Mediating Factors in Within-Person Developmental Cascades ...

MEDIATORS IN DEVELOPMENTAL CASCADES1

Mediating Factors in Within-Person Developmental Cascades of

Externalising, Internalising and ADHD Symptoms in Childhood

Lydia Gabriela Speyera, Ingrid Obsuthb, Denis Ribeaudd, Manuel Eisnerc,d, Michelle

Lucianoa, Bonnie Auyeunga,e, and Aja Louise Murraya,c

Affiliations: aDepartment of Psychology, University of Edinburgh, Edinburgh, United

Kingdom; bClinical Psychology Department, University of Edinburgh, Edinburgh, United

Kingdom; cViolence Research Centre, Institute of Criminology, University of Cambridge,

Cambridge, United Kingdom; dJacobs Center for Productive Youth Development,

University of Zurich, Zurich, Switzerland; and eAutism Research Centre, Department ofPsychiatry, University of Cambridge, Cambridge, United Kingdom;

Address correspondence to: Lydia Gabriela Speyer, Department of Psychology, University

of Edinburgh, 7 George Square, Edinburgh, EH8 9JZ, United Kingdom, [[email protected],

00436506914172].

Conflicts of Interest: The authors have no conflicts of interest relevant to this article to disclose.

Acknowledgements: The authors are grateful to the children, parents and teachers who

provided data for the z-proso study and the research assistants involved in its collection.

mailto:[email protected]

mailto:[email protected]


Abstract

Previous studies have offered evidence for peer problems and academic achievement as

mediators in developmental cascades from externalising to internalising problems, and from

ADHD symptoms to internalising problems and externalising problems. However, these

mediators have not been found to fully account for these cascades indicating that there may

be additional mediators involved. This study investigated the role of parental involvement

alongside academic achievement and peer problems in mediating within-person

developmental cascades from externalising to internalising problems and from ADHD

symptoms to internalising and externalising problems using autoregressive latent trajectory

models with structured residuals. Models were fit for parent- and teacher-reports on

children’s psychosocial development as measured by the Social Behaviour Questionnaire

(SBQ) collected over ages 7, 9, and 11 in an ethnically diverse Swiss longitudinal cohort

study (zproso; N=1387, 51% male). Results indicated that none of the considered factors

acted as significant mediators in longitudinal relations between ADHD, internalising and

externalising problems; hence, a full picture of the mediating mechanisms in developmental

cascades remains to be identified.

Keywords: Developmental cascades; Internalising; Externalising; ADHD Symptoms; z-

proso;


According to the World Health Organisation (WHO, 2018) up to 20% of children and

adolescents globally suffer from a mental health problem and about 50% of all mental health

conditions have their onset by age 14. The leading mental health concerns among young

people in this age group are Attention deficit hyperactivity disorder (ADHD), behaviour (or

externalising) problems, anxiety and depression (or internalising problems) (e.g. Ghandour et

al., 2019). Importantly, these mental health conditions commonly co-occur (e.g. Reale et al.,

2017). Available longitudinal studies suggest that the negative consequences of experiencing

a number of childhood mental health problems are more serious than experiencing a single

condition (e.g. Biederman, 2004). Comorbidity also significantly impacts the prognosis,

treatment, and outcomes of therapeutic interventions with young people (Sibley et al., 2011).

For these reasons it is crucial to understand the processes through which different childhood

mental health problems are linked over time. To add to this understanding, in this study we

examined the roles of academic achievement, peer problems and parental involvement as

potential mechanisms of longitudinal links between ADHD, internalising and externalising

problems.

One theory that seeks to explain the high externalising-internalising longitudinal

comorbidity within the framework of developmental cascade models is the dual failure model

(Capaldi, 1992). Developmental cascade models hypothesise that links between mental health

domains are the result of problems in one domain causally affecting the other domain,

potentially through a number of linking mechanisms (Masten & Cicchetti, 2010). According

to the dual failure model, externalising problems lead to the development of internalising

problems through unidirectional cascades from externalising problems to educational

underachievement (academic failure) and social difficulties such as peer problems (social

failure) which in turn increase the risk for developing internalising problems such as anxiety

or depression (Capaldi, 1992).


While the focus on linking factors between developmental cascades from

externalising to internalising problems has been mainly on peer relationships and academic

achievement with a number of studies finding support for the dual failure model (e.g. van

Lier et al., 2012), these factors have not been found to fully explain externalising-

internalising cascades (e.g. Evans & Fite, 2019; Poirier et al., 2019). Given the complexity of

the development of mental health problems, it is highly likely that other factors play a

mediating role in these cascades. To date, few studies have investigated the role of parenting

behaviours as a potential mechanism for linking externalising-internalising cascades, even

though Capaldi’s (1992) paper originally hypothesised dual failure model explicitly included

parental relationships in the social domain. Considering that there is substantial evidence

highlighting the role of parenting behaviours for children’s socio-emotional development (for

a review see McKee, Colletti, Rakow, Jones, & Forehand, 2008) as well as the role of

externalising problems in eliciting worse parenting through a coercive cycle of negative

parent-child interaction (Patterson, 2002), parenting behaviours are a strong candidate for

linking the developmental relations of externalising and internalising problems.

One of the few studies that included a parental factor in a developmental cascade

model found that maternal dissatisfaction with their child was a significant mediator in a

developmental cascade from externalising problems at age 5 to internalising problems at age

12 (Wertz et al., 2015). A parental factor that could be particularly relevant in mediating

developmental cascades is parental involvement. A large body of research has highlighted

that parental involvement is not only associated with externalising problems but is also an

important factor in the development of internalising problems (Aboobaker, Jangam, Sagar,

Amaresha, & Jose, 2019; Hawes, Dadds, Frost, & Russell, 2013; Kirkhaug, Drugli, Klöckner,

& Mørch, 2013; Wang & Sheikh-Khalil, 2014). For instance, using the Alabama Parenting

Questionnaire which measures six domains of parenting behaviour, Aboobaker et al. (2019)


found that, out of all measured parenting domains, parental involvement was the only

parenting domain associated with emotional problems in a sample of 10 to 18 year olds.

Further research is needed to examine the role of parental involvement in the links between

externalizing and internalizing problems.

Internalising and externalising problems also commonly co-occur with ADHD, with

up to 50% of children suffering from ADHD symptoms also suffering from co-occurring

internalising and externalising problems (Gnanavel, Sharma, Kaushal, & Hussain, 2019). One

explanation of these comorbidities has been offered by Beauchaine and McNulty (2013)

through the ontogenic process model of externalising psychopathology which hypothesises

that early-onset ADHD symptoms lead to later externalising problems as a result of

longitudinal transaction between contextual risk factors and individual vulnerabilities (e.g.

genetic factors). The theory implies that ADHD symptoms might set up individuals for a

developmental trajectory characterised by increasingly escalating externalising behavioural

problems through coercive parent-child interactions (Theule, Wiener, Tannock, & Jenkins,

2013), poorer peer relationships (Diamantopoulou, Henricsson, & Rydell, 2005; Murray,

Zych, Ribeaud, & Eisner, 2020) and academic underachievement (Barry, Lyman, & Klinger,

2002). While some studies have found evidence for developmental cascades from ADHD

symptoms to externalising problems (e.g. Obsuth, Murray, Di Folco, Ribeaud, & Eisner,

2020), few studies to date have investigated potential linking factors in these relations (e.g.

Sevincok, Ozbay, Ozbek, Tunagur, & Aksu, 2020), hence, further research is needed.

Longitudinal investigations of the relations between ADHD and internalising

problems have also found support for developmental cascades from ADHD symptoms to

internalising problems with some also finding support for cascades in the opposite direction

(Murray, Caye, et al., 2020). Since ADHD has been found to be associated with similar risk


factors as externalising problems, a number of studies have also applied the dual failure

model to ADHD-internalising cascades, finding support for mediating effects of both peer

problems and academic attainment (Powell et al., 2020; Roy, Hartman, Veenstra, &

Oldehinkel, 2015). While the focal linking factors have also been peer relationships and

academic achievement, there has also been some evidence for parental relationships as a link

between ADHD symptoms and depression (Humphreys et al., 2013).

One limitation common to most studies on developmental cascades relates to their

statistical operationalisation. The most frequently used method to investigate developmental

cascades and potential mediating factors is the cross-lagged panel model (CLPM), which

examines the relations between variables over time after adjusting inter-individual stability in

those same variables. However, the CLPM does not separate between-person differences

from within-person relations, which means that its parameters represent a mixture of within-

and between-person effects that provide ambiguous results (Berry & Willoughby, 2017).

