Emotional Reactivity and Regulation Following Citalopram Therapy in Children and Adolescents with Anxiety Disorders

Abstract

Objective:

Emotional dysregulation is an important element in the pathophysiology of childhood anxiety disorders and can distinguish anxious subjects from controls. Treatment with selective serotonin inhibitors (SSRIs) has been found to lessen anxiety, but its effects on emotional reactivity and regulation are less documented. The aim of the study was to prospectively assess changes in emotional reactivity and regulation in response to citalopram in children and adolescents with anxiety disorders, with special focus on the mechanism of reappraisal.

Methods:

The sample included 70 children and adolescents (38 boys, 32 girls) 10–17 years of age, divided into three groups: Those with anxiety disorder treated with citalopram for 8 weeks (n = 35); untreated subjects with anxiety disorder on the waiting list for cognitive behavioral therapy (CBT) (n = 15); and subjects without anxiety disorder (controls) (n = 20). Emotional reactivity and regulation (i.e., reappraisal), were assessed at baseline and after 8 weeks (follow-up) with validated computer-based instruments, Reactivity and Regulation-Situations (REAR-S) and Reactivity and Regulation-Images (REAR-I).

Results:

Citalopram-treated subjects showed significantly greater improvement in reappraisal ability than CBT-waitlisted subjects. Improvement in the ability to reappraise threatening images correlated significantly with the decrease in anxiety. There was a decrease in negative emotional reactivity between assessments, which was positively correlated with clinical improvement. Higher intensity of baseline reactivity (on the REAR-S) predicted more severe symptoms at follow-up.

Conclusions:

Citalopram therapy improves reappraisal ability in children and adolescents with anxiety. However, the improvement in other examined emotional reactivity indices occurred in both medicated and waitlisted groups. It is possible that these findings may have implications for understanding the pathophysiology of anxiety in children and adolescents.

Introduction

Anxiety disorders can be debilitating, affecting well-being, self-esteem, social, academic, and familial functioning, as well as the ability to attain normal developmental emotional goals (Lyneham and Rapee 2005). Moreover, despite their high prevalence, substantial morbidity, and other deleterious long-term effects, anxiety disorders in childhood remain underrecognized and undertreated (Walkup et al. 2008). Monotherapy with cognitive behavioral therapy (CBT) or selective serotonin-reuptake inhibitors (SSRIs) has been found efficacious. Combination of the two treatments is more efficacious than monotherapy with either CBT or SSRIs (Ginsburg et al. 2011, 2014; Piacentini et al. 2014). However, better understanding of the mechanisms underlying the response to treatment is needed (James et al. 2005; Bridge et al. 2007).

The affective approach to the study of emotional problems is based on the assumptions that emotional processes comprise both generative and regulative elements (Gross 2013), and that abnormality in these processes plays an important role in most forms of psychopathology, including anxiety and mood disorders (Rottenberg et al. 2003; Lonigan et al. 2004).

The emotion-generative process begins when an external or internal event signals to the individual that something important may be at stake (e.g., an upcoming separation from parent through the eyes of a child). When attended to and interpreted in certain ways (e.g., something bad will happen to both self and parent following separation), these emotional cues trigger a coordinated set of response tendencies that involve experiential (e.g., feeling anxious), behavioral (e.g., crying and clinging to parent), central (e.g., amygdala activation), and peripheral physiological systems (e.g., increased heart rate). People vary significantly in the quality and intensity of their emotional responding (Davidson 1998).

Emotional reactivity can be defined as the threshold of stimuli needed to generate emotional response, and the intensity of the various components of the emotional response. The measured emotional response can be either positive or negative and is reflected as a change from baseline affect (Davidson 1998; Bylsma et al. 2008). Emotion regulation refers to attempts that individuals make to influence which emotions they have, when they have them, and how these emotions are experienced or expressed (Gross 2002).

The role of emotion regulation in healthy development is supported by developmental studies showing that emotion regulation is central to social adjustment, attaining personal goals, and overall emotional health (Cicchetti and Toth 1995; Gross and John 2003; Mayer et al. 2003; Zeman et al. 2006; Domes et al. 2010). Attention to the structure of emotion and its relation to psychopathology has been increasing over the past decades. This line of research has been applied to specific types of psychopathology (Davidson et al. 2002; Rottenberg et al. 2002; Lonigan et al. 2004; Suveg and Zeman 2004; Bloch et al. 2015). Early findings in anxiety disorders suggest that anxious individuals may be identified by the expression of intense and frequent negative emotional responses to stimuli that would normally evoke no or only mild negative emotions in others, combined with an inability to effectively regulate these emotions when they arise (McLaughlin et al. 2007; Carthy et al. 2010a,b; Blechert et al. 2012; Gross 2013).

Recent studies have sparked interest in the concept of reappraisal as a key mechanism in emotional regulation. Reappraisal is defined as a change in the trajectory of an emotional response based on a reinterpretation of the meaning of the emotional stimulus. Reappraisal alters the experiential, behavioral, and physiological aspects of the emotional response (Blechert et al. 2012). For example, in healthy adults, heightened use of cognitive reappraisal served as a mechanism through which promotion-focused regulation (i.e., pursuing desirable end states by promoting positive behaviors/cognitions concerned with achievement and advancement) protected against anxiety (Llewellyn et al. 2013).

In previous publications (Carthy et al. 2010a,b), we reported new instruments for the real-time assessment of emotional reactivity and regulation (with special focus on reappraisal), which can distinguish anxious children from controls. Specifically, the results indicated that children with anxiety disorders have negative emotional hyperreactivity and diminished reappraisal ability and self-efficacy, and that they exhibit less frequent use of reappraisal and problem solving in everyday activities. Their use of emotional regulation strategies is characterized by higher reliance on avoidance and on seeking help from others (Carthy et al. 2010a.b).

