Somatic Symptoms in Children and Adolescents With

Anxiety Disorders

GOLDA S. GINSBURG, PH.D., MARK A. RIDDLE, M.D., AND MARK DAVIES, M.P.H.

ABSTRACT
Objective: To evaluate the prevalence of somatic symptoms (SSs) in children and adolescents with anxiety disorders; the
relationship between SSs and anxiety severity, impairment, and child global functioning; and the impact of fluvoxamine
(FLV) versus pill placebo (PBO) on reducing SSs. Method: As part of a double-blind, placebo-controlled trial, 128 children
(mean age, 10.8 years; range, 6Y17) with DSM-IV anxiety disorders (i.e., social, separation, and generalized anxiety) were
assessed by expert clinicians on 16 SSs using the Pediatric Anxiety Rating Scale. Results: The most common SSs at
baseline were restlessness (74%), stomachaches (70%), blushing (51%), palpitations (48%), muscle tension (45%),
sweating (45%), and trembling/shaking (43%). Older children (age 12 and older) reported more SSs than younger
children, boys and girls reported similar numbers of SSs, and SSs were higher among children with than without
generalized anxiety disorder. SSs were significantly and positively correlated with anxiety severity, impairment, and global
functioning. Pre-/postreductions in SSs were statistically significant in both PBO and FLV conditions; however, FLV was
superior to PBO in reducing SSs. Conclusions: SSs are highly prevalent among children and adolescents with anxiety
disorders and are associated with greater anxiety severity and impairment. Treatment with FLV was effective in reducing
rather than increasing SSs. The high rates of SSs in youths with each of the three anxiety disorders suggest a re-
evaluation of SSs in the DSM-IV diagnostic criteria for the most common anxiety disorders among children and
adolescents. J. Am. Acad. Child Adolesc. Psychiatry, 2006;45(10):1179Y1187. Key Words: anxiety disorders, somatic
symptoms, fluvoxamine.

Accepted May 2, 2006.

Somatic symptoms (SSs), such as muscle tension and
Drs. Ginsburg and Riddle are with the Division of Child and Adolescent stomachaches, are hallmark clinical features of anxiety
Psychiatry, Department of Psychiatry, Johns Hopkins University School of disorders in children and frequently interfere with their
Medicine; and Dr. Davies is with Columbia University and New York State
Psychiatric Institute. daily functioning. For instance, SSs are often a reason for
The authors acknowledge the Research Units in Pediatric Psychopharmacology children_s absence from school, calls to pediatricians, and,
(RUPP) Anxiety Group sites and NIMH, which supported the data collection for in more extreme cases, visits to hospital emergency
this study: John T. Walkup, M.D., and Michael J. Labellarte, M.D., Johns
Hopkins University; Daniel S. Pine, M.D., Laurence Greenhill, M.D., Rachel departments. Clinical observations of the link between SSs
Klein, Ph.D., and Michael Sweeney, Ph.D., Columbia University and New and anxiety symptoms in youths have been supported by
York State Psychiatric Institute; Howard Abikoff, Ph.D., Sabine Hack, M.D.,
empirical research. Specifically, somatic complaints have
and Brian Klee, M.D., New York University; James McCracken, M.D., Lindsey
Bergman, Ph.D., and John Piacentini, Ph.D., University of California, Los been found to be more common among children with
Angeles; John March, M.D., M.P.H., and Scott Compton, Ph.D., Duke anxiety disorders relative to their nonanxious peers (Beidel
University; and Ben Vitiello, M.D., NIMH.
Correspondence to Dr. Golda S. Ginsburg, Division of Child and Adolescent
et al., 1991) and among anxious and/or depressed youths
Psychiatry, Johns Hopkins Children_s Center, 600 North Wolfe Street, CSMC compared with youths with other mental disorders (Masi
340, Baltimore, MD 21287-3325; e-mail: [email protected]. et al., 2000). Somatic symptoms also appear to have a
0890-8567/06/4510Y1179Ó2006 by the American Academy of Child
and Adolescent Psychiatry.
unique and independent association with anxiety relative
DOI: 10.1097/01.chi.0000231974.43966.6e to depression. In a study comparing SSs among adolescent

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006 1179

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
GINSBURG ET AL.

