Editorial: The role of digital technology in children and young people's mental health – a triple-edged sword?

Abstract

The rapid expansion of access to, and engagement with, the Internet and digital technology over the past 15 or so years has transformed the social, educational and therapeutic space occupied by children and young people in contemporary society in remarkable ways. First, it has created previously unimaginable opportunities for learning and development and personal exploration and growth. Second, it seems that the very same qualities and characteristics of the Internet that make these positive contributions possible, such as its immediacy, portability, intimacy, unconstrained reach and lack of supervision and regulation of content, has opened children and young people up to a range of serious social, intellectual and mental health risks. Finally, over and above these 'effects', the digital space is increasingly successfully being harnessed for the identification and treatment of mental health problems. Accordingly, the Internet is not so much a double-, as a triple-edged sword, with regard to children’s mental health.

The rapid expansion of access to, and engagement with, the Internet and digital technology over the past 15 or so years has transformed the social, educational and therapeutic space occupied by children and young people in contemporary society in remarkable ways. First, it has created previously unimaginable opportunities for learning and development and personal exploration and growth. Second, it seems that the very same qualities and characteristics of the Internet that make these positive contributions possible, such as its immediacy, portability, intimacy, unconstrained reach and lack of supervision and regulation of content, has opened children and young people up to a range of serious social, intellectual and mental health risks. Finally, over and above these ‘effects’, the digital space is increasingly successfully being harnessed for the identification and treatment of mental health problems. Accordingly, the Internet is not so much a double-, as a triple-edged sword, with regard to children’s mental health.

The Journal of Child Psychology and Psychiatry (JCPP) has previously made substantive contributions to the discussion of the triple-edged nature of the digital revolution for children‘s mental health and well-being through the publication of authoritative reviews. Recognising the relative infancy of this field these reviews, in part, represented something of a rallying call for researchers in the field to step up and conduct better science – to develop new ways of theorising the role of the digital environment in children and young people’s mental health and undertake studies with greater methodological rigour and power to test those theories using sophisticated designs. In judging that the field has reached a level of sufficient productivity and scientific quality required for us to publish Young People’s Mental Health in the Digital Age – From Understanding Risk and Resilience to Promoting Therapeutic Innovation Potential as the first of two JCPP special issues in 2020, the JCPP editors have given the field a big vote of confidence. As you can see from the eight papers included in this special issue, they will not be disappointed.

Conceptualising and studying digital risk and resilience

The first five of the papers focus on digital risk and resilience in children’s mental health. It is now some years since Livingstone and Smith (2014) reviewed this literature in their 2014 JCPP Annual Research Review (ARR). In that period, social media in particular have become not only commonplace but almost infrastructural in everyday life – taken for granted as the routine means by which family and peer relationships are sustained, identities expressed and networks built. The same period has seen rising public anxiety, resulting in a series of policy reports regarding children and young people’s mental health, personal safety and well-being in what is fast becoming, especially during COVID-19, a thoroughly digital world. Recognising that not all digital risks need necessarily result in actual harm, the hunt is on to identify the key protective and risk factors that exacerbate existing vulnerabilities and enable young people’s resilience. As Livingstone and Smith summarised in 2014, the evidence regarding vulnerability to aggressive and sexual risks pointed to ‘personality factors (sensation-seeking, low self-esteem, psychological difficulties), social factors (lack of parental support, peer norms) and digital factors (online practices, digital skills, specific online sites)’ (p. 635). For the most part, these are familiar (and sadly often intractable) causes of vulnerability, centring on problems with family and peers, in turn often linked to socio-economic or other forms of disadvantage, though personality factors also contribute.

Strikingly, in that review, and in others conducted since (see Odgers & Jensen, 2020), such ‘offline’ (or nondigital) problems are shown to be the largest predictors of children’s online experiences of risk and harm and yet most of the variance in these experiences across children remains unexplained. On the one hand, research must continue to examine how longstanding sources of vulnerability manifest online, as the digital environment becomes ever more established in young people’s lives. On the other hand, it should explore the tantalising prospect of discovering a role for distinctively ‘online’ factors, such as distinctive online practices, or the design features of particular digital sites or services. These, arguably, could become the focus of policy or professional intervention and redesign. While this prospect is being actively explored (as discussed below), in the interim it is often left to online friends or communities to harness the affordances of platforms in ways that provide support for vulnerable young Internet users.

