The aim of this study is to conduct an outcome and process evaluation of the REWARD intervention, a serious game intervention using reward-based strategies as a main method to trigger healthy snacking in Flemish adolescents.

As a high intake of energy-dense snacks is common in adolescents and has been related to overweight, the REWARD-project aims to develop and evaluate an intervention focusing on improving adolescents' snacking patterns. Since healthy nutrition promoting strategies used until now had limited effects, the REWARD-project proposes a paradigm shift: the development of an innovative framework to facilitate healthy food choices in adolescents by introducing the concept of reward sensitivity (RS) to explain food choices and using reward-based strategies in a serious game intervention.

1.1 General aim and objectives of the evaluation study

The aim of this study is to conduct an outcome and process evaluation of the REWARD intervention, a serious game using reward-based strategies as a main method to trigger healthy snacking in Flemish adolescents.

1. To assess if a serious game rewarding healthy food choices improves the healthy snack intake (i.e., measured with a healthy snack index) in 14- to 16-year old Flemish adolescents.
2. To assess if adolescents' reward sensitivity moderates the effect of the intervention on the healthy snack index of consumed snacks.

1.2 Methodology

1.2.1 Study design and setting The design consists of a pair-matched controlled pre-post design. The intervention will be conducted in three secondary schools in a specific city in Flanders (Belgium), while three matching schools from a similar city (comparable socio-economical, population density, size) will be selected as a control setting. A specific city was selected at the start of the project as the intervention city of which all secondary schools will participate in the intervention study of the REWARD project. In order to examine intervention effects, the results of the adolescents in the intervention city need to be compared to adolescents who will not play the game from schools in a similar city, which is referred to as the control city.

1.2.2 Study population and recruitment The study population will consist of 14- to 16- year-old Flemish adolescents. Most of the adolescents of this age are in the 3rd and 4th grade of the secondary school. From each school, at least 12 classes with at least 15 adolescents (i.e., the mean number of students expected per class given the mixture of general, vocation, and technical education classes) need to participate in the evaluation study, resulting in a total of 1080 adolescents. Taking into account a possible drop-out of 33% at the adolescent level, 33% oversampling will be applied. Thus, 1620 adolescents (i.e., 24 classes per school) will be recruited to participate in the evaluation study (the sample size calculation is described in 2.7). No exclusion criteria will be applied.

The initial contact with the intervention schools took place before the start of this study via an invitation letter and call for a meeting. As this is an innovative health promotion project using smartphones in a school context, the school principals of the three secondary schools were invited together with the mayor of the intervention city for an extensive presentation and discussion of the project. After this meeting, principals were further informed and asked to participate in the development phase of the intervention as well as in the outcome and process evaluation study. Moreover, the school principals and some teachers per school were also involved in the intervention development. All schools in the matching control city (i.e., 6 schools) will be contacted by means of a formal email and a telephone call, subsequently the control schools will be further informed through a visit by the researchers in order to convince at least three schools to participate.

Following acceptance of the school to participate in the study, parents will receive a letter explaining the purpose over the study and will be asked for passive consent for participation of their child in the study. This letter will brought home by the adolescent and returned by the adolescent when parents do not want their participation. In addition, active consent of the adolescent will be asked at the beginning of the study.

1.2.3 Development of the materials / theoretical basis

To date, most health promotion efforts to improve the dietary behaviors in youth were focusing on changing cognitive determinants and environmental factors in order to improve the dietary behaviors. However, interventions used until now have had only limited success in changing adolescents' eating patterns and anthropometrics. Therefore, one of the aims of the REWARD-project was to develop a new approach (i.e., a reward-based serious game intervention) to train adolescents to opt for healthy snack choices. Snacks are defined as all the food products (no beverages) consumed between the three main meals (breakfast, lunch, and dinner), food products consumed less than 30 minutes before or after the meal are considered as part of the meal.