SSRIs have been found to be effective in the treatment of anxiety (Walkup et al. 2008), and their impact on emotional reactivity and regulation has been investigated (Faria et al. 2014). Most research focused on changes in the reactivity of the amygdala to negative emotional stimuli, usually facial images (Sheline et al. 2001; Drevets et al. 2002; Fu et al. 2004). Serotonin seems to be involved in inhibition of the amygdala by the prefrontal cortex, both areas associated with emotional reactivity and regulation (Gross 1998; Ochsner et al. 2004; Liddell et al. 2005; Ray et al. 2008; Fisher et al. 2009; Goldin et al. 2009; Ray et al. 2010; Ochsner et al. 2012).

Prompted by studies reporting reduced activation of the amygdala in response to threatening stimuli in adults treated with SSRIs (Furmark et al. 2002; Szeszko et al. 2004; Furmark et al. 2005), we sought to determine if biological intervention with SSRIs can also reduce abnormalities in emotional reactivity and regulation in anxious children and adolescents and if this improvement is related to remission from anxiety. Given that anxious children and adolescents who receive medication show greater clinical improvement than those who do not, we hypothesized that anxious children and adolescents with an SSRI-treated anxiety disorder would have decreased negative emotional reactivity and improved cued and uncued (i.e., spontaneous) reappraisal ability.

Methods

Participants

Study participants included 50 children and adolescents (27 boys and 23 girls) 10–17 years of age attending the Anxiety Disorder Clinic of a tertiary pediatric medical center. To ensure that intelligence level and reading speed and accuracy in all study participants did not differ from normal controls, 20 children and adolescents (11 boys, 9 girls) without anxiety disorder from the community served as a control group (nonanxious children [NAC]). The study and control groups did not differ in average age and gender distribution.

The subjects with anxiety disorder were further divided into 35 patients (21 boys, 14 girls) being treated with the SSRI citalopram (medication group [MED]) and 15 patients (6 boys, 9 girls) who had refused treatment with citalopram and were on the waiting list for CBT (waiting list group [WL]). The assessment and treatment procedures in our clinic are described in detail subsequently. Other inclusion criteria for all study participants were normal intelligence and routine school attendance. Because of the small number of participants, we excluded patients with reading disabilities, psychotic symptoms, current psychological or psychiatric treatment, or a major life stressor within the past 6 months. Subjects with a primary diagnosis of obsessive-compulsive disorder or posttraumatic stress disorder were excluded as well. All participants and their parents provided informed assent/written consent, respectively, to participate in the study. The study was approved by the Institutional Review Board.

Table 1 presents the participants' characteristics and tabulates the specific differences among the three subgroups (MED, WL, NAC) at baseline assessment on clinical, demographic, and cognitive measures. Parent ratings of child anxiety were higher in the medication group. It is possible that parents' assessments of symptom severity played a causal role in condition assignment, such that parents who perceived their child's anxiety as more severe were more likely to choose medication for their child. As expected, both anxious subgroups had significantly higher scores in all clinical measures than the NAC.

Table 1.

Clinical and Cognitive Measures in Citalopram-Treated and Untreated (CBT Waitlisted) Children and Adolescents with Anxiety Disorder and Controls

	Anxious group - medicated (n = 35) mean (SD)	Anxious group - waitlist (n = 15) mean (SD)	Nonanxious control group (n = 20) mean (SD)	df	F
Age (years)	13.4 (2.1)	13.5 (2.4)	13.0 (2.1)	67	0.2
Reading speed (WISC-IV)	57.5 (16.5)	62.2 (20.6)	66.6 (12.6)	66	1.6
Reading accuracy (percentage of mistakes) (WISC-IV)	4.3 (4.8)	3.5 (4.5)	6.0 (4.1)	66	0.15
Vocabulary subscale (WISC-IV)	33.3 (4.8)	32.2 (4.3)	34.2 (4.6)	66	1.3
Severity of anxiety- child report (SCARED-C)	33.1^a (12.9)	31.2^a (16.4)	8.3^b (5.5)	66	27.9^***
Severity of anxiety-parent report (SCARED-P)	38.8^a (14.3)	27.4^a (9.3)	5.4^b (4.6)	66	54.8^*
Severity of depression (CDI/BDI-II, t score)	57.1^a (8.3)	52.4^a (12.4)	44.5^b (2.6)	67	14.6^*
Negative affect (STAI-C)	30.1^a (5.4)	28.2^a (6.8)	23.1^b (5.5)	65	10.9^*
Global illness severity (CGI-S)	4.7^a (0.83)	4.2^a (1.0)	1.1^b (0.36)	67	135.0^*

Different lower case letters denote significant differences between specific groups on post-hoc tests.

p < 0.05, ^*** p < 0.001 (two tailed).

CBT, cognitive behavioral therapy; WISC, Wechsler Intelligence Scale for Children; SCARED, Screen for Anxiety and Related Emotional Disorders; CDI, Children's Depression Inventory; BDI, Beck Depression Inventory; STAI, State-Trait Anxiety Inventory for Children; CGI-S, Clinical Global Impressions Severity Scale.

Procedure

All subjects attending the Anxiety Disorder Clinic underwent a screening interview by the same senior psychiatrist (i.e., the standard protocol at the clinic). On clinical diagnosis, they were offered pharmacological treatment with an SSRI (citalopram), which required no waiting period, or psychological treatment (CBT) at the clinic, which had a mean waiting time of 10 weeks. In addition, they were also offered the opportunity to join the study. When families agreed, informed consents were obtained from parents and subjects as approved by Schneider Children's Hospital Institutional Review Board. Within a few days they met with one of the two child psychologists who were part of the study group and went through the baseline assessment, which included a diagnostic interview for primary and secondary diagnosis (The Anxiety Disorders Interview Schedule for Children [ADIS-C]), self-report questionnaires, affect questionnaires, and a computer-based assessment of emotional reactivity and regulation. The instruments used are detailed in the next section.

Medication management

The subjects who received citalopram and those who were waitlisted were followed by the same senior psychiatrist (A.A.). The appointments were scheduled in week 1 (starting medication), week 2 (follow-up and increasing to target dose), and week 5 (follow-up). Starting dosage for all patients was citalopram 10 mg/day for the 1st week, which was then increased to fixed dose of 20 mg/day through week 8.

