Assessment and Treatment of Anxiety Among Children and Adolescents

Anxiety is a feeling of fear, worry, or nervousness that also can involve distress, helplessness, and a somatically aroused central nervous system (1). It often represents a response to perceived danger or threat that is real, imagined, or exaggerated, and it can lead to avoidance of the stimuli causing the sense of danger. As such, anxiety frequently is normal and adaptive, including among children and adolescents and particularly during specific developmental periods. For example, stranger anxiety (the distress an infant may demonstrate when presented with unfamiliar people) is a common and expected phase of early childhood development, typically peaking between the ages of six and 12 months. The presence of stranger anxiety typically supports simultaneously the infant’s safety and the development of strong relationships between the infant and his or her primary caregivers, notably during a developmental period in which the vulnerable infant is becoming increasingly aware of and curious about his or her surroundings (2).

However, when anxiety is persistent, is excessive, or occurs in inappropriate contexts (including during atypical developmental periods or in nonthreatening settings), it can cause significant distress or impairment in a youth’s abilities to function in day-to-day life. In such instances, this type of intense, maladaptive, and disruptive anxiety no longer represents normal or nonclinical anxiety but instead represents anxiety in the clinical or pathological realm. The broader category of pathological anxiety is subdivided into various narrower anxiety disorder diagnoses, each of which is defined according to both distress- or fear-inducing stimuli, on the one hand, and physiologic, psychoemotional, and behavioral manifestations of anxiety, on the other. This review focuses on the assessment and treatment of the following anxiety disorders among children and adolescents: generalized anxiety disorder (GAD), social phobia or social anxiety disorder, separation anxiety disorder, and panic disorder.

GAD typically consists of diffuse, difficult-to-control worries and general unease that pertain to multiple domains (e.g., school performance, health of self and others, the future). Among youths, GAD frequently can involve chronic irritability, restlessness, impairment of concentration, difficulties sleeping, and a repeated need for reassurance from parents or teachers. Social phobia, also termed social anxiety disorder, involves the experience of significant anxiety or fear in social situations (including performance situations) in which there are concerns of potential negative judgment by others. Children and adolescents with social phobia frequently demonstrate restricted interactions with peers, avoidance of social gatherings, and limited participation in class.

Separation anxiety disorder occurs when there is excessive fear or worry of losing or being separated from major attachment figures. Youths with separation anxiety disorder may follow their parents around the house, insist on sleeping with their parents, and avoid separation from their parents (e.g., attending school, a sleepover, or sleep-away camp). Furthermore, when such youths are separated from their parents, they repeatedly may call their parents to “check in.”

Panic disorder, described as a “fear of fear,” involves fear of the physical sensations accompanying the fight-or-flight response that characterizes panic attacks. With recurrent panic attacks, the individual fears the occurrence and consequences of sensations such as dyspnea, tachycardia, parasthesias, and other autonomic sensations. Youths with panic disorder may avoid either settings in which they previously experienced a panic attack or activities that involve physical sensations that approximate those of a panic attack (e.g., vigorous exercise with associated increased heart rate and labored breathing). Table 1 outlines the distinguishing characteristics of these specific anxiety disorder diagnoses.

Epidemiology and Natural Course

Among children and adolescents, anxiety disorders are very common psychiatric disorders (3, 4). Between 15% and 30% of youths are diagnosed as having an anxiety disorder before adulthood (3–9). Indeed, the onset of anxiety disorders is most commonly in childhood, with a median onset between six and 11 years of age (3, 7). Table 1 outlines the lifetime prevalence values for several separate anxiety disorder diagnoses (3, 4, 7, 9–11). Of note, youths frequently demonstrate cumulative comorbidity with several anxiety disorders (12, 13). In particular, the disorders of the “anxiety triad”—GAD, social phobia, and separation anxiety disorder—are frequently studied together, because they have the highest prevalence, have high comorbidity, share similar responses to cognitive-behavioral therapy (CBT) and selective serotonin reuptake inhibitors (SSRIs), and thus have been considered distinct from other anxiety disorders (e.g., obsessive-compulsive disorder [OCD] or posttraumatic stress disorder) (14). In a study by Kendall et al., 75% of the treatment-seeking sample of youths met criteria for two or more of these three anxiety disorders (12). However, Copeland et al. determined in a longitudinal study that, with respect to concurrent comorbidity, only 13.8% of youths with an anxiety disorder met criteria for two or more comorbid anxiety disorders (9).

