Sleep-Related Problems Among Children and

Adolescents With Anxiety Disorders

CANDICE A. ALFANO, PH.D., GOLDA S. GINSBURG, PH.D.,
AND JULIE NEWMAN KINGERY, PH.D.

ABSTRACT
Objective: The present study examined sleep-related problems (SRPs) among a large sample (n = 128) of youth with
anxiety disorders (i.e., generalized, separation, and social). The frequency of eight specific SRPs was examined in relation
to age, gender, type of anxiety disorder, anxiety severity, and functional impairment. The impact of pharmacological
treatment (fluvoxamine versus pill placebo) in reducing SRPs also was examined. Method: As part of a large, double-blind,
randomized, controlled trial (Research Units on Pediatric Psychopharmacology Anxiety Study Group), clinician and parent
reports of SRPs were examined among children and adolescents, ages 6 to 17 years, before and after treatment.
Results: Eighty-eight percent of youth experienced at least one SRP, and a majority (55%) experienced three or more.
Total SRPs were positively associated with anxiety severity and interference in family functioning. Significantly greater
reductions in SRPs were found among children treated with fluvoxamine compared with placebo. Conclusions: These
findings indicate that SRPs are commonly associated with childhood anxiety disorders and suggest a need for the
assessment of and attention to these problems in research and clinical settings. J. Am. Acad. Child Adolesc. Psychiatry,
2007;46(2):224Y232. Key Words: sleep problems, childhood anxiety, anxiety severity, impairment, treatment.

Adequate sleep is essential for health and normal

growth and development in children. Because insufficient sleep has been associated with impairments such
Accepted July 18, 2006.
Dr. Alfano is with the Department of Psychiatry, Childrens National
Medical Center, Washington, DC; Drs. Ginsburg and Kingery are with the
Department of Psychiatry and Behavioral Sciences, Division of Child and
Adolescent Psychiatry, Johns Hopkins University School of Medicine, Baltimore.
Preparation of this paper was supported by NIMH grant K23-MH63427-02
awarded to Dr. Ginsburg.
The authors wish to acknowledge the Research Units on Pediatric
Psychopharmacology Anxiety Group (RUPP) sites that supported the data
collection for this study: Mark A. Riddle, M.D., John T. Walkup, M.D., and
Michael J. Labellarte, M.D., Johns Hopkins University; Daniel S. Pine, M.D.,
Laurence Greenhill, M.D., Rachel Klein, Ph.D., and Michael Sweeney, Ph.D.,
Columbia University and New York State Psychiatric Institute; Howard
Abikoff, Ph.D., Sabine Hack, M.D., 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 Angeles; John March, M.D.,
M.P.H., and Scott Compton, Ph.D., Duke University; and Ben Vitiello, M.
D., National Institute of Mental Health.
Correspondence to Dr. Candice Alfano, Department of Psychiatry, Children_s
National Medical Center, 111 Michigan Avenue, NW, Washington, DC
20010; e-mail: calfano@cnmc.org.
0890-8567/07/4602-02242007 by the American Academy of Child
and Adolescent Psychiatry.
DOI: 10.1097/01.chi.0000242233.06011.8e

224

as decreased attention, impulsivity, behavioral problems, and decrements in school performance (Mindell
et al., 1999), there is a need to better understand the
association between sleep disturbances and psychiatric
symptoms in children. The term sleep disturbance is, in
fact, quite broad and often is used to refer to a range of
sleep problems that may be influenced by intrinsic (e.g.,
difficulty initiating/maintaining sleep) and/or extrinsic
(e.g., poor sleep hygiene, bedtime resistance) factors.
Moreover, compared with adults, sleep disturbances
among children generally encompass a wider range of
problem behaviors. For example, in addition to
problems initiating sleep, anxious children commonly
experience nonspecific nighttime fears, nightmares, and
difficulty sleeping alone/away from home, all of which
may disrupt sleep continuity and quality and result in
excessive daytime somnolence. Thus, we refer to this
range of potential nighttime difficulties among anxious
children more broadly as sleep-related problems (SRPs).
Among children with anxiety disorders, research
examining SRPs and their potential impact on daytime
functioning is extremely limited. However, data based
on community samples of children reveal an important

