Introduction
Internalizing disorders of childhood and adolescence are serious problems. While conditions such as anxiety and depression were once viewed as benign, transient, or even barely-existent, they are now recognized as common disorders that need identifying and treating promptly. This review will cover some of the most important developments in our understanding of anxiety and depression that have taken place in recent years. Since the study of these disorders in younger populations has tended not to be disorder-specific, most of what we describe is relevant to all types of depression or anxiety. Where research is focused on a specific disorder, this is made clear.
Only since the 1980s (Kovacs, Gatsonis, Paulauskas and Richards, Reference Kovacs, Gatsonis, Paulauskas and Richards1989) have these conditions received serious scientific scrutiny, and only very recently have outcome researchers published treatment trials that conform to CONSORT criteria. Much of the emerging research is from a cognitive-behavioural perspective, and published NICE guidelines recommend CBT for depression in children and for adult anxiety (there are currently no child anxiety guidelines) (National Institute for Health and Clinical Excellence, 2007; National Institute for Clinical Excellence, 2005). This article will outline what we know about anxiety and depression in children and adolescents, what is becoming understood, and what now needs to be done.
Early research
The first trials of CBT for anxiety emerged in the 1990s, with the pioneering work of Philip Kendall's group. Their first trials used basic CBT techniques that were well-established with adult clients. The programme consisted largely of behavioural elements, particularly exposure and relaxation. The cognitive element consisted mainly of “coping self-talk”, wherein the client repeated coping statements during exposure. In these initial trials, no parental component was included. Despite this, results were promising, and subsequent trials have extended these good results by adding components. In particular, over the past decade the cognitive element has been upgraded to include identification and challenging of unhelpful cognitions. A number of creative ways of teaching young clients these techniques have evolved, and trials of CBT for childhood anxiety now have success rates of around 60% (Cartwright-Hatton, Roberts, Chitsabesan, Fothergill and Harrington, Reference Cartwright-Hatton, Roberts, Chitsabesan, Fothergill and Harrington2004). Treatment of depression in children and adolescents has followed a similar pathway, with adult models being used to develop treatments, often simplifying components or focusing on behavioural elements, and adding in parental psycho-educative or co-therapist components (Clarke, Hops, Lewinsohn and Andrews, Reference Clarke, Hops, Lewinsohn and Andrews1992; Lewinsohn, Clarke, Hops and Andrews, Reference Lewinsohn, Clarke, Hops and Andrews1990; Wells and Albano, Reference Wells and Albano2005). Earlier treatment trials tended to target adolescents, with later trials including children as young as nine or ten (Asarnow, Scott and Mintz, Reference Asarnow, Scott and Mintz2002; Harrington, Whittaker and Shoebridge, Reference Harrington, Whittaker and Shoebridge1998; Lewinsohn et al., Reference Lewinsohn, Clarke, Hops and Andrews1990; Temple, Reference Temple1997; Treatment for Adolescents With Depression Study (TADS) Team et al., 2004; Wood, Harrington and Moore, Reference Wood, Harrington and Moore1996).
Before long, outcome researchers recognized that parents had an important role to play in their children's treatment, and it is now unusual for treatment not to include parents in some capacity. However, the role of parents in treatment varies tremendously, and there is controversy about their optimal role (Barmish and Kendall, Reference Barmish and Kendall2005). It appears that this depends on a complex array of factors, including, in particular, the child's age, and the parents’ own mental health (Creswell and Cartwright-Hatton, Reference Creswell and Cartwright-Hatton2007).
What is happening now?
The treatments that have been used until recently have been derived, almost exclusively, from adult treatment protocols. Little attention was paid to whether these protocols (or the models on which they were based) were applicable to younger populations. In recent years this work has begun and, in addition, some researchers have begun to establish new models that are derived not from adult models, but from novel research with children.
Are adult models and techniques appropriate for children?
