HHS Public Access

Author manuscript
Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
Author Manuscript

Published in final edited form as:

Clin Child Psychol Psychiatry. 2020 January ; 25(1): 189–199. doi:10.1177/1359104518822685.

Novel insights from actigraphy: Anxiety is associated with sleep

quantity but not quality during childhood
Emily M. Cohodes, Aviva Abusch, Paola Odriozola, Dylan G. Gee
Yale University, Department of Psychology

Abstract
Anxiety and sleep function change dynamically across development, and sleep dysfunction has
Author Manuscript

emerged as a correlate and predictor of anxiety in pediatric clinical samples. Despite this, previous
research has not investigated how associations between qualitative and quantitative measures of
sleep function change with anxiety across development, specifically from childhood to
adolescence. The present study used actigraphy collection to examine whether associations
between quantitative and qualitative sleep function and anxiety differed as a function of
developmental stage in a community pediatric sample (8–17 years old; N = 92). Age moderated
the association between anxiety and sleep quantity, but not sleep quality. Contrary to hypotheses,
higher anxiety was related to increased sleep for children, but not adolescents. Results suggest
age-related changes in the association between sleep function and anxiety across development,
with implications for targeting sleep-related interventions for youth with anxiety.
Author Manuscript

Keywords
Sleep; actigraphy; anxiety; childhood; adolescence

Sleep-wake functions are a critical target of research on arousal and modulatory systems,
highlighting the pervasiveness of sleep dysfunction in psychiatric disorders (Cuthbert &
Insel, 2010). This focus on sleep suggests that tracing the roots of sleep dysfunction—and its
association with psychopathology—throughout development may lead to greater
optimization of treatment for mental illness across the lifespan (El-Sheikh and Buckhalt,
2015). Up to 90% of children diagnosed with anxiety disorders report sleep-related
problems, and an estimated 88% of clinically anxious youth have a comorbid sleep disorder
diagnosis (Alfano, Ginsburg, & Newman Kingery, 2007; Storch et al., 2008). Although
specific patterns of sleep-related problems appear to be dependent on type of anxiety
Author Manuscript

disorder diagnosis (Alfano, Pina, Zerr, & Villalta, 2010), numerous studies have documented
that children and adolescents with anxiety disorders experience sleep-related problems
including more difficulty falling asleep and maintaining sleep, going to sleep later, and
getting less sleep than their non-anxious peers (Hudson, Gradisar, Gamble, Schniering, &
Rebelo, 2009; Stein, Mendelsohn, Obermeyer, Amromin, & Benca, 2001).

Previous research on associations between sleep and anxiety across development has
focused on the transition from childhood to adolescence, when individuals experience
significant shifts in both hormonal factors and social contexts that have been hypothesized to
underlie sleep disturbance (McMakin & Alfano, 2015), and when rates of psychiatric
 Cohodes et al. Page 2

disorder onset peak for both males and females (Kessler et al., 2005). Higher rates of sleep-
Author Manuscript

related problems (e.g., sleep disturbance, bedtime resistance, sleep anxiety, and parasomnia)
are observed in younger children diagnosed with an anxiety disorder, as compared to
adolescents (Alfano et al., 2010), which is consistent with previous reports that sleep-related
problems may more commonly co-occur with anxiety disorders earlier in development
(Alfano, Zakem, Costa, Taylor, & Weems, 2008; Dahl, 1996). However, both sleep function
and the prevalence of anxiety symptomatology—as well as frontolimbic circuitry that
underlies anxiety—undergo dynamic changes during adolescence (Casey, Heller, Gee, &
Cohen, 2017), with adolescence representing a peak in anxiety disorder onset (Kessler et al.,
2005), as well as a developmental period characterized by significant changes in circadian
rhythms (Darchia & Cervena, 2014).

In addition to discrepant associations between anxiety and sleep during childhood versus
adolescence, developmental changes in the biological processes underlying the emergence of
Author Manuscript

both sleep disturbance and anxiety suggest that age may be an important moderator of
associations between sleep quality and quantity and anxiety-related symptoms, particularly
during the transition from childhood to adolescence (Gee et al., 2013; Gabard-Durnam et al.,
2014; Hare et al., 2008; Kessler et al., 2005; Skeldon, Derks, & Dijk, 2016). Furthermore,
sleep dysfunction has emerged as both a correlate and predictor of anxiety in pediatric
clinical and community samples (Gregory, Van der Ende, Willis, & Verhulst, 2008; Roberts
& Duong, 2017; Shanahan, Copeland, Angold, Bondy, & Costello, 2014). Such findings
have provided evidence for the bidirectional nature of the relation between sleep problems
and anxiety disorders and prompted further investigation of how the association between
sleep and anxiety may change during development (Sadeh, Tikotzky, & Kahn, 2014). These
findings have also promoted the theory that sleep problems in childhood may contribute risk
for the onset of anxiety-related symptomatology in adolescence (Cowie et al., 2014; Kelly &
Author Manuscript

El-Shiekh, 2015; Willis & Gregory, 2015).

