How effective are mobile mental health application in preventing the escalation of subthreshold anxiety to clinical levels among UK university students age 18-24?
Student’s name
Student’s number
Date
Abstract
Aim: The review will place available evidence on the effectiveness of mobile mental-health applications in preventing the development of clinical Generalised Anxiety Disorder (GAD) in students aged 18 to 24 years at UK higher education institutions, following subthreshold anxiety.
Methods: We examined five peer-reviewed studies that used randomised controlled trials, qualitative thematic analysis, and mixed-methods research designs. The inclusion criteria were that the participants had to report subthreshold anxiety, use a digital application based on transdiagnostic risk factors, and report longitudinal outcomes. Data were extracted based on cognitive processes, engagement outcomes, and clinical outcomes and the risk of bias was assessed using the CASP tool.
Results: Repetitive Negative Thinking (RNT) apps showed a statistically significant decrease in the long-term incidence of GAD (RR=0.68, 95 per cent CI = 0.52-0.90). Nevertheless, an engagement-outcome paradox emerged: the therapeutic effect was dose-dependent, but motivational impediments and a sense of burden increased attrition to 45%. The moderation results showed that younger (18-20 years old) and patients with the shortest anxiety history (< 6 months) had the highest gains, especially with the use of apps in a blended care model, such as self-monitoring and infrequent therapist interactions. Qualitative themes revealed the significance of user agency, personalised content, and real-time feedback loops, and the inability to sustain it was exacerbated by a lack of digital literacy.
Conclusion: Mobile applications have untapped potential as preventive measures, but only when person-centred design, minimal-dose effects, and inclusion in stepped-care pathways are in place. Universities are supposed to integrate apps into the broader mental-health provision, rather than implement them independently, so that their use supports the introduction of digital literacy.
Future implication: Future studies should be diverse in their demographic representation and develop Just-in-time adaptive interventions to alleviate the motivational drop-off and improve retention and clinical outcomes. Overall, the concept of digital prevention is quite reasonable, though its practical effectiveness warrants consideration when applied to the existing clinical ecosystem.
1. Introduction
Globally, one in seven 10-19-year-olds experiences a mental disorder, accounting for 15% of the global burden of disease in this age group, with depression, anxiety, and behavioural disorders being the leading causes of illness (WHO, 2025). The students of the university, especially between the ages of 18 and 24 years, are highly susceptible to mental health issues, and anxiety is one of the most prevalent cases observed in this group of people. In the UK, the percentage of university students reporting a mental health condition rose to 5.8% in 2022/23 compared to less than 1% in 2010/11 (Lewis and Stiebahl, 2025). Nevertheless, confidential surveys indicate far greater prevalence rates of mental health problems, with 57% of students self-reporting a mental health problem, and 27% with a diagnosed issue (Lewis and Stiebahl, 2025). Unattended subthreshold anxiety may develop into overt clinical anxiety or other forms of mental illness, further impacting academic performance, social performance, and general quality of life.
In recent years, mental health provision in higher education has undergone tremendous development and advancements, with the enactment of the Special Educational Needs and Disability Act of 2001 and the Equality Act of 2010, which imposed legal obligations on institutions in providing services to students with mental health conditions (Royal College of Psychiatrists, 2011). Mental Health Support Teams (MHSTs) have been introduced in schools and colleges in England since 2018/19, with more than 600 teams expected to be in place by Spring 2025, serving 52% of the school population (NHS England, 2025). These teams will be extended to every school by 2029/30 by the government (NHS England, 2025).
