View
224
Download
1
Category
Preview:
Citation preview
Running head: DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Developmental Pathways linking Childhood and Adolescent Internalizing, Externalizing,
Academic Competence, and Adolescent Depression
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Abstract
This study examined longitudinal pathways through three domains of adjustment from ages 4-5
to 14-15 (internalizing problems, externalizing problems, and academic competence) towards
depressive symptoms at age 16-17. Participants were 6,425 Canadian children followed bi-
annually as part of the National Longitudinal Study of Children and Youth. Within-domain (i.e.,
stability) effects were moderate in strength. We found longitudinal cross-domain effects across
one time point (i.e., one-lag cascades) between internalizing and externalizing in early childhood
(positive associations), and between academic competence and externalizing in later childhood
and adolescence (negative associations). We also found cascade effects over multiple time points
(i.e., multi-lag cascades); lower academic competence at age 4-5 and greater internalizing at age
6-7 predicted greater age 12-13 externalizing, and greater age 6-7 externalizing predicted greater
age 16-17 depression. Important pathways towards adolescent depression include a stability path
through childhood and adolescent internalizing, as well as a number of potential paths involving
all adjustment domains, highlighting the multifactorial nature of adolescent depression.
Keywords: depression, longitudinal, cascades, internalizing, externalizing, academic
competence
2
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Recent research in developmental psychopathology suggests that three general domains of
adaptation are of particular importance in understanding child and adolescent adjustment:
internalizing problems (or internalizing; i.e., anxiety, depression), externalizing problems (or
externalizing; i.e., conduct problems, aggression), and academic competence (e.g., school
grades, test scores) (Bornstein, Hahn, & Suwalsky, 2013; Burt & Roisman, 2010; Masten et al.,
2005; Moilanen, Shaw, & Maxwell, 2010; Obradovic, Burt, & Masten, 2010; Vaillancourt,
Brittain, McDougall, & Duku, 2013). Problems in one domain of adaptation may beget problems
in other domains (Masten, 2006), potentially starting a chain reaction of subsequent problems—
a set of processes referred to most commonly as developmental cascades (Masten & Cicchetti,
2010).
The associations among internalizing, externalizing, and academic competence may be
particularly important for understanding the development of depression, which typically has its
earliest onset in mid-to-late adolescence (Kessler & Bromet, 2013) and represent one of the most
common psychiatric problems experienced during this age period (Thapar, Collishaw, Pine, &
Thapar, 2012). Indeed, prospective studies have positively linked internalizing (Reinherz,
Paradis, Giaconia, Stashwick, & Fitzmaurice, 2003), externalizing (Chronis-Tuscano et al.,
2010) in childhood with adolescent depression. Other research suggests a negative association
between school grades and depression in adolescence (Frojd et al., 2008) and protective effects
of higher IQ in childhood against adolescent depression (Glaser et al. 2011; Harpur et al. 2015).
Therefore, adolescent depression may be seen as a natural extension of the co-
development of internalizing, externalizing, and academic competence. Nevertheless, although a
number of studies have examined the longitudinal inter-associations among these domains of
adaptation, both within domains (i.e., continuity effects) and across domains (i.e., cascade
3
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
effects) (Bornstein et al., 2013; Burt & Roisman, 2010; Masten et al., 2005; Moilanen et al.,
2010; Obradovic et al., 2010; Vaillancourt et al., 2013), no studies have examined adolescent
depressive symptoms as an outcome in developmental cascades models. Previous research has
also focused on one-lag cascade effects (i.e., cascade effects separated by only one time point).
However, it is also important to test higher-order (i.e., multi-lag) effects in order to examine the
unique contribution of early adjustment domains on later adjustment. Indeed, socio-emotional
adjustment early in life has been shown to have a significant impact on adolescent adjustment,
including adolescent depression (Chronis-Tuscano et al., 2010).
Previous Findings
The existing research suggests the relationships among internalizing, externalizing, and
academic competence are complex, involving bidirectional and transactional effects over time. It
has been suggested that externalizing interferes with academic competence, which in turn may
lead to greater internalizing (Patterson & Capaldi, 1990). Indeed, a handful of developmental
cascades studies have found evidence in favor of this pattern (Masten et al., 2005; Moilanen et
al., 2010; Obradovic et al., 2010; van Lier et al., 2012). However, in addition to externalizing,
Moilanen et al. (2010) have argued that internalizing may interfere with academic competence,
which then can negatively impact other domains of adjustment. Indeed, some research has shown
associations from early internalizing to subsequent externalizing (Englund & Siebenbruner,
2012) as well as subsequent math and reading ability (Grover, Ginsburg, & Ialongo, 2007).
Lower academic competence may also undermine subsequent externalizing problems, in that
students showing lower performance may be placed in classrooms with more problem behaviors,
which could increase their own externalizing behaviors (Moilanen et al., 2010). In fact, there is
evidence of negative cascade effects from standardized test scores to subsequent externalizing
4
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
(Burt & Roisman, 2010; Englund & Siebenbruner, 2012; Moilanen et al., 2010; Vaillancourt et
al., 2013).
Limitations of existing cascades research. Developmental cascade modeling is a robust
and conservative approach, allowing for specific cascade effects while controlling for cross-
sectional correlations between domains and continuity effects within domains (Bornstein, Hahn,
& Haynes, 2010). However, there are several limitations in the existing literature. First and
foremost, developmental cascades research typically includes relatively small samples, typically
with fewer than 300 participants (Masten et al., 2005), and no such studies have included
samples that are representative of the larger population. Thus, the use of population surveys can
greatly enhance the generalizability of findings and overcome the limitations of using small,
highly variable, and unrepresentative samples.
Second, in order to understand the complex developmental pathways among these
domains over time, it is important to examine cascade effects from childhood through to
adolescence with the inclusion of multiple time points. However, studies using time points in
close proximity have included only three or four time points and age ranges of three or four years
during either childhood (van Lier et al., 2012) or early adolescence (Vaillancourt et al., 2013).
Moreover, research following children through adolescence and into adulthood has included only
four time points that are spaced at seven, three, and ten years apart (Masten et al., 2005;
Obradovic et al., 2010). Such large gaps between time points suggest that continuity effects may
be under-controlled in modeling.
