In recent years, heightened attention has been directed toward mental health in young people, primarily prompted by the widespread prevalence and profound impact of various mental health conditions. The aggregate estimated prevalence of mental disorders reported for individuals aged 5 to 24 years indicates that more than 1 in 10 children and youth around the world, which equates to 293 million individuals, live with at least one diagnosable mental disorder [1]. The majority of mental disorders appear by the age of 14 years but often remain undiagnosed and untreated in adulthood [2]. The pooled incidence of mental health conditions in children and adolescents has been reported to be 13.4% globally [3], and a recent systematic review found that the incidence rates were higher in Europe at 15.5% [2]. Over the entire life course, 25% of all years lived with disability attributable to mental disorders were recorded before the age of 25 years [1]. The gravity of this matter is striking when looking at various rates of mental ill health around the world [4]. The prevalence of mental disorders varies considerably in different countries, for example, 20% in North America, 12% in Europe and Asia, and 8% in Africa [4]. Differences in gender and age are also present as males are more likely to receive a diagnosis of attention-deficit/hyperactivity disorder compared to females, while females are more likely to be diagnosed with depression compared to males, and adolescents are more susceptible than children to affective disorders and behavior disorders [5]. It is worrying that only 58% of young adults aged 18-25 years with severe mental disorders are accessing treatment [6]. Anxiety, mood disorders, such as depression, and behavioral disorders are among the most prominent contributors [7]. Since 2020, state interventions in the context of the COVID-19 pandemic, such as enforced isolation and school closure, have most certainly increased the burden on young people’s well-being and increased the likelihood of acquiring mental ill health or experiencing worse mental ill health [8]. Lower health-related quality of life and higher anxiety levels are now more common than before COVID-19, particularly among people with a poor socioeconomic position, a migration background, or limited living space [9]. Despite these statistics, the global median of government health spending on mental health is less than 2%, and demand for face-to-face therapy continues to surpass capacity [10,11].

These trends highlight the urgent need for comprehensive strategies to address the complex interplay of mental health challenges using scalable solutions, as reflected in the health policies and strategy reports at both European [12] and global [13,14] levels. The evolution of mental health services to incorporate digital interventions stands out as a promising avenue for effecting positive change on a global scale. One such scalable response has focused on the use of information and communication technology to boost capacity to support and improve young people’s mental health [15].

Digital mental health interventions (DMHIs) refer to the digital delivery of well-established psychological treatments by leveraging the use of technology, including digital devices such as computers and smartphones [16]. DMHIs can include text, video, or audio-based technology, and the most commonly used DMHI is internet-delivered cognitive behavioral therapy (CBT) [17]. Internet-delivered CBT has demonstrated high usability and acceptability [18] as well as effectiveness at reducing the symptoms of depression [19] and anxiety [20] among young people.

There are several barriers and facilitators for the use of DMHIs. From the perspective of young people, perceived facilitators include anonymity, accessibility, and prompt feedback [21]. The reported advantages of digital technologies include greater reach to geographically isolated populations, flexible access, increased convenience, fewer visits to specialist clinics, greater privacy and anonymity, enhanced treatment fidelity, rapid scalability, and low-cost delivery [22].

Barriers include lack of personalization to individual needs, fear of misdiagnosis, and issues with the effectiveness of DMHIs [23]. Previous studies have reported low adherence and high dropout rates among adolescents and young adults [24]. Human interaction has been highlighted as an important factor influencing engagement, as low adherence rates have been reported in online interventions without therapeutic guidance [25,26]. Therefore, guided interventions usually have higher engagement rates than unguided interventions [27]. In this context, blended interventions refer to the integration of digital interventions with face-to-face mental health care [28,29], and they have previously been found to be effective in improving students’ mental health [30]. Nonetheless, a previous report has highlighted heterogeneity across meta-analyses in the level of detail regarding the nature of interventions, the target populations, and the type of support delivered, making it challenging to draw strong conclusions with regard to the circumstances under which human support is most effective [31].