Since conventional face-to-face treatment is not always possible or adequate, mobile mental health applications (mHealth apps) have emerged as a potential solution to support students with anxiety symptoms in a more personalised and convenient manner (Borjalilu et al., 2019). Apps may include evidence-based approaches such as Cognitive Behavioural Therapy (CBT), relaxation, and mindfulness and aim to decrease anxiety and prevent its development to clinical levels (Apolinário-Hagen et al., 2020). McCloud et al. (2020) studied whether a self-guided mobile app, Feel Stress Free, is an effective intervention in the treatment of anxiety and depression symptoms in university students. These outcomes showed that the app had a significant impact on the symptoms of depression during week six. The trend of mHealth usage in the UK has been facilitated by the national efforts. For example, NHS England has a range of free wellbeing apps available to help ease mental health. The Student Health App is a PIF TICK-certified, free self-care application used in UK higher education institutions that offers students evidence-based health information, support tools, and early intervention (Expert Self Care, 2025). The WellMind app is an NHS mental health and wellbeing application that should aid in stress, anxiety, and depression (Stand‑by‑me, 2025)
Following these findings, this article aims to identify how effective mobile mental health applications are in preventing the escalation of subthreshold anxiety to clinical levels among UK university students aged 18–24. The research question was framed within the PIO framework to assess the existing evidence on the effectiveness of these mobile interventions, focusing on their potential positive and negative impacts on students’ mental health outcomes, through a scoping review of the current literature.
2. Literature selection and Identification
This study aimed to examine the way mobile mental health apps can reduce the development of subthreshold anxiety to a clinical level in UK university students aged 18-24. To facilitate the search process, the research question was further subdivided into the following aspects: the population (university students aged 18-24), the intervention (mobile mental health applications), and the outcome (prevention of anxiety escalation and improvement in well-being). Based on this, a set of specific keywords was identified, such as “young adult,” “university student,” “mobile mental health,” “smartphone app,” “anxiety,” “subthreshold anxiety,” “anxiety prevention,” and “early intervention.” The search strategy used Boolean operators (AND, OR) to optimise the search for keywords such as mobile mental health apps, university students, and subthreshold anxiety.
The searches used the CINAHL and MEDLINE databases via the EBSCOhost platform, as they are reputable sources for health and psychological studies. The choice of these databases is due to their access to peer-reviewed, high-quality academic literature directly related to the current research topic. Articles older than 2015 were then excluded to focus on recent publications, which reflect the most up-to-date trends and developments in mobile mental health interventions. The inclusion criteria for the studies required that participants be university-aged students (18-24 years) or young adults, with a focus on anxiety or subthreshold anxiety. The studies were also required to investigate the application of mobile mental health applications or interventions aimed at treating anxiety symptoms, either by randomised controlled (or) qualitative research or mixed-methods approaches.
The PRISMA framework by Page et al. (2021) (Figure 1) guided a systematic search to evaluate the suitability of mobile mental health apps to prevent the development of subthreshold anxiety among students aged 18-24 years at a UK university. The CINAHL and MEDLINE databases on the EBSCOhost platform were searched, and 125 records were identified. No other sources detected any additional records, so the total was 125. After eliminating duplicates, 115 records remained. A total of 80 records were filtered based on relevance; 35 full-text articles were evaluated against eligibility criteria, and 30 were excluded because they did not address mobile mental health apps, subthreshold anxiety, or university students. A total of 5 studies met the inclusion criteria and were included in the synthesis.
|
||||
3. Critical appraisal
3.1 Critical appraisal of the studies
Based on the Critical Appraisal Skills Programme (CASP) framework, the evidence from past chosen studies regarding the spectrum of methodological rigour, internal validity, and applicability to the target population was reviewed (See Appendix C). The CASP checklist applied to the primary studies focuses on the significance of study design, sample, data collection, analysis and ethical issues. Topper et al. (2017) and Funk et al. (2025) were the strongest in this respect because they were randomised controlled trials (RCTs), which are generally considered the standard for assessing the effectiveness of interventions. The sample size (251 participants) and the 12-month longitudinal follow-up indicated that Study 2 by Topper et al. (2017) had high internal validity, and it was possible to state that the observed patterns in the prevalence of generalised anxiety disorder (GAD) resulted from the intervention and not from other factors. This sample size provided sufficient statistical power to conduct a credible mediation analysis linking decreases in rumination-negative thought (RNT) with lower GAD rates, thereby fulfilling the CASP requirement for rigorous analysis.
