Journal of Abnormal Child Psychology

https://doi.org/10.1007/s10802-018-0495-6

Early Childhood Predictors of Anxiety in Early Adolescence

Jennifer L. Hudson 1 & Kou Murayama 2,3 & Lotte Meteyard 2 & Talia Morris 1 & Helen F. Dodd 2

# Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract
This longitudinal study examined a multitude of early childhood predictors of anxiety symptoms and disorders over an 8-year
period. The purpose of the study was to identify early life predictors of anxiety across childhood and early adolescence in a sample
of at-risk children. The sample included 202 preschool children initially identified as behaviorally inhibited or uninhibited between
the ages of 3 years 2 months and 4 years 5 months. Temperament and familial environment variables were assessed using
observation and parent report at baseline. Anxiety symptoms and disorders were assessed using questionnaires and diagnostic
interviews at baseline (age 4), and at age 6, 9 and 12 years. In line with our hypotheses, the findings showed that preschool children
were more likely to experience anxiety symptoms and disorders over time i) when the child was inhibited, ii) when there was a
history of maternal anxiety disorders or iii) when mothers displayed high levels of overinvolvement. Further, the study identified a
significant interaction effect between temperament and maternal overvinvolvement such that behaviorally inhibited preschoolers
had higher anxiety symptoms at age 12, only in the presence of maternal overinvolvement at age 4. The increased risk of anxiety in
inhibited children was mitigated when mothers demonstrated low levels of overinvolvement at age 4. This study provides evidence
of both additive and interactive effects of temperament and family environment on the development of anxiety and provides
important information for the identification of families who will most likely benefit from targeted early intervention.

Keywords Temperament . Anxiety . Internalising . Parenting . Attachment . Behavioral inhibition

Anxiety disorders are highly prevalent in children and adoles- of these disorders remains comparatively limited.
cents, negatively impact multiple domains of functioning and Understanding the factors that place an individual at risk of
have ominous long-term implications for adjustment anxiety, early in life, can provide valuable information regard-
(Polanczyk et al. 2015; Rapee et al. 2009). In addition to their ing the ultimate prevention of anxiety. Of particular interest
high prevalence, anxiety disorders have the earliest age of are risks that occur during the preschool years, the period prior
onset compared to other major mental health disorders and, to the typical age of onset.
if left untreated, persist into adulthood resulting in significant One of the key early life predictors of anxiety disorders is a
personal and societal costs (Erskine et al. 2015; Merikangas behaviorally inhibited temperament. Kagan and colleagues
et al. 2010). Despite well-documented efficacy of treatments defined Behavioral Inhibition (BI) as reactions of withdrawal,
for child anxiety (James et al. 2015), research into the causes wariness, avoidance and shyness in unfamiliar situations and
suggest that roughly 10–15% of infants can be identified as BI
(Garcia Coll et al. 1984). Findings from longitudinal studies of
Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s10802-018-0495-6) contains supplementary BI children suggest that an inhibited child is significantly
material, which is available to authorized users. more likely than an uninhibited child to have an anxiety dis-
order at baseline and also more likely to develop an anxiety
* Jennifer L. Hudson disorder over time (Frenkel et al. 2015; Hirshfeld et al. 1992;
[email protected] Hudson et al. 2011a, b; Hudson and Dodd 2012; Prior et al.
2000; Schwartz et al. 1999). Research consistently demon-
1
Centre for Emotional Health, Department of Psychology, Macquarie strates that BI marks an increased risk for anxiety disorders,
University, Sydney, NSW 2109, Australia in particular, a sevenfold increased risk for social anxiety dis-
2
School of Psychology and Clinical Language Sciences, University of order (Clauss and Blackford 2012).
Reading, Reading, UK Alongside BI, a number of developmental psychopathology
3
Research Institute, Kochi University of Technology, Kochi, Japan models identify key familial environments that increase a
 J Abnorm Child Psychol

child’s risk for disorder (Dodd et al. 2017). Some of the most specifically, Colonnesi and colleagues (Colonnesi et al.
widely studied familial factors include parental anxiety and 2011) concluded that attachment security showed a moderate
parenting related constructs such as over-involvement, negativ- relationship (r = 0.30) with child anxiety, particularly during
ity, and attachment security. In addition to the genetic transmis- adolescence. In contrast, in our previous work, mother-child
sion of anxiety from parent to child, parent anxiety may impact attachment assessed during the strange situation at age 4 was
the development of offspring anxiety via modelling or verbal not predictive of later anxiety symptoms and disorder
transmission of threat and coping. An anxious parent may be (Hudson and Dodd 2012; Hudson et al. 2011a, b). More re-
more likely to model anxious behaviors or provide the child cently however, Lewis-Morrarty and colleagues (Lewis-
with information that increases perceptions of threat and avoid- Morrarty et al. 2015) examined anxiety symptoms in adoles-
ance behavior. In support of this, a number of studies have cents predicted by temperament and attachment at age
shown the impact of a mother’s fearful facial expressions, be- 14 months demonstrating that attachment moderated the rela-
haviors or communications on child avoidance and fear (Dubi tion between temperament and social anxiety in adolescents.
et al. 2008; Percy et al. 2016). Consistent with these findings, That is inhibited children, particularly boys, with an insecure
using data from the sample reported in the current study, ma- attachment had an increased risk for social anxiety.
ternal anxiety disorders measured at age 4 significantly predict- In further support of the role of the familial environment in
ed child anxiety at age 6 and 9, even after controlling for the the development of child anxiety, a recent study using a pow-
child’s anxiety at age 4 (Hudson and Dodd 2012). Taken to- erful ‘children of twins’ design, demonstrated that transmis-
gether, these results suggest that maternal anxiety plays an im- sion of anxiety from parent to child was almost entirely ex-
portant role in predicting child anxiety over time. plained by direct environmental transmission (Eley et al.
Another familial environment variable that may shape the 2015). Research that rigorously assesses the familial environ-
development of anxiety is parental overinvolvement ment using multiple methods of assessment is essential in
(Ollendick and Grills 2016). A child whose parent is order for us to obtain knowledge about environmental trans-
overinvolved or controlling is likely to have reduced opportu- mission of these frequent and impairing disorders.
nities for exposure to novelty or potentially difficult situations, Identification of environmental factors can then lead to the
thereby reducing the child’s opportunities to determine accu- development of enhanced prevention programs. One of the
rate information about threat and coping. In a review of current knowledge gaps is the impact of these early life factors
parenting behaviors associated with anxiety in young on anxiety during adolescence, given this is a period of in-
people, McLeod et al. (2007) reported that controlling, creased onset for anxiety disorders. Only a handful of studies
overinvolved parenting behaviors were more strongly associ- have examined adolescent outcomes for inhibited children
ated with anxiety in children than other aspects of parenting (Chronis-Tuscano et al. 2009; Lewis-Morrarty et al. 2015;
such as negativity or rejection. Parenting that is low in warmth Schwartz et al. 1999) and although some have identified an
has been inconsistently associated with child anxiety and ac- interaction between BI and parenting factors predicting social
counts for a smaller proportion of variance in anxiety symp- anxiety, we have limited knowledge about these variables
toms. This construct is more strongly associated with exter- predicting a range of anxiety disorders.
nalizing behavior (Asbrand et al. 2017; Rubin et al. 2003). In The purpose of the current study was to identify early life
support of these findings, our longitudinal work confirmed predictors of anxiety across childhood and early adolescence,
that, maternal overinvolvement, but not negativity was asso- in a sample of at-risk children. It was hypothesised that family
ciated with increased anxiety disorders and symptoms at age 6 environment variables (maternal overinolvement, maternal
and 9 years for inhibited and uninhibited children (Hudson negativity, attachment insecurity, maternal anxiety) observed
and Dodd 2012; Hudson et al. 2011a, b). Others have also at age 4 would be significantly and positively associated with
shown the longitudinal association between parental overpro- increased youth anxiety disorders and symptoms across the
tection and anxiety over time (Johnson et al. 2016; Muris et al. four assessment points: baseline, 2-year, 5-year, and 8-year
2011; Rubin et al. 2002). follow-up. Although our previous research with this sample
The absence of a secure attachment relationship between identified these as additive predictors, that is, these environ-
infant and caregiver has been described as a non-specific risk ments increased risk for all children, we also explored the
factor for both internalising and externalising disorders in chil- theorised notion (supported in other empirical research, e.g.,
dren (Ainsworth et al. 2015). When a child is unable to reli- Lewis-Morrarty et al. 2012; Lewis-Morrarty et al. 2015) that
ably elicit caregiver attention when needed, the child then the familial environment variables would moderate the rela-
develops an internal working model in which they may view tionship between BI and anxiety across the four time points.
themselves as incapable, the world as unsafe and others as Specifically, we hypothesised that BI children with higher
untrustworthy, placing them at risk for anxiety disorders maternal overinvolvement, higher maternal negativity, attach-
(Manassis 2001). In a meta-analysis examining the relation- ment insecurity and higher maternal anxiety would experience
ship between attachment security and child anxiety more anxiety in early adolescence. By identifying these early
J Abnorm Child Psychol

