. We are separately interviewing parents and teachers about their perceptions of VR exergaming and PA.

Awareness of any government guidelines around PA was low but desire for knowledge was high. Our participants (particularly girls) felt that an educational component would be desirable, and presenting information about benefits of PA in a visual format was recommended. This is in line with our previous work suggesting less than 20% of parents knew the government PA guidelines for their children [31]. A review of 17 adolescent PA trials found that education alone did not result in behavior change. However, multicomponent studies incorporating education found strong effects on PA [10]. Multicomponent studies can be labor intensive (eg, requiring alteration of environmental infrastructure), so their potential for wide-scale implementation is questionable. The present digital revolution and the ubiquity and frequency of technology use by adolescents, highlighted in our study, has greatly increased potential reach.

Rewards were suggested as being important to encourage PA engagement (particularly in boys, but also for some girls), in line with previous studies [15,32]. The rewards suggested were always material (usually financial). Financial rewards are effective in motivating PA in adults [33], and a trial using incentives and gamification to promote PA in families had positive effects [34]. In line with theories such as the self-determination theory, fun and enjoyment are intrinsically motivating and key motivators for gaming [35], and most adolescents engage in gaming without a material reward. Fun was consistently reported as a reason for engaging in exergaming by our participants.

Our data suggest that a home-based PA digital intervention is appealing to our target population. Most previous interventions targeting adolescent PA have been at least partially delivered in schools, but a more recent cohort found 70% to 81% boys and girls said they would choose to be active at home or in a gym/leisure center, as opposed to in school or outdoors [32]. Particularly for the girls in our study, home-based PA had the potential to overcome cultural and social barriers. In addition, they were viewed as appealing because they harnessed behaviors (such as gaming) that were already being performed. Many suggested including additional intervention elements such as community-based meet ups or groups to enhance the social elements and foster competition. In the aforementioned quantitative survey, adolescents also reported they would choose to be active with their friends (over, eg, family) [32]. However, some girls in our study felt the ability to exercise privately in their own home was appealing, so an optional social element would be best.

At the time of writing this paper, the cost of the high-end VR equipment required to create a fully immersive exergaming experience was prohibitive (around £2000 for the equipment and necessary computer), and this was recognized as a barrier by our participants. However, participants expect costs to fall. This is likely and ownership will increase [36]. We are aware that the novelty, reported as important in our study and others [15], will therefore also diminish. This strengthens the argument for developing a digital intervention within a theory-based framework, so the active ingredients can be replicated using other platforms in future. This also emphasizes the importance of making digital interventions intrinsically fun and enjoyable so that they are not reliant only on novelty. Indeed, in our study the participants who were most positive were those who had tried VR.

Our study had specific and narrow aims and around 60 min worth of discussion per participant. Interviews were conducted by researchers experienced in working with adolescents. Therefore, we believed our sample size held sufficient information power to address the research questions [37].

Our participants desired continual updates and additional challenge levels, and these are key elements of gaming that can map to specific behavior change techniques in a digital health intervention [38]. Embedding the digital platform at the center of a multicomponent intervention, and working with professional developers who understand how to keep a game engaging and challenging, is therefore crucial for success.

The results were obtained from a reasonable number of participants in comparison with other studies using thematic analysis [29], with a diverse range of ethnicities and more girls than boys. It is possible that those who were more interested in health agreed to participate. However, as less than 10% to 15% of the UK adolescent population meet PA recommendations [7], it is very unlikely that we only recruited active participants. The focus on technology as well as PA was also likely to attract a mix of interests. All participants were aged between 13 and 15 years, and results may have been different if opinions were from older adolescents. It is likely that our intervention will target younger adolescents given the value in intervening early (VR gaming is not recommended in those younger than 13 years). We will also continue to work with our public engagement groups and adolescent steering committee to gather information from adolescents from diverse backgrounds.

The results of this study suggest that an intervention to promote PA in adolescents that has VR exergaming at the core is promising. However, it is likely that additional elements will be required to produce sustained behavior change including educational elements, tangible rewards, prompts to encourage breaks, and parental support.