Although recent advances in technology have facilitated the objective measurement of sleep,
the majority of studies examining associations between sleep and anxiety disorders in
pediatric samples have employed subjective (typically parent-report) measures of sleep
quality and sleep-related problems (Willis & Gregory, 2015) with fewer studies examining
associations between objective measures of sleep-related problems and anxiety disorders.
Among studies employing objective measures of sleep function—including
polysomnography and actigraphy—findings have been inconsistent and appear to vary by
type of objective measurement. Polysomnography studies have identified disruptions in
sleep among anxious youth. One study reported that clinically anxious children and
adolescents (age 7–17) exhibited more nighttime awakenings, less slow-wave sleep, and
Author Manuscript

longer sleep onset latency, as measured with polysomnography, than depressed and control
groups (Forbes et al., 2008). In a separate sample of children age 7–11, children diagnosed
with generalized anxiety disorder (GAD) showed increased sleep onset latency and reduced
latency to REM sleep, also measured by polysomnography (Alfano, Reynolds, Scott, Dahl,
& Mellman, 2012). However, prior research employing actigraphy as an objective measure
of sleep quality has not identified sleep-related differences between clinically anxious youth
and non-anxious controls. A comparison of youth with social anxiety disorder and controls
revealed no differences on actigraphy-based measurements of sleep quality (Mesa, Beidel, &

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 3

Bunnell, 2014) and, similarly, a comparison of children with GAD and same-aged controls
Author Manuscript

found no differences on actigraphy-based sleep function (Alfano, Patriquin, De Los Reyes,

2015).

Prior evidence has also highlighted discordance between subjective and objective reports of
sleep quality and sleep-related problems across development. For example, previous studies
have found weak and non-significant associations between subjective reports of sleep quality
(parent- and child-report) and objective measures of sleep quality (e.g., actigraphy; Alfano et
al., 2015; Gregory et al., 2011). Taken together, findings spanning studies employing
subjective and multiple types of objective measurement of sleep imply a strong need for
further investigation using objective measures of sleep quality in both clinical and
community samples to determine whether associations between anxiety and subjectively-
reported sleep function are consistent with associations between anxiety and objectively-
measured sleep function.
Author Manuscript

In this vein, the present study builds on the extant literature in several major ways—by
contributing to the limited literature examining associations between objective sleep function
and anxiety-related symptomatology in community samples, and specifically, by testing
associations between sleep function and anxiety-related symptomatology across the
transition from childhood to adolescence, a developmental period when frontoamygdala
circuitry, which is hypothesized to be influenced by circadian rhythms (Dayan et al., 2017;
Pace-Schott et al., 2015), is particularly in flux (Gee et al., 2013; Gabard-Durnam et al.,
2014; Hare et al., 2008). The current study aimed to examine the following questions: 1)
What is the relation between anxiety and objective measures of sleep quantity and quality in
a community sample of children and adolescents? and 2) How does the relation between
objective sleep function and anxiety symptomatology differ by age among 8–17-year-olds?
Author Manuscript

Given previous reports of discrepant findings between subjective and objective measures of
sleep quality, these associations were also tested using a subjective measure of sleep for a
subset of adolescents who had available data on self-reported sleep function. It was
hypothesized that lower quantitative and qualitative sleep function would be associated with
higher levels of anxiety and that this association would be strongest in adolescence.

Methods
Participants
Participants were a subset of subjects from the publicly available Enhanced Nathan Kline
Institute Rockland Sample (NKI-RS; Nooner et al., 2012). Cross-sectional behavioral data
from 1178 participants ages 6 – 80 were retrieved from the NKI-RS sample; data were
Author Manuscript

collected from 2012–2016. Based on the inclusion and exclusion criteria described below,
the final sample consisted of 92 participants (54.30% male) age 8–17 (Mage = 12.73, SD =
2.89). Participants were excluded if they exhibited low intellectual functioning (IQ < 80;
measured using Wechsler Abbreviated Scales of Intelligence-II; Wechsler, 1999) or if they
had missing data on the behavioral measures of interest. After assessing for all exclusionary
criteria, outliers on each behavioral measure of interest (defined as three standard deviations
or more from the mean) were also excluded from analyses. All subjects provided written
informed consent (assent for minors) to unrestricted distribution of anonymous data through

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 4

the International Neuroimaging Data-Sharing Initiative (INDI1; 1000 Functional

Author Manuscript

Connectomes Project [FCP], 2009) according to the procedures set forth by the NKI
Institutional Review Board. 14.1% of participants in the final sample met DSM-IV criteria
for a current anxiety disorder diagnosis (n = 13).

