Only 8% of Canadian youth meet current physical activity (PA) recommendations [1] and they spend up to 9 hours sitting daily, often in screen-time pursuits [1,2]. Youth who do not meet PA recommendations are unlikely to benefit from the positive effects of PA on health [3,4], and sedentary youth may experience a wide range of negative physical and mental health issues [5]. Effective programs and public health policy are needed to address these issues because co-occurrence of low PA and high sedentary time contributes to childhood obesity and its numerous, deleterious health sequelae [6-9]. Despite possible negative impacts on health, screens are an indisputable reality in today’s increasingly high-tech world. Underscoring its popularity, 83% of US youth aged 8 to 18 years have a traditional video game console at home, and many of these consoles having add-on capacity for exergaming (ie, video games that are also a form of exercise) equipment. In addition, approximately 50% of adolescents own smartphones, which can download apps for mobile exergaming [10]. Many smartphone apps are free or low-cost, making exergaming and augmented reality (ie, a real-world view of the environment which is “augmented” by computer-generated input) highly accessible to youth.

In 2010, nearly 40% of US students in grades 9 to 12 exergamed at least 1 day per week [10], 25% of Canadian youth aged 15 to 18 years reported exergaming in the past week [11], and up to 20% of young adults exergamed 1 to 3 times per month or more [12]. In addition, in 2014, a representative sample of 20,122 Canadian youth in grades 6 to 10 reported exergaming for an average of 30 min per day [13]. The Xbox 360’s Kinect unit (ie, a motion-sensing device used to capture bodily movement and promote PA) sold approximately 25 million units in 2013 [14], and Pokémon Go, released on July 6, 2016 by Nintendo, had 25 million active players daily after only 1 month of the official release [15]. Leveraging benefit from screen fervor by incorporating PA into video gaming and designing exergames [16-19] that are easily available and appealing holds promise in terms of increasing PA and reducing sedentary time [16-19]. Exergaming may be a healthier pastime than traditional video games because they elicit more energy expenditure (EE) [20-22], and they may improve physical fitness, body composition, and cognitive health [20,23-25]. The effect of exergaming on reducing sedentary behavior has also been studied, but to a much lesser extent [26]. Overall, the data suggest that exergaming may be a viable method to increase PA, change body composition, and improve mental health, and it represents an improvement in EE over sedentary behavior [16,18,19,27].

There are important methodological challenges in the current literature evaluating exergaming including that many intervention studies report high attrition. Short intervention periods (ie, usually 6 or 12 weeks with no follow-up post intervention) may hinder detection of longer-term effects [24,28,29], and most studies are performed under controlled conditions that may not apply to “real-life” settings. For example, a small pilot study that provided African-American and Hispanic children with access to exergaming at home and during an after-school program indicated that fitness improved after 12 weeks [30]. However, 42% of participants lost interest in exergaming within 3 months, indicative that use related partly to the novelty of the consoles. This suggests that improving understanding of PA motivation pertaining to exergaming and sustained exergaming is needed.

To inform the development of exergaming interventions that have the potential to be sustained, evidence is needed on the determinants of sustained exergaming in naturalistic settings. Although the study by O’Loughlin et al [12]. reported that exergamers were more likely to be girls, to play nonactive video games, to watch TV 2 hours per day or more, to report weight-related stress, and to be nonsmokers, no reports to date describe factors associated with sustained exergaming in “real-world” contexts. The objectives of this study were (1) to describe the sustainability of exergaming over 2 years in a population-based sample of adolescents and (2) to identify factors associated with sustained exergaming.

Data were drawn from AdoQuest I (2005-11), a 6-year longitudinal study of grade 5 students (n=1843 aged 10-11 years at baseline), which investigated the natural course of the co-occurrence of health-compromising behaviors such as smoking, PA, sedentary behavior, and substance use in children [11]. A random sample of French-language elementary schools with more than 90 students in grade 5 was recruited in the greater Montreal area. To balance representation of students in high, middle, and low socioeconomic status, all schools were first stratified into groupings defined by tertile cut-offs of a school deprivation indicator based on maternal education, parental employment, and a measure of low family income that accounts for family size and area of residence [30]. An equal number of schools was then selected into each grouping, and 10 schools in the first, 10 in the second, and 9 in the third grouping (72.5% (29/40) of schools invited) agreed to participate. All students in all grade 5 classes in the 29 participating schools were eligible for recruitment. Of 2946 grade 5 students, 61.13% (1801/2946) participated at baseline (42 students joined after baseline data were collected). No data were collected from students who did not participate. Characteristics of the AdoQuest sample were comparable to those of 2 provincially representative samples of similarly aged Québec youth [31]. Parents completed mailed self-report questionnaires in 2006-07 and again in 2008-09. Participants provided assent, and their parents or guardians provided informed consent. The study received approval from the ethics and protection review boards of Concordia University and the Centre de Recherche du Centre Hospitalier de l’Université de Montréal.

Sustainability of exergaming over 2 years was investigated using data collected in the fall and winter of 2008-09 (Time 1 [T1]) when most participants were in grades 8 and 9 (aged 12-14 years) and 2 years later in the fall and winter of 2010-11 (Time 2 [T2]) when most participants were in grades 10 and 11 (aged 14-19 years). Data on exergaming and potential correlates and predictors were collected in mailed self-report questionnaires at both T1 and T2. Most questions used in AdoQuest were drawn from ongoing surveys and studies of youth including the Canadian Youth Smoking Survey [32] and the Nicotine Dependence in Teens study [33].