In the natural course of anxiety, youths frequently do not retain a specific anxiety disorder diagnosis over time. Many childhood anxiety disorders remit within three to four years (15), and stability rates of anxiety disorders among community youths are only low to moderate (16). Some researchers have argued that this fact and the aforementioned high rate of comorbidity with other anxiety disorder diagnoses suggest that the narrower anxiety diagnoses should be subsumed under a larger and broader diagnosis of “anxiety” (17); however, others have reflected that individual anxiety disorders differ in important ways (i.e., developmental course, familial aggregation, comorbidity patterns, risk factors, biology, and prediction of adult disorders) (5, 9, 18, 19). In particular, some have argued that anxiety as a phenomenon may be expressed differently over the course of development, thereby meeting criteria for different anxiety disorders at different points of development (12). Copeland et al. observed that separation anxiety was common early in their longitudinal cohort’s development and that generalized anxiety, panic, and agoraphobia were common in young adulthood (9).

Despite the instability of specific anxiety diagnoses of youths over time, early-life diagnoses of anxiety affect long-term outcomes. Unaddressed or incompletely treated, anxiety disorders among youths predict future anxiety disorders, depression, substance misuse, and psychiatric hospitalization (11, 20–23). In addition, the long-term effects of anxiety seem to extend beyond psychiatric functioning. Inadequately addressed anxiety also has been shown to lead to notable financial and societal costs through increased service use and lost productivity in the workplace (24). Bodden et al. determined that a family with a clinically anxious child requires societal funds that are 21 times higher than those required by a family from the general population (25).

Biopsychosocial Underpinnings

Numerous factors in an individual’s biological makeup and early psychological and social (including familial) environments have been identified to confer risk for future anxiety. Additionally, no single factor confers risk for anxiety; rather, the accumulation of risk across various domains leaves a person most vulnerable. Furthermore, the largely unexplored and minimally understood interplay among these factors likely significantly influences an individual’s overall risk of developing anxiety in youth.

Multiple studies have demonstrated familial aggregation of anxiety. A large population study showed that children of parents with anxiety disorders were significantly more likely to meet criteria for anxiety disorders than were children of parents without anxiety disorders (26). When both parents had received diagnoses of anxiety while hospitalized, the risk for the children increased even further. A birth cohort study using the Danish Registry System concluded that both a maternal history and a paternal history of any anxiety disorder increased the risk of diagnosis of anxiety between the ages of 10 and 21 years (odds ratio [OR]=2.38 and 1.84, respectively) (27).

In exploring the potential genetic transmission of anxiety disorders, twin studies have been a major contributor to the present understanding. In their meta-analysis, Hettema et al. determined that genetic heritability—estimated to be 30%−40% across the various anxiety disorders—is the major source of familial risk for anxiety (28). The authors noted that this represents only a modest level of heritability and emphasized that the majority of an individual’s risk for anxiety is determined by factors in his or her environment. Supporting the earlier-referenced argument favoring an approach to anxiety more broadly (rather than to anxiety subtypes or specific anxiety disorders), Shimada-Sugimoto et al. proposed in their review that any genetic heritability may pertain to clinical or pathological anxiety in general and may “transcend” the classification of specific anxiety disorder diagnoses (29).

In addition to genetic determinants of anxiety, researchers are exploring complex potential epigenetic determinants of anxiety—particularly gene-environment interactions. The combination of adverse childhood environmental experiences and particular genetic variants of the serotonin transporter gene (SLC6A4) and brain-derived neurotrophin factor gene have been associated with anxiety symptoms (30, 31). DNA methylation frequently is identified as the pathway by which the environment affects gene expression, and this mechanism of methylation alterations may also pertain to anxiety disorders (29, 32, 33).

Temperament is often recognized as a potential contributor to the eventual development of anxiety among youths. Behavioral inhibition is one type of discrete temperament category. It reflects a tendency to display signs of fear and wariness in response to unfamiliar stimuli and may be genetically linked to an allele of the corticotropin-releasing hormone (34). As outlined by Pérez-Edgar and Fox (1) and by Paulus et al. (35), multiple studies have demonstrated that behavioral inhibition in early childhood is a risk factor for the development of anxiety in childhood and adolescence.

The role that parents play in the etiology and maintenance of anxiety disorders among youths has been an area of interest and research. Lebowitz and colleagues reviewed the evidence supporting social learning theory and observational learning as mechanisms for the cross-generational transmission of anxiety—namely, that children can learn to fear a stimulus or assess the danger of a situation by observing the emotional reactions of others (e.g., parents) (36). They also highlighted the manners by which parental responses to children’s anxiety symptoms may affect the child’s risk of developing or maintaining an anxiety disorder. Regarding both nonclinical anxiety and already existing clinical anxiety, parental responses can influence the intensity of the child’s anxiety, the child’s sensitivity toward experiencing anxiety symptoms, and the strategies by which the child manages the anxiety.