psychiatric inpatients diagnosed with major depressive Despite the accumulating data linking SSs and
disorder alone (n = 32) or major depressive disorder and childhood anxiety in clinic and community samples,
comorbid anxiety (n = 64), significantly higher levels of replication is needed and additional critical questions await
SSs were found among depressed teens with comorbid examination. For instance, few studies have examined the
anxiety (Jolly et al., 1994). frequency and presentation of individual SSs. Moreover,
Although SSs have been consistently linked to anxiety few data exist establishing the association between SSs and
disorders, there are inconsistent findings as to whether overall anxiety severity and/or impairment. Finally, no
specific SSs (e.g., stomachaches, muscle tension) vary as a study has yet to assess the impact of pharmacological
function of specific demographic (e.g., age, gender), or treatment for anxiety disorders on SSs. Research on these
clinical characteristics (e.g., type of anxiety disorder, overall critical questions is important because it has the potential
severity). One of the earliest and largest studies to examine to inform assessment, treatment, and estimates for
some of these issues was conducted by Last (1991). In this prognosis. For instance, to the extent that SSs are related
study, 158 clinic-referred youths with anxiety disorders to anxiety severity and treatment response, additional
completed a comprehensive assessment of various clinical monitoring and intervention may be needed to target these
characteristics. Children were categorized as somatic or symptoms. Alternatively, if psychopharmacological inter-
nonsomatic based on their responses to a semistructured ventions have minimal impact on SSs, their use for
diagnostic interview (i.e., the Schedule for Affective reducing SSs among youths may need to be re-evaluated.
Disorders and Schizophrenia for School-Age Children The present study attempted to address these important
[K-SADS]). Sixty percent (95 of 158) of these children issues via three specific aims. The first aim was to replicate
were categorized as somatic, reporting at least one clinically previous studies by examining the frequency of a broad
significant SS (no information was provided on the specific range of SSs among a large clinical sample of children with
SS). When SSs were examined by type of anxiety disorder, anxiety disorders using clinician assessments. Specifically,
the authors found that SSs were reported by 100% of
the presence of 16 SSs were examined in relation to age,
youths with panic disorder (n = 16), 78% with separation
gender, and type of anxiety disorder (i.e., GAD, SAD,
anxiety disorder (SAD; n = 40), 56% with overanxious
SOP). The second aim examined whether SSs were asso-
disorder (n = 18), 44% with obsessive-compulsive disorder
ciated with overall anxiety severity and impairment (e.g.,
(n = 9), and 43% for those with phobic disorders (social
interference in home, personal distress, avoidance, global
phobia [SOP], n = 54). SSs were less common among
functioning). The third aim examined the impact of
younger (younger than age 13) compared with older
children, and in contrast to previous studies (Bernstein fluvoxamine (FLV), a selective serotonin reuptake inhibitor
et al., 1989; Livingston et al., 1988), no gender differences (SSRI), versus pill placebo (PBO) on each of the 16 SSs.
were found in the reporting of SSs.
METHOD
In a more recent study using a population-based
sample (N = 3,733) of youth ages 9 to 16 from the Great Participants
Smoky Mountains Study (Costello et al., 1996; Egger
Participants were enrolled in a double-blind, placebo-controlled
et al. 1999) examined the presence of SSs (specifically clinical trial of FLV for youth with SOP, SAD, and/or GAD (Research
stomachaches, headaches, and musculoskeletal pains) and Units on Pediatric Psychopharmacology Anxiety Study Group
their relation to SAD and generalized anxiety disorder [RUPP], 2001). Five RUPP sites were involved in the study: Duke
(GAD). SSs and diagnostic status were assessed using the University, Johns Hopkins University, New York State Psychiatric
Institute/Columbia University, New York University, and University
Child and Adolescent Psychiatric Assessment (Angold of California, Los Angeles. Across these sites, 128 children, ages 6 to
et al., 1995). Relevant findings indicated that among 17 years (mean, 10.8 years; 49% female; 63% white; Table 1), with a
girls, SSs were associated with having an anxiety disorder current DSM-IV diagnosis of SOP, SAD, and/or GAD were enrolled.
Diagnoses were determined based on information from both child and
and 60% of girls with an anxiety disorder reported at least parent K-SADS interviews, administered by a trained clinician. These
one SS compared with 12% without an anxiety disorder. clinicians were experienced child and adolescent psychiatrists or psy-
For boys, SSs were generally unrelated to having an anxiety chologists and were blind to treatment condition. Sites recruited
disorder. These findings support other community-based participants from a variety of sources, including mental health settings
and newspaper advertisements. Exclusion criteria included current use
studies that have found an association between somatic of any illicit or prescribed psychoactive substance; current diagnoses of
complaints and anxiety symptoms (Beidel et al., 1991; major depressive disorder, Tourette_s disorder, obsessive-compulsive
Muris and Meesters, 2004). disorder, posttraumatic stress disorder, conduct disorder, or