life predictors, prevention programs can be further tailored to Table 1 Demographic data at 8-year follow-up across groups
reduce a child’s risk for the development of anxiety disorders Demographic BI BUI
across childhood and early adolescence. N = 61 N = 86

Mean Age in years (SD) 11.8 (0.48) 11.8 (0.36)

Method Mother’s age in years (SD) 44.98 (4.61) 44.39 (4.56)
Gender (% female) 50.8 48.8
This study is an 8-year follow-up of a sample of behaviorally Family Structure (baseline)
inhibited (BI) and behaviorally uninhibited (BUI) preschool % Two-parent 91.8 88.4
children and their parents. A detailed description of the sample, Family Income
measures and assessments conducted at baseline, 2- and 5-year % > $80,000 75.5 71.4
assessments can be found in our earlier papers (Hudson et al. %$40–80,000 18.9 17.1
2011a, b; Hudson and Dodd 2012; Hudson et al. 2011a, b). %$0–40,000 5.7 11.4
Ethnicity
Participants %Australian/Caucasian 57.4 78.8
%Asian 21.3 4.7
At baseline, 102 BI and 100 BUI children aged between %European 11.5 11.8
3 years 2 months and 4 years 5 months (M = 4 years, SD = %American 3.3 1.2
4 months) participated in assessments. Of these, 72.8% par- %African 3.3 1.2
ticipated in the 8-year follow-up (61 BI and 86 BUI). The %Middle Eastern 3.3 2.4
mean time between baseline assessment and 8-year follow-
up was 7 years 10 months (SD = 4.9 months). Participants
were initially recruited through local preschools and via an
advertisement in a free parenting magazine. Initial BI classifi- number of approaches to the peer. A participant was defined
cation was made at baseline on the basis of mothers’ report as BI based on observation if they scored above a pre-
using the Short Temperament Scale for Children (STSC), de- determined cut-off on three or more of these five behaviors
scribed below. Children scoring more than one standard devi- (Rapee et al. 2005). Inter-rater reliability for observed BI was
ation above or less than one standard deviation below the determined by having a second trained coder independently
normative mean on the Approach Scale were classified as BI score the videotapes for 25% of the baseline sample. The
or BUI respectively. For families participating in the 8-year inter-rater reliability for number of cutoffs exceeded was
follow-up (n = 147), there were no significant differences be- ICC = 0.91, and for overall BI classification was kappa = 0.79.
tween BI groups on age, sex, family income, maternal age,
family structure (See Table 1). The BI group was significantly Maternal Anxiety Disorders At baseline, mothers were
more likely to classify themselves as being of Asian ethnicity interviewed with the Anxiety Disorders Interview Schedule
than the BUI group (compared to other ethnic categories), χ2 for DSM-IV (DiNardo et al. 1994) to assess current and past
(5, N = 127) = 12.39, p = 0.03. diagnoses. These were combined to give a measure of the
number of anxiety disorders mothers had met criteria for in
Measures their lifetime, to capture anxiety severity as well as clinical
status. Diagnoses were assigned by trained clinicians unaware
Behavioral Inhibition (BI) BI was assessed at baseline using the of the child’s group and anxiety status. A total of 20 cases
approach scale of STSC, a parent-report measure containing (10%) were coded by a second clinician from videotape.
30 items. The STSC has adequate validity, good internal con- Interrater agreement for the number of lifetime anxiety diag-
sistency and reliability (Sanson et al. 1994). The internal con- noses was high (ICC = 0.91).
sistency for the approach scale in the present sample at base-
line was α = 0.92. Maternal Overinvolvement and Negativity Maternal
BI was also assessed at baseline using observed laboratory overinvolvement and negativity were assessed at baseline
tasks similar to those used by Kagan and colleagues (Garcia using a speech preparation task and the Five-Minute Speech
Coll et al. 1984). Children’s responses to a new room, novel Sample (FMSS). Additionally, overinvolvement was assessed
toy, masked experimenter dressed in a strange suit and a same- using the Parent Protection Scale (PPS). Each of these mea-
sex unfamiliar peer were observed. Behaviors used to deter- sures is described briefly below. Further details are provided
mine inhibition status included: i) time spent proximal to the in our earlier paper (Hudson et al. 2011a, b). After converting
mother; ii) amount of time starting at the peer; iii) time spent the data from these measures to z-scores, means were calcu-
talking; iv) number of approaches to the stranger; and v) lated to construct a single overinvolvement variable and a
 J Abnorm Child Psychol