After 8 weeks, both the waitlisted and treated patients returned to the clinic for follow-up assessment with the same questionnaires and computer-based tasks.

For the present study, the psychological instruments administered to the subjects with anxiety disorder were also administered to the NAC at the same interval. In appreciation for their participation, the NAC received a gift valued at $20.

Measures

Clinical symptoms

The ADIS-C is a widely used Diagnostic and Statistical Manual for Mental Disorders, 4th ed. (DSM-IV)-based clinical interview for the assessment of anxiety disorders and the differential diagnoses of common childhood psychological problems. Studies have reported acceptable reliability and validity (American Psychiatric Association 1994; Silverman et al. 2001).

The Screen for Anxiety and Related Emotional Disorders (SCARED), Child (SCARED–C) and Parent (SCARED-P) forms were used to assess the severity of anxiety symptoms (Birmaher et al. 1997). Internal consistency of the SCARED-C and SCARED-P in the current study was α = 0.92 and α = 0.80, respectively.

The Children's Depression Inventory (CDI) (Smucker et al. 1986) was used to assess depressive symptoms in children <13 years of age, and the Beck Depression Inventory (BDI-II) was used in adolescents (Beck et al. 1996; Osman et al. 2004). Internal consistency of the CDI and BDI-II in the current study was α = 0.91 and α = 0.80, respectively.

The Clinical Global Impressions–Severity Scale (CGI-S) was used to assess illness severity, and the Clinical Global Impressions-Improvement Scale (CGI-I) (Guy 1976) to assess clinical improvement.

Cognitive and reading abilities

The Wechsler Intelligence Scale for Children–IV (Wechsler 2003), vocabulary section, was used to check for normal intelligence. In addition, a popular measure of reading speed and accuracy that includes age norms (Shani et al. 2006) was administered to ensure normative reading abilities.

Negative affect

The state part of the State-Trait Anxiety Inventory for Children (STAI-C) (Spielberger 1973) was used to assess baseline negative affect. The scale has been found to have acceptable internal consistency (Papay and Spielberger 1986). Internal consistency of the STAI-C in the current sample was α = 0.75.

Emotional reactivity and regulation

The Reactivity and Regulation-Situations (REAR-S) and the Reactivity and Regulation-Images (REAR-I) are psychometrically sound computerized instruments for the assessment of emotional reactivity and emotion regulation. They were both developed and validated by our group (Carthy et al. 2010a,b). Briefly, for the REAR-S, subjects are seated in front of a computer screen and are presented with descriptions of 16 ambiguous situations taken from daily life, which have potentially threatening meanings (e.g., violence, dangerous animals, angry faces, accidents) in a familial, social, physical or performance context (e.g., “Your teacher hands you back your test and says she is surprised”). All items are one sentence long. To encourage self-reference, they are written in the present tense and second person. The task comprised two conditions, “view” and “reappraise.” In the first condition, patients are asked to state their initial thoughts and rate their emotional reaction on a visual scale from 1 (do not feel negative at all) to 8 (feel extremely negative). In the second condition patients are asked to reappraise aloud the presented situation, and rate the reappraisal's efficacy. The task enabled evaluation of two aspects of negative emotional reactivity frequency and intensity, and three aspects of reappraisal: Uninstructed, cued, and self-efficacy.

For the REAR-I, subjects are seated at a computer screen and are presented with 28 images of threatening content (e.g., violence, dangerous animals, angry faces, accidents) selected from the International Affective Picture System (Lang et al. 1999). Six parents of children in the sample's age group independently reviewed the images for age appropriateness. The task comprised two conditions, “view” and “reappraise.” Each 20 second trial included the following components: Instruction (“view” or “reappraise,” 2 seconds), image presentation (8 seconds), emotion rating (8 seconds), and relaxation (2 seconds). The REAR-I is designed to assess the intensity of negative emotional reactivity, cued reappraisal ability, and reappraisal efficacy.

Our earlier study found good internal consistency and interrater reliability for both the REAR-S and REAR-I. Internal correlations between the measures of emotional reactivity and the aspects of reappraisal were significant, supporting the construct validity of the tasks (REAR-S α = 0.89, REAR-I, α = 0.9) (Carthy et al. 2010a,b). In the present study, the test–retest reliability of the REAR-S and REAR-I indices was evaluated in the control group (r = 0.67 and r = 0.75, respectively).

Statistical analysis

The clinical results are presented as means and standard deviation (SD). Differences in baseline measures between groups were analyzed by ANOVA with post-hoc analyses. The extent of change in clinical parameters, emotional regulation, and reappraisal over time within groups was analyzed by ANOVA with repeated measures. Remission was defined as a CGI-S score ≤2 after 8 weeks. Differences between the groups in the likelihood of remission were tested via Fisher's exact test. The relationship between changes in clinical and emotional parameters was analyzed by Pearson correlations.

We constructed a composite measure of symptoms at both baseline and follow-up by averaging standardized parent, child, and clinician ratings of anxiety. This was justified by principal components analyses, which showed that among the anxious subjects, all three anxiety measures loaded on the same component at both baseline and follow-up. After calculating composite scores, we examined whether baseline characteristics were related to symptoms at follow-up after controlling for treatment group. Additionally, we examined whether baseline characteristics interacted with treatment group to predict follow-up symptoms. For all hypothesis tests, the significance level was set to 0.05.

Results

Baseline tests

Findings for the baseline clinical and cognitive parameters in all three groups are shown in Table 1. The distribution of the specific anxiety disorders (Silverman et al., 2001) in the two study groups is shown in Table 2. There was no significant difference in the mean number of disorders between the MED subjects (2.2 ± 0.93) and the WL group (2.1 ± 1.0). Mean age and sex distribution were similar in the anxious and NAC groups. Moreover, the two study groups did not statistically differ in baseline scores (SCARED, CDI/BDI-II, STAI-C, CGI-S).

Table 2.