1180 J. AM. ACAD. CH ILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
 SOMATIC SYMPTOMS AND CHILD ANXIETY

TABLE 1 For the present study, 16 items representing SSs were examined
Demographic and Clinical Characteristics of Entire Sample (items listed in Table 2). Thirteen items were based on criteria listed
(N = 128) in the DSM-IV for panic attacks (10 items) and GAD (3 items). An
additional three items (blushing, recurrent urge to go to the
No. % bathroom, paralyzed feeling) were added by the RUPP investiga-
Age, yr tional team and were based on clinical and research experience. SSs
6Y11 86 67 were rated by a trained clinician based on information from both
12Y17 42 33 child and parent interviews.
K-SADS. Inclusion/exclusion criteria for psychiatric diagnoses were
Female 63 49.2
assessed using the K-SADS (Ambrosini, 2000; Kaufman et al., 1997),
Ethnicity administered by experienced clinicians. The K-SADS possesses good
White 81 63.3 reliability and validity (Ambrosini, 2000). The entire K-SADS
Hispanic 24 18.8 interview was administered before treatment.
Black (non-Hispanic) 9 7.0 Children_s Global Assessment Scale (CGAS). The CGAS (Shaffer
Other 14 10.9 et al., 1983) is a modification of the adult Global Assessment Scale
Total family income and provides a measure of global impairment and functioning. The
G$25,000 19 14.8 scale ranges from 1 (lowest) to 100 (highest). Ample evidence exists
$25,000Y$39,999 20 15.6 supporting the psychometric properties of the scale.
$40,000Y$59,999 18 14.1 Clinician_s Global Impressions-Severity (CGI-S) Scale. The CGI-S
(Guy, 1976) was used to assess global severity of anxiety using a 7-
9$60,000 55 43.0
point scale, ranging from normal to most extreme. This measure was
Refused/unknown 16 12.5 completed before and after treatment to assess global changes in
Diagnosis (based on K-SADS) anxiety severity.
SOP only 26 20.3
SAD only 18 14.1 Procedure
GAD only 5 3.9 Before participating in this study, all of the parents signed informed
SOP and SAD only 11 8.6 consent; children older than age 6 signed assents. A comprehensive
SOP and GAD only 21 16.4 pretreatment evaluation was conducted that included the diagnostic
SAD and GAD only 21 16.4 interview and additional measures to assess inclusion/exclusion criteria
SOP, SAD, and GAD 26 20.3 and primary outcomes. The pretreatment evaluation was completed
during three visits, which spanned a 3- to 4-week period. During these
Note: K-SADS = Schedule for Affective Disorders and Schizo- visits, all measures (described above) were completed. During the treat-
phrenia for School-Age Children; SOP = social phobia; SAD = ment phase, the treating child and adolescent psychiatrist, who was blind
separation anxiety disorder; GAD = generalized anxiety disorder. to treatment condition, saw each child and parent on a weekly basis for
the first 6 weeks of the trial and again at week 8. At each visit, child and
panic disorder; any history of mania, psychosis, or pervasive adolescent psychiatrists completed four tasks: (1) they administered
developmental disorder; current suicidal ideation; mental retardation supportive psychoeducational therapy to children and families, (2) they
as assessed with the Kaufman Brief Intelligence Test (IQ G 70); conducted a clinical assessment of anxiety symptoms, (3) they assessed
previous use of an SSRI in appropriate doses; and a diagnosis of adverse events, and (4) they determined medication dose following a
attention-deficit/hyperactivity disorder that required pharmacologi- flexible, forced dose titration (i.e., a fixed dose escalation schedule was
cal treatment. Table 1 summarizes the characteristics of participants used that could be modified by the clinical investigator in response to
in the entire sample. After complete description of the study to the adverse events).
subjects, written informed consent was obtained.
Statistical Analyses
Measures
For aim 1, t test and # 2 analyses were conducted to compare the
Pediatric Anxiety Rating Scale (PARS). The PARS (Research Units frequency (total score and individual SSs) of SSs across gender, age, and
on Pediatric Psychopharmacology Anxiety Study Group, 2002) is a primary diagnoses. For aim 2, Pearson correlations were conducted
clinician-rated instrument for assessing the severity of anxiety between SSs and anxiety severity (as measured by CGI-S and PARS total
symptoms associated with common DSM-IV anxiety disorders in score) and impairment (as measured by the CGAS and PARS
children ages 6 to 17. The instrument has two sections. The first Impairment items). Aim 3 was tested using t test and # 2 analyses to
section includes a 50-item symptom checklist that the clinician rates examine the frequency of SSs (total score and individual SSs) at baseline
as present or absent within the past week. The second section is and posttreatment within and between groups.
composed of seven severity/impairment items that are rated on a 5-
point Likert scale. A total score was calculated by adding five of the seven RESULTS
severity items. The two severity items omitted from the total score were
total number of symptoms and total number of somatic/physical Aim 1: Descriptive Statistics: Gender, Age, and Primary
symptoms. These items were excluded from the total score in the RUPP
study because of a ceiling effect for total number of symptoms and the Anxiety Disorder
potential overlap of somatic/physical symptoms with side effects of Frequencies of the 16 SSs are presented in Tables 2
SSRIs. Internal consistency, test-retest and interrater reliability, and
validity (convergent, divergent) were found to be acceptable (Research and 3 for the total sample, by gender, age (6Y11 versus
Units on Pediatric Psychopharmacology Anxiety Study Group, 2002). 12Y17 years), and diagnosis (GAD, SAD, and SOP).

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006 1181

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
GINSBURG ET AL.

TABLE 2
Somatic Symptoms by Gender and Age
% Yes
Symptom Total (N = 128) Male (n = 65) Female (n = 63) Age 6Y11 (n = 86) Age Q12 (n = 42)
Restlessness 74.2 82.8 76.2 72.9 78.6
Feels sick to stomach 70.1 72.3 57.1** 72.1 64.3
Blushing 50.8 58.5 52.4 45.9 64.3*
Palpitations 48.0 53.1 47.6 44.2 64.3
Muscle tension 45.2 51.6 42.9 41.7 61.9
Sweating 44.9 39.1 42.9 36.5 52.4**
Trembling or shaking 43.0 38.5 38.1 32.6 52.4***
Easily fatigued 35.2 38.5 37.1 31.8 47.6
Feels paralyzed 31.5 38.1 31.7 30.2 38.1
Chills or hot flashes 31.0 29.7 23.8 29.1 35.7
Difficulty breathing 26.8 28.1 22.6 27.1 34.1
Recurrent urge bathroom 25.2 27.7 22.2 22.4 29.3
Feels dizzy 25.0 26.2 22.2 18.6 26.2*
Chest pain or discomfort 18.8 26.2 12.5* 17.4 23.8
Problems swallowing 17.2 21.5 11.1 12.8 21.4
Paresthesias 14.1 20.0 7.9 9.3 21.4*
Total no. (SD) of SSs 6.0 (3.5) 6.5 (3.7) 5.5 (3.2) 5.4 (3.4) 7.2** (3.6)
Note: SSs = somatic symptoms.
*p G 0.05; **p G 0.01; ***p G 0.001.

