Treatment of Separation, Generalized, and Social Anxiety Disorders in Youths

Separation anxiety disorder, generalized anxiety disorder, and social anxiety disorder (or social phobia) are among the most common psychiatric illnesses in children and adolescents (1, 2). They have an earlier age at onset than other internalizing psychiatric disorders (2), are associated with significant functional impairment, and tend to persist into adulthood (3–6). In general, much of adult psychopathology can present first in childhood and adolescence, with childhood-onset disorders predicting adolescent disorders (7–10) and both child and adolescent problems and diagnoses predicting adult disorders (3, 11–13). The anxiety disorders that present in childhood (3) place children and adolescents at risk of later anxiety disorders, mood disorders, substance use disorders, and disruptive behavior (3, 4, 7).

Separation, generalized, and social anxiety disorders are a heterogeneous group of disorders with overlapping symptoms (e.g., somatic symptoms, problems sleeping, and avoidance behavior). They are commonly comorbid (14–16), and they have a positive treatment response to cognitive-behavioral therapy (CBT) and antidepressant medications (16). As a result, many studies of childhood-onset anxiety disorders, including treatment studies, group together children and adolescents with these three disorders and evaluate treatment effects on overall anxiety severity.

There is a substantial and growing literature supporting the use of psychosocial and psychopharmacological treatments of childhood anxiety disorders. Early detection of youth anxiety and application of the evidence base is critical to adequately and effectively address anxiety before anxious children develop secondary problems and dysfunction into adulthood. In the following, we discuss, first, the current evidence regarding pharmacotherapies, and then psychosocial interventions, focusing on CBT.

Pharmacotherapy

There is substantial evidence of the efficacy of antidepressant medications for separation, generalized, and social anxiety disorders in children and adolescents as young as age 6, yet no medication currently has U.S. Food and Drug Administration (FDA) approval for this use. Despite the implications of the FDA’s approval for safety and efficacy, the FDA labeling process does not reflect the existing and evolving evidence base for childhood anxiety disorders. It is important to note that a medication will not receive labeling for an indication if neither the manufacturer nor the FDA completes the FDA’s labeling process. As a result, clinicians who wish to practice evidence-based medicine for children with anxiety disorders will have to prescribe off-label.

There have been numerous short-term, double-blind, placebo-controlled trials assessing pharmacologic treatment of childhood-onset separation, generalized, and social anxiety disorders alone or in combination. The inclusion of patients with one of a combination of these disorders is consistent with findings that co-occurring anxiety disorders are the rule rather than the exception. Studies of patients with a combination of these disorders are usually investigator initiated. Because the FDA does not generally sanction trials of co-occurring conditions, industry-sponsored studies typically focus on a specific primary anxiety disorder but allow for symptoms of other anxiety disorders to co-occur. Given the rates of comorbidity among the childhood anxiety disorders, it makes sense to first discuss studies that include patients with co-occurring anxiety disorders.

Pharmacotherapy for Co-Occurring Anxiety Disorders

The largest and most rigorous study of comorbid anxiety disorders to date is the Child/Adolescent Anxiety Multimodal Study (CAMS) (16), which compared sertraline (up to 200 mg/day), pill placebo, CBT, and their combination over 12 weeks in 488 children ages 7–17 years presenting with separation, generalized, and/or social anxiety disorder. Participants were allowed to enroll with comorbid attention deficit hyperactivity disorder (ADHD) (if it was well managed with stimulant medications), oppositional defiant disorder, conduct disorder, and dysthymia. CBT plus sertraline was superior to CBT alone and sertraline alone (p<0.001, effect size=0.86). CBT alone (effect size=0.31) and sertraline alone (effect size=0.45) were equally effective, and all active treatments were superior to placebo (p<0.001). Sertraline was generally well tolerated, and there was no statistical difference in rates of adverse events between the sertraline and placebo groups. There was no evidence of harm-related adverse events in any participants.

In advance of CAMS, the Research Unit on Pediatric Psychopharmacology Anxiety Study Group completed one of the earliest investigator-initiated, multisite, double-blind placebo-controlled trials (17), comparing fluvoxamine (up to 300 mg/day) to placebo in 128 children ages 6–17 years with separation, generalized, and/or social anxiety disorder. Participants could enroll with comorbid oppositional defiant disorder and dysthymia. Over 8 weeks of study treatment, participants in the fluvoxamine group showed statistically greater rates of response (defined as an improvement score <4 on the Clinical Global Impressions Scale) than those in the placebo group (76% and 29%, respectively; p<0.01). Adverse events were generally mild, with abdominal discomfort and increased activity level being more common in the fluvoxamine group than in the placebo group. In a subsequent 6-month open-label follow-up study (18), treatment gains were maintained by 94% of participants.

