Background: Anxiety disorders are common in children, with a prevalence of 10 to 20 percent; however, they are often unrecognized and undertreated. Cognitive behavior therapy (CBT) and selective serotonin reuptake inhibitors (SSRIs) are effective, but up to one half of children will not respond to short-term treatment with either modality. There are few randomized controlled trials comparing these treatments. Walkup and colleagues evaluated the relative effectiveness of CBT, sertraline (Zoloft), a combination of the two therapies, and a placebo.

The Study: The 12-week randomized controlled trial included children and adolescents seven to 17 years of age who had a primary diagnosis of separation anxiety disorder, generalized anxiety disorder, or social phobia. Patients with coexisting diagnoses, such as attention-deficit/hyperactivity disorder, obsessive-compulsive disorder, and conduct disorders, were allowed to participate. Patients were excluded from the study if they had concurrent major depressive or bipolar disorder or psychosis, or if they had not previously responded to two adequate trials of SSRIs or an adequate trial of CBT.

Participants were randomized to receive sertraline alone, CBT alone, a combination of sertraline and CBT, or placebo. The starting dosage for patients receiving sertraline was 25 mg per day. Through the eighth week of the trial, doses were titrated until patients' symptoms improved or until they received 200 mg per day. CBT consisted of 14 one-hour sessions, including training in anxiety-management skills. Anxiety levels were quantified every four weeks using the Clinical Global Impression-Improvement scale, the Pediatric Anxiety Rating Scale, and the Children's Global Assessment Scale.

Results: The authors randomized 488 patients with a mean age of 10.7 years. The three intervention groups had significant improvements in their anxiety compared with placebo. The percentage of patients whose conditions were rated as much improved or very much improved was 80.7 percent in the combination therapy group, 59.7 percent in the CBT group, and 54.9 percent in the sertraline group, compared with 23.7 percent in the placebo group. Combination therapy also yielded higher improvement rates than either monotherapy group. Symptom improvement was statistically equivalent between the sertraline and CBT monotherapy groups. One participant had a serious adverse event (worsening of behavior) that was possibly related to sertraline use. One child in the sertraline group was withdrawn from the study because of nonsuicidal self-harm and homicidal ideation. No children attempted suicide during the study.

Conclusion: The authors conclude that although CBT and sertraline improved symptoms in children with anxiety disorder, combination therapy produced the greatest rate of improvement.