Effectiveness of gamified digital interventions in mental health prevention and health promotion among adults: a scoping review

Background Though still a young field of research, gamified digital interventions have demonstrated potential in exerting a favourable impact on health and overall well-being. With the increasing use of the internet and digital devices, the integration of game elements presents novel opportunities for preventing mental disorders and enhancing mental health. Hence, this review aims to assess the effectiveness of gamified interventions focusing on preventing mental disorders or promoting mental health among adults. Methods Based on a scoping review across four databases (MEDLINE, Embase, PsycInfo and Web of Science), 7,953 studies were initially identified. After removing duplicates and screening titles, abstracts and full texts, 16 studies were identified as suitable for inclusion in a narrative synthesis of findings. We included interventional studies encompassing an intervention and a control group aiming to investigate the effectiveness of the use of gamified digital mental health interventions and the use of gamified digital elements. Results Overall, positive effects of gamified interventions on mental health-related outcomes were identified. In particular, beneficial consequences for psychological well-being and depressive symptoms were observed in all studies. However, further outcomes, such as resilience, anxiety, stress or satisfaction with life, showed heterogenous findings. Most game elements used were reward, sensation and progress, whilst the quantity of elements was not consistent and, therefore, no substantiated conclusion regarding the (optimal) quantity or composition of game elements can be drawn. Further, the outcomes, measurements and analyses differed greatly between the 16 included studies making comparisons difficult. Conclusion In summary, this review demonstrates the potential of integrating digital game elements on mental health and well-being with still a great gap of research. A taxonomy is needed to adequately address relevant game elements in the field of mental health promotion and prevention of mental disorders. Therefore, future studies should explicitly focus on the mechanisms of effect and apply rigorous study designs. Supplementary Information The online version contains supplementary material available at 10.1186/s12889-023-17517-3.


Background
The mental health of individuals is affected by a variety of determinants on different levels such as individual or social factors, economics and culture, as well as living and working conditions, environmental and biological factors [1].In 2016, about 16% of the global population was affected by mental or addictive disorders [2].The most common mental disorders are depression (prevalence per 100,000: 3,627) and anxiety disorders (prevalence per 100,000: 3,715) [2].In 1990, the global prevalence of depression and anxiety disorders was about 12.7%, proving them to be the most common mental disorders for at least 30 years [3].Moreover, within the COVID-19 pandemic the prevalence has considerably increased in most countries [4].Nochaiwong et al. [5] have also indicated the impact of the COVID-19 pandemic on the global prevalence of mental health problems among the general population.
For counteracting these challenges, digital technologies may be supportive.At the moment, the whole health sector is undergoing a monumental shift: Digital technologies are shaping the present and the future of health care.Additionally, the interest in online health information continues to grow.Along with the growing use and relevance of the Internet worldwide (2005: 1 billion; 2022: 5.3 billion) [6], the increasing relevance of digital applications and media results in new possibilities and potentials for health promotion and prevention, especially in terms of mobile and web-based applications [7].Moreover, the digital transformation reveals changes on a social, organisational and individual level [8].For targeting changes on those levels, innovative digital methods and concepts such as gamification have been utilized increasingly.Gamification is defined as the utilization of game-design techniques and elements outside of a game-context to positively impact user behaviour [9,10].The game elements cannot always be clearly separated, but largely refer to typical characteristics of a game [9].In contrast to serious games -which we do no focus upon in this contribution -, it is not about the complete game, but about playful elements.Positive effects on health and well-being through gamification have been observed [11][12][13][14][15]. Thus, e.g., Johnson et al. [11] identified within their systematic review that the majority of studies found positive effects on health and well-being through gamified elements.In addition, Bostock and colleagues [15] found within their randomized controlled trial a significant positive association with gamification and well-being, likewise with stress.Further, in an multi-centre interventional study, gamification has been proven to be an effective strategy for prevention of diseases and helps reducing expenses in prevention [16].Thus, gamified interventions can induce behaviour change by improving self-determination and self-management skills [17].In addition, continuous use of such applications increases also the satisfaction and self-esteem [18].Overall, gamification seems to be an effective strategy to promote health.Previous reviews already investigated gamified interventions and the effect on mental health [11][12][13][14].However, current literature has not comprehensively focused on prevention and health promotion [11][12][13].Previous research such as by Six et al. [13] examined the effectiveness of gamification in mental health apps to reduce depression symptoms of adults, regardless of whether they are sick or not.Cheng et al. [14] in turn analysed which game elements and mental health and well-being domains are most commonly utilized and targeted in interventions of gamification for mental health and well-being.In this respect, further research is required in terms of whether and how gamified digital interventions can promote mental health and prevent mental disorders among adults.
In this context, we consider two research questions to be relevant: How effective is the use of gamified interventions, measured by relevant indicators for improving mental health or preventing mental diseases, for working-age adults?And secondly: Which game elements are most commonly used within the interventions identified?