Measures
Participants were administered a comprehensive battery of questionnaire, biological, and
behavioral measures as part of the broader NKI-RS procedures. In the present study, we
utilized assessments of anxiety symptomatology and sleep. Anxiety symptomatology. The
Multidimensional Anxiety Scale for Children (MASC; March & Parker, 2004) was used to
assess anxiety symptomatology. The MASC is a 39-item self-report questionnaire that
assesses anxiety in youth and yields the following subscales: physical symptoms, harm
avoidance, social anxiety, separation/panic, and total anxiety. Participants were asked to rate
Author Manuscript

their own behavior on a 4-point scale ranging from 0 (“Never true about me”) to 3 (“Often
true about me”). The present study used the total anxiety raw score, which had good internal
consistency (39 items, Cronbach α = 0.89). The MASC has been shown to have good
reliability and validity (March, Parker, Sullivan, & Stallings, 1997). A higher score on the
MASC indicates higher self-reported anxiety-related symptomatology.

Sleep function.
Objective sleep function.: Actigraphy data were collected using a Philips Respironics
Actiwatch 2. Participants wore the waterproof unit continuously on their non-dominant wrist
for 1–18 days (mean = 7.96 days) depending on the spacing of participants’ first and second
laboratory visits, in order to collect data on sleep function, respiration, and heart rate.
Participants were instructed to press the event marker button immediately before going to
Author Manuscript

sleep. The Philips Respironics Actiwatch 2 has been shown to have good interunit reliability
and criterion validity (Gironda, Lloyd, Clark, & Walker, 2007). Mean hours of sleep per
night was used as the primary measure of quantitative objective sleep function in the present
study. The following variables were used as measures of qualitative objective sleep function:
average sleep onset (minutes), average sleep efficiency percentage (percentage of time
asleep), average duration of awakenings after sleep onset (minutes), and average number of
awakenings.

Subjective sleep function.: The Pittsburgh Sleep Quality Index (PSQI; Buysse, Reynolds,
Monk, Berman, & Kupfer, 1989), a well-validated and widely used measure that is
correlated with other subjective measurements of sleep function (e.g., sleep diaries) (Buysse
et al. 1989; Backhaus et al. 2002), was used to assess subjective sleep function in
Author Manuscript

participants age 13–17. The overall sleep quality scale of the PSQI was used as the primary
subjective qualitative sleep variable for a subset of participants who completed the
questionnaire. A higher score on the PSQI indicates worse self-reported sleep quality.

Socioeconomic status.: To control for socioeconomic status, participants’ parents completed

the Hollingshead Four Factor Index of Socioeconomic Status (Hollingshead, 1975), which is
designed to assess social status based on marital status, employment status, educational
attainment, and occupation prestige. Parental education was rated on a 7-point scale ranging

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 5

from 7 (graduate or professional training) to 1 (less than 7th grade). Parental occupation was
Author Manuscript

rated on a 9-point scale ranging from 9 (higher executive, proprietor of large businesses,
major professional) to 1 (farm laborers, menial service workers, students, housewives,
dependent on welfare, no regular occupation). Scores were summed across these dimensions
to yield the child SES composite score, which was used in the present study.

Data Analysis Plan

Multiple linear regression was conducted using the PROCESS Macro for SPSS (Hayes,
2015) to test the hypothesis that sleep function would be negatively correlated with anxiety
symptomatology, with age examined as a potential moderator of this association. Overall
anxiety symptomatology was the dependent variable in one family of regression analysis
that examined age as a moderator of the association between quantitative sleep function and
anxiety symptomatology (independent variables: age, average number of hours of sleep per
Author Manuscript

night). In a second family of tests that examined age as a moderator of the association
between qualitative sleep function and anxiety, four separate models were run to examine
each qualitative sleep variable of interest (i.e., a separate model for each of the following
variables: average sleep onset (minutes), average sleep efficiency percentage (percentage of
time asleep) average duration of awakenings after sleep onset (minutes), and average number
of awakenings). Because there were four sleep variables of interest within the family of
regression analyses focusing on qualitative sleep function, Bonferroni correction for multiple
comparisons was applied (.05/4 = p < .0125). Sleep function and age were coded
continuously and were mean-centered (Aiken & West, 1991) and multiplied to yield the
interaction term in the model. In order to better assess for R2 change, the analysis was
conducted as a hierarchical multiple linear regression in which the initial step included the
primary independent variables, followed by the two-way interactions (sleep × age) in the
Author Manuscript

second step. Standardized Beta coefficients were used as an estimate of effect size.

For a subsample of adolescent participants with subjective, self-reported sleep function data
available (n = 43), hierarchical multiple linear regression [initial step: primary independent
variables (overall sleep quality [PSQI], age), second step: two-way interaction (overall sleep
quality × age) with overall anxiety symptoms as the dependent variable] was conducted to
test whether age moderated the association between subjective sleep quality and anxiety
symptomatology. Due to recommendations that participants wear actigraphy monitors for at
least five days during collection of objective sleep function data (Sadeh & Acebo, 2002), all
analyses were re-run in the subsample of participants who wore the actigraphy monitor for at
least five days (n = 73). All findings in this subsample were consistent with the results
reported in the overall sample.
Author Manuscript

Results
Preliminary Analyses
Seven outliers were excluded from the sample (remaining n = 92). Skew and kurtosis cutoffs
(West, Finch, & Curran, 1995) indicated normal distributions across all main study variables
in the final sample. Descriptive statistics for main demographic, predictor, and outcome
variables of the final sample are presented in Table 1.