Distribution of Diagnosis in Citalopram-Treated and Untreated (CBT-Waitlisted) Children and Adolescents with Anxiety Disorders

	Primary diagnosis (%)		Secondary diagnosis (%)
	Anxious group – medicated (n = 35)	Anxious group – waitlist (n = 15)	Anxious group – medicated (n = 35)	Anxious group – waitlist (n = 15)
Generalized anxiety disorder	40	33	14	20
Social anxiety disorder	20	20	23	26
Separation anxiety disorder	11	20	14	13
Specific phobia	20	13	51	40
Panic disorder	3	7	17	6
Anxiety disorder not otherwise specified	6	7	-	-
Obsessive-compulsive disorder	-	-	5	-
Eating disorder	-	-	2	13
Major depressive disorder	-	-	20	6
Attention-deficit/hyperactivity disorder	-	-	2	13

CBT, cognitive behavior therapy.

Effect of medication on clinical parameters

Table 3 compares the extent of clinical change from baseline to the 8 week follow-up between the anxiety groups. There was a significant decrease in both illness severity and parental rating of child anxiety in the MED group compared with the WL group. The MED group also showed significantly greater clinical improvement as measured by score on the CGI-I (t[48] = −4.4, p < 0.001, ANOVA). Analysis of the subject self-reports of anxiety yielded a significant effect of time, with a decrease in scores in both groups between the baseline and follow-up assessments, with no between-group difference.

Table 3.

Extent of Clinical Improvement from Baseline to 8 Weeks' Follow-Up in Citalopram-Treated and Untreated (CBT-Waitlisted) Children and Adolescents with Anxiety Disorder

	Time point	Anxious group-medicated (n = 35) mean (SD)	Anxious group waitlist (n = 15) mean (SD)	F time effect	F time × group
Severity of anxiety-child (SCARED-C)	Baseline	34.7 (14.0)	31.4 (16.3)	13.6^*	2.07
Severity of anxiety-child (SCARED-C)	Follow-up	23.0 (13.5)	26.3 (15.9)
Severity of anxiety-parent (SCARED-P)	Baseline	38.9 (14.3)	27.4 (9.3)	3.6	11.5^*
Severity of anxiety-parent (SCARED-P)	Follow-up	30.2 (13.1)	29.8 (12.9)
Severity of depression (CDI/BDI-II, t score)	Baseline	57.4 (8.3)	53.6 (12.3)	1.6	1.8
Severity of depression (CDI/BDI-II, t score)	Follow-up	53.4 (9.5)	53.6 (9.7)
Global illness severity (CGI-S)	Baseline	4.7 (.8)	4.2 (1.0)	41.7^**	15.4^**
Global illness severity (CGI-S)	Follow-up	2.8 (1.2)	3.7 (1.0)

p < 0.01, ^** p < 0.001.

CBT, cognitive behavioral therapy; SCARED, Screen for Anxiety and Related Emotional Disorders; CDI, Children's Depression Inventory, BDI, Beck Depression Inventory; CGI-S, Clinical Global Impressions Severity Scale.

Remission was defined as a CGI-S score ≤2 after 8 weeks. Remission was documented in 16 of the 35 in the MED group (44%) and 2 of the 15 in the WL group (13%). The likelihood of remission was significantly greater in the MED group (p = 0.02, one tailed Fisher exact test).

Effect of medication on negative emotional reactivity

In the cohort of anxious subjects as a whole, emotional reactivity scores decreased between assessments, with no significant difference in the extent of decrease between the MED and WL groups (i.e., no significant group by time interaction). The results of the repeated ANOVA are shown in Table 4.

Table 4.

Changes in Emotional Reactivity from Baseline to 8 Weeks' Follow-Up in Citalopram-Treated and Untreated (CBT-Waitlisted) Children and Adolescents with Anxiety Disorder

	Time point	Anxious group-medicated (n = 35) mean (SD)	Anxious group waitlist (n = 15) mean (SD)	F time effect	F time × group
Frequency of negative emotional response (REAR-S)	Baseline	9.7 (4.3)	10.8 (3.2)	9.0^**	0.30
Frequency of negative emotional response (REAR-S)	Follow-up	8.7 (3.6)	9.2 (3.7)
Intensity of negative emotional response (REAR-S)	Baseline	3.0 (1.3)	3.6 (1.4)	16.9^***	0.74
Intensity of negative emotional response (REAR-S)	Follow-up	2.4 (1.1)	2.8 (1.2)
Intensity of negative emotional response (REAR-I)	Baseline	3.3 (1.9)	4.0 (1.7)	7.1^*	0.62
Intensity of negative emotional response (REAR-I)	Follow-up	2.8 (1.9)	3.2 (1.6)

p < 0.05, ^** p < 0.01, ^*** p < 0.001.

CBT, cognitive behavioral therapy; REAR-S, Reactivity and Regulation-Situations; REAR-I, Reactivity and Regulation-Images.

Effect of medication on reappraisal

Significant time by group interactions were found for two REAR-I reappraisal parameters, ability and efficacy (Table 5). There was a significantly greater improvement in cued reappraisal ability in the MED than in the WL group. Within-group paired post-hoc analysis of the reappraisal indices showed that the medicated group had a significant increase over time in reappraisal ability on the REAR-I (t[34] = −2.6, p < 0.02) but not on the REAR-S. In the WL group, none of the parameters significantly improved. The only significant change was a decrease in reappraisal efficacy between assessments (t[12] = 2.2, p < 0.05).

Table 5.

