Gene-environment interactions on the course of Attention-Deficit/Hyperactivity Disorder
(ADHD) symptoms
Brinksma, Djûke Maaike
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Brinksma, D. M. (2018). Gene-environment interactions on the course of Attention-Deficit/Hyperactivity Disorder (ADHD) symptoms: From early into late adolescence. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter 4
ADHD symptoms in middle adolescence
predict exposure to person-related
life stressors in late adolescence in
5-HTTLPR S-allele homozygotes
Djûke M. Brinksma Pieter J. Hoekstra Annelies De Bildt Jan. K. Buitelaar Barbara van den Hoofdakker
Catharina A. Hartman & Andrea Dietrich
ABSTRACT
Background
Literature suggests that life stressors predict Attention-Deficit/Hyperactivity Disorder (ADHD) symptoms and that this relationship is moderated by the serotonin transporter polymorphism (5-HTTLPR). It is less clear whether, on reverse, ADHD symptoms may influence the risk of exposure to life stressors. Furthermore, the role of life stressors may vary across development depending on the type of life stressor.
Methods
We used three-wave longitudinal data of 1,306 adolescents from the general popula-tion and clinic-referred cohort of the TRacking Adolescents’ Individual Lives Survey. The
5-HTTLPR genotype (SS, LS, LL), parent-reported ADHD symptoms at three time points
(T1: Mage = 11.2; T2: Mage = 13.5; T3: Mage = 16.2 years), and the number of person-related (‘dependent’) and environment-person-related (‘independent’) life stressors occurring between measurements (T1-T2, T2-T3) were assessed. Using path analyses, we examined bidirectional relations between exposure to these life stressors and ADHD symptoms between the separate waves moderated by 5-HTTLPR status.
Results
Exposure to life stressors did not predict ADHD symptoms. Rather, we found that in
5-HTTLPR S-allele homozygotes, ADHD symptoms in middle adolescence (T2) predicted
exposure to the number of person-related life stressors later in adolescence (T2-T3, p = .001). There was no relation with environment-related life stressors.
Conclusions
Our study suggests that S-allele homozygotes with higher levels of ADHD symptoms in middle adolescence are more vulnerable to becoming exposed to person-related (‘de-pendent’) life stressors in late adolescence. Findings emphasize the need to be aware of social-emotional adversities that may occur in genetically vulnerable adolescents with ADHD symptoms in the transition into adulthood.
4
INTRODUCTION
Attention-Deficit/Hyperactivity Disorder (ADHD) may be viewed as the extreme end of behaviors varying along a continuum within the population rather than as a categorical disorder with discrete determinants (Levy et al., 1997). Adolescents with ADHD symp-toms frequently encounter stressors in many areas of daily life, including academic and social-emotional functioning, such as difficulties dealing with school tasks, presence of family and peer conflict, peer rejection, and social isolation (Nijmeijer et al., 2008).
These life stressors may both impact on and result from ADHD symptoms (Combs et al., 2015; Nijmeijer et al., 2008). While exposure to life stressors has indeed been shown to prospectively predict severity of ADHD (Gunther et al., 2007; Sasser et al., 2016) or related externalizing symptoms (Kim et al., 2003), still little is known about prospective bidirectional relationships. Surprisingly few studies have addressed the impact of ADHD symptoms on subsequent exposure to life stressors. The existing studies merely focused on single specific stressors (Lifford et al., 2009; Matthews et al., 2015; Verlinden et al., 2015). For example, in a large prospective cohort, 5-year-olds with ADHD symptoms and problem behaviors were at elevated risk of becoming more socially isolated at age 12 (Matthews et al., 2015). The effects of ADHD symptoms on life stress may be understood in that an individual’s personal characteristics and behaviors may evoke negative re-sponses in the environment leading to stress exposure; also, getting involved in high-risk environments may lead to future stress exposure (Dick et al., 2006; Reiss, 2010).
Interestingly, studies have shown moderation of the relation between life stressors and ADHD symptoms by the serotonin transporter polymorphism (5-HTTLPR; Van der Meer et al., 2014; Müller et al., 2008; Retz et al., 2008). Only S-allele carriers and not L-allele homozygotes of the 5-HTTLPR genotype showed a prospective positive as-sociation between life stressors and severity of ADHD symptoms in a large two-wave study of adolescents with (subthreshold) ADHD and controls (Van der Meer et al., 2014). This gene-environment (G×E) interaction is consistent with the extensive literature on internalizing disorders (Caspi et al., 2010; Nugent et al., 2011), which provides strong evi-dence for 5-HTTLPR moderating the stress-depression relationship in a comprehensive meta-analysis (Karg et al., 2011). Specifically, the S-allele has consistently been shown to modify an individual’s response to stress and has been linked to increased stress sensitivity and reactivity compared to L-allele homozygotes of the 5-HTTLPR (Caspi et al., 2003, 2010). This may be due to increased amygdala activation in response to emotional stimuli (Hariri et al., 2002; Drabant et al., 2012).
The potential role of 5-HTTLPR genotype in the ADHD-life stressor pathway is less straightforward. While on one hand a meta-analysis showed that it is the 5-HTTLPR L-allele that represents increased risk for ADHD (Gizer et al., 2009), on the other hand it is conceivable that S-allele carriers of the 5-HTTLPR evoke more interpersonal conflict
as part of their personal characteristics and behaviors (Dick et al., 2006; Shaw, 2015). In these individuals also higher stress sensitivity may play a role (Caspi et al., 2003; Drab-ant et al., 2012), resulting in increased exposure to life stressors (Kendler et al., 2003; Nijmeijer et al., 2008).