Methods
Originally planned as a systematic literature review, we performed a screening in four major databases, namely MEDLINE (via PubMed), Embase, PsycInfo and Web of Science.The primary objective was to identify intervention studies featuring at least one control condition that have been published between 2010 and 2022.The search was executed in January 2023.To ensure consistency and comparability, we used the following complete search algorithm, based on similar reviews [11,14]: • "mental health" and "well-being" were included as terms with positive connotations since well-being was seen as a mental health-related outcome in this study; • "mental disorders" and "mental illness" as terms with negative connotations; • and "depression" as well as "anxiety" as the most common disorders in this context.These two indications were chosen due to their high prevalence and importance in the field of mental disorders as already described in the background section, but also other indications were included if identified by the search algorithm.
Thus, the following search strategy was utilised: The database search yielded a total of 7,953 records, of which 3,024 duplicates were removed.The study selection process started with the screening of titles and abstracts.Two authors (LA and PAS) independently carried out the initial title and abstract screening, which led to an interrater-agreement of 96.4%.In case of inconsistency a third party (FF or KW) screened those abstracts.The screening of titles and abstracts led to the exclusion of 4,871 studies.The screening of references of systematic reviews identified within the database search and studies included in the full text screening has not led to further hits.Subsequently, two authors (LA and PAS) independently appraised the full-texts against the inclusion and exclusion criteria (Table 1) and any discrepancies (n = 8) were resolved by consensus.Finally, the full text screening for the remaining 58 records resulted in 16 included studies (see Fig. 1).
Results are presented in accordance with the PRISMA statement [19].Due to the high heterogeneity of studies, we were unable to compare effect estimates as originally planned.For that reason, we decided to shift the systematic literature review to a scoping review for synthesizing the results.We extracted information on the study design (e.g., sample size, drop-outs, number and time of followups, primary and secondary outcomes as well as scales for measuring these outcomes) and intervention (e.g., country where the study was conducted, duration and characteristics of intervention and control group).For the latter, we additionally prepared a documentation of game mechanics based on the studies by Toda et al. [20] and Hervas et al. [21] to investigate the use of game elements in mental health promotion and prevention.Thus, the following game elements were included: Reward, Sensation, Progress, Challenges, Surprise, Storytelling/Narration, Social sharing, Level, Leaderboard, Goals, Avatar.For an explanation of the elements used in this study, see supplementary material.The results are described as a qualitative overview, allowing for a systematization of the outcomes and categorization of game elements.
An assessment of the methodological quality of the included randomized controlled studies was conducted by two authors (LA and FF) independently.The assessment was based on the revised Cochrane risk-of-bias tool for randomized studies (RoB2), which also allows to assess the quality of cluster-randomized studies [22].No discrepancies in assessment were observed.

Results
A total of 16 studies were identified and included in the synthesis on the effectiveness of gamified digital interventions on mental health prevention and promotion among adults (Table 2).The measurements, statistical analyses and outcomes differed greatly; however, all studies showed overall a positive impact on mental health outcomes in at least one related outcome measurement.

