Identifying candidate mechanisms linking ADHD symptoms and internalising problems in adolescence: An exploratory mediation analysis

doi:10.21203/rs.3.rs-4869282/v1

Download PDF

Research Article

Identifying candidate mechanisms linking ADHD symptoms and internalising problems in adolescence: An exploratory mediation analysis

https://doi.org/10.21203/rs.3.rs-4869282/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

ADHD and internalising symptoms are known to be associated in adolescence and understanding the mechanisms is important for improving mental health outcomes for adolescents with ADHD symptoms. Across the literature, a wide range of mechanisms have been proposed; however, studies to date have typically focused narrowly on only a handful at once. Multiple mediation models that evaluate numerous candidate mediators simultaneously offer the opportunity to provide a more comprehensive picture of the possible mechanisms and identify which mediators represent those with the strongest unique effects that could be prioritised for future longitudinal and intervention research. In the current study, we used exploratory mediation analysis with regularised structural equation modelling to examine >50 candidate mediators of the ADHD-internalising association at ages 11,14, and 17 in the Millennium Cohort Study (n=4918-12597). There were a large number of significant mediators of small-to-moderate effect. Peer problems, conduct problems, school problems, and self-esteem emerged as those with the largest unique effects. Results suggest that there are potentially a wide range of mechanisms contributing to the association between ADHD and internalising problems in adolescence and imply a need for more encompassing theories that address a multitude of pathways between ADHD symptoms and internalising problems. They also point to those which can be prioritised in longitudinal and interventional studies.

ADHD

internalising problems

exploratory mediation

regularised structural equation modelling

Young people with ADHD symptoms commonly experience internalising problems such as anxiety and depression, occurring at rates in these young people exceeding the rates in the general population [1, 2]. Estimates suggest that approximately 25% of young people with ADHD also meet diagnostic criteria for anxiety disorder [3] and approximately 40% experience a depressive episode before the age of 30 [4]. Illuminating the mechanisms underlying this association is important for informing interventions to reduce their co-occurrence, especially through prevention of internalising problems secondary to ADHD symptoms. Adolescence is a critical period in this respect. It is associated with a range of challenges, including increased expectations of independence, academic and social stressors, and emotional changes that can increase the risk of internalising problems [5]. These challenges may compound pre-existing vulnerabilities present among young people with ADHD symptoms.

Many and varied factors have been proposed to mediate the links between ADHD symptoms and internalising problems. Theoretical models such as the cognitive behavioural and dual failure models propose, for example, that impairments or ‘failures’ in various life domains can lead to a negative self-concept and increased risk of depression [6, 7]. Indeed, studies have suggested that facing challenges such as social skills and peer difficulties such as peer rejection, victimisation or dislike; problems with school work; difficulties in relationships with parents, and behavioural problems [6–12] mediate associations between ADHD symptoms and internalising problems. Other models focus on co-occurring cognitive-behavioural features of ADHD and/or internalising problems as mediators. For example, studies have identified factors such as emotion (dys-)regulation [13–16], reward processing, self-control [9], maladaptive cognitions and cognitive-behavioural avoidance [17], sluggish cognitive tempo [18], locus of control [19], and rejection sensitivity [20] as mediators of ADHD-internalising problem associations. Finally, a number of studies have identified a potential role of parenting e.g., overprotective parenting [21] or parent behaviour management [19].

Previous studies analysing mediators have, however, often found that they individually explain only a modest proportion of the association between ADHD symptoms and internalising problems [7, 12]. This suggests that there may be a large number of mediators each of small effect. However, existing studies have typically focused on only a handful at a time, meaning that a comprehensive picture of their combined effects remains lacking.

More comprehensive analyses can be used to build more encompassing theories of the relations between ADHD and internalising problems in adolescence. It can also play a key role in determining which mediators are most promising to investigate further as intervention targets. Even when a candidate mediator is identified, confirming a causal role is challenging and cannot be done conclusively without resource intensive research designs (such as randomised controlled experiments). There is thus considerable value in providing initial exploratory evidence on which mediators represent promising targets for further study before embarking on confirmatory research [22].