Considering that developmental cascades refer to within-person processes, more appropriate

methods are needed to confirm that the established relations between externalising problems

as well as ADHD symptoms and internalising problems still hold when appropriately

accounting for between-person differences. One modelling technique well-suited to overcome

this limitation of the CLPM, is the autoregressive latent trajectory model with structured

residuals (ALT-SR) (Curran, Howard, Bainter, Lane, & McGinley, 2014). ALT-SRs combine

the CLPM with a parallel process latent growth curve model that accounts for between-

person effects through allowing growth intercepts of different variables to covary. Thus, they

are well suited to disaggregate within- from between-person effects. Relatively few studies

have applied ALT-SRs to the study of developmental cascades (Murray, Eisner, & Ribeaud,

2020; Oh et al., 2020), and even fewer to the study of potential mediators (Murray et al.,

2021). The only study on linking factors using an ALT-SR so far has not provided strong


evidence for the dual failure model. Using data from the same cohort as the current study,

Murray et. al (2021) investigated the mediating role of peer- and teacher-relationships in

developmental cascades between self-reported internalising problems and aggression during

adolescence. Murray et al. did not identify any significant mediation effects nor did they find

support for developmental cascades between aggression and internalizing problems.

However, they only investigated these relations during a specific developmental period, that

is adolescence (ages 11, 13 and 15) and did not consider the role of parental involvement or

academic achievement as well as ADHD symptoms in these relations. Since previous

research has shown that the within-person developmental relations between internalising and

externalising problems unfold differently in childhood compared to adolescence (Murray,

Eisner, et al., 2020), it is highly important to also investigate mediators in the dual-failure

model during childhood.

In the current study, we aimed to extend the dual failure model by investigating

whether within-person developmental cascades from externalising problems to internalising

problems as well as cascades from ADHD symptoms to internalising and externalising

problems during childhood (median-ages 7, 9 and 11) were mediated by parental

involvement, peer problems (as indicated by negative social roles in the classroom) and

academic achievement. To overcome one of the main limitations of previous research—the

conflation of within- and between-person effects—we used an autoregressive latent trajectory

model with structured residuals which gives insights into the within-person dynamics of the

studied developmental relations. Since previous research has highlighted the need for multi-

informant approaches to the studies of children’s psychosocial development (De Los Reyes,

2013) we further built two ALT-SRs, one using parent-reported data on internalising,

externalising and ADHD symptoms and one using teacher-reported data to evaluate the

stability of our results across informants. This was deemed to be particularly important since


previous research using the ADHD data used in the current study already highlighted that

agreement on children’s ADHD symptom trajectories was only low to moderate across

parents and teachers (Murray, Booth, Ribeaud, & Eisner, 2018; Murray, Ribeaud, Eisner,

Murray, & McKenzie, 2019).

Methods

Participants

The participants of this study included 1387 (51% male) children who were part of the

Zurich Project on Social Development from Childhood to Adulthood (z-proso), a Swiss

longitudinal cohort study. Z-proso has been following the lives of 1675 children from 2004

when the children entered school at age 7 up until today with data collection still ongoing. To

date, there have been ten waves of data collection at ages 7, 8, 9, 10, 11, 12, 13, 15, 17 and

20. Participants attended one of 56 schools in Zurich that were chosen based on a stratified

random sampling procedure that included the school’s size and location as stratification

variables to ensure representativeness in terms of area-based deprivation. While the official

language of the canton of Zurich is German, Zurich is a very culturally diverse city, thus, in

order to recruit and retain as many of the non-German speaking participants as possible,

contact letters and parent interviews were translated into an additional nine languages.

Further details on assessment procedures, recruitment, retention, and attrition can be found in

the literature (M. Eisner & Ribeaud, 2005, 2007; N. L. Eisner, Murray, Eisner, & Ribeaud,

2019). Further details on the z-proso study can be found on the study website:

https://www.jacobscenter.uzh.ch/en/research/zproso/aboutus.html. This study uses data from

waves at median-ages 7, 9 and 11 since at these ages data on internalising, externalising and

ADHD symptoms as well as on peer problems, academic achievement and parental

involvement were available.


Measures

Psychosocial Development

Internalising problems, externalising problems and ADHD symptoms were measured

using the Social Behaviour Questionnaire (SBQ; Tremblay et al., 1991) which assesses

children’s psychosocial development across five domains: aggression, non-aggressive

externalizing problems, anxiety/depression, ADHD symptoms and prosocial behaviour. At

median-ages 7, 9 and 11, teachers completed a German paper-and-pencil version of the SBQ

while parents completed a computer-assisted personal interview. Most of the interviews were

conducted in German, but parents with other native-languages were able to complete the

interview in an additional nine languages (M. Eisner & Ribeaud, 2007). Items in the paper-

and-pencil version of the SBQ were presented in the same order to all teachers, whereas the

computer assisted interview presented items in a randomised order. Self-reported SBQs were

also available, but those were not used in the current study as they were collected in the form

of an adapted computer-based multimedia version of the SBQ with children answering ‘yes’

or ‘no’ to a series of questions relating to their psycho-social development. Internalising

problems were measured using four items on depression and three items on anxiety. All items

were rated on a 5-point Likert scale from Never to Very Often and subsequently averaged to

create a composite score (parent-report: Cronbach’s αage 7 = 0.71; αage 9 = 0.76; αage 11 = 0.79;

teacher-report: Cronbach’s αage 7 = 0.89; αage 9 = 0.91; αage 11 = 0.91). Higher scores indicate

more mental health problems. Composite scores for externalising problems were created by

averaging item scores of four items related to non-aggressive conduct problems, two items

measuring symptoms of oppositional defiant disorder, five items on physical aggression,

three items on proactive aggression and 3 items on reactive aggression, thus, reflecting the

multi-dimensional nature of externalising problems (parent-report: Cronbach’s αage 7 = 0.81;

αage 9 = 0.82; αage 11 = 0.84; teacher-report: Cronbach’s αage 7 = 0.94; αage 9 = 0.94; αage 11 = 0.94).


The ADHD symptom composite score was derived from four items on inattention and four

items on hyperactivity/impulsivity (parent-report: Cronbach’s αage 7 = 0.79; αage 9 = 0.84; αage 11

= 0.85; teacher-report: Cronbach’s αage 7 = 0.94; αage 9 = 0.95; αage 11 = 0.95). Psychometric

analyses of the SBQ in the current sample have found support for developmental invariance,

factorial and criterion validity and have shown that the SBQ is a reliable measure of

moderately low to very high levels of psychopathology in the general population (Murray,

Eisner, Obsuth, & Ribeaud, 2017; Murray, Eisner, & Ribeaud, 2019).

Mediating Factors

Academic achievement was measured using a composite score of arithmetic skills and

reading and verbal language performance scores which were provided by teachers who

compared the child’s abilities in both domains to an average student on a five-point Likert

scale from ‘worse’ to ‘better than average’. These scores were averaged to create a composite

score with higher scores indicating better academic performance (Cronbach’s αage 7 = 0.79;

αage 9 = 0.80; αage 11 = 0.83).

Peer problems were based on a measure of children’s social roles in the classroom.

Teachers were asked to rate the child’s role on four items relating to popularity, bullying

victimisation, rejection and domination on a 5-point Likert scale from ‘very untrue’ to ‘very

true’. Negative social roles, such as being a bully or a bullying victim, have previously been

identified to be associated with internalising as well as externalising problems (Kelly et al.,

2015). In addition, children with multiple negative social roles (e.g., being a bully and a

bullying victim) have been found to have more internalising and externalising problems than

children with just one negative role (being a bully or a bullying victim) (Kelly et al., 2015).

Before averaging all items to create a composite score, the item on popularity was reverse

coded in order to be consistent with the other items. Higher scores indicate that a child has


more negative social roles in the classroom, thus more peer problems (Cronbach’s αage 7 =

0.67; αage 9 = 0.73; αage 11 = 0.69).

Parental involvement was measured using the Alabama Parenting Questionnaire

(APQ) which measures parenting behaviours in six domains: parental involvement, positive

parenting, poor monitoring, inconsistent discipline, corporal punishment, and other discipline

(Shelton, Frick, & Wootton, 1996). Since past research has offered evidence that parental

involvement may be involved in both the development of internalising and externalising

problems (Aboobaker et al., 2019; Hawes et al., 2013; Kirkhaug et al., 2013; Wang &

Sheikh-Khalil, 2014), we here focus on the parental involvement subscale as a potential

mediator in our models. The involvement subscale includes ten items that parents rated on a

5-point Likert scale ranging from ‘never’ to ‘always’. Scores on all items were averaged to

create a composite score (Cronbach’s αage 7 = 0.63; αage 9 = 0.67; αage 11 = 0.72) with higher

scores being indicative of more parental involvement.