Research is beginning to examine whether the premises on which adult CBT is based, and the techniques that it employs, are appropriate for younger clients. For example, in anxiety, key cognitive behavioural processes, such as attentional biases towards threat, have been demonstrated in children (Hadwin, Garner and Perez-Olivas, Reference Hadwin, Garner and Perez-Olivas2006). More specifically, key tenets of major adult models, for example OCD (Salkovskis, Reference Salkovskis1985), Social Phobia (Clark and Wells, Reference Clark, Wells, Heimberg and Liebowitz1995), and GAD (Wells, Reference Wells1995) have been explored in younger populations. For adolescents at least, it seems that the key principles of these models are applicable. For instance, the responsibility and thought-action-fusion components of Salkovskis’ OCD model seem to be applicable for adolescents and older children (Libby, Reynolds, Derisley and Clark, Reference Libby, Reynolds, Derisley and Clark2004; Mather and Cartwright-Hatton, Reference Mather and Cartwright-Hatton2004). Key aspects of the social phobia model have also been confirmed, including inappropriately negative evaluations of one's social skills (Cartwright-Hatton, Hodges and Porter, Reference Cartwright-Hatton, Hodges and Porter2003; Cartwright-Hatton et al., Reference Cartwright-Hatton, Tschernitz and Gomersall2005), and negative interpretation of ambiguous social situations (Vine and Stopa, Reference Vine and Stopa2008). Early evidence suggests that cognitive techniques designed for socially anxious adults are also effective in adolescence (Morgan and Banerjee, Reference Morgan and Banerjee2006; Parr and Cartwright-Hatton, Reference Parr and Cartwright-Hatton2008). Finally, aspects of the GAD model, particularly the role of metacognitive beliefs, have been confirmed in adolescence (Cartwright-Hatton et al., Reference Cartwright-Hatton, Roberts, Chitsabesan, Fothergill and Harrington2004). In depressed children, the evidence for the applicability of adult models is sparse but emerging. For instance, there is evidence of a similar over general memory bias to that found in adults (Bennett, Cartwright-Hatton and Stavrinos, Reference Bennett, Cartwright-Hatton and Stavrinos2008; Kuyken, Howell and Dalgleish, Reference Kuyken, Howell and Dalgleish2006), for Beck's negative cognitive triad (Jacobs and Joseph, Reference Jacobs and Joseph1997), for the role of rumination in generating negative mood (Park, Goodyer and Teasdale, Reference Park, Goodyer and Teasdale2004), and for the role of cognitive change mediating mood change (Kaufman, Rohde, Seeley, Clarke and Stice, Reference Kaufman, Rohde, Seeley, Clarke and Stice2005). Needless to say, much of the research has focused on older children and adolescents, and very little is known about these processes in younger children. To our knowledge, there is no published research showing any aspect of an adult model that does not appear to apply to children. However, we must be aware that “negative” results are less likely to be published than “positive” ones.
Developmental and family models of internalizing disorders
Rather belatedly, research is now exploring processes that are not directly derived from adult models of disorders and might be used to inform new, developmentally appropriate treatments. Examples include the work of the University of Reading group, who are exploring intergenerational processes in anxiety disorders. Their early outputs have shown, for example, that socially phobic mothers encourage their infants to interact less with a friendly stranger than mothers with GAD, and this is predictive of the extent to which the child subsequently shows anxiety in the presence of the stranger (de Rosnay, Cooper, Tsigaras and Murray, Reference de Rosnay, Cooper, Tsigaras and Murray2006; Murray, Cooper, Creswell, Schofield and Sack, Reference Murray, Cooper, Creswell, Schofield and Sack2007) and clinical interventions based on these findings are underway. Other examples of pioneering developmentally led work into child anxiety include research by Andy Field, another UK based psychologist, which has produced insights into how children learn to be anxious, in particular by exploring how children use fearful verbal information, or model fearful reactions from others (Field, Reference Field2006). Such work has wide-ranging implications for developmentally appropriate interventions and preventative approaches.
Similarly, the role of the family is now receiving increasing attention. In both the anxiety and depression literatures there is evidence that parents of impaired children are more likely to display overcontrolling or overprotective behaviours, and are more likely to display hostile behaviours (Rapee, Reference Rapee1997). However, there is increasing recognition that parenting is a complex phenomenon, which interacts with many other factors, not least the child's personality. Increasingly, attention is being paid to parents’ cognitions, and this is being woven into new models of childhood difficulties (Creswell, Brewin and O'Connor, in press; Creswell, Schniering and Rapee, Reference Creswell, Schniering and Rapee2005; Field and Cartwright-Hatton, Reference Field and Cartwright-Hatton2008; Stark, Schmidt and Joiner, Reference Stark, Schmidt and Joiner1996). Interventions that take these more sophisticated parenting factors into account are just being explored, and will emerge in the literature in the coming years. There are implications here for both child-focused and adult-focused clinicians, with evidence emerging that treating one family member's difficulties (whether parent or child) can have secondary effects on other family members’ difficulties (Bogels and Siqueland, Reference Bogels and Siqueland2006; Verduyn, Barrowclough, Roberts, Tarrier and Harrington, Reference Verduyn, Barrowclough, Roberts, Tarrier and Harrington2003). Research is now examining the effect on the child of treating parental difficulties (with or without treating the child in parallel), although it will be some time before the outcomes of this research are known (Creswell, Reference Creswell2007).