J. AM . ACAD. CHILD ADOLESC. PSYCH IATRY, 46:2, FEBRUARY 2007

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

SLEEP PROBLEMS AND ANXIOUS CHILDREN

association between sleep problems and anxiety. For

example, Mindell and Barrett (2002) reported incremental increases in trait anxiety relative to an increased
frequency of nightmares among 60 children (ages 5Y11
years) recruited from elementary school classrooms. For
children reporting three or more nightmares per week,
parent-reported anxiety scores approached the clinically
significant range. Using longitudinal data, Gregory and
O_Connor (2002) examined the relation between parent report of early sleep problems and the development
of emotional/behavioral problems during midadolescence in a large community sample of children (n = 490).
The presence of sleep problems at age 4 significantly
predicted problems of anxiety/depression at ages 13 to 15.
A second longitudinal investigation used a large community sample (n = 943) to examine early sleep problems
as predictors of later psychiatric diagnoses (Gregory
et al., 2005). Among the 12% of children whose parents
reported persistent sleep problems from ages 5 to 9,
almost half (46%) developed an adult anxiety disorder.
In contrast, persistent sleep problems during childhood
did not significantly predict the development of adult
depressive disorders.
Only a few studies have examined SRPs among
clinically anxious youth. Using a large sample of children
(n = 157) with generalized anxiety disorder (GAD), Masi
and colleagues (2004) found that 56% of children and
49% of adolescents experienced a Bsleep disturbance[
based on combined parent and child reports. Similarly,
Pina and colleagues (2002) reported that 42% of children
and 57% of adolescents with GAD/overanxious disorder
(n = 111; ages 7Y16) experienced Btrouble sleeping,[ and
Kendall and Pimentel (2003) reported the presence of
Bsleep disturbance[ among 66% of children with GAD
(n = 47; ages 9Y13) based on child and parent report.
Further information regarding the specific nature of sleep
problems experienced by anxious youth was not
provided. Thus, this small body of research is limited
to children with GAD and is restricted to broad,
nonspecific indexes of sleep disturbance.
Limited data based on the use of polysomnography
also provide preliminary support for the presence of
disrupted sleep among anxious youth. Rapoport et al.
(1981) reported reduced sleep efficiency and increased
sleep latency among nine adolescents (ages 13Y17) with
obsessive-compulsive disorder compared with matched
healthy controls. Adolescents with obsessive-compulsive
disorder required twice as long as control adolescents to

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 46:2, FEBRUARY 2007

initiate sleep onset. Although research using objective

measures of sleep among youth with other anxiety
disorders has not been conducted, studies of anxietydisordered adults indicate different forms of sleep
disruption to be highly prevalent (see Uhde, 2000).
Research examining the frequency of specific SRPs
among anxious youth, including potential associations
with age, gender, type of anxiety disorder, anxiety
severity, and impaired daytime functioning, may therefore provide important information for both researchers
and clinicians.
Similar to a need for information concerning sleep
disruption among anxious youth, data examining the
impact of effective treatments for childhood anxiety on
co-occurring SRPs also are needed. One study reported
general decreases in somatic symptoms (including one
Bsleep disturbance[ item) among children with GAD
after cognitive-behavioral treatment, although reductions in specific SRPs were not examined (Kendall and
Pimentel, 2003). With regard to pharmacological
treatment studies, outcomes related to sleep are
particularly relevant. Because the neurotransmitter
systems involved in the modulation of anxiety also are
implicated in the regulation of sleep and because
pharmacological agents, including selective serotonin
reuptake inhibitors (SSRIs), that reduce anxiety also
have been shown to alter sleep (see Sandor and Shapiro,
1994), improvement in symptoms across both domains
may be observed. Several published studies have
reported significant decreases in children_s anxiety
symptoms after treatment with an SSRI (March et al.,
1998; Research Unit on Pediatric Psychopharmacology
Anxiety Study Group, 2001; Wagner et al., 2004), yet
sleep-related outcomes have not been examined.
The present study begins to address these gaps in the
literature through preliminary examination of several
types of SRPs among a large sample of youth with
anxiety disorders. In particular, the frequency of eight
specific SRPs was examined in relation to age, gender,
and type of anxiety disorder (i.e., GAD, separation
anxiety [SAD], and social anxiety [SOC]) on the basis
of both parent and clinician assessment. Second,
associations between SRPs, anxiety severity, and
impairment in daytime functioning both within and
outside the home were examined. On the basis of
findings from community-based samples of children
(e.g., Mindell and Barrett, 2002; Mindell et al., 1999),
we hypothesized that the presence of SRPs would be

225

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

ALFANO ET AL.

related to higher levels of anxiety severity and impairments across both domains. Finally, we examined
the impact of pharmacological treatment in reducing
SRPs among anxious youth. In particular, we examined
clinician report of three specific SRPs after 8 weeks of
treatment with either fluvoxamine (FLV) or pill placebo (PBO). We hypothesized that treatment with an
SSRI would produce a significant decrease in SRPs
relative to placebo.
METHOD
Participants
Participants were 128 children, 6 to 17 years of age (mean, 10.8
years), who met DSM-IV criteria for GAD, SAD, and/or SOC on
the basis of child and parent Schedule for Affective Disorders and
Schizophrenia for School-Aged Children (Kaufman et al., 1997)
interviews with a trained clinician (see Research Units on Pediatric
Psychopharmacology Anxiety Study Group, 2001). All children
were enrolled in a double-blind, placebo-controlled, clinical trial of
FLV for youth with anxiety disorders (i.e., SOC, SAD, and/or
GAD). Children were recruited from five sites designated Research
Units in Pediatric Psychopharmacology (RUPP): Duke University,
Johns Hopkins School of Medicine, New York State Psychiatric
Institute/Columbia University, New York University, and University of California, Los Angeles. Sample characteristics are presented
in Table 1.
Exclusion criteria included current use of any illicit or prescribed
psychoactive substance; current diagnoses of major depressive
disorder, Tourette_s disorder, obsessive-compulsive disorder, posttraumatic stress disorder, conduct disorder, or panic disorder; any
past or current history of mania, psychosis, or pervasive developmental disorder; suicidal ideation; mental retardation; previous
treatment with a selective serotonin reuptake inhibitors; and a
diagnosis of attention-deficit/hyperactivity disorder that required
pharmacological treatment.
Measures
Pediatric Anxiety Rating Scale. The Pediatric Anxiety Rating Scale
(PARS; Research Units on Pediatric Psychopharmacology Anxiety
Study Group, 2002) is a clinician-rated instrument for assessing the
severity of anxiety symptoms associated with childhood anxiety
disorders. The instrument has two sections. The first section
includes a 50-item symptom checklist in which items are rated as
present or absent during the past week. The second section is made
up of five severity items and two impairment items (rated on a 5point Likert scale), with higher scores reflecting greater severity/
impairment. Internal consistency, test-retest reliability, and validity
for the PARS have been found to be acceptable (Research Units on
Pediatric Psychopharmacology Anxiety Study Group, 2002).
Hamilton Anxiety Rating Scale. The Hamilton Anxiety Rating Scale
(HAM-A; Hamilton, 1959) is a 14-item, clinician-rated instrument for
assessing the severity of anxiety symptoms. Items are scored from none
(0) to very severe (4). The psychometric properties of the HAM-A have
been shown to be acceptable (Maier et al., 1988).
Child Behavior Checklist-Parent Version. The Child Behavior
Checklist-Parent Version (CBCL; Achenbach and Edelbrock,