To better understand the role of exposure to life stressors associated with ADHD symptoms, it is important to distinguish between person-related (‘dependent’) and environment-related (‘independent’) life stressors (Johnson et al., 2013). Interestingly, the genetic and environmental origins of person-related and environment-related life stressors have been found to vary over adolescent development (Johnson et al. 2013). This may suggest different developmental risk pathways of stress exposure in relation to psychopathology over time (Johnson et al., 2013). Person-related life stressors are thought to be mainly ‘dependent’ on a person’s behavior and personality traits (Amone-P’Olak et al., 2009; Kendler et al., 1995) and are more genetically influenced than environment-related stressors, with genetic factors increasing in magnitude during the transition to adolescence (Johnson et al., 2013). Environment-related life stressors are thought to occur largely ‘independent’ of the person’s direct involvement and personal traits (Amone-P’Olak et al., 2009; Kendler et al., 1995) and are less genetically influenced than dependent stressors; such shared environmental influences appear to play a greater role in exposure to life stressors in childhood, with decreasing influences dur-ing adolescence (Johnson et al., 2013). This distinction is supported by knowledge on adolescents’ social-emotional development towards personality building and growing into independence (Arnett, 1999; Erikson, 1968). Adolescence represents a particularly sensitive period of life, characterized by increasing peer influences and social relations, self-awareness, sensitivity to social environmental cues (particularly social acceptance), inter-personal conflicts, and mental health problems, resulting from personal experi-ences, maturation, and neurobiological changes (Arnett, 1999; Schriber & Guyer, 2016). This may make the developing adolescent more susceptible to person-related life stressors with increasing age. Indeed, the role of genetic and environmental risk factors for ADHD has been suggested to vary across development (Pingault et al., 2015; Shoval et al., 2014; Thapar et al., 2007).
The current study
The aim of this longitudinal study was to examine bidirectional relations of exposure to different types of life stressors with ADHD symptoms in relation to three 5-HTTLPR geno-types (SS, LS, LL) between early (T1: Mage = 11.2 years), middle (T2; Mage = 13.5 years), and
late adolescence (T3; Mage = 16.2 years), using data from a large pooled general
popula-tion and clinic-referred cohort (n = 1,306). This allowed us to test both the direcpopula-tionality and specificity (environment versus person dependency) of stress-ADHD relations and to detect possible age-dependent effects. We expected to find that (1) exposure to
4
life stressors would predict ADHD symptom levels only in S-allele carriers but not in L-allele homozygotes of the 5-HTTLPR genotype; (2) ADHD symptom levels would pre-dict later exposure to life stressors, also only in S-allele carriers; and (3) person-related (‘dependent’) life stressors would play a more prominent role in the relation with ADHD symptoms, particularly during later adolescence in contrast to environment-related (‘independent’) stressors.
METHODS
Participants
Our study included 1,306 participants in the first (T1; Mage = 11.2, range 10.0-12.6 years),
second (T2; Mage = 13.5, range 12.6-14.9 years), and third (T3; Mage = 16.2, range
14.9-18.4 years) wave of the Tracking Adolescents’ Individual Lives Survey (TRAILS). TRAILS is an ongoing prospective study of Dutch adolescents with the aim to chart and explain the development of mental health from early adolescence into adulthood. Data were derived from the TRAILS population cohort including both urban and rural areas in the North of the Netherlands, and from the parallel clinic-referred cohort of children who had been referred to the Groningen university child and adolescent psychiatry outpa-tient clinic at least once in their life. The sampling procedures, descriptive statistics, and response rates of both cohorts have been well-documented elsewhere (Oldehinkel et al., 2015; De Winter et al., 2005).
At baseline, 2,773 adolescents participated in the population-based (n = 2,230) or clinic-referred cohort (n = 543), with response rates for both cohorts over 80% at each follow-up assessment. Genotyping data of 5-HTTLPR were available of 1,788 of the 1,922 adolescents who had donated DNA. We further excluded 293 adolescents who missed more than half of the items on the number of life stressors at T1-T2 and/or T2-T3, and another 189 individuals whose age at the subsequent wave was within the range of the age range of the previous measurement wave. Our final sample consisted of 1,306 participants, 85.8% of which stemmed from the general population cohort (90.5% Dutch ancestry; n = 105 with clinical or subclinical levels of ADHD symptoms based on ASEBA cut-off scores [Achenbach 1991]) and 14.2% from the clinic-referred cohort (98.9% Dutch ancestry; n = 97 with clinical or subclinical levels of ADHD symptoms).
The child’s parents or legal guardian and adolescents (≥ 12 years) provided both writ-ten informed consent prior to each wave, whereas younger participants provided verbal assent. The TRAILS study was approved by the Central Committee on Research Involving Human Subjects (Dutch CCMO).
Measures
ADHD symptoms. At all waves, the DSM-IV-oriented subscale
Attention-Deficit/Hyper-activity Problems of the Child Behavioral Checklist (CBCL; Achenbach, 1991; Verhulst & Achenbach, 1995) consisting of 7 items (3 inattention and 4 hyperactivity-impulsivity items) was used to measure parent-reported ADHD symptoms during the previous six months. Items were scored on a 3-point Likert-scale ranging from 0 (‘not true’) to 2 (‘very true or often true’). The DSM-oriented subscale of the CBCL has shown good reliability as well as convergent and discriminative validity in a clinical sample of adolescents (Nakamura et al. 2009).