Statistical Analysis

Autoregressive Latent Trajectory Models with Structured Residuals (ALT-SR) were

fitted separately for parent- and teacher-reported data on children’s psychosocial

development to investigate the effects of academic achievement, peer problems and parental

involvement on developmental cascades from externalising problems and ADHD symptoms

to internalising problems. We decided on running separate models for each informant since

teacher- and parent-reports on children’s psychosocial functioning tend to show limited

convergence with cross-informant reports correlating at a low-to-moderate magnitude (De

Los Reyes et al., 2015). Additionally, combining an ALT-SR with a latent variable model

would have likely resulted in estimation difficulties. ALT-SRs combine cross-lagged panel

models with latent growth curve models to optimally account for between-person effects. The


latent growth curve part of the model included random intercepts that are allowed to co-vary

and fixed slopes. Slope factor loadings were fixed proportional to the time intervals between

measurement occasions. Gender was further regressed onto each intercept factor to adjust for

gender differences. The autoregressive and cross lagged effects are defined between the

variables’ residuals, reflecting deviations from the person-specific means and growth curves

at certain time points and, thus, offering insights into within-person dynamics. The cross-

lagged part of the model included all first-order autoregressive and cross-lagged effects as

well as within-time residual covariances. In addition, mediation effects were estimated for

cascades from externalising problems to internalising problems via peer problems, parental

involvement and academic achievement as well as for ADHD symptoms to internalising and

externalising problems via the same potential mediators. Second-order cross-lagged effects

between externalising at time 1 and internalising problems at time 3 as well as between

ADHD symptoms at time 1 and internalising and externalising at time 3 were further

included to allow for testing of longitudinal mediation. To assess statistical significance of

estimated indirect effects, standard errors were computed using the delta method. However,

since the delta method incorrectly assumes a symmetric sampling distribution of indirect

effects (MacKinnon, Warsi, & Dwyer, 1995), we also computed bootstrapped 95%

confidence intervals using standard maximum likelihood estimation.

Models were fit in Mplus 8.5 (Muthén & Muthén, 2018) using a robust maximum

likelihood estimator (MLR) which uses full information maximum likelihood (FIML) to

handle missing data (Enders, 2001). Model fit was judged to be acceptable based on the

following criteria: Tucker Lewis Index (TLI) >.90, Comparative Fit Index (CFI) >.90 and

Root Mean Squared Error of Approximation (RMSEA) <.05 (Kline, 2005). Mplus code and

full model results are provided on the Open Science Framework: https://osf.io/d7wfb/?

view_only=1e61e55ade1f4a53b18c6f376991ffbf.

https://osf.io/d7wfb/?view_only=1e61e55ade1f4a53b18c6f376991ffbf



Results

Descriptive statistics are provided in Table 1. For a correlation matrix of all observed

variables, see the online supplementary Table S1. Correlations between parent- and teacher-

reported variables on psychosocial development were relatively small, ranging from .11

to .22 for internalising problems, from .19 to .23 for externalising problems and from .36

to .42 for ADHD symptoms. Model fit indices for the ALT-SR using parent reported data

showed good fit with RMSEA = .038, CFI = .992 and TLI = .954. For a summary of

significant standardised autoregressive and cross-lagged parameters see Figure 1 and for full

results see the online supplementary Table S2.

[TABLE ONE HERE]

[FIGURE ONE HERE]

While internalising and externalising problems were stable across both lags, ADHD

symptoms only showed evidence for homotypic continuity across the second lag. In terms of

cross-lagged effects, ADHD symptoms at age 7 and at age 9 were found to be predictive of

internalising problems at age 9 and age 11 respectively. Age 9 internalising was also

predictive of age 11 ADHD, thus showing evidence for a reciprocal relation between ADHD

and internalising difficulties during late childhood. The model further identified a significant

cascade from ADHD symptoms at age 7 to externalising problems at age 9 and from age 9

externalising to age 11 ADHD. Results did not indicate any statistically significant cascades

from externalising to internalising problems. Analyses of indirect effects indicated that

neither cascades from externalising to internalising problem nor cascades from ADHD

symptoms to internalising or externalising problems were significantly mediated by peer

problems, academic achievement or parental involvement, see Table 2 (a). Their combined

mediating effect was also not statistically significant which was further confirmed using


bootstrapped confidence intervals (see: https://osf.io/d7wfb/?

view_only=1e61e55ade1f4a53b18c6f376991ffbf). In fact, none of the considered mediating

factors shared any statistically significant longitudinal relations with internalising problems.

However, all considered mediating factors did share at least some concurrent relations with

the considered domains of psychosocial functioning. See Table 3 (a) for residual correlations.

[TABLE TWO HERE]

[TABLE THREE HERE]

The teacher-reported ALT-SR also showed good fit (RMSEA = .050, CFI = .987, TLI

= .927). Significant standardised autoregressive and cross-lagged parameters visualised in

Figure 2 and presented in full in the online supplementary Table S3. In contrast to the parent-

reported model, internalising problems, externalising problems and ADHD symptoms all

showed homotypic continuity over the first lag but not over the second. The model also did

not identify any significant developmental cascades involving psychosocial difficulties. In

line with the parent-reported model, neither externalising-internalising cascades nor ADHD-

internalising or ADHD-externalising cascades were significantly mediated by peer problems,

academic achievement or parental involvement. Their combined mediating effect was also

not statistically significant based on bootstrapped confidence intervals, see Table 2 (b). For

full results see: https://osf.io/d7wfb/?view_only=1e61e55ade1f4a53b18c6f376991ffbf).

Residual correlations are summarised in Table 3 (b).

We also conducted additional exploratory analyses using a standard cross-lagged

panel model in order to provide insights into why our results may have differed from

previous studies. The full model results for both parent-and teacher-reported models can be

found here: https://osf.io/d7wfb/?view_only=1e61e55ade1f4a53b18c6f376991ffbf, The

results of the parent-reported CLPM model suggested that a cascade from ADHD symptoms






at age 7 to externalising problems at age 11 was significantly mediated by peer problems at

age 9. The teacher-reported model identified additional developmental cascades from

externalising problems at age 9 to internalising problems at age 11 and found support for

mediating effects of peer problems and academic achievement in cascades from ADHD to

internalising problems. These differences in results highlight the need for appropriately

disaggregating within- from between-person effects.

[FIGURE TWO HERE]

Discussion

The aim of the current study was to evaluate whether developmental cascades from

externalising to internalising problems and from ADHD symptoms to internalising and

externalising problems were mediated by parental involvement, peer problems and academic

underachievement when appropriately disaggregating within- from between-person effects.

Results indicated that none of these factors acted as significant mediators, thus, calling the

dual failure model into question (Capaldi, 1992). Additionally, we found only limited support

for developmental cascades between the studied domains of psychosocial functioning. While

the parent-reported ALT-SR suggested that there were developmental cascades from ADHD

symptoms to internalising problems across all studied time intervals as well as a

developmental cascade from internalising problems at age 9 to ADHD symptoms at age 11,

these patterns were not found in the model using teacher-reported data on internalising,

externalising and ADHD symptoms. Similarly, the only significant ADHD-externalising

developmental cascade identified from age 7 to age 9 in the parent-reported model was also

not significant in the teacher-reported model. However, concurrent relations of the different

domains of psychosocial functioning and the hypothesised mediating factors did show a


number of significant associations in both teacher- and parent-reported models, suggesting

that these factors may play a role in the risk for mental health problems.

While many previous studies have found evidence for the dual failure model (e.g. van

Lier et al., 2012), this is not the first study to report null-findings. For example, Poirier et al.

(2019) found no evidence for a mediating effect of academic skills on developmental

cascades from conduct problems to depressive symptoms. Similarly, in their study of 912 1st,

2nd and 3rd graders, Evans and Fite (2019) found only limited and mixed support for the dual

failure model in linking aggressive behaviour to depressive symptoms over time. These

authors concluded that the pathways via peer rejection and academic performance may be

observed in younger children but not in children at later elementary school age. The only

other study to date investigating mediating factors in externalising-internalising cascades also

using an ALT-SR which appropriately accounts for between-person effects (Curran et al.,

2014) also found no evidence for peer relationships or teacher relationships to mediate

developmental cascades from aggression to internalising problems (Murray et al., 2021).

Using data from the same cohort as the current study, but with a focus on adolescence (ages

11, 13, and 15), Murray et al. also did not identify any significant developmental cascades

from aggression to internalising problems. This could indicate that previously identified

developmental cascades and mediating relations are potentially not due to within-person

processes but could have been the result of between-person confounding. In fact, the

comparison of results with a standard CLPM in the current study also showed that when

conflating within- and between-person effects, some of the hypothesised mediators were

indeed identified as significant. Nevertheless, there has also been some research showing

evidence for externalising-internalising and ADHD-internalising cascades when

appropriately accounting for between-person effects (Flouri et al., 2019; Murray, Caye, et al.,

2020; Murray, Eisner, & Ribeaud, 2016; Oh et al., 2020). However, these studies account for


only a minority of studies and more research using appropriate statistical tools are needed to

evaluate developmental cascades and their potential mediators.