226

TABLE 1
Demographic and Clinical Characteristics of Entire Sample
Characteristic
n
%
Age, y
6Y11
12Y17
Female gender
Ethnicity
White
Hispanic
Black (non-Hispanic)
Other
Total family income, $
<25,000
25,000Y39,999
40,000Y59,999
960,000
Refused/unknown
K-SADS diagnosis
SOC only
SAD only
GAD only
SOC and SAD only
SOC and GAD only
SAD and GAD only
SOC, SAD, and GAD

86
42
63

67
33
49.2

81
24
9
14

63.3
18.8
7.0
10.9

19
20
18
55
16

14.8
15.6
14.1
43.0
12.5

26
18
5
11
21
21
26

20.3
14.1
3.9
8.6
16.4
16.4
20.3

Note: K-SADS = Schedule for Affective Disorders and

Schizophrenia for School-Aged Children; SOC = social anxiety
disorder; SAD = separation anxiety disorder; GAD = generalized
anxiety disorder. N = 128.
1991) is a factor analytically derived checklist including a broad
range of child behaviors. Parents are asked to rate 113 items as not
true (0), sometimes true (1), or often/always true (2) on the basis of
their child_s behavior during the past 6 months. The CBCL yields a
total problems score, internalizing and externalizing problems
scores, and eight subscales. Ample evidence supports the psychometric properties of the CBCL.
Children_s Global Assessment Scale. The Children_s Global
Assessment Scale (CGAS; Shaffer et al., 1983), a modification of
the adult GAS, is a widely used measure of global impairment in
functioning. Scores range from 1 to 100, with higher scores reflecting
higher levels of functioning.
Clinician_s Global Impression-Severity of Illness. The Clinician_s
Global Impression-Severity of Illness (CGI-S; Guy, 1976), a widely
used measure to assess global severity of illness, was used to assess
severity of anxiety (based on a 7-point scale). Higher scores reflect
more severe anxiety.
Sleep-Related Problems. Because a standardized measure of sleep
problems was unavailable as part of the original RUPP anxiety data
set, information on eight specific SRPs was gathered from three
separate measures. Specifically, the PARS includes two sleep items
assessing refusal/reluctance to sleep alone and refusal/reluctance to
sleep away from home; the HAM-A includes one item assessing the
presence of insomnia (defined as difficulty initiating or maintaining
sleep); and the CBCL includes five sleep items assessing nightmares,
being overtired without good reason, sleeping less than most kids,
sleeping more than most kids, and talking/walking during sleep. To

J. AM . ACAD. CHILD ADOLESC. PSYCH IATRY, 46:2, FEBRUARY 2007

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

SLEEP PROBLEMS AND ANXIOUS CHILDREN

examine overall levels of sleep disturbance, these eight sleep items

were combined to create a total SRP score, computed by adding
responses for all items across measures (scored as either 0 = not
present or 1 = present) for a total possible score ranging from 0 to 8.
To standardize total SRP scores, some items had to be recoded. For
the CBCL items, responses based on a 3-point scale were
dichotomized so that a response of either sometimes true (1) or
often/always true (2) was coded as a yes. For the HAM-A items,
responses based on a 5-point scale were dichotomized so that a
response of moderate (2), severe (3), or very severe (4) was coded as
a yes. Internal consistency was computed for the total SRP scale and
was acceptable (! = .64).
Anxiety Severity. In addition to the CGI-S, which assesses global
severity of illness, three severity items from the PARS were
combined to create a total anxiety severity score. These items
included overall anxiety severity, severity of somatic/physical
symptoms of anxiety, and frequency of anxiety symptoms. Two
severity items from the PARS were not used in the present study
because the calculation of one item (total number of anxiety
symptoms) is directly influenced by the endorsement of sleep items,
and the conceptual basis for examining the other item (overall
avoidance of anxiety-provoking situations) in association with SRP
is unclear. The HAM-A total anxiety score (minus the insomnia
item) also was used to assess overall severity of anxiety.
Impairment. Two items from the PARS were used to assess
impairment: interference at home/family functioning (including
relationships with parents/siblings, impact on family activities,
completion of chores, etc.) and interference outside the home/peer
relationships (including, e.g., school performance, extracurricular
activities, relationships with peers). The CGAS provided a measure
of global impairment of functioning.