The rationale for this innovative intervention can be described as follows. Current eating behaviors in the population are often driven by hedonic rather than homeostatic hunger. Hedonic hunger is driven by the palatability or the reward value of foods in the absence of energy need. It has been identified as a crucial determinant for disruption of homeostatic intake regulation often resulting in a long-term positive energy balance (when energy input exceeds energy expenditure). The present 'obesogenic' environment mostly includes fat- and sugar-rich food cues with a low nutritional value. This highly rewarding food environment might trigger and shape eating behavior in a more or less automatic way following the principles of instrumental and classical conditioning. Nevertheless, not all individuals are equally sensitive to the rewarding aspects of the present food environment. These individual differences in reward sensitivity (RS) refer to one's ability to experience pleasure or reward when exposed to appetitive stimuli, including palatable foods. High RS has been linked to unhealthy food preferences and eating behaviors both in adults and children. Moreover, it has been found that adolescents have a peak in reward driven eating behavior while their regulative control matures at a slower pace. Therefore, they are more vulnerable to the development of selective reward driven food preferences. While research has suggested to take into account adolescents' individual RS as a biological and psychological predisposition that guides eating behavior, little is known about the specific role of RS in food cue reactivity and learning processes to adopt healthy food preferences and healthy eating behavior. Both learning and personality theories provide valuable insights on eating behavior in their own way. However, when they propose a single determining factor, they are at risk of being overly simplistic. Contemporary learning theories stress that the role of any given factor (i.e. learning processes) can only be understood when additional factors, like personality traits, are taken into account. The REWARD-project therefore aims to integrate learning theories and personality traits, and so it steps away from the traditional main effects model by investigating how individual characteristics (i.e., RS) interact with learning processes related to eating behavior. Therefore, the goal of this project was to develop a serious game intervention with the provision of rewards and positive reinforcement as main behavioral change technique.

However, earlier research indicated that energy balance-related behaviors are the result of a joint function between conscious and unconscious processes. Dual-system models or dual-process models explain health behaviors as two interconnected mental systems, each operating according to different principles. On the one hand, a reflective system can be utilized including elaboration and cognitive efforts to build beliefs and decisions. On the other hand, behaviors can be the consequence of an impulsive/automatic system in which certain stimuli or cues are linked to certain behaviors based on earlier learned associations. Based on this previous research and theoretical evidence, it was decided to incorporate also behavioral change strategies/methods to influence the cognitive/reflective pathway in addition to the reward-strategies. The central idea in the REWARD game -clear from the start- is that participants will earn credits based on the nutrition value of the snacks they consume (see the following chapter for more details). The Intervention Mapping protocol was used to put this idea in practice and to identify additional important (cognitive) behavioral change strategies.

The intervention is developed according to the stepwise and iterative principles of the Intervention Mapping Protocol and included an intensive participatory approach with several significant stakeholders throughout all phases of the intervention development.

The Intervention Mapping protocol starts with an extensive problem analysis using existing and newly generated evidence. Preliminary studies informing the intervention development and organized in the REWARD-project were 1) a large-scale cross-sectional study, 2) a large-scale focus group research, and 3) experimental research to investigate how non-food reward schemas can compete with the rewarding character of palatable foods in adolescents. A large-scale cross-sectional study investigating dietary and game behaviors and its individual and environmental determinants among 1100 Flemish adolescents was first implemented. Subsequently, focus group research was conducted among 100 adolescents to investigate the nutrition-related results of this cross-sectional study into depth. Next, focus groups were conducted with the same pool of adolescents to brainstorm about the conceptual model of the serious game (i.e., preferred game dynamics, game use, and game motivations).

Based on extensive literature reviews as well as the results of the preliminary REWARD studies, the following correlates/determinants at the individual level were found to be associated with the intake of snacks: knowledge, attitude, self-efficacy, and taste preferences. More knowledge about the healthiness of snacks was associated with a higher intake of healthy snacks. A positive attitude towards healthy or unhealthy snacks correlated with a higher intake of the respective snacks. A lower self-efficacy to eat healthy snacks was associated with a higher intake of fat- and/or sugar-rich snacks and vice versa. Taste preference is one of the strongest predictors of snack intake, however, difficult to change so this was not included in our intervention.

The selected determinants (knowledge, attitude, and self-efficacy) are targeted via five methods, namely active learning, advance organizers, goal setting, and monitoring (the latter two both include feedback).