Characteristics of primary studies
Overall, four studies originated from the United Kingdom (UK), three each from Australia and New Zealand, two from the Netherlands, one from Portugal and one from the United States of America (USA).One study included participants from eleven countries; for another study, the location could not be determined.In most studies, generally healthy working-aged adults were observed (n = 13), while some focused specifically on university students (n = 3).Overall, 3,585 participants were included in the studies, with higher percentage of women (62%).The duration of the interventions ranged from 10 min (one break) up to 12 weeks, with most studies lasting four to six weeks (n = 8).Moreover, the majority of the studies provided an active control condition (n = 12).Of those active control conditions, five studies included access to meditation guides, the provision of information or the filling out of diaries.Further, other active interventions, such as similar applications to the intervention group or apps based on cognitive behavioural therapy (CBT), were visible in four studies.Other comparison groups framing the same intervention just the other way around, or the same intervention with another design or layout (n = 3).Seven studies had a waitlist or inactive control condition (without any intervention).While most studies were twoarmed, four studies used a three-arm design.All studies identified used randomization for allocation to the groups).A comprehensive overview of the interventions, including their used psychological techniques or strategies, can be found in the supplementary material.

Outcomes investigated within primary studies
A broad variety of outcomes investigated within the selected studies was observed.Beyond measuring various dimensions of well-being (n = 7), the studies also analysed resilience (n = 4) and mindfulness (n = 1).In addition, stress (n = 6), depression (n = 5) or anxiety (n = 5), as well as other mental health outcomes (n = 11) such as satisfaction with life, quality of life, or positive and negative affect were examined within the included studies.Due to the large number of different outcomes, the measurement instruments used were highly heterogeneous.Even for the same outcome, various instruments were used (Table 3).A detailed overview of the outcomes examined can be found in Table 3, while all included outcomes within the studies are presented in the supplementary material.

Resilience and mindfulness
In total, four studies included resilience and one study mindfulness.While significant effects were found with respect to resilience for most studies, one of the four studies measuring resilience did not show significant improvements [29].Moreover, the study by Flett et al.
[29] also failed to identify any significant effects on mindfulness.

Well-being
Overall, seven studies investigated well-being within their studies.Most of these studies showed significant positive   -Participants had 12 weeks of access to the OL@-OR@ m-Health program Comparison group: -Participants received a control version of the OL@-OR@ tool which was similar in visual design but that only collected baseline and outcome data  This study reports results from a pilot experiment and real-life experiment.We only present results for the real-life experiment here  [25,26,28,31,37], while all studies examining psychological or mental well-being (n = 5) found significant improvements [15,28,33,34,36].Spiritual well-being (n = 1) [28], interpersonal well-being (n = 1) [34], community well-being (n = 2) [28,34], and economic well-being (n = 1) [34] showed also significant progresses through a game-based  intervention.In contrast, for occupational well-being no significant effects could be observed [34].Physical wellbeing was significantly improved in one study [28], while no effects could be observed in another [34].

Stress, depression, and anxiety
In total, six studies examined stress and five studies each anxiety and depression.There is some evidence suggesting that game-based interventions have significant positive effects on (psychological di-)stress (n = 3) [15,23,35].
With regard to the internalizing mental health problems, positive effects on depression (n = 5) [26,29,30,35,36], and anxiety (n = 3) [33,35,36] were observed.Results related to stress were very mixed, with three studies showing significant improvements [15,23,35] and three other studies without significant effects [27,29,31].All studies used different scales.The results regarding anxiety are similar: Three studies reported significant positive effects, while two studies could not identify any significant improvements.In contrast, all interventions had a significant positive impact on the prevention of depression regardless of the measurements used.