Nevertheless, a demographic structure of 83.7% females introduces a significant gender bias, undermining the external validity of the study for male students and not necessarily representing the university population as a whole. Funk et al. (2025), who used a slightly younger group (16-22 years old), with an even larger sample (365), also contributed to the evidence base since they were able to isolate active therapeutic elements by using two versions of the apps, such as concreteness training as opposed to an entire RNT-oriented treatment. This design is consistent with CASP’s focus on internal consistency and the isolation of mechanisms’ causal effects. However, the high attrition rate in the study weakened statistical power. It raised questions about the sample’s representativeness, since a high dropout rate is likely to systematically exclude participants who are less engaged or more anxious at baseline and, consequently, bias estimates of the effect. More importantly, both RCTs recognised the shortcomings of demographic diversity, a consistent criticism of digital mental health studies that might overlook minority or less educated groups of students whose needs and usage behaviours may vary significantly from the majority-female, highly educated sample. Marshall et al. (2021), in contrast, used a digitally enhanced single-case experimental design (SCED) comprising 39 participants, provided deep qualitative data on the experiences of an individual, but faced multiple CASP limitations. For instance, the average age of respondents (34.04 years) was well outside the 18-24 range, rendering the study’s relevance to the university situation questionable.
Furthermore, unregulated external stress, represented by the opportunistic influence of the COVID-19 pandemic, might have complicated symptom trajectories, thereby limiting the study’s internal validity. Although SCEDs are appreciated for their ecological validity and ability to measure behavioural change, the limited age cohort and confounding factors due to the pandemic reduce confidence that the results will reflect the target population. Valentine et al. (2024) used semi-structured interviews with 15 study participants aged 16-25 years, thereby enhancing relevance. Thematic analysis of these interviews helped to shed light on a certain cognitive impediment, the negative thoughts, which quantitative scales do not measure, and thus added to the picture of how the app works in minimising the psychological distress and anxiety of the patients. In Balaskas et al. (2022), 600 app-store reviews were paired with 15 semi-structured interviews to provide a mixed-methods approach, which meets the CASP requirement for methodological triangulation. The range of data available in the review offered a broad scope of user feelings and perceived usefulness. Still, the demographic data on the reviewers was unavailable, which affected the credibility of the sampling and the representativeness of the findings. The interviews have supported the themes of user burden and identified the requirement for personalised, person-centred design as one of their findings, which has also been confirmed by the wider body of literature on digital mental health adherence. In the evidence set, the handling of data and equipment is also another contributor to methodological strengths and weaknesses. The first is methodological heterogeneity, as studies employed a variety of tools, including customised just-in-time adaptive interventions such as Mello and commercially available applications like Smiling Mind or Mindshift. In Studies such as Marshall et al. (2021) and Topper et al. (2017), the credibility of causal inferences was improved through the application of advanced statistical methods, including time-series analysis and mediation models, respectively. However, the different levels of technical complexity across the studies highlight the importance of uniform standards for data quality, particularly when handling sensitive mental health data.
3.2 Thematic analysis
3.2.1 Study table
| Study Number | Author’s name | Sample Population | Method Type |
| Study 1 | Marshall et al. 2021 | 39 participants (27 female, 12 male); Mean age: 34.04 years. | Single-Case Research Design (SCED): Digitally enhanced, multiple baseline across individuals. |
| Study 2 | Topper et al. 2017 | 251 participants (83.7% female); Adolescents and young adults with elevated RNT. | Randomised Controlled Trial (RCT): 3-arm (Group intervention, Internet intervention, Waitlist). |