• Knowledge about the nutritional quality of a food item will be enhanced via the credit system used to reward the players for the healthy choice. This credit system includes a continuum ranging from 0 (i.e., the most unhealthy snack item) to 50 (i.e., the most healthy food item). In the game, we will refer to preferred and non-preferred snacks instead of healthy and unhealthy snacks, respectively. By the provision of these credits in the game and the game advancement linked to these credits, participants will learn about the location of specific snacks on the healthy-unhealthy continuum. The methods selected for changing the practical knowledge are based on the Elaboration Likelihood Model, namely active learning and advance organizers. The Elaboration Likelihood Model suggests that people only process information centrally (i.e., careful consideration which can increase knowledge) when they are motivated to do so, when information is personally relevant, and repeated. The credit system used in the game can be seen as an external motivator to change their snacking behaviour, however, insights about the (un)healthiness of snack items based on this credit system can stimulate central processing of this information as more knowledge about the specific credits per snack item and the application of this knowledge can enhance their progress in the game. Moreover, these points are also personally relevant (i.e., linked to the own snacks consumed) and are repeatedly presented. In addition, this healthy-unhealthy continuum ranging from 0-50 and the placement of the different snack items on this continuum based on the credits, can act as a mental reminder (i.e., a mental axis) regarding the nutritional value of snack items. Furthermore, similar to the application of these two methods to the nutritional quality, gratuities will be used in the serious game related to limitation of snack intake (≤ six snacks per day) and preference for the healthy snacks (≥2/3 healthy snacks per day) to learn adolescents about a well-balanced snacking pattern. This cut-off of six snacks and more than 2/3 healthy snacks was based on Flemish guidelines of necessary food and nutrient intake. It is advised to consume maximum 10% energy from snacks (defined here as unhealthy snacks). Fourteen to 16- year old girls and boys have a general advised total energy intake per day of 2100 and 2600 kcal, respectively. So this corresponds with a maximum intake of 210-250 kcal from unhealthy snacks (i.e., one or two unhealthy snacks per day depending on the size and nutritional content). Additionally, it is advised for adolescents to eat three pieces of fruit and four dairy products per day (i.e., not only via snacks but also via meals). The cut-off of maximum six snacks and more than 2/3 healthy snacks per day was deduced from these guidelines, an intake of three or four healthy snacks together with one or two unhealthy snacks is acceptable for adolescents as they need a lot of energy to grow.
• Positive attitudes regarding healthy snacks can be activated via the method of mere exposure. Earlier research indicated that participants have a more positive attitude about stimuli the more times they are exposed to them (i.e., in our intervention we expect them to increase exposure to healthy snacks as participants receive more credits for them), even if they are not consciously aware of the process. Moreover, the association of positive reinforcement and healthy snacks can also result in more positive attitudes, previous research stated that attitudes can be automatically activated: exposure to stimuli with a positive evaluation activates positive attitudes.
• Self-efficacy will be targeted via goal setting and monitoring. Goal setting has a high likelihood to improve health behaviors because persons with goals exert themselves to a greater extent, persevere in their aims, concentrate more, and develop strategies to be able to perform the necessary behavior. In addition, earlier research has indicated that self-monitoring is the most successful behavioral change technique in energy balance-related intervention research. During three weeks of the intervention, the participants need to choose one out of four -personal, feasible and measurable- snacking goals (from week 3 till week 5: every week one specific goal should be chosen and this goal should be reached daily). The chosen goal will be evaluated per day and credits/rewards will be provided in case of success. At the end of every game week, feedback will be provided via an in-game week-report that portrays all the consumed snacks per weekday with their matching credits. This week-report will also provide feedback about the fail or success regarding the goals. By providing them a weekly self-monitoring tool regarding their snack intake, they will have more insight into and awareness of their snacking behavior. This week-report can create elaboration about the snacking behavior. In case of success (i.e., linked to the goals), it will increase self-efficacy/capability of eating more healthy snacks. In case of failure, the feedback and the pending reward can stimulate elaboration and another try and a possible increase in self-efficacy after success in the future.

Additionally, the influence of peers and parents has also been indicated as important. Peer' modeling, social support, and social pressure as well as the physical and social home environment have been associated with snack intake in adolescents. However, it was not possible to target the environment through the chosen -individual- serious game channel. Nevertheless, indirect positive effects of this intervention on the peers' modeling and social support pressure can be expected given the chosen game motivations and dynamics in the serious game, such as corporation and competition (see further).

The results of the Uses and Gratifications Questionnaire measured in the REWARD cross-sectional study, indicated that challenge (i.e., to push yourself to a higher level of skill or personal accomplishment) was the most important motivation to play games for adolescents. Additionally, competition (i.e. proving who has the best skills and can react and think the fastest) and social interaction (i.e. playing together with other both online as offline) were also shown to be important motivations to engage in game play. Furthermore, the REWARD cross-sectional study examined which game dynamics or structural characteristics of a game are indicated as important aspects of a game by adolescents using the list of game dynamics. These game dynamics refer to inherent features of games which can itself facilitate initiation of game play and maintenance of game play over time. The results showed that adolescents regard the following dynamics as the most crucial and important characteristics of a game: (1) competitive aspects / playing against other people / leaderboard rankings, (2) cooperation / working together to reach goals, (3) different story outcomes based on your player actions, (4) 'Levelling up' a game character, and (5) earning points or other rewards.