Procedure
After a full description of the study, all of the parents and children
signed informed consent/assent (see RUPP, 2001 for details). A
comprehensive pretreatment evaluation, including a diagnostic
interview and additional measures to assess inclusion/exclusion
criteria and primary outcomes, was conducted. During the 8-week
treatment phase, child psychiatrists who were nave about children_s
treatment assignment saw each child and parent on a weekly basis
for the first 6 weeks and again at week 8. At each visit, child

SRP
Insomnia
Reluctance/refusal to sleep alone
Reluctance/refusal to sleep away from home
Nightmares
Overtired without good reason
Sleeps less than most kids
Sleeps more than most kids
Talks/walks in sleep
Mean no. SRPs (SD)

psychiatrists administered supportive psychoeducational therapy to

children and families, conducted a clinical assessment of anxiety
symptoms, assessed adverse events, and determined medication
dosage following a flexible, forced dose titration (i.e., psychiatrists
followed a fixed schedule unless clinical factors indicated the need
for a slower titration).
Data Analyses
Data were analyzed with SPSS 13.0 statistical software. Pairedsample t tests and x2 tests were used to examine differences in total
and individual SRPs based on age, gender, and type of anxiety
disorder. Pearson correlation coefficients were used to examine
associations between SRPs, anxiety severity, and functional
impairment, and a hierarchical linear regression model was used
to examine SRPs as a predictor of impaired functioning. A repeatedmeasures analysis of variance was used to examine changes in total
SRPs from before to after treatment, and follow-up x2 tests were
used to examine changes across individual SRP.
RESULTS
Gender, Age, and Anxiety Disorder

Frequencies of the eight sleep items are presented in

Tables 2 and 3 for the total sample and by gender, age
group, and diagnosis (GAD, SAD, and SOC). Eightyeight percent of the total sample reported at least one
SRP; 55% reported three or more SRPs. The most
common SRPs were insomnia, nightmares, and refusal/
reluctance to sleep alone.
Gender. The difference for mean number of SRPs
among girls compared with boys was not statistically
significant. In terms of individual SRPs, nightmares
[x2(117) = 4.12, p < .05] were significantly more
common among girls than boys.
Age. The difference for mean number of SRPs
among younger children (ages 6Y11) compared with

TABLE 2
Sleep-Related Problems by Gender and Age
Total
Male
Female
(N = 128), %
(n = 65), %
(n = 63), %
66.6
47.9
40.9
54.5
43.2
36.9
15.1
22.7
2.9 (1.9)

60.7
46.5
44.8
44.0
40.7
33.3
15.0
21.7
2.7 (1.9)

72.1
50.8
40.0
62.7*
45.8
40.7
15.3
23.7
3.1 (2.0)

Ages 6Y11
(n = 86), %

Ages 12Y17
(n = 42), %

70.7
61.0**
44.7
63.8**
42.5
36.3
11.3
25.0
3.1 (1.8)

59.5
26.2
38.1
31.6
44.7
38.5
23.1
17.9
2.5 (2.1)

Note: SRPs = sleep-related problems. Some SRP items were missing for some subjects.
* p < 0.05; **p < 0.01.

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 46:2, FEBRUARY 2007

227

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

ALFANO ET AL.

SRP
Insomnia
Reluctance/refusal to sleep alone
Reluctance/refusal to sleep away from home
Nightmares
Overtired without good reason
Sleeps less than most kids
Sleeps more than most kids
Talks/walks in sleep
Mean no. SRPs (SD)

TABLE 3
Sleep-Related Problems by Anxiety Diagnosis
SOCj
SOC+
GADj
(n = 52)
(n = 76)
(n = 67)
77.3
66.7*
60.0*
62.5
43.8
42.9
16.3
26.5
3.5 (1.7)

60.3
37.0
32.4
46.4
43.5
31.9
14.5
18.8
2.5 (2.0)*

37.3
48.4
40.3
45.9
35.5
30.6
16.1
16.1
2.5 (1.8)

GAD+
(n = 61)

SADj
(n = 52)

SAD+
(n = 76)

56.9*
49.1
45.5
61.4
51.8
43.9
14.0
29.8
3.3 (1.9)*

52.1
22.9
16.3
37.5
37.5
33.3
14.6
16.7
2.0 (1.8)

76.8**
66.2**
61.8**
64.3**
47.1
39.4
15.5
26.8
3.5 (1.8)**

Note: SOC = social anxiety disorder; SAD = separation anxiety disorder; GAD = generalized anxiety disorder; j = not present; + = present.
* p < 0.05; **p < 0.01.

older children (ages 12Y17) was not statistically

significant. However, younger anxious children were
more likely to exhibit refusal/reluctance to sleep alone
[x2(119) = 13.21, p < .001] and to experience nightmares
[x2(118) = 10.72, p < .01] than older children.
Anxiety Diagnosis. The limited number of children
with a single anxiety disorder precluded statistically
meaningful comparisons of SRPs for children with only
one anxiety disorder. Thus, Table 3 displays the means
for total number of SRPs and the frequencies for each
individual SRP among youth with and without GAD,
SAD, and SOC.
Ninety-eight percent of children with a GAD
diagnosis had at least one SRP. Children with a
diagnosis of GAD had a significantly greater number of
SRPs compared with youth without a GAD diagnosis
[t (1, 125) = 2.55, p < .01]. In addition, children with GAD
were more likely to experience insomnia [x2(117) = 4.51,
p < .05] than children without GAD.
Ninety-seven percent of children with a SAD
diagnosis had at least one SRP. Children with a
diagnosis of SAD had a significantly greater number of
SRPs compared with youth without an SAD diagnosis
[t (1, 125) = 4.55, p < .001]. Analyses of individual SRP
revealed that children with a diagnosis of SAD were
more likely to experience insomnia [x2(117) = 6.11,
p < .025], reluctance/refusal to sleep alone [x2(119) =
21.47, p < .001], reluctance/refusal to sleep away from
home [x2(117) = 24.03, p < .001], and nightmares
[x2(118) = 8.21, p < .01] than children without SAD.
Ninety percent of children with an SOC diagnosis
had at least one SRP. Overall, children with a diagnosis
of SOC had significantly fewer SRPs compared with