Third, research suggests boys exhibit greater externalizing (Baillargeon et al., 2007)
while girls exhibit greater internalizing (Lewinsohn, Gotlib, Lewinsohn, Seeley, & Allen, 1998)
as well as higher rates of adolescent depression (Wade, Cairney, & Pevalin, 2002). However,
5
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
existing studies have either not examined gender (Moilanen et al., 2010) or have only tested
overall gender invariance of cascade models rather than examining gender invariance of specific
longitudinal effects (Burt & Roisman, 2010; Masten et al., 2005; Vaillancourt et al., 2013). In
fact, overall gender invariance between models does not preclude gender non-invariance of
particular pathways, which may have important research implications.
Finally, it is important to determine whether associations observed among domains of
adjustment are robust to the inclusion of possible confounders. However, developmental cascade
studies have typically examined only a small number of additional covariates, most commonly
child IQ/cognitive function, parenting factors, and socio-economic status (Bornstein et al., 2010;
Burt & Roisman, 2010; Masten et al., 2005; Moilanen et al., 2010), while others have not
examined any control variables (van Lier et al., 2012). Past research suggests several other
possible confounders, including maternal depression (Leve, Kim, & Pears, 2005) family
environment (Leve et al., 2005), stressful life events (Cheung, Harden, & Tucker-Drob, 2014)
and childhood physical health problems (Mesman & Koot, 2001).
The Current Study
The purpose of the current study was to examine the relationships among internalizing,
externalizing, and academic competence from age 4-5 to 14-15, and depressive symptoms at age
16-17, in a large and representative sample of Canadians. We expected to find positive continuity
effects from age 4-5 to 14-15, based on previous research (Burt & Roisman, 2010; Englund &
Siebenbruner, 2012; Masten et al., 2005; van Lier et al., 2012). Findings from previous studies
regarding one-lag effects are inconsistent, with only mixed support for specific models and most
studies suggesting more complicated effect patterns. However, based on the findings of previous
studies (Masten et al., 2005; Moilanen et al., 2010), we expected internalizing and externalizing
6
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
to be positively associated cross-sectionally and longitudinally. We also expected internalizing
and externalizing to exhibit negative cross-sectional and longitudinal associations with academic
competence, based on previous research (Masten et al., 2005; Moilanen et al., 2010; Obradovic
et al., 2010; van Lier et al., 2012). Whereas multi-lag effects have not been included in similar
cascades studies, one recent study found a multi-lag effect from early externalizing to adolescent
depression (Nilsen et al., 2013). Thus, our examination of cascade effects among internalizing,
externalizing, and academic competence in the current study was exploratory, particularly with
regards to multi-lag paths.
Although our tests of effects predicting depressive symptoms at age 16-17 were also
exploratory, we expected internalizing at age 14-15 to be positively associated with adolescent
depressive symptoms, due to the conceptual overlap between these measures. Also, in line with
previous research, we anticipated positive cascade effects from externalizing, and negative
cascade effects from academic competence, to adolescent depressive symptoms (Chronis-
Tuscano et al., 2010; Frojd et al., 2008; Reinherz et al., 2003).
Method
Study Population
Participants were from the National Longitudinal Survey of Children and Youth
(NLSCY), a nationally representative prospective study of Canadians from 1994/1995 until
2008/2009 (Statistics Canada and Human Resources and Skills Development Canada, 2009). Our
sample included 6,425 children (3144 girls; 3281 boys) who were age 4-5 in either 1994/95 or
1996/97, followed every two years across seven waves. Additional demographic details have
been reported elsewhere (Findlay, Garner, & Kohen, 2013).
Procedure
7
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Measures of internalizing, externalizing, and academic competence were available every
two years from age 4-5 until age 14-15. Additionally, a measure of self-reported depressive
symptoms was available at age 16-17. Informed consent was provided by the Person Most
Knowledgeable (PMK) about the child, who was the mother in roughly 90% of cases. All
measures were assessed via computer-assisted face-to-face interviews, with the exception of
measures of academic competence (see below). The current study received ethical approval from
the Ottawa Health Science Network Research Ethics Board.
Measures
Internalizing and externalizing. Internalizing and externalizing were assessed using
items assessing symptoms of anxiety/depression (7 items; e.g., “My child seems to be unhappy,
sad or depressed,” “My child is too fearful or anxious”) and physical aggression/conduct
problems (7 items; e.g., “My child gets into many fights,” “My child threatens people”) on a
scale from 0 (never) to 2 (often). The items were derived from the Ontario Child Health Study
(OCHS-R) (Boyle et al., 1987) and the Child Behavior Checklist (CBCL) (Achenbach &
Edelbrock, 1981) with the intent of operationalizing DSM-III disorders (Boyle et al., 1993).
From ages 4 to 11, the PMK reported on these symptoms. From ages 12 to 15, the child self-
reported on the same set of items, albeit with slightly different wording (e.g., “I am unhappy or
sad,” “I get into many fights”). Subscale items were summed to create total scores ranging from
0 to 14, with Cronbach’s alphas between 0.79 and 0.81 across all cycles of the NLSCY
(Statistics Canada and Human Resources and Skills Development Canada, 2002, 2008).
Academic competence. Children aged 4-5 completed the Peabody Picture Vocabulary
Test- Revised (PPVT-R) (Dunn & Dunn, 2006), which assesses verbal ability and scholastic
8
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
aptitude. The interviewer administered the test in the child’s home. The child was asked to
identify pictures corresponding to the word the interviewer read out.
Under the supervision of their teacher, children age 6-7 through 14-15 completed a
Mathematics Computation Test (MCT), a shortened version of the MCT component of the
Canadian Achievement Tests (CAT/2) that measures understanding of addition, subtraction,
multiplication, and division of whole numbers. For the PPVT-R and MCT, standard scores were
calculated based on norm samples from across Canada (Statistics Canada and Human Resources
and Skills Development Canada, 1998).
Adolescent depressive symptoms. Adolescents aged 16-17 completed a 12-item version
of the Center for Epidemiologic Studies Depression scale (CES-D) (Radloff, 1977, 1991).
Respondents were asked the following question: “How often have you felt or behaved this way
during the past week (7 days)?”, and rated each item (e.g., “I felt depressed”) on a scale from 0
(rarely or none of the time) to 3 (most or all of the time). Items were summed to create a total
depression score, with scores ranging from 0 to 36. A Cronbach’s alpha of 0.83 was reported in
cycle 8 of the NLSCY (Statistics Canada and Human Resources and Skills Development Canada,
2008).
Additional covariates. A number of covariates were measured at baseline (age 4-5).
Socioeconomic status (SES) was measured with a ratio of household income adjusted for the
low-income cut off, which considers an individual’s income relative to the community in which
they live and the size of their family (Cotton, 2001).