Ninety-six percent of the total sample reported at statistically meaningful comparisons of SSs for children
least one SS (only five children reported no SSs), and an with only one anxiety disorder). Children with and
average of six SSs were reported per child. The most without SOP reported a similar number of SSs (t = 0.19,
common SSs in the total sample were restlessness df = 126, p 9 .20). Among youths with SOP (n = 84), the
(74%), stomachaches (70%), blushing (50%), palpita- three most common SSs were restlessness, stomachaches,
tions (48%), muscle tension (45%), sweating (45%), and blushing. Compared with those without a diagnosis
and trembling/shaking (43%). of SOP (n = 44), socially anxious youths reported
Gender. The mean numbers of SSs for girls and boys significantly more sweating and less problems swallowing.
were 5.5 and 6.5, respectively, and were not signifi- Children with and without SAD also reported a similar
cantly different (t = 1.54, df = 126, p G .13). Among the number of SSs (t = 1.21, df = 126, p 9 .20). With respect
16 specific SSs, however, boys reported significantly to the individual 16 SSs, the three most common SSs
more stomachaches (# 2 = 8.79, p G .001) and chest pain among youths with SAD (n = 76) were restlessness,
(# 2 = 3.82, p G .05) compared with girls. stomachaches, and palpitations, and when compared with
Age. Older youths reported a significantly higher those without a diagnosis of SAD (n = 52), these three
number (t = 2.77, df = 126, p G .01) of SSs (mean = 7.2 T symptoms were significantly more common.
3.6) than younger children (mean = 5.4 T 3.4). Among Finally, children with GAD reported a significantly
the 16 individual SSs, older children reported signifi- higher number of SSs compared with those without
cantly more blushing (#2 = 3.79, p G .05), sweating (#2 = GAD (t = 3.59, df = 126, p G .001) and the three most
6.36, p G .01), trembling/shaking (#2 = 10.33, p G .001), common SSs among children with a diagnosis of GAD
feeling dizzy (#2 = 4.72, p G .03), and paresthesias (#2 = (n = 73) were restlessness, stomachaches, and palpita-
3.79, p G .05). tions. Three of the 16 SSs were more common in
Diagnosis. Table 3 displays the mean total number of youths with GAD compared with youths without GAD
SSs and frequencies for each of the 16 SSs for youths with (n = 65): restlessness, stomachaches, and chills or hot
and without SOP, GAD, and SAD (the limited number flashes. Of note, restlessness and stomachaches were the
of children with a single anxiety disorder precluded most frequently reported SSs across all three diagnoses.

1182 J. AM. ACAD. CH ILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
 SOMATIC SYMPTOMS AND CHILD ANXIETY

TABLE 3
Somatic Symptoms (Percentage Yes) by Diagnosis
% Yes
SOPj (n = 44) SOP+ (n = 84) GADj (n = 65) GAD+ (n = 73) SADj (n = 52) SAD+ (n = 76)
Restlessness 75.0 73.8 56.4 87.7*** 63.5 81.6*
Feels sick to stomach 75.0 67.5 53.7 82.2*** 56.9 78.9**
Blushing 43.2 54.8 47.3 53.4 55.8 47.4
Palpitations 52.3 45.8 38.9 54.8 35.3 56.6*
Muscle tension 54.5 40.2 37.0 51.4 46.3 46.7
Sweating 27.3 54.2** 40.0 48.6 51.9 40.0
Trembling or shaking 45.5 41.7 40.0 45.2 48.1 39.5
Easily fatigued 31.8 36.9 29.1 39.7 32.7 36.8
Feels paralyzed 25.0 34.9 32.7 30.6 35.3 28.9
Chills or hot flashes 30.2 31.3 20.0 39.4* 24.0 35.5
Difficulty breathing 27.3 26.5 18.5 32.9 21.6 30.3
Recurrent urge bathroom 25.6 25.0 18.2 30.6 15.4 32.0
Feels dizzy 20.5 27.4 16.4 31.5 25.0 25.0
Chest pain or discomfort 18.2 19.0 10.9 24.7 13.5 22.4
Problem swallowing 27.3 11.9* 10.9 21.9 15.4 18.4
Paresthesias 13.6 14.3 9.1 17.8 17.3 11.8
Total no. (SD) of SSs 5.9 (3.2) 6.0 (3.7) 4.8 (3.3) 6.9*** (3.4) 5.5 (3.3) 6.3 (3.6)
Note: SOP = social phobia; GAD = generalized anxiety disorder; SAD = separation anxiety disorder; SSs = somatic symptoms.
*p G 0.05; **p G 0.01; ***p G 0.001.