single negativity variable. Overinvolvement during the speech independently coded 42 (21%) cases and reliability for classi-
preparation task was significantly correlated with Emotional fication was kappa = 0.72.
Over-Involvement on the FMSS (EOI; r = 0.178, p < 0.05)
and the PPS (r = 0.151, p < 0.05). EOI was also significantly Child Anxiety Disorders Child anxiety diagnoses were
correlated with the PPS (r = 0.174, p < 0.05). The negativity assessed at baseline, 2-year, 5-year and 8-year follow-up using
scale was made up of criticism on the FMSS and warmth the Anxiety Disorders Interview Schedule for DSM-IV,
during the speech preparation task. These measures were not parent/child version (ADIS-P-IV) (Silverman and Albano
significantly correlated (r = −0.054, p > 0.05). 1996). At baseline and 2-year follow-up, only the parent was
interviewed. At 5-year and 8 –year follow-up both the parent
Parent Protection Scale The PPS was used to assess maternal and child were interviewed and composite diagnoses were
behaviors related to overprotection and autonomy granting at assigned taking both responses into consideration, that is, a
baseline (Thomasgard et al. 1995). The PPS contains 25 items diagnosis was given if either parent OR child provided infor-
(on a scale 0–3) and four subscales: Supervision, Separation, mation that the symptoms were present at a clinical level.
Dependence and Control. The Control scale was of interest to Diagnoses and Clinical Severity Ratings (CSRs on a scale of
the current study and includes items such as ‘I determine who 0–8) were assigned by trained clinicians unaware of the
my child will play with’ and ‘I dress my child even if he/she child’s group membership. Diagnoses were only considered
can do it alone’. The PPS has adequate internal reliability, re- ‘clinical’ if the CSR was 4 or greater. To assess reliability, 20%
test reliability, criterion and content validity (Thomasgard and of the interviews were coded by a second clinician. Interrater
Metz 1999; Thomasgard et al. 1995). The internal consistency agreement was as follows: presence of clinical anxiety diag-
was α = 0.65. nosis (baseline kappa = 0.86, 2-year follow-up kappa = 0.80,
5-year follow-up kappa = 0.85, 8-year = 1.0). In the current
Speech Preparation Task At baseline, mothers were observed study, anxiety disorders were defined as the presence or ab-
interacting with their child during a three-minute speech prep- sence of a diagnosis.
aration task adapted from Hudson and Rapee (2001). The
tasks were videotaped and maternal involvement and maternal Child Anxiety Symptoms Parent-reported symptoms of child
negativity were coded by two postgraduate students in psy- anxiety were measured across the four assessment points using
chology, trained in the coding system. Both coders were un- either the Preschool Anxiety Scale (PAS: baseline and 2- year
aware of participants’ diagnostic status and rated each inter- follow-up)(Spence et al. 2001) or the Spence Children’s
action. The reliability for the average of these ratings at base- Anxiety Scale (SCAS: 5-year and 8-year follow-up)(Spence
line was ICC = 0.94 for the overinvolvement factor and ICC = 1998) depending on the time point. The SCAS comprises 38
0.73 for the negativity factor. items and the PAS 28 items, each on a four-point scale (never,
sometimes, often, always) assessing overall anxiety as well as
Five Minute Speech Sample The FMSS was conducted and specific aspects of anxiety (e.g., social anxiety, generalized
coded according to the method described by Magana and col- anxiety, separation anxiety, obsessive-compulsive, physical in-
leagues (Magana et al. 1986). At baseline, parents were asked jury fears, panic/agoraphobia). To allow comparisons over
to talk about their child and their relationship uninterrupted for time, the raw scores were converted to z-scores according to
5 min. The speech samples were videotaped, transcribed and available normative data. For the PAS, the z-scores were based
coded for criticism and over-involvement as outlined in the on published normative data for 3 (baseline) and 5 year olds
coding manual (Magana et al. 1986). Coders were unaware of (2 year follow-up) (Spence et al. 2001). For the SCAS-P, the z-
participants’ diagnostic status or group membership. A subset scores were based on age and gender means published online
of 48 (24%) transcripts were assessed for inter-rater reliability (https://www.scaswebsite.com).
at baseline: Overinvolvement (kappa = 0.63), Criticism (kap-
pa = 0.96). Procedures

Attachment At baseline, child-mother attachment was Macquarie University Human Ethics Committee approved the
assessed using the preschool version of the Strange Situation study. Following the initial screen using the STSC, children
procedure (Cassidy and Marvin 1992). Children were classi- meeting entry criteria were invited to take part in the full study
fied as having secure, insecure-avoidant, insecure-ambivalent, and mothers provided written informed consent. At baseline
disorganised-controlling or insecure-other attachment follow- and follow-up assessments, participants visited the university
ing coding of videotaped interactions by one of two certified for approximately 2-h sessions. In the follow-up assessments,
coders trained in the Cassidy-Marvin (Macarthur) Preschool child anxiety diagnoses were assessed and various other tasks,
Attachment Classification System. Insecure-other was com- not reported here, were completed. Questionnaire measures
bined with disorganised for the analyses. Both coders were typically completed prior to the assessment either in hard
J Abnorm Child Psychol

copy or online. After each assessment, participants were For the anxiety diagnosis, we also attempted to apply a
rewarded with $50 and a small gift for the child. series of similar latent growth curve models using the four
time-points to mirror the analyses for anxiety symptoms.
However, because this variable is dichotomous and requires
Analysis Plan
non-normal linking functions, the models consistently pro-
duced convergence errors. We therefore decided to run logistic
There were two dependent variables for anxiety: anxiety
regressions with anxiety diagnosis at 8-year follow-up as the
symptoms and presence of an anxiety diagnosis. Analyses
dependent variable to be in keeping with the analyses reported
for these DVs are conducted and reported separately. In addi-
in Hudson and Dodd (2012) for the 5-year follow-up of this
tion, there were a number of predictor variables: BI group and
4 family environment variables including number of lifetime sample.
Logistic regression analyses were conducted for anxiety
(current and past) maternal anxiety diagnosis; maternal over-
diagnosis as a dependent variable using the base package in
involvement; maternal negativity; and mother-child attach-
ment security (2 levels: secure vs insecure). R. Each of the risk factors (BI group and the five family
environment factors) was included in a separate analysis to
For the anxiety symptoms, we first conducted latent growth
ensure sufficient power and avoid multi-collinearity. First,
curve modelling (LGM) to examine predictors of change in
scores over the four time-points using mixed-effects model- analyses were conducted without the corresponding baseline
measure of anxiety and subsequently ran again with baseline
ling. This was done using R (R Core Team 2013) with lmer()
anxiety controlled for. We also included Asian ethnicity as a
function from the lme4 package (Bates et al. 2014) in combi-
nation with lmerTest package (Kuznetsova et al. 2014). We controlling variable in all analyses. To examine the interplay
between BI and other risk factors, four additional logistic re-
first ran an initial LGM model that included the linear and
gressions were conducted. These included a main effect of BI
quadratic effects of time as fixed predictors of anxiety symp-
tom z-scores and their corresponding random-participants ef- group and its interaction with family environment variables.
Therefore, these models assess whether each family environ-
fects (random intercepts for participants and random slopes
ment variable predicts anxiety at follow-up after controlling
for linear and quadratic effects of time varying by partici-
for BI group, as well as any interactions between BI group and
pant).1 Time was anchored such that zero represented the ini-
family environment variables.2
tial assessment point (Biesanz et al. 2004). The model includ-
All reported analyses were conducted with BI group
ed BI group (0 = BUI; 1 = BI) and its interaction with the
according to parent report as participants were initially
linear and quadratic effects of time to model different growth
recruited on this basis. To examine whether the pattern
curves between groups. Given differences in ethnicity be-
of results was consistent with the reduced sample of par-
tween BI groups at baseline and follow-up, we also included
ticipants whose parent report BI grouping matched their
Asian ethnicity (0 = No; 1 = Yes) as a control variable.
observed BI group allocation, all analyses were conduct-
Next, we ran four LGM models that separately assessed
ed again. There were some minor differences in statisti-
family environment variables: the number of lifetime maternal
cal significance between the original analyses and these
anxiety diagnoses at baseline; maternal over-involvement;
analyses, likely to due to reduced power, but the overall
maternal negativity; and mother-child attachment security.
pattern of results was the same. Results from the analy-
These models examined whether (and how) these variables
ses conducted with the reduced sample are provided, and
can explain the individual differences in the growth curves.
discrepancies in statistical significance are highlighted, in
Each variable was added to the initial model above, starting
Supplementary Tables S1 and S2.
with main effects and then adding higher order interactions
with BI group, linear and quadratic effects of time. To ensure
sufficient power and avoid multi-collinearity, we ran this mod-
el separately for each family environment variable. This
means that, for each model, BI group, a family environmental
Results
variable (e.g., the number of lifetime maternal anxiety diag-
Descriptive statistics for all three anxiety dependent variables
noses), and their interactions were included as fixed effect
at each time point are shown in Table 2. Table 3 provides the
predictors.
descriptive details maternal anxiety disorders, maternal over-
involvement, maternal negativity, and mother-child attach-
ment security at baseline.
1
We also explored the models that omitted the random and/or quadratic effect
2
but the fit of these models was significantly worse than the reported model To examine the effect of sex, we included sex in the above models, and also
(log-likelihood ratio test, ps < .05), indicating the importance of incorporating the interaction between BI and sex. Neither Sex nor the interaction between BI
these effects in the LGM to account for growth curves. and sex were significant and did not change the outcome of the models.
 J Abnorm Child Psychol