Many reviews have been undertaken to evaluate the wide range of digital mental health supports that are accessible, but there is a dearth of up-to-date reviews, with many published in 2021 or earlier, thus failing to consider the change in the digital world since COVID-19 [20,32-35]. Most prior systematic and scoping reviews focused on specific topics only, such as technologies (eg, apps only [36,37]), specific aspects of mental health (eg, depression and anxiety only [20,32,33,35,38] and mental health promotion only [8]), study populations (eg, university students for convenience [38,39]), and research techniques (eg, randomized controlled trials [RCTs] alone [37]). Previous reviews [20,32,33,37-40] primarily focused on determining effectiveness rather than on assessing the implementation features of existing approaches, such as delivery formats, modalities, and retention rates, across the spectrum of interventions from promotion and prevention to early intervention and treatment, which is the focus of this review. Given that previous studies often focused on a single population type, this scoping review explored the inclusion of marginalized populations in the digital mental health support context. UNICEF defines disadvantaged, vulnerable, or marginalized adolescents as “individuals aged 10-19, who are excluded from social, economic, or educational opportunities enjoyed by other adolescents in their community due to numerous factors beyond their control” [41]. These factors include those operating at the social level (eg, economic inequality, violence, stigma, racism, and migration), family level (eg, neglect and abuse), and individual level (eg, disability and ethnicity). Examples of disadvantaged, vulnerable, or marginalized young people include immigrants, refugees, orphans, and those who belong to stigmatized indigenous, ethnic, or religious groups. They also include individuals who identify as belonging to “sexual minorities” (eg, gay, lesbian, bisexual, and queer) or “gender minorities” (eg, transgender and gender diverse), which will be referred to as lesbian, gay, bisexual, transgender, and queer (LGBTQ+) in the context of this review.

The purpose of this study is to report on a scoping review that was undertaken to investigate research findings on the range of DMHIs and supports available for a broad range of young people aged 16-25 years, with a particular focus on digital tools, modalities, delivery formats, target populations, and retention rates.

Specific research questions are as follows:

A review protocol for this study has been registered with Open Science Framework (OSF) [42]. The search process was guided by the main stages outlined in the Arksey and O’Malley framework [43]. A search of the literature was initially conducted on December 16, 2022, and updated on June 7, 2024, in the following databases: PubMed, Web of Science, Scopus, MEDLINE, Cochrane Library, and PsychInfo. The search combined terms related to 4 key concepts, including mental health, technology, young people, and interventions. The search strategy can be found in Table 1, and examples of search results can be found in Multimedia Appendix 1. Reporting was guided by the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) checklist (Multimedia Appendix 2).

Articles published in English only were considered for inclusion. The selection criteria are presented in Textbox 1. In the registered protocol on OSF, the original criteria stated that articles reporting on a study population of those aged 12-25 years would be included; however, this was later refined to 16-25 years. Given that there is no universal agreement on the definition of the age group of youth, populations between 16 and 25 years were selected to capture the out-of-school population of young people in the United Kingdom and Ireland as per education policies, and these ages fall within the range for young people reported by the World Health Organization [44]. Grey literature was not included in the search strategy. Initially, the protocol on OSF stated that the search would include articles from the last 5 years, from 2017 to 2022. However, the search was updated in 2024 and thus included articles published from 2019 onwards. The search was restricted to studies published in the last 5 years (2019-2024) to ensure the review included the most relevant up to date literature, considering the rapid adoption and assessment of digital mental health technologies throughout the sector in recent years, particularly during the COVID-19 pandemic [45] and beyond.

All records were imported to Covidence [46]. Duplicates automatically detected by Covidence were removed prior to screening. Title and abstract screening were conducted by 4 independent reviewers (CP, TW, CK, and AMC). During the screening, articles were sorted by title in alphabetical order to compare the title, author name, and abstract with consecutive articles to manually delete duplicates. Three reviewers (CP, TW, and CK) completed the full-text review. For both title and abstract screening and full-text review, 2 independent reviewers screened all articles. Any conflicts were discussed and resolved between these 2 reviewers to reach consensus.