Life stressors. At T2 and T3, the Long-Term Difficulties questionnaire developed by
TRAILS (Amone-P’Olak et al., 2009; Oldehinkel et al., 2008) was used to assess the pres-ence of parent-reported person-related (‘dependent’; 6 items) and environment-related (‘independent’; 5 items) life stressors their child may have been exposed to in the past two years (see Supplementary Information 4.1). We categorized adolescents as having been exposed to no (0), one (1), two (2), or three or more (3) life stressors in order to reduce the influence of extreme or rare scores (Zandstra et al., 2015).
Covariates. As covariates we assessed past-year ADHD medication use (i.e.,
methyl-phenidate, dexamphetamine, and/or atomoxetine) at any time in the preceding year; socioeconomic status (SES) based on five indicators (household income and both parents’ professional occupation and educational attainment) and categorized as low (<P25), intermediate (between P25-P75), or high SES (>P75); two genetic principal com-ponent analysis scores to correct for genetic population stratification; and internalizing problems based on the 32-item DSM-oriented broadband internalizing scale of the CBCL (Achenbach, 1991; Verhulst & Achenbach, 1995) with scores from 0-2 over the past six months, measured at all three waves. Aggression problems based on the 18 items of the empirical CBCL aggressive behavior scale (Achenbach, 1991; Verhulst &Achenbach, 1995) were used as an additional covariate as part of a post-hoc analysis (see results section).
Genotyping. DNA was extracted from blood samples (n = 1,211) or buccal swabs
with a Cytobrush (n = 195) using a manual salting out procedure as described by Miller, Dykes, and Polesky (1988). Genotyping of the length polymorphism 5-HTTLPR was done by simple sequence length analysis (call rate 91.6%). Furthermore, samples were geno-typed for the single-nucleotide polymorphism (SNP) rs25331 (an A to G substitution within the5-HTTLPR L allele by a custom-made TaqMan assay (Applied Biosystems; call rate 96.5%). Genotyping was done at the Research lab for Multifactorial Diseases within the Human Genetics department of the Radboud University Nijmegen Medical Centre in Nijmegen, The Netherlands. Concordance between DNA replicates showed an accuracy of 100%. As presence of the rs25331 G SNP is known to reduce transcriptional efficiency of the L allele, comparable with the S allele (Hu et al. 2006), lg alleles were classified as S, while la alleles were classified as L. Based on these alleles, we refer to the functionality
4
of the expressed transporter as low (SS), intermediate (LS), and high (LL) expression. The
5-HTTLPR genotype distribution was well within Hardy-Weinberg equilibrium in both
cohorts (p > .99).
Statistical analysis
Descriptive analyses were conducted using PASW Statistics 19 whereas Structural Equa-tion Modeling (SEM) with robust Maximum Likelihood estimaEqua-tion (MLR) was done using Mplus Version 6.12 (Muthén & Muthén, 1998-2012). The following fit indices were evalu-ated to test model fit: the Chi-square (χ2) goodness-of-fit test, the Comparative Fit Index
(CFI), and the Root-Mean-Square of Approximation (RMSEA). Values for the CFI should preferably be larger than .95 (Hu & Bentler, 1999) and RMSEA should be below .08 and preferably below .05 (Kline, 2005). Missing data were handled using the Full Information Maximum Likelihood (FIML) method. Chi-square goodness of fit tests were used for test-ing differences in model fit between nested models (Satorra & Bentler, 2001).
We carried out multigroup analyses with regard to the 5-HTTLPR genotypes (SS, LS, and LL) on the longitudinal bidirectional associations between ADHD symptoms and both person-related and environment-related life stressors between T1-T2 and T2-T3. We specified a multigroup model with stability paths for respectively ADHD symptoms and life stressors between available time points, as well as cross-lagged paths between ADHD symptoms and life stressors across subsequent time points. In this model we included the covariates gender (0 = female, 1 = male), age, SES (0 = high, 1 = intermedi-ate, 2 = low), past-year ADHD medication use (0 = no use, 1 = use), and internalizing problems. Adjustment for internalizing symptoms was done given their importance in previous work related to stress × 5-HTTLPR interactions (Caspi et al., 2010). Of note, a priori analyses of quadratic growth did not point to non-linear functions of life stressors (p-values > .10). A p < 0.05 was used to indicate statistical significance.
Model fitting analyses were conducted in two steps. First, we checked the equiva-lence and model fit of the stability paths and relations with covariates across the three
5-HTTLPR genotypes. To the extent that initial model fit of this baseline model was not
acceptable we added additional parameters based on the modification indices. Second, having established adequate model fit of this baseline model, we tested whether the strength of the paths between the respective variables differed significantly between the SS, LS, and LL genotypes using the Δχ2 difference test (Satorra & Bentler, 2001). When
the Δχ² difference test was significant, we used subsequent Wald tests to determine whether the strength of the individual parameters differed between the SS, LS, and LL genotypes.
We additionally conducted a number of sensitivity analyses to check whether (1) results differed when stratified for boys and girls, (2) for subdimensions of ADHD
symp-toms (i.e., the inattention and hyperactivity-impulsivity sympsymp-toms separately), and (3) when excluding subjects with non-Dutch ancestry.