Table 2 Descriptive statistics for child anxiety variables across time

Age 4 Age 6 Age 9 Age 12

BI BUI BI BUI BI BUI BI BUI

Parent reported child anxiety 1.08 (1.24) −0.83 (0.73) 1.32 (1.36) −0.056 (0.84) 0.48 (1.22) −0.42 (0.71) −0.11 (1.19) −0.59 (0.76)
(PAS/SCAS) z-scores M (SD)
Presence of anxiety diagnosis- 74 17 (17%) 52 (60%) 12 (12%) 38 (54%) 16 (18%) 22 (36%) 16 (19%)
total number (% of group) (73%)

BI, behaviorally inhibited; BUI, behaviorally uninhibited

Missing Data to deal with missing data and pooled results from five
imputations are reported. MI was conducted following
The following percentages indicate the amount of missing Enders (2010) using the mice package in R (van Buuren and
data for anxiety diagnoses across the 4 time points for BI Groothuis-Oudshoorn 2011). This method also rests on the
children: 0% (baseline); 15% (2-year); 30% (5-year); 40% MAR assumption as outlined above.
(8-year); and, BUI children: 0% (baseline); 9% (2-year);
11% (5-year); 14% (8-year). The following percentages indi- Anxiety across Time
cate the amount of missing data for anxiety symptoms across
the 4 time points for BI children: 3% (baseline); 16% (2-year); Figure 1a and b show the correlations between anxiety across
34% (5-year); 43% (8-year); and, BUI children: 1% the four time-points. As expected the strength of the associa-
(baseline); 6% (2-year); 9% (5-year); 25% (8-year). Data were tions decreases as the time between assessments increases.
missing either because the family could not be contacted or
because they chose not to participate at follow-up. Predicting Anxiety Symptoms
To deal with missing data, we took two strategies. For the
latent growth curve modelling, we used a maximum likeli- The results of the growth curve analysis are summarized in
hood method to estimate parameters. This method requires Table 4. Figure 2 presents the predicted growth curve of anx-
that data are missing at random. For data to be considered iety scores for the BI and BUI groups. As expected, BI group
missing at random missingness can be conditional on other positively predicted the intercept (beta = 0.94, p < 0.01), indi-
variables included in the analyses but cannot depend on the cating that the BI group had higher anxiety scores at age 4 (as
values of variables that are missing. Thus, the fact that missing reported in Hudson et al. 2011a, b). In addition, there was a
data can be predicted by BI group, which is also measured and significant interaction between the linear effect of time and BI
included in the analyses means that the data can be considered group (beta = −0.13, p < 0.01), this suggests that the initial
missing at random (MAR). As Marsh et al. (in press) argued, linear trend of anxiety scores differ across the groups; whereas
in a longitudinal panel design, this assumption is unlikely to the BUI group initially exhibited an increase in anxiety, the BI
be seriously violated because, even if missingness is depen- group showed an initial decrease in anxiety.
dent on the variable itself, this dependency is likely to be
accounted for by the same variable assessed at a different time Maternal Anxiety Table 4 reports the results for the model
point. For the logistic regression, we used multiple imputation examining maternal anxiety. The number of current and past

Table 3 Family environment

variables by group for full sample BI BUI
at baseline
Maternal lifetime anxiety diagnosis at baseline n = 74 (73%) n = 50 (50%)
Maternal overinvolvement (z score) 0.14 (0.70) −0.14 (0.59)
Maternal negativity (z score) 0.13 (0.70) −0.16 (0.67)
Attachment security
Secure n = 45 n = 59
45% 60%
Insecure n = 54 n = 39
55% 40%

Mean and (standard deviation) shown; for categorical variables the number of individuals in each category and
(percentage) is provided. BI = behaviorally inhibited; BUI = behaviorally uninhibited
J Abnorm Child Psychol

a (beta = −0.02, p < 0.05) slopes for time. Maternal over-

involvement also had a main effect on the linear slopes for
time (beta = 0.11, p < 0.05). To interpret these multiple inter-
action effects, we plotted the predicted growth curve of chil-
dren with high- and low- maternal over-involvement for both
BI and BUI groups. Figure 4 suggests that maternal
overinvolvement does not impact anxiety trajectory for BUI
children but does affect anxiety in BI children over time.
Although maternal overinvolvement was not associated with
baseline child anxiety within the BI group, when mothers had
low levels of overinvolvement, anxiety in the BI children rap-
idly declined such that by age 9 it resembled that of BUI
children. In contrast, when mothers had high levels of
overinvolvement, high anxiety levels in BI children declined
less rapidly.
For maternal negativity, there was a significant positive
association with initial anxiety scores (beta = 0.29, p < 0.01)
b as reported elsewhere (Hudson et al. 2011a, b); children
whose mothers were more negative at baseline had high levels
of anxiety at baseline. In addition, parental negativity signifi-
cantly predicted the linear slopes (beta = −0.12, p < 0.05). This
indicates that children whose mothers were more negative at
baseline had higher baseline anxiety (previously reported) and
then for both BI and BUI groups, the slope value is more
negative if negativity is high relative to when it is low. That
is, children whose mothers were more negative at baseline had
a steeper decrease in anxiety over time than those whose
mothers were less negative. This must be interpreted in the
context of the children with more negative mothers having
higher baseline anxiety scores. Interactions with BI were not
significant.

Attachment Security The results for attachment security are

Fig. 1 Correlation matrices showing continuity in anxiety over time. a
shows anxiety symptom z-scores, b shows kappa for presence of an
shown in Table 4. As can be seen, attachment security did not
anxiety diagnosis. Cells in green are significant p < 0.001, cells in white show any significant effects in the LGM analysis.
are not significant p > 0.05. Colour bar indicates strength of the
coefficient Predicting Presence of an Anxiety Diagnosis

maternal anxiety diagnoses at baseline was a significant pre- Table 2 provides the descriptive data for anxiety disorders
dictor of the initial level of anxiety symptom scores (beta = across the four time points. Frequency of specific anxiety
0.23, p < 0.01), with a greater number of maternal anxiety disorders at 4, 6 and 9 have been reported elsewhere and hence
diagnoses associated with higher child anxiety at baseline we provide specific anxiety disorders at age 12. BI children
(as reported elsewhere, Hudson et al. 2011a, b). The three- were more likely than BUI children to have social anxiety
way interaction with linear time and BI group was significant disorder (BI: 26.2%; BUI: 5.8%, χ2 (1, N = 147) = 12.15, p
(beta = 0.05, p < 0.05) but the effect size was small. Figure 3 <. 001) and OCD (BI: 4.9% BUI: 0% χ2 (1, N = 147) = 4.32,
shows this interaction. Scrutiny of the plot suggests that the p = 0.04. There was a non-significant trend for BI children to
interaction is subtle and is overshadowed by large main effects be more likely to have SAD (BI: 3.3; BUI: 0%, χ2 (1, N =
of BI and maternal anxiety. 147) = 2.86, p < 0.1). There were no significant differences
between BI and BUI children in the prevalence of GAD (BI:
Maternal Overinvolvement and Negativity The results for ma- 13.1%; BUI: 5.8%), Specific Phobia (BI: 11.5%; BUI: 9.3%)
ternal parenting are also shown in Table 4. Maternal over- or Panic Disorder (BI: 1.6%; BUI: 0%; ps > 0.05).
involvement at baseline significantly interacted with BI group Table 5 summarizes the results of the logistic regression
to explain the linear (beta = 0.14, p < 0.01) and quadratic predicting the presence of an anxiety diagnoses at age 12.
 J Abnorm Child Psychol