One author (CP) created the data extraction template, which was later refined with 2 additional authors (TW and CK). The data extraction template included the main study characteristics (year, aim, study design, and location), participant details for experimental and control groups (age, gender identity, inclusion of marginalized young people, number of participants, where participants were recruited, method of recruitment [eg, convenience sampling], and participant type [eg, general population]), intervention details (target area, delivery [ie, digital only or blended digital and human support], blended intervention details, type of digital tool, primary and secondary outcome measures, scales used, follow-up, psychological modality used in interventions, duration, incentives, promotion/prevention or intervention, intervention name, and features), and retention details (number of those who started and completed the study). One author (CP) extracted data from all articles included in the scoping review. A second author (CK) completed data extraction for one-third of the articles, and the first author (CP) completed a consensus template for these studies and discussed any discrepancies with the second author.

Data were analyzed in R using R version 4.3.2 (R Project for Statistical Computing).

A total of 36,591 records were identified across the 6 databases that were searched; however, 23,285 articles were removed as they were identified as duplicates (Figure 1). Through abstract and full-text screening, 12,893 articles were removed, and the reasons for removal are outlined in Figure 1. Eventually, 145 articles were included in the final review (Figure 1).

Table 2 provides an overview of the country of origin, research methodology, and year of publication for each study. Of the 145 studies, 80 (55.2%) reported on digital tools for primary prevention or mental health promotion and the remaining 65 (44.8%) reported on digital tools for early intervention or treatment of mental ill health symptoms.

The interventions ranged from one-off single sessions with a digital tool to sessions lasting up to 26 weeks (6 months) (Multimedia Appendix 3). The most common study length was 4 weeks (Multimedia Appendix 3), and the mean study length was 6 weeks. Other prescribed intervention durations were 1-2 weeks, 1-9 weeks, 3-6 weeks, and 3-12 weeks. Two articles did not detail the study length. Less than half of the articles (63/145, 43.4%) recorded participant follow-up outside of the conventional pre-post study data. A total of 48 of these 63 studies recorded only 1 follow-up anywhere from 2 weeks after the intervention period to 52 weeks. The most common follow-up period was 12 weeks, and the mean follow-up period was 19 weeks. Ten studies followed up with participants at 2 time points (eg, at 4 and 12 weeks), and 2 studies followed up with participants at 3 time points (eg, at 8, 12, and 24 weeks).

Ten interventions were used across more than one study, including Calm [71,120], EMIcompass [148,158], Grow It! [150,151], imi [75,85], LifeBuoy [94,101], Mindfulness Virtual Community [139,141], Spark [55,93], Step-by-Step [128,130], Tita [162,164], and Whitu [156,157]. Thus, this section of the results reports on all unique interventions (n=135) after excluding duplicate interventions.

In studies that recruited participants from the general population (n=63), the most common features used in digital interventions were psychoeducation or education (44/63, 70%), mindfulness or breathing exercises (37/63, 59%), and mood logging (20/63, 32%) (Figure 9).

In studies that recruited people experiencing mental ill health symptoms (n=38), the top features in digital interventions were the same, including psychoeducation or education (32/38, 86%), mindfulness or breathing exercises (17/38, 45%), and mood logging (16/38, 43%) (Figure 9).

In studies that recruited people with a mental health diagnosis or those who met diagnostic criteria (n=27), the most common features in digital interventions were psychoeducation or education (21/27, 78%); mood logging, check-in, or self-monitoring of symptoms (16/27, 59%); and mindfulness or breathing exercises (5/27, 25%) (Figure 9).

Personalization or customization was a more common feature in interventions for people experiencing mental ill health symptoms (6/38, 16%) or those with a mental health diagnosis (7/27, 26%) compared to the general population (5/63, 8%) (Figure 9). Self-help plans or relapse plans were used more commonly in people experiencing mental ill health symptoms (4/38, 10%) or those with a mental health diagnosis (4/27, 15%) compared to the general population (1/63, 1%) (Figure 9). Gamification or reward incentives were used more often in interventions for the general population (15/63, 24%) and for people experiencing mental ill health symptoms (7/37, 19%) compared to interventions for those with a mental health diagnosis (1/27, 4%) (Figure 9).