The current study offered some evidence for developmental cascades from ADHD

symptoms to internalising and externalising problems; however, these cascades were not

robust across informants. A number of previous studies have highlighted that teacher and

parent-reports are limited in their agreement on children’s psychosocial development (e.g.

Cheng et al., 2018) with cross-informant correlation having been shown to average .25 for

internalising problems, 30 for externalising problems (De Los Reyes, 2013) and range

from .09 to .52 for ADHD symptoms (Narad et al., 2015). Cross-informant correlations in the

present study showed a pattern in line with these findings (.11 to .22 for internalising

problems,.19 to .23 for externalising problems, .36 to .42 for ADHD symptoms). The large

inconsistencies between the teacher- and parent-reported ALT-SR reported here further

highlight the need for future research to include multiple informants when investigating the

development of mental health problems in children. One potential reason for the discrepancy

found in the current study could be that parents and teachers observe children in different

context with some difficulties likely being tied to specific environments. For instance,

previous research has indicated that inattentive ADHD symptoms are primarily noticeable in

the school context (Murray, Ribeaud, et al., 2019). Another reason for the observed

discrepancy could relate to the fact the questionnaire items in the parent-reported SBQ were

presented in a randomised order whereas they were presented in the same order to all

teachers. This would have addressed potential item position effects in the parents but not

teachers.

One of the developmental relations identified consistently across both time intervals

in both teacher- and parent-reported models was cascades from academic achievement to

parental involvement with higher academic achievement being predictive of less parental


involvement at later time points. We also identified a significant negative developmental

cascade from parental involvement to academic achievement across both models, suggesting

that these two domains share reciprocal relations in later childhood. While a large body of

literature has investigated cascades from parental involvement to academic achievement,

finding evidence for negative, positive and null associations (for a review see Boonk,

Gijselaers, Ritzen, & Brand-Gruwel, 2018), cascades in the other direction have not been

extensively investigated yet. The results of the current study suggest that children who

performed well in school, might need less parental support later as is reflected in less parental

involvement. These findings need to be replicated in future studies.

Even though we found little evidence for longitudinal effects of peer problems,

academic achievement and parental involvement on childhood mental health problems, we

identified significant concurrent relations between all domains of psychosocial functioning

and these factors at multiple time points. This indicates that mental health problems are often

accompanied by peer problems, academic underachievement and a lack of parental

involvement; however, any directional effects between these variables are potentially at a

shorter timescale then it was possible to evaluate in the current study.

Strengths and Limitations

The key strength of this study is that we used a method suitable of overcoming the

main limitation of previous studies investigating the dual failure model, that is an ALT-SR,

which allowed us to clearly separate within- from between-person effects (Curran et al.,

2014). In addition, while most previous studies focused on one focal mediator and two

domains of psychosocial functioning, we jointly modelled internalising, externalising and

ADHD symptom cascades and potential mediators, thus, capturing the most commonly

occurring and co-occurring conditions in childhood within one comprehensive model. We


were also able to investigate the studied cascades using both parent- and teacher-reported

data on children’s psychosocial development, hence, providing a more complete picture of

their development than when only using a single-informant approach.

However, the current study also has to be interpreted in light of some limitations.

First, the included measures on peer problems and academic achievement were relatively

brief. In particular, the measure on academic achievement consisted of only two items that

were based on a teacher comparing students’ performance on mathematics and language

skills to other students. Consequently, it does not provide an objective indication of a child’s

academic achievement. Ideally, future studies should replicate the current study using

standardised assessments of children’s educational attainment. In addition, it is possible that

the lack of identified longitudinal relations between peer problems and internalising

difficulties is due to the fact that only specific aspects of peer problems are related to the

development of internalising difficulties (e.g. bullying victimisation) (Roy et al., 2015).

However, the measure included in the current study was only a very general measure of peer

problems, focusing on whether the child had any negative social roles in the classroom, and

did not allow for detailed investigations of individual factors. Furthermore, we only

investigated one specific aspect of parenting behaviours, that is parental involvement. Other

dimensions of parenting behaviours, such as harsh parenting, could also be a potential linking

factor in the development of mental health problems and should therefore be investigated in

future research. Also, at some of the included time points, the measures of peer problems and

parental involvement had slightly lower reliability than desired as indicated by Cronbach’s

alphas below .70 which could have impacted our results. Finally, another main limitation of

the current study is that we used time intervals of two years to investigate developmental

cascades. As the results of concurrent relations have shown, most domains were related

within the same time point, thus, shorter time lags could have potentially provided different


results as these cascades might act over shorter periods (M. P. Eisner & Malti, 2015). Future

studies should address this limitation by employing shorter time intervals to the study of

mediating factors in developmental cascades of mental health problems (Dormann & Griffin,

2015). In particular, experience sampling methods could be a useful approach to gain insights

into the short term relations between internalising problems, externalising problems, ADHD

symptoms and social and academic failure (Verhagen, Hasmi, Drukker, van Os, & Delespaul,

2016). This could provide valuable clarification on which difficulties act as antecedents to

other difficulties and consequently inform interventions. Finally, while ALT-SRs address

between-person confounds that tend to be stable over time they are vulnerable to time-

varying confounding. Future evaluations that can account for these and other complementary

work using, for example, causal inference frameworks such as mendelian randomisation or

matching-based methods could provide further insights into which mental health and

psychosocial issues have a causal impact on others during development.

Conclusion

In conclusion, findings of the current study do not provide support for the dual failure

model, neither for the original hypothesised mediating role of peer problems and academic

underachievement nor for parental involvement as an additional potential mediator. Since this

is one of only few studies appropriately disentangling within- from between-person effects

when investigating mediating factors in developmental cascade models, further research is

needed to re-evaluate the empirical evidence for the dual failure model. The results of the

current study do however highlight that peer problems, academic underachievement and lack

of parental involvement are associated with worse psychosocial development within the same

time points, thus these factors might still represent valuable intervention targets in children

suffering from psychosocial difficulties.


References

Aboobaker, S., Jangam, K. V., Sagar, K. J. V., Amaresha, A. C., & Jose, A. (2019).

Predictors of emotional and behavioral problems among Indian adolescents: A clinic-

based study. Asian Journal of Psychiatry, 39, 104–109.

https://doi.org/10.1016/j.ajp.2018.12.002

Barry, T. D. S., Lyman, R. D., & Klinger, L. G. (2002). Academic underachievement and

Attention-Deficit/Hyperactivity Disorder: The negative impact of symptom severity on

school performance. Journal of School Psychology, 40(3), 259–283.

https://doi.org/10.1016/S0022-4405(02)00100-0

Beauchaine, T. P., & McNulty, T. (2013). Comorbidities and continuities as ontogenic

processes: Toward a developmental spectrum model of externalizing psychopathology.

Development and Psychopathology, 25(4 PART 2), 1505–1528. https://doi.org/10.1017/

S0954579413000746

Berry, D., & Willoughby, M. T. (2017, July 1). On the Practical Interpretability of Cross-

Lagged Panel Models: Rethinking a Developmental Workhorse. Child Development,

Vol. 88, pp. 1186–1206. Blackwell Publishing Inc. https://doi.org/10.1111/cdev.12660

Biederman, J. (2004). Impact of Comorbidity in Adults With Attention-Deficit/Hyperactivity

Disorder. The Journal of Clinical Psychiatry, 65(suppl 3), 3–7.

Boonk, L., Gijselaers, H. J. M., Ritzen, H., & Brand-Gruwel, S. (2018). A review of the

relationship between parental involvement indicators and academic achievement.