The degree to which parents adjust their own behaviors to minimize their children’s anxiety-related distress (a concept referred to as family accommodation or overprotection) also has been a particular area of inquiry regarding the development and persistence of anxiety disorders among youths. Although the causal pathways have not been delineated, family accommodation has been shown to be quite common among the parents of children with anxiety disorders (37, 38). The first of these two studies found that as the degree of family accommodation increased, so did the severity of the children’s anxiety symptoms (37). Two apparent particular factors contributing to this correlation were active participation in symptom-driven behaviors (e.g., providing verbal reassurance or cosleeping) and modification of family routines. Thus, some researchers have recommended that reducing family accommodation may lead to improvement of the child’s anxiety and may increase the likelihood of successful treatment (38).

Early childhood adversity has been identified as a consistent predictor of psychopathology during adolescence and early adulthood. Exposure to physical and sexual abuse early in life has been found to increase lifetime rates of anxiety disorders (39). Hyland et al. determined that a history of placement into an institution or foster home in childhood predicted a later diagnosis of anxiety between the ages of 10 and 21 (OR=1.47) (27).

Anxiety Assessments

Comprehensive assessment of child and adolescent anxiety disorders benefits from a multimethod approach to evaluation and diagnosis. The Anxiety Disorders Interview Schedule for DSM-IV-TR, Child and Parent Report (ADIS-C/P) (40), is considered the gold-standard assessment tool for evaluating anxiety disorders among youths (41). The ADIS-C/P is a semistructured interview that assesses the presence of each of the anxiety disorders listed in the DSM. The clinician conducts separate interviews with the child and the parents, and diagnostic determinations are based on the information obtained from both reports. The clinician assigns a clinical severity rating on an 8-point scale for each relevant diagnostic category, with clinical severity ratings of 4 and above indicating a diagnosis at a clinical level of severity. In addition to assessing the presence of anxiety disorders, the ADIS-C/P may also be used to assess for OCD, traumatic stressors, mood disorders, and disruptive behavior problems. An updated version of the ADIS corresponding to diagnostic criteria from the DSM-5 is currently in press (42).

For practitioners seeking a more abbreviated semistructured assessment tool, the Pediatric Anxiety Rating Scale (PARS) (43) assesses symptom severity, frequency, and associated functional impairment for social anxiety disorder, separation anxiety disorder, and GAD. The PARS comprises 50 items assessing the presence or absence of symptoms of these disorders within the past week and is administered to children and their parents separately. The PARS also includes a global index of anxiety disorder severity, which is useful for evaluating change over time during treatment and in the context of comorbid conditions.

A number of child-report questionnaires are available to assess anxiety disorder symptoms. For example, the Revised Children’s Anxiety and Depression Scale (44, 45) is a 47-item child self-report measure that assesses the frequency of symptoms associated with anxiety and depression among youths ages six to 18. The measure generates disorder-specific subscales, as well as a Total Anxiety Scale and a Total Internalizing Scale.

In addition to assessing anxiety symptoms more broadly, it may be useful to assess syndrome-specific symptoms with brief child self-report measures. For example, the Penn State Worry Questionnaire for Children (46) assesses excessive worry, which is often associated with GAD, and the Panic Disorder Severity Scale for Children (47) assesses the frequency and severity of symptoms associated with panic disorder among youths.

Although symptom remission below full diagnostic criteria is an important indicator of treatment effectiveness, assessing changes in functional impairment is an equally important criterion by which to establish meaningful therapeutic change (48). To this effect, measures such as the Child Anxiety Impact Scale (49, 50) and the Strengths and Difficulties Questionnaire (51, 52) provide assessment of the degree to which anxiety symptoms interfere in school, social, and family settings. Both child- and parent-report versions of the Child Anxiety Impact Scale are available, which may be useful when researchers are evaluating parent and child perspectives of anxiety-related functional impairment. Parents, teachers, pediatricians, school psychologists, and other adults or caretakers who interact regularly with the child may provide vital information that informs accurate diagnosis and treatment planning. Collateral histories from these sources should be obtained.