The PMK reported on statements addressing aspects of family functioning such as
problem solving, communications, and affective involvement (12 items; e.g., “We don’t get
along well together”). The PMK indicated the degree to which each item describes their family
9
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
on a scale from 0 (strongly disagree) to 3 (strongly agree), with scores ranging from 0 to 36 and
higher total scores indicating greater family dysfunction.
The PMK completed an adapted version of the Parent Practices Scale (Strayhorn &
Weidman, 1988) to rate the degree to which they engaged in hostile parenting (7 items; e.g.,
“how often do you get angry when you punish your child?”). Items were rated on a scale from 0
(never) to 4 (many times each day), with scores ranging from 0 to 28.
The PMK was asked the following question to assess the presence of a chronic illness in
their child: “Does [child] have any of the following long-term conditions that have been
diagnosed by a health professional”? (16 conditions; e.g., diabetes, high blood pressure).
Two additional variables were used as time-varying covariates (measured at baseline
through age 14-15). The PMK completed the same short version of the CES-D (Radloff, 1977,
1991) as completed by children at age 16-17, as a measure of maternal depression. Also, the
PMK reported whether their child had experienced any of 14 stressful life events (SLEs) (e.g.,
death of a family member). Previous research indicates that two or more childhood traumatic and
stressful events are particularly predictive of adolescent depression compared to one event or no
events (Colman et al., 2013). We thus dichotomized this variable such that the presence of two or
more SLEs indicated ‘high stress’.
Statistical Analysis
Structural equation modeling was used to examine the associations among internalizing,
externalizing, and academic competence over time. Three models were compared in terms of
relative fit, starting with the most constrained (i.e., most parsimonious). In all models, cross-
sectional correlations (i.e., cross-domain associations at each time point) were estimated. Model
1 (autoregressive model) included cross-sectional correlations at each age, as well as continuity
10
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
paths. Due to conceptual and measurement similarity, a continuity path from internalizing at age
14-15 to depression at age 16-17 was included in these models. Model 2 (one-lag cascades
model) included paths from Model 1, with the addition of one-lag cascade paths (i.e., variables at
one time point could only predict cross-domain variables at the next time point). Model 3 (multi-
lag cascades model) included all paths in the previous models, and also included all multi-lag
cascade paths.
Models were fitted using Mplus Version 6 (Muthén & Muthén, 1998-2010). Relative fit
of models was assessed using a scaled chi-square difference test (Satorra & Bentler, 2001).
Absolute model fit was evaluated using the comparative fit index (CFI), with values > .90
indicating acceptable fit and values > .95 indicating a close fit (Hu & Bentler, 1999), the root
mean square error of approximation (RMSEA), with values < .05 indicating a good fit
(MacCallum, Browne, & Sugawara, 1996), and the standardized root mean square residual
(SRMR), with values < .08 indicating a good fit (Hu & Bentler, 1998). Missing data were
handled using full information maximum likelihood (FIML) estimation under a missing at
random (MAR) assumption.
Survey responses were weighted using sampling weights provided by Statistics Canada.
These weights have been adjusted to account for non-response at each cycle, resulting in
representativeness among responders and non-responders (Statistics Canada and Human
Resources and Skills Development Canada, 2008). In order to accommodate the use of these
sampling weights, as well as to account for non-normal data, MLR estimators with robust
standard errors were used in model estimation.
Following previous cascades research (Vaillancourt et al., 2013; van Lier et al., 2012),
overall gender invariance of cascade paths in Model 3 was tested by comparing a model with all
11
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
one-lag and multi-lag paths constrained to be equal across gender to an unconstrained model.
Model 3 was then examined separately for boys and girls to identify potential non-invariance of
significant cascade paths. Finally, scaled chi-square tests were used to compare a constrained
model (i.e., gender-specific path constrained across gender) to an unconstrained model for each
path of interest.
Finally, in order to test the robustness of observed paths in Model 3, the model was
examined with the inclusion of covariates. Indeed, FIML is optimized with the inclusion of
additional variables. All observed variables were regressed on baseline covariates, and observed
variables at each age were regressed on the two time-varying covariates measured at that age
(with those at age 14-15 in the case of depression at age 16-17).
Results
Preliminary Analysis
Descriptive statistics for the main study variables are presented in Table 1. As shown, the
number of responses decreases at each subsequent time point. We examined the degree to which
dropout was associated with study covariates as well as the main cascade variables. Results
indicated that dropout was associated with the presence of a chronic illness (OR = 1.17; 95% CI:
1.00, 1.37) and 2 or more SLEs (OR = 1.59, 95% CI: 1.13, 2.25), as well as lower SES (b = -
0.12, SE = 0.03, p < .0001) and less hostile parenting (b = -0.08, SE=0.11, p < .0001), at
baseline. Also, greater academic competence at age 4-5 was negatively associated with dropout
(OR = 0.98, 95% CI: 0.96, 0.99).
Cascade Analysis
Results for structural equation models are presented in Table 2. Results indicated that the
‘one-lag cascades’ model (Model 2) showed a significant improvement in fit over the
12
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
‘autoregressive’ model (Model 1). However, the ‘multi-lag cascades’ model (Model 3) showed a
significant improvement in fit as compared to Model 2. Model 3 also had acceptable absolute fit
statistics, and was thus chosen as the best fitting model.
Significant cross-sectional correlations among variables are shown in Table 3.
Internalizing and externalizing were positively correlated at each time point, whereas
externalizing and academic competence were negatively correlated at ages 12-13 and 14-15.
Internalizing and academic competence were only correlated cross-sectionally at age 12-13.
Figure 1 shows the standardized path coefficients for all significant pathways in Model 3.
All continuity effects within each domain were positive and significant. A number of significant
one-lag cascade effects were identified. Greater internalizing at age 4-5 predicted greater
externalizing at age 6-7 and greater externalizing at age 4-5 predicted greater internalizing at age
6-7. Academic competence negatively predicted externalizing from age 8-9 to age 10-11, and
from age 10-11 to 12-13. Externalizing at age 8-9 negatively predicted academic competence at
age 12-13. Externalizing at age 12-13 negatively predicted academic competence at age 14-15
and positively predicted depression at age 16-17.
A number of multi-lag cascades were also identified. Greater age 4-5 internalizing
predicted greater academic competence at age 10-11, greater age 6-7 externalizing predicted
greater depression at age 16-17, and lower age 4-5 academic competence predicted greater
internalizing and externalizing at age 12-13.