228

youth without an SOC diagnosis [t (1, 124) = j3.05,

p < .01]. Children with a diagnosis of SOC were less
likely to exhibit reluctance/refusal to sleep alone
[x2(118) = 9.82, p < .01] and reluctance/refusal to
sleep away from home [x2(116) = 8.56, p < .01] than
children without a diagnosis of SOC.
Anxiety Severity and Impairment

Pearson correlation coefficients between total SRP

scores and measures of anxiety severity and impairment
in functioning at baseline are presented in Table 4. Total
SRPs were positively and significantly associated with
CGI-S scores, PARS anxiety severity, HAM-A total scores,
and PARS interference at home/family functioning.
Do SRPs Independently Predict Impairment in Family
Functioning?

Because correlations between total SRPs, anxiety

severity, and interference at home/family functioning
were found to be significant, we were interested in
examining whether SRPs independently predicted
interference at home/family functioning or alternatively
whether this association may be better explained by
overall higher levels of anxiety severity among children
with SRPs. A hierarchical linear regression examining
the prediction of interference at home/family functioning was therefore conducted. PARS anxiety severity,
HAM-A total score, and CGI-S score were entered and
controlled for in step 1 of the model. Total SRP score
was then entered into the model in the step 2. Results
indicated that total SRP score was a significant predictor of impairment in home functioning after controlling for severity of anxiety (" = .26, p < .01, R 2 = 0.21).

J. AM . ACAD. CHILD ADOLESC. PSYCH IATRY, 46:2, FEBRUARY 2007

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

SLEEP PROBLEMS AND ANXIOUS CHILDREN

TABLE 4
Bivariate Correlations for Sleep-Related Problems, Anxiety Severity, and Impairment
Total
PARS Anxiety
HAM-A Total
At-Home
SRP
CGI-S
Severity
Anxiety
CGAS
Impairment
Total SRP
CGI-S
PARS anxiety
severity
HAM-A total
anxiety
CGAS
At-home
impairment
Out-of-home
impairment

V
0.18*
0.34**

V
0.41**

0.54**

0.22*

j0.18
0.39**
j0.05

Out-of-Home
Impairment

V
0.59**

j0.58**
0.30**

j0.32**
0.24**

0.42**

0.30**

V
j0.15
0.34**

V
j0.22*

0.09

j0.35**

0.12

Note: CGI-S = Clinical Global Impression Scale-Severity of Illness; PARS = Pediatric Anxiety Rating Scale; HAM-A = Hamilton Anxiety
Rating Scale; CGAS = Children_s Global Assessment Scale.
* p < 0.05, **p < 0.01.

Impact of Treatment

Mean total scores and percentages for individual SRP

items at baseline and after treatment for children in the
FLV and PBO groups are presented in Table 5. Because
CBCL data were available only at baseline, changes in
SRP from pre- to posttreatment were examined for
three clinician-rated SRP items only (from the PARS
and HAM-A), with total SRP scores ranging from 0 to
3. At baseline, children in the FLV and PBO groups did
not differ in terms of total SRP scores or on any of the
individual SRP items.
A 2  2 repeated-measures analysis of variance
(treatment group by time) examining changes in total
TABLE 5
Mean Scores and Percentages for Sleep-Related Problems Among
Fluvoxamine and Placebo Children at Baseline and After Treatment
FLV (n = 54)
PBO (n = 51)
SRP
Insomnia, %
Reluctance/refusal to
sleep alone, %
Reluctance/refusal to
sleep away from
home, %
Total mean SRPs
(SD), no.

After
After
Baseline Treatment Baseline Treatment
50.8
53.3

11.1*
27.3*

43.1
44.1

33.9
39.3

44.1

22.2*

41.4

27.8*

1.4
(1.2)

0.61
(9.4)*

1.2
(1.1)

0.90
(1.0)

Note: FLV = fluvoxamine; PBO = placebo.

* Between-group change from baseline to after treatment, p < .05.