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 6

Correlation analyses were conducted to examine associations between the main study
Author Manuscript

variables. Zero-order correlations among main study variables are presented in Table 2. Of
note, anxiety was positively correlated with the quantitative index of sleep function, average
hours of sleep per night, r(90) = 0.24, p = .024. There was no significant association between
anxiety and any of the measures of qualitative sleep function (all p-values > .50). Age was
negatively correlated with average hours of sleep per night, r(90) = −0.34, p = .001 and with
anxiety, r(90) = −0.23, p = .031, respectively. In a subsample of adolescents (age 13–17)
with available subjective sleep function data (n = 43), self-reported overall sleep quality was
not associated with objective hours of sleep per night (as measured by actigraphy), r(41) =
−0.18, p = .25, but was significantly positively correlated with average sleep onset latency
(as measured by actigraphy), r(41) = −0.40, p = .008, suggesting that participants who took
longer to fall asleep reported poorer sleep quality. Subjective sleep function was not
significantly associated with any of the other qualitative indices of sleep quality, as measured
Author Manuscript

by the actigraphy monitor (all p-values > .30).

Preliminary analyses were also conducted to test whether participant sex or SES were
related to any predictor variables (quantitative and qualitative indices of sleep function,
anxiety, age). Results showed no differences by sex (all p-values > .05), with the exception
of anxiety, which was higher for females (M = 46.95, SD = 18.17) than males (M = 36.94,
SD = 15.45), t(90)= −2.86, p = .005. Additionally, there were no significant associations
between SES with quantitative or qualitative indices of sleep quality or age (all p-values > .
10). Anxiety was significantly positively correlated with SES, r(90) = 0.34, p = .001, such
that individuals with higher SES reported higher levels of anxiety symptomatology. Given
these preliminary findings, all analyses were conducted with and without SES and sex
included as covariates in the model (covariates were entered in the first block before the
dependent variable).
Author Manuscript

Regression Analyses
In the first family of tests examining sleep quantity, a two-way interaction (quantitative
indicator of sleep function × age) emerged as a significant predictor of anxiety, B = −0.47,
SE = 0.23, p = .039. Among children (ages 8–11, based on median split), sleep was a
significant predictor of anxiety, B = 2.13, SE = 0.79, p = .010, whereas, in contrast, there
was no significant association between quantitative sleep function and anxiety among
adolescents (ages 12–17), B = −0.11, SE = 0.76, p = .883. In the second family of tests
examining sleep quality, no significant interactions between any of the qualitative indicators
of sleep function and age emerged as a significant predictor of anxiety. Specifically, the two-
way interactions between average onset sleep latency and age did not predict anxiety, B =
−0.04, SE = 0.05, p = .414, nor did the interaction between average sleep efficiency and age
Author Manuscript

predict anxiety, B = −0.03, SE = 0.07, p = .685. Similarly, the two-way interaction between
average number of awakenings after sleep onset and age was not a significant predictor of
anxiety, B = −0.00, SE = 0.01, p = .920, nor was the interaction between average number of
awakenings and age, B = 0.01, SE = 0.07, p = .853. In the subsample of participants with
self-reported qualitative sleep function data available, the age × self-reported sleep quality
interaction did not predict anxiety, B = 3.21, SE = 3.27, p = .331. Results did not differ when

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 7

analyses were conducted in the subsample of participants who wore the actigraphy monitor
Author Manuscript

for at least 5 days (n = 73).

Discussion
In their recent review of a forum on sleep across child development sponsored by the
Society for Research on Child Development, El-Sheikh and Buckhalt (2015) outline a model
for integrating investigations of the role of sleep in the developmental trajectories of
psychopathology and other health-related outcomes, underscoring the numerous
implications of such work for clinical intervention and policy. The present study—which
highlights age as a key moderator of patterns of association between sleep and anxiety—
contributes to our understanding of the role of sleep in the development of anxiety during the
transition from childhood to adolescence. Consistent with previous studies that have
identified childhood, as compared to adolescence, as a developmental period characterized
Author Manuscript

by more sleep-related problems for individuals with anxiety disorders (Alfano et al., 2010),
the present findings suggest that sleep function was associated with anxiety-related
symptomatology in childhood, but not adolescence. In addition, contrary to our hypotheses,
results of the current study suggest that relatively increased sleep, rather than decreased
sleep, obtained by an objective measurement of sleep function, is associated with higher
anxiety-related symptomatology in childhood, highlighting increased somnolence in
childhood as a potential clinical marker of anxiety symptomatology onset.