Descriptive statistics were used to compare participants who did and did not sustain exergaming. The analysis considered investigation of each potential correlate or predictor as an independent study that addressed a specific hypothesis so that only 3 statistical tests (ie, univariate, partially adjusted model, and fully adjusted model) were performed for each potential correlate or predictor [34]. Potential confounders were retained in the model if they were not on the causal pathway and if they were correlated with the potential correlate or predictor and outcome at r>0.20 [35,37]. A total of 3 logistic regression models including (1) a univariate model examining the unadjusted association between the potential correlate/predictor and outcome with no covariates; (2) a partially adjusted model accounting for age and sex; and (3) a fully adjusted model including age, sex, and (other) potential confounders. Data were analyzed using SPSS version 20.0 (released 2011, SPSS Statistics for Windows; IBM Corp). All statistical tests were 2-sided, with the significance level set at 0.05.

The purpose of this study was to describe sustained exergaming in a population-based sample of adolescents and to identify factors associated with sustained exergaming. To our knowledge, this is the first investigation of sustained exergaming in a population-based sample. Results indicated that 44% of exergamers sustained exergaming for at least 2 years. Female sex and having higher introjected PA behavior regulation were associated with sustained exergaming.

In a cross-sectional study of 200 children who owned consoles, Simons et al [38] found that 11% never exergamed and only 32% exergamed regularly (ie, >1 hour per week). Because of concerns that the novelty of exergaming dissipates over time, there have been numerous calls to investigate the sustainability of exergaming [17,24,28,39-42]. The 44% (81/186) of sustained exergamers in our study compares with the 41% of girls (but not the 69% of boys) who remained involved in team sports for over 5 years, as observed by Belanger et al [41]. Specifically, of 1276 adolescents age 12-13 years who completed a 7-day PA recall every 3 months for 5 years, the authors reported that between 14% and 53% re-engaged in a specific PA after discontinuation. Although not investigated in the Belanger study, exergaming could be an intermittent activity in adolescents linked to the release of new games and consoles. Future studies will need more frequent follow-up to assess the stop-start aspect of exergaming over time.

Three-quarters (77% [143/186]) of participants exergamed at light or moderate intensity. Exergaming at these intensities may be more enjoyable, practical, and achievable among young persons. Thus, the flexibility of exergaming in level of intensity may contribute to sustainability.

Previous studies have reported that females are more likely to exergame than males [12], and in this analysis, girls were more likely to sustain exergaming. The reasons for sex differences in sustained exergaming may mirror the reasons for sex difference in exergaming. Girls, especially those with body image challenges, may be more comfortable being active at home away from the scrutiny of others, while still enjoying the social interaction provided by exergaming. Boys may be attracted by the novelty of exergaming but return to nonactive video games as the novelty wears off. Although sex differences in exergaming should be further investigated, our results suggest that exergaming may be a viable option to help girls in particular remain physically active.

A specific focus of this work was to assess whether PA motivation is associated with sustained exergaming. Self-determination theory (SDT) has been used as a framework to predict intentions to engage in and sustain traditional PA, and investigators have reported that those with higher intrinsic regulation (ie, motivation) report increased intentions to engage in PA than those with external PA regulation (eg, [43-46]. Introjected regulation (ie, a type of PA motivation indicative of internalization of PA as a behavior) was a correlate of sustained exergaming. Introjected regulation, as described in the SDT [47-49], is motivation from an internal pressure that usually drives short-term behavior change, but does not foster sustainable behavior change [48,50]. It is generally negatively associated with or unrelated to PA levels, although there appears to be sex differences such that girls who report higher introjected regulation also report more PA [49]. The link between behavior regulation and exergaming has been studied in clinical settings and in specific populations such as overweight youth [51,52], but few studies investigate this association in population-based samples of youth using the SDT [53], and it is not fully understood if exercise behavior regulation differs between those who do and do not sustain exergaming. Sustained exergamers may not be highly motivated intrinsically toward traditional PA and are drawn toward exergaming as a PA alternative. They may be interested in developing skills for fitness and weight change, which is fostered through fitness exergames (ie, body weight, body alignment, placement, feeling, and speed captured by motion and sensor-captures and displayed in real time on the screen) [52], or aspects of the advertising or marketing of fitness exergames may play on or increase guilt, particularly among female users. Future research should focus on better understanding the differences between motivation for PA and for exergaming, which in turn lead to exergaming interventions that have a greater impact on PA and sedentary behavior.

No predictors of sustained exergaming were identified in this study. Rather than being a planned behavior, exergaming may be triggered by events in the immediate present such as purchasing a new console or friends coming over, and therefore, has few predictors. It is also possible that the sample size was too small to detect factors associated with sustained exergaming because exergaming is a relatively new area of research, not all relevant predictors may not have been investigated.

Limitations of this study include the small sample size, that self-report data are subject to misclassification, that loss to follow-up may have resulted in selection bias, and that restriction of the sample to francophones may have limited external generalizability. As data were not collected more frequently, it was not possible to confirm whether participants exergamed between data collection cycles.

Exergaming may represent a novel approach to help adolescents remain physically active during a period of life which is notable for sharp declines in PA. Exergaming may be more sustainable if games include components that foster intrinsic PA motivation such as providing more choices in the games offered (ie, with whom the game can be played, whether the setting is collaborative or competitive, how intensely the game is played) or having a coach and/or other social support while learning a new game. Finally, clinicians and practitioners can counsel parents to encourage their children to choose exergaming over more traditional sedentary video games or to exergame as a family.