Table 4 Summary of the growth curve analysis on anxiety symptom scores

No baseline Maternal lifetime Maternal Maternal Attachment

predictors anxiety diagnosis over- negativity security
involvement

Intercepts 0.20** −0.12 0.21** 0.20** 0.28*

BI group 0.94** 0.84** 0.93** 0.91** 0.96**
Family environment 0.23** 0.09 0.29** −0.16
Family Environment x BI group 0.02 0.06 0.11 −0.06
Ethnicity 0.00 0.19 −0.14 0.01 −0.00
Linear effects for time 0.17** 0.13** 0.15** 0.16** 0.13**
x BI group −0.13** −0.19** −0.15** −0.11** −0.11*
x Family environment 0.02 0.11* −0.12* 0.05
x Family environment x BI group 0.05* 0.14** −0.00 −0.05
Quadratic effects for time −0.03** −0.03** −0.03** −0.03** −0.03**
x BI group 0.01 0.01* 0.01 0.00 0.00
x Family environment −0.00 −0.01 0.01 −0.01
x Family environment x BI group −0.00 −0.02** −0.00 0.00

Intercepts 0.62 0.59 0.73 0.66 0.70

Linear slope for time 0.02 0.02 0.02 0.02 0.02
Quadratic slope for time 0.00 0.00 0.00 0.00 0.00

Each column represents a single model, with a different family environment factor included as a predictor. The baseline growth curve model (the leftmost
column) includes the intercepts, linear slopes for time, and quadratic slopes for time with their respective random effects. Asian ethnicity was also added
as a control variable. The rest of the models included the BI group, one family environment factor, and their interactions with the linear and quadratic
effects for time. * < 0.05, ** < 0.01

Number of maternal lifetime anxiety disorders, and maternal eliminating the interaction effects but keeping BI in the
over-involvement predicted the presence of an anxiety diag- models. The obtained results are reported in Table 5. BI, ma-
nosis at follow-up, even after controlling for baseline anxiety. ternal anxiety and overinvolvement all significantly predicted
For analyses that included BI together with each family envi- the presence of anxiety diagnosis at age 12. After controlling
ronment factor and their interaction, none of the interaction for anxiety and BI at baseline, only number of maternal life-
terms were significant (ps > 0.05). Given the absence of sig- time anxiety disorders remained significant. Maternal negativ-
nificant interaction effects, we re-ran the models after ity and attachment did not predict anxiety at follow-up, re-
gardless of whether baseline anxiety was controlled for.

1.0
Discussion

This eight-year longitudinal study examined early life predic-

0.5
tors of anxiety disorders in early adolescence. The study uti-
lized observational techniques and questionnaires to assess
Anxiety

Group
BI
BUI
familial and temperamental variables, as well as both struc-
0.0 tured diagnostic interviews and questionnaires to examine the
presence of anxiety. This study extends the previous literature
by uniquely assessing multiple family environment factors to
predict early adolescent anxiety including maternal anxiety,
-0.5
maternal negativity, maternal overinvolvement and mother-
child attachment. Consistent with earlier studies, preschool
children were more likely to experience anxiety symptoms
4 6 9 12
Age and disorders in early adolescence when the child was
Fig. 2 Predicted growth curve for anxiety symptoms of the behaviorally inhibited, when there was a history of maternal anxiety disor-
inhibited (BI) and uninhibited (BUI) children ders and when mothers displayed high levels of
J Abnorm Child Psychol

1.0
that this relationship may be present for total anxiety symp-
toms and holds even after controlling for baseline anxiety
0.5 symptoms. This finding is also consistent with theoretical
models that highlight the importance of restrictive and over-
Group protective parenting in reducing the child’s opportunities for
0.0
Anxiety

BI & High Parental Anxiety

BI & Low Parental Anxiety
BUI & High Parental Anxiety
exposure to novelty or potentially difficult situations (Hudson
BUI & Low Parental Anxiety
and Rapee 2004; Ollendick and Grills 2016). Providing too
-0.5 much involvement in an inhibited child’s activities therefore
further reduces opportunities for the child to discover accurate
-1.0
data about threat and coping, increasing the child’s vulnera-
bility to anxiety.
4 6 9 12 Consistent with our earlier findings, a history of maternal
Age
anxiety was identified as a significant additive predictor of
Fig. 3 Predicted growth curve for anxiety symptoms across time for the
child anxiety symptoms and disorders, such that children
BI and BUI children with high vs low maternal current and past anxiety
whose mothers had an anxiety disorder were more likely to
experience higher anxiety symptoms and disorders over time.
overinvolvement. These findings held even after accounting
This effect of maternal anxiety was robust across all analyses.
for baseline symptoms. Further, this study identified a moder-
The results indicate that this is primarily a main effect of
ation effect between temperament and maternal behavior. That
maternal anxiety such that BUI and BI children who have an
is, preschool children with an inhibited temperament were at
anxious mother are at similarly increased risk for anxiety over
risk for increased anxiety symptoms in early adolescence
time. There was a significant three-way interaction between
when their mothers displayed high over-involvement.
maternal lifetime diagnoses, BI and the linear effect of time
Inhibited children whose mothers were not overinvolved dur-
but the effect size was small and there is little evidence from
ing the preschool years showed similar anxiety levels in ado-
the plot (Fig. 3) that this interaction is meaningful. It is possi-
lescence to uninhibited peers. Although this interaction effect
ble that BI children whose mothers were anxious had a slight-
was not observed when anxiety disorder status was measured,
ly shallower decline in their slope across time than BI children
maternal overinvolvement remained a significant predictor of
whose mothers were not anxious but this is a very subtle effect
child anxiety disorder status.
and should be interpreted with caution. The results provide
These findings are consistent with the large body of litera-
strong support for the important role maternal anxiety plays
ture that consistently identifies an association between paren-
in affecting child anxiety over time.
tal overinvolvement and anxiety symptoms (McLeod et al.
Parenting high in negativity was associated with increased
2007; van der Bruggen et al. 2008). The current study extends
anxiety symptoms at baseline (as reported earlier; Hudson
the existing literature as it highlights the moderating role par-
et al. 2011a, b), followed by a decline in symptoms over time
enting plays in shaping the development of anxiety symptoms
such that by age 12, early maternal negativity had little effect
in inhibited children. Previous longitudinal studies have iden-
on anxiety symptoms. Maternal negativity did not interact
tified a moderating relationship between inhibition and mater-
with BI indicating that the effects of negativity are consistent
nal control specifically for social anxiety symptoms (Lewis-
for both inhibited and uninhibited children. Maternal negativ-
Morrarty et al. 2012). Importantly, the current study shows
ity also did not predict disorder status in early adolescence.
Taken together there is little evidence that maternal negativity
1.0 at age 4 significantly affects anxiety risk through to early
adolescence. With this in mind, the positive association be-
tween high negativity and anxiety symptoms found at baseline
0.5
may have been due to mothers responding negatively to their
Group
child’s anxiety rather than maternal negativity playing a causal
Anxiety