| Study 3 | Valentine et al. 2024 | 15 participants (aged 16–25); 60% women, 27% men, 13% nonbinary. | Qualitative: Semistructured interviews and Thematic Analysis. |
| Study 4 | Balaskas et al. 2022 | 600 app reviews (Phase 1) + 15 user interviews (Phase 2). | Mixed-Methods: Content analysis of reviews and qualitative interviews. |
| Study 5 | Funk et al. 2025 | 365 participants (aged 16–22); Mostly female and highly educated. | Randomised Controlled Trial (RCT): 3-arm, parallel group (Full app, Concreteness app, Waitlist). |
3.2.2 Theme 1: Targeting transdiagnostic risk factors in relation to Repetitive Negative Thinking (RNT)
The initial key theme observed from the collected data of the chosen studies refers to Repetitive Negative Thinking (RNT) as the major transdiagnostic factor, which does not allow sub-clinical levels of anxiety to develop into clinical ones in the age group (18-24) years. When comparing the results of Studies 2, 3, and 5, it is possible to state that mobile health (mHealth) applications have the most significant effect when they become active interruptions of this cognitive cycle, because they do not just track the mood but manage it. Study 2 by Topper et al. (2017) provides the strongest quantitative support for the mediation analysi,s, which further shows that decreases in RNT directly reduced the prevalence of Generalised Anxiety Disorder (GAD) at 12 months of follow-u,pcompared topared with the control (wgroup, ait-list), to 16.0% in the intervention group. It act upon the core of the problem (the process of rumination), not merely by focusing on the reasons of the anxiety. This causal conclusion coincides with the theoretical concept of Response Styles Theory, according to which rumination is a consistent predictor of the levels of depression and anxiety (Ulitzsch et al. 2022). While in comparison, the qualitative findings presented in the work of Valentine et al. (2024) as a supplement to quantitative ones showed that adolescents and young adults who use applications like Mello have documented a shift between passive avoidance and active coping with negative thought cycles. In contrast, these self-reported experiences are in line with the empirical results of Zhang et al. (2022) that internet-based cognitive-behavioural therapy has the potential to yield similar effects to face-to-face treatment of anxiety disorders.
Whereas Study 2 described extensive effectiveness of an internet-based intervention, while, Funk et al. (2025) did not find any significant reduction in symptoms in the general sample with the help of a self-help application; only frequent users who reached a certain threshold of minimum dose benefits. It is important to note that this observation aligns with Marshall et al. (2021) that reported that mHealth tools are much more effective with people with a smaller history of mental illness, and so the subthreshold stage was found to be the best period to implement preventive RNT-based approaches. The engagement crisis identified in Funk et al. (2025) is similar of Eysenbach’s (2005) Law of Attrition, an established phenomenon in eHealth in which the large dropout rates make even theoretically sound interventions appear (Hurmuz-Bodde et al. 2025). In addition to this, Valentine et al. (2024) indicates that gamification can potentially reduce attrition through the improvement of motivation, but the literature by Zakaria et al. (2024) warns that more does not necessarily lead to an equivalent clinical outcome, meaning that a simple increase in usage may not be enough. Academic pressures in the context of a UK university student population, where the issue of user burden may contribute to make the concept of static and self-help modules less efficient compared to dynamic and just-in-time adaptive interventions described in Studies 3 and 4 (Valentine et al. 2024; Balaskas et al. 2024). These adaptive models adapt content to situational cues in real time, and so match therapeutic intensity to times of increased need and may avoid the problem of attrition. As a result, the synthesis of the selected studies above have highlighted that although RNT is a theoretically and empirically based transdiagnostic target, its implementation in scalable and real-life mHealth interventions requires a delicate sense of the engagement dynamics. Future studies can seek to thus query the way of dose calibration of interventions, bringing the adaptive delivery processes together and the incorporation of user-centered design features which safeguard the conflicting requirements of the academic life whilst maintaining the mechanistic emphasis of disrupting rumination and worry.