The gender-constrained model had a significantly worse fit than the unconstrained model
(∆χ2=161.79 (118), p= .005), suggesting overall gender non-invariance. Gender-stratified models
indicated that four paths remained significant only for girls, while four paths remained
significant only for boys. Constrained versus unconstrained models were examined for each
13
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
apparent gender-specific path (see Table 4). In two cases, the constrained model provided a
significantly worse fit, suggesting gender non-invariance. Specifically, the path from
externalizing at age 12-13 to depression at age 16-17 was found to be stronger among girls,
while the path from internalizing at age 8-9 to externalizing at age 12-13 was found to be
stronger among boys. These two paths are noted in Figure 1 (the G superscript).
Finally, Model 3 was examined with the inclusion of baseline and time-varying
covariates (Model 4). This model showed an acceptable absolute fit (CFI= 0.926; RMSEA=
0.020; SRMR= 0.03), and all significant paths from Model 3 remained significant. However,
Model 4 also had a relatively poor fit compared to Model 3 (∆χ2=245.20 (192), p= .006) and thus
Model 3 was retained as the final model. However, for contextual purposes, significant pathways
involving additional covariates from Model 4 are presented in Supplemental Table 1.
Discussion
We used a representative sample of Canadians to examine developmental cascade
pathways among externalizing, internalizing, and academic competence from age 4-5 to age 14-
15 in the prediction of age 16-17 adolescent depressive symptoms. We identified unique effects
of early adjustment on later adjustment, while controlling for significant cross-sectional
correlations and continuity effects. These effects were also robust to the inclusion of a large
number of additional covariates. Our findings add to the growing evidence of complex inter-
associations among these domains in childhood and adolescence, as well as their role in
predicting adolescent depressive symptoms.
Continuity Paths
Significant continuity paths were consistent with previous research showing the relative
stability of internalizing and externalizing in early childhood and adolescence (Pianta &
14
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Caldwell, 1990; Reitz, Dekovic, & Meijer, 2005) and of academic achievement in later
childhood (Feshbach & Feshbach, 1987), as well as previous developmental cascade studies
(Bornstein et al., 2010; Englund & Siebenbruner, 2012; Vaillancourt et al., 2013; van Lier et al.,
2012). These results highlight the central role of previous adjustment in each of these domains of
development in predicting later adjustment in the same domain.
Cascade Paths
Internalizing and externalizing. Several significant cascade paths were also identified.
Internalizing and externalizing showed reciprocal positive effects between age 4-5 and 6-7,
consistent with previous studies examining the co-development of internalizing and externalizing
(Gilliom & Shaw, 2004). A possible explanation is that internalizing and externalizing in early
childhood are being driven by common factors, such as high negative emotionality (Gilliom &
Shaw, 2004). Additionally, our subsequent examination of covariates indicated that greater
hostile parenting predicted greater internalizing and externalizing at age 4-5, suggesting that both
factors might be driven by parent-child interactions. In fact, previous research has shown that
increases in negative parenting are associated with the development of internalizing and
externalizing throughout childhood (Braza et al., 2015). These findings add to previous cascades
research, wherein almost no studies have followed children any earlier than age 6 and thus have
been unable to examine these pathways.
Greater internalizing at age 6-7 predicted greater externalizing at age 12-13. Although
this findings is consistent with effects shown in one previous cascades study (Bornstein et al.,
2010), most similar studies have not shown longitudinal associations from internalizing to
externalizing. Importantly, the current study included an examination of multi-lag effects. It is
15
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
not clear whether similar effects from internalizing to externalizing would have been found in
previous research if such effects had been examined.
Externalizing and academic competence. We also found negative cascade effects
between academic competence and externalizing in from age 8-9 to age 12-13. These findings
are consistent with some previous cascades research (Englund & Siebenbruner, 2012; Moilanen
et al., 2010; van Lier et al., 2012). However, our findings also highlight the potential importance
of academic success as a protective factor against antisocial behavior beyond childhood. The
bidirectional nature of these effects supports the findings of one cascades study with a similar
age range as in the current study (Moilanen et al., 2010). Notably, lower academic competence at
age 4-5 predicted greater externalizing at age 12-13. Although there were no cross-sectional
associations between these variables before age 12-13, there may be lingering effects of early
cognitive problems on behavior problems many years later. Also, our subsequent analysis of
covariates showed that lower SES at age 4-5 predicted lower academic competence at age 4-5
and age 12-13 as well as greater internalizing and externalizing at age 12-13, which supports past
research (Letourneau, Duffett-Leger, Levac, Watson, & Young-Morris, 2013). Therefore, this
finding may reflect a more overarching effect of economic disadvantage.
Internalizing and academic competence. Lower academic competence at age 4-5 also
predicted greater internalizing at age 12-13, which is consistent with a recent study showing
protective effects of childhood IQ on depressive symptoms in early adolescence (Glaser et al.,
2011) and which might explained by previously mentioned socioeconomic factors. In contrast,
we found no evidence that internalizing interfered with academic competence, consistent with
some cascades studies (Masten et al., 2005; Vaillancourt et al., 2013; van Lier et al., 2012). As
suggested by Masten and colleagues (Masten et al., 2005), internalizing might only interfere with
16
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
academic outcomes for those in the clinical range of internalizing symptoms. The only
significant cascade path from internalizing to academic competence was a positive association
from age 4-5 to age 10-11. One possible explanation is that children with elevated internalizing
symptoms are more likely to worry about academic performance, which could potentially result
in more diligent school work and, in turn, greater subsequent academic achievement. However,
this result contradicts previous research showing a negative association between internalizing
and academic achievement in childhood (Rapport, Denney, Chung, & Hustace, 2001) and may
reflect our use of different measures of academic competence, as well as a younger age range,
than in previous research.
Cascades predicting adolescent depression. Greater externalizing at age 6-7 predicted
greater depression at age 16-17, consistent with research showing that aggressive behavior in
childhood predicts greater odds of mood disorders 14 years later (Hofstra, van der Ende, &
Verhulst, 2002), as well as a recent study that showed a direct link from externalizing at age 4.5
to depressive symptoms in adolescence (ages 14.5 – 16.5), over and above the effects of
internalizing and maternal distress (Nilsen et al., 2013). These authors suggest that this effect
might be due to externalizing behavior becoming increasingly non-normative through childhood,
as evidenced by trajectories research (Cheung et al., 2014). However, apart from the gender-
modified path from age 12-13, our findings suggest that externalizing in later childhood and
adolescence were not predictive of subsequent adolescent depressive symptoms, also mirroring
the findings of Nilsen and colleagues (2013). One possible explanation is that delinquency is
more normative during adolescence than in childhood (Moffitt, Caspi, Harrington, & Milne,
2002), as noted by Nilsen et al. (2013).