RESULTS
Descriptive statistics
The characteristics of the study sample are shown in Table 1. Correlations between person-related and environment-related life stressors were r = 0.22 (p < .001) at T1-T2 and r = 0.33 (p < .001) at T2-T3. Correlations between ADHD symptoms and internal-izing behavior at the same time points ranged from .37 to .45 (all p-values < .001) for variations of the 5-HTTLPR genotype. There were no significant gene-environment cor-relations (p-values ≥ .05).
Path Analyses
Person-related life stressors. The initial baseline model did not meet the
recommend-ed thresholds for acceptable model fit (χ²(223) = 515.73, p < .001, CFI = .923, RMSEA = .055). Based on modification indices we added an additional stability path from ADHD symptoms at T1 to ADHD symptoms at T3. An invariant path across the three 5-HTTLPR genotypes was preferred above the variant path (Δχ²(2) = 1.95, p = .38). Since the model fit of the baseline model with one additional path was still not satisfactory (χ²(222) = 471.30, p < .001, CFI = .934, RMSEA = .051), a second additional path from ADHD medica-tion at T1 to ADHD symptoms at T2 had to be included. Again, the invariant path across the three groups was preferred above the variant path (Δχ²(2) = 0.58, p = .75). This final model, presenting good model fit (χ²(221) = 414.58, p < .001, CFI = .949, RMSEA = .047) is displayed in Figure 1. To simplify graphical presentation, we do not show the second additional path from ADHD medication use at T1 to ADHD symptoms at T2 (ß = .14, p < .001), pointing to slightly higher ADHD symptom levels at T2 for ADHD medication users at T1.
The stability paths of ADHD symptoms between T1 and T2 (Δχ²(2) = 1.48, p = .48) and between T2 and T3 (Δχ²(2) = 3.96, p = .14) were not moderated by 5-HTTLPR status. Over-all, we found moderate-to-strong positive associations of ADHD symptoms between T1-T2 and between T2-T3, and small-to-moderate associations between T1-T3. Similarly, the stability paths of the number of person-related life stressors between T1-T2 and T2-T3 were positive with a moderate effect size, and were not moderated by 5-HTTLPR status (Δχ²(2) = 2.34, p = .31).
The number of person-related life stressors did not predict subsequent ADHD symp-toms according to 5-HTTLPR status (from T1-T2 to T2: Δχ²(2) = 1.48, p = .48; from T2-T3 to T3: Δχ²(2) = 1.83 p = .40). However, we found that ADHD symptoms at T2 predicted
4
Table 1 Sample char ac teristic s p er 5-HT TLPR status (t otal n = 1,306) SS allele c arriers n = 323 LS allele c arriers n = 659 LL allele c arriers n = 324 Test st atistic P-value Popula tion-based c ohor t, N (%) 259 (80.2%) 571 (86.6%) 291 (89.8%) χ²(2)= 9.77 .001 Age in y ears , M (SD ) T1-T3 13.64 (0.44) 13.65 (0.45) 13.61 (0.45) F(2) = 0.29 .75 M ale gender , N (%) 168 (52.0%) 322 (48.9%) 168 (51.9%) χ²(2) = 1.23 .54 D ut ch anc estr y a, N (%) 289 (89.5%) 609 (92.4%) 298 (92.3%) χ²(2) = 2.66 .27 Socio -ec onomic sta tus b , N (%) lo w 100 (7.7%) 220 (17.0%) 107 (8.2%) medium 156 (12.0%) 330 (25.4%) 166 (12.8%) χ² (4) = 3.41 .49 high 64 (4.9%) 106 (8.2%) 48 (3.7%) ADHD medica tion use c, N (%) T1 22 (6.8%) 50 (7.6%) 12 (3.7%) χ²(2) = 5.55 .06 T2 34 (10.5%) 64 (9.7%) 37 (11.4%) χ²(2) = 0.70 .71 T3 22 (6.8%) 52 (7.9%) 17 (5.2%) χ²(2) = 2.36 .31 In ter nalizing pr oblems d , M (SD ) T1 0.26 (0.21) 0.27 (0.22) 0.26 (0.21) F(2) = 0.40 .67 T2 0.22 (0.20) 0.22 (0.19) 0.21 (0.20) F(2) = 0.47 .62 T3 0.23 (0.20) 0.20 (0.21) 0.22 (0.22) F(2) = 0.52 .59 Main v ariables ADHD sympt oms e,f , M (SD ) T1 0.67 (0.53) 0.59 (0.51) 0.59 (0.48) F(2) = 2.77 .06 T2 0.52 (0.48) 0.46 (0.47) 0.46 (0.44) F(2) = 2.14 .12 T3 0.48 (0.45) 0.44 (0.43) 0.41 (0.42) F(2) = 1.97 .14 Person-r ela ted lif e str essors f,g, M (SD ) T1-T2 0.41 (0.65) 0.39 (0.66) 0.40 (0.63) F(2) = 0.44 .96 T2-T3 0.78 (1.02) 0.66 (0.87) 0.65 (0.87) F(2) = 2.07 .13 En vir onmen t-r ela ted lif e str essors f,g, M (SD ) T1-T2 0.32 (0.51) 0.32 (0.57) 0.25 (0.47) F(2) = 2.73 .07 T2-T3 0.48 (0.75) 0.44 (0.70) 0.38 (0.67) F(2) = 1.75 .17 Not e. ADHD = A tt en tion-D eficit/H yper ac tivit y Disor der a B oth par en ts w er e bor n in the Nether lands . b Groups based on sum sc
or
e of fiv
e indica
tors (household inc
ome and both par
en
ts
’ oc
cupa
tion and educa
tion) c M eth ylphenida te , de xamphetamine and a tomo xetine use a t an y time dur
ing the past y
ear (1) v ersus non-use (0) d M ean of 32-it em DSM -or ien ted br oadband in ter
nalizing scale of the Child B
eha vior Check list ( CBCL; A chenbach, 1991; sc or e r ange 0-2) e M ean of 7-it ems DSM -IV -or ien
ted ADHD subscale of the CBCL (
Achenbach, 1991; sc or e r ange 0-2) f T1-T2 < T2-T3 a t p < .001. g M ean number of e xposur e t o lif e str
essors in the past t
w o y ears bef or e T2 and T3 (sc or e r ange 0-3) h SS < LS, SS < LL a t p < .05.
