Gamification for the Improvement of Diet, Nutritional Habits, and Body Composition in Children and Adolescents: A Systematic Review and Meta-Analysis

Currently, one of the main public health problems among children and adolescents is poor adherence to healthy habits, leading to increasingly high rates of obesity and the comorbidities that accompany obesity. Early interventions are necessary, and among them, the use of gamification can be an effective method. The objective was to analyse the effect of game-based interventions (gamification) for improving nutritional habits, knowledge, and changes in body composition. A systematic review and meta-analysis were performed in CINAHL, EMBASE, LILACS, MEDLINE, SciELO, and Scopus databases, following the PRISMA recommendations. There was no restriction by year of publication or language. Only randomized controlled trials were included. Twenty-three articles were found. After the intervention, the consumption of fruit and vegetables increased, as well as the knowledge on healthy food groups. The means difference showed a higher nutritional knowledge score in the intervention group 95% CI 0.88 (0.05–1.75). No significant effect of gamification was found for body mass index z-score. Gamification could be an effective method to improve nutritional knowledge about healthier nutritional habits. Promoting the development of effective educational tools to support learning related to nutrition is necessary in order to avoid and prevent chronic diseases.


Introduction
Nowadays, the absence of a physical exercise routine and adherence to a balanced diet are two major public health problems. In 2019, 38 million children under 5 years old were overweight or obese [1]. Furthermore, more than 80% of the adolescent population in the world does not do enough physical activity [2].
The WHO emphasizes that unhealthy diets and physical inactivity are two key risk factors to develop non-communicable diseases such as, cardiovascular diseases, cancer, and diabetes [3]. In addition, the intake of fruits and vegetables in the child population is under the recommended levels and that of sugar is well above the established limits, which increases the risk of developing these diseases [4].
A systematic review and meta-analysis was carried out following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines [25].
The following databases were consulted: CINAHL, EMBASE, LILACS, MEDLINE, SciELO, and Scopus. The search was carried out in April 2021. There was no restriction by year of publication. Using the Mesh terms, the search equation was "(game OR gamification) AND (child OR adolescent) AND (nutrition OR feeding behaviour OR food OR diet OR body composition OR body mass index OR health) AND (RCT OR randomized controlled trial)".
For the selection of the study sample, the PICO (population, intervention, comparison, and outcome) strategy was used. The population were children and adolescents; the intervention was through different gamification programs (defined as organized games with a set of rules for playing and achieving goals or objectives by providing feedback and interaction); the comparison was made between the pre-and post-intervention groups, or the intervention and control groups; and the outcomes were diet and body composition improvement.

Inclusion and Exclusion Criteria
The studies with the following characteristics were included: (1) randomized controlled trials, (2) study sample comprising children and adolescents, (3) intervention as a playful game component, (4) gathering data on the effect of the intervention on eating habits, knowledge, and body composition. There was no restriction by language or by year of publication.
The exclusion criteria were (1) studies related to improvement in other health habits, (2) interventions applied exclusively to certain diseases/pathologies, (3) combination with other types of interventions such as physical activity, (4) studies without a control group.

Study Selection, Quality Appraisal, and Risk of Bias
The selection was carried out by two authors independently. First, the title and abstract were read. Then the full text was read. Finally, a critical reading of the selected studies was carried out to assess the risk of bias. A third author was consulted in case of disagreement.
The quality of the studies was assessed following the levels of evidence and grades of recommendation of the Oxford Centre for Evidence-Based Medicine (OCEBM) (Howick et al., 2011) ( Table 1). Risk of bias was analysed by pairs of independent reviewers using the Cochrane Collaboration risk of bias tool [26]. All the articles reached a quality level according to the quality assessment tools; therefore, no study was excluded. To examine an online video-game to promote fruit-vegetable consumption changes CG: no intervention IG: "Squire's Quest II" online video-game. To examine the effect of advergames that promote intake of energy-dense snacks or fruit on children CG: no intervention IG1: advergame that promoted energy-dense snacks IG1: advergame that promoted fruit IG3: non-food products -Caloric intake Children who played an advergame that promoted food (energy-dense snacks or fruit) ate significantly more than did the children who played an advergame that promoted non-food products (p < 0.01) and also ate. Sex (male) (p < 0.05), hunger (p < 0.01), and age (p < 0.05) were significantly related to energy-dense calorie intake

Data Abstraction
A descriptive analysis was performed to extract the data from each included study, consulting with a third author in case of disagreement. The variables obtained from the selected articles were (1) author, year of publication, country, (2) design, (3) sample size, (4) aim, (5) type of intervention and duration, (6) measuring instruments, and (7) main results.

Data Analysis
The meta-analysis compared the standardized means difference between the gamification group and the control group. The heterogeneity of the sample was assessed with the I 2 index, if this was greater than 50% a random-effects analysis was selected [26]. Publication bias was assessed using the funnel plot, and a sensitivity analysis was also performed. RevMan Web software was used.

Study Characteristics
The database search comprised a total of 1433 articles. The final sample was n = 23 articles. The selection process is shown in Figure 1. All included studies were clinical trials . The total sample was 11,280 children and adolescents. The oldest article dates from 2005, although most studies were published from 2010 (n = 20). Most of the studies were conducted in the USA (n = 8) and Italy (n = 4). The adherence rate to the intervention programs was high, from 96.4% [40] to 91% [42,46]. The main characteristics of all the included studies are listed in Table 1.