With this in mind, Lavis and Winter (2020) explore the ambiguous world of online peer-to-peer interactions around self-harm. While the instinctive reaction of adults, from parents to politicians, is to try to shut down online sites where self-harm is discussed (fearing social contagion or copycat problems), online ethnographic methods and qualitative interviews reveal that the young people who engage with these discussions not only tend to be already self-harming (implying that the causes lie elsewhere), but also that online peers are primarily supportive. As Odgers and Jensen (2020) found in their ARR, the evidence for online activities causing or worsening young people's mental health is both minimal (in the sense of identifying very small effect sizes) and contested (because of a lack of adequate statistical controls and an insensitivity to differences among youth, mental health conditions and digital services). However, as they further consider, the question of whether digital technologies can provide new opportunities for support (either from a community of peers, trained professionals, or even particular affordances of site design) is now pressing. Lavis and Winter advance just this line of thinking through their ethically sensitive analysis, showing how seemingly small acts – sending emojis for emotional support, commenting in ways that offer recognition not judgement, checking in on community members regularly – are recognised as part of an ethic of care. Yet this comes at an emotional cost to the participants, and while it may possibly prevent an escalation of harm, it may equally hold the community together and impede members’ recovery. Moreover, the role of the platforms (in this case, Instagram, Reddit and Twitter) in algorithmically escalating the reach of more extreme images and messages must not be underestimated, hence, the importance of ongoing developments in platform regulation against online harms.

Not only peers but also parents may play a supportive role for teenagers, now that family relationships are increasingly mediated. Just as peers’ online influence is popularly seen as problematic rather than supportive, Modecki et al. (2020) observe that parents’ use of digital devices is also widely criticised as poor role modelling or as distracting them from actively parenting. Their large online survey of diverse parents found that, especially in families where ‘technoference’, as they term it (meaning technology use specifically linked to family displacement or conflict) was fairly low, parents’ smartphone use was, contra popular expectations, linked to higher parenting quality and warmth. Meanwhile, the findings lead the authors to suggest that parents who do not describe themselves as high in warmth may even be better off displacing family time with their smartphone use than not doing so. Only where parental smartphone use was high – described as immersive – did the findings suggest a negative impact on parent–child relations. As before, further research is needed to identify the deeper causes of such differential family patterns.

Perret et al. (2020) advance our understanding of the potentially causal relation between young people’s online activities and mental health using a longitudinal design. Their prospective birth cohort study included measures at ages 13, 15 and 17 of cybervictimisation, face-to-face victimisation, suicidal ideation and suicide attempts. Even with their initial sample size of 2000+, the Ns for those reporting suicidal ideation or attempts are small, so the findings should be treated with caution. Using cross-sectional analyses, they found that both cybervictimisation and cybervictimisation combined with face-to-face victimisation were correlated with suicidal ideation/attempts, after controlling for mental health, family and other circumstances including prior face-to-face victimisation. However, any adverse effects of cybervictimisation appear to be short-term only, far more notable was the finding that only face-to-face victimisation was associated with suicidal ideation/attempt two years later. Reflecting on the cross-sectional findings, it would seem that cybervictimisation can compound the adverse effects of face-to-face victimisation; as is often said, the effect is that the victim cannot easily escape their persecutor even at home or in other supposedly safe spaces for, via digital technologies, they can be reached 24/7. Why then is it face-to-face victimisation only that can predict later suicidal ideation or attempts? The authors suggest this to be more chronic in nature, compared with cybervictimisation (which they suggest is less repeated, though it may be intense in the moment). Presumably too, although the authors controlled for a range of circumstantial factors, there may be other predictors of face-to-face victimisation yet to be discovered and these, in turn, may account for later suicidal ideation or attempts. Again, the conclusion appears to be that digital technologies can amplify or intensify both risk and protective factors, but that the root causes of young people’s mental health difficulties tend to lie deeper, most likely in their (nondigital) personal circumstances.