Retention rates ranged from 16% to 100%, with a mean retention rate of 66% (SD 23%) across all studies. Table 7 highlights the retention rate across study type, delivery, type of digital tool, study population, and incentive versus no incentive.

This study sought to investigate the literature on DMHIs for young people aged 16-25 years in order to assess (1) the characteristics of studies, (2) the characteristics of DMHIs offered to young people and the level of human support in combination with digital mental health, (3) the target populations of DMHIs for young people, and (4) the retention rates across studies involving DMHIs for young people.

Studies were spread out globally, mostly originating from the United States of America (43/145, 29.7%), Australia (16/145, 11.0%), and the United Kingdom (14/145, 9.7%). There was a range of study types (mostly RCTs: 81/145, 55.9%), and the majority of studies had a control group (88/145, 60.7%). Just under half of the studies (65/145, 44.8%) focused on treatment, and the rest (80/145, 55.2%) focused on mental health promotion or prevention of mental health problems. The mean prescribed length of using an intervention was 6 weeks, and 43.4% (63/145) of studies recorded follow-up data after the intervention period, most commonly at 12 weeks. Participant numbers varied greatly across studies from 8 to 2222, with a mean of 167 but with a large SD of 303. Over half of the studies (80/145, 55.2%) provided reimbursements for study participation, and these were mostly in the form of financial incentives.

Interventions mostly targeted general mental well-being, depression, anxiety, or stress. Apps (51/135, 37.8%), web-based resources (45/135, 33.3%), and websites (19/135, 14.1%) were the most popular digital tools. Psychoeducational content (97/135, 71.9%) and mindfulness or breathing exercises (70/135, 51.9%) were most common across digital interventions. Moreover, 48.1% (65/135) of interventions involved a blended approach combining human support with a digital intervention, while 51.9% (70/135) involved a digital-only approach.

The largest group of participants across studies included young people from the general population without reported mental health issues (63/145, 43.4%). The next most common group included those experiencing mental ill health symptoms (38/145, 26.2%), followed by those with a mental health diagnosis or meeting diagnostic criteria (27/145, 18.6%). Among those diagnosed, most had anxiety or depression; 5 studies included participants with depression or anxiety along with other co-morbid mental health conditions and 5 studies included participants with other mental health conditions. Convenience sampling was prevalent (140/145, 96.6%), with universities and colleges being the primary recruitment location. Three-quarters of the studies (108/145, 74.5%) included young people from marginalized groups. Digital-only approaches were common for the general population, while blended human-digital support was more often used for those with mental ill health symptoms or a mental health diagnosis. CBT, positive psychology, and mindfulness were the most popular approaches in interventions for the general population, whereas digital interventions for mental health symptoms or diagnosis often used CBT or other specific therapeutic modalities. Gamification or reward incentives were more prevalent in interventions for the general population (15/63, 24%) and populations experiencing mental health symptoms (7/37, 19%) and less prevalent in interventions for individuals with mental health diagnoses (1/27, 4%). The mean age of participants was 20.8 years across all studies. Females were largely overrepresented (71% of study populations), while males typically constituted 30% of study populations. Only 36.6% (53/145) of studies included LGBTQ+ participants, including transgender and nonbinary individuals. In these studies, LGBTQ+ representation averaged 13%.

The average retention rate across all studies was 66% (SD 23%). Retention rates were higher than the mean for pilot studies (mean 79%), apps (mean 76%), studies that recruited participants with a mental health diagnosis (mean 76%), and studies that offered any type of incentive (mean 73%).

In this review, some information was missing during data extraction as indicated in the results. For example, 12 studies reported age ranges instead of average age, but most data were largely reported. As a scoping review methodology was employed, a risk of bias assessment or an assessment of the efficacy or effectiveness of DMHIs was not conducted. As a quality assessment was not conducted, the interpretation of findings regarding retention rates might be affected. In particular, 44% of studies were not RCTs, which are considered the gold standard in research for evaluating effectiveness. Therefore, the finding that retention rates for pilot studies were higher (79% on average) than those for RCTs (70% on average) should be interpreted with caution. Given the heterogeneity of DMHIs included in this study, a systematic review and meta-analysis was not feasible, and thus, it was not possible to assess effectiveness. However, future work could focus on specific areas, for example, apps for specific cohorts of young people, to assess effectiveness. Despite these limitations, this scoping review provides extensive insights into the current landscape of DMHIs for young people, ranging from mental health promotion and primary prevention to more targeted treatment interventions, and indicates useful areas for future research.