BI & High Parental Involvement

BI & Low Parental Involvement
0.0 BUI & High Parental Involvement
BUI & Low Parental Involvement
role in their child’s anxiety. These findings are in keeping with
previous work showing that parental warmth and rejection are
weakly and inconsistently associated with anxiety disorders
-0.5
(McLeod et al. 2007). Perhaps the inconsistent finding in the
current study is a result of the sample size not being sufficient-
-1.0 ly large to detect such small effects.
4 6 9 12
Age Mother-child attachment security did not significantly pre-
Fig. 4 Parental over-involvement and anxiety symptom scores over time dict anxiety symptoms or disorders over time. Although this is
and BI group consistent with results from earlier assessments utilizing this
 J Abnorm Child Psychol

Table 5 Results of logistic regressions to assess the effect of each risk factor on the presence of an anxiety diagnosis at 8-year follow-up (results based
on Multiple Imputation to handle missing data). Parent report BI groups

Before controlling After controlling After controlling for

for baseline anxiety for baseline anxiety baseline anxiety and
BI group

Behavioral Inhibition 0.96* 1.00*

Number of maternal lifetime anxiety disorders 0.52* 0.49* 0.51*
Over-involvement 0.56* 0.49* 0.37
Negativity 0.12 0.04 −0.04
Attachment security −0.14 −0.07 0.02

* p < 0.05; ** p < 0.01. All other values p > 0.1

sample (Hudson and Dodd 2012), the finding conflicts with a at greater risk of developing OCD by early adolescence and a
recent study showing inhibited children with an insecure non-significant trend was observed for separation anxiety dis-
mother-child attachment were at increased risk of adolescent order. In our previous work with this sample we have shown
social anxiety symptoms (Lewis-Morrarty et al. 2015). that BI predicts not only social anxiety disorder but also sep-
Specifically, Lewis-Morrarty and colleagues found that infant aration anxiety disorder and generalised anxiety disorder
inhibited males with an insecure-resistant attachment were at (Hudson and Dodd 2012). The current findings indicate that
increased risk of adolescent social anxiety. In the current BI is associated with a clear increased risk for social anxiety
study, we only identified 12 children in the insecure ambiva- disorder (Clauss and Blackford 2012), but that this tempera-
lent category and thus we had to collapse across types of mental construct also places the child at risk for other anxiety
insecurity. It therefore remains possible that children who related disorders. In this sample, early adolescence was a pe-
have an ambivalent attachment may be at increased risk for riod of marked remission of anxiety disorders for inhibited
anxiety disorders. We also assessed attachment at age 4 children. Although high rates of anxiety disorders were ob-
whereas Lewis-Morrarty and colleagues assessed attachment served in the inhibited sample during childhood (54–73%),
much earlier (14 months), which may explain some of the only 36% of inhibited pre-schoolers were highly anxious in
inconsistency in findings. early adolescence. This finding highlights that, although inhi-
Perhaps most surprising was the limited value of early anx- bition is a risk factor for anxiety disorders, many inhibited
iety as a predictor of anxiety disorders in early adolescence. children, in fact the majority, do not experience long-term
Previous research has demonstrated that high anxiety is rela- problems with mental health (Degnan and Fox 2007).
tively stable over time, yet stability rates of disorders have been A clear strength of this study is the use of multiple assess-
shown to vary dramatically between studies (Weems 2008). In ment methods to assess the child’s early family environment
previous longitudinal assessments of this sample, anxiety dis- as well as assessment of anxiety symptoms and disorders over
orders in early childhood predicted anxiety disorders during an 8-year period. A significant limitation of the study is the
middle childhood. Our current analysis showed continuity in sole focus on maternal behaviours and the absence of critical
anxiety symptoms and disorder over the short term but anxiety data on the role of fathers. Recent data suggest that fathers also
disorders at age 4 were not significantly related to anxiety play a key role in the development of anxiety symptoms
disorders at age 12. One possible explanation for the absence (Lazarus et al. 2016) and thus the exclusion of fathers signif-
of a significant effect from preschool to pre-adolescence here is icantly limits the conclusions that can be made about parent-
the greater attrition in the inhibited sample at this follow-up, ing in general. Another significant methodological and con-
with only 60% of the sample participating compared to 86% of ceptual limitation is the extreme groups design, that ignores
the uninhibited sample. This significantly reduced the number the ‘typical’ presentation of children. As is the case with much
of children with anxiety disorders in the final sample. The of the work examining the category of behavioural inhibition,
finding should therefore be interpreted with some caution. the child who is neither eager to explore nor avoidant of ex-
As expected, BI at age 4 was a significant predictor of ploration is not captured within this study design and thus we
anxiety across time and this was relatively robust across out- are unable to generalise these findings to typical children in-
come measures. The comparisons by individual anxiety diag- stead of children at either end of the inhibition spectrum.
nosis indicate that whilst BI children are significantly more Further, although the sample included a range of ethnically
likely to meet criteria for an anxiety diagnosis at age 12, this is diverse families, indicative of the Sydney population, the chil-
primarily driven by an increase in social anxiety disorder di- dren are primarily from intact families with a moderate socio-
agnoses. Behaviorally inhibited preschool children were also economic advantage, thus limiting the generalisations of these
J Abnorm Child Psychol