One of the key messages from contemporary child psychology and psychiatry is that children’s behaviour and development is heavily influenced by variations in sociodemographic and cultural processes operating within social eco-systems. Yet time after time we, as researchers and policymakers, continue to act as if we can generalise effects found in wealthy Western countries (especially Europe and the United States) to all other settings irrespective of these factors (Sonuga-Barke, 2014). That seems to have certainly been the case in studies of digital risk and resilience. The paper by Kardefelt-Winther et al. (2020) tackles this head on by reporting new data from the Global Kids Online network. They compared the link between time online and mental health in four culturally and socio-economically different countries – Bulgaria, Chile, Ghana and the Philippines. The study sends out an important cautionary message to researchers and policymakers alike loud and clear – do not generalise findings about digital risk from one country or cultural group to another. True, there was a clear, though small, negative association between time online and well-being in two countries (Bulgaria and Chile – perhaps the more socio-economically advantaged in the study) as often seen in US studies. However, there was a more ambiguous finding in a third country (Ghana) and little or no effect in a fourth (Philippines). Just as importantly – in all countries nondigital family and relationship factors seemed to have a much more consistent effect on children’s well-being than levels of digital exposure. It will be fascinating to track secular trends in the effects of digital engagement in these countries as levels of digital usage increases (between cohorts) as well as how things change for the children studied as they grow (within cohort).

With the paper by Wold Hygen et al. (2020), we shift conceptual gears by moving from looking at a person’s interactions with, and exposure to, the digital environment as either creating risk to, or providing protection for, their mental health – to considering those digital-related behaviours as signs and symptoms of mental disorder itself. The authors here focus on what has been termed in a prototype DSM-5 formulation as internet gaming disorder (IGD): a disorder characterised by an addiction-like obsession with gaming that disrupts everyday functioning and negatively impacts social relationships. The authors do not directly address the question of whether IGD is a distinct mental disorder. Rather they explore the reasons for its oft observed association with other mental disorders. Using sophisticated cross-lagged analysis of longitudinal data gathered from a large general population sample at ages 10, 12 and 14 years their study found: (a) small concurrent correlations between IGD symptoms and psychopathology (consistent with past reports) – especially with externalising (ADHD/conduct problems) and anxiety; and (b) no evidence that IGD symptoms prospectively drove increased risk for later psychopathology. In fact, there was some evidence the IGD might even be protective when it comes to anxiety – perhaps, as the authors suggest, it reduces worry/rumination or facilitates online friendships for shy people. More generally, the authors see their data as supporting the idea that the manifest associations between IGD and other psychopathologies are the result of shared underlying causes – such as genes. At first sight, one might think that this sort of conclusion undermines the distinctiveness of the IGD concept – that perhaps IGD is just a contemporary digital expression of established disorders (i.e. anxiety, ADHD) and not a new disorder at all. However, this is not necessarily the case as we now know that diverse disorders with very different clinical presentations, not just IGD, share aetiological elements yet retain clinical utility (Posner, Polanczyk & Sonuga-Barke, 2020). The clinical value of the IGD concept in terms of helping to understand its course and treatment outcome remains to be discovered.

Harnessing digital technology to intervene to promote better mental health

Three papers in the special issue illustrate how the research community has risen to the challenges for digital mental health intervention science set out by Hollis et al. in the 2016 JCPP Annual Research Review (Hollis et al., 2016). He and his colleagues have argued that of all fields within health care, the opportunity for digital transformation in child and adolescent mental health is probably the greatest (Hollis et al., 2015). This reflects both the high level of digital engagement in young people’s daily lives and the enormous potential to transform mental health services through improved access to evidence-based digital resources and interventions, and by automating parts of diagnostic, monitoring and treatment pathways. Digital mental health interventions and tools have the potential both to close the mental health treatment gap for children and young people and, importantly, to go beyond simply replicating traditional face-to-face interventions online by providing interventions that are more effective, more accessible and more cost-effective. However, to date this potential is largely unrealised (Hollis et al., 2016). Few well-researched academic-led digital interventions are included in routine care and poor uptake, and adherence is seen in those that are. In contrast, industry-led innovations, while delivered at scale, often lack a research evidence-base and youth codesign (crucial to ensure that products fit with needs and lifestyles, and to tackle nonadherence). Effective, usable and accessible digital innovations could close the mental health treatment gap and ensure evidence-based mental health interventions reach young people who need them most (Hollis et al., 2018).