The current findings add to a limited but growing body of literature testing concordance
between objective measures of sleep function and anxiety in community pediatric samples.
In stark contrast to robust patterns of association between subjective sleep function and
anxiety in previous studies (e.g., Alfano et al., 2007), actigraphy was not negatively
Author Manuscript

associated with all indices of anxiety symptomatology, as hypothesized. In addition, in a

subsample of adolescent participants with available self-reported sleep function data,
subjective sleep function and actigraphy measures were not associated. Building on several
other studies that have noted a dissociation between subjective and objective reports of sleep
function (e.g., Alfano et al., 2015), these findings suggest that there may be divergence in
associations between objectively- and subjectively-measured sleep and anxiety
symptomatology, particularly in developmental samples.

With regard to discrepancies between qualitative and quantitative indices of actigraphy

measurement, results of the present study suggest that the number of hours of sleep a child
gets may be a more important indicator of anxiety-related symptomatology than the quality
of sleep during those hours. Discord between self-reported and actigraphy-based indices of
sleep function may be due, in part, to the fact that sleep patterns change dramatically across
Author Manuscript

development in the age range of the present sample (Colrain & Baker, 2011); therefore,
youths may not have an accurate “baseline” from which to record deviations in sleep in a
self-report questionnaire. In other words, adolescents’ perception of their own sleep quality
may be less accurate given that they are experiencing ongoing changes in sleep during this
period of development. Additionally, differences in specific measures used to assess sleep
quality may have contributed to discrepant findings across the literature and, since only 13–
17-year-olds in our study had qualitative sleep quality data, the restricted age range may

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 8

have contributed to the observed findings. Further research is needed to continue to test
Author Manuscript

these associations in a broader age range.

Though the current study suggests that the association between sleep and anxiety may be
age-dependent, longitudinal research is needed to further probe the directionality of this
association. Previous research testing associations between daytime affect and objectively-
measured sleep quality sheds light on the bidirectional nature of the relation between sleep
quality and symptomatology. Within a sample of clinically anxious youth, increased negative
affect was associated with more time spent awake the following night (measured by
actigraphy), and more time spent awake at night predicted higher levels of negative affect the
following day (Cousins et al., 2011), suggesting that impaired sleep function predicts and
results from negative affect, which is characteristic of anxiety disorders (Watson, Clark, &
Tellegen, 1988). Examining these processes longitudinally, with a particular focus on the
transition from childhood to adolescence, would further elucidate the role of sleep in the
Author Manuscript

emergence of anxiety disorders. Specifically, longitudinal designs will clarify whether

decreased sleep results in development or exacerbation of symptoms of anxiety, symptoms
of anxiety result in less sleep attainment, or whether both mechanisms contribute to the
phenomenon observed in this study.

Limitations of the present study include a variable sampling period for actigraphy data,
though results did not change when analyses were conducted in a subsample of participants
with at least five days of actigraphy collection. Measurement of actigraphy data may be
varied due to the nature of data collection (wearing device during sleep), duration of
monitoring (and whether monitoring is continuous or disrupted), weight, and sex (e.g.,
Brown, Smolensky, D’Alonzo, & Redman, 1990; Tyron, 1987). Further, as previously noted,
subjective sleep quality data were only assessed in a subset of participants (age 13–17),
Author Manuscript

precluding an investigation of age-related changes in the association between subjective

sleep quality and anxiety-related symptomatology during the transition from childhood to
adolescence.

Future research should continue to incorporate both subjective (parent- and child-report) and
objective measures of sleep function to assess potential dissociations between objective and
subjective measures of sleep in developmental samples. Furthermore, future studies would
benefit from assessing additional predictors of discordance between subjective and objective,
as well as quantitative versus qualitative, measures of sleep function. Consistent with
McMakin & Alfano’s (2015) call for consideration of the dimensionality of anxiety
disorders in studies of the association between sleep and anxiety, in order to assess whether
the significant association between anxiety and sleep quantity observed in children in the
Author Manuscript

present study is driven by a specific anxiety-related symptom cluster, future studies could
assess associations between sleep function and specific clusters of anxiety-related symptoms
(e.g., somatic symptoms of anxiety, panic, worry). Furthermore, previous research has
posited that sleep and circadian rhythm disruption may interfere with extinction learning via
alterations in frontoamygdala circuitry (Dayan et al., 2017; Pace-Schott et al., 2015). Given
dynamic changes in this circuitry across development (Gee et al., 2013; Gabard-Durnam et
al., 2014; Hare et al., 2008), it is critical to further explore the role of sleep on the
developing brain and the emergence of anxiety disorders.

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 9

In summary, this study utilized actigraphy to objectively measure sleep function, and results
Author Manuscript

were inconsistent with previous findings on subjectively-rated sleep function and anxiety,
highlighting the need for future studies to employ both subjective and objective measures of
sleep function during development. Based on the unique associations between sleep and
anxiety-related symptomatology during childhood and adolescence observed in the present
study, these findings lay the groundwork for future research to further probe associations
between multiple domains of sleep function and anxiety across development.