3.2.3 Theme 2: Personalisation and accessibility as an alternative to traditional therapy
The second theme recognises flexibility, person-centered design (PCD) as a necessary factor of the effectiveness of mobile interventions designed to prevent the growth of anxiety, especially in 18-24-year-old students of universities (Valentine et al. (2024); Balaskas et al. (2022). With regards to the results of the Valentine et al. (2024) that identified a higher-order theme of doing therapy own way, they observed that participants praised the self-directed character of the apps like Mello by offering them autonomy and accessibility unattainable by traditional face-to-face therapy. The participants observed that autonomy allows them to engage in just-in-time support and target their subthreshold symptoms when they occur (Valentine et al. 2024). Nevertheless, Balaskas et al. (2022) provides the required amount of complexity, although the idea of an app is attractive, the interface design and high user burden as perceived after 600 reviews of the app and 15 in-depth interviews are serious obstacles to continued usage of the intervention for anxiety management (Balaskas et al. (2022). The service users clearly required the variation of content, customisation, and individual experience to remain engaged. The absence of these HCD elements will make the app look like a digital burden instead of a preventative tool. The focus on flexibility is critiqued as per existing literature, in that Person-Based Approach by Alhasani and Orji, (2025) argues that the clinical efficacy of digital interventions is secondary to their persuasiveness and its ease of use, in case a student does not deem the tool to be relevant to them personally. They have not reached the minimum dose necessary to have a clinically significant impact on their psychological issue. Besides, the use of digital autonomy as compared to traditional therapy, which is evident in Valentine et al. (2024), can also be justified by Mohr et al. (2011), who identified that digital technologies have the ability to minimise the stigma and accessibility logistics of university counselling. However, the gamification and tracking of progress demands as seen in Valentine et al. (2024) and Balaskas et al. (2022) is worth cautioning under the paradigm of deterrence theory in UX. Although gamification is capable of increasing short-term engagement, Castellano-Tejedor and Cencerrado, (2024) suggest that gamification can lead to the minimisation of serious mental health outcomes and can result in the shallow engagement in which people are driven by the points instead of therapeutic reflection. In the case of a UK university student, the design should compromise between low-friction entry needed by a person in distress and the clinical sophistication necessary to actually prevent the escalation of anxiety. Overall, flexibility is not just an attribute; to prevent clinical intensification, an app should integrate into the ever-changing and often disordered life of a student such as suggested by the Person-centered design principles in Balaskas et al. (2022).
3.2.4 Theme 3: Engagement-outcome relationship
Although applications are only useful when used regularly, the symptoms of anxiety themselves tend to be the greatest obstacle to user engagement. Combining the findings of Funk et al. (2025), Valentine et al. (2024), and Balaskas et al. (2022), it is evident that the effectiveness of mHealth in the age group of 18 to 24 years old individuals is dose-dependent. Study 5 has a rigorous quantitative basis of this, showing that whereas an RNT -based approach to the app did not indicate any significant reduction in symptoms in the general sample, a minimum dose criterion showed that those who did participate frequently reported significant reductions in anxiety symptoms. This implies that the inability of various apps in averting clinical escalation is not the inability of the therapeutic content, but the inability to sustain its use. This gap in utilisation can be explained by the underlying causes disclosed by the qualitative data in Study 3 by Valentine et al. (2024) where the young people said they felt stuck in negative thoughts. The mental effort to engage with an application can become overwhelming during times of high anxiety, and this may build a paradoxical situation where students most require intervention cannot engage in it. Study 4 by Balaskas et al. (2024) combines that by finding user burden as a determining factor; when the app is perceived as a burden, or in the format of a personalised interface, the already stressed students will get rid of it quickly. To overcome these issues, the study 3 and 4 suggest the use of Just-in-Time Adaptive Interventions (JITAIs) and gamification to reduce entry point. This method is correlated with the Person-Based Approach by Birch, (2024), who states that to reduce anxiety escalation a digital tool should be regarded as a persuasive one and convenient at the time of acute distress. Moreover, the study recommends that to avoid the transition to clinical anxiety, it will be necessary to shift away on the current treatment interventions with respect to the use of modules that are static and towards adaptive and low-friction modules that have the ability to approach a student out of the state of ruminating without imposing additional cognitive load on the student. Overall, the Engagement Outcomes helps to emphasise the idea that with UK university students, the effectiveness of digital mental health initiatives does not only depend on the content of the therapeutic process but also on the long-term low-effort usage that is sensitive enough to react to the exact times when anxiety intensifies.