Educational Research Review, 24, 10–30. https://doi.org/10.1016/j.edurev.2018.02.001

Capaldi, D. M. (1992). Co-occurrence of conduct problems and depressive symptoms in early

adolescent boys: II. A 2-year follow-up at Grade 8. Development and Psychopathology,

4(1), 125–144. https://doi.org/10.1017/S0954579400005605

Cheng, S., Keyes, K. M., Bitfoi, A., Carta, M. G., Koç, C., Goelitz, D., … Kovess-Masfety,


V. (2018). Understanding parent–teacher agreement of the Strengths and Difficulties

Questionnaire (SDQ): Comparison across seven European countries. International

Journal of Methods in Psychiatric Research, 27(1). https://doi.org/10.1002/mpr.1589

Curran, P. J., Howard, A. L., Bainter, S. A., Lane, S. T., & McGinley, J. S. (2014). The

separation of between-person and within-person components of individual change over

time: A latent curve model with structured residuals. Journal of Consulting and Clinical

Psychology, 82(5), 879–894. https://doi.org/10.1037/a0035297

De Los Reyes, A. (2013). Strategic objectives for improving understanding of informant

discrepancies in developmental psychopathology research. Development and

Psychopathology, 25(3), 669–682. https://doi.org/10.1017/S0954579413000096

De Los Reyes, A., Augenstein, T. M., Wang, M., Thomas, S. A., Drabick, D. A. G., Burgers,

D. E., & Rabinowitz, J. (2015). The validity of the multi-informant approach to

assessing child and adolescent mental health. Psychological Bulletin, 141(4), 858–900.

https://doi.org/10.1037/a0038498

Diamantopoulou, S., Henricsson, L., & Rydell, A.-M. (2005). ADHD symptoms and peer

relations of children in a community sample: Examining associated problems, self-

perceptions, and gender differences. International Journal of Behavioral Development,

29(5), 388–398. https://doi.org/10.1177/01650250500172756

Dormann, C., & Griffin, M. A. (2015). Optimal time lags in panel studies. Psychological

Methods, 20(4), 489–505. https://doi.org/10.1037/met0000041

Eisner, M. P., & Malti, T. (2015). Aggressive and Violent Behavior. In Handbook of Child

Psychology and Developmental Science (pp. 1–48). John Wiley & Sons, Inc.

https://doi.org/10.1002/9781118963418.childpsy319

Eisner, M., & Ribeaud, D. (2005). A Randomised Field Experiment to Prevent Violence -

The Zurich Intervention and Prevention Project at Schools, ZIPPS. European Journal of


Crime, Criminal Law and Criminal Justice, 13. Retrieved from

https://heinonline.org/HOL/Page?handle=hein.journals/eccc13&id=31&div=&collection

=

Eisner, M., & Ribeaud, D. (2007). Conducting a Criminological Survey in a Culturally

Diverse Context. European Journal of Criminology, 4(3), 271–298.

https://doi.org/10.1177/1477370807077183

Eisner, N. L., Murray, A. L., Eisner, M., & Ribeaud, D. (2019). A practical guide to the

analysis of non-response and attrition in longitudinal research using a real data example.

International Journal of Behavioral Development, 43(1), 24–34.

https://doi.org/10.1177/0165025418797004

Enders, C. K. (2001). The performance of the full information maximum likelihood estimator

in multiple regression models with missing data. Educational and Psychological

Measurement, 61(5), 713–740. https://doi.org/10.1177/00131640121971482

Evans, S. C., & Fite, P. J. (2019). Dual Pathways from Reactive Aggression to Depressive

Symptoms in Children: Further Examination of the Failure Model. Journal of Abnormal

Child Psychology, 47(1), 85–97. https://doi.org/10.1007/s10802-018-0426-6

Flouri, E., Papachristou, E., Midouhas, E., Ploubidis, G. B., Lewis, G., & Joshi, H. (2019).

Developmental cascades of internalising symptoms, externalising problems and

cognitive ability from early childhood to middle adolescence. European Psychiatry, 57,

61–69. https://doi.org/10.1016/j.eurpsy.2018.12.005

Ghandour, R. M., Sherman, L. J., Vladutiu, C. J., Ali, M. M., Lynch, S. E., Bitsko, R. H., &

Blumberg, S. J. (2019). Prevalence and Treatment of Depression, Anxiety, and Conduct

Problems in US Children. Journal of Pediatrics, 206, 256-267.e3.

https://doi.org/10.1016/j.jpeds.2018.09.021

Gnanavel, S., Sharma, P., Kaushal, P., & Hussain, S. (2019, September 1). Attention deficit


hyperactivity disorder and comorbidity: A review of literature. World Journal of

Clinical Cases, Vol. 7, pp. 2420–2426. Baishideng Publishing Group Co. https://doi.org/

10.12998/wjcc.v7.i17.2420

Hawes, D. J., Dadds, M. R., Frost, A. D. J., & Russell, A. (2013). Parenting practices and

prospective levels of hyperactivity/inattention across early- and middle-childhood.

Journal of Psychopathology and Behavioral Assessment, 35(3), 273–282. https://doi.org/

10.1007/s10862-013-9341-x

Humphreys, K. L., Katz, S. J., Lee, S. S., Hammen, C., Brennan, P. A., & Najman, J. M.

(2013). The association of ADHD and depression: Mediation by peer problems and

parent-child difficulties in two complementary samples. Journal of Abnormal

Psychology, 122(3), 854–867. https://doi.org/10.1037/a0033895

Kelly, E. V., Newton, N. C., Stapinski, L. A., Slade, T., Barrett, E. L., Conrod, P. J., &

Teesson, M. (2015). Suicidality, internalizing problems and externalizing problems

among adolescent bullies, victims and bully-victims. Preventive Medicine, 73, 100–105.

https://doi.org/10.1016/j.ypmed.2015.01.020

Kirkhaug, B., Drugli, M. B., Klöckner, C. A., & Mørch, W.-T. (2013). Association between

parental involvement in school and child conduct, social, and internalizing problems:

teacher report. Educational Research and Evaluation, 19(4), 346–361.

https://doi.org/10.1080/13803611.2013.771893

Kline, R. B. (2005). Principles and Practice of Structural Equation Modeling, Second

Edition. Guilford Publications. Retrieved from https://books.google.at/books?

id=EkMVZUxZrgIC

MacKinnon, D. P., Warsi, G., & Dwyer, J. H. (1995). A Simulation Study of Mediated Effect

Measures. Multivariate Behavioral Research, 30(1), 41–62.

https://doi.org/10.1207/s15327906mbr3001_3


Masten, A. S., & Cicchetti, D. (2010, August). Developmental cascades. Development and

Psychopathology, Vol. 22, pp. 491–495. Dev Psychopathol.

https://doi.org/10.1017/S0954579410000222

McKee, L., Colletti, C., Rakow, A., Jones, D. J., & Forehand, R. (2008, June). Parenting and

child externalizing behaviors: Are the associations specific or diffuse? Aggression and

Violent Behavior, Vol. 13, pp. 201–215. NIH Public Access.

https://doi.org/10.1016/j.avb.2008.03.005

Murray, A. L., Booth, T., Ribeaud, D., & Eisner, M. (2018). Disagreeing about development:

An analysis of parent teacher agreement in ADHD symptom trajectories across the ‐

elementary school years. International Journal of Methods in Psychiatric Research,

27(3). https://doi.org/10.1002/mpr.1723

Murray, A. L., Caye, A., McKenzie, K., Auyeung, B., Murray, G., Ribeaud, D., … Eisner, M.

(2020). Reciprocal Developmental Relations Between ADHD and Anxiety in

Adolescence: A Within-Person Longitudinal Analysis of Commonly Co-Occurring

Symptoms. Journal of Attention Disorders, 108705472090833.

https://doi.org/10.1177/1087054720908333

Murray, A. L., Eisner, M., Obsuth, I., & Ribeaud, D. (2017). Situating violent ideations

within the landscape of mental health: Associations between violent ideations and

dimensions of mental health. Psychiatry Research, 249, 70–77.

https://doi.org/10.1016/j.psychres.2017.01.005

Murray, A. L., Eisner, M., & Ribeaud, D. (2016). The Development of the General Factor of

Psychopathology ‘p Factor’ Through Childhood and Adolescence. Journal of Abnormal

Child Psychology, 44(8), 1573–1586. https://doi.org/10.1007/s10802-016-0132-1

Murray, A. L., Eisner, M., & Ribeaud, D. (2019). Can the social behavior questionnaire help

meet the need for dimensional, transdiagnostic measures of childhood and adolescent


psychopathology? European Journal of Psychological Assessment, 35(5), 674–679.

https://doi.org/10.1027/1015-5759/a000442

Murray, A. L., Eisner, M., & Ribeaud, D. (2020). Within person analysis of developmental ‐

cascades between externalising and internalising problems. Journal of Child Psychology

and Psychiatry, 61(6), 681–688. https://doi.org/10.1111/jcpp.13150

Murray, A. L., Obsuth, I., Speyer, L., Murray, G., McKenzie, K., Eisner, M., & Ribeaud, D.

(2021). Developmental Cascades from Aggression to Internalizing Problems via Peer

and Teacher Relationships from Early to Middle Adolescence. Journal of Youth and

Adolescence. https://doi.org/10.1007/s10964-021-01396-1

Murray, A. L., Ribeaud, D., Eisner, M., Murray, G., & McKenzie, K. (2019). Should We

Subtype ADHD According to the Context in Which Symptoms Occur? Criterion

Validity of Recognising Context-Based ADHD Presentations. Child Psychiatry and

Human Development, 50(2), 308–320. https://doi.org/10.1007/s10578-018-0842-4

Murray, A. L., Zych, I., Ribeaud, D., & Eisner, M. (2020). Developmental relations between

ADHD symptoms and bullying perpetration and victimization in adolescence.