Aim 2: Somatic Symptoms and Anxiety Severity significant reductions in both groups. However, there
and Impairment was a significantly larger reduction in the total number of
Pearson correlations were conducted between SSs SSs as well as 6 of the 16 SSs in the FLV versus PBO group
and overall anxiety severity (as measured by CGI-S and (i.e., blushing, palpitations, muscle tension, sweating,
PARS total score) and impairment, as measured by the feeling paralyzed, chills or hot flashes).
CGAS and PARS Impairment items of Avoidance (of
anxiety-provoking situations), Interference at Home, DISCUSSION
and Interference Out of Home (i.e., at school, with The present study examined SSs among children and
peers, other activities) at baseline. Correlations were adolescents with anxiety disorders (i.e., SOP, SAD,
positive and statistically significant between SSs and the GAD). Three specific questions were explored: (1)
PARS total score (r = 0.34; p G .001), CGI-S (r = 0.32; What were the most common SSs among anxious
p G .001); Interference at Home (r = 0.22; p G .05); and youths and did SSs differ by age, gender, and anxiety
Avoidance (r = 0.18; p G .05). SSs were significantly and diagnosis? (2) What was the relationship between SSs,
negatively correlated with the CGAS (r = j0.27; p G .01). anxiety severity, impairment, and global child func-
SSs were unrelated to Interference Out of the Home/Peers tioning? (3) What was the impact of psychopharmaco-
(r = 0.12; p 9 .05). logical treatment on SSs? Findings for each of these
aims are discussed below, along with limitations and
Aim 3: SSRI versus PBO and Somatic Symptoms implications of the study_s findings.
Mean total scores and percentages of SSs reported at
baseline and posttreatment (using the last observation Aim 1: Prevalence of Somatic Symptoms
carried forward) for youths in the FLV and PBO groups Virtually every child with an anxiety disorder in the
are presented in Table 4. Comparisons at baseline revealed present study reported at least one SS and the majority of
no differences between the groups on the total number of children reported more than one. This finding suggests
SSs or any of the 16 specific SSs. At posttreatment, the that SSs are a key feature of anxiety disorders in children
total number, and several individual SSs, showed and highlights the need to assess for and treat SSs on a

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006 1183

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
GINSBURG ET AL.

TABLE 4
Somatic Symptoms by Treatment Group, Time
Baseline Posttreatment Pre-/Post-Changea Luvox vs.
Symptom (n = 63) Luvox (n = 65) Placebo Luvox Placebo Luvox Placebo Placebob (Post)
Restlessness 73.0 75.4 39.3 50.8 *** *** V
Feels sick to stomach 66.1 73.8 26.2 40.3 *** *** V
Blushing 50.8 50.8 18.0 35.5 *** *
Palpitations 50.8 49.2 16.4 35.5 *** * *
Muscle tension 46.8 47.7 14.8 34.4 *** *
Sweating 44.4 45.3 13.1 32.3 *** **
Trembling or shaking 38.1 40.0 13.1 30.6 *** * V
Easily fatigued 33.8 38.5 13.1 27.9 V
Feels paralyzed 32.3 32.3 11.7 25.8 *** **
Chills or hot flashes 30.6 30.8 9.8 14.5 *** **
Difficulty breathing 29.5 27.7 9.8 11.3 * * V
Recurrent urge bathroom 28.6 23.1 8.2 9.7 * V
Feels dizzy 22.6 23.1 6.6 8.1 ** * V
Chest pain or discomfort 17.5 21.5 4.9 6.5 * ** V
Problems swallowing 15.9 20.0 4.9 3.2 * V
Paresthesias 11.1 12.3 1.6 2.1 * V
Total score
Mean 5.9 6.1 2.0 3.9 *** *** ***
SD 3.6 3.4 2.3 3.1
a
Significance level of change in rates from pre- to posttreatment
b
Significance level for difference in rates at posttreatment.
*p G .05; **p G .01; ***p G .001.

routine basis among clinically anxious youth. High rates criteria for panic disorder (PD; palpitations, sweating,
of SSs have been found in previous studies of clinically and trembling/shaking), and two are currently included
anxious youth. For instance, Last (1991) found that 60% as criteria for GAD (restlessness, muscle tension). Inter-
of anxious youths reported Bclinically significant[ estingly, both restlessness and stomachaches were the
somatic complaints and Masi et al. (2000) found 69% most common SSs among youths across each anxiety
of psychiatrically ill youths reported at least one SS. disorder (GAD, SOP, and SAD), suggesting that these
Although the prevalence rates of SSs in the current study may be key somatic features of all three anxiety disorders.
were somewhat higher, these differences were likely Blushing, often associated with SOP, and stomachaches,
caused by variations in the methods used (e.g., in- common to all of the anxiety disorders, are not part of the
formants, time frame) and the specific SS assessed. In current diagnostic criteria for any anxiety disorder.
comparison, somatic complaints in the general pediatric Few SSs differentiated the anxiety disorders. Youths
population have been estimated to range between 4% with SAD (versus without SAD) reported significantly
and 11% for young children and between 5% and 30% more restlessness, stomachaches, and palpitations with
among adolescents (Dhossche et al., 2002). Future rates as high as 81%, yet these symptoms are not part of
research is needed comparing clinically anxious youths the diagnostic criteria for SAD. Similarly, youth with
with community controls, using the same measures of SOP reported more sweating than their anxious peers
SSs, to clarify the magnitude of differences in prevalence without SOP. Again, these SSs were common, yet are not
rates of SSs. part of diagnostic criteria. Finally, youths with GAD
Among the 16 SSs assessed in the present study, the reported more restlessness, stomachaches, and chills or
most common, experienced by Q40% of all children, hot flashes than youths without GAD. Although
were restlessness, stomachaches, blushing, palpitations, restlessness is part of the GAD criteria, none of the
muscle tension, sweating, and trembling/shaking. Three other symptoms are, despite being highly prevalent
of these symptoms are consistent with current DSM-IV among this group. Interestingly, as part of a larger study