Informant-report measures are a key component of comprehensive assessment of child anxiety disorders. The Achenbach Systems of Empirically Validated Assessment (53) is a widely used cross-informant measure assessing a wide range of child problem behaviors. This system includes the parent-report Child Behavior Checklist as well as the Teacher Report Form and a Youth Self-Report version, all of which present comparable items assessing the frequency of problematic child behaviors. Regarding parent-report measures that are specific to child anxiety symptoms, the Revised Child Anxiety and Depression Scale—Parent Version (54) is a parent version of the child self-report measure and may provide useful data to inform diagnostic impressions and symptom change over treatment. Assessing parental accommodation of children’s anxiety symptoms is an important factor that should be considered throughout treatment. The Family Accommodation Scale Anxiety (37, 38) assesses the extent to which parents of anxious children participate in their child’s symptoms and modify their behavior accordingly.

Behavioral observations provide important information to corroborate diagnostic impressions. Although we are unaware of formalized, psychometrically validated behavioral assessment tools to measure anxiety symptoms among children, naturalistic observations made during the course of intake evaluation and early sessions in treatment provide valuable information regarding patient diagnosis and functioning. For example, noting whether a child struggles to separate from his or her parent during the initial intake evaluation may inform clinicians’ understanding of separation anxiety symptoms.

Structured behavioral tasks in which children are exposed to unfamiliar individuals or novel tasks relevant to their presenting concern can also provide useful data. For example, bringing an unfamiliar adult into the playroom with a child who presents with social anxiety provides a useful opportunity to observe anxiety-related behavior. Moreover, these data can be quantified (i.e., recording the child’s latency to approach the novel stimulus or individual), serving as a useful baseline indicator of the behavioral consequences of anxiety (e.g., 55, 56) and also response to treatment over time.

Differential Diagnosis

Anxiety symptoms are manifested broadly (Box 1) and can include avoidance behaviors, somatic complaints, social difficulties, and sleep disturbances. Anxiety disorders commonly co-occur (e.g., 3, 57), and many anxiety disorders share the same symptoms. Despite the prevalence of these disorders, as well as the range in symptomatology and associated impairment, child anxiety disorders are often underrecognized. Parents, caregivers, and practitioners may struggle to differentiate clinically interfering anxiety symptoms from normative childhood fears or behavioral inhibition, particularly given children’s likelihood to report somatic symptoms (e.g., stomach ache) instead of verbalizing worry or fear (58). When somatic complaints are the primary symptom reported, practitioners should rule out medical conditions that may underlie such symptoms to increase confidence in accurate anxiety disorder diagnosis. In particular, when evaluating panic attacks, practitioners should consider general medical conditions, including hyperthyroidism, hyperparathyroidism, pheochromocytoma, vestibular dysfunctions, seizure disorders, cardiac conditions (e.g., arrhythmias or supraventricular tachycardia), and pulmonary conditions (e.g., asthma or chronic obstructive pulmonary disease) (59). As with all psychiatric conditions, the prevalence of symptom overlap highlights the need for comprehensive assessment to increase the likelihood of accurate diagnosis.

Anxiety disorders are often comorbid with a number of other mental health conditions, including depressive disorders, OCD, and externalizing conditions, among others. Anxiety and depressive disorders are highly comorbid and can be difficult to distinguish from each other, although data suggest that the two are, indeed, distinct conditions. Consistent with Clark and Watson’s (60) tripartite model of anxiety and depression, De Bolle and colleagues (61) demonstrated that depression among youths is associated with high negative affect and low positive affect. In contrast, anxiety among youths is associated with high negative affect, but anxious youths maintain the ability to experience positive affect (which typically is not shared with depression). Thus, although youths with depression are likely to demonstrate depressed mood and anhedonia across situations, anxious youths are more likely to demonstrate heightened negative affect, primarily in the context of feared stimuli and situations.

GAD and OCD are often difficult to differentiate (62). Patients with both conditions are likely to report that intrusive or uncontrollable thoughts are “stuck” in their head. In addition, research suggests that worry may serve the same function as compulsive behavior by decreasing anxious arousal (e.g., 63), which makes it even more difficult to distinguish worry from mental compulsions (64). Considering the content of the aversive cognitive activity may be helpful in determining the appropriate diagnosis. For example, worries associated with GAD tend to be about “normal” forthcoming problems, such as concerns about academic performance or the health of loved ones, and it is the excessiveness of the worry—rather than the content of the worry—that is considered abnormal (65). In contrast, obsessional beliefs are often irrational or unrealistic—such as fears that a parent will die if the child does not say his or her prayers perfectly. It is also important to assess the vividness of the cognitive intrusion when differentiating worries from obsessions. Obsessions tend to be experienced in imaginal form (e.g., intrusive images of mutilated bodies, urges to push someone into traffic), whereas worry is more often experienced as verbal or abstract (64).