Previous research also highlighted the importance of challenges, competition, and interaction as game motivations and cooperation, and rewards, as game motivations / dynamics. Previous studies indicated the importance of story lines to increase immersion and intrinsic motivation to play the game. Moreover, the inclusion of three psychological needs based on the Self-Determination Theory, namely autonomy, competence, and relatedness, have been linked to increased intrinsic motivation to play games. From this viewpoint, it is important that participants can make their own choices in the game that lead to different endings (i.e., autonomy), can customize their own avatar according to their preferences (i.e., autonomy and relatedness), and experience different difficulty levels based on their performances and have indicators for their achievements as this better satisfies players' needs of competence. The next chapter will provide the details about the application of the abovementioned methods and game motivations and dynamics in the REWARD serious game.

1.2.4 Study duration and intervention components

The study period will consist of a pretest, the four-week 'The Snack Track High School" serious game intervention, a post-test immediately after the end of the intervention. The intervention group will receive the four-week serious game intervention, no extra components will be added in the intervention schools. The control schools will continue the usual school curriculum. Smartphones (Lumia 435) will be provided to the adolescents without smartphone so all adolescents in the intervention schools can participate.

As mentioned above until now, health promotion interventions have had only limited effects. Therefore, a new innovative approach was chosen in the REWARD-project as well as a more innovative intervention channel, i.e., a mobile serious game. Several reasons could be provided why this channel has a high likelihood for success. First, playing games is a popular pastime among adolescents, earlier studies found that adolescents spend on average 18 hours a month on playing games. Second, gaming is intrinsically motivating to adolescents so a higher engagement in the intervention can be achieved. Third, video games are interactive, which can lead to more satisfaction. Finally, games allow for practicing the desired behavior in a safe environment.

1.2.5 Implementation and evaluation of the intervention

The serious game will be launched on a Monday in February 2016 by the researchers. Weekly process evaluation moments will provide inside into the participation of the adolescents in the game and will also be used as a moment to encourage participation, this will be done by the teachers or researchers depending of the preference of the schools. During the study period, it will be possible to contact the researchers by phone or email. If necessary, visits of the researchers will be possible. In addition, a help function is built in the game so participants can always contact the researchers and game developers in case of technical or other problems.

The outcome evaluation of the study includes pre-post measurements. These measurements will consist of anthropometrical measurements as well as an adolescent questionnaire measuring the primary outcome (the healthy snacking index; BMI), secondary outcomes (personal and peer' variables), and controlling variables.

1.2.6 Sample size calculation

The sample size was calculated on the healthy snacking index (i.e., the primary outcome of the intervention study). The ICC's, mean and standard deviation regarding the healthy eating index were based on the earlier cross-sectional Reward study and the test-retest of the REWARD FFQ to measure snack intake. Assuming an ICC of 0.02 at school level and an ICC of 0.03 at class level with a mean and standard deviation of the Healthy Snacking Index of 37.8±20.2, we need to involve at least 12 classes with 15 adolescents per school (3 intervention schools and 3 control schools) to obtain a power of 80% to detect a difference of 20% between intervention and control at the 5% significance level. To account for possible loss to follow-up, an oversampling of 33% is applied. A total sample of 1620 adolescents (control and intervention) will be recruited.

1.2.7 Data handling After every data collection wave, the adolescent questionnaires will be scanned using the scanning software package TELEForm (version 6.1, Cardiff Software Inc., San Marcos, California, USA). Afterwards the data will be translated into an SPSS file. The school management questionnaire, the teacher post-questionnaire, the weekly logs, the anthropometrical measures will be entered manually by the research assistants after data collection.

1.2.8 Statistics Descriptive analyses will be used at the individual level to compare the intervention and control group at baseline. Repeated measure analyses taking into account the clusters (child-class-school) will be used to evaluate differences in the primary outcome variables healthy snacking index and anthropometrics. In addition, subgroup analyses, and mediating and moderating analyses will be conducted to provide more insight into the intervention effects.

Inclusion criteria:

• Adolescents attending 3th and 4th grade of secondary school in Belgium (14- to 16- year olds)
• Adolescents understanding Dutch.