Exploratory mediation within a SEM framework that implements regularisation on the mediation parameters can provide this evidence (a technique also known as ‘regSEM’). The approach has been used in several previous psychosocial health studies to facilitate mediator selection where there are a large number of candidate mediators [23, 24]. For example, Ammerman et al. (2018) used regSEM to identify a subset of 11 of 46 variables that contributed to the effects of childhood maltreatment on suicidality in females, while none of the 46 candidate variables were selected in males [23]. Similarly, Casini et al. (2021) used regSEM to select emotion regulation mediators of the association between rejection sensitivity and the outcomes of aggression, withdrawal, and prosociality [25].

These applications illustrate the value of the regSEM for identifying candidate mediators for further study (e.g., in longitudinal or intervention research). In this study we, therefore, apply the technique to identify the most promising mediators in the association between ADHD symptoms and internalising problems in adolescence in a large UK-based study with adolescent participants. Based on data availability and past literature we evaluated a wide range of candidate mediators for inclusion in our models.

Participants

We used data from the waves of Millennium Cohort Study (MCS) [26] taking place during participants’ adolescence (ages 11, 14, and 17). MCS is a UK-based longitudinal birth cohort study fully documented and available at: https://ukdataservice.ac.uk/. In brief, the first wave of MCS took place in 2000–2002 when participants were aged 9 months old. Participants were sampled using a stratified random sampling design in which individuals were clustered geographically and disproportionately sampled from the three smaller nations of the UK (Scotland, Wales, and Northern Ireland), disadvantaged areas and ethnic minorities. From the age 11 wave of data collection, data were utilised from the 5287 participants (2556 male, 2731 female), at age 14 we used data from 12597 participants (6246 male, 6351 female) and at age 17 we used data from 4918 participants (2891 male, 2126 female).

Measures

ADHD symptoms

ADHD symptoms were measured with the hyperactivity/inattention subscale of the Strengths and Difficulties Questionnaire (SDQ) [27]. The SDQ is one of the most widely used and well-validated behavioural screening instruments for children and [28], including in the current sample where it has shown good psychometric properties, including a high degree of gender, informant and developmental invariance [29, 30]. Though there is some debate around the appropriate cut-point, the hyperactivity/inattention subscale has shown good discrimination with respect ADHD diagnosis [31, 32]. The subscale includes five items that refer to behaviour during the last six months with reference to the following behaviours: ‘restless, overactive, cannot stay still for long’; ‘constantly fidgeting or squirming’; ‘easily distracted, concentration wanders’; ‘thinks things out before acting’; and ‘sees tasks through to the end, good attention span’. Responses were provided on a 3-point Likert-type scale with options: not true (0), somewhat true (1), and certainly true (2). Positively worded items were reverse-coded, and item responses were summed to produce an overall hyperactive/inattentive score with higher scores indicating greater hyperactivity/inattentiveness (possible range = 0–10). We used the parent (age 14), teacher (age 11 and 14), and self-reported (age 17) versions of the SDQ, as available.

Internalising problems

Internalising symptoms were measured with the emotional problems subscale of the SDQ (described above). The emotional problems items refer to: often complaining of headaches, stomach-aches or sickness; having many worries; being often unhappy, down-hearted, or tearful; being nervous or clingy in new situations; and having many fears, being easily scared.

Candidate mediators

Full details of the candidate mediator measures are provided in Supplementary Materials. Mediators included in specific models are listed in the relevant section for each model.

Statistical procedure

Overview

Given that available mediators varied by wave, we fit cross-sectional models to maximise the inclusion of mediators for each age. We began by fitting basic SEMs to estimate the raw association between ADHD symptoms and internalising symptoms at each wave. We then proceeded to add mediators to each model and implement model selection using regularised SEM.

ADHD-internalising problems correlation

For each wave, we first fit a SEM with only ADHD symptoms and internalising problems using diagonally weighted least squares (DWLS) estimation to account for the ordered-categorical nature of the indicators. From these we estimated the correlation between the latent ADHD symptoms and internalising problems factors. These models were fit in lavaan in R statistical software [33].