3.3 Substances and Co-morbidities
The findings of the chosen study showed that the ability to prevent clinical anxiety intensification with the help of mobile applications cannot be discussed separately of the complex interdependence of substance use and co-morbidities. Although the studies by Topper et al. (2017) and Marshall et al. (2021) focus on Repetitive Negative Thinking (RNT), the results suggest that the subthreshold anxiety does not appear as a single pathology, but rather is found to be frequently co-morbid with depression, as well as with other affective disturbances. Within the framework of the UK university students, more specifically comorbidity of anxiety and depression, appears as the expectable effect of app efficacy. Marshall et al. (2021) observed that mixed anxiety and depression participants experienced better results as compared to single-diagnosis depression, which indicates that transdiagnostic apps are especially well-adapted to multi-faceted clinical picture almost always observed in student groups. Nevertheless, it is the intersection of mental health care and substance use that creates an additional level of clinical risk, which these digital tools are often unprepared to address. Substance use as a key variable is not directly measured in the studies, but the causes of stuckness and motivational barriers found in Valentine et al. (2024) and the user burden as addressed by Balaskas et al. (2022) are often worsened by substance related co-morbidities. The self-medication hypothesis would apply to a university setting, as students might consume alcohol or other drugs to reduce the physiological arousal of the subthreshold anxiety. This places an acute challenge with the deliberate introspection demanded by applications like Mello in the study by Valentine et al., (2024); when a student happens to be intoxicated in order to evade bad thoughts, they will feel less chances of reaching the minimum dose of cognitive engagement that Funk et al. (2025) deemed suitable. Moreover, the blended-care approach highlighted by Marshall et al. (2021) is necessary in the context of dealing with co-morbidities. Since apps are largely unguided, they might lack the diagnosing capabilities to detect whether an aspect of anxiety is motivated or concealed by drug abuse in a student. This is consistent with previous research done by Kessler (2004) who observed that co-morbidities are significant contributors to the continuation and intensity of mental illness. In turn, although the applications that address RNT offer a more scalable, preventative approach, their effectiveness in a clinical environment in a university can only be achieved with the incorporation of the larger care framework, which can capture the multifaceted interaction between anxiety, depressive symptoms, and possible substance use.
4. Implications for practice
The findings of the study have provided a significant but a complex solution to the research question that the mobile mental health apps are theoretically efficient in avoiding the development of subthreshold anxiety. But their actual performance depends upon user use, clinical characteristics, and the combination of cognitive processes being directed by these apps strictly. Combining the evidence of Topper et al. (2017) and Funk et al. (2025) it can be seen that application developed specifically to tackle the Repetitive Negative Thinking (RNT) can significantly decrease the progression to clinical Generalised Anxiety Disorder, given that the users can reach the minimum dose of interaction; otherwise the therapeutic effect will not be achieved. This discovery has a critical implication on the clinical practice, notably the paradigm toward a blended-care model. According to Marshall et al. (2021), in the case of high-risk students, apps cannot be recommended as a solitary self-help tool; instead, the UK university mental-health services should incorporate such digital interventions into a stepped-care model, where low-intensity, instant support is provided through apps, but there is a clear connection between this type of support and health practitioners, which is where the strategy of Mohr et al. (2011) theory of supportive accountability comes in and proposes that such a strategy can significantly increase the level of engagement (Kwok et al. 2025). Also, comorbidities and substance use should be screened by the practitioners because Valentine et al. (2024) note a type of stuckness in students with complicated clinical histories, which makes the independent digital interventions ineffective when an acute anxiety spike occurs. As far as health education is concerned, it is time that university programs cease to market the wellness apps and create digital literacy that preempts awareness of cognitive processes (Mohanty et al. 2025). Topper et al. (2017) show that educational interventions that show the students how to view RNT as a discrete risk factor that leads to clinical escalation positively rebrand the use of apps as a skill-development activity in the intentional reflection (Valentine et al., 2024) but not a passive mood tracker, which then allows students to value the dose-response relationship of such tools. The education should also focus on the user burden described by Balaskas et al. (2022), where the realistic expectations are established, and they should be viewed as preventative exercises, similar to physical health routines, where a consistent effort is required. Regardless of these innovations, there are still major gaps in the literature of the diversity of user base and the utilisation gap across the demographics. The future studies will consequently need to leave the largely female and highly educated samples of Topper et al. (2017) and Funk et al. (2025), and examine the engagement patterns of male students, as well as students belonging to a minority group. In particular, the research should be done with regard to Just-in-Time Adaptive Interventions (JITAIs); in the same footsteps as Valentine et al., (2024), the ecological momentary assessment studies may help determine whether nudging students in specific situations when anxiety is at its highest point can help overcome motivational barriers that become evident in cases of being stuck. Also, the gamified versus standard interface should be compared at future in order to determine the reality of clinical effects like progress tracking to anxiety or merely superficial interaction as warned by Balaskas et al. (2022). There is also a need of longitudinal research that monitors the effect of alcohol and drug use on UK students in the efficacy of cognitive-based application as the comorbidity usually determines the course of subthreshold symptoms.