findings to the broader population. The sample size across the References
four follow-up assessments also limited the study’s capacity to
examine all of the familial variables together. Although the Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. N. (2015).
Patterns of attachment: A psychological study of the strange
analyses tell us the predictive power of each family environ-
situation. Psychology Press.
ment variables in predicting anxiety over and above baseline Asbrand, J., Hudson, J. L., Schmitz, J., & Tuschen-Caffier, B. (2017).
symptoms and temperament, they are unable to determine the Maternal Parenting and Child Behaviour: An Observational Study
unique contribution of these variables over and above other of Childhood Social Anxiety Disorder. Cognitive Therapy and
Research, 41(4), 562–575. https://doi.org/10.1007/s10608-016-
family environment variables. Future research would benefit
9828-3.
from examining the unique value of these variables as well as Bates, D., Maechler, M., Bolker, B., & Walker, S. (2014). lme4: Linear
the potential moderation and mediation between family envi- mixed-effects models using Eigen and S4. R package version, 1, 1–6
ronment variables that may exist. http://cran.r-project.org/package=lme4.
Despite these limitations, the results of this study present Biesanz, J. C., Deeb-Sossa, N., Papadakis, A. A., Bollen, K. A., &
Curran, P. J. (2004). The role of coding time in estimating and
clear clinical implications for the prevention of anxiety disor- interpreting growth curve models. Psychological Methods, 9(1),
ders in children and early adolescents. Targeted interventions 30–52. https://doi.org/10.1037/1082-989x.9.1.30.
should be made available for families of children with an Cassidy, J., & Marvin, R. S. (1992). Attachment organization in pre-
inhibited temperament and specifically, these interventions school children: Procedures and coding manual (4th ed.).
University of Virginia: Unpulished manuscript.
should empower mothers to encourage their child’s indepen- Chronis-Tuscano, A., Degnan, K. A., Pine, D. S., Perez-Edgar, K.,
dence and reduce high levels of overinvolvement. Henderson, H. A., Diaz, Y., et al. (2009). Stable early maternal
Interventions that specifically reduce this parenting behaviour report of behavioral inhibition predicts lifetime social anxiety disor-
in inhibited children will be important for reducing anxiety in der in adolescence. Journal of Amer Academy of Child & Adolescent
Psychiatry, 48(9), 928–935. https://doi.org/10.1097/CHI.
the long term. These findings also make a strong case for the 1090b1013e3181ae1009df.
delivery of interventions to mothers who experience signifi- Clauss, J. A., & Blackford, J. U. (2012). Behavioral inhibition and risk for
cant anxiety. Adult anxiety disorders are treatable and prevent- developing social anxiety disorder: a meta-analytic study. Journal of
able (Craske et al. 2017; Hudson 2017). Thus, the familial the American Academy of Child & Adolescent Psychiatry, 51(10),
1066–1075.e1061. https://doi.org/10.1016/j.jaac.2012.08.002.
transmission of anxiety from parent to child can be partially
Colonnesi, C., Draijer, E. M., Jan, J. M., Stams, G., Van der Bruggen, C.
prevented by providing evidence-based treatments to parents O., Bögels, S. M., & Noom, M. J. (2011). The relation between
who experience anxiety disorders (Ginsburg et al. 2015). Such insecure attachment and child anxiety: a meta-analytic review.
treatment can reduce the likelihood parents will provide their Journal of Clinical Child & Adolescent Psychology, 40(4), 630–
children with biased negative information about the world, or 645. https://doi.org/10.1080/15374416.2011.581623.
Core Team, R. (2013). R: A language and environment for statistical
specific aspects of the world, and reduce the likelihood that computing. Vienna, Austria: R Foundation for Statistical
they will model anxious and avoidant behaviors to their child. Computing http://www.r-project.org.
Targeting these key variables will likely reduce the prevalence Craske, M. G., Stein, M. B., Eley, T. C., Milad, M. R., Holmes, A., Rapee,
of anxiety disorders in children and adolescents. R. M., & Wittchen, H.-U. (2017). Anxiety disorders. Nature
Reviews Disease Primers, 3, 17,024. https://doi.org/10.1038/nrdp.
2017.24.
Degnan, K. A., & Fox, N. A. (2007). Behavioral inhibition and anxiety
Compliance with Ethical Standards disorders: Multiple levels of a resilience process. Development and
Psychopathology, 19(3), 729–746.
Funding This project was supported by the Australian Research DiNardo, P. A., Brown, T. A., & Barlow, D. H. (1994). Anxiety Disorders
Council Discovery Grant (DP0342793) and Future Fellowship grant Interview Schedule: Lifetime Version (ADIS-IV-L). Client Interview
(Hudson: FT20100217). The study was also supported in part by JSPS Schedules. Oxford: Oxford University Press.
KAKENHI (#16H06406 to K.M.) and the Leverhulme Trust (Grant # Dodd, H.F., Hudson, J.L., & Rapee, R.M. (2017). Temperament in Youth
RL-2016-030 to K.M.). Internalizing Disorders. In Treatments for Psychological Problems
and Syndromes (pp. 504–524): John Wiley & Sons, Ltd.
Dubi, K., Rapee, R. M., Emerton, J. L., & Schniering, C. A. (2008).
Conflict of Interest The authors report no financial interests or potential
Maternal Modeling and the Acquisition of Fear and Avoidance in
conflicts of interest.
Toddlers: Influence of Stimulus Preparedness and Child
Temperament. Journal of Abnormal Child Psychology, 36(4),
Ethical Approval The Macquarie University Human Ethics Committee 499–512. https://doi.org/10.1007/s10802-007-9195-3.
approved the methods of the study. Eley, T. C., McAdams, T. A., Rijsdijk, F. V., Lichtenstein, P., Narusyte, J.,
Reiss, D., et al. (2015). The intergenerational transmission of anxi-
Informed Consent Mothers provided initial consent to participate in the ety: A children-of-twins study. American Journal of Psychiatry,
study. Mothers and children both provided written, informed consent to 172(7), 630–637.
participate in the 8 year follow-up. Enders, C. K. (2010). Applied missing data analysis: Guilford Press. New
York: NY.
Publisher’s Note Springer Nature remains neutral with regard to jurisdic- Erskine, H. E., Moffitt, T. E., Copeland, W. E., Costello, E. J., Ferrari, A.
tional claims in published maps and institutional affiliations. J., Patton, G., et al. (2015). A heavy burden on young minds: the
global burden of mental and substance use disorders in children and
 J Abnorm Child Psychol