Effects of Gamification on Food Groups and Food Habits
Providing nutritional information through gamification interventions had a positive influence on food selection in children and adolescents. After the intervention, the consumption of certain food groups increased. Weekly intake of fruit [28,35,39,49] and vegetables [27,33,35,39,49] was augmented with an increase of about 0.67 servings per day up to 1 serving [30,32]. Intake of nutrients related to fruits and vegetables such as vitamin C, beta-carotene, potassium, and dietary fibre improved [31,46]. The consumption of whole and protein-rich foods was also increased [35], and the sugar intake decreased significantly [42]. Some authors when analysing water consumption did not find significant differences [30] although they increased motivation [39].
Knowledge about information related to food groups increased from 4.8% [29] up to 34.10% [36]. In addition, knowledge about the five major macronutrients improved after intervention, although not at follow-up [37]. It also improved self-efficacy in the adoption of healthy eating practices such as the preparation of healthy meals [31].
Regarding eating habits, after the intervention, the frequency of eating while watching television or studying as well as eating in fast food restaurants was reduced [31].

Effects of Gamification in Body Composition
Regarding changes in body composition, some authors found significantly lower changes after intervention in waist circumference and BMI z-score after intervention and follow-up [47][48][49]. Although other authors did not find significant differences after intervention [27,30].

Meta-Analysis Results and Risk of Bias
There was no publication bias, and no studies were removed after sensitivity analysis. Regarding nutritional knowledge variable, six studies provided the necessary data, with a final sample of n = 2574 subjects in the intervention group and n = 2649 in the control group. The standardized means difference, with the 95% confidence interval, was 0.88 (0.05-1.75) and displayed a statistically significant higher knowledge score in the values of the experimental group (p < 0.05). These data mean that using gamification helps to improve nutritional knowledge about healthier nutritional habits. Although, in real clinical practice, that difference was not large enough to be statistically significant. Analysing BMI z-score, only two studies had the necessary information for the meta-analysis and showed no significant effect of gamification in this outcome. The forest plot and the risk of bias of each study are shown in Figures 2 and 3.

Discussion
To our knowledge, this is the first systematic review and meta-analysis that analyses the benefits of interventions based on the improvement of nutritional habits, knowledge, and changes in body composition in children and adolescents.
Game-based interventions showed improvements in the selection of healthy food groups within this population. As indicated in other studies, the consumption of fruits and vegetables increased [50][51][52][53][54][55]. Although other studies also found positive effects in knowledge about food groups, these were not reflected in an increase in the intake of this food group [53,56].
Knowledge about the five macronutrient groups also increased, as noted by other authors, although without maintained effects [57]. Other studies after online game interventions in nutritional programs showed improvements in calcium and vitamin D intake [58]. In addition, the frequency and quantity of sugar-sweetened beverage intake were also reduced [59]; while water intake, as corroborated by other studies, did not improve [59].
The gamification of nutrition can lead to improvements in dietary behaviour among adolescents in the short term [60]. Some studies that used card games found a 23.1% reduction in the number of students who did not eat breakfast, maintaining this habit up to 3 months later [61]. Others even found an improvement in adherence to the Mediterranean diet [54].
Several studies showed that gamification was effective in teaching nutrition and weight management knowledge, as well as in the intention to follow a healthy diet [62,63]. However, in this study, there were few articles that analysed changes in body composition, although the results found showed benefits in waist circumference and BMI. Similar studies conducted in children with obesity and pre-type II diabetes after mobile application interventions found improvements in BMI, waist circumference, and percentage of body fat maintained over time [64], and even in adults they found improved weight, BMI, fat mass, waist circumference, and cholesterol [63].
The meta-analysis showed a higher nutritional knowledge score after the intervention, as pointed out by other authors [65][66][67] and that the psychoeducational multimedia games had the potential to substantially change dietary behaviour [68].
Participants preferred to select healthier foods after playing [66]. At these ages, game-based interventions could exert a very positive influence in improving health. Through gamification, improvements have been made in sexual education [13] and smoking habits [69]. In addition, these strategies have also shown benefits in teaching processes in students of different educational levels [10,70].
Gamification was a useful method for improving health habits. The gamification process gives participants the possibility to learn and face the challenge through a different and exciting process that allows them to increase the degree of commitment [8]. In addition, motivation increases thanks to obtaining prizes and rewards [71,72]. All of this improved self-efficacy to select healthy foods and adherence to healthy lifestyles [73].
According to current international standards, the child and adolescent population eat insufficient fruit and vegetables and a lot of processed foods with added sugars [74]. The healthy eating habits that are acquired in childhood are maintained in adulthood, so it is essential to promote an adequate intake. Given that more traditional health interventions have limited success, health policies could focus on the implementation of gamification projects in the school environment.

Limitations
This study has several limitations. First, although all studies use gamification as an intervention, the great variability in the duration of the intervention and minutes of play may influence the heterogeneity of the results. On the other hand, the interventions were relatively short in time. Furthermore, the sample size of the clinical trials was small and the monitoring of the effect of the intervention over time was little studied. Finally, it was not possible to perform a meta-analysis about important outcomes such as fruit/vegetable intake, due to the heterogeneity in the units of measurement and presentation of the studies' data.
Future research would be necessary to analyse the improvement in all food groups by age, as well as the changes in body parameters. Furthermore, more clinical trials would be necessary to determine the lasting effects of the enhancement over time.

Conclusions
Gamification was a positive influence on dietary behaviour and nutritional knowledge. The choice of food groups improved, highlighting an increase in the consumption of fruit and vegetables. Furthermore, the results of the meta-analysis showed an increase in the level of nutritional knowledge, but a significant effect of gamification was not found for body mass index z-score. Game-based interventions could be very helpful in promoting healthy habits. Promotion of the development of effective educational tools to support children in nutrition learning is necessary in order to avoid and prevent chronic diseases related to nutrition from childhood.