Aggressive Behavior. https://doi.org/10.1002/ab.21930

Muthén, L. K., & Muthén, B. (2018). Mplus. The Comprehensive Modelling Program for

Applied Researchers: User’s Guide, 5.

Narad, M. E., Garner, A. A., Peugh, J. L., Tamm, L., Antonini, T. N., Kingery, K. M., …

Epstein, J. N. (2015). Parent-teacher agreement on ADHD symptoms across

development. Psychological Assessment, 27(1), 239–248.

https://doi.org/10.1037/a0037864

Obsuth, I., Murray, A. L., Di Folco, S., Ribeaud, D., & Eisner, M. (2020). Patterns of

Homotypic and Heterotypic Continuity Between ADHD Symptoms, Externalising and

Internalising Problems from Age 7 to 15. Journal of Abnormal Child Psychology, 48(2),


223–236. https://doi.org/10.1007/s10802-019-00592-9

Oh, Y., Greenberg, M. T., Willoughby, M. T., Vernon-Feagans, L., Blair, C. B., Burchinal,

M. R., … Mills-Koonce, W. R. (2020). Examining Longitudinal Associations between

Externalizing and Internalizing Behavior Problems at Within- and Between-Child

Levels. Journal of Abnormal Child Psychology, 48(4), 467–480. https://doi.org/10.1007/

s10802-019-00614-6

Patterson, G. R. (2002). The early development of coercive family process. In Antisocial

behavior in children and adolescents: A developmental analysis and model for

intervention. (pp. 25–44). American Psychological Association.

https://doi.org/10.1037/10468-002

Poirier, M., Temcheff, C. E., Déry, M., Toupin, J., Verlaan, P., & Lemelin, J.-P. (2019). The

Role of Academic Skills in the Evolution of Conduct Problems and Depressive

Symptoms Among Children With and Without Early Clinically Significant Conduct

Problems. The Journal of Early Adolescence, 39(3), 340–370.

https://doi.org/10.1177/0272431618757679

Powell, V., Riglin, L., Hammerton, G., Eyre, O., Martin, J., Anney, R., … Rice, F. (2020).

What explains the link between childhood ADHD and adolescent depression?

Investigating the role of peer relationships and academic attainment. European Child

and Adolescent Psychiatry, 29(11), 1581–1591. https://doi.org/10.1007/s00787-019-

01463-w

Reale, L., Bartoli, B., Cartabia, M., Zanetti, M., Costantino, M. A., Canevini, M. P., … Rossi,

G. (2017). Comorbidity prevalence and treatment outcome in children and adolescents

with ADHD. European Child and Adolescent Psychiatry, 26(12), 1443–1457.

https://doi.org/10.1007/s00787-017-1005-z

Roy, A., Hartman, C. A., Veenstra, R., & Oldehinkel, A. J. (2015). Peer dislike and


victimisation in pathways from ADHD symptoms to depression. European Child and

Adolescent Psychiatry, 24(8), 887–895. https://doi.org/10.1007/s00787-014-0633-9

Sevincok, D., Ozbay, H. C., Ozbek, M. M., Tunagur, M. T., & Aksu, H. (2020). ADHD

symptoms in relation to internalizing and externalizing symptoms in children: the

mediating role of sluggish cognitive tempo. Nordic Journal of Psychiatry, 74(4), 265–

272. https://doi.org/10.1080/08039488.2019.1697746

Shelton, K. K., Frick, P. J., & Wootton, J. (1996). Assessment of parenting practices in

families of elementary school-age children. Journal of Clinical Child Psychology, 25(3),

317–329. https://doi.org/10.1207/s15374424jccp2503_8

Sibley, M. H., Pelham, W. E., Molina, B. S. G., Gnagy, E. M., Waschbusch, D. A., Biswas,

A., … Karch, K. M. (2011). The delinquency outcomes of boys with ADHD with and

without comorbidity. Journal of Abnormal Child Psychology, 39(1), 21–32.

https://doi.org/10.1007/s10802-010-9443-9

Theule, J., Wiener, J., Tannock, R., & Jenkins, J. M. (2013). Parenting Stress in Families of

Children With ADHD. Journal of Emotional and Behavioral Disorders, 21(1), 3–17.

https://doi.org/10.1177/1063426610387433

Tremblay, R. E., Loeber, R., Gagnon, C., Charlebois, P., Larivée, S., & LeBlanc, M. (1991).

Disruptive boys with stable and unstable high fighting behavior patterns during junior

elementary school. Journal of Abnormal Child Psychology, 19(3), 285–300.

https://doi.org/10.1007/BF00911232

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 problems. Child Development, 83(5), 1775–1788. https://

doi.org/10.1111/j.1467-8624.2012.01802.x

Verhagen, S. J. W., Hasmi, L., Drukker, M., van Os, J., & Delespaul, P. A. E. G. (2016). Use


of the experience sampling method in the context of clinical trials. Evidence-Based

Mental Health, 19(3), 86–89. https://doi.org/10.1136/ebmental-2016-102418

Wang, M.-T., & Sheikh-Khalil, S. (2014). Does Parental Involvement Matter for Student

Achievement and Mental Health in High School? Child Development, 85(2), 610–625.

https://doi.org/10.1111/cdev.12153

Wertz, J., Zavos, H., Matthews, T., Harvey, K., Hunt, A., Pariante, C. M., & Arseneault, L.

(2015). Why some children with externalising problems develop internalising

symptoms: Testing two pathways in a genetically sensitive cohort study. Journal of

Child Psychology and Psychiatry and Allied Disciplines, 56(7), 738–746. https://doi.org/

10.1111/jcpp.12333

WHO. (2018). WHO | Child and adolescent mental health. WHO. Retrieved from

http://www.who.int/mental_health/maternal-child/child_adolescent/en/


Table 1. Descriptive statistics

N Mea

n

SD Min Max Rang

e

Ske

w

Kurtosis

Age 7 internalising (teacher) 1302 1.86 0.76 1.00 5.00 4.00 0.99 0.72



Age 7 internalising (parent) 1201 1.61 0.49 1.00 3.86 2.86 1.08 1.36



Age 7 ADHD symptoms (teacher) 1299 2.24 0.99 1.00 5.00 4.00 0.61 -0.43

Age 9 ADHD symptoms (teacher) 1281 2.06 0.95 1.00 5.00 4.00 0.84 0.12

Age 11 ADHD symptoms

(teacher)

1053 2.07 0.99 1.00 5.00 4.00 0.84 -0.09

Age 7 ADHD symptoms (parent) 1186 2.20 0.67 1.00 4.75 3.75 0.48 0.04



Age 7 externalising (teacher) 1263 1.48 0.57 1.00 4.35 3.35 1.73 3.02



Age 7 externalising (parent) 1087 1.58 0.37 1.00 3.71 2.71 0.87 1.61



Age 7 peer problems 1321 1.68 0.62 1.00 4.00 3.00 1.20 1.16



Age 7 academic achievement 1338 3.23 1.01 1.00 5.00 4.00 -0.15 -0.49



Age 7 parental involvement 1229 3.19 0.42 1.50 4.00 2.50 -0.46 0.12



Table 2. Indirect effects

(a) Parent-reported ALT-SR β SE p CIlower CIupper


Age 7 externalising to age 11 internalising

via age 9 academic achievement 0.007 0.010 .278 -0.008 0.038

via age 9 peer problems 0.006 0.009 .464 -0.009 0.030

via age 9 parental involvement -0.000 0.005 .923 -0.021 0.014

sum of indirect effect 0.013 0.014 .365 -0.018 0.059

Age 7 ADHD to age 11 internalising


via age 9 peer problems -0.002 0.004 .620 -0.014 0.007

via age 9 parental involvement 0.000 0.001 .958 -0.012 0.014


Age 7 ADHD to age 11 externalising





(b) Teacher-reported ALT-SR β SE p CIlower CIupper

Age 7 externalising to age 11 internalising

via age 9 academic achievement -0.000 0.004 .771 -0.014 0.011




Age 7 ADHD to age 11 internalising



via age 9 parental involvement -0.012 0.011 .274 -0.038 0.009

sum of indirect effect -0.007 0.014 .620 -0.040 0.024

Age 7 ADHD to age 11 externalising





Note. Confidence Intervals (CIs) are based on 1000 bootstrapped samples using standard

maximum likelihood estimation

MEDIATORS IN DEVELOPMENTAL CASCADES 32

Table 3. Residual correlations

(a) Parent-reported ALT-SR

Age 7 Age 9 Age 11

1. 2. 3. 4. 5. 6. 1. 2. 3. 4. 5. 6. 1. 2. 3. 4. 5. 6.