Exploratory mediation analysis

We next conducted exploratory mediation analysis, adding the set of candidate mediators to the basic SEMs described above. We followed the exploratory mediation analysis with SEM described by [34]. In this method, a path mediation model is fit in which regression paths are specified from ADHD symptoms to the mediators, from the mediators to internalising problems, and from ADHD symptoms to internalising problems. Where relevant, latent variable measurement models for constructs were used (e.g., ADHD symptoms, internalising problems; see Methods section for details). Figure 1 provides a simplified example of the multiple path mediation models fitted. Paths a1-ak label the paths from ADHD symptoms to the mediators, paths b1-bk label the paths from the mediators to internalising problems, and c’ is the direct effect from ADHD symptoms to internalising problems. From the a and b parameters, indirect effects via each mediator are derived by finding the product a*b. The total effect of ADHD symptoms on internalising problems is then the sum of the indirect effects and c’.

In the approach proposed by Serang et al.(2017) the above-described multiple model is fit using regularised SEM [35]. Regularised SEM incorporates a penalty term into the maximum likelihood function used for model fitting:

$$\:F=\text{log}\left(\left|\varSigma\:\right|\right)+tr\left(C*{{\Sigma\:}}^{-1}\right)-\text{log}\left(\left|C\right|\right)-p+\lambda\:P\left(.\right)$$

(1)

where $\:\varSigma\:$ is the expected covariance matrix, C is the observed covariance matrix, p is the total number of manifest variables, $\:\lambda\:\:$is a tuning parameter, and $\:P\left(.\right)$ is a function for summing over the absolute value of the penalised coefficients. In exploratory mediation applications of SEM this is the a and b coefficients. We used a general purpose optimiser (rsolnp) which provides an approximate solution and thresholding to set parameters very close to 0 to 0 [36]. This was implemented within the regsem R package [37].

First, all continuous variables standardised (for the constructs specified using latent variables this was achieved by fixing the latent variable variance to 1). Next, the tuning parameter value was found by testing 100 $\:\lambda\:\:$ values (0 to 1 in 0.01 increments) and identifying the model with the smallest Bayesian Information Criterion (BIC). BIC provides a more practical solution to tuning by cross-validation, which can become impractical for large models due to the computation time involved. The model was then re-fit without regularisation, using only the subset of mediators with non-zero indirect effects from the model with the optimal $\:\lambda\:\:$ value. Statistical significance of the indirect effects was assessed using the 95% confidence interval of the bootstrapped standard errors of the indirect effect coefficients (using 1000 bootstrap samples).

Age 11 exploratory mediation model

At age 11, the following mediators were included in the cross-sectional model: prosociality, peer problems, conduct problems (all with parent and teacher reports); school happiness and parental discipline (parent reports), self-esteem, school motivation, well-being (self-reports), academic performance, injuries, parental closeness, physical activity, gaming screen time, surfing screen time, messaging screen time, social networking screen time, happiness, peer conflict, antisocial behaviour, self-reported academic performance, victimisation, aggression and risk-taking based on the Cambridge Gambling Task (risk-taking, quality of decision-making, deliberation time, risk adjustment, delay aversion) and parental mental health problems latent factor (parent self-reports).

Age 14 exploratory mediation model

At age 14, the following mediators were included in the cross-sectional model : risk decision making based on the Cambridge Gambling Task (risk-taking, quality of decision-making, deliberation time, risk adjustment, delay aversion), physical activity, screen time (TV time, gaming time, social media time), academic self-concept, educational motivation, relationships with parents (maternal conflict, maternal closeness, paternal conflict, paternal closeness), parental monitoring, parental discipline, social support, smoking, gambling, bullying (traditional bullying perpetration, cyberbullying perpetration, traditional bullying victimisation, traditional bullying perpetration), victimisation, antisocial behaviour, willingness to take risks, patience, trust, wellbeing, self-esteem, general health, parental mental health, prosociality, peer problems, and conduct problems.

Age 17 exploratory mediation model

At age 17, the cross-sectional model candidate mediators were: self-reported prosociality conduct problems, peer problems, self-esteem, smoking, alcohol use, binge drinking, victimisation, risky behaviours, and antisocial behaviour, self-control, TV screen time, gaming screen time, social media screen time, gambling, sleep quality, social provision, overall health, physical activity, and parental psychological distress.

Some candidate mediators that were considered in these models such as criminal justice involvement and the use of specific drugs other than cannabis were not included due to high levels of missingness or low category endorsement. Further, while the measurement models included some categorical indicators, for pragmatic reasons (the large number of mediators) the models were fit with maximum likelihood estimation, treating the categorical indicators as if continuous. Complete case analysis was used based on pragmatic considerations given the computation time that would be needed to combine FIML or multiple imputation with tuning (and bootstrapping).