The first investigator-initiated, single-site randomized controlled trial (19) compared fluoxetine (20 mg/day) to placebo over 12 weeks in 74 children ages 7–17 years with separation, generalized, and/or social anxiety disorder. Participants were allowed to enroll with comorbid ADHD, posttraumatic stress disorder, oppositional defiant disorder, conduct disorder, and dysthymia. The fluoxetine group had a significantly greater response rate (61%) than the placebo group (35%) by week 9 of the study. Reported side effects were minimal, with abdominal pain being more common in the fluoxetine group (44%) than in the placebo group (22%). Behavioral disinhibition was more common in the fluoxetine group (N=7) than in placebo group (N=4), but the difference was not statistically significant. Treatment gains were maintained in a 1-year open-label follow-up.

Pharmacotherapy for Individual Disorders

Most studies of childhood anxiety disorders have not enrolled participants with co-occurring anxiety disorders. All trials of medications other than selective serotonin reuptake inhibitors (SSRIs) for childhood anxiety disorders have evaluated only their efficacy for treatment of individual disorders. Many studies evaluating treatment of individual disorders are large industry-sponsored trials, and others are smaller investigator-initiated studies.

Social Anxiety Disorder (Social Phobia)

A number of studies have evaluated medications for childhood social phobia. In one randomized controlled trial (20), 80 children ages 7–17 years with social phobia were randomly assigned to receive 12 weeks of fluoxetine (up to 40 mg/day), placebo, or Social Effectiveness Therapy for Children. Children receiving social effectiveness therapy showed the greatest response rate (72.7%), and those receiving fluoxetine showed a greater response rate than those in the placebo group (30.1% compared with 6.3%).

A large, industry-sponsored, 16-week, double-blind, multisite study of paroxetine (up to 50 mg/day) in 322 children and adolescents ages 8–17 years with social phobia (21) demonstrated greater response rates in the paroxetine group (77.6%) than in the placebo group (38.3%). The medication was generally well tolerated, with insomnia, vomiting, and reduced appetite occurring more commonly in the paroxetine group than in the placebo group.

Beyond the SSRIs, a small number of trials have been conducted with serotonin-norepinephrine reuptake inhibitors and other antidepressants to treat childhood-onset anxiety disorders. One trial compared extended-release venlafaxine (up to 225 mg/day) to placebo in social phobia in 293 children and adolescents ages 8–17 over 16 weeks (22). The response rate was greater in the venlafaxine group (56%) than in the placebo group (37%). While most adverse events were mild to moderate, including nausea, anorexia, weight loss, pharyngitis, and mydriasis, three participants in the venlafaxine group developed new suicidal ideation, compared with none in the placebo group. A small (N=18) 8-week open-label trial of mirtazapine in children with social phobia (23) found improvement in 56% (10/18) of participants. Four of the participants discontinued because of adverse effects, including fatigue and irritability. There was also significant weight gain on the medication. No other studies have been reported to date on the use of mirtazapine in children with anxiety. However, mirtazapine is a unique antidepressant; it comes in a fast-dissolve formulation, and at low dosages it has pronounced sedative effects and enhances appetite, effects that tend to decrease with higher dosages. Mirtazapine’s potential benefits and combination of attributes may make it an ideal choice for the anxious child who becomes activated on SSRIs, has trouble swallowing pills (common in anxious children), and has appetite inhibition due to anxiety.

Selective mutism, most commonly observed in young children, is considered to be closely linked with, and a more severe form of, social phobia (24). Thus, the efficacy of antidepressants for older children with social phobia informs the pharmacological approach to young children with selective mutism, and its treatment can be considered together with that of social phobia. A small 9-week open-label trial of fluoxetine (up to 60 mg/day) (25) demonstrated improvement in symptoms among 76% of the 12 participants with selective mutism, with an inverse correlation between response and age. A small (N=5) 16-week double-blind single-case research trial of sertraline (26) showed a nonsignificant improvement in the sertraline group. However, two of the five participants no longer met criteria for selective mutism after 10 weeks on sertraline (at 100 mg/day), and a third participant was asymptomatic at 20 weeks poststudy. A small (N=15) 2-week controlled trial of fluoxetine in children with selective mutism (27) found improvement with fluoxetine treatment after nonresponse with placebo, but the difference, based on overall clinician-rated improvement, was not statistically significant (effect size=0.21). The study was limited by its small sample size and short duration.

Generalized Anxiety Disorder

Only four studies to date have evaluated pharmacologic treatment of childhood generalized anxiety disorder. Two industry-sponsored, 8-week, double-blind, placebo-controlled trials (28) evaluated extended-release venlafaxine (up to 225 mg/day) in children ages 6–17 years with generalized anxiety disorder. Pooled data from the 320 participants identified greater response rates in the venlafaxine group (68%) than in the placebo group (47%). Venlafaxine was generally well tolerated. The most common adverse events were asthenia, pain, anorexia, and somnolence. A small investigator-initiated double-blind trial (N=22) (29) demonstrated the efficacy of a fixed dose of sertraline (50 mg/day) in patients ages 5–17 years over 9 weeks. An industry-sponsored trial of buspirone for children and adolescents with generalized anxiety disorder (not published but described briefly in the buspirone product information) did not demonstrate a greater response in the buspirone group compared with the placebo group; response rates to both active drug and placebo were elevated.