1184 J. AM. ACAD. CH ILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
 SOMATIC SYMPTOMS AND CHILD ANXIETY

on pediatric GAD, Kendall and Pimentel (2003) and sweating, are related to increases in social anxiety.
reported the prevalence rates for restlessness to be 33% Additional research is needed to evaluate these hypotheses.
(child report) and 64% (parent report) and muscle aches
to be 15% (child report) and 50% (parent report), Aim 2: Somatic Symptoms and Anxiety Severity
suggesting that prevalence rates may depend on infor- and Impairment
mant. Taken together, the high number of SSs reported
The second aim of the present study assessed whether
across disorders and evidence that specific symptoms
the presence of SSs was related to overall anxiety severity,
may be uniquely associated with specific disorders
impairment in familial and social domains of the child_s
suggests that a re-evaluation of the diagnostic criteria
life, and global child functioning. Findings revealed that
for these anxiety disorders in children and adolescents
youths with higher levels of SSs reported higher levels of
may be useful.
overall anxiety severity (based on clinician ratings of
With respect to gender, no differences were found
anxiety severity), avoidance, and interference with family
between girls and boys in the total number of SSs
relationships. The presence of SSs was not associated with
endorsed. However, an examination of individual SSs
interference with peer relationships. Overall, however,
revealed that boys reported more stomachaches and
youths with higher levels of SSs were rated as more
chest pain compared with girls. Thus, although girls
globally impaired than youths with fewer SSs. This
often score higher on measures of anxiety symptoms
finding replicates previous studies that have found SSs to
that reflect affect and cognitions, it may be that boys are
be positively correlated with anxiety symptoms
more apt to report specific physical symptoms. The
(Garber et al., 1991) and highlights that SSs are likely
literature on SSs has been mixed with respect to gender
to have negative sequelae. It also suggests that targeting
differences. Last (1991) and Masi et al. (2000), for
SSs in treatment may lead to reductions in impairment
example, found no gender differences in the frequency
and improved functioning.
of SSs among clinical populations of youths. In contrast,
several others have reported a higher prevalence of SSs for
girls compared with boys (Bernstein et al., 1989; Escobar Aim 3: Somatic Symptoms and
et al., 1987; Garber et al., 1991; Livingston et al., 1988; Psychopharmacological Treatment
Muris and Meesters, 2004; Rauste-von Wright and von The final aim of the present study was to examine the
Wright, 1981). The differences may result from sample impact of psychopharmacological treatment on SSs.
characteristics (clinical versus community), age of Significant reductions in the total number of SSs as well
participants, and the specific types of SSs assessed. as on individual SSs from pre- to posttreatment were
In terms of age differences, findings indicated that found among children taking both FLV and PBO.
adolescents reported a higher number of physical symp- However, children taking FLV reported an average
toms compared with younger children. Adolescents re- reduction of four SSs compared with two for children
ported five specific symptoms more often than children: in the PBO group. In addition, among children taking
blushing, sweating, trembling/shaking, feeling dizzy, and PBO, 9 of 16 SSs were significantly reduced from pre-
paresthesias. Consistent with previous studies (Last, to posttreatment; 13 of 16 were significantly reduced in
1991), older children (12 years and older) also reported the FLV group. Comparisons between the two groups
a significantly higher number of SSs (on average, about revealed that relative to PBO, children in the FLV
two more SSs) compared to younger children group reported fewer SSs at posttreatment (four versus
(6Y11 years). There are several interpretations for this two) as well as lower rates of six specific SSs (i.e., muscle
finding. First, it is possible that older youths are better able tension, blushing, feeling paralyzed, palpitations, chills
to articulate the physical sensations they experience in their or hot flashes, and sweating). Of note, four of these
body (trembling, dizziness) than their younger counter- symptoms were among the seven most common SSs
parts. Alternatively, physical manifestations of anxiety may reported. It is also noteworthy that common SSRI side
increase with age and/or pubertal development. Finally, effects, such as Bfeels sick at stomach[ or restlessness,
given that adolescents tend to have higher levels of social were reduced, not increased, by SSRI treatment. Thus,
anxiety (and social anxiety disorder; Costello et al., 2004), an SSRI, specifically FLV, appears to be effective in
it is plausible that the higher rates of SSs, such as blushing reducing the severity of certain physical symptoms.

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006 1185

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
GINSBURG ET AL.