Distinguishing externalizing disorders, such as attention-deficit hyperactivity disorder (ADHD) and oppositional defiant disorder, from anxiety disorders presents another common diagnostic difficulty. Youths with anxiety disorders may become oppositional or argumentative during attempts to escape or avoid feared situations, which may be misconstrued as a disruptive behavior disorder if the function of the behavior is not fully understood. Differentiating anxiety concerns from externalizing conditions is further complicated by symptom overlap in diagnostic criteria. In particular, both GAD and ADHD include restlessness, difficulty concentrating, and inattention as part of their separate diagnostic criteria (65). Despite this symptom overlap, the causes and correlates of these symptoms differ. In GAD, intrusive worries and hypervigilance to threat cues often manifest as symptoms of inattention and restlessness (66), and internal preoccupation leads youths with anxiety to become distracted and have difficulty focusing on daily activities. In contrast, youths with ADHD demonstrate these symptoms as a result of preoccupation with external activities or stimuli.

When attempting to distinguish between ADHD and anxiety, clinicians should consider whether the child experiences symptoms of inattention outside of the context of anxiety-provoking situations. If the child is able to maintain attention and focus when unencumbered by stress, a diagnosis of an anxiety disorder may be more appropriate. In contrast, a diagnosis of ADHD requires functional interference across situations, and therefore symptoms of inattention should be observed in activities other than those provoking anxiety. Despite these differences, ADHD and anxiety disorders are highly comorbid, with comorbidity estimates ranging from 11% to 40% (67). This high degree of comorbidity highlights the need for comprehensive assessment to make accurate diagnostic determinations.

In addition, distinguishing social phobia from autism spectrum disorder (ASD) among children often presents a diagnostic challenge (68). social phobia and ASD are associated with deficits in interpersonal interactions, often with early onset, and these deficits may be associated with profound difficulties establishing and maintaining relationships. For clinicians attempting to distinguish between these conditions, obtaining a full developmental history is crucial, because symptoms of ASD typically appear earlier than those associated with social phobia. When deficits in social interactions are present, the clinician should assess for co-occurring restricted or repetitive patterns of behavior or interests (e.g., echolalia, hyper- or hyporeactivity to sensory input, inflexible adherence to routines), which are present among youths with ASD at an early age (i.e., emerging as early as six months) but absent among youths with social phobia. Language and communication delays, as well as deficits in joint attention, may also be important markers of ASD versus social phobia (65). It is important to note that individuals with ASD do not demonstrate fear of social interactions, as do those with social phobia (65), and therefore it is critical to assess the degree to which youths with social difficulties believe that interpersonal interactions are threatening.

Differentiating between psychotic symptoms and anxiety-driven hallucinations also warrants comment. Hallucinations are fairly common among nonpsychotic youths (69), and benign hallucinations—that is, tactile or visual hallucinations that present at night, are anxiety-based or related to phobias, and are short-lived—have been observed among preschool and early school-aged children with no additional psychotic symptoms (70). For example, hallucinations reported by a five-year-old child who sees a man in her bedroom when the lights are off or hallucinations of a seven-year-old child who claims that he feels bugs crawling on him in bed are likely attributable to anxiety concerns rather than to psychosis in the absence of additional psychotic symptoms (e.g., delusional beliefs, language disturbances, decreased motor activity, incongruous mood, bizarre behaviors, and social withdrawal) or medication reaction (70, 71). Examining the content of the hallucination, particularly when it appears in the absence of other psychotic symptoms, may be critical in determining the underlying psychological issues at play (71).

Psychosocial Treatment of Anxiety Among Children and Adolescents

Research efforts in psychosocial treatments over the past 50 years have greatly advanced the treatment of anxiety disorders among youths, leading to the development of a number of options for children and adolescents struggling with these conditions (72). Among the available psychosocial interventions, CBT approaches that incorporate exposure-based therapies have emerged as the most well-established treatment approaches for addressing anxiety disorders of children and adolescents (see 72, 73). Research indicates that CBT plus antidepressant medication is most effective in the treatment of anxiety for youths ages seven to 17, compared with either CBT or medication alone (74). However, 50%−70% of children and adolescents receiving CBT for anxiety disorders demonstrate significant improvements in symptoms without medication (e.g., 74–76), which makes CBT a promising option for families who prefer psychosocial treatment when available. Similar rates of efficacy have been demonstrated across child and adolescent samples (77), and these gains were maintained at long-term follow-up (78, 79).