76
person-related life stressors at T2-T3 (Δχ²(2) = 12.58, p = .002) depending on 5-HTTLPR status. Subsequent analyses showed a significant difference between the SS and LS (Wald’s statistic: χ²(1) = 11.39, p < .001), and SS and LL (Wald’s statistic: χ²(1) = 10.28, p = .001), but not between the LS and LL (Wald’s statistic: χ²(1) = 0.73, p = .39) genotypes of the 5-HTTLPR. Thus, only for adolescents with the SS genotype higher levels of ADHD symptoms at T2 significantly predicted exposure to a higher number of person-related life stressors at T2-T3 (ß = .20, p < .001), but not for those with the LS (ß = .03, p = .48) or LL genotype (ß = -.04, p = .58) of the 5-HTTLPR. To illustrate this finding, of the S-allele homozygotes with clinical or subclinical levels of ADHD symptoms at T2 50.9% experi-enced two or more person-related life stressors at T3 as compared to 15.0% in L-allele homozygotes with clinical or subclinical levels of ADHD symptoms. Moreover, although the 5-HTTLPR status did not differentiate associations between ADHD symptoms at T1 and person-related life stressors at T1-T2 (Δχ²(2) = .44, p = .80), there were small signifi-cant associations within each of the SS, LS, and LL genotypes (all ß = .08, p-values < .001).
Chapter 4: Figure 1
Figure 1. Standardized path coefficients for the associations between ADHD symptoms and the number of
person-related life stressors between early (T1), middle (T2), and late (T3) adolescence in relation to
5-HT-TLPR status.
Note. ADHD = Attention-Deficit/Hyperactivity Disorder symptom severity. Left, middle, and right path
co-efficients represent the SS (n = 323), LS (n = 659), and LL genotypes (n = 324) of the 5-HTTLPR genotype, respectively.
***p < .001 indicate significant within-group differences for corresponding path coefficients; brackets rep-resent significant between-group differences in relation to the SS, LS and LL genotypes of 5-HTTLPR (p < .05).
Environment-related life stressors. As with the model examining the number of
person-related life stressors, the initial baseline model had no acceptable fit (χ2(223) =
481.21, p < .001, CFI = .924, RMSEA = .052). Also here two additional invariant paths were added to the model: a stability path from ADHD symptoms at T1 to ADHD symptoms at T3 and one from ADHD medication at T1 which was regressed on ADHD symptoms at T2. This single adjustment to the initial baseline model resulted in an adequate model
4
fit (χ²(221) = 379.54, p < .001, CFI = .953, RMSEA = .041). The final model is displayed in Figure 2 without the second additional path from ADHD medication use at T1 to ADHD symptoms at T2 (ß = .14, p < .001), pointing to slightly higher ADHD symptom levels at T2 in ADHD medication users at T1.
None of the Δχ2 difference tests indicated significant differences across the three 5-HTTLPR genotypes regarding the bidirectional associations of ADHD symptoms with
environment-related life stressors, and vice versa, between any time points (p-values between .06 and .83). Thus, environment-related life stressors did not predict ADHD symptoms, nor the other way around.
Chapter 4: Figure 2
Figure 2. Standardized path coefficients for the associations between ADHD symptoms and the number
of environment-related life stressors between early (T1), middle (T2), and late (T3) adolescence in relation to 5-HTTLPR status.
Note. ADHD = Attention-Deficit/Hyperactivity Disorder symptom severity. Left, middle, and right path
co-efficients represent the SS (n = 323), LS (n = 659), and LL genotypes (n = 324) of the 5-HTTLPR genotype, respectively.
***p < .001 indicate significant within-group difference for corresponding path coefficients; there are no significant between-group differences in relation to the SS, LS and LL genotypes of 5-HTTLPR (p < .05).
Sensitivity analyses. All sensitivity analyses confirmed the results of the main analyses,
leaving the prospective association between ADHD symptoms at T2 and person-related life stressors at T2-T3 moderated by 5-HTTLPR intact both for boys (Δχ²(2) = 8.30, p = .02) and girls (Δχ²(2) = 6.74, p = .001); for inattention (Δχ²(2) = 11.45, p = .003) and hyperac-tivity-impulsivity symptoms (Δχ²(2) = 9.21, p = .01); and when excluding adolescents of non-Dutch ancestry (Δχ²(2) = 10.35, p = .006).