As models of childhood disorders become more sophisticated, this should have an impact on treatment. However, although we may develop intricate models of the cognitive processes that characterize childhood disorders, do we have the techniques to tackle these? This important question has been neglected. Although outcome researchers include many different techniques in their protocols, they rarely have the complex designs or the statistical power to examine their individual efficacy. We now need basic research into the cognitive capabilities of children across developmental stages, and systematic examination of their ability to benefit from different techniques. This research has barely begun. However, early work from the University of East Anglia (Doherr, Reynolds, Wetherly and Evans, Reference Doherr, Reynolds, Wetherly and Evans2005; Quakley, Coker, Palmer and Reynolds, Reference Quakley, Coker, Palmer and Reynolds2003; Quakley, Reynolds and Coker, Reference Quakley, Reynolds and Coker2004) has begun to unpick the ages at which children can carry out some of the fundamental tasks of CBT. For example, it appears that prepubescent children can often name emotions, generate post-event attributions, and link thoughts, feelings and behaviours. Moreover, these skills may be enhanced when sensitive cueing is used. More of this type of research is now needed. Similarly, although parental cognition is now implicated in the maintenance of internalizing disorders, we have yet to establish that the techniques used to modify adults’ cognitive distortions carry over to situations when the index client is not them but their child. It seems likely that this will be the case, but research to establish this empirically is just beginning.
Recent developments in treatment
Prevention
One recent development is work aimed at preventing internalizing disorders. In contrast to the treatment trial data, which are largely very positive, results from prevention trials have been disappointing. Briefly, the results of recent, well-conducted trials of CBT-based prevention programmes for childhood depression (Merry, McDowell, Hetrick, Bir and Muller, Reference Merry, McDowell, Hetrick, Bir and Muller2007; Sheffield et al., Reference Sheffield, Spence, Rapee, Kowalenko, Wignall, Davis and McLoone2006; Spence, Sheffield and Donovan, Reference Spence, Sheffield and Donovan2005; Spence and Shortt, Reference Spence and Shortt2007) and anxiety (Dadds et al., Reference Dadds, Holland, Laurens, Mullins, Barrett and Spence1999; Dadds, Spence, Holland, Barrett and Laurens, Reference Dadds, Spence, Holland, Barrett and Laurens1997) have not shown substantial benefits when applied in universal settings. The FRIENDS programme for anxiety (Barrett, Farrell, Ollendick and Dadds, Reference Barrett, Farrell, Ollendick and Dadds2006) has produced statistically significant changes in symptoms, but effect sizes are often small, and it is unclear whether resources would be better used to treat children who have already developed a disorder.
New format CBT
Several groups are developing new methods of delivering CBT to children and families, which fit with busy lifestyles and short attention spans. In particular, there is evidence that CBT for anxiety can be delivered to children and their families reasonably effectively in “summer camp” compressed form, by telephone, and even without much therapist input at all, for example by computer or bibliotherapy (Cunningham, Rapee and Lyneham, Reference Cunningham, Rapee and Lyneham2006; Ehrenreich and Santucci, Reference Ehrenreich and Santucci2007; Rapee, Abbott and Lyneham, Reference Rapee, Abbott and Lyneham2006). Self-help books for depression and anxiety (mainly aimed at parents) are readily available, although the effectiveness of this mode of treatment for depression has not been established.
Third wave interventions
The “third wave” therapies that are at the cutting-edge of CBT for adults are attracting interest from a number of child-focused researchers, particularly with the growing evidence of similar meta-cognitive biases and information-processing in adults and adolescents. The appeal of interventions such as mindfulness and attention-training is that they appear to be deliverable in ways that place modest demands on an immature cognitive system. This assumption is yet untested, but several groups are carrying out small trials of these types of interventions with children or adolescents. More established therapy models such as Dialectical Behaviour Therapy, which utilize mindfulness components, have been shown to have significant effects on suicidal ideation and mood in depressed adolescents (Miller, Rathus and Linehan, Reference Miller, Rathus and Linehan2007), and there is some interesting early evidence about the potential effectiveness of Acceptance and Commitment Therapy approaches with depressed adolescents too (Murrell, Coyne and Wilson, Reference Murrell, Coyne, Wilson, Hayes and Strosahl2005).
What needs to happen next?
Although there is now much attention focused toward treatment outcome research, the evidence suggests that we are a long way from universally effective treatment for children with internalizing disorders. A recent systematic review demonstrated that less than 60% of children receiving CBT for anxiety disorders were recovered at the end of treatment (Cartwright-Hatton et al., Reference Cartwright-Hatton, Roberts, Chitsabesan, Fothergill and Harrington2004). The figures for depression are no better (Watanabe, Munot, Omori, Churchill and Furukawa, Reference Watanabe, Munot, Omori, Churchill and Furukawa2007; Wood et al., Reference Wood, Harrington and Moore1996). It is likely that this situation will only improve substantially once child-centred, developmentally appropriate models are built. Furthermore, once clearer models are available, considerable effort will need to go into designing cognitive-behavioural techniques that are appropriate for our young clients and their families.
Comments
No Comments have been published for this article.