ADHD-internalising problem correlations

Age 11

Initial CFA models fit to the teacher-reported (CFI = .98, TLI = .98, RMSEA = .068, SRMR = .06) and parent-reported CFI = .93, TLI = .91, RMSEA = .10, SRMR = .06) ADHD symptoms and internalising problems data showed reasonable fit overall. The correlation between ADHD and internalising symptoms was r = .53 (p < .001) in the parent-reported data and was r = .34 (p < .001) in the teacher-reported data.

Age 14

Initial CFA models fit to the parent-reported (ADHD symptoms and internalising problems data showed good fit (CFI = .95, TLI = .93, RMSEA = .085, SRMR = .062). ADHD and internalising symptoms were correlated at r = .47 (p < .001).

Age 17

An initial model CFA model fit to the self-reported ADHD and internalising problems data fit well (CFI = .95, TLI = .92, RMSEA = .08, SRMR = .06). ADHD and internalising symptoms were correlated at r = .44 (p < .001).

Exploratory mediation models

Age 11

In the age 11 exploratory mediation analyses, model tuning yielded an optimal $\:\lambda\:\:$ of 0.087. Applying this penalty indicated no mediators for de-selection (i.e., none of the corresponding coefficients were driven to 0) and the final model, therefore, included all mediators. The indirect effects from this model are provided in Table 1. The pattern of mediation was characterised by many mediators of small to moderate effect. Of the 124 mediating pathways tested (and selected), fewer than half (56) were statistically significant. The mediating effects of the largest magnitudes tended to pathways via well-being, peer problems, conduct problems, and school motivation.

Age 14

In the model tuning phase of the regularised SEM, one model did not converge and was discarded. Among the remainder, the optimal $\:\lambda\:\:$ value based on the minimum BIC was 0, which implies the selection of all mediators. The indirect effects from this model are provided in Table 2. Of the 37 mediating pathways tested 27 were statistically significant; however, most had small standardised effects sizes. The largest mediating effects were through school motivation, peer problems, and conduct problems.

Age 17

Fitting the multiple mediation model with regularised SEM identified a $\:\lambda\:\:$ value of 0.89 as optimal based on BIC. With this penalty, there were 5 mediators (conduct problems, risky behaviour, self-control, TV screen time, and social provision) selected. Re-fitting a model with these mediators but without regularisation gave the indirect effects provided in Table 1. All selected mediators were statistically significant. The largest effect sizes were observed for prosociality, peer problems, and self-esteem.

The present study provided a comprehensive exploratory analysis of candidate mediators in the association between ADHD and internalising symptoms across adolescence. These findings indicate which mediators may be most promising to examine in future longitudinal and intervention research. Lasso within regularised SEM selected a large number of mediators of small to moderate effect at across ages 11,14, and 17 suggesting that ADHD symptoms are likely to engender risk for internalising problems via a wide range of pathways. However, there were some key mediators with comparatively larger magnitudes for their unique effects. At age 11 this included wellbeing, school problems, peer problems, and conduct problems; at age 14 they included peer and conduct problems; and at age 17 they included prosociality, peer problems, and self-esteem.

The findings of the current study imply that focusing on just a handful of mediators will provide only a limited piece of the full picture. However, from a practical perspective there are difficulties in measuring and modelling large numbers of mediators simultaneously. Space is at a premium in data collections and researchers will typically need to keep the number of measures as few as possible to minimise participant burden, especially for young people with ADHD [38]. In providing a means to apply principled variable selection to large models, regularised SEM in pre-existing data may help to optimise the use of limited space in studies by pointing to the most promising mediators to measure [34]. By providing preliminary indications of which candidate mediators may be involved in the ADHD-internalising problems association, our findings can also help focus the resources of more rigorous but resource intensive research designs on mediators that are likely to be the most promising. This includes longitudinal research with suitable controls for both time-invariant and time-varying confounding, causal mediation analyses, or experimental designs, which are typically expensive and logistically challenging to implement [39–41].