Post-hoc analyses. As results gave rise to the question whether the association
be-tween ADHD symptoms and person-related stressors might have been accounted for by co-occurring aggression problems (e.g., by initiating fights with peers), we additionally covaried for aggression problems, also confirming the main results (Δχ²(2) = 13.36, p = .001). Thus, results including examination of person-related life stressors and all other subanalyses were similar to the main results, both in terms of significance levels as well
as direction and magnitude of effects. Finally, to verify that the lack of a prospective asso-ciation between life stressors and ADHD symptoms in our path analysis models was not an artefact deriving from reduced variance in the life stressor variables, we performed separate linear regression analyses for the 5-HTTLPR genotypes (i.e., regressing T2 ADHD symptoms on T1-T2 stressors, and T3 ADHD symptoms on T2-T3 stressors) adjusted for covariates. In line with our path analyses, we did not find significant associations (all
p-values above .28).
DISCUSSION
This longitudinal three-wave study investigated bidirectional relations between expo-sure to (in)dependent life stressors and ADHD symptoms moderated by the 5-HTTLPR genotype in a large pooled general population and clinic-referred cohort spanning early to late adolescence. In contrast to our expectations, our results do not support recent prospective evidence for a G×E interaction between the 5-HTTLPR genotype and expo-sure to life stressors predicting ADHD symptoms (Van der Meer et al., 2014), nor associa-tions with environment-related (‘independent’) life stressors. However, as hypothesized, ADHD symptoms predicted exposure to person-related (‘dependent’) life stressors (e.g., conflicts with parents or peers, lack of friends, being bullied, school pressures) at two-year follow-up. In particular, adolescents homozygous for the S-allele with higher levels of ADHD symptoms during middle adolescence were exposed to more person-related life stressors in late adolescence. Overall, our findings suggest differential developmen-tal risk pathways to stress exposure associated with psychopathology, varying as per type of stressor, adolescent developmental stage, and underlying genetic vulnerability (Johnson et al., 2013).
ADHD symptoms predict person-related life stressors
Our findings substantially add to the sparse longitudinal studies which predominantly focused on children reporting unidirectional effects of ADHD symptoms on exposure to specific person-related stressors. For example, Lifford and colleagues (2009) found that boys’ ADHD symptoms prospectively impacted upon hostile mother-son relations, and not vice versa. Further, it has been reported that 5-year-olds with ADHD symptoms were at elevated risk of social isolation at age 12 years, but not the other way around (Matthews et al., 2015). Also, the presence of ADHD symptoms in preschoolers was as-sociated with the risk of becoming a bully-victim in elementary school (Verlinden et al., 2015). We expand this literature to an adolescent population in the transition into adulthood, while investigating a set of stressors, and pointing to the developmental influence of the 5-HTTLPR genotype.
4
An explanation for the prospective relation between ADHD symptoms and person-related life stressors is that exposure to these life stressors is a consequence of an individual’s personal characteristics, for example, through evoking certain responses in others that may lead to social adversities in interrelationships (Caspi et al., 2005; Dick ,2011; Hampson, 2008; Nijmeijer et al., 2008; Reiss, 2010). Beyond early adolescence, conflicts with parents and peers gain in affective importance (Laurens et al., 1998), given the increasing social sensitivity during adolescence (Schriber & Guyer, 2016), strive for independence, and greater significance of and time spent with peers (Arnett, 1999). Moreover, an individual’s characteristics may be associated with the likelihood to get into high risk environments that may set off further problems (Dick et al., 2006; Dick, 2011), such as placement in special education school, in which the child may become exposed to other children with problem behaviors and peer conflicts. On another note, increasing cognitive demands throughout middle and high school may place adoles-cents with ADHD symptoms at risk for experiencing high pressures at school, another frequent person-related stressor.
No significant gene-environment correlations were found between ADHD symptoms and 5-HTTLPR genotype, which implies that our finding represents a true G×E interaction (Dick, 2011). Thus, being S-allele homozygous of the 5-HTTLPR genotype and having high levels of ADHD symptoms appears to increase exposure to more person-related life stressors in later adolescence; this points to genetic influences on becoming exposed to this type of stressors during development, consistent with Johnson and colleagues (2013). One explanation for this finding is that S-allele homozygotes with a higher number of ADHD symptoms might possess specific personal characteristics, such as ag-gression or hostility, trait impulsivity, less efficient response inhibition, or autistic traits that regulate social communication (Nijmeijer et al., 2008), known to be related to the S-allele of 5-HTTLPR genotype (Brune et al., 2006; Landrø et al., 2015; Lesch & Merschdorf, 2000; Oades et al., 2008), and that may impact on life stress exposure. Another possibility is that, in line with the comprehensive stress literature regarding the role of 5-HTTLPR genotype (Caspi et al., 2010; Karg et al., 2011), adolescent S-allele homozygotes show an increased sensitivity and reactivity to environmental cues, reflecting an underlying vulnerability to encounter life problems on a broader scale. Individuals are not just pas-sive recipients of life stressors, but there is a continuous interplay between person and environment shaping one’s course of life (Cox et al., 2010; Hampson, 2008; Lerner, 1991; Shanahan & Boardman, 2009). To illustrate, adolescents with ADHD symptoms who have become bully-victims and who exhibit high stress sensitivity may develop even more withdrawn, unadjusted, or aberrant behavior, placing them at further risk of being bul-lied, isolated, having poor-quality relationships, or underachieving at school. Eventually, this may result in a vicious cycle with poor long-term outcomes in multiple domains of daily living.