5. Conclusion
Overall, this literature review shows that mobile mental health apps provide a statistically significant yet operationally complicated avenue of averting the intensification of sub threshold anxiety in university students. The evidence synthesis helps to confirm that the most effective cognitive strategy in terms of the reduction of the 12-month prevalence of Generalised Anxiety Disorder is the targeting of transdiagnostic mechanisms, namely Repetitive Negative Thinking (RNT) (Topper et al., 2017). Nonetheless, the practical usefulness of these tools is conditioned by the Paradox of Engagement-Outcome, according to which the efficacy of an intervention is absolutely dependent on a minimum dose of its use that is often undercut by the symptoms of anxiety that it is aimed at (Funk et al., 2025; Valentine et al., 2024). The results also point to the fact that the person-centered design and the flexibility are not only aesthetic values but clinical requirements; the apps should provide low-friction and personalised user interfaces to reduce the user burden that contributes to the high rate of attritions in the student population (Balaskas et al., 2022). Furthermore, the effectiveness of such digital interventions depends on individual attributes, with the highest effectiveness observed with the implementation of interventions during childhood and early adulthood among younger groups and as a component of a blended care model that takes into consideration comorbidities and the specifics of the student lifestyle (Marshall et al. 2021). Finally, although the answer to the research question can be regarded as positive in terms of the possibility of mHealth, the existing evidence points to a discrepancy between the efficacy hypothesised and the one observed in practice. The research process of this field demonstrates that the cognitive basics of preventing anxiety are properly developed; the systems of delivery should become more adaptive and inclusive. The way forward in the area should be to close the utilisation gap by implementing the just-in-time interventions and making the digital mental health support more widespread to the whole UK university population to encourage the future engagement and decrease mental health disparities in the long term among the student body, and be more resilient in the future.
\
References
Alhasani, M. and Orji, R., 2025. Promoting stress management among students in higher education: Evaluating the effectiveness of a persuasive time management mobile app. International Journal of Human–Computer Interaction, 41(1), pp.219-241.
Apolinário-Hagen, J., Drüge, M., and Fritsche, L. (2020). Cognitive behavioral therapy, mindfulness-based cognitive therapy and acceptance commitment therapy for anxiety disorders: integrating traditional with digital treatment approaches. Anxiety disorders: Rethinking and understanding recent discoveries, 291-329.
Balaskas, A., Schueller, S.M., Cox, A.L. and Doherty, G., 2022. Understanding users’ perspectives on mobile apps for anxiety management. Frontiers in digital health, 4, p.854263.
Birch, H.A., 2024. Exploring the Acceptability and Feasibility of Interventions to Signpost Students to University Mental Health and Wellbeing Support: A Mixed-Methods Approach (Doctoral dissertation, University of Northumbria at Newcastle (United Kingdom)).
Borjalilu, S., Mazaheri, M. A., and Talebpour, A. (2019). Effectiveness of mindfulness-based stress management in the mental health of Iranian university students: a comparison of blended therapy, face-to-face sessions, and mHealth app (Aramgar). Iran J Psychiatry Behav Sci, 13(2), e84726.
Castellano-Tejedor, C. and Cencerrado, A., 2024. Gamification for mental health and health psychology: Insights at the first quarter mark of the 21st century. International journal of environmental research and public health, 21(8), p.990.
Expert Self Care. (2025). Student Health App (UK) [Web page]. Expert Self Care Ltd. https://www.expertselfcare.com/health-apps/student-health-app/
Funk, J., Kopf-Beck, J., Takano, K., Watkins, E. and Ehring, T., 2025. Can an app designed to reduce repetitive negative thinking decrease depression and anxiety in young people? Results from a randomised controlled prevention trial. Journal of Behavior Therapy and Experimental Psychiatry, 87, p.102014.