If the findings of the present study are confirmed by these designs, they would suggest that interventions to reduce the risk of developing co-occurring internalising problems among adolescents with ADHD may be most effective if they address multiple mechanisms. These may include cognitive-behavioural features associated with ADHD (e.g., risk-taking tendencies), socio-environmental stressors (e.g., school and peer problems), and their impacts (e.g., self-esteem). However, some mediators appeared to be particularly key based on the present analysis. In particular, peer, conduct, school, and self-esteem emerged as having the largest unique effects at different developmental stages. These findings are consistent with developmental cascade models which propose that ADHD symptoms can lead to academic and peer issues which undermine young people’s self-esteem and lead to an increased risk of anxiety and depression [7]. Peer problems showed consistently strong effects across adolescence, supporting calls for more attention to this as a key intervention target for adolescents with ADHD [42].

Limitations

It is important to consider the limitations of the present study. First, not all possible mediators that merit exploration were included due to data availability. Mediators such as emotion regulation, sluggish cognitive tempo, rumination, and executive functions are also likely to play a mediating role and would be valuable to include in future comprehensive mediation analyses [43, 44]. There were also some measurement limitations in the mediators, for example, some were measured with a single item that could not capture the complexity of the concepts involved. Others were measured only at some ages making it difficult to compare findings across development and to conduct longitudinal mediation analyses, despite well-known limitations to cross-sectional mediation analyses [45]. As such, as previously emphasised, the current findings should only be considered exploratory hypothesis-generating findings to help highlight the most promising mediators to explore further in more rigorous but resource-intensive designs. Finally, although MCS was collected according to a complex survey design, design-adjusted models (incorporating weighting, stratification and clustering variables) are not readily implementable in the context of regularised SEM. As such, the analyses were not design-adjusted and may therefore be limited in their generalisability to the underlying UK population.

Based on comprehensive cross-sectional mediation analyses with lasso-based variable selection there are likely to be a large number of mediators with small-to-moderate effects responsible for linking ADHD symptoms to internalising problems. Key mediators included peer, conduct, school and self-esteem and well-being issues. These and the range of other mediators selected would represent priority mediators to include in future more resource-intensive longitudinal and interventional research.

Funding The research described within this report was supported by a Medical Research Foundation (MRF) grant (Reference: MRF-001-0014-RG-MURR-C0932)

Conflicts of interest/Competing interests The authors have no conflicts of interest to disclose.

Availability of data and material The Millennium Cohort Study is deposited with the UK Data Service: UK Data Service

Code availability Code is available on OSF: OSF | Identifying candidate mechanisms linking ADHD symptoms and internalising problems in adolescence: An exploratory mediation analysis

Ethics approval The MCS data collections received full ethical approval from the National Health Service MultiCentre Research and Ethics Committee at each wave.

Informed consent Informed consent was obtained in relation to all data used in the study.