Changes in Reappraisal in the Medicated and the Waitlist Groups Between Baseline and Follow-Up Assessments

	Time point	Anxious group-medicated (n = 35) mean (SD)	Anxious group waitlist (n = 15) mean (SD)	F time effect	F time × group
Cued reappraisal ability REAR-S	Baseline	0.85 (0.19)	0.78 (0.19)	1.4	0.0
Cued reappraisal ability REAR-S	Follow-up	0.90 (0.15)	0.83 (0.22)
Cued reappraisal ability REAR-I	Baseline	0.86 (0.17)	0.84 (0.12)	0.09	5.3^*
Cued reappraisal ability REAR-I	Follow-up	0.93 (.08)	0.79 (0.17)
Uninstructed reappraisal REAR-S	Baseline	0.21 (0.18)	0.25 (0.23)	3.0	0.1
Uninstructed reappraisal REAR-S	Follow-up	0.29 (0.24)	0.31 (0.32)
Reappraisal efficacy^† REAR-I	Baseline	0.30 (0.23)	0.39 (0.23)	3.9	4.6^*
Reappraisal efficacy^† REAR-I	Follow-up	0.31 (0.17)	0.22 (0.25)
Reappraisal self-efficacy REAR-S	Baseline	0.66 (0.17)	0.70 (0.23)	0.72	1.6
Reappraisal self-efficacy REAR-S	Follow-up	0.75 (0.20)	0.69 (0.24)
Everyday use of reappraisal ERQ	Baseline	4.2 (1.2)	4.0 (1.9)	0.36	1.1
Everyday use of reappraisal ERQ	Follow-up	4.0 (1.5)	4.4 (1.2)

p < 0.05

†

Percentage of decrease in negative emotional response between view and reappraisal.

CBT, cognitive behavioral therapy; REAR-S, Reactivity and Regulation-Situations; REAR-I, Reactivity and Regulation-Images; ERQ, Emotion Regulation Questionnaire.

Association between changes in clinical and emotional indices

Pearson tests showed moderate but significant correlations of the extent of increase in reappraisal ability on the REAR-I with a greater extent of overall clinical improvement between assessments (r[35] = −0.30 p < 0.04) and a decrease in anxiety (r[34] = −0.38 p < 0.02) in the MED group.

Prompted by our findings of a decrease in negative emotional reactivity in the subjects with anxiety as a whole, we examined the possible relationship of these changes with the clinical improvement. Our analysis showed that the decrease in the intensity of emotional response on the REAR-S significantly correlated with the extent of decrease in anxiety on the SCARED-C (r[48] = 0.44, p < 0.001) and with overall clinical improvement (r[50] = 0.27, p < 0.05). There was also a borderline significant correlation between the decrease in intensity of the negative emotional response on the REAR-I and the decrease in anxiety (r[47] = 0.25, p = 0.06).

Baseline measures and emotional reactivity as predictors of clinical outcome

The principal components analyses showed that among the anxious subjects, all three anxiety measures loaded on the same component at both baseline (loadings: Parent 0.58, child 0.89, clinician 0.89) and follow-up (parent 0.65, child 0.91, clinician 0.90). The first component explained the majority of the variance in the three measures (64% at baseline, 69% at follow-up), and no other component explained >30% of the variance. The three indices combined into an internally consistent measure at both baseline (α = 0.70) and follow-up (α = 0.76).

After calculating the composite scores, we examined whether the baseline characteristics could predict symptoms at follow-up after controlling for treatment group and whether baseline characteristics interacted with treatment group to predict follow-up symptoms. We found that both baseline symptoms (t[46] = 4.7, p < 0.001) and treatment (t[46] = 3.1, p = 0.004) were highly significant predictors of follow-up symptoms. However, treatment and baseline symptoms did not interact (t[46] = 0, p > 0.5).

A primary question of this study was whether measures of emotional reactivity and regulation at baseline would predict follow-up symptoms. The REAR-S negative emotional reactivity was significantly related to follow-up symptoms after controlling for baseline symptoms and treatment group, such that individuals with higher intensity of baseline reactivity had more severe symptoms at follow-up (frequency: t[44] = 2.5, p < 0.05; intensity: t[44] = 2.4, p < 0.05). However, REAR-I intensity score was not related to follow-up symptoms (t[44] = 1.2, p = 0.24). In addition, neither demographic variables (age, gender, cognitive score) nor reappraisal ability and efficacy were found to be predictors of symptoms on follow-up (p > 0.1 for all). They also did not interact with treatment (p > 0.1 for all).

Characteristics of medicated subjects who achieved remission

The subjects who met the criteria for remission after receiving SSRIs for 8 weeks had lower baseline clinician ratings of overall illness than nonremitters (t[33] = −2.5, p < 0.02). There were no significant differences between these subgroups for any of the other clinical parameters. Remitters also had a lower intensity and frequency of negative emotional reactivity on the REAR-S than nonremitters (frequency: t[33] = −3.3, p < 0.003; intensity: t[33] =−3.4, p < 0.003).

Interestingly, on ANOVA with baseline clinician rating as the covariate, the differences in negative emotional reactivity between remitters and nonremitters remained significant (frequency: F[1,35] = 7.7, p < 0.01; intensity: F[1,35] = 5.9, p < 0.03). However, when baseline negative emotional negativity score was the covariate, the differences between remitters and nonremitters in clinician ratings of overall severity lost their significance, suggesting that the distinctive trait at baseline between remitters and nonremitters was negative emotional reactivity.

Discussion

This pilot study sought to investigate changes in emotional reactivity and regulation in children and adolescents treated with citalopram, a widely used SSRI, for anxiety disorder. Our major finding, although preliminary, was a significantly greater increase in cued reappraisal ability over time in the MED group than in the WL group. Therefore, SSRI treatment seems to be associated with improvement in at least one behavioral aspect of emotion regulation: The ability to provide alternative explanations of threatening visual stimuli.

We also found a significant correlation between the extent of increase in reappraisal ability on the REAR-I from baseline to follow-up and the extent of clinical improvement, as manifested by scores on measures of anxiety, depression, and overall illness severity. Reappraisal in the REAR-I task was more sensitive to SSRI medication than reappraisal in the REAR-S task. This finding might be explained by the involvement of different neural pathways in the different tasks. The REAR-I is a visual task and may activate more “bottom-up” brain processes whereas the REAR-S is a situation-related task and may activate more “top-down” processes associated with regulatory activity (Ochsner et al. 2004). There is some biological evidence to support this supposition (Ochsner et al. 2009; Faria et al. 2012; Outhred et al. 2013; Phan et al. 2013).