Interestingly, in early adolescence ADHD symptoms significantly predicted person-related life stressors independently from the 5-HTTLPR genotype. This points to devel-opmental serotonin-related changes across adolescence, as suggested by previous re-search (e.g., Shoval et al., 2014; Thissen et al., 2015). In particular, the 5-HTTLPR genotype mediated vulnerability may operate during sensitive developmental periods such as late adolescence and young adulthood (Uher & McGuffin, 2008). Taken together, findings are in line with the notion that there is a complex and dynamic interplay between genes and environment changing over the life course (Shanahan & Boardman, 2009).
Environment-related life stressors were not associated with ADHD
symptoms
Environment-related (‘independent’) life stressors, which mainly occur independently from the individual’s own behavior, did not predict ADHD symptoms moderated by the
5-HTTLPR genotype. Although previous literature has indicated associations between
in-dependent life stressors and the broader domain of externalizing problems (e.g., Ciu et al., 2007; Rudolph et al., 2000; Schneiders et al., 2003), also negative findings have been reported (Conway et al., 2014). In our study population, stress exposure (e.g., neighbor-hood violence) may not have been severe enough to impact on ADHD symptoms (Karg et al. 2011). Moreover, as it has been shown that shared environmental influences (e.g., within the family system) decrease in magnitude as children transition into adolescence (Johnson et al., 2013), we suspect that our study population may already have passed the critical susceptible age to show an association between (family-based) environment-related stressors and ADHD symptoms.
Life stressors did not predict ADHD symptoms
Our nullfinding of an effect of life stressors on ADHD symptoms moderated by the
5-HT-TLPR genotype appeared not to be a consequence of our pathway model or the presence
of non-linear relationships. A few factors may have contributed to our nullfinding com-pared to previous evidence (Van der Meer et al., 2014; Retz et al., 2008). Most notably, we investigated long-term difficulties as stress measure and not classic incidental life events (e.g., abuse, death of a loved one), although the long-term stressors were similar as in Van der Meer and colleagues (2014). Also, there were some differences with the study population of the previous studies (Van der Meer et al., 2014; Retz et al., 2008), we included the full range of ADHD symptoms from a population-based and clinic-referred sample and exclusively focused on adolescence, whereas the other samples also had young adults and only consisted of persons with more severe symptomatology. Perhaps life stressors have a more pronounced influence with increasing age and/or increasing severity of symptoms. Further studies are warranted to confirm the earlier potentially
4
interesting findings taking the nature, timing, and severity of life stressors and ADHD symptomatology into account.
Strengths and limitations
A strength of this study was the use of a large, longitudinal dataset enabling us to exam-ine bidirectional effects between ADHD symptoms and the exposure to (in)dependent life stressors in relation to 5-HTTLPR status spanning early to late adolescence. Findings were independent from co-occurring internalizing and aggression problems. Additional analyses confirmed that the results were not influenced by sex and ADHD symptom subdimensions (i.e., inattention and hyperactivity-impulsivity). While the use of a CBCL subscale to assess ADHD symptoms may be seen as suboptimal as it does not capture all 18 DSM ADHD symptoms, it has been shown to have adequate discriminative valid-ity (Nakamura et al., 2009). Other limitations have been the use of a small number of parent-reported life stressors, assessing long-term stressors rather than severe adverse life events through which important aspects of life stress may not have been captured. Also the retrospective recall of life stressors may be a potential limitation because of under- and overestimations of stress exposures. While one may question the relevance of some of the person-related stressors (e.g., chronic illness/handicap) we kept with the original scale classification; however, confounding of results is unlikely to have hap-pened, as these stressors typically do not arise as new stressors within the time of our study. Further of note, although the prospective design and the involvement of genetic moderation, as well as the theoretical background may suggest causative pathways be-tween ADHD symptoms and exposure to life stressors, we can neither exclude the influ-ence of other unmeasured traits associated with ADHD symptoms, nor specify whether ADHD symptoms in itself are a necessary or sufficient cause for evoking life stressors. Yet, we could exclude the possibility that our findings were driven by aggression problems. It also deserves mentioning that our pathway model with only two time points of stress exposure does not allow for partialling out within- and between individual patterns and therefore limits causal inferences (Hamaker et al., 2015). Moreover, it should be noted that present findings on the full spectrum of ADHD symptoms in adolescents may not generalize to clinical samples with a diagnosis of ADHD, nor to other age groups. Finally, given the novelty of the findings and concerns about replication in candidate gene stud-ies, future replication is warranted.
Conclusions
In conclusion, our study emphasizes that genetically vulnerable adolescents homozy-gous for the S-allele of the 5-HTTLPR genotype with higher ADHD symptom levels in middle adolescence are at increased risk of becoming exposed to person-related life stressors in late adolescence. Findings support the idea that exposure to person-related
life stressors is not merely a matter of ‘fateful coincidence’, but is influenced by an in-dividual’s personal characteristics and genetic predisposition. This vulnerability likely varies over the life course. Importantly, and also of direct clinical relevance, our study highlights the need for assessment and prevention particularly of social-emotional adversities associated with interpersonal and academic problems (i.e., person-related life stressors) that may occur in the transition from adolescence into adult life in indi-viduals with ADHD symptoms. In other words, clinicians, parents, and teachers should also be aware of the future consequences of problem behavior. Future research should identify the most vulnerable individuals investigating long-term outcomes in multiple domains of functioning far into adulthood including the role of the 5-HTTLPR. Future research could also address the potential role of variation in positive environments given evidence that 5-HTTLPR may represent a plasticity gene rather than a risk gene per se (Belsky & Pluess, 2009).