Clinical, Research, and Diagnostic Implications among adults with these anxiety disorders requires
Findings from the present study have several additional examination.
implications for practicing clinicians. For instance, in Several other directions for future research are
light of findings that SSs are highly common among implicated including the extent to which physical symp-
anxious youth and are associated with both overall toms occurred during discrete panic or anxiety attacks
anxiety severity and impairment in functioning, (versus chronically and nonepisodically) and the degree
increased attention to their assessment and treatment to which SSs are related to anxiety sensitivity (i.e., the
is warranted. For clinical situations that preclude catastrophic misinterpretation of bodily sensations),
lengthy assessments, focusing on the seven most which has been found to be a predictor of anxiety
common SSs measured in this study will likely identify disorders and panic attacks/disorders (Ginsburg and
most children with impairing anxiety. Moreover, Drake, 2002; Kearney et al., 1997; Lau et al., 1996).
because patients, parents, and clinicians may attribute
SSs to medication side effects, it is important to carefully Limitations
assess SSs before, during, and after treatment. Findings of the present study need to be considered in
The assessment of SSs may help guide treatment de- the context of several methodological limitations. First,
cisions. Specifically, although the presence of SSs only 16 SSs were assessed. Thus, although these SSs were
decreased from pre- to posttreatment with FLV, almost common among anxious youth, there are likely to be
40% of youths continued to report restlessness and 26% other SSs that may be of equal or greater importance.
reported continued stomach aches after an 8-week course Indeed, headache is often cited as one of the most
of FLV. Thus, additional treatments, both psychosocial common SSs among children with psychiatric illnesses
and pharmacological, may be needed to reduce somatic and was not measured in the present study (Masi et al.,
complaints to return the child to Bnormal[ functioning. 2000). Another potential limitation of the present study
In addition, findings revealed that treatment with FLV is the method for assessing SSs. Although the PARS has
led to a reduction rather than increase in several SSs (i.e., the advantage of relying on expert clinical judgment, the
muscle tension, blushing, feeling paralyzed, palpitations, time frame for reporting these symptoms was within the
chills or hot flashes, sweating) during the 8-week trial. past week. This time frame may be too brief to capture
In addition, common SSRI side effects, such as Bfeels impairing SSs that wax and wane or occur as part of
sick at stomach[ or restlessness, which may show initial discrete episodes at less regular intervals. Finally,
increases, are likely to decrease in the majority of youths interpretations of the present findings are limited by
after 8 weeks. the sample characteristics. Children in the present study
With respect to research, findings highlight the need to (1) were volunteers agreeing to participate in a medica-
standardize the method of assessing SSs. Studies have tion trial, (2) met strict inclusion/exclusion criteria, (3)
examined different SSs and have used different methods, were primarily white, and (4) had diagnoses that were
making comparisons difficult. A measure that assesses a limited to GAD, SOP, and SAD. Thus, whether findings
broad range of SSs such as the Children_s Somatization can be generalized to other clinic samples (e.g., those
Inventory (Garber et al., 1991) or a side-effect checklist unwilling to take medication, have comorbid condi-
commonly used in clinical practice are potential tions), racial/ethnic minorities, and youth with other
candidates. Broad measures of SSs are recommended anxiety disorders (e.g., obsessive-compulsive disorder)
because even the least common symptom in the current awaits investigation.
study, paresthesias, was endorsed by 14% of children.
These findings also have potential implications for CONCLUSIONS
the development of DSM-V. In DSM-IV, SSs are only Somatic symptoms, particularly restlessness and
included as criteria in GAD and panic disorder, yet the stomachaches, are highly prevalent among children and
present findings revealed that physical symptoms were adolescents with anxiety disorders. Somatic symptoms
common in all three anxiety disorders studied: SAD, were more common among older youths and among
GAD, and SOP. For DSM-V, the addition of SSs, such children with GAD. SSs are associated with anxiety
as restlessness or feels sick at stomach, needs to be severity and impairment in children_s lives. Fluvoxamine
considered for child populations. The presence of SSs successfully reduced most SSs.

1186 J. AM. ACAD. CH ILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
 SOMATIC SYMPTOMS AND CHILD ANXIETY

Guy W (1976), In: ECDEU Assessment Manual for Psychopharmacology.

