Editorial: taking our time – a long-term perspective on child and adolescent mental health provides invaluable insights for scientists, clinicians and policy makers


  • David Coghill

  • … studies can not only provide strong support for therapeutic interventions but also challenge popularly held scientific theories and political beliefs

The need to take a developmental perspective that embraces change over time at both the individual and group level is a key reference point for all involved in child and adolescent mental health. However, for most of us, our lives are dominated by the present. Time pressures are inevitable and something that most of us struggle with on a daily basis. The academic needs to meet the deadline for the next grant application (or editorial!), to recruit subjects into their studies, and of course publish the results of these studies as quickly as possible in high-impact journals. As the recognition of mental health problems in children and adolescents increases, so do the referral rates to specialist child and adolescent mental health services. Clinicians are often required to provide increasingly sophisticated evidence-based treatments to increased numbers of children and young people, but without increased staffing levels. As politics becomes increasingly dominated by the media, it often feels as if policy decisions are driven by the culture of the sound bite and the need for instant, rather than long-term, results. As a consequence of these constant time pressures we can often resort to a kind of short-termism, with the researcher conducting compact cross-sectional studies, the clinician focusing on short-term treatment outcomes, and policy makers on the instant fix without due recourse to the long-term picture. Several articles in this month’s journal firmly buck this trend and illustrate, the important insights that can be gained by taking time to look at the long-term perspective, and that such studies can not only provide strong support for therapeutic interventions but also challenge popularly held scientific theories and political beliefs.

Using data from the well-known Great Smoky Mountains Study, Sterba and colleagues demonstrate how longitudinal data can be used to test, and ultimately challenge, one of the key theoretical concepts of developmental psychology – the differentiation hypothesis. This hypothesis proposes that development involves differentiation from a relatively unsophisticated and undifferentiated state in infancy, with fewer underlying latent dimensions, towards increasing differentiation with clearer boundaries and separation between different aspects of functioning, such that by late adolescence/early adulthood there are several clearly defined, and relatively uncorrelated, dimensions. The differentiation hypothesis has been used to explain the particularly high rates of comorbidity observed among mainly differentiated disorders in early childhood compared to adolescence and adulthood as a consequence of increasing differentiation of psychopathology across development, from an almost undifferentiated state in infancy, which could be either unidimensional or with several different but highly correlated dimensions of psychopathology, to a fully differentiated and multidimensional state in late adolescence where the various dimensions recognised in adulthood (e.g., depression, anxiety, antisocial behaviours) are present, clearly separable, and relatively uncorrelated with each other. While seductive in its logic and with considerable evidence to support it, the (increasing) differentiation hypothesis is, however, far from confirmed and other pathways are possible. For example, there could be no differentiation in infancy and this could continue into adolescence, or there could be an early differentiation that remains stable over time. Several studies that have combined data across cross-sectional studies have provided some preliminary evidence for increasing syndrome differentiation with age for the disruptive disorders; there is less support within the emotional disorders. However, as Sterba and colleagues point out, the considerable limitations associated with the cross-sectional comparisons mean that firm conclusions are difficult to make. The strengths of the Great Smoky Mountain Study data set lie not only in its size, duration of follow-up and impressive retention rate, but also the breadth of information available at each time point. Together these allowed the study team to use a sophisticated modelling framework to investigate the latent structure of psychopathology across development and test for stability of the model and of the interactions between the latent variables over time. Their findings, that by 9 years of age psychopathology is already relatively well differentiated, with evidence for separable domains of social anxiety, separation anxiety, oppositional defiant conduct, attentional deficit/hyperactivity syndromes, and a unidimensional syndrome of depression and generalised anxiety, and, maybe more importantly, that this pattern remains relatively stable over time, provide a strong challenge to the differentiation hypothesis. From a clinical perspective the finding of a single general anxiety/depression syndrome should have particular resonance and supports the development of integrated care pathways for managing emotional disorders in children and adolescents.

This resonates with the study of Saavedra and colleagues into the long-term outcomes of cognitive behavioural treatments (CBT) for children with anxiety disorders. Few clinicians now doubt the short-term efficacy of CBT in childhood but far less is known about the long-term outcomes of these treatments into late adolescence and early adulthood. Also, few studies have investigated whether treatment aimed at reducing anxiety symptoms also has an impact on other symptoms and problem behaviours such as depression and substance misuse. Although there are limitations in the data (particularly the lack of a non-active comparison group), their findings, that 8–13 years after treatment there was long-term remission for anxiety disorder and targeted anxiety symptoms, as well as reductions in depressive and substance use disorders, are striking. Also important is the finding that there were more similarities than differences between the long-term outcomes for individual and group approaches or between contingency management and self-control approaches. These data will obviously be viewed positively by clinicians and should support the continuing development of skills and increase of service capacity to provide CBT for children and young people with anxiety disorders. They should also be taken seriously by policy makers. Although the situation is slowly improving, it is still the case that many children and young people with emotional disorders are denied the opportunity to engage in structured CBT owing to a shortage of trained staff to provide such therapies. While investment is difficult during periods of economic instability, it may be a false economy to deny access to treatments that are effective in the long term, especially where these can also prevent the development of future difficulties.

Policy makers and politicians will also be interested in the third and final study to be highlighted in this editorial. Collishaw and colleagues describe results from the Youth Trends study. They demonstrate that while there has been a clear increase in emotional problems during adolescence over a 20-year period (1986–2006), these increases were not explained either by ‘the disintegration of family life’ or an increase in ethnic diversity. While parental divorce and separation can clearly have important negative impacts on children and young people, so can the experience of continuing to live in a family where hostile and critical emotions predominate. It is, however, sometimes too simplistic to blame increases in mental health problems over time on increased rates of divorce or separation or ethnic diversity. In the light of these data a simple sentiment that ‘a return to family values’ would result in a reduction in the mental health problems of children and young people appears rather naïve and simplistic. Politically, such strategies are appealing as they win the hearts of the people; however, enshrining such strategies in policy without fully understanding the evidence is unlikely, in itself, to achieve the long-term goal of significantly reducing the burden of poor mental health on children and young people. There is also the danger that State-sponsored funders of research will be more reluctant to invest in studies designed to identify the, as yet unknown, causes for the observed increases in mental health difficulties that have occurred in the recent past. The situation is almost certain to be much more complex and require a finer-grained measurement to identify which environmental factors are indeed responsible for these changes. It remains possible that family factors such as relationship or communication patterns between parents and their children or increased use of technology within the home will be important. It is also possible that community factors such as educational expectations or community cohesion play a key role in the observed increase in reported symptoms; clearly all of these possibilities need to be explored.

Taken together these studies emphasise the importance, and benefits, of taking a long-term view of mental health problems in children and adolescence. Such studies are expensive, complex to design properly and extremely difficult to execute effectively; however, when these three factors can be successfully negotiated they have huge potential to move the field forwards and impact on our understanding at the scientific, clinical and policy levels.