Anxiety disorders are prevalent and heterogeneous. They include generalized anxiety—characterized by excessive worry; panic disorder—characterized by an exaggerated physical and emotional fear response; and the various phobias of social interaction (social phobia), of being separated from a caregiver (separation anxiety), of speaking (selective mutism), and other specific phobias. In this article, I will summarize the latest thinking on the use of psychotherapy for childhood anxiety. Because recent issues of CCPR have explored OCD and PTSD in detail, these will not be included in this review.

While there are many different anxiety disorders, the mainstay of treatment for nearly all of them is the same: cognitive behavioral therapy (CBT). The wealth of empirical evidence supporting the use of CBT for childhood anxiety disorders is impressive. A recent Cochrane review documented a 56% remission rate, compared to 28% for controls, for all of the previously mentioned anxiety disorders with the exception of simple phobias, which were not included in the review. The number needed to treat (NNT) was three. (For a reminder on NNT, see the article on psychopharmacology in this issue.) These results were consistent across all CBT formats, including individual psychotherapy, group-based interventions, and protocols that include significant family involvement (James A et al, Cochrane Review 2005;Issue 4). Various meta-analyses have come to the same conclusion. In fact, the term “empirically supported treatment” is virtually synonymous with CBT when it comes to the treatment of childhood anxiety (Silverman W et al, J Clin Child & Adolesc Psychology 2008;37(1):105–130).

Despite its singular position among empirically-supported treatments, there are a wide range of CBT protocols for children. The most popular of these is probably the Coping Cat, developed by Philip Kendall at Temple University, which is used in an individual, group-based, and family format for social phobia, generalized anxiety disorder, and separation anxiety disorder. This 16-session method offers a mix of psychoeducation and “real life” practice to help kids recognize and appropriately respond to their anxiety triggers.

Because of the differences among the anxiety disorders, protocols vary and typically include elements that address the particular characteristics of each disorder. For example, the treatment of generalized anxiety has more of an emphasis on cognition, the treatment of phobias center on exposure, the treatment of social phobia contains social skills practice, and the treatment of panic disorder contains a heavier dose of psychoeducation about the physiology of panic symptoms. Still, all CBT protocols contain the same basic elements: psychoeducation, cognitive examination and restructuring, exposure, and relapse prevention.

End of story? Not quite. A closer look at the CBT research does raise at least one question: is CBT a singularly effective treatment for anxiety or does it merely have the good fortune to be the subject of multiple, well-designed studies? Moreover, do we really know how CBT holds up against alternate treatments?

A striking feature of the 2005 Cochrane review is that all included studies but one used a wait-list control (the remaining study had control subjects keep diaries): none were compared with an active psychotherapy. Surprisingly, studies including active controls are hard to find.

One recent study compared CBT to “usual care”: a group that received an eclectic set of treatments including psychodynamic, client-centered, and family-based therapy. This very interesting study not only randomized youths with various anxiety disorders to the two treatment arms, but also randomized the therapists, providing specific training in the CBT protocol to those therapists assigned to that group.

The findings? Both treatments were equally efficacious—CBT did not outperform usual care on any measure. Both groups did well, with 67% remission in the CBT group versus 74% in the usual care group for treatment completers (both numbers dropped about 10 percentage points when treatment dropouts were factored into the analysis). There were also no differences between the groups in the use of psychotropic medications (Southam-Gerow M et al, J Am Acad Child and Adolesc Psychiatry 2010;49(10):1043–1052). This study brings up the possibility that CBT is helpful in a non-specific way. There have been few studies evaluating non-CBT therapies for childhood anxiety.

Studies of psychodynamic treatments for anxiety are sparse. A retrospective chart review was conducted at the Anna Freud Centre in London. While these cases included a diagnostic mix of children with various internalizing disorders, the authors concluded that about 75% of children who participated in treatment for at least six months showed improvement, with simple phobias the most likely to remit (Target M and Fonagy P, J Am Acad Child and Adolesc Psychiatry 1994;33(3):361–371).

A more recent study compared an 11-session structured psychodynamic psychotherapy protocol to community care, a heterogeneous group that received either no treatment, individual psychotherapy at another facility, or school-based services. The effects for anxiety disorders are again somewhat difficult to tease out because the children had anxiety, depression, or oppositional defiant disorders. However, the treatment group showed more improvement on the internalizing subscale of the CBCL (Child Behavior Scale) after six months and again after two years, as well as fewer externalizing problems or need for additional treatment (Muratori F et al, J Am Acad Child Adolesc Psychiatry 2003;42:331–339). These results are difficult to compare with the CBT studies, most of which use more diagnostically focused patient populations and more anxiety-specific outcome measures.

Unlike psychodynamic psychotherapy, PCIT (parent-child interaction therapy) has strong empirical support as a treatment for externalizing disorders such as ADHD, ODD, and CD, in young children. In this technique, parents are taught the principles of play therapy in addition to authoritative and positive parenting techniques by receiving remote coaching through a wireless earphone from the therapist who watches from behind a two-way mirror. To date, only small pilot studies exist, one for separation anxiety (Choate M et al, Cog Behav Pract 2005;12(1):126–135) and the other for a heterogeneous anxiety disorder sample (Comer J, J Anx Dis 2011;online ahead of print). Both studies found benefit for PCIT, although neither included a control group. However, in reading these studies more closely, the original PCIT treatment has been modified to include CBT elements including psychoeducation and exposure, after early work found a lack of benefit when these components were absent.