Other mental health outcomes
Next to the mental health outcomes reported, a diverse array of emotional outcomes could be identified within eleven studies.Flourishing partly used as a proxy for well-being, was investigated in two studies.However, no significant results were identified [29,30].Three studies examined satisfaction with life and one study quality of life, while findings are heterogeneous.Using the same satisfaction with life scale, two studies detected significant improvements [23,35], while one study did not [30].Meanwhile, quality of life was significantly improved within two gamified interventions [25,35].However, one study did not show any significant improvements [37].
Personal growth (= 1) [33], sleep problems (n = 1) [37], emotional cognition (n = 1) [36], frustration and irritability (n = 1) [27], cognitive engagement (n = 1) [32], cognitive and affective engagement (n = 1) [32], energetic arousal (n = 1) [24], and recovery (n = 1) [24], were identified in one study each and had a significant association with gamified interventions.At the same time, positive and negative affect, sometimes used as proxy for wellbeing, were included in three studies as a combination and in one study separately.Accordingly, there was a significant association for positive and negative affect visible in two studies [30,32].Interestingly, Howells et al. [30] reported significant improvements for positive affect but not negative affect.In turn, Schakel et al. [37] could not identify any significant effects, neither for negative nor for positive affect (Table 3).

Game elements within primary studies
In total, eleven game elements were applied within the studies included in this review.The most frequently utilized elements were reward (n = 11), progress and sensation (n = 9), followed by challenges (n = 6), surprise (n = 5) social sharing and storytelling/narration (n = 4).Less frequently used were avatars, goals, leaderboards and levels (n = 2) (Fig. 2).Overall, at least three game elements were integrated in the interventions, while Kelders et al. [32] used most game elements (n = 6).
Regarding the use of gamification, an analysis of the elements shows great variations.For example, Routledge et al. [36] showed with the integration of progress a significant improvement in psychological well-being.At the same time, Myers et al. [34] demonstrated that an integration of four elements (Challenges, Progress, Social sharing, and Sensation) also improve psychological Fig. 2 Game elements observed in the studies well-being.Moreover, when integrating challenges exclusively, significant improvements for well-being and depression were observed but not for anxiety [26].Costa et al. [25] pointed out, that the integration of four elements (Challenges, Storytelling, Social sharing, and Sensation) similarly improve well-being.

Quality appraisal of included studies
Overall, the studies showed some or high risk of bias, particularly due to deviations from the intended interventions or missing outcome data.In contrast, the randomization only led to low risk of bias in most of the studies.There was a study which showed only low risk for all five dimensions, whereas one study showed high risk in four dimensions and some concerns in terms of the randomization process (Fig. 3).
In addition to this quality assessment, we have extracted the drop-out rates.Drop-outs varied from 0 to 67.0% per study.Four studies did not report on dropouts, and two studies attributed the drop-out explicitly to technical issues.Most studies performed per-protocol analysis or are based on complete data.Only three studies conducted both an intention-to-treat and perprotocol analysis; one further study performed an intention-to-treat analysis merely.