youth. Psychological medicine, 45(07), 1551–1563. https://doi.org/ adolescent social anxiety symptoms. Child Development, 86(2),
10.1017/S0033291714002888. 598–613. https://doi.org/10.1111/cdev.12336.
Frenkel, T. I., Fox, N. A., Pine, D. S., Walker, O. L., Degnan, K. A., & Magana, A. B., Goldstein, M. J., Karno, M., Miklowitz, D. J., Jenkins, J.,
Chronis-Tuscano, A. (2015). Early childhood behavioral inhibition, & Falloon, I. R. H. (1986). A brief method for assessing expressed
adult psychopathology and the buffering effects of adolescent social emotion in relatives of psychiatric patients. Psychiatry Research, 17,
networks: a twenty-year prospective study. Journal of Child 203–212.
Psychology and Psychiatry, 56(10), 1065–1073. https://doi.org/10. Manassis, K. (2001). Child–parent relations: Attachment and anxiety
1111/jcpp.12390. disorders. Anxiety disorders in children and adolescents, 255–272.
Garcia Coll, C., Kagan, J., & Reznick, J. S. (1984). Behavioral inhibition Marsh, H. W., Pekrun, R., Parker, P. D., Murayama, K. Guo, J., Dicke, T.,
in young children. Child Development, 55(3), 1005–1019. https:// & Arens, A. K. (in press). The Murky Distinction Between Self-
doi.org/10.2307/1130152. Concept and Self-Efficacy: Beware of Lurking Jingle-Jangle
Ginsburg, G. S., Drake, K. L., Tein, J. Y., Teetsel, R., & Riddle, M. A. Fallacies. Journal of Educational Psychology.
(2015). Preventing onset of anxiety disorders in offspring of anxious McLeod, B. D., Wood, J. J., & Weisz, J. R. (2007). Examining the asso-
parents: a randomized controlled trial of a family-based intervention. ciation between parenting and childhood anxiety: A meta-analysis.
American Journal of Psychiatry, 172(12), 1207–1214. Clinical Psychology Review, 27(2), 155–172.
Hirshfeld, D. R., Rosenbaum, J. F., Biederman, J., Bolduc, E. A., Merikangas, K. R., He, J.-P., Burstein, M., Swanson, S. A., Avenevoli, S.,
Faraone, S. V., Snidman, N., et al. (1992). Stable behavioral inhibi- Cui, L., et al. (2010). Lifetime prevalence of mental disorders in U.S.
tion and its association with anxiety disorder. Journal of the adolescents: Results from the National Comorbidity Survey
American Academy of Child & Adolescent Psychiatry, 31(1), 103– Replication-Adolescent Supplement (NCS-A). Journal of the
111. American Academy of Child & Adolescent Psychiatry, 49(10),
Hudson, J. L. (2017). Prevention of anxiety disorders across the lifespan. 980–989. https://doi.org/10.1016/j.jaac.2010.05.017.
JAMA Psychiatry. https://doi.org/10.1001/jamapsychiatry.2017. Muris, P. E. H. M., Brakel, A. M. L. V., Arntz, A., & Schouten, E. (2011).
2430. Behavioral inhibition as a risk factor for the development of child-
Hudson, J. L., & Rapee, R. M. (2001). Parent–child interactions and hood anxiety disorders: a longitudinal study. Journal of Child and
anxiety disorders: an observational study. Behaviour Research and Family Studies, 20, 157–170.
Therapy, 39, 1411–1427. Ollendick, T. H., & Grills, A. E. (2016). Perceived control, family envi-
Hudson, J. L., & Dodd, H. F. (2012). Informing Early Intervention: ronment, and the etiology of child anxiety—revisited. Behavior
Preschool Predictors of Anxiety Disorders in Middle Childhood. Therapy, 47(5)), 633–642. https://doi.org/10.1016/j.beth.2016.01.
PLoS ONE, 7(8), e42359. https://doi.org/10.1371/journal.pone. 007.
0042359. Percy, R., Creswell, C., Garner, M., O’Brien, D., & Murray, L. (2016).
Hudson, J. L., & Rapee, R. M. (2004). From Temperament to Disorder: Parents’ verbal communication and childhood anxiety: a systematic
An Etiological Model of Generalized Anxiety Disorder. In R. G. review. Clinical Child and Family Psychology Review, 19(1), 55–
Heimberg, C. C. Turk, & D. S. Menin (Eds.), Generalized Anxiety 75. https://doi.org/10.1007/s10567-015-0198-2.
Disorder: Advances in Research and Practice. New York: Guildford Polanczyk, G. V., Salum, G. A., Sugaya, L. S., Caye, A., & Rohde, L. A.
Press. (2015). Annual research review: a meta-analysis of the worldwide
Hudson, J. L., Dodd, H., & Bovopoulos, N. (2011a). Temperament, prevalence of mental disorders in children and adolescents. Journal
Family Environment and Anxiety in Preschool Children. Journal of Child Psychology and Psychiatry, 56(3), 345–365.
of Abnormal Child Psychology, 39(7), 939–951. https://doi.org/10. Prior, M., Smart, D., Sanson, A., & Oberklaid, F. (2000). Does shy-
1007/s10802-011-9502-x. inhibited temperament in childhood lead to anxiety problems in
Hudson, J. L., Dodd, H. F., Lyneham, H. J., & Bovopoulous, N. (2011b). adolescence? Journal of the American Academy of Child &
Temperament and family environment in the development of anxi- Adolescent Psychiatry, 39(4), 461–468.
ety disorder: two-year follow-up. Journal of the American Academy Rapee, R. M., Kennedy, S., Ingram, M., Edwards, S., & Sweeney, L.
of Child and Adolescent Psychiatry, 50(12), 1255–1264.e1251. (2005). Prevention and early intervention of anxiety disorders in
James, A.C., James, G., Cowdrey, F.A., Soler, A., & Choke, A. (2015). inhibited preschool children. Journal of Consulting and Clinical
Cognitive behavioural therapy for anxiety disorders in children and Psychology, 73, 488–497.
adolescents. The Cochrane Library. Rapee, R. M., Schniering, C. A., & Hudson, J. L. (2009). Anxiety disor-
Johnson, V. C., Olino, T. M., Klein, D. N., Dyson, M. W., Bufferd, S. J., ders during childhood and adolescence: Origins and treatment.
Durbin, C. E., et al. (2016). A longitudinal investigation of predic- Annual Review of Clinical Psychology, 5, 335–365.
tors of the association between age 3 and age 6 behavioural inhibi- Rubin, K. H., Burgess, K. B., & Hastings, P. D. (2002). Stability and
tion. Journal of Research in Personality, 63, 51–61. https://doi.org/ social-behavioral consequences of toddlers’ inhibited temperament
10.1016/j.jrp.2016.04.008. and parenting behaviors. Child Development, 73(2), 483–495.
Kuznetsova, A., Brockhoff, P.B. & Bojesen, R.H. (2014) lmerTest: Tests Rubin, K. H., Burgess, K. B., Dwyer, K. M., & Hastings, P. D. (2003).
for random and fixed effects for linear mixed effect models (lmer Predicting preschoolers’ externalizing behaviors from toddler tem-
objects of lme4 package). R package version 2.0–6. http://CRAN.R- perament, conflict, and maternal negativity. Developmental psychol-
project.org/package=lme4 ogy, 39(1), 164.
Lazarus, R., Dodd, H., Majdandzić, M., Vente, W., Morris, T., Byrow, Y.,. Sanson, A., Smart, D., Prior, M., Oberklaid, F., & Pedlow, R. (1994). The
.. Hudson, J.L. (2016). The relationship between challenging par- structure of temperament from age 3 to 7 years: age, sex, and
enting behaviour and childhood anxiety disorders (Vol. 190). sociodemographic influences. Merrill-Palmer Quarterly, 40, 233–
Lewis-Morrarty, E., Degnan, K. A., Chronis-Tuscano, A., Rubin, K. H., 252.
Cheah, C. S., Pine, D. S., et al. (2012). Maternal over-control mod- Schwartz, C. E., Snidman, N., & Kagan, J. (1999). Adolescent social
erates the association between early childhood behavioral inhibition anxiety as an outcome of inhibited temperament in childhood.
and adolescent social anxiety symptoms. Journal of Abnormal Child Journal of the American Academy of Child & Adolescent
Psychology, 40(8), 1363–1373. Psychiatry, 38(8), 1008–1015.
Lewis-Morrarty, E., Degnan, K. A., Chronis-Tuscano, A., Pine, D. S., Silverman, W. K., & Albano, A. M. (1996). The Anxiety Disorders
Henderson, H. A., & Fox, N. A. (2015). Infant attachment security Interview Schedule for Children for DSM-IV: Child and Parent
and early childhood behavioral inhibition interact to predict Versions. San Antonio: Psychological Corporation.
J Abnorm Child Psychol

Spence, S. H. (1998). A measure of anxiety symptoms among children. van Buuren, S., & Groothuis-Oudshoorn, K. (2011). Mice: multivariate
Behaviour Research & Therapy, 36(5), 545–566. imputation by chained equations in R. Journal of Statistical
Spence, S. H., Rapee, R., McDonald, C., & Ingram, M. (2001). The Software, 45(3), 1–67 URL https://www.jstatsoft.org/v45/i03/.
structure of anxiety symptoms among preschoolers. Behaviour van der Bruggen, C. O., Stams, G. J. J., & Bogels, S. M. (2008). Research
Research & Therapy, 39(11), 1293–1316. review: The relation between child and parent anxiety and parental
Thomasgard, M., & Metz, W. (1999). Parent-child relationship disorders: control: A meta-analytic review. Journal of Child Psychology and
what do the child vulnerability scale and the parent protection scale Psychiatry, 49(12), 1257–1269. https://doi.org/10.1111/j.1469-
measure? Clinical Pediatrics, 38, 347–356. 7610.2008.01898.x.
Thomasgard, M., Metz, W., Edelbrock, C., & Shonkoff, J. P. (1995). Weems, C. F. (2008). Developmental trajectories of childhood anxiety:
Parent-child relationship disorders: I. Parental overprotection and Identifying continuity and change in anxious emotion.
the development of the Parent Protection Scale. Journal of Developmental Review, 28(4), 488–502. https://doi.org/10.1016/j.
Developmental & Behavioral Pediatrics, 16, 244–250. dr.2008.01.001.