Disclosure: Dr. Riddle has an unrestricted research grant from Pfizer
Washington, DC: U.S. Department of Health, Education, and Welfare,
and is a consultant to Johnson & Johnson and Shire. Dr. Davies holds pp 76Y338
stock in Merck, Pfizer, Wyeth, GSK, Johnson & Johnson, Amgen, Bard, Jolly JB, Wherry JN, Wiesner DC, Reed DH, Rule JC, Jolly JM (1994),
and Amgen in excess of $50,000. Dr. Ginsburg has no financial The mediating role of anxiety in self-reported somatic complaints of
relationships to disclose. depressed adolescents. J Abnorm Child Psychol 22:691Y702
Kaufman J, Birmaher B, Brent D et al. (1997), Schedule for Affective
Disorders and Schizophrenia for school-age children-Present and
REFERENCES Lifetime version (K-SADS-PL): initial reliability and validity data.
J Am Acad Child Adolesc Psychiatry 36:980Y988
Ambrosini PJ (2000), Historical development and present status of the Kearney CA, Albano AM, Eisen AR, Allan WD, Barlow DH (1997), The
schedule for affective disorders and schizophrenia for school-age children phenomenology of panic disorder in youngsters: an empirical study of a
(K-SADS). J Am Acad Child Adolesc Psychiatry 39:49Y58 clinical sample. J Anxiety Disord 11:49Y62
Angold A, Prendergast M, Cox A, Harrington R, Simonoff E, Rutter M Kendall PC, Pimentel SS (2003), On the physiological symptom
(1995), The Child and Adolescent Psychiatric Assessment (CAPA). constellation in youth with generalized anxiety disorder (GAD).
Psychol Med 25:739Y753 J Anxiety Disord 17:211Y221
Beidel DC, Christ MAG, Long PJ (1991), Somatic complaints in anxious Last CG (1991), Somatic complaints in anxiety disordered children.
children. J Abnorm Child Psychol 19:659Y670 J Anxiety Disord 5:125Y138
Bernstein GA, Garfinkel BD, Hoberman HM (1989), Self-reported anxiety Lau JJ, Calamari JE, Waraczynski M (1996), Panic attack symptomatology
in adolescents. Am J Psychiatry 146:384Y386 and anxiety sensitivity in adolescents. J Anxiety Disord 10:355Y364
Costello EJ, Angold A, Burns BJ, Erkanli A, Stangl DK, Tweed DL (1996), Livingston R, Taylor JL, Crawford SL (1988), A study of somatic complaints
The Great Smoky Mountains Study of Youth. Functional impairment and psychiatric diagnosis in children. J Am Acad Child Adolesc Psychiatry
and serious emotional disturbance. Arch Gen Psychiatry 53:1137Y1143 27:185Y187
Costello EJ, Egger HL, Angold A (2004), Developmental epidemiology of Masi G, Favilla L, Millepiedi S, Mucci M (2000), Somatic symptoms in
anxiety disorders. In: Phobic and Anxiety Disorders in Children and children and adolescents referred for emotional and behavioral disorders.
Adolescents, OllendickTH, MarchJS, eds. New York: Oxford University Psychiatry 63:140Y149
Press, pp 61Y91 Muris P, Meesters C (2004), Children_s somatization symptoms, correla-
Dhossche D, van der Steen F, Ferdinand R (2002), Somatoform disorders in tions with trait anxiety, anxiety sensitivity, and learning experiences.
children and adolescents: a comparison with other internalizing Psychol Rep 94:1269Y1275
disorders. Ann Clin Psychiatry 14:23Y31 Rauste-von Wright M, von Wright J (1981), A longitudinal study of
Egger HL, Costello EJ, Erkanli A, Angold A (1999), Somatic complaints and psychosomatic symptoms in healthy 11Y18 year old girls and boys.
psychopathology in children and adolescents. J Am Acad Child Adolesc J Psychosom Res 25:525Y534
Psychiatry 38:852Y860 Research Unit on Pediatric Psychopharmacology (RUPP) Anxiety Study
Escobar JI, Burnam MA, Karno M, Forsythe A, Golding JM (1987), Group (2001), Fluvoxamine for the treatment of anxiety disorders in
Somatization in the community. Arch Gen Psychiatry 44:713Y718 children and adolescents. N Engl J Med 344:1279Y1285
Garber J, Walker LS, Zemnan J (1991), Somatization symptoms in a Research Units on Pediatric Psychopharmacology Anxiety Study Group
community sample of children and adolescents: further validation of the (RUPP) (2002), The Pediatric Anxiety Rating Scale (PARS): Develop-
children_s somatization inventory. Psychol Assess 3:588Y595 ment and Psychometric Properties. J Am Acad Child Adolesc Psychiatry
Ginsburg GS, Drake K (2002), Anxiety sensitivity and panic attack 41:1061Y1069
symptomatology among low-income African American adolescents. Shaffer D, Gould MS, Brasic J et al. (1983), A children_s global assessment
J Anxiety Disord 16:83Y96 scale (CGAS). Arch Gen Psychiatry 40:1228Y1231

How Do Psychological Factors Influence Adolescent Smoking Progression? The Evidence for Indirect Effects Through
Tobacco Advertising Receptivity Janet Audrain-McGovern, PhD, Daniel Rodriguez, PhD, Vaishali Patel, BA, Myles S. Faith,
PhD, Kelli Rodgers, BA, and Jocelyn Cuevas, BA
Objectives: To determine whether novelty seeking and depressive symptoms had mediated or indirect effects on adolescent
smoking progression through tobacco advertising receptivity. Methods: More than 1000 adolescents were monitored from 9th
grade to 12th grade and completed annual surveys that measured demographic characteristics, smoking behavior, tobacco
advertising receptivity, novelty-seeking personality, depressive symptoms, family and peer smoking, alcohol use, and marijuana
use. Results: Latent growth modeling indicated that novelty seeking had a significant indirect effect on smoking progression
through baseline tobacco advertising receptivity. For each 1-SD increase in novelty seeking, the odds of being more receptive to
tobacco advertising increased by 12% (ie, being in a specific category or higher), which in turn resulted in an 11% increase in the
odds of smoking progression from 9th grade to 12th grade. The indirect effect from depressive symptoms to smoking progression
did not reach significance. Conclusions: These findings may inform future research on other factors that influence tobacco
advertising receptivity, as well as programs aimed at preventing adolescent smoking initiation and progression. Pediatrics
2006;117:1216Y1225.

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:10, OCTOBER 2006 1187

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.