Acknowledgments
This research was supported by a Brain and Behavior Research Foundation Young Investigator Award to D.G.G.,
NIH Director’s Early Independence Award (grant number DP5OD021370) to D.G.G., and an NSF GRFP award to
E.M.C. Thank you to F. Xavier Castellanos, Bennett Leventhal, and Michael Milham for providing the Nathan
Kline Institute/Rockland Sample.
Author Manuscript

References
Aiken LS, & West SG (1991). Multiple regression: Testing and interpreting interactions. Newbury Pak:
Sage.
Alfano CA, Ginsburg GS, & Newman Kingery JN (2007). Sleep-related problems among children and
adolescents with anxiety disorders. Journal of the American Academy of Child & Adolescent
Psychiatry, 46(2), 224–232. doi:10.1097/01.chi.0000242233.06011.8e [PubMed: 17242626]
Alfano CA, Patriquin MA, & De Los Reyes A (2015). Subjective–objective sleep comparisons and
discrepancies among clinically-anxious and healthy children. Journal of Abnormal Child
Psychology, 43(7), 1343–1353. doi:10.1007/s10802-015-0018-7 [PubMed: 25896729]
Alfano CA, Pina AA, Zerr AA, & Villalta I (2010). Pre-sleep arousal and sleep problems of anxiety-
disordered youth. Child Psychiatry & Human Development, 41, 156. doi:10.1007/
s10578-0090158-5 [PubMed: 19680805]
Alfano CA, Reynolds K, Scott N, Dahl RE, & Mellman T (2012). Polysomnographic sleep patterns of
non-depressed, non-medicated children with generalized anxiety disorder. Journal of Affective
Author Manuscript

Disorders, 147, 379–384. doi:10.1016/j.jad.2012.08.015 [PubMed: 23026127]

Alfano CA, Zakem AH, Costa NM, Taylor LK, & Weems CF (2009). Sleep problems and their relation
to cognitive factors, anxiety, and depressive symptoms in children and adolescents. Depression and
Anxiety, 26(6), 503–512. doi: 10.1002/da.20443 [PubMed: 19067319]
Backhaus J, Junghanns K, Broocks A, Riemann D, & Hohagen F (2002). Test–retest reliability and
validity of the Pittsburgh Sleep Quality Index in primary insomnia. Journal of Psychosomatic
Research, 53(3), 737–740. doi: 10.1016/S00223999(02)00330-6 [PubMed: 12217446]
Brown AC, Smolensky MH, D’Alonzo GE, & Redman DP (1990). Actigraphy: A means of assessing
circadian patterns in human activity. Chronobiology International, 7(2), 125–133. doi:
10.3109/07420529009056964 [PubMed: 2242506]
Buysse DJ, Reynolds CF, Monk TH, Berman SR, & Kupfer DJ (1989). The Pittsburgh Sleep Quality
Index: A new instrument for psychiatric practice and research. Psychiatry Research, 28(2), 193–
213. doi: 10.1016/0165-1781(89)90047-4 [PubMed: 2748771]
Casey BJ, Heller AS, Gee DG, & Cohen AO (2017). Development of the Emotional Brain.
Author Manuscript

Neuroscience letters. doi: 10.1016/j.neulet.2017.11.055

Colrain IM, & Baker FC (2011). Changes in sleep as a function of adolescent development.
Neuropsychology review, 21(1), 5–21. doi: 10.1007/s11065-010-9155-5 [PubMed: 21225346]
Cousins JC, Whalen DJ, Dahl RE, Forbes EE, Olino TM, Ryan ND, & Silk JS (2011). The
bidirectional association between daytime affect and nighttime sleep in youth with anxiety and
depression. Journal of Pediatric Psychology, 36(9), 969–979. doi:10.1093/jpepsy/jsr036 [PubMed:
21795377]
Cowie J, Alfano CA, Patriquin MA, Reynolds KC, Talavera D, & Clementi MA (2014). Addressing
sleep in children with anxiety disorders. Sleep Medicine Clinics, 9(2), 137–148.

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 10

Cuthbert BN, & Insel TR (2013). Toward the future of psychiatric diagnosis: the seven pillars of
RDoC. BMC Medicine, 11(1), 126. doi:10.1186/1741-7015-11-126 [PubMed: 23672542]
Author Manuscript

Dahl RE (1996). The regulation of sleep and arousal: Development and psychopathology.
Development and Psychopathology, 8(01), 3–27. doi:10.1017/S0954579400006945
Darchia N, & Cervena K (2014). The journey through the world of adolescent sleep. Reviews in the
Neurosciences, 25(4), 585–604. doi: 10.1515/revneuro-2013-0065 [PubMed: 24717334]
Dayan J, Rauchs G, & Guillery-Girard B (2017). Rhythms dysregulation: A new perspective for
understanding PTSD? Journal of Physiology-Paris. doi:10.1016/j.jphysparis.2017.01.004
El-Sheikh M, & Buckhalt JA (2015). II. Moving sleep and child development and child development
research forward: Priorities and recommendations from the SRCD sponsored forum on sleep and
child development. Monographs of the Society for Research in Child Development, 80(1), 15–32.
doi: 10.1111/mono.12142 [PubMed: 25704733]
Forbes EE, Bertocci MA, Gregory AM, Ryan ND, Axelson DA, Birmaher B, & Dahl RE (2008).
Objective sleep in pediatric anxiety disorders and major depressive disorder. Journal of the
American Academy of Child and Adolescent Psychiatry, 47(2), 148–155. doi:10.1097/chi.
0b013e31815cd9bc [PubMed: 18176336]
Author Manuscript