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 46:2, FEBRUARY 2007

SRP scores revealed a significant main effect for time

[F(1, 104) = 34.41, p < .001]. The treatment group-bytime interaction also was significant [F(1, 104) = 9.43,
p < .01], revealing a significantly greater reduction in
SRP among children treated with FLV. Treatment
results were not moderated by gender or age. Follow-up
x2 analyses examining change in status (i.e., present at
baseline versus present/absent after treatment) for each
of the three SRP items from baseline to posttreatment
revealed a significant difference between the groups
for two of the three items. Both insomnia [x2(102) =
10.55, p < .05] and reluctance/refusal to sleep alone
[x2(104) = 10.18, p < .05] were significantly reduced
with FLV compared with PBO. Significant reductions
in reluctance/refusal to sleep away from home were
found for both FLV and PBO children after treatment.
DISCUSSION
Prevalence of SRPs Among Anxious Youth

Eighty-eight percent of anxious youth experienced at

least one SRP, and more than half experienced three or
more. The most common SRPs were insomnia, nightmares, and reluctance/refusal to sleep alone. Rates of
insomnia found in the present study are consistent with
data indicating insomnia to be present among 60% to
70% of adults with anxiety disorders (Ohayon, 1997;
Uhde, 2000). Although there were no gender or age
differences in total number of SRPs, nightmares were

229

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

ALFANO ET AL.

significantly more common among girls compared with

boys, and younger children were more likely to refuse to
sleep alone and experience nightmares than adolescents.
Although some specific SRPs represent appropriate
developmental fears for many young children, these
problems were experienced by almost two thirds of girls/
young anxious children, a rate that is considerably higher
than that found in the general population (Mindell,
1993). It should be noted, however, that because many of
the younger children in the present study had a diagnosis
of SAD, age-based differences in SRPs may have been
influenced by anxiety diagnosis. Future research will need
to further examine these differences.
SRPs Across Anxiety Diagnoses

SRP were most prevalent among youth with GAD

and SAD. For children with a GAD diagnosis, insomnia was the most common SRP. Because complaints of
difficulty initiating/maintaining sleep are highly prevalent in adults with GAD, insomnia is considered a
core feature of the illness (Monti and Monti, 2000).
Such findings have led to the hypothesis that symptoms
of hypervigilance and hyperarousal associated with GAD
may directly lead to problems initiating/maintaining
sleep (Saletu-Zyhlarz et al., 1997). Indeed, a similar
process may be present for youth with the disorder.
Because the self-regulatory skills of children are likely to
be underdeveloped compared with adults, sleep onset
and maintenance, which require the ability to selfregulate and self-soothe, may pose a particular challenge
for children and adolescents.
Among children with an SAD diagnosis, insomnia,
nightmares, and refusal to sleep alone were commonly
reported. However, specific overlap in some SRPs
examined in the present study and DSM-IV diagnostic
criteria for SAD are noteworthy. Nighttime may be
particularly difficult for children with SAD because of
associations with darkness and separation from caregivers. Nonetheless, because feelings of fear and arousal
are inherently incompatible with sleep onset and
maintenance, the potential exists for these SRPs to
disrupt sleep continuity and quality on an ongoing
basis. The extent to which nighttime fears and other
SRPs among anxious children may result in chronic
sleep disruption is unknown.
Although examination of the shared mechanisms
underlying sleep disruption and anxiety in children has
yet to be conducted, SRPs also may be influenced by

230

environmental/familial factors. Parents of anxious

children may reinforce children_s nighttime fears and
avoidant behaviors through the use of ineffective
parenting strategies, including excessive reassurance,
extension of bedtimes, and/or permitting cosleeping
with parents or siblings (e.g., Dadds et al., 1996). One
recent study found that not putting children to bed
while awake and cosleeping were associated with
childhood sleep disorders among children at risk for
the development of anxiety (Warren et al., 2003).
Among clinically anxious children, such parenting
behaviors may interfere with the development of selfregulatory skills and contribute to increased levels of
arousal at bedtime, ultimately resulting in an increase in
both anxiety and sleep disruption.
SRPs, Anxiety Severity, and Impairment

Consistent with community-based research (Mindell

and Barrett, 2002), these findings suggest that the
presence of SRPs is associated with higher levels of anxiety
severity. Although the causal nature of this relationship
cannot be determined from these data, several possibilities remain to be addressed. Longitudinal studies
indicate sleep problems to be early markers for the later
development of psychopathology (Ford and Kamerow,
1989), including anxiety disorders (Gregory et al., 2005),
and the presence of increased levels of anxiety may
similarly lead to the development of SRP in children. As
discussed by Dahl (1996), a reciprocal relationship
between sleep disruption and psychopathology, in
general, is most likely. Because sleep, arousal, and affect
represent overlapping regulatory systems, the presence of
sleep disruption during critical periods of maturational
development may provide a pathway toward later
affective dysregulation and vice versa (see Dahl, 1996
for a review). At present, specific mechanisms and
processes mediating the regulation of sleep and affect are
poorly understood, and data examining this overlap in
symptoms among clinic-referred children are needed.
The presence of SRPs also was associated with
impairment within the home. In fact, SRPs independently predicted interference at home/in family
functioning after controlling for severity of anxiety.
Contrary to expectations, SRPs and impairment outside the home/in peer relationships were not associated.
This finding is somewhat surprising in light of research
linking childhood sleep disruption with impairments in
academic and behavioral functioning (Bruni et al.,

J. AM . ACAD. CHILD ADOLESC. PSYCH IATRY, 46:2, FEBRUARY 2007

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

SLEEP PROBLEMS AND ANXIOUS CHILDREN

2006; Wolfson and Carskadon, 1998). One possibility

for these divergent findings includes the limited
measures of interference used in the present study.
Future research examining SRPs among anxious youth
should include more specific measures of academic,
behavioral, and interpersonal functioning to begin to
elucidate the potential daytime sequelae of SRP in this
population of children.
It also is possible that among anxious children,
functioning within the home and in family relationships may be affected more adversely by the occurrence
of SRPs. Previous studies examining family and
parenting variables in connection with childhood
anxiety have highlighted the presence of dysfunctional
processes, including overcontrol, enmeshment, reinforcement of anxious behaviors, and conflict (see Ginsburg
et al., 2004). Children_s SRPs may be particularly
disruptive within these environments because of their
impact on other family members. Over time, these
nighttime difficulties may become a major component
of parenting stress and family discord within the home.
Changes in SRPs After Pharmacological Treatment