CBT can be delivered in multiple formats and across many settings, with demonstrated efficacy when delivered in individual, family, group, intensive, school-based, and Internet-delivered formats (73, 80–87). Primary treatment strategies in CBT for youth anxiety include psychoeducation, relaxation training, modeling, contingency management, problem solving, cognitive restructuring, exposure, and relapse prevention (83, 88, 89). CBT also includes the provision of between-sessions assignments, during which youths are expected to practice implementing skills learned in treatment to enhance generalization to naturalistic settings (90). Many CBT protocols also address the ways parents’ responses to and accommodation of their child’s anxious behavior may exacerbate and maintain child anxiety. Parents receive psychoeducation about the impact of their behaviors on their child’s symptoms and are coached by therapists in how best to support their children during exposure practices that occur outside of sessions in a manner that encourages and rewards approach behavior.

Among the many CBT protocols available to address anxiety among anxious youths, the Coping Cat program is one of the most well-established. The Coping Cat program was developed to address three of the most common pediatric anxiety disorders—separation anxiety disorder, GAD, and social phobia—among youths ages eight to 13. In the initial randomized controlled trial (RCT) evaluating the efficacy of the Coping Cat program, 47 children were randomly assigned either to receive the intervention or to a wait-list control condition. At posttreatment, 66% of treated children no longer met diagnostic criteria for their primary anxiety disorder (91), and these gains were maintained at long-term follow-up (92). These results were replicated in a second RCT (93), the results of which indicated that 53% of intervention participants were diagnosis free at posttreatment.

Social Effectiveness Therapy for Children (SET-C) is another well-validated treatment targeting social phobia among youths. It includes psychoeducation and in-vivo exposure as well as social skills training delivered in a group format. The initial RCT evaluating the efficacy of SET-C randomly assigned 67 children between the ages of eight and 12 either to the SET-C group or to a nonspecific treatment control group. At posttreatment, 67% of participants in the SET-C group no longer met criteria for social anxiety disorder, compared with 5% of the control group, and these results were maintained at six-month follow-up (94). Both the Coping Cat and SET-C programs have been used in research evaluating the relative efficacy of psychosocial and pharmacological treatments for anxiety disorders of youths (74, 95).

Although fewer RCTs of non-CBT therapies have been conducted, preliminary support has been identified for several other psychosocial treatment options. Child and Adolescent Anxiety Psychodynamic Psychotherapy is a brief, manual-based psychodynamic therapy approach for youths ages eight to 16 with GAD, separation anxiety disorder, and social phobia that has demonstrated promising initial results (96, 97). In addition, promising preliminary support has been demonstrated for CBT with attachment-based family therapy (ABFT) for the treatment of adolescents with GAD, separation anxiety disorder, and social phobia. CBT-ABFT addresses aspects of the parent-child relationship that contribute to youth anxiety symptoms. Treatment targets parental beliefs about anxiety and parental overcontrol and overprotection, with the goal of repairing ruptures in the attachment relationship and promoting adolescent autonomy (98). Although initial support has been identified for treatments such as biofeedback (99) and stress inoculation training (100), a robust evidence base has not yet emerged to support the use of these treatments as first-line treatment approaches for child anxiety (72). Additional research is needed to strengthen the literature evaluating alternative treatments for children with anxiety who refuse or do not benefit from CBT (72).

Psychopharmacologic Treatment of Anxiety Among Children and Adolescents

Numerous RCTs have been published to support the use of SSRIs as a first-line medication treatment of impairing anxiety disorders among children and adolescents (Table 2). Additionally, some studies support the use of serotonin-norepinephrine reuptake inhibitors (SNRIs) and tricyclic antidepressants for the treatment of pediatric anxiety disorders (107–110). SSRIs have been established to be safe and efficacious for treatment of pediatric anxiety and are considered the medication of choice for this population (111). To date, there are limited data to establish the efficacy and safety of the use of benzodiazepines or buspirone; thus, these studies are referenced but are not reviewed in further detail (74, 95, 101–106, 112).

Many of these studies have focused on the conditions that make up the “anxiety triad,” which includes separation anxiety disorder, GAD, and social phobia. As previously discussed, the disorders in the anxiety triad are highly comorbid and share several symptoms. Some studies have focused specifically on GAD or social phobia. To date and to the best of our knowledge, no RCTs have been successfully conducted to study psychopharmacologic interventions for panic disorder among children and adolescents.