We expected that the MED group would also show greater decreases in negative emotional reactivity than the WL group. However, although negative emotional reactivity decreased significantly in the MED group between baseline and follow-up, it also decreased in the WL group. Moreover, we found correlations between the decrease in the intensity of negative emotional response and the improvement in anxiety symptoms in both anxious groups. Therefore, the decrease in negative emotional reactivity over time may be associated with attenuation of anxiety and does not appear to be specific to SSRI medication. The lack of statistically significant effect between the MED and WL groups in child self-reported anxiety underline the multifactorial complexity of an anxiety disorder in children and adolescents. It is of note that in the MED group, parent self-reports showed significant improvement in their children's anxiety.

A number of studies have examined the effects of medication on emotional reactivity in anxiety and, to an even greater extent, in depression. Many focused on changes in the reactivity of the amygdala to negative emotional stimuli, usually facial images (Sheline et al. 2001; Drevets et al. 2002; Fu et al. 2004; Bloch et al. 2015). Notably, one study of anxious subjects, using a similar design to ours, examined possible differences in functional changes in medicated children with anxiety disorder and controls (Whalen et al. 2008). The authors found no association of medication with the decrease in amygdala reactivity over time, thereby providing biological support for our psychological finding that medication does not have a specific effect on emotional reactivity.

Other important findings of this study included an association of higher negative emotional reactivity on the REAR-S at baseline with more severe clinical symptoms at follow-up, even after controlling for baseline symptoms and treatment group. Moreover, secondary analyses of remitters and nonremitters in the medicated group yielded a difference only in negative emotional reactivity at baseline. This may suggest that susceptibility to higher negative emotional reactivity may serve as a prognostic sign beyond overt symptoms of anxiety. However, there was no such association in the REAR-I. Therefore, the different results on the two tests need to be further investigated.

Limitations

The major limitation of the present study was the small number of participants, especially in the WL group. As it was an open trial, participants were not randomly assigned to citalopram versus waitlist. However, the two study groups did not statistically differ in baseline scores (SCARED, CDI/BDI-II, STAI-C, CGI-S). In addition, CGI-S and CGI-I were administrated by two child psychologists who could not be blind to study condition. Therefore, there might have been an expectancy bias toward those on citalopram compared with those on the waitlist.

Nevertheless, the major strengths of this study are its prospective design and use of online measures of emotional reactivity and regulation.

Conclusions

There is an overall improvement in negative emotional reactivity measures over time. Nevertheless, citalopram therapy specifically improves reappraisal ability on the REAR-I in children with anxiety, and the extent of increase in reappraisal ability correlated with extent of clinical improvement. In addition, The REAR-S negative emotional reactivity at baseline was significantly related to follow-up symptoms.

Clinical Significance

We believe that our findings have some implications for the assessment and treatment of anxiety disorders in childhood and adolescence. In terms of assessment, they suggest that a differentiated and detailed evaluation of both emotional reactivity and regulatory processes may provide a useful means of understanding patients' difficulties (Rottenberg and Gross 2003). Negative emotional reactivity seems to be not only a predictor of the severity of anxiety symptoms, but may also be a predictor of prognosis. Therefore, the REAR-I and REAR-S tasks may serve as an indirect measure of the clinical severity of anxiety disorders in children and adolescents. In terms of treatment, replication of our finding that remitters have below cutoff scores at baseline may suggest a potential role for baseline negative emotional reactivity in identifying children and adolescents with anxiety disorders who may or may not be suitable for SSRI treatment. Further, our findings suggest a distinction between less directly controlled aspects of anxiety (i.e., negative emotional hyperreactivity) and aspects that may be more subject to volition (e.g., emotion regulation). This could help patients become more aware of their emotional reactivity and better at identifying potential triggers (objects or situations) of an intense negative emotional response. Patients might also be taught emotion regulation strategies that they can adapt to particular situations. For example, reappraisal may be a better choice when emotional intensity is not overwhelming.

The tasks administered here address automatic thoughts in response to specific external, emotionally provocative stimuli presented in an interactive computerized setting. Recreating similar conditions in clinical settings may provide anxious children and adolescents the opportunity to improve their ability to change their automatic appraisals for eventual application to threatening situations encountered in daily life.

Footnotes

Acknowledgments

The authors thank Charlotte Sachs and Gloria Ginzach (Editorial Board, Rabin Medical Centre, Israel) and Doc Edge (Department of Biology Stanford University) for their assistance.

Disclosures

No competing financial interests exist.

References

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC: American Psychiatric Association; 1994.

Beck

, Steer

, Ball

, Ranieri

: Comparison of Beck Depression Inventories -IA and -II in psychiatric outpatients. J Pers Assess, 67:588–597, 1996.

Birmaher

, Khetarpal

, Brent

, Cully

, Balach

, Kaufman

, Neer

: The Screen for Child Anxiety Related Emotional Disorders (SCARED): Scale construction and psychometric characteristics. J Am Acad Child Adolesc Psychiatry, 36:545–553, 1997.

Blechert

, Sheppes

, Di Tella

, Williams

, Gross

: See what you think: Reappraisal modulates behavioral and neural responses to social stimuli. Psychol Sci, 23:346–353, 2012.

Bloch

, Aviram

, Braw

, Gvirts

, Rabany

, Walter

: Attention improves after clinical improvement in acutely depressed adolescents. J Neuropsychiatry Clin Neurosci, 27:153–157, 2015.

Bridge

, Iyengar

, Salary

, Barbe

, Birmaher

, Pincus

, Ren

, Brent

: Clinical response and risk for reported suicidal ideation and suicide attempts in pediatric antidepressant treatment: A meta-analysis of randomized controlled trials. JAMA, 297:1683–1696, 2007.

Bylsma

, Morris

, Rottenberg

: A meta-analysis of emotional reactivity in major depressive disorder. Clin Psychol Rev, 28: 676–691, 2008.

Carthy

, Horesh

, Apter

, Edge

, Gross

: Emotional reactivity and cognitive regulation in anxious children. Behav Res Ther, 48:384–393, 2010a.