17
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Due to the large number of possible indirect pathways, such pathways were not tested in
the current study. However, the pattern of findings suggests that adolescent depression can be the
end result of a number of indirect pathways, which highlights the concept of equifinality (i.e., the
same effect resulting from a number of different events) (Cicchetti & Rogosch, 1996), and
reinforces the notion that adolescent depression is a multifactorial phenomenon. However, our
findings also suggest that externalizing at 12-13 may be particularly important in terms of
understanding the associations among these variables. It is worth noting that this age period
coincides with the transition to middle school and the onset of puberty. Middle school transition
is known to be a potentially stressful period often marked by changes in peer relationships
(Hirsch & Rapkin, 1987), distress levels (Crockett, Peterson, Graber, Schulenberg, & Ebata,
1989), and academic achievement (Simmons & Blyth, 1987), all of which could negatively affect
child adjustment. Also, children who mature sooner are more likely to exhibit externalizing,
possibly as a result of affiliation with delinquent peers (Ge, Brody, Conger, Simons, & Murry,
2002).
Gender differences
Two paths were shown to vary by gender. First, although greater internalizing at age 8-9
predicted lower age 12-13 externalizing, this path was stronger among boys, consistent with
previous studies showing similar effects (Masten et al., 2005). In fact, other studies have found
similar effects in male-only samples (Kerr, Tremblay, Pagani, & Vitaro, 1997) or have found this
effect only among boys (Mesman, Bongers, & Koot, 2001). As has been suggested (Moilanen et
al., 2010), this negative association may be due to the potentially attenuating effect of inhibition
on risk-taking behaviors, which are more typical among males (Byrnes, Miller, & Schafer,
1999). Second, greater externalizing at age 12-13 predicted greater depression at age 16-17, but
18
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
was stronger among girls. This result is consistent with the idea of the “gender paradox” effect
(Loeber & Keenan, 1994), where gender non-normative behaviors such as physical aggression in
girls are proposed to result in poorer outcomes. Also, one previous study showed that anger
expression was related to depressed mood among girls, but not boys (Clay, Hagglund, Kashani,
& Frank, 1996). There is also a well-documented gender divergence in the prevalence of
depression at age 14 (Wade et al., 2002), and our findings highlight a possible mechanism
underlying this gender difference.
Limitations
This study had a number of important limitations that are noteworthy. First, measures of
childhood externalizing and internalizing were completed by parents until age 10-11, and were
subsequently self-reported. Also, academic competence was measured with the PPVT-R at age
4-5 and with the MCT subsequently, and these measures may not be comparable to each other.
Although all continuity paths were significant for each adjustment domain, these differences in
construct measurement may have weakened continuities within domains, as can be seen by the
relatively smaller path coefficients between ages 4-5 to 6-7 for academic competence and
between age 10-11 and 12-13 for internalizing and externalizing. Differences in construct
measurement may also have affected the consistency of cascade effects in the current study.
Second, the measures of academic competence in the current study were limited to standardized
measures of verbal ability and math ability. Although standardized measures have been included
in other cascades research (Burt et al., 2010; Masten et al., 2005), this same research also
included additional measures of achievement such as school grades. Thus, some caution is need
in comparing our results to those of previous cascades studies. Third, the chance of Type I errors
is increased due to the large number of associations tested.
19
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Contributions
Despite these limitations, our study also contributed to the existing literature in a number
of important ways. First, we used a large and nationally representative sample, giving our results
greater generalizability than those of previous research that included much smaller and less
representative samples. Ours is also among the very few studies that have examined these
adjustment domains from early childhood into adolescence. This allowed for the examination of
direct longitudinal effects over multiple time points. Although the focus of developmental
cascades research has been on one-lag cascade effects, our identification of multi-lag effects
suggests these may be important components of cascade models that are generally overlooked.
Implications
Finally, our findings have implications for intervention programs that may impact
adolescent outcomes such as depressive symptoms. We identified not only stable adjustment
domains, but also unique effects on adjustment several years later, highlighting the potential
long-term impact of early childhood interventions. In particular, early targeting of externalizing
may have beneficial effects on the risk of depression in adolescence. Also, the presence of
complex associations among domains suggests that multiple areas of adjustment should be
jointly targeted in such programs.
20
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
References
Achenbach, T. M., & Edelbrock, C. S. (1981). Behavioral-Problems and Competencies Reported by Parents of Normal and Disturbed-Children Aged 4 through 16. Monographs of the Society for Research in Child Development, 46, 1-82.
Baillargeon, R. H., Zoccolillo, M., Keenan, K., Cote, S., Perusse, D., Wu, H. X., Boivin, M., & Tremblay, R. E. (2007). Gender differences in physical aggression: A prospective population-based survey of children before and after 2 years of age. Dev Psychol, 43, 13-26.
Bornstein, M. H., Hahn, C. S., & Haynes, O. M. (2010). Social competence, externalizing, and internalizing behavioral adjustment from early childhood through early adolescence: Developmental cascades. Development and Psychopathology, 22, 717-735.
Bornstein, M. H., Hahn, C. S., & Suwalsky, J. T. D. (2013). Language and internalizing and externalizing behavioral adjustment: Developmental pathways from childhood to adolescence. Development and Psychopathology, 25, 857-878.
Boyle, M. H., Offord, D. R., Hofmann, H. G., Catlin, G. P., Byles, J. A., Cadman, D. T., Crawford, J. W., Links, P. S., Raegrant, N. I., & Szatmari, P. (1987). Ontario Child Health Study .1. Methodology. Archives of General Psychiatry, 44, 826-831.
Boyle, M. H., Offord, D. R., Racine, Y., Sanford, M., Szatmari, P., & Fleming, J. E. (1993). Evaluation of the Original Ontario-Child-Health-Study Scales. Canadian Journal of Psychiatry-Revue Canadienne De Psychiatrie, 38, 397-405.
Braza, P., Carreras, R., Munoz, J. M., Braza, F., Azurmendi, A., & Sagastizabal. (2015). Negative maternal and paternal parenting styles as predictors of children's behavioral problems: Moderating effect of the child's sex. Journal of Child and Family Studies, 24, 847-856.