Discussion
This scoping review investigated the effects of gamified digital interventions in mental health promotion and prevention of mental diseases among working-age adults.Further, it investigated which game elements were most commonly used.Overall, positive effects of gamified interventions on mental health were identified, in particular on psychological well-being and depressive symptoms.However, further outcomes indicated heterogenous findings.Most game elements used were reward, sensation, and progress.However, due to missing information in the primary studies, no substantiated conclusion about the (optimal) quantity or composition of game elements in an intervention can be drawn.
According to our research, this is the first scoping review which investigates the effectiveness of gamified digital interventions on mental health in adults of working age from a health promotion and disease prevention perspective.Previous reviews which, however, did not concentrate on health promotion and disease prevention, underline our findings in a very general way [11,12,40,41].However, when interpreting the results in the context of mental health, one needs to take into account that we explicitly focussed on persons who showed no mental impairments and, therefore, people who have high scores Fig. 3 Quality appraisal of includes studies on a well-being scale and low scores on a depression scale.These floor or ceiling effects may -in contrary to treatment in mentally ill persons -only lead to a limited room for improvement.For that reason, one should not merely focus on statistical significance, but also on effect sizes.However, also small effect sizes may be considered as relevant.
When interpreting the overall results, several further aspects need to be considered.First, the intervention period is relatively short in all included studies.Only two studies had an intervention period of 12 weeks [28,32].Five studies of four weeks, or 30 days respectively [23,26,27,34,36], and four studies of only ten days or less [24,[29][30][31].There are indications that interventions with longer duration are more effective than those of shorter length [ [31], e.g.[26,30,34]]; e.g.interventions with a duration of more than one month did not show nonsignificant results in mental health-related outcomes [e.g.[15,25,33]].Moreover, taking into consideration that half of the studies did not have an active control group [15,23,31,[35][36][37], the findings need to be interpreted with caution.The inactive control group can be compared with the intervention group, however, in these cases no assertion can be made if the gamified intervention is more effective compared to a non-gamified intervention.
Although it has to be acknowledged that all intervention studies included in the synthesis used randomization for allocation purposes, high dropout rates in the majority of the studies should be kept in mind.Drop-out rates might be higher in health promotion and disease prevention than in treatment, due to low psychological strain and lower motivation [42].However, only four studies conducted an intention-to-treat analysis to avoid systematic error caused by drop-outs.
Mental health is influenced by numerous risk and protective factors that interact with each other [43].In this respect, different determinants, such as social conditions, working or living conditions, could influence the effectiveness of game elements in terms of well-being.Myers et al. [34], for instance, found a statistically significant relation with income as well as community and economic well-being.Thus, high-income earners were 2.34 times more likely than low-income earners to comply with the programme.These effects -as already described by Dahlgren and Whitehead [44] -was taken up in a model on digital determinants of health equities [45].Beyond that, it might be interesting to examine a person's individual characteristics in the context of gamification.It is thereby possible that e.g., personality, level of knowledge, experiences or even level of motivation, may influence the effectiveness of an intervention.As an example, interventions could be more effective for individuals who already enjoy playing games in their free time [46].Since too little information on other variables was provided, no conclusive statement regarding these indicators can be drawn.
As a matter of fact, a (long-term) impact of an intervention is one of the most important aspects.The longest follow-up in the included studies was 12 months [26].Other longest follow-up periods were 12 weeks [35], 60 days [34], 30 days [29], and four weeks [37].However, most studies do not report follow-up measurements.For this reason, none of the studies investigated whether well-being is increased in the long term or whether intrinsic motivation is maintained.This aspect is a key element in the health care sector.Plugmann [16] emphasizes that gamification can help to reduce costs in the field of prevention.The authors argue that new products and services with gamification can lead to a breakthrough as an innovative prevention strategy.Thereby, however, it is a prerequisite that individuals are willing to share their data.A survey examining the usage of big data and in relation the protection of the privacy indicated that two third of respondents believed that too little attention is paid to the enforcement of data protection and that it will therefore fail.On a positive note, however, the healthcare industry has the highest level of trust compared to other industries, at over 20% [47].None of the studies included investigates the aspect of cost savings and trusts in digital interventions.In this respect, a long-term view of cost savings and the presentation of the tolerance level when opening private data is an important aspect that could be decisive for the success of gamified interventions in mental health.