Carthy

, Horesh

, Apter

, Gross

: Patterns of emotional reactivity and regulation in children with anxiety disorders. J Psychopathol Behav Assess. 23–36, 2010b.

10.

Cicchetti

, Toth

: A developmental psychopathology perspective on child abuse and neglect. J Am Acad Child Adolesc Psychiatry, 34:541–565, 1995.

11.

Davidson

: Affective style and affective disorders: Perspectives from affective neuroscience. Cogn Emot, 12:307–330, 1998.

12.

Davidson

, Pizzagalli

, Nitschke

, Putnam

: Depression: Perspectives from affective neuroscience. Annu Rev Psychol, 53:545–574, 2002.

13.

Domes

, Schulze

, Bottger

, Grossmann

, Hauenstein

, Wirtz

, Heinrichs

, Herpertz

: The neural correlates of sex differences in emotional reactivity and emotion regulation. Hum Brain Mapp, 31:758–769, 2010.

14.

Drevets

, Bogers

, Raichle

: Functional anatomical correlates of antidepressant drug treatment assessed using PET measures of regional glucose metabolism. Eur Neuropsychopharmacol, 12:527–544, 2002.

15.

Faria

, Ahs

, Appel

, Linnman

, Bani

, Bettica

, Pich

, Wahlstedt

, Fredrikson

, Furmark

: Amygdala-frontal couplings characterizing SSRI and placebo response in social anxiety disorder. Int J Neuropsychopharmacol, 17:1149–1157, 2014.

16.

Faria

, Appel

, Ahs

, Linnman

, Pissiota

, Frans

, Bani

, Bettica

, Pich

, Jacobsson

, Wahlstedt

, Fredrikson

, Furmark

: Amygdala subregions tied to SSRI and placebo response in patients with social anxiety disorder. Neuropsychopharmacology, 37:2222–2232, 2012.

17.

Fisher

, Meltzer

, Price

, Coleman

, Ziolko

, Becker

, Moses–Kolko

, Berga

, Hariri

: Medial prefrontal cortex 5-HT(2A) density is correlated with amygdala reactivity, response habituation, and functional coupling. Cereb Cortex, 19:2499–2507, 2009.

18.

, Williams

, Cleare

, Brammer

, Walsh

, Kim

, Andrew

, Pich

, Williams

, Reed

, Mitterschiffthaler

, Suckling

, Bullmore

: Attenuation of the neural response to sad faces in major depression by antidepressant treatment: A prospective, event-related functional magnetic resonance imaging study. Arch Gen Psychiatry, 61:877–889, 2004.

19.

Furmark

, Appel

, Michelgard

, Wahlstedt

, Ahs

, Zancan

, Jacobsson

, Flyckt

, Grohp

, Bergstrom

, Pich

, Nilsson

, Bani

, Langstrom

, Fredrikson

: Cerebral blood flow changes after treatment of social phobia with the neurokinin-1 antagonist GR205171, citalopram, or placebo. Biol Psychiatry, 58:132–142, 2005.

20.

Furmark

, Tillfors

, Marteinsdottir

, Fischer

, Pissiota

, Langstrom

, Fredrikson

: Common changes in cerebral blood flow in patients with social phobia treated with citalopram or cognitive-behavioral therapy. Arch Gen Psychiatry, 59:425–433, 2002.

21.

Ginsburg

, Becker

, Keeton

, Sakolsky

, Piacentini

, Albano

, Compton

, Iyengar

, Sullivan

, Caporino

, Peris

, Birmaher

, Rynn

, March

, Kendall

: Naturalistic follow-up of youths treated for pediatric anxiety disorders. JAMA Psychiatry, 71:310–318, 2014.

22.

Ginsburg

, Kendall

, Sakolsky

, Compton

, Piacentini

, Albano

, Walkup

, Sherrill

, Coffey

, Rynn

, Keeton

, McCracken

, Bergman

, Iyengar

, Birmaher

, March

: Remission after acute treatment in children and adolescents with anxiety disorders: Findings from the CAMS. J Consult Clin Psychol, 79:806–813, 2011.

23.

Goldin

, Manber

, Hakimi

, Canli

, Gross

: Neural bases of social anxiety disorder: Emotional reactivity and cognitive regulation during social and physical threat. Arch Gen Psychiatry, 66:170–180, 2009.

24.

Gross

: Antecedent- and response-focused emotion regulation: divergent consequences for experience, expression, and physiology. J Pers Soc Psychol, 74:224–237, 1998.

25.

Gross

: Emotion regulation: Affective, cognitive, and social consequences. Psychophysiology, 39:281–291, 2002.

26.

Gross

: Emotion regulation: Taking stock and moving forward. Emotion, 13:359–365, 2013.

27.

Gross

, John

: Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. J Pers Soc Psychol, 85:348–362, 2003.

28.

Guy

: ECDEU Assessment Manual for Psychopharmacology. Rockville MD: 1976.

29.

James

, Soler

, Weatherall

: Cognitive behavioural therapy for anxiety disorders in children and adolescents. Cochrane Database Syst Rev, 4:CD004690, 2005.

30.

Lang

, Bradley

, Cuthbert

: International Affective Picture System (IAPS): Instruction Manual and Affective Ratings. Gainesville, FL: The Center for Research in Psychophysiology, University of Florida; 1999.

31.

Liddell

, Brown

, Kemp

, Barton

, Das

, Peduto

, Gordon

, Williams

: A direct brainstem–amygdala–cortical ‘alarm’ system for subliminal signals of fear. Neuroimage, 24:235–243, 2005.

32.

Llewellyn

, Dolcos

, Iordan

, Rudolph

, Dolcos

: Reappraisal and suppression mediate the contribution of regulatory focus to anxiety in healthy adults. Emotion, 13:610–615, 2013.

33.

Lonigan

, Vasey

, Phillips

, Hazen

: Temperament, anxiety, and the processing of threat-relevant stimuli. J Clin Child Adolesc Psychol, 33:8–20, 2004.

34.