As predicted, children treated with FLV evidenced

greater decreases in SRPs after treatment relative to
children receiving PBO. The greatest reduction was
found for insomnia. Although several treatment studies
have reported insomnia as an adverse effect of SSRI
treatment in children, the present results suggest that
insomnia and other SRPs are diminished after 8 weeks
of treatment with FLV. However, it is important to
note that only three SRPs were examined after
treatment, that significant improvement in one SRP
was found among PBO children, and that approximately one fourth of children treated with FLV continued to experience difficulty sleeping alone and/or
away from home after treatment. As such, the potential
role of neurobiological, biobehavioral, and environmental factors in the presentation of SRPs among
anxious children requires further investigation. As a first
step toward gathering these data, both pharmacological
and behavioral treatment studies should include
structured assessment of children_s sleep problems at
baseline, after treatment, and at follow-up intervals.
Limitations

We examined eight specific SRPs in the present

study, but many other potential SRPs were not

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 46:2, FEBRUARY 2007

examined (e.g., bedtime resistance, poor sleep hygiene)

and should be included in future research. The present
study did not include a standardized instrument to
assess sleep behaviors/problems but instead relied on
data from several parent- and clinician-reported measures as part of an existing database. Thus, findings
should be interpreted with caution because, in general,
sleep problems are common in children and do not
necessarily indicate the presence of a sleep disorder.
Along these lines, the extent to which parents of anxious
youth may have simply endorsed SRPs as symptoms
confirming their child_s anxiety diagnosis also is unclear.
Ultimately, because parent, child, or clinician report
does not provide an objective assessment of children_s
sleep, data based on the use of polysomnography and
actigraphy methods are needed to better determine sleep
characteristics of anxious youth. We also note that
because some sleep items and measures of anxiety
severity/impairment were drawn from the same instruments, the possibility of measurement effects exists.
Finally, because anxious children in the present study
volunteered to participate in a psychopharmacological
treatment trial and met specific inclusion/exclusion
criteria, the generalizability of phenomenological findings to all anxious youth awaits further study.
Clinical Implications

Findings from the present study indicate SRPs to be

a common feature of childhood anxiety disorders and
underscore the need for research in this area that
includes objective assessment of sleep. Results suggest
that clinicians should obtain detailed information
related to both sleep and anxiety in children presenting
with difficulties in either domain. Research in this area
may provide crucial information on the temporality of
the relationship between childhood SRPs and anxiety
because emerging data suggest sleep problems to be
early markers for nascent psychopathology, including
anxiety disorders (Gregory et al., 2005). The significant
association found between SRPs and impaired family
functioning also is worthy of further investigation. A
better understanding of the nature of this relationship
may provide guidance in designing effective interventions for both childhood anxiety and sleep problems.
Finally, data suggest that FLV may be an effective
treatment for children with comorbid anxiety and sleep
problems, although prospective data based on standardized instruments for assessing sleep are needed.

231

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

ALFANO ET AL.

Disclosure: The authors have no financial relationships to disclose.

REFERENCES
Achenbach TM, Edelbrock C (1991), Manual for the Child Behavior
Checklist. Burlington, VT: Queen City Printers, Inc.
Bruni O, Ferini-Strambi L, Russo PM (2006), Sleep disturbances and
teacher ratings of school achievement and temperament in children. Sleep
Med 7:43Y48
Dadds MR, Barrett PM, Rapee RM, Ryan S (1996), Family process and
child anxiety and aggression: an observational analysis. J Abnorm Child
Psychol 24:715Y734
Dahl RE (1996), The regulation of sleep and arousal: development and
psychopathology. Dev Psychopathol 8:3Y27
Ford DE, Kamerow DB (1989), Epidemiologic study of sleep disturbances
and psychiatric disorders: an opportunity for prevention? JAMA
262:1479Y1484
Ginsburg GS, Siqueland L, Masia-Warner C, Hedtke KA (2004), Anxiety
disorders in children: family matters. Cog Behav Pract 11:28Y43
Gregory AM, Caspi A, Eley TC, Moffitt TE, O_Connor TG, Poulton R
(2005), Prospective longitudinal associations between persistent sleep
problems in childhood and anxiety and depressive disorders in
adulthood. J Abnorm Child Psychol 33:157Y163
Gregory AM, O_Connor TG (2002), Sleep problems in childhood: a
longitudinal study of developmental change and association with
behavioral problems. J Am Acad Child Adolesc Psychiatry 41:964Y971
Guy W (1976), The clinical global impression scale, ECDEU Assessment
Manual for Psychopharmacology-Revised (DHEW Publication No.
ADM 76-338). Rockville, Md: Department of Health, Education and
Welfare Public Health Service, Alcohol, Drug Abuse, Mental Health
Administration, NIMH Psychopharmacology Research Branch, Division of Extramural Research
Hamilton M (1959), The assessment of anxiety states by rating. Br J Med
Psychol 32:50Y55
Kaufman J, Birmaher B, Brent D (1997), Schedule for affective disorders
and schizophrenia for school-aged children-present and lifetime version
(K-SADS-PL): initial reliability and validity. J Am Acad Child Adolesc
Psychiatry 36:980Y988
Kendall PC, Pimentel SS (2003), On the physiological symptom
constellation in youth with generalized anxiety disorder (GAD). J
Anxiety Disord 17:211Y221
Maier W, Buller R, Philipp M, Heuser I (1988), The Hamilton Anxiety
Scale: reliability, validity and sensitivity to change in anxiety and
depressive disorders. J Affect Disord 14:61Y68