Gabard-Durnam LJ, Flannery J, Goff B, Gee DG, Humphreys KL, Telzer E, … & Tottenham N (2014).
The development of human amygdala functional connectivity at rest from 4 to 23 years: A cross-
sectional study. Neuroimage, 95, 193–207. doi: 10.1016/j.neuroimage.2014.03.038 [PubMed:
24662579]
Gee DG, Humphreys KL, Flannery J, Goff B, Telzer EH, Shapiro M, … & Tottenham N (2013). A
developmental shift from positive to negative connectivity in human amygdala–prefrontal circuitry.
Journal of Neuroscience, 33(10), 4584–4593. doi: 10.1523/JNEUROSCI.3446-12.2013 [PubMed:
23467374]
Gironda RJ, Lloyd J, Clark ME, & Walker RL (2007). Preliminary evaluation of reliability and
criterion validity of Actiwatch-Score. Journal of Rehabilitation Research and Development, 44(2),
223–230. doi:10.1682/JRRD.2006.06.0058 [PubMed: 17551874]
Gregory AM, Cousins JC, Forbes EE, Trubnick L, Ryan ND, Axelson DA, … & Dahl RE (2011).
Sleep items in the child behavior checklist: a comparison with sleep diaries, actigraphy, and
polysomnography. Journal of the American Academy of Child & Adolescent Psychiatry, 50(5),
499–507. doi: 10.1016/j.jaac.2011.02.003 [PubMed: 21515199]
Author Manuscript

Gregory AM, Van der Ende J, Willis TA, & Verhulst FC (2008). Parent-reported sleep problems during
development and self-reported anxiety/depression, attention problems, and aggressive behavior
later in life. Archives of Pediatrics & Adolescent Medicine, 162(4), 330–335. doi:10.1001/
archpedi.162.4.330 [PubMed: 18391141]
Hare TA, Tottenham N, Galvan A, Voss HU, Glover GH, & Casey BJ (2008). Biological substrates of
emotional reactivity and regulation in adolescence during an emotional go-nogo task. Biological
Psychiatry, 63(10), 927–934. doi: 10.1016/j.biopsych.2008.03.015 [PubMed: 18452757]
Hayes AF (2015). The PROCESS macro for SPSS and SAS. Retrieved December 10, 2015, from
www.processmacro.org.
Hollingshead AB (1975). Four factor index of social status. New Haven, CT: Author.
Hudson JL, Gradisar M, Gamble A, Schniering CA, & Rebelo I (2009). The sleep patterns and
problems of clinically anxious children. Behaviour Research and Therapy, 47(4), 339–344. doi:
10.1016/j.brat.2009.01.006 [PubMed: 19233345]
Kelly RJ, & El-Sheikh M (2014). Reciprocal relations between children’s sleep and their adjustment
Author Manuscript

over time. Developmental Psychology, 50(4), 1137. [PubMed: 24188035]

Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, & Walters EE (2005). Lifetime prevalence
and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey
Replication. Archives of General Psychiatry, 62(6), 593–602. doi: 10.1001/archpsyc.62.6.593
[PubMed: 15939837]
March JS, & Parker JD (2004). The multidimensional anxiety scale for children (MASC) In Maruish
M(Ed.), The use of psychological testing for treatment planning and outcomes assessment; Vol. 2,
Instruments for children and adolescents (3rd ed, pp. 39–62). Mahwah, NJ: Erlbaum.

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 11

March JS, Parker JD, Sullivan K, Stallings P, & Conners CK (1997). The Multidimensional Anxiety
Scale for Children (MASC): Factor structure, reliability, and validity. Journal of the American
Author Manuscript

Academy of Child & Adolescent Psychiatry, 36(4), 554–565. doi:

10.1097/00004583-199704000-00019 [PubMed: 9100431]
McMakin DL, & Alfano CA (2015). Sleep and anxiety in late childhood and early adolescence.
Current Opinion in Psychiatry, 28(6), 483. doi: 10.1097/YCO.0000000000000204 [PubMed:
26382163]
Mesa F, Beidel DC, & Bunnell BE (2014). An examination of psychopathology and daily impairment
in adolescents with social anxiety disorder. PloS one, 9(4), e93668. doi: 10.1371/journal.pone.
0093668 [PubMed: 24691406]
Nooner KB, Colcombe SJ, Tobe RH, Mennes M, Benedict MM, Moreno AL, … Milham MP (2012).
The NKI-Rockland Sample: A model for accelerating the pace of discovery science in psychiatry.
Frontiers in Neuroscience, 6, 152. doi:10.3389/fnins.2012.00152 [PubMed: 23087608]
Pace-Schott EF, Germain A, & Milad MR (2015). Sleep and REM sleep disturbance in the
pathophysiology of PTSD: The role of extinction memory. Biology of Mood & Anxiety Disorders,
5(1). doi:10.1186/s13587-015-0018-9
Author Manuscript