Lyneham

, Rapee

: Evaluation and treatment of anxiety disorders in the general pediatric population: A clinician's guide. Child Adolesc Psychiatr Clin N Am, 14:845–861, 2005.

35.

Mayer

, Salovey

, Caruso

, Sitarenios

: Measuring emotional intelligence with the MSCEIT V2.0. Emotion, 3:97–105, 2003.

36.

McLaughlin

, Mennin

, Farach

: The contributory role of worry in emotion generation and dysregulation in generalized anxiety disorder. Behav Res Ther, 45:1735–1752, 2007.

37.

Ochsner

, Ray

, Cooper

, Robertson

, Chopra

, Gabrieli

, Gross

: For better or for worse: Neural systems supporting the cognitive down- and up-regulation of negative emotion. Neuroimage, 23: 483–499, 2004.

38.

Ochsner

, Ray

, Hughes

, McRae

, Cooper

, Weber

, Gabrieli

, Gross

: Bottom-up and top-down processes in emotion generation: Common and distinct neural mechanisms. Psychol Sci, 20:1322–1331, 2009.

39.

Ochsner

, Silvers

, Buhle

: Functional imaging studies of emotion regulation: A synthetic review and evolving model of the cognitive control of emotion. Ann N Y Acad Sci, 1251:E1–24, 2012.

40.

Osman

, Kopper

, Barrios

, Gutierrez

, Bagge

: Reliability and validity of the Beck depression inventory–II with adolescent psychiatric inpatients. Psychol Assess, 16:120–132, 2004.

41.

Outhred

, Hawkshead

, Wager

, Das

, Malhi

, Kemp

: Acute neural effects of selective serotonin reuptake inhibitors versus noradrenaline reuptake inhibitors on emotion processing: Implications for differential treatment efficacy. Neurosci Biobehav Rev, 37:1786–1800, 2013.

42.

Papay

, Spielberger

: Assessment of anxiety and achievement in kindergarten and first- and second-grade children. J Abnorm Child Psychol, 14:279–286, 1986.

43.

Phan

, Coccaro

, Angstadt

, Kreger

, Mayberg

, Liberzon

, Stein

: Corticolimbic brain reactivity to social signals of threat before and after sertraline treatment in generalized social phobia. Biol Psychiatry, 73:329–336, 2013.

44.

Piacentini

, Bennett

, Compton

, Kendall

, Birmaher

, Albano

, March

, Sherrill

, Sakolsky

, Ginsburg

, Rynn

, Bergman

, Gosch

, Waslick

, Iyengar

, McCracken

, Walkup

: 24- and 36-week outcomes for the Child/Adolescent Anxiety Multimodal Study (CAMS). J Am Acad Child Adolesc Psychiatry, 53:297–310, 2014.

45.

Ray

, McRae

, Ochsner

, Gross

: Cognitive reappraisal of negative affect: Converging evidence from EMG and self-report. Emotion, 10:587–592, 2010.

46.

Ray

, Wilhelm

, Gross

: All in the mind's eye? Anger rumination and reappraisal. J Pers Soc Psychol, 94:133–145, 2008.

47.

Rottenberg

, Gross

: When emotion goes wrong: Realizing the promise of affective science. Clin Psychol Sci Pract, 10: 227–232, 2003.

48.

Rottenberg

, Kasch

, Gross

, Gotlib

: Sadness and amusement reactivity differentially predict concurrent and prospective functioning in major depressive disorder. Emotion, 2:135–146, 2002.

49.

Rottenberg

, Wilhelm

, Gross

, Gotlib

: Vagal rebound during resolution of tearful crying among depressed and nondepressed individuals. Psychophysiology, 40:1–6, 2003.

50.

Shani

, Lachman

, Shalem

, Bahat

, Zeiger

: Single Words Reading Test Manual. Tel-Aviv: Sfarim; 2006.

51.

Sheline

, Barch

, Donnelly

, Ollinger

, Snyder

, Mintun

: Increased amygdala response to masked emotional faces in depressed subjects resolves with antidepressant treatment: An fMRI study. Biol Psychiatry, 50:651–658, 2001.

52.

Silverman

, Saavedra

, Pina

: Test–retest reliability of anxiety symptoms and diagnoses with the Anxiety Disorders Interview Schedule for DSM-IV: Child and parent versions. J Am Acad Child Adolesc Psychiatry, 40:937–944, 2001.

53.

Smucker

, Craighead

, Green

: Normative and reliability data for the Children's Depression Inventory. J Abnorm Child Psychol, 14:25–39, 1986.

54.

Spielberger

: Manual for the State-Trait Anxiety Inventory for Children. Palo-Alto, CA: Consulting Psychologists Press; 1973.

55.

Suveg

, Zeman

: Emotion regulation in children with anxiety disorders. J Clin Child Adolesc Psychol, 33:750–759, 2004.

56.

Szeszko

, MacMillan

, McMeniman

, Lorch

, Madden

, Ivey

, Banerjee

, Moore

, Rosenberg

: Amygdala volume reductions in pediatric patients with obsessive-compulsive disorder treated with paroxetine: preliminary findings. Neuropsychopharmacology, 29:826–832, 2004.

57.

Walkup

, Albano

, Piacentini

, Birmaher

, Compton

, Sherrill

, Ginsburg

, Rynn

, McCracken

, Waslick

, Iyengar

, March

, Kendall

: Cognitive behavioral therapy, sertraline, or a combination in childhood anxiety. N Engl J Med, 359:2753–2766, 2008.

58.

Wechsler

: Wechsler Intelligence Scale for Children–IV. San Antonio: Psychological Corporation; 2003.

59.

Whalen

, Johnstone

, Somerville

, Nitschke

, Polis

, Alexander

, Davidson

, Kalin

: A functional magnetic resonance imaging predictor of treatment response to venlafaxine in generalized anxiety disorder. Biol Psychiatry, 63:858–863, 2008.

60.

Zeman

, Cassano

, Perry–Parrish

, Stegall

: Emotion regulation in children and adolescents. J Dev Behav Pediatr, 27:155–168, 2006.