1. Internalising problems - - -2. Externalising problems .03 - .32* - .32* -3. ADHD symptoms .56* .09 - .50* .40* - .48* .29* -4. Peer problems -.10 -.09 -.10 - -.05 -.06 -.04 - .10* .13* .04 -5. Academic achievement .02 .05 .08 -.10 - .03 .22* .05 -.07 - -.01 -.12* -.22* -.21* -6. Parental involvement -.07 .03 -.11 .07 .01 - -.21* -.07 -.18* .13* -.36* - -.09* -.22* -.06 .01 -.06 -

(a) Teacher-reported ALT-SR

1. Internalising problems - - -2. Externalising problems .37* - .43* - .36* -3. ADHD symptoms .32* .46* - .44* .53* - .33* .45* -4. Peer problems .35* .54* .34* - .34* .45* .32* - .47* .45* .31* -5. Academic achievement -.08 -.07 -.17* -.13* - -.16* -.06 -.10 -.11 - -.28* -.24* -.40* -.19* -6. Parental involvement -.03 .04 .06 .07 -.02 - .10 .13 .15* .13* -.42* - -.09* -.03 -.01 .01 -.06 -

Note. *Significant at p<.05

Fig. 1. Standardised autoregressive and cross-lagged parameters from ALT-SR using parent-reported data. Int = internalising,

Academic = academic achievement, Peer = peer problems, Involvement = parental involvement, ADHD = ADHD symtomps, Ext =

externalising. Only statistically significant paths are shown. Latent growth curve and covariance parameters are omited for clarity.

33MEDIATORS IN DEVELOPMENTAL CASCADES

Fig. 2. Standardised autoregressive and cross-lagged parameters from ALT-SR using teacher-reported data. Int = internalising,

Academic = academic achievement, Peer = peer problems, Involvement = parental involvement, ADHD = ADHD symtomps, Ext=

externalising. Only statistically significant paths are shown. Latent growth curve and covariance parameters are omited for clarity.



Supplementary Materials

Table S1. Correlation matrix of observed variables1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

1. Age 7 internalising (teacher)2. Age 9 internalising (teacher) .503. Age 11 internalising (teacher) .24 .264. Age 7 internalising (parent) .11 .03 .045. Age 9 internalising (parent) .07 .12 .12 .536. Age 11 internalising (parent) .06 .10 .22 .43 .557. Age 7 ADHD (teacher) .40 .28 .20 .05 .07 .128. Age 9 ADHD (teacher) .26 .45 .13 -.03 .02 .06 .659. Age 11 ADHD (teacher) .21 .22 .37 -.00 .09 .12 .53 .5310. Age 7 ADHD (parent) .11 .09 .15 .38 .29 .28 .36 .27 .3111. Age 9 ADHD (parent) .08 .12 .18 .26 .39 .33 .33 .34 .36 .6512. Age 11 ADHD (parent) .09 .17 .18 .18 .29 .45 .34 .36 .42 .61 .6913. Age 7 externalising (teacher) .37 .28 .12 .05 .02 .03 .63 .49 .41 .26 .19 .2314. Age 9 externalising (teacher) .24 .41 .08 -.03 .01 .06 .45 .69 .37 .16 .21 .24 .5915. Age 11 externalising (teacher) .06 .10 .33 -.02 .02 .02 .31 .30 .59 .14 .14 .18 .39 .3316. Age 7 externalising (parent) -.01 -.0

2.02 .37 .24 .26 .15 .16 .13 .47 .40 .35 .19 .14 .15

17. Age 9 externalising (parent) -.01 .05 .11 .25 .35 .30 .15 .18 .15 .37 .53 .39 .12 .17 .16 .6718. Age 11 externalising (parent) .01 .08 .11 .22 .27 .38 .16 .19 .20 .35 .44 .47 .15 .21 .23 .58 .7519. Age 7 peer problems .37 .25 .17 .08 .05 .09 .41 .29 .25 .19 .14 .20 .57 .36 .21 .07 .05 .1220. Age 9 peer problems .27 .46 .18 .01 .11 .10 .37 .42 .30 .19 .16 .19 .38 .52 .19 .02 .08 .12 .4521. Age 11 peer problems .17 .18 .54 .07 .07 .15 .23 .20 .39 .25 .15 .21 .26 .22 .55 .16 .14 .21 .20 .2322. Age 7 academic achievement -.21 -.1

3-.11 -.07 -.0

5-.10 -.3

7-.24 -.26 -.1

5-.17 -.18 -.1

7-.11 -.04 .05 .09 .03 -.2

6-.19 -.0

923. Age 9 academic achievement -.16 -.2

3-.15 -.02 -.0

9-.09 -.2

6-.28 -.26 -.0

9-.14 -.14 -.0

8-.10 .02 .10 .06 .06 -.0

7-.19 -.1

1.56

24. Age 11 academic achievement

-.17 -.15

-.28 .02 -.03

-.07 -.32

-.28 -.42 -.14

-.17 -.24 -.17

-.14 -.11 .10 .09 .05 -.16

-.18 -.17

.58 .59


25. Age 7 parental involvement -.07 -.06

.02 -.13 -.13

-.09 -.07

-.09 -.06 -.10

-.08 -.07 -.02

-.07 -.03 -.19

-.13 -.14 .03 .01 -.02

.05 .03 .02

26. Age 9 parental involvement -.02 .02 -.02 -.05 -.17

-.08 .04 -.00 -.06 -.06

-.10 -.06 .00 -.00 -.05 -.24

-.22 -.16 .04 .02 -.02

-.08 -.06 -.09

.61

27. Age 11 parental involvement -.02 .00 -.03 -.09 -.18

-.16 .01 .01 -.03 -.13

-.18 -.14 .01 -.03 -.00 -.25

-.25 -.28 -.02

.04 -.01

-.02 -.02 -.08

.54 .65

Note. Correlations in bold are significant at p < .05


Table S2. Standardised autoregressive and cross-lagged parameters for the parent-reported ALT-SR

Parameter Estimate SE p

Age 11 internalising on age 9 internalising 0.303 0.067 <.001*Age 11 internalising on age 9 externalising -0.051 0.066 .438Age 11 internalising on age 9 ADHD 0.234 0.073 .001*Age 11 internalising on age 9 peer problems -0.032 0.041 .429Age 11 internalising on age 9 academic achievement 0.040 0.049 .411Age 11 internalising on age 9 parental involvement 0.003 0.050 .948Age 11 internalising on age 7 externalising -0.028 0.060 .636Age 11 internalising on age 7 ADHD 0.141 0.074 .055Age 11 externalising on age 9 internalising 0.016 0.081 .842Age 11 externalising on age 9 externalising 0.354 0.084 <.001*Age 11 externalising on age 9 ADHD -0.007 0.084 .930Age 11 externalising on age 9 peer problems 0.039 0.050 .440Age 11 externalising on age 9 academic achievement 0.108 0.057 .057Age 11 externalising on age 9 parental involvement 0.011 0.066 .863Age 11 externalising on age 7 ADHD -0.169 0.087 .053Age 11 ADHD on age 9 internalising 0.241 0.084 .004*Age 11 ADHD on age 9 externalising -0.172 0.063 .007*Age 11 ADHD on age 9 ADHD 0.325 0.086 <.001*Age 11 ADHD on age 9 peer problems 0.000 0.049 .994Age 11 ADHD on age 9 academic achievement 0.004 0.050 .942Age 11 ADHD on age 9 parental involvement -0.011 0.061 .861Age 11 peer problems on age 9 internalising 0.005 0.073 .949Age 11 peer problems on age 9 externalising -0.010 0.072 .892Age 11 peer problems on age 9 ADHD -0.049 0.070 .484Age 11 peer problems on age 9 peer problems 0.023 0.056 .686Age 11 peer problems on age 9 academic achievement -0.015 0.051 .764Age 11 peer problems on age 9 parental involvement 0.033 0.062 .596Age 11 academic achievement on age 9 internalising 0.049 0.079 .536Age 11 academic achievement on age 9 externalising 0.098 0.073 .180Age 11 academic achievement on age 9 ADHD -0.129 0.075 .086Age 11 academic achievement on age 9 peer problems -0.120 0.049 .015*Age 11 academic achievement on age 9 academic achievement 0.030 0.057 .596Age 11 academic achievement on age 9 parental involvement -0.231 0.064 <.001*Age 11 parental involvement on age 9 internalising -0.062 0.072 .383Age 11 parental involvement on age 9 externalising 0.030 0.064 .640Age 11 parental involvement on age 9 ADHD -0.109 0.073 .134Age 11 parental involvement on age 9 peer problems 0.007 0.053 .890Age 11 parental involvement on age 9 academic achievement -0.111 0.054 .039*