The Child and Adolescent Multimodal Study (CAMS) was a large, multisite study examining the relative efficacy of treatment options for children and adolescents with anxiety disorders that provided important data about youth response rates to SSRIs, CBT, and the combination of these modalities. Children and adolescents (N=488) ages seven to 17 with a primary diagnosis of separation anxiety disorder, GAD, or social phobia (or a combination thereof) were randomly assigned to receive either sertraline (mean±SD dose of 146±61 mg/day), 14 sessions of CBT, a combination of CBT and sertraline (133.7±59.8 mg/day; COMB group), or a placebo medication for 12 weeks. Pharmacotherapy was initiated with 25 mg sertraline daily and adjusted up to 200 mg daily through week 8 on the basis of tolerability and response. CBT was based on the Coping Cat program (91), modified to be appropriate given the age of the participant and 12-week duration of the study.

The primary outcome measure for the CAMS trial was the Clinical Global Impression–Improvement (CGI-I) scale. Results indicated that all active treatment arms were superior to the placebo condition at posttreatment. Children and adolescents in the COMB condition demonstrated the highest response rates, with 80.7% of participants rated as “very much improved” or “much improved” on the CGI-I scale, followed by 59.7% of participants in the CBT condition and 54.9% of participants receiving sertraline. Only 23.7% of those receiving the placebo medication were rated as “very much improved” or “much improved” posttreatment.

Further analysis of this study with attention to remission rates demonstrated 46%−68% remission rate for the COMB group, 34%−46% for the sertraline-only group, 20%−42% for the CBT group, and 15%−27% for the placebo condition. Rates of adverse events, including suicidal and homicidal ideation, were not significantly greater in the sertraline versus placebo conditions, and no child attempted suicide during the study period. Common adverse events associated with SSRIs among children and adolescents are similar to those observed among adults and include nausea, headache, and sleep disturbance.

The authors concluded that the combination therapy and monotherapies studied were all effective short-term treatments and further suggested that combination therapies provide the best chance of symptom improvement for youths struggling with anxiety disorders. They recommended that clinicians consider the family’s treatment preference as well as treatment availability, cost, and time burden when making treatment choices for patients.

Analysis of this study’s remission rates demonstrated a 46%−68% remission rate for COMB, 34%−46% for sertraline, 20%−42% for CBT, and 15%−27% for placebo (113). This group also investigated predictors and moderators of treatment response and identified that younger age, nonminority race and ethnicity status, lower severity of anxiety, absence of comorbid anxiety or depressive disorder, and absence of social phobia were all factors associated with improved response rates. Thus, practitioners should also consider these variables when determining treatment recommendations for a given patient.

Although the CAMS trial provides strong evidence for combination treatment, it bears importance to reinforce the safety and efficacy of psychopharmacological monotherapy, especially in the cases when evidence-based psychosocial interventions are not available. A recently published meta-analysis of RCTs focused on SSRIs in pediatric depression, OCD, and non-OCD anxiety disorders demonstrated strong effects for SSRIs in non-OCD anxiety (114). Although pooled increased risk for medication versus placebo was noted for suicidal ideation and suicidal behavior, no statistically significant difference was found between the groups. Furthermore, the number needed to treat for non-OCD anxiety was three, whereas the number needed to harm was 143.

Another comparative trial specifically evaluated the efficacy of another common SSRI, fluoxetine (40 mg daily), relative to SET-C among youths ages seven to 17 with a primary diagnosis of social phobia (95). Participants (N=122) were randomly assigned to receive fluoxetine-only treatment, SET-C, or medication placebo. Fluoxetine was initiated at 10 mg and increased through week 7 to 40 mg, which was maintained to week 12. Dose reduction was allowed in the case of side effects, but no adjustments were necessary among study participants. SET-C included weekly social skills training and peer generalization conducted in small groups (150-min sessions) and individual sessions for in-vivo exposures (on average, 60 min) for 12 weeks. Results indicated that at posttreatment, significantly more of the SET-C group (79%) were categorized as treatment responders on the CGI-I scale, compared with those who received fluoxetine (36.4%) or placebo (6.3%). Additionally, 53% of the SET-C group no longer met diagnostic criteria for social phobia at posttreatment, which was significantly higher than the fluoxetine group (21.2%) or placebo group (3.1%). No suicidal ideation was reported in the study.

Analysis of secondary outcome measures (with the social phobia and Anxiety Inventory for Children and the PARS) demonstrated that although both SET-C and fluoxetine were effective in diminishing behavioral avoidance and distress, youths treated with SET-C had significantly higher ratings of social skills than those receiving fluoxetine only or the placebo. Treatment gains in both active treatments were maintained at one-year follow-up assessment. On the basis of the results of this study, the authors concluded that although both treatments were efficacious, SET-C offered an advantage by enhancing social skills.