Roberts RE, & Duong HT (2017). Is there an association between short sleep duration and adolescent
anxiety disorders? Sleep Medicine, 30, 82–87. doi:10.1016/j.sleep.2016.02.007 [PubMed:
28215269]
Sadeh A, & Acebo C (2002). The role of actigraphy in sleep medicine. Sleep Medicine Reviews, 6(2),
113–124. [PubMed: 12531147]
Sadeh A, Tikotzky L, & Kahn M (2014). Sleep in infancy and childhood: implications for emotional
and behavioral difficulties in adolescence and beyond. Current Opinion in Psychiatry, 27(6), 453–
459. doi: 10.1097/YCO.0000000000000109 [PubMed: 25247458]
Shanahan L, Copeland WE, Angold A, Bondy CL, & Costello EJ (2014). Sleep problems predict and
are predicted by generalized anxiety/depression and oppositional defiant disorder. Journal of the
American Academy of Child & Adolescent Psychiatry, 53(5), 550–558. doi:10.1016/j.jaac.
2013.12.029 [PubMed: 24745954]
Skeldon AC, Derks G, & Dijk DJ (2016). Modelling changes in sleep timing and duration across the
lifespan: changes in circadian rhythmicity or sleep homeostasis?. Sleep Medicine Reviews, 28, 96–
107. doi: 10.1016/j.smrv.2015.05.011 [PubMed: 26545247]
Author Manuscript

Stein MA, Mendelsohn J, Obermeyer WH, Amromin J, & Benca R (2001). Sleep and Behavior
Problems in School-Aged Children. Pediatrics, 107(4). doi:10.1542/peds.107.4.e60
Storch EA, Murphy TK, Lack CW, Geffken GR, Jacob ML, & Goodman WK. (2008). Sleep-related
problems in pediatric obsessive compulsive disorder. Journal of Anxiety Disorders, 22, 877–885.
doi:10.1016/j.janxdis.2007.09.003 [PubMed: 17951025]
Tryon WW (1987). Activity as a function of body weight. The American Journal of Clinical Nutrition,
46(3), 451–455. doi: 10.1093/ajcn/46.3.451 [PubMed: 3630964]
Watson D, Clark LA, & Tellegen A (1988). Development and validation of brief measures of positive
and negative affect: the PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063.
doi: 10.1037/0022-3514.54.6.1063 [PubMed: 3397865]
Wechsler D (1999). Wechsler Abbreviated Scale of Intelligence. San Antonio, TX: The Psychological
Corporation.
West SG, Finch JF, & Curran PJ (1995). Structural equation models with nonnormal variables:
Problems and remedies In Hoyle RH (Ed.), Structural equation modeling: Concepts, issues, and
Author Manuscript

applications (pp. 56–75). Thousand Oaks, CA: Sage Publications, Inc.

Willis TA, & Gregory AM (2015). Anxiety disorders and sleep in children and adolescents. Sleep
Medicine Clinics, 10(2), 125–131. doi:10.1016/j.jsmc.2015.02.002 [PubMed: 26055860]

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Cohodes et al. Page 12

Table 1.

Descriptive statistics of study variables

Author Manuscript

Variable Min-Max Mean SD Variance Skewness Kurtosis

Child age 8.27–17.94 12.73 2.88 8.34 0.01 −1.31
Overall symptoms of anxiety 2.00–94.00 41.51 17.39 302.45 0.36 0.15
Average hours of sleep per night 0.17–10.45 5.64 3.26 10.68 −0.80 −0.87
Average onset latency (minutes) 0.00–61.17 13.60 13.56 184.00 1.88 3.71
Average sleep efficiency percentage 53.21–95.21 81.06 9.82 96.44 −1.12 0.67
Average wakefulness after sleep onset (minutes) 14.00–219.69 70.73 46.01 2116.89 1.28 1.26
Average number of awakenings 16.00–56.14 36.53 9.67 93.49 0.09 −0.59
Author Manuscript
Author Manuscript
Author Manuscript

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
 Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Table 2.

Correlations among all main study variables

Average
Average sleep wakefulness after Average
Cohodes et al.

Overall symptoms of Average hours of Average sleep onset efficiency sleep onset number of
Child age anxiety sleep per night latency (minutes) percentage (minutes) awakenings
Child age --
Overall symptoms of anxiety −.23* --
Average hours of sleep per night −.34* .24 --
Average sleep onset latency (minutes) .27* .01 .03 --
Average sleep efficiency percentage −.17 −.05 .22* −.29** --
Average wakefulness after sleep onset
−.01 −.01 −.11 −.08 −.85** --
(minutes)
Average number of awakenings −.13 −.01 .19 −.14 −.23* .37** --

Clin Child Psychol Psychiatry. Author manuscript; available in PMC 2020 January 01.
Page 13