Hurmuz-Bodde, M.Z., Jansen-Kosterink, S.M., Hermens, H.J. and van Velsen, L., 2025. Attrition of older adults in web-based health interventions: Survival analysis within an observational cohort study. Journal of health psychology, 30(8), pp.1768-1779.
Kwok, G., Cheung, S.P.Y., Duffecy, J. and Devine, K.A., 2025. Application of the supportive accountability model in digital health interventions: scoping review. Journal of medical Internet research, 27, p.e72639.
Lewis, J., and Stiebahl, S. (2025, April 28). Student mental health in England: Statistics, policy, and guidance (House of Commons Library Research Briefing No. CBP‑8593). UK Parliament. https://commonslibrary.parliament.uk/research-briefings/cbp-8593/
Marshall, J.M., Dunstan, D.A. and Bartik, W., 2021. Smartphone psychological therapy during COVID-19: a study on the effectiveness of five popular mental health apps for anxiety and depression. Frontiers in Psychology, 12, p.775775.
McCloud, T., Jones, R., Lewis, G., Bell, V., and Tsakanikos, E. (2020). Effectiveness of a mobile app intervention for anxiety and depression symptoms in university students: randomised controlled trial. JMIR mHealth and uHealth, 8(7), e15418.
Mohanty, S., Satpathy, I., Patnaik, B.C.M. and Chakravarty, D., 2025. Enhancing University-Workforce Collaboration Through Digital Literacy, Technology Integration, and Mental Well-Being Support. In Institutes of Higher Education (IHE) and Workforce Collaboration for Digital Literacy (pp. 125-152). IGI Global Scientific Publishing.
NHS England. (2025). Mental health support teams (Trailblazers) in schools and colleges [Web page]. NHS England. https://www.england.nhs.uk/mental-health/cyp/trailblazers/
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., … and Moher, D. (2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. bmj, 372.
Royal College of Psychiatrists. (2011). The mental health of students in higher education: College Report CR166. Royal College of Psychiatrists. https://www.rcpsych.ac.uk/docs/default-source/improving-care/better-mh-policy/college-reports/college-report-cr166.pdf?sfvrsn=d5fa2c24_2
Stand‑by‑me. (2025). WellMind [Web page]. Stand‑by‑me. https://stand‑by‑me.org.uk/resource/wellmind/
Topper, M., Emmelkamp, P.M., Watkins, E. and Ehring, T., 2017. Prevention of anxiety disorders and depression by targeting excessive worry and rumination in adolescents and young adults: A randomised controlled trial. Behaviour research and therapy, 90, pp.123-136.
Ulitzsch, E., Pohl, S., Khorramdel, L., Kroehne, U. and von Davier, M., 2022. A response-time-based latent response mixture model for identifying and modeling careless and insufficient effort responding in survey data. psychometrika, 87(2), pp.593-619.
Valentine, L., Arnold, C., Nicholas, J., Castagnini, E., Malouf, J., Alvarez-Jimenez, M. and Bell, I.H., 2024. A Personalised, Transdiagnostic Smartphone App (Mello) Targeting Repetitive Negative Thinking for Depression and Anxiety: Qualitative Analysis of Young People’s Experience. Journal of Medical Internet Research, 26, p.e63732.
World Health Organization. (2025, September 1). Mental health of adolescents [Fact sheet]. World Health Organization. https://www.who.int/news-room/fact-sheets/detail/adolescent-mental-health
Zakaria, A.L.J., Anas, B. and Malki, M.O.C., 2024. A Systematic Review of Gamification in MOOCs: Effects on Student Motivation, Engagement, and Dropout Rates. Journal of Educators Online, 21(2), p.n2.
Zhang, W., Du, Y., Yang, X., Wang, E., Fang, J., Liu, Z., Wu, S., Liu, Q. and Hu, Y., 2022. Comparative efficacy of face-to-face and internet-based cognitive behavior therapy for generalised anxiety disorder: a meta-analysis of randomised controlled trial. Frontiers in Psychiatry, 13, p.832167.
Appendices
Appendix A Prisma Flowchart.
Appendix B Data Extraction table