Consent for publication N/A

Baxter AJ, Scott KM, Vos T, Whiteford HA (2013) Global prevalence of anxiety disorders: a systematic review and meta-regression. Psychological medicine 43:897–910
Krone B, Newcorn JH (2015) Comorbidity of ADHD and anxiety disorders. Attention-deficit hyperactivity disorder in adults and children 98:
D’Agati E, Curatolo P, Mazzone L Comorbidity between ADHD and anxiety disorders across the lifespan. Int J Psychiat Clin. https://doi.org/10.1080/13651501.2019.1628277
Meinzer MC, Lewinsohn PM, Pettit JW, et al (2013) Attention–deficit/hyperactivity disorder in adolescence predicts onset of major depressive disorder through early adulthood. Depression and Anxiety 30:546–553
Rapee RM, Oar EL, Johnco CJ, et al (2019) Adolescent development and risk for the onset of social-emotional disorders: A review and conceptual model. Behaviour Research and Therapy 103501
Eadeh H-M, Bourchtein E, Langberg JM, et al (2017) Longitudinal evaluation of the role of academic and social impairment and parent-adolescent conflict in the development of depression in adolescents with ADHD. Journal of child and family studies 26:2374–2385
Roy A, Hartman CA, Veenstra R, Oldehinkel AJ (2015) Peer dislike and victimisation in pathways from ADHD symptoms to depression. European child & adolescent psychiatry 24:887–895
Eddy LD, Dvorsky MR, Molitor SJ, et al (2018) Longitudinal evaluation of the cognitive-behavioral model of ADHD in a sample of college students with ADHD. Journal of Attention Disorders 22:323–333
Feldman JS, Tung I, Lee SS (2017) Social skills mediate the association of ADHD and depression in preadolescents. Journal of Psychopathology and Behavioral Assessment 39:79–91
Gordon CT, Hinshaw SP (2017) Parenting stress as a mediator between childhood ADHD and early adult female outcomes. Journal of Clinical Child & Adolescent Psychology 46:588–599
Humphreys KL, Katz SJ, Lee SS, et al (2013) The association of ADHD and depression: Mediation by peer problems and parent–child difficulties in two complementary samples. Journal of abnormal psychology 122:854
Powell V, Riglin L, Hammerton G, et al (2020) What explains the link between childhood ADHD and adolescent depression? Investigating the role of peer relationships and academic attainment. European child & adolescent psychiatry 1–11
Antony EM-A, Pihlajamäki M, Speyer LG, Murray AL (2022) Does emotion dysregulation mediate the association between ADHD symptoms and internalizing problems? A longitudinal within-person analysis in a large population-representative study. Journal of Child Psychology and Psychiatry
Bodalski EA, Knouse LE, Kovalev D (2019) Adult ADHD, emotion dysregulation, and functional outcomes: Examining the role of emotion regulation strategies. Journal of Psychopathology and Behavioral Assessment 41:81–92
Murray AL, Wong S-C, Obsuth I, et al (2020) An ecological momentary assessment study of the role of emotional dysregulation in co-occurring ADHD and internalising symptoms in adulthood. Journal of Affective Disorders
Seymour KE, Chronis-Tuscano A, Halldorsdottir T, et al (2012) Emotion regulation mediates the relationship between ADHD and depressive symptoms in youth. Journal of abnormal child psychology 40:595–606
Knouse LE, Zvorsky I, Safren SA (2013) Depression in adults with attention-deficit/hyperactivity disorder (ADHD): The mediating role of cognitive-behavioral factors. Cognitive Therapy and Research 37:1220–1232
Sevincok D, Ozbay HC, Ozbek MM, et al (2020) ADHD symptoms in relation to internalizing and externalizing symptoms in children: The mediating role of sluggish cognitive tempo. Nordic journal of psychiatry 74:265–272
Ostrander R, Herman KC (2006) Potential cognitive, parenting, and developmental mediators of the relationship between ADHD and depression. Journal of Consulting and Clinical Psychology 74:89
Bondü R, Esser G (2015) Justice and rejection sensitivity in children and adolescents with ADHD symptoms. European Child & Adolescent Psychiatry 24:185–198
Meyer A, Kegley M, Klein DN (2022) Overprotective parenting mediates the relationship between early childhood ADHD and anxiety symptoms: Evidence from a cross-sectional and longitudinal study. Journal of Attention Disorders 26:319–327
Serang S, Jacobucci R (2020) Exploratory mediation analysis of dichotomous outcomes via regularization. Multivariate Behavioral Research 55:69–86
Ammerman BA, Serang S, Jacobucci R, et al (2018) Exploratory analysis of mediators of the relationship between childhood maltreatment and suicidal behavior. Journal of adolescence 69:103–112
O’Loughlin CM, Culianos D, Park Y, et al (2021) Implementing exploratory mediation to clarify the nonsuicidal self-injury–suicidality connection. Journal of Psychopathology and Behavioral Assessment 43:206–215