The most recently published RCT examining the efficacy of psychiatric medication treatment for children and adolescents with anxiety disorders warrants discussion in this review, because this study has informed U.S. regulatory authorization to establish the indication of duloxetine for the treatment of GAD among children and adolescents (101). To date, of all SSRIs, SNRIs, and tricyclic antidepressants, only duloxetine has received Food and Drug Administration approval for use for GAD of patients under age 18.

The double-blind trial randomly assigned 272 children ages seven to 11 years and adolescents ages 12 to 17 years to active or placebo groups (101). Duloxetine was initiated at 30 mg and titrated on the basis of CGI Severity scores (mean duloxetine dose 53.6 mg/day). After week 10, the placebo group was transitioned to receive 30 mg duloxetine, and then all participants at 12 weeks were transitioned to 60 mg duloxetine daily. After week 14, open-label flexible dosing was used, and the study continued through a 28-week extension.

To maintain consistency in assessment across studies, the researchers used the PARS to assess treatment response and remission rates, as in the CAMS trial. In acute treatment, children in the duloxetine group demonstrated a statistically significantly greater improvement than the placebo group; however, for adolescents, results for the duloxetine and placebo groups were not significantly different. Because the treatment × age category interaction for the PARS severity for GAD was not statistically significant, the authors concluded that there were no differential treatment effects among children versus adolescents.

During the extension treatment, both the group that was initiated on duloxetine and the group started on placebo and then transitioned to duloxetine showed mean improvement in PARS severity for GAD, as well as anxiety symptom severity, CGI Severity score, and Clinical Global Assessment Scale score. The Columbia Suicide Severity Rating Scale was used to assess for suicidal ideation and behavior. Suicidal ideation, suicide attempts, and aborted suicides were reported among both the participants with a history of suicidal ideation and those with no prior events. Results indicated that there were no statistically significant differences between medication and placebo groups for suicidal ideation or behavior.

Statistically significant differences between the duloxetine and placebo groups in mean change in pulse (mean change increase) and weight (mean change decrease), but not blood pressure, were observed. The authors suggested that because hemodynamic changes may be noradrenergically medicated and related to norepinephrine reuptake inhibition, monitoring blood pressure and pulse before initiating treatment and periodically throughout treatment with duloxetine is advisable.

No long-term studies assessing the developmental effects or risks of SSRIs or SNRIs have been published; however, investigators are currently recruiting for a long-term study of the impact of treatment with sertraline among children with a diagnosis of anxiety disorder, depressive disorder, or OCD (115). As such, the American Academy of Child and Adolescent Psychiatry recommends that clinicians consider a trial off medication after significant reduction in anxiety symptoms has been maintained for approximately one year. The medication-free trial should be considered at a low-stress period of the youth’s life, and medication may be reinitiated if symptom relapse is observed (111). During the medication-free period, CBT booster sessions, continued parents’ work, and classroom-based accommodations may continue to be important interventions (111).

Future Directions

Research efforts over the past several decades have supported the efficacy, effectiveness, and acceptability of a range of psychosocial and pharmacological treatments for youths with anxiety disorders, allowing practitioners to tailor treatment approaches to meet the unique needs and preferences of individual patients. Although many promising treatment avenues are available for these conditions, additional research is needed to broaden the available evidence-based intervention options for youths struggling with anxiety.

Several investigations are examining use of glutamatergic modulators (e.g., memantine, minocycline, and d-cyloserine) to augment medication, CBT, or both (116–119). In addition, antiadrenergics have also been implicated in enhancing central adrenergic sensitivity among pediatric patients with anxiety disorders in a manner that may promote treatment response rates (119); this is a promising avenue for further exploration. Efforts must also continue to prioritize the dissemination and implementation of evidence-based treatments for child anxiety beyond the institutions where they are developed to reach more children in need of services. Computer- and Internet-delivered treatments, school-based interventions, treatments delivered in primary care settings, and intensive treatments provide creative solutions to expanding the scope and accessibility of high-quality mental health care for youths (81, 120); however, much work remains in this area. Furthermore, additional research identifying mediators, moderators, and predictors of treatment response is crucial to advance the understanding of treatment elements most responsible for meaningful therapeutic change (121, 122). To this effect, dismantling studies are needed to better understand the additive impact and optimal timing of various practice elements (72).

Key Points