School Refusal and Separation Anxiety Disorder

Although school refusal is not a DSM-recognized disorder, it is a frequent concern of parents presenting with their children for evaluation, is often seen in the context of childhood-onset anxiety disorders, and has been studied in children with separation anxiety disorder. One of the earliest psychopharmacological treatment trials and earliest trials comparing medication and psychotherapy (30) randomly assigned a small sample (N=35) of children with school refusal to a 6-week trial of imipramine (up to 200 mg/day) or placebo with or without behavioral treatment. School attendance improved in a greater number of children in the imipramine group (81%) than in the placebo group (47%). All of the children in the imipramine group reported feeling much better after 6 weeks, compared with 21% of those in the placebo group (p<0.005). A subsequent attempt to replicate this study in children with separation anxiety disorder (31) did not find a benefit for imipramine. In a study comparing CBT plus imipramine to CBT plus placebo in 24 depressed adolescents with school refusal (32), the combination of CBT and imipramine was superior to CBT plus placebo (p<0.001) over 8 weeks. A double-blind placebo-controlled trial comparing clomipramine (up to 75 mg/day) to placebo in 51 youths with school refusal over 12 weeks (33) found no difference between groups. In an 8-week double-blind placebo-controlled trial of 24 children and adolescents with school refusal (34) comparing imipramine (up to 275 mg/day), alprazolam (up to 4 mg/day), and placebo, no differences were found between groups.

Psychosocial Interventions

Many patients do not respond to medication alone, and the evidence suggests that combining antidepressants and CBT can be more powerful than either intervention alone (16). In addition, milder symptoms of anxiety may not warrant medication and the attendant risk of side effects. Some parents and children are reluctant to consider medication regardless of symptom severity. Given these considerations, it is important to understand the level of evidence supporting psychosocial interventions, as well as the underlying theories and components of commonly used treatments. It is valuable to understand these concepts as a clinician treating children with anxiety disorders, even if another clinician is performing the therapy. This improves effective communication between providers in split-treatment scenarios and provides a common language when speaking with children and parents.

CBT has been found to be at least as effective for pediatric anxiety disorders as standard pharmacotherapy (16). Psychodynamic approaches to childhood-onset anxiety have been proposed and are being studied (35), but to date few empirical studies have demonstrated a benefit for psychodynamic psychotherapy in anxious youths (36, 37). CBT is the psychosocial treatment with the broadest evidence base for pediatric anxiety disorders and thus is the focus here.

CBT

Dozens of studies have demonstrated the efficacy of CBT for pediatric anxiety disorders (38, 39). Citing this empirical support, the American Academy of Child and Adolescent Psychiatry anxiety practice parameter highlights the importance of CBT as a first-line treatment for mild to moderate anxiety in youths (40). CBT for pediatric anxiety utilizes a number of intervention strategies, including psychoeducation, relaxation training, cognitive restructuring, problem solving, gradual exposure, and relapse prevention (41, 42). Treatment course is typically brief (10–16 weekly 50-minute sessions) and emphasizes the importance of homework or out-of-office practice activities.

The goal of psychoeducation is for the patient to gain a better understanding of the phenomenology of anxiety and the rationale for CBT. The relationship between cognitions, somatic experiences, and behavior in anxiety-provoking situations is emphasized as a model for understanding CBT. Cognitive restructuring strategies help the patient identify anxious cognitive biases that result in unrealistic, unhelpful, or inaccurate appraisal of anxiety-provoking situations, with the goal of using objective evidence to come up with a more helpful or realistic appraisal or “coping thought.” Relaxation strategies, such as deep breathing and progressive muscle relaxation, are used to manage somatic symptoms, such as gastrointestinal distress, muscle tension, tachycardia, and difficulty sleeping. Patients are instructed to practice these skills in between sessions to improve their capacity to tolerate and regulate uncomfortable affect.