Limitations
There are some limitations which have to be taken into account when interpreting the results.First, the included sample covers a wide age range.We focused on people in the working age group (18-65), whereas young individuals have different habits and needs compared to, for example, individuals nearing the age of retirement.Accordingly, a differentiation of age would be necessary in order to address the effectiveness more specifically.Along these lines, the whole age range was not included in any of these studies.Participants in the study of Costa et al. [25] for example, had a mean age of 73 years, which is attributable to their inclusion criteria (study population should be 50 years or older).Thus, participants above the maximum age of 65 years were included.In addition, some studies focused on university students, with a mean age of e.g.21.48 years in Keeman et al. [31] and 22.8 years in Kelders et al. [32].So, within this review, no differentiation of age groups was done.It is therefore critical to consider whether the sample was defined too broadly or whether a differentiated presentation is necessary.
Second, about half of the studies comprised small sample sizes of fewer than 100 participants [23-25, 27, 31, 32, 37].However, some studies investigated more than 300 participants [33][34][35][36], whereas Firestone et al. [28] counted actually 794 participants.The heterogeneous number of included participants was not taken into account and thus no weighting of the results was performed.Due to this and the heterogenic target sample, the generalisation of findings is limited.
Third, a limitation arise from the lack of comparability of the studies, as different survey and evaluation instruments were used.For instance, the overall well-being counts four different scales for four measurements.Similar observations were made for stress, depression, and anxiety.
Fourth, all studies reported at least one significant result.This might be an indication of publication bias.
Furthermore, a variety of statistical analyses were used.This results in a fifth and major limitation.Ratings were presented as a plus (significant positive effect), if one out of various analysis found a significant impact.Since the analysis were very heterogeneous, no differentiation between the analysis and the number of significant results has been made.Therefore, the conclusion about the strength of the effects is limited.
Sixth, several studies used HeadSpace as a gamified intervention [15,23,24,27,29,31].Thus, due to the high number of studies based on the HeadSpace, the variety of gamified interventions is limited, resulting in a potential for bias as this intervention and its effects are given a higher weighting.
Seventh, the usage of game-based elements is diverse and no constant findings were observed.In some cases, especially concerning HeadSpace, the elements used are not described clearly, and different wordings are used.Thus, findings addressing these elements are insufficient, which indicates the need to a well-developed categorization or taxonomy for game elements to make clear statements about the number, type, and combinations of elements which are effective for promoting mental health and preventing mental diseases.
Finally, the quality assessment revealed some or high risk of bias, which must be taken into account when interpreting the results.Although we decided to not only focus on randomized controlled trials, but to include all studies with a control condition, all studies used randomization for allocation.The study quality is heterogeneous, but not that bad as one might expect from experiences of digital interventions in the previous decades.Therefore, as more studies become available, a more detailed perspective on study design and statistical analysis should be considered.

Conclusions
There is evidence of the effectiveness of digital game elements in improving mental health among workingage adults.However, findings are still limited, and results are heterogeneous, which can be traced back to the different interventions and designs applied within the included studies.A variety of eleven elements was used, and it was not clear whether some features or combinations are more important in the context of mental health promotion and prevention.
Despite the limited research field, the present review indicates important insights and tendencies for research and practice.Thus, especially sociodemographic variables, such as a differentiation between age groups, should be considered in future research.In the course of the next few years, it will be important to identify the long-term effects in order to expand the innovation capability described above.Accordingly, this means for policy to support digital gamified interventions, in research and in practice, to promote one's mental health.The implementation of digital interventions for treatment presents one important step in the right direction.Now, gamification should become more of a focus.Another important aspect that was taken up is the satisfaction of basic needs.In future work, the reference to self-determination theory could be more strongly focussed, whereby a more well-founded statement could be made with regard to the number and combination of game elements.Along these lines, a taxonomy is needed to adequately address relevant game design elements in the field of mental health promotion and prevention.
For the practice of health promotion and prevention, the increasing digital innovations result in new interfaces that need to be linked in the future.For instance, data protection is more important than ever, and the success of gamified interventions is related to the trust of users.Among other things, the current information overload is an important issue.Therefore, high-quality interventions need to be made transparent.With regard to reaching specific groups of people and associated effects, prevention and health promotion also face innovative strategies.In conclusion, some gaps with considerable potential for further research and practice in health promotion and prevention can be identified in this still very young field of research.
• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year

•
At BMC, research is always in progress.

Learn more biomedcentral.com/submissions
Ready to submit your research Ready to submit your research ?Choose BMC and benefit from: ? Choose BMC and benefit from:

Table 1
Inclusion and exclusion criteria

Table 2 (
continued) RCT Randomized control trial, cRCT Cluster randomized control trial, M Median, f Female, m Male, CBT Cognitive behavioural therapy, ICBT Internet-based cognitive behavioural therapy 1 Numbers of participants refer to the time of randomization 2 Information taken from [38, 39] 3 In this case only the HeadSpace group is identified as the intervention group because of gamified elements 4 Only results from the real-life experiment are reported because Study 2 has no control group 5

Table 3
Outcomes and scales used within primary studies effects through game-based interventions.The general well-being significantly improved in five studies