Casini E, Glemser C, Premoli M, et al (2022) The mediating role of emotion regulation strategies on the association between rejection sensitivity, aggression, withdrawal, and prosociality. Emotion 22:1505
Connelly R, Platt L (2014) Cohort profile: UK millennium Cohort study (MCS). International journal of epidemiology 43:1719–1725
Goodman R (1997) The Strengths and Difficulties Questionnaire: a research note. Journal of child psychology and psychiatry 38:581–586
Kersten P, Czuba K, McPherson K, et al (2016) A systematic review of evidence for the psychometric properties of the Strengths and Difficulties Questionnaire. International Journal of Behavioral Development 40:64–75
Murray AL, Speyer LG, Hall HA, et al (2021) A Longitudinal and Gender Invariance Analysis of the Strengths and Difficulties Questionnaire Across Ages 3, 5, 7, 11, 14, and 17 in a Large UK-Representative Sample. Assessment 10731911211009312
Murray AL, Speyer LG, Hall HA, et al (2021) Teacher versus parent informant measurement invariance of the Strengths and Difficulties Questionnaire. Journal of pediatric psychology 46:1249–1257
Algorta GP, Dodd AL, Stringaris A, Youngstrom EA (2016) Diagnostic efficiency of the SDQ for parents to identify ADHD in the UK: a ROC analysis. European child & adolescent psychiatry 25:949–957
Riglin L, Collishaw S, Thapar AK, et al (2016) Association of genetic risk variants with attention-deficit/hyperactivity disorder trajectories in the general population. Jama psychiatry 73:1285–1292
Rosseel Y (2012) Lavaan: An R package for structural equation modeling and more. Version 0.5–12 (BETA). Journal of statistical software 48:1–36
Serang S, Jacobucci R, Brimhall KC, Grimm KJ (2017) Exploratory mediation analysis via regularization. Structural equation modeling: a multidisciplinary journal 24:733–744
Jacobucci R, Grimm KJ, McArdle JJ (2016) Regularized structural equation modeling. Structural equation modeling: a multidisciplinary journal 23:555–566
Orzek JH, Arnold M, Voelkle MC (2023) Striving for Sparsity: On Exact and Approximate Solutions in Regularized Structural Equation Models. Structural Equation Modeling: A Multidisciplinary Journal 30:956–973. https://doi.org/10.1080/10705511.2023.2189070
Li X, Jacobucci R, Ammerman BA (2021) Tutorial on the Use of the regsem Package in R. Psych 3:579–593
Murray AL, Speyer LG, Thye M, et al (2023) Illuminating the daily life experiences of adolescents with and without ADHD: Protocol for an ecological momentary assessment study
Imai K, Keele L, Tingley D (2010) A general approach to causal mediation analysis. Psychological methods 15:309
Keogh RH, Daniel RM, VanderWeele TJ, Vansteelandt S (2018) Analysis of longitudinal studies with repeated outcome measures: adjusting for time-dependent confounding using conventional methods. American journal of epidemiology 187:1085–1092
Streeter AJ, Lin NX, Crathorne L, et al (2017) Adjusting for unmeasured confounding in nonrandomized longitudinal studies: a methodological review. Journal of clinical epidemiology 87:23–34
Breaux R, Langberg JM (2020) Development and refinement of the RELAX intervention, an intervention targeting emotion dysregulation and interpersonal conflict in adolescents with ADHD: Results from a pilot study. Evidence-Based Practice in Child and Adolescent Mental Health 5:147–163
Broletti MC, Efthymiou C, Murray AL, et al (2023) Investigating the Mediating Role of Executive Function in the Relationship Between ADHD and DCD Symptoms and Depression in Adults. J Autism Dev Disord. https://doi.org/10.1007/s10803-023-06148-7
Kandeğer A, Ünal SO, Ergün MT (2023) Mentation processes such as excessive mind wandering, rumination, and mindfulness mediate the relationship between ADHD symptoms and anxiety and depression in adults with ADHD. European Psychiatry 66:S113–S114
Maxwell SE, Cole DA (2007) Bias in cross-sectional analyses of longitudinal mediation. Psychological methods 12:23

No competing interests reported.

SupplementaryMaterials.docx

Download PDF

Editor assigned by journal
08 Aug, 2024
Submission checks completed at journal
08 Aug, 2024
First submitted to journal
06 Aug, 2024

You are reading this latest preprint version

Identifying candidate mechanisms linking ADHD symptoms and internalising problems in adolescence: An exploratory mediation analysis

Status:

Version 1

Abstract

Figures

Introduction

Methods

Participants

Measures

ADHD symptoms

Internalising problems

Candidate mediators

Statistical procedure

Overview

ADHD-internalising problems correlation

Exploratory mediation analysis

(1)

Age 11 exploratory mediation model

Age 14 exploratory mediation model

Age 17 exploratory mediation model

Results

ADHD-internalising problem correlations

Age 11

Age 14

Age 17

Exploratory mediation models

Age 11

Age 14

Age 17

Discussion

Limitations

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1