Exposure therapy is the most “active” ingredient and the most effective intervention in CBT. Exposure involves a gradual and systematic reacquaintance with anxiety-provoking stimuli and situations. Typically an exposure hierarchy is constructed, allowing the patient to rate potential exposure challenges on a scale from 1 to 10, creating a rank order from the easiest or least anxiety provoking to the most challenging and maximally anxiety provoking. The patient participates in the exposure task, often with the support of the therapist, a parent, or another trusted support person. The goal of each exposure task is to experience anxiety, or other uncomfortable affect, and remain in the situation long enough to master distress tolerance, habituate to the anxiety, and learn that the feared outcome is not going to come true. Exposure tasks are often repeated in order to experience decreased anxiety and increased mastery. Exposure tasks can include imaginal exposure (imagining the feared experience in one’s mind), narrative (writing about an anxiety-provoking situation), or in vivo exposure (actually confronting the feared stimulus or situation). Exposure can include looking at pictures or videos of the stimulus before completing the in vivo task. It can also include more general emotional material, such as a song or movie clip that brings up uncomfortable affect but is not necessarily specific to the patient’s symptoms. Often the patient is later asked to reflect on the exposure experience, citing what was learned and rating the change in anxiety over the course of the practice. Home practice tasks are assigned, and over time the benefits of exposure typically generalize across other potentially anxiety-provoking contexts. Treatment ends with a review of skills learned, exposure tasks mastered, and identification of potential future triggers to maintain treatment gains and prevent relapse.

Several published manuals are available that provide a structured overview of CBT delivery. The Coping Cat (43), a well-established CBT approach for treating pediatric separation, generalized, and social anxiety disorders that was used in the CAMS study, includes the following treatment components: identifying feelings and somatic symptoms, restructuring negative thoughts into “coping self-talk,” relaxation, problem solving, self-monitoring and reinforcement, graded exposure therapy, and skills practice. More recent clinical research efforts have taken a transdiagnostic approach, using CBT to improve emotion regulation across all anxiety and depressive disorders (44). Group CBT, such as the form used in Social Effectiveness Therapy for Children (20), has also demonstrated efficacy, although further study may help elucidate which children would benefit most from group or individual interventions.

Despite the strong evidence for CBT as an effective treatment for childhood anxiety, not all children improve, and some are only partial responders. In addition, research on CBT in children under age 7 is still lacking. Children may have poorer CBT outcomes when their parents are experiencing their own psychopathology that impairs their ability to support their child’s treatment (45–47). For example, parental anxiety may interfere with parents facilitating and following through with exposure tasks for their child. Thus, including parent education and training in the child’s treatment, and/or referring parents for their own mental health treatment, may improve children’s clinical outcomes (46).

Family education and training in CBT should include understanding the positive and negative reinforcement patterns that often maintain a child’s anxiety symptoms and impairing behaviors. Family accommodation of child anxiety symptoms, in the form of attention to anxious behavior (external positive reinforcement) or support of avoidance behavior (external negative reinforcement), is associated with greater symptom severity, poorer treatment outcomes, and greater functional impairment (48). For instance, when a child stays home from school because of anxiety, the anxiety relief experienced is an example of an external (parent-driven) negative reinforcement pattern that may worsen anxiety and avoidance in the long term. When a parent provides attention, support, and comfort in response to a child’s anxiety symptoms, this external positive reinforcement may inadvertently maintain the anxious behavior pattern. Functional assessment and intervention identifies these unhelpful reinforcement patterns and trains parents to pay more attention to mastery of brave behavior and decrease accommodation of the anxiety symptoms.

Conclusions

In the case of Jon, the patient described in the vignette, earlier identification and treatment of his anxiety symptoms might have prevented the accumulated disability seen in his history—social isolation, school failure, and demoralization. His parents would have been educated and provided with parenting strategies to avoid the accommodation of his anxiety, which led to external positive reinforcement (i.e., increased parental attention and soothing in the face of anxiety) and negative reinforcement (i.e., facilitated avoidance of anxiety-provoking situations, such as not going to other children’s homes for play dates) of his symptoms. Jon would have learned strategies to relax in the face of anxiety and ways of recognizing and challenging his anxious thoughts. Through targeted exposures, he would have learned distress tolerance skills and seen that the feared outcomes did not come true. Treatment with medication might have reduced his anxiety symptoms, “turning down the volume” on his anxiety, thus making it easier to engage in the CBT, persist in the face of difficult exposure tasks, and maintain treatment gains.

Jon developed limited coping and adaptation skills throughout development because of his anxiety symptoms, placing him at risk of failure in the transition to young adulthood. Even with effective pharmacotherapy and targeted CBT for his anxiety disorders, this lack of coping and adaptation skills will remain a barrier requiring a specific, targeted therapeutic intervention. Earlier intervention might have prevented development of these accumulated deficits, making treatment much more straightforward.

Early identification and treatment of the childhood-onset anxiety disorders is critical to preventing future disability and avoiding the scenario illustrated by Jon's case. There is overwhelming evidence supporting the efficacy of pharmacotherapy, particularly antidepressants, and CBT for the childhood-onset anxiety disorders. Furthermore, the evidence strongly supports the use of an SSRI and CBT in combination when both are available. Delaying treatment places the patient at increased risk of significant life impairment, symptom persistence, and development of additional psychiatric illnesses.