While roughly half of the studies included in this review focused on the treatment of common mental health problems, future work could direct more resources toward promotion and prevention measures to halt the escalation of mental ill health. The period between 18 and 25 years of age represents a transitional phase marked by increased exposure to risk factors like instability, employment search, identity exploration, and heightened self-focus, elevating the risk of mental health conditions [195]. Thus, for young people, it is important to focus on mental health promotion and prevention strategies, which include universal, whole population approaches irrespective of current mental health status; selective approaches for specific risk factors; indicated approaches that focus on early subclinical symptoms; and tertiary approaches that target specific mental health conditions [196]. A recent meta-analysis found that the use of digital mental health promotion tools among youth aged between 11 and 18 years resulted in small significant improvements in general well-being and small-to-medium significant improvements in anxiety [8]. The current literature indicates that universal approaches used within psychological or psychoeducational interventions may improve symptoms of anxiety but may not prevent depressive or anxiety disorders [197]. Further research into digital mental health promotion for young people would be beneficial to strengthen these findings, given the heterogeneity of studies in this area.

Psychoeducational content and mindfulness or breathing exercises were most commonly used across interventions for all populations, and throughout the studies, there was a major focus on CBT and mindfulness-based approaches. Although CBT is considered the gold standard for treatment, as a previous systematic review and meta-analysis found that digitally delivered CBT interventions had greater effect sizes compared to other modalities [198], future interventions could use a broader spectrum of therapeutic models. Acceptance and commitment therapy, which is another type of third-wave cognitive treatment, was used in a small proportion of studies, and it has been previously demonstrated to be effective in enhancing and sustaining mental health outcomes across various demographics; however, evidence of the clinical significance of these effects is lacking [199]. Given the scarcity of research, a future direction could be to run large-scale trials to determine the effectiveness of other therapeutic modalities outside of CBT or combined approaches, particularly in the digital mental health context. In addition to these therapeutic modalities, there is limited research on theoretical models informing DMHIs, particularly for young people. A commentary paper provided a conceptual overview of how established behavior theories and models, such as Health Belief Model, Theory of Planned Behavior, Transtheoretical Model, and Social Cognitive Theory, can inform the development of a digital intervention for individuals with mental ill health [200]. Naslund et al [200] suggested that individual characteristics should inform intervention design and shape the content, while theory should inform strategies to support behavior change, which can be modified in real-time based on user feedback, and that theory can also guide outcomes to inform behavior mechanisms and intervention modifications. In 2020, the World Health Organization published a framework for planning, developing, and implementing youth-centered DMHIs, which mentions developing a theory-driven approach but does not provide explicit details on how to approach this using psychological theory [201]. Further research is needed on how these theoretical approaches can be adapted for digital delivery and their appropriateness for young people.

One of the most prominent results was the exclusion of more marginalized groups outside of ethnic minorities. However, there may be underreporting of the inclusion of marginalized groups as studies may not always capture this information when collecting participant demographic information. Based on the data used in this review, there was a lack of gender diversity and lack of representation from certain populations, including young people who were unemployed, were living in a rural area, were socially or economically disadvantaged, had a low education level, were neurodivergent, were living with a physical health condition, were migrants or immigrants, had a low income, were homeless, were experiencing current or previous substance use issues, and were not educated or employed. These groups are often excluded from digital mental health studies, as convenience sampling is most often used, with participants most commonly being recruited through universities or colleges. In particular, LGBTQ+ young people often experience victimization and have poorer mental health and higher rates of self-harm and suicidal ideation compared to their cisgender and heterosexual peers [202,203]. Looking more broadly across the whole population in a UK mental health service, a recent study highlighted that certain vulnerable groups, including ethnic minorities, individuals with disabilities, those born outside of the UK, and those with lower academic attainment, were underrepresented in psychological therapies at the national level [204]. Other previous work also identified that more research is needed on preventative DMHIs in young people having poor or underserved backgrounds [205]. It is clear that future work in this area should aim to be more inclusive, with representation from marginalized and vulnerable young people.