The paper by O’Dea et al. (2020) addresses a number of the issues highlighted in the 2016 ARR. First, they studied a single session relationship-focused intervention for adolescents delivered on a mobile phone app. This represents an important advance from the majority of ‘first-generation’ digital health intervention studies which have largely focussed on traditional 10 to 12 session cognitive behavioural therapy (CBT) interventions delivered in a web-based format usually not accessible on mobile devices which the majority of young people use. Although depression (the primary outcome) did not improve ratings of welling and helping seeking intensions did increase. Second, a major strength of the study was that all recruitment and assessment were conducted online and that a randomised controlled trial (RCT) of the app was completed in just six months. The use of online recruitment (in this case through Facebook) enabled this sample of Australian youth to be recruited in less than 10 weeks. Given the rapid pace of change in digital technologies and the cost and duration of traditional RCTs, some have argued that RCT methodology is no longer relevant to study digital interventions. This paper helps to refute this argument by providing an example that RCTs of mental health apps are feasible and can be conducted quickly and efficiently. Third, the paper raises tantalising questions about how to ensure that digital interventions are accessible and appealing to the whole population of youth, including those who experience digital poverty and are often those most at high risk of mental health problems.

The paper by Spence et al. (2020) tackles the issue of personalisation or ‘precision’ digital interventions. Understanding the heterogeneity of therapeutic response by identifying factors that interfere with or promote successful treatment outcome may enable researchers and clinicians to adapt and enhance interventions in order to increase the proportion of young people who benefit from them. Spence et al. examine which factors (including family functioning) predict response to unsupported digital CBT for childhood anxiety. Despite this study identifying very few predictors of treatment response for therapist-assisted iCBT (only parent relationship quality and maternal age were significant negative predictors), this line of research remains important for tailoring and optimising interventions such as self-help iCBT which may be more easily scalable, but typically have higher rates of dropout and lower effect sizes from treatment, than supported digital interventions.

Finally, the paper by Bevan Jones et al. (2020) provides a detailed description of the methods and processes of ‘coproduction’ and youth codesign for digital interventions. While the concepts of youth coproduction and codesign are increasingly discussed and viewed as essential to ensure interventions fit young people’s needs and lifestyle and are used – these concepts are rarely defined or operationalised specifically in relation to mental health interventions. This practitioner review should assist both researchers unfamiliar with implementing and reporting these methods, but also reviewers evaluating how well principles of coproduction have been adhered to. Their paper describes a range of approaches to involve children and young people and other stakeholders in the codevelopment of technologies throughout the research cycle. The review should prove invaluable to practitioners/researchers interested in the codesign of digital technologies for children and young people and sets a new benchmark for how technologies of high quality are developed and how digital intervention studies report the process of youth codesign.

In summary

These papers in this special issue are a testament to how high-quality science is essential if we are to challenge and test the veracity of the powerful cultural myths about the impact of the digital environment on children and young people’s mental health (its cost, benefits and therapeutic promise) that are circulating unchecked in contemporary society. They also illustrate brilliantly what will inevitably emerge over the next few years as more sophisticated theories and conceptualisations are developed and more powerful research methods and designs emerge. The relationship between digital life and mental health is best characterised by a complex mix of positive and negative influences varying over time both within and between individuals – conditioned and moderated by personal characteristics and cultural, historical and socio-economic factors. It will be the urgent job of the next generation of research to parse this complexity and heterogeneity to identify new and innovative ways to reduce risk, increase resilience and exploit digital therapeutic opportunities.

These papers in this special issue are a testament to how high-quality science is essential if we are to challenge and test the veracity of the powerful cultural myths about the impact of the digital environment on children and young people’s mental health (its cost, benefits and therapeutic promise) that are circulating unchecked in contemporary society.

Acknowledgements

E.S.B is the Editor-in-Chief of the Journal of Child Psychology and Psychiatry. He has visiting chairs at Aarhus University in Denmark. He is also a consultant for Neurotech solutions, Aarhus University, Copenhagen University, Berhanderling, and KU Leuven. Further, the author is a member of the speakers’ advisory groups for Shire Pharmaceuticals and has received speaker’s fees from numerous scientific organisations. Grants: Shire Pharmaceuticals; MRC; ESRC; EU; Wellcome Trust; DoH/NHS; Nuffield Foundation; Fonds Wetenschappelijk Onderzoek-Vlaanderen (FWO); MQ – Transforming Mental Health. All authors contributed equally to this editorial. The authors have declared that they have no potential or competing conflicts of interest in relation to this editorial. The authors’ shared membership of Nurture Network (ES/S004667/1) has faciliated this editorial collaboration. C.H. is a member of the NIHR Mindtech/Medtech Coperative and the NIHR Nottingham BRC. E.S.B. is a member of the SLaM NIHR BRC.