Age 11 parental involvement on age 9 parental involvement 0.216 0.076 .005*Age 9 internalising on age 7 internalising 0.287 0.105 .006*Age 9 internalising on age 7 externalising -0.107 0.086 .213Age 9 internalising on age 7 ADHD 0.223 0.088 .011*Age 9 internalising on age 7 peer problems -0.031 0.057 .583Age 9 internalising on age 7 academic achievement 0.036 0.066 .588Age 9 internalising on age 7 parental involvement -0.029 0.060 .629Age 9 externalising on age 7 internalising -0.025 0.092 .786Age 9 externalising on age 7 externalising 0.252 0.091 .005*Age 9 externalising on age 7 ADHD -0.199 0.097 .039*Age 9 externalising on age 7 peer problems -0.117 0.063 .064Age 9 externalising on age 7 academic achievement 0.156 0.073 .033*Age 9 externalising on age 7 parental involvement 0.133 0.066 .044*Age 9 ADHD on age 7 internalising 0.212 0.112 .058Age 9 ADHD on age 7 externalising -0.028 0.085 .740Age 9 ADHD on age 7 ADHD 0.192 0.099 .054Age 9 ADHD on age 7 peer problems -0.034 0.060 .566Age 9 ADHD on age 7 academic achievement 0.029 0.074 .689Age 9 ADHD on age 7 parental involvement 0.000 0.063 .999Age 9 peer problems on age 7 internalising -0.116 0.083 .163Age 9 peer problems on age 7 externalising -0.196 0.071 .005*Age 9 peer problems on age 7 ADHD 0.055 0.077 .472Age 9 peer problems on age 7 peer problems 0.163 0.063 .009*Age 9 peer problems on age 7 academic achievement -0.011 0.059 .850Age 9 peer problems on age 7 parental involvement 0.071 0.060 .239Age 9 academic achievement on age 7 internalising 0.001 0.104 .989Age 9 academic achievement on age 7 externalising 0.170 0.100 .091Age 9 academic achievement on age 7 ADHD 0.177 0.122 .146Age 9 academic achievement on age 7 peer problems 0.089 0.073 .222Age 9 academic achievement on age 7 academic achievement -0.058 0.117 .619Age 9 academic achievement on age 7 parental involvement -0.074 0.085 .385Age 9 parental involvement on age 7 internalising 0.028 0.094 .766Age 9 parental involvement on age 7 externalising -0.095 0.080 .237Age 9 parental involvement on age 7 ADHD -0.010 0.101 .921Age 9 parental involvement on age 7 peer problems 0.119 0.065 .067Age 9 parental involvement on age 7 academic achievement -0.225 0.073 .002*Age 9 parental involvement on age 7 parental involvement 0.121 0.088 .170

Note. *Significant at p<.05.


Table S3. Standardised autoregressive and cross-lagged parameters for the teacher-reported ALT-SR

Parameter Estimate SE P

Age 11 internalising on age 9 internalising 0.038 0.077 .620Age 11 internalising on age 9 externalising -0.093 0.077 .230Age 11 internalising on age 9 ADHD 0.011 0.082 .889Age 11 internalising on age 9 peer problems 0.051 0.062 .412Age 11 internalising on age 9 academic achievement -0.050 0.051 .321Age 11 internalising on age 9 parental involvement -0.075 0.062 .226Age 11 internalising on age 7 externalising 0.029 0.056 .605Age 11 internalising on age 7 ADHD 0.035 0.066 .594Age 11 externalising on age 9 internalising 0.219 0.122 .073Age 11 externalising on age 9 externalising -0.142 0.149 .342Age 11 externalising on age 9 ADHD -0.065 0.108 .546Age 11 externalising on age 9 peer problems -0.054 0.075 .470Age 11 externalising on age 9 academic achievement 0.146 0.075 .052Age 11 externalising on age 9 parental involvement 0.100 0.102 .324Age 11 externalising on age 7 ADHD -0.137 0.077 .077Age 11 ADHD on age 9 internalising 0.072 0.103 .481Age 11 ADHD on age 9 externalising -0.120 0.102 .243Age 11 ADHD on age 9 ADHD 0.080 0.104 .437Age 11 ADHD on age 9 peer problems 0.109 0.066 .100Age 11 ADHD on age 9 academic achievement 0.019 0.057 .739Age 11 ADHD on age 9 parental involvement 0.040 0.074 .582Age 11 peer problems on age 9 internalising 0.068 0.071 .339Age 11 peer problems on age 9 externalising -0.070 0.078 .365Age 11 peer problems on age 9 ADHD 0.033 0.071 .640Age 11 peer problems on age 9 peer problems 0.027 0.059 .643Age 11 peer problems on age 9 academic achievement -0.005 0.052 .929Age 11 peer problems on age 9 parental involvement 0.056 0.060 .354Age 11 academic achievement on age 9 internalising 0.005 0.071 .941Age 11 academic achievement on age 9 externalising -0.019 0.078 .810Age 11 academic achievement on age 9 ADHD -0.043 0.078 .579Age 11 academic achievement on age 9 peer problems -0.080 0.058 .166Age 11 academic achievement on age 9 academic achievement 0.000 0.062 .996Age 11 academic achievement on age 9 parental involvement -0.298 0.068 <.001*Age 11 parental involvement on age 9 internalising -0.030 0.064 .642Age 11 parental involvement on age 9 externalising -0.105 0.071 .142Age 11 parental involvement on age 9 ADHD 0.117 0.066 .076Age 11 parental involvement on age 9 peer problems 0.019 0.059 .742Age 11 parental involvement on age 9 academic achievement -0.124 0.053 .020*


Age 11 parental involvement on age 9 parental involvement 0.229 0.075 .002*Age 9 internalising on age 7 internalising 0.313 0.056 <.001*Age 9 internalising on age 7 externalising 0.125 0.067 .063Age 9 internalising on age 7 ADHD 0.047 0.071 .502Age 9 internalising on age 7 peer problems -0.005 0.050 .916Age 9 internalising on age 7 academic achievement 0.062 0.054 .248Age 9 internalising on age 7 parental involvement -0.013 0.055 .808Age 9 externalising on age 7 internalising 0.103 0.054 .054Age 9 externalising on age 7 externalising 0.282 0.074 <.001*Age 9 externalising on age 7 ADHD -0.040 0.064 .534Age 9 externalising on age 7 peer problems -0.025 0.052 .630Age 9 externalising on age 7 academic achievement 0.034 0.052 .514Age 9 externalising on age 7 parental involvement -0.029 0.051 .569Age 9 ADHD on age 7 internalising 0.042 0.060 .490Age 9 ADHD on age 7 externalising -0.003 0.073 .967Age 9 ADHD on age 7 ADHD 0.310 0.073 <.001*Age 9 ADHD on age 7 peer problems 0.082 0.054 .128Age 9 ADHD on age 7 academic achievement 0.144 0.056 .010*Age 9 ADHD on age 7 parental involvement -0.010 0.056 .858Age 9 peer problems on age 7 internalising -0.005 0.049 .926Age 9 peer problems on age 7 externalising 0.014 0.063 .819Age 9 peer problems on age 7 ADHD 0.105 0.053 .050Age 9 peer problems on age 7 peer problems 0.168 0.062 .007*Age 9 peer problems on age 7 academic achievement -0.006 0.053 .907Age 9 peer problems on age 7 parental involvement 0.065 0.055 .240Age 9 academic achievement on age 7 internalising -0.008 0.067 .906Age 9 academic achievement on age 7 externalising 0.025 0.082 .763Age 9 academic achievement on age 7 ADHD 0.002 0.080 .985Age 9 academic achievement on age 7 peer problems 0.037 0.072 .613Age 9 academic achievement on age 7 academic achievement -0.026 0.099 .789Age 9 academic achievement on age 7 parental involvement -0.129 0.081 .111Age 9 parental involvement on age 7 internalising -0.052 0.059 .379Age 9 parental involvement on age 7 externalising -0.072 0.074 .328Age 9 parental involvement on age 7 ADHD 0.165 0.066 .013*Age 9 parental involvement on age 7 peer problems 0.119 0.066 .074Age 9 parental involvement on age 7 academic achievement -0.244 0.065 <.001*Age 9 parental involvement on age 7 parental involvement 0.100 0.082 .222

Note. *Significant at p<.05.

Mediating Factors in Within-Person Developmental Cascades ...

Documents