In addition, there was a lack of inclusion of a range of mental health conditions outside of anxiety and depression, given that only 5 studies focused solely on study populations of people with diagnoses that did not include anxiety or depression. Previous systematic reviews on digital interventions for young people with anxiety and depression have found small to medium effect sizes when comparing digital interventions, most commonly using CBT modalities, with interventions used in a control group [32,33]. In addition, studies targeting those with mental health diagnoses were relatively underpowered, with sample sizes falling short of the previously reported median of 106 participants in efficacy and effectiveness trials [206].

The average prescribed length of using an intervention was 6 weeks, but there were large variations from a single one-off session to a use period of 6 months, and only 43.4% (63/145) of studies recorded follow-up data after the intervention period at different time points. Compared to traditional therapy, there are no consistent prescribed lengths for DMHIs. Depending on the type of DMHI being delivered, future work could explore intervention lengths akin to that prescribed by psychological therapy services. For example, in the UK, the National Health Service provides “NHS Talking Therapies,” formerly known as Improving Access to Psychological Therapies, which follows a stepped care model. Within this model, individuals experiencing mild-to-moderate mental ill health symptoms (step 2) would receive low-intensity treatment, which could be up to 6 sessions with a mental health professional. There was also a lack of long-term follow-up data in over half of the studies, which would make it difficult to track sustained changes in mental health over time. Future work should look at standardizing the length of treatment within DMHIs and including long-term follow-up data to help determine the efficacy or effectiveness of digital tools compared to traditional face-to-face mental health support.

The average retention rate across all studies was 66%, which falls slightly short of the weighted average of 69% reported for the retention of internet-based mental health interventions [207]. In comparison to other studies in this field, 74.8% of young people aged 11-18 years completed mental health promotion interventions with a digital component [8] and 80% of children and young people used digital health interventions or completed follow-up measures [21]. The average rate identified in this study is also comparable with that in other studies on youth mental health interventions generally that did not involve a digital component. For example, a study that examined interventions for young people at risk of psychosis reported a pooled retention rate of 66% [208]. For long-term clinical studies, a rate of 80% or higher is considered acceptable or high completion [209], so there is still room for improvement. Sustaining engagement with DMHIs in real-world settings outside of RCTs is even more challenging, as highlighted by Fleming et al [210], and only 0.5% to 28.7% of users from the general population completed or continually used self-help interventions for anxiety and depression. Some previous strategies have been identified to recruit young people and keep them engaged in longitudinal health research, such as recruitment through social media, including financial incentives for taking part, joining with peer groups, and incorporating data collection that is user friendly and flexible, and young people also emphasized the importance of social connection with both peers and the research team [211]. Another study that examined barriers and facilitators to engagement with digital health interventions among those aged 2-25 years found that children and young people preferred digital health interventions with features such as videos, limited text, personalization, ability to connect with others, and options to receive text message reminders [21]. An important area for future work will be to identify or incorporate better strategies to improve engagement and completion, particularly for marginalized populations. The main recommendations for the field based on key findings in this review can be found in Table 8, which include general recommendations for future research, as well as relevant findings for clinicians, technology developers, and policy makers.

This scoping review has provided a comprehensive overview of DMHIs for young people aged 16-25 years. It has highlighted the global spread of studies, focusing on the treatment, prevention, and promotion of mental health. The scoping review sheds light on the need for greater inclusivity in research, given the high proportion of females who are often recruited from universities through convenience sampling. The review also highlights a lack of representation from marginalized groups and emphasizes the importance of long-term follow-up data to assess the efficacy and effectiveness of these interventions. Additionally, the review highlights challenges in sustaining engagement and completion, particularly outside of controlled trials, and suggests the importance of incorporating better engagement strategies, especially for marginalized populations. Overall, this review calls for a more inclusive and comprehensive approach to DMHIs for young people.