March JS, Biederman J, Wolkow R (1998), Sertraline in children and

adolescents with obsessive-compulsive disorder: a multicenter randomized controlled trial. JAMA 280:1752Y1756
Masi G, Millepiedi S, Mucci M, Poli P, Bertini N, Milantoni L (2004),
Generalized anxiety disorder in referred children and adolescents. J Am
Acad Child Adolesc Psychiatry 43:752Y760
Mindell JA (1993), Sleep disorders in children. Health Psychol 12:151Y162
Mindell JA, Barrett KM (2002), Nightmares and anxiety in elementary aged
children: is there a relationship? Child Care Health Dev 28:317Y322
Mindell JA, Owens JA, Carskadon MA (1999), Developmental features of
sleep. Child Adol Psychiatr Clin N Am 8:695Y725
Monti JM, Monti D (2000), Sleep disturbance in generalized anxiety
disorder and its treatment. Sleep Med Rev 4:263Y276
Ohayon MM (1997), Prevalence of DSM-IV diagnostic criteria of insomnia:
distinguishing insomnia related to mental disorders from sleep disorders.
J Psychiatr Res 31:333Y346
Pina A, Silverman WK, Alfano CA, Saavedra L (2002), Diagnostic efficiency
of symptoms in the diagnosis of DSM-IV generalized anxiety disorder in
youth. J Child Psychol Psychiatry 43:959Y967
Rapoport J, Elkins R, Langer DH (1981), Childhood obsessive compulsive
disorder. Am J Psychiatry 138:1545Y1554
Research Units on Pediatric Psychopharmacology Anxiety Study Group
(RUPP) (2002), The Pediatric Anxiety Rating Scale (PARS): development and psychometric properties. J Am Acad Child Adolesc Psychiatry
41:1061Y1069
Research Unit on Pediatric Psychopharmacology Anxiety Study Group
(RUPP) (2001), Fluvoxamine for the treatment of anxiety disorders in
children and adolescents. N Engl J Med 344:1279Y1285
Saletu-Zyhlarz G, Saletu B, Anderer P (1997), Non-organic insomnia in
generalized anxiety disorder. Neuropsychobiology 36:117Y129
Sandor P, Shapiro CM (1994), Sleep patterns in depression and anxiety:
theory and pharmacological effects. J Psychosom Res 38:125Y139
Shaffer D, Gould MS, Brasic J (1983), A children_s global assessment scale
(CGAS). Arch Gen Psychiatry 40:1228Y1231
Uhde TW (2000), Anxiety disorders. In: Principles and Practice of Sleep
Medicine, Kryger MH, Roth T, Dement WC, eds. Philadelphia: W.B.
Saunders
Wagner KD, Berard R, Stein MB (2004), A multicenter, randomized,
double-blind, placebo-controlled trial of paroxetine in children and
adolescents with social anxiety disorder. Arch Gen Psychiatry
61:1153Y1162
Warren SL, Gunnar MR, Kagan J (2003), Maternal panic disorder: infant
temperament, neurophysiology, and parenting behaviors. J Am Acad
Child Adolesc Psychiatry 42:814Y825
Wolfson AR, Carskadon MA (1998), Sleep schedules and daytime
functioning in adolescents. Child Dev 69:875Y887

Associations Between Sleep Problems, Anxiety, and Depression in Twins at 8 Years of Age Alice M. Gregory, PhD, Fruhling V.
Rijsdijk, PhD, Ronald E. Dahl, MD, Peter McGuffin, PhD, Thalia C. Eley, PhD
Objectives: Associations between sleep and internalizing problems are complex and poorly understood. To better understand these
covarying difficulties, genetic and environmental influences were estimated by using a twin design. Methods: Three hundred 8-year-old
twin pairs reported on their anxiety and depression by completing the Screen for Childhood Anxiety Related Emotional Disorders and
the Children_s Depression Inventory. Parents reported on their children_s sleep problems by completing the Child Sleep Habits
Questionnaire. Results: Children reported by their parents to have different types of sleep problems self-reported more depression
symptoms than those without. The correlation between total sleep-problem score and depression was moderate. That between sleep
problems and anxiety was smaller and was not examined further. The association between sleep problems and depression was mainly
explained by genes, and there was substantial overlap between the genes influencing sleep problems and those influencing depression.
There was smaller influence from environmental factors making family members alike, and environmental factors making family
members different decreased the association between sleep problems and depression. Conclusions: A range of sleep difficulties are
associated with depression in school-aged children, and the overall association between the 2 difficulties may be largely influenced by
genes. Pediatrics 2006;118:1124Y1132.

232

J. AM . ACAD. CHILD ADOLESC. PSYCH IATRY, 46:2, FEBRUARY 2007

Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.