Burt, K. B., & Roisman, G. I. (2010). Competence and psychopathology: cascade effects in the NICHD Study of Early Child Care and Youth Development. Development and Psychopathology, 22, 557-567.
Byrnes, J. P., Miller, D. C., & Schafer, W. D. (1999). Gender differences in risk taking: A meta-analysis. Psychological Bulletin, 125, 367-383.
Cheung, A. K., Harden, K. P., & Tucker-Drob, E. M. (2014). GenexEnvironment Interactions in Early Externalizing Behaviors: Parental Emotional Support and Socioeconomic Context as Moderators of Genetic Influences? Behavior Genetics.
Chronis-Tuscano, A., Molina, B. S. G., Pelham, W. E., Applegate, B., Dahlke, A., Overmyer, M., & Lahey, B. B. (2010). Very Early Predictors of Adolescent Depression and Suicide Attempts in Children With Attention-Deficit/Hyperactivity Disorder. Archives of General Psychiatry, 67, 1044-1051.
Cicchetti, D., & Rogosch, F. A. (1996). Equifinality and multifinality in developmental psychopathology. Development and Psychopathology, 8, 597-600.
Clay, D. L., Hagglund, K. J., Kashani, J. H., & Frank, R. G. (1996). Sex differences in anger expression, depressed mood, and aggression in children and adolescents. Journal of Clinical Psychology in Medical Settings, 3, 79-92.
Colman, I., Garad, Y., Zeng, Y., Naicker, K., Weeks, M., Patten, S. B., . . . Wild, T. C. (2013). Stress and development of depression and heavy drinking in adulthood: Moderating effects of childhood trauma. Social Psychiatry and Psychiatric Epidemiology, 48, 265-274.
Cotton, C. (2001). Developments in the Low Income Cutoffs. In Cotton, C. Ottawa, ON: Statistics Canada.
Crockett, L. J., Peterson, A. C., Graber, J. A., Schulenberg, J. E., & Ebata, A. (1989). School transitions and adjustment during early adolescence. Journal of Early Adolescence, 9, 181–210.
Dunn, L., & Dunn, L. (2006). Peabody Picture Vocabulary Test. In: Pearson.
21
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Englund, M. M., & Siebenbruner, J. (2012). Developmental pathways linking externalizing symptoms, internalizing symptoms, and academic competence to adolescent substance use. Journal of Adolescence, 35, 1123-1140.
Feshbach, N. D., & Feshbach, S. (1987). Affective Processes and Academic-Achievement. Child Development, 58, 1335-1347.
Findlay, L., Garner, R., & Kohen, D. (2013). Trajectories of parental monitoring knowledge and their associations with adolescents' substance use, poor academic outcome, and behaviour problems. International Journal of Child, Youth and Family Studies, 4, 403-424.
Frojd, S. A., Nissinen, E. S., Pelkonen, M. U. I., Marttunen, M. J., Koivisto, A. M., & Kaltiala-Heino, R. (2008). Depression and school performance in middle adolescent boys and girls. Journal of Adolescence, 31, 485-498.
Ge, X., Brody, G. H., Conger, R. D., Simons, R. L., & Murry, V. M. (2002). Contextual amplification of pubertal transition effects on delinquent peer affiliation and externalizing behavior among African American children. Developmental Psychology, 38, 42–54.
Gilliom, M., & Shaw, D. S. (2004). Codevelopment of externalizing and internalizing in early childhood. Development and Psychopathology, 16, 313-333.
Glaser, B., Gunnell, D., Timpson, N. J., Joinson, C., Zammit, S., Smith, G. D., & Lewis, G. (2011). Age- and puberty-dependent association between IQ score in early childhood and depressive symptoms in adolescence. Psychol Med, 41, 333-343.
Grover, R. L., Ginsburg, G. S., & Ialongo, N. (2007). Psychosocial outcomes of anxious first graders: A seven-year follow-up. Depression and Anxiety, 24, 410-420.
Harpur, L. J., Polek, E., & van Harmelen, A. (2015). The role of timing of maltreatment and child intelligence in pathways to low symptoms of depression and anxiety in adolescence. Child Abuse and Neglect, 47, 24-37.
Hirsch, B. J., & Rapkin, B. D. (1987). The transition to junior high school: A longitudinal study of self-esteem, psychological symptomatology, school life and social support. Child Development, 58, 1235–1243.
Hofstra, M. B., van der Ende, J., & Verhulst, F. C. (2002). Child and adolescent problems predict DSM-IV disorders in adulthood: A 14-year follow-up of a Dutch epidemiological sample. Journal of the American Academy of Child and Adolescent Psychiatry, 41, 182-189.
Hu, L. T., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification. Psychological Methods, 3, 424-453.
Hu, L. T., & Bentler, P. M. (1999). Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria Versus New Alternatives. Structural Equation Modeling-a Multidisciplinary Journal, 6, 1-55.
Kaltiala-Heino, R., Marttunen, M., Rantanen, P., & Rimpela, M. (2003). Early puberty is associated with mental health problems in middle adolescence. Social Science and Medicine, 57, 1055 – 1064.
Kerr, M., Tremblay, R. E., Pagani, L., & Vitaro, F. (1997). Boys' behavioral inhibition and the risk of later delinquency. Archives of General Psychiatry, 54, 809-816.
Kessler, R. C., & Bromet, E. J. (2013). The epidemiology of depression across cultures. Annual Review of Public Health, 34, 119-138.
Kim, K. J., Conger, R. D., Elder, G. H., Jr., & Lorenz, F. O. (2003). Reciprocal influences between stressful life events and adolescent internalizing and externalizing. Child Development, 74, 127-143.
Letourneau, N. L., Duffett-Leger, L., Levac, L., Watson, B., & Young-Morris, C. (2013). Socioeconomic Status and Child Development: A Meta-Analysis. Journal of Emotional and Behavioral Disorders, 21, 211-224.
Leve, L. D., Kim, H. K., & Pears, K. C. (2005). Childhood temperament and family environment as predictors of internalizing and externalizing trajectories from ages 5 to 17. J Abnorm Child Psychol, 33, 505-520.
22
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Lewinsohn, P. M., Gotlib, I. H., Lewinsohn, M., Seeley, J. R., & Allen, N. B. (1998). Gender differences in anxiety disorders and anxiety symptoms in adolescents. J Abnorm Psychol, 107, 109-117.
Loeber, R., & Keenan, K. (1994). Interaction between Conduct Disorder and Its Comorbid - Effects of Age and Gender. Clinical Psychology Review, 14, 497-523.
MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for covariance structure modeling. Psychological Methods, 1, 130-149.
Masten, A. S. (2006). Developmental psychopathology: Pathways to the future. International Journal of Behavioral Development, 30, 47-54.
Masten, A. S., & Cicchetti, D. (2010). Developmental cascades. Development and Psychopathology, 22, 491-495.
Masten, A. S., Roisman, G. I., Long, J. D., Burt, K. B., Obradovic, J., Riley, J. R., Boelcke-Stennes, K., & Tellegen, A. (2005). Developmental cascades: linking academic achievement and externalizing and internalizing symptoms over 20 years. Dev Psychol, 41, 733-746.
Mesman, J., Bongers, I. L., & Koot, H. M. (2001). Preschool developmental pathways to preadolescent internalizing and externalizing. Journal of Child Psychology and Psychiatry and Allied Disciplines, 42, 679-689.
Mesman, J., & Koot, H. M. (2001). Early preschool predictors of preadolescent internalizing and externalizing DSM-IV diagnoses. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1029-1036.
Moffitt, T. E., Caspi, A., Harrington, H., & Milne, B. J. (2002). Males on the life-course-persistent and adolescence-limited antisocial pathways: follow-up at age 26 years. Development and Psychopathology, 14, 179-207.
Moilanen, K. L., Shaw, D. S., & Maxwell, K. L. (2010). Developmental cascades: Externalizing, internalizing, and academic competence from middle childhood to early adolescence. Development and Psychopathology, 22, 635-653.
Muthén, L. K., & Muthén, B. O. (1998-2010). Mplus User's Guide (Sixth Edition ed.). Los Angeles, CA: Muthén & Muthén.
Nilsen, W., Gustavson, K., Roysamb, E., Kjeldsen, A., & Karevold, E. (2013). Pathways from maternal distress and child problem behavior to adolescent depressive symptoms: a prospective examination from early childhood to adolescence. Journal of Developmental and Behavioral Pediatrics, 34, 303-313.
Obradovic, J., Burt, K. B., & Masten, A. S. (2010). Testing a dual cascade model linking competence and symptoms over 20 years from childhood to adulthood. J Clin Child Adolesc Psychol, 39, 90-102.
Patterson, G. R., & Capaldi, D. M. (1990). A mediational model for boys' depressed mood. In J. Rolf, A. S. Masten, D. Cicchetti, K. H. Nuechterlein & S. Weintraub (Eds.), Risk and protective factors in the development of psychopathology (pp. 141-163). New York: Cambridge University Press.
Pianta, R. C., & Caldwell, C. B. (1990). Stability of externalizing symptoms from kindergarten to first grade and factors related to instability. Development and Psychopathology, 2, 247-258.
Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401.
Radloff, L. S. (1991). The Use of the Center for Epidemiologic Studies Depression Scale in Adolescents and Young-Adults. Journal of Youth and Adolescence, 20, 149-166.
Rapport, M. D., Denney, C. B., Chung, K. M., & Hustace, K. (2001). Internalizing behavior problems and scholastic achievement in children: Cognitive and behavioral pathways as mediators of outcome. Journal of Clinical Child Psychology, 30, 536-551.
Reinherz, H. Z., Paradis, A. D., Giaconia, R. M., Stashwick, C. K., & Fitzmaurice, G. (2003). Childhood and adolescent predictors of major depression in the transition to adulthood. American Journal of Psychiatry, 160, 2141-2147.
Reitz, E., Dekovic, M., & Meijer, A. M. (2005). The structure and stability of externalizing and internalizing problem behavior during early adolescence. Journal of Youth and Adolescence, 34, 577-588.
23
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Romano, E., Babchishin, L., Pagani, L. S., & Kohen, D. (2010). School readiness and later achievement: replication and extension using a nationwide Canadian survey. Dev Psychol, 46, 995-1007.
Satorra, A., & Bentler, P. M. (2001). A scaled difference chi-square test statistic for moment structure analysis. Psychometrika, 66, 507-514.
Simmons, R. G., & Blyth, D. A. (1987). Moving into adolescence: The impact of pubertal change and school context. Hawthorne, NY: Aldine de Gruyter.
Statistics Canada and Human Resources and Skills Development Canada. (1998). National Longitudinal Survey of Children and Youth - Cycle 1 Data Dictionary. In Statistics Canada and Human Resources and Skills Development Canada. Ottawa, ON: Statistics Canada.
Statistics Canada and Human Resources and Skills Development Canada. (2002). National Longitudinal Survey of Children and Youth (NLSCY), Cycle 3 - Microdata User Guide. In Statistics Canada and Human Resources and Skills Development Canada. Ottawa, ON: Statistics Canada.
Statistics Canada and Human Resources and Skills Development Canada. (2008). National Longitudinal Survey of Children and Youth (NLSCY), Cycle 8 - Microdata User Guide. In Statistics Canada and Human Resources and Skills Development Canada. Ottawa, ON: Statistics Canada.
Statistics Canada and Human Resources and Skills Development Canada. (2009). National Longitudinal Survey of Children and Youth: Survey Overview for the 2008/2009 Data Collection, Cycle 8. In Statistics Canada and Human Resources and Skills Development Canada. Ottawa, ON: Statistics Canada.
Strayhorn, J. M., & Weidman, C. S. (1988). A Parent Practices Scale and Its Relation to Parent and Child Mental-Health. Journal of the American Academy of Child and Adolescent Psychiatry, 27, 613-618.
Thapar, A., Collishaw, S., Pine, D. S., & Thapar, A. K. (2012). Depression in adolescence. Lancet, 379, 1056-1067.
Vaillancourt, T., Brittain, H. L., McDougall, P., & Duku, E. (2013). Longitudinal Links Between Childhood Peer Victimization, Internalizing and Externalizing, and Academic Functioning: Developmental Cascades. Journal of Abnormal Child Psychology, 41, 1203-1215.
van Lier, P. A. C., Vitaro, F., Barker, E. D., Brendgen, M., Tremblay, R. E., & Boivin, M. (2012). Peer Victimization, Poor Academic Achievement, and the Link Between Childhood Externalizing and Internalizing ProblemsInternalizing. Child Development, 83, 1775-1788.
Wade, T. J., Cairney, J., & Pevalin, D. J. (2002). Emergence of gender differences in depression during adolescence: national panel results from three countries. Journal of the American Academy of Child and Adolescent Psychiatry, 41, 190-198.
24
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Table 1 Means (SDs) or Proportions of Internalizing, Externalizing, Academic Competence, Depressive Symptoms, and Maternal Depression, 1 Stressful Life Event, 2 or More Stressful Life Events, Family Dysfunction, Hostile Parenting, Presence of a Chronic Illness, and Socioeconomic Status
Time point(age in years)
# of cases Int Ext Acad Dep Mat Dep 1 SLE ≥ 2
SLEsFamily
DysfuncHostile
ParChronic Illness SES
1 (4-5) 6425 2.10(2.2)
1.61(2.0)
99.24(15.2)
— 4.74(5.2)
20.02 2.99 8.12 (5.2)
9.08 (3.7)
36.71 1.78(1.2)
2 (6-7) 4939 2.49(2.5)
1.40(1.8)
311.6(48.0)
— 4.48(5.4)
19.22 2.41 — — — —
3 (8-9) 4645 2.64(2.5)
1.31(1.8)
364.6(56.3)
— 4.28(5.5)
18.54 1.95 — — — —
4 (10-11) 4281 2.51(2.4)
1.16(1.7)
439.0(57.2)
— 3.96(5.3)
19.24 1.49 — — — —
5 (12-13) 3975 3.30(2.7)
1.05(1.8)
494.8(70.0)
— 3.67(5.0)
17.93 1.91 — — — —
6 (14-15) 3741 3.39(2.8)
1.01(1.7)
578.0(91.9)
— 3.67(5.1)
16.25 2.12 — — — —
7 (16-17) 3539 — — — 8.04 (5.84)
— — — — — — —
Int = Internalizing; Ext = Externalizing; Aca = Academic Competence; Dep = Depressive symptoms; Mat Dep = Maternal Depression; SLE = Stressful Life Events; Family Dysfunc = Family Dysfunction; Hostile Par = Hostile Parenting; SES = Socioeconomic Status.
25
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Table 2 Nested Model Comparisons and Fit Statistics
Path analyses Difference test of relative fit Absolute fit statistics
Model df c χ2 Comp. cd ∆χ2 ∆df p CFI RMSEA SRMR
1 135 3.070 798.24 2 vs. 1 3.256 92.20 30 .000 .892 .028 .0742 105 3.017 712.77 3 vs. 2 3.068 100.86 70 .009 .901 .030 .0623 35 2.916 631.36 .903 .052 .055
c = weighting constant for computing the chi-square statistic using the robust estimation method; Comp. = model comparison; cd = weighting constant for the difference between two chi-square statistics using the robust estimation method; CFI = Comparitive Fit Index; RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual.
26
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Table 3
Significant cross-sectional associations for the final model (Model 3)
Time point(age in years)
Estimated Path β SE p
1 (4-5) Int1 Ext1 0.36 0.02 .000
2 (6-7) Int2 Ext2 0.35 0.02 .000
3 (8-9) Int3 Ext3 0.31 0.03 .000
4 (10-11) Int4 Ext4 0.34 0.04 .000
5 (12-13) Int5 Ext5 0.30 0.30 .000Int5 Aca5 -0.07 0.03 .045Ext5 Aca5 -0.09 0.03 .002
6 (14-15) Int6 Ext6 0.23 0.03 .000Ext6 Aca6 -0.10 0.03 .000
Ext = Externalizing; Int = Internalizing; Aca = Academic Competence.
27
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Table 4
Model 3 comparisons for gender analysis
Gender-specific pathwaysPath-specific gender invariance tests
cd ∆χ2 ∆df p
Significant paths among girls only:
Ext1 Int2 3.05 1.16 1 .281
Int1 Ext1 3.05 0.57 1 .450
Aca1 Int5 2.77 1.89 1 .170
Ext5 Dep7 2.11 62.26 1 .000
Significant paths among boys only:
Ext2 Dep7 2.98 0.35 1 .556
Ext5 Aca6 2.20 1.15 1 .283
Aca4 Ext5 2.70 0.25 1 .614
Int3 Ext5 1.63 7.21 1 .007
Ext = Externalizing; Int = Internalizing; Aca = Academic Competence; Dep = Depressive symptoms. Numbers denote time point (1= age 4-5; 2 = age 6-7; 3 = age 8-9; 4 = age 10-11; 5 = age 12-13; 6 = age 14-15; 7 = age 16-17).
28
DEVELOPMENTAL CASCADES AND ADOLESCENT DEPRESSION
Supplemental Table 1
Significant pathways involving additional covariates from Model 4
Estimated Path β SE p
Hostile parenting Ext1 0.43 0.02 .000SLEs Ext1 0.35 0.02 .000
Chronic condition Int1 0.11 0.03 .000Hostile parenting Int1 0.31 0.02 .000
Maternal depression Int1 0.17 0.03 .000SLEs Int1 0.08 0.02 .000SES Aca1 0.22 0.03 .000
Family dysfunction Aca1 -0.09 0.02 .000Maternal depression Aca1 -0.07 0.03 .000
SES Ext2 -0.05 0.02 .000Hostile parenting Ext2 0.09 0.02 .000
Maternal depression Ext2 0.12 0.03 .000SLEs Ext2 0.05 0.02 .015
Maternal depression Int2 0.21 0.02 .000SLEs Int2 0.08 0.02 .000
Chronic condition Aca2 0.18 0.06 .002Hostile parenting Ext3 0.12 0.02 .000
Maternal depression Ext3 0.09 0.02 .000Hostile parenting Int3 0.07 0.03 .016
Maternal depression Int3 0.14 0.03 .000SLEs Int3 0.05 0.02 .003
Hostile parenting Ext4 0.10 0.03 .000Maternal depression Ext4 0.08 0.03 .019
Hostile parenting Int4 0.10 0.03 .000Maternal depression Int4 0.16 0.03 .000
SLEs Int4 0.09 0.04 .012SES Ext5 -0.04 0.02 .046
Maternal depression Ext5 0.07 0.03 .013SES Int5 -0.08 0.02 .000
SES Aca5 0.06 0.02 .006Maternal depression Ext6 0.06 0.03 .015Maternal depression Int6 0.13 0.03 .000Chronic condition Aca6 -0.09 0.02 .008
Maternal depression Dep7 0.09 0.03 .007
Ext = Externalising problems; Int = Internalising problems; Aca = Academic Competence; Dep = Depressive symptoms. Numbers denote time point (1= age 4-5; 2 = age 6-7; 3 = age 8-9; 4 = age 10-11; 5 = age 12-13; 6 = age 14-15; 7 = age 16-17).
29
Recommended