4
REFERENCES
Achenbach, T. (1991). Manual for the Child Behavior Checklist/4-18 and 1991 Profile. Burlington, VT: Univer-sity of Vermont Department of Psychiatry.
Amone-P’Olak, K., Ormel, J., Huisman, M., Verhulst, F.C., Oldehinkel, A.J., & Burger, H. (2009). Life stressors as mediators of the relation between socioeconomic position and mental health problems in early adolescence: The TRAILS study. Journal of the American Academy of Child and Adolescent
Psychiatry, 48(10), 1031-1038.
Arnett, J.J. (1999). Adolescent storm and stress, reconsidered. American Psychology, 51(4), 317-326. Belsky, J., & Pluess, M. (2009). Beyond diathesis stress: Differential susceptibility to environmental
influ-ences. Psychological Bulletin, 135(6), 885-908.
Brune, C.W., Kim, S-J., Salt, J., Leventhal, B.L., Lord, C., & Cook, E.H. (2006). 5-HTTLPR genotype-specific phenotype in children and adolescents with autism. American Journal of Psychiatry, 163(12), 2148-2156.
Caspi, A., Hariri, A.R., Holmes, A., Uher, R., & Moffitt, T.E. (2010). Genetic sensitivity to the environment: The case of the serotonin transporter gene and its implications for studying complex diseases and traits. The American Journal of Psychiatry, 167(5), 509-527.
Caspi, A., Roberts, B.W., & Shiner, R.L. (2005). Personality development: Stability and change. Annual
Review of Psychology, 56, 453-484.
Caspi, A., Sugden, K., Moffitt, T.E., Taylor, A., Craig, I.W., Harrington, H., ... Poulton, R. (2003). Influence of life stress on depression: Moderation by a polymorphism in the 5-HTT gene. Science, 301(5631), 386-389.
Ciu, M., Donnellan, M.B., & Conger, R.D. (2007). Reciprocal influences between parents’ marital problems and adolescent internalizing and externalizing behavior. Developmental Psychology, 43(6), 1544-1552.
Combs, M.A., Canu, W.H., Broman-Fulks, J.J., Rocheleau, C.A., & Nieman, D.C. (2015). Perceived stress and ADHD symptoms in adults. Journal of Attention Disorders, 19(5), 425-434.
Conway, C.C., Slavich, G.M., & Hammen, C. (2014). Daily stress reactivity and serotonin transporter gene (5-HTTLPR) variation: internalizing responses to everyday stress as a possible transdiagnostic phenotype. Biology of Mood & Anxiety, 4(2), 2.
Cox, M.J., Mills-Koonce, R., Propper, C., & Garpiéry, J.L. (2010). Systems theory and cascades in develop-mental psychopathology. Development and Psychopathology, 22, 497-506.
De Winter, A.F., Oldehinkel, A.J., Veenstra, R., Agnes, J., Verhulst, F.C., & Ormel, J. (2005). Evaluation of non-response bias in mental health determinants and outcomes in a large in a large of non-non-response bias in mental health determinants and outcomes sample. European Journal of Epidemiology, 20(2), 173-181.
Dick, D.M. (2011). Gene-environment interaction in psychological traits and disorders. Annual Review of
Clinical Psychology, 7, 383-409.
Dick, D.M., Rose, R.J., & Kaprio, J. (2006). The next challenge for psychiatric genetics: Characterizing the risk associated with identified genes. Annals of Clinical Psychiatry, 18(4), 223-231.
Drabant, E.M., Ramel, W., Edge, M.D., Hyde, L.W., Kuo, J.R., Goldin, P.R., ... Gross, J.J. (2012). Neural mecha-nisms underlying 5-HTTLPR related sensitivity to acute stress. American Journal of Psychiatry,
169(4), 397-405.
Erikson, E. H. (1968). Identity: Youth and crisis. New York: Norton.
Gizer, I.R., Ficks, C., & Waldman, I.D. (2009). Candidate gene studies of ADHD: A meta-analytic review.
Gunther, N., Drukker, M., Feron, F., & Van Os, J. (2007). No ecological effect modification of the association between negative life experiences and later psychopathology in adolescence: A longitudinal community study in adolescents. European Psychiatry, 22(5), 296-304.
Hamaker, E.L., Kuiper, R.M., & Grasman, R.P. (2015). A critique of the cross-lagged panel model.
Psychologi-cal Methods, 20(1), 102-116.
Hampson, S.E. (2008). Mechanisms by which childhood personality traits influence adult well-being.
Current Directions in Psychological Science, 17, 264-268.
Hariri, A.R., Mattay, V.S., Tessitore, A., Kolachana, B., Fera, F., Goldman, D., ... Weinberger, D.R. (2002). Sero-tonin transporter genetic variation and the response of the human amygdala. Science, 297(5580), 400-403.
Hu, L.T. & Bentler, P.M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1-55.
Hu, X.Z., Lipsky, R.H., Zhu, G., Akhtar, L.A., Taubman, J., Greenberg, B.D., ... Weinberger, D.R.(2006). Se-rotonin transporter promoter gain-of-function genotypes are linked to obsessive-compulsive disorder. American Journal of Human Genetics, 78 (5), 815-826.
Johnson, D.P., Rhee, S.H., Whisman, M.A., Corley, R.P., & Hewitt, J.K. (2013). Genetic and environmental influences on negative life events from late childhood to adolescence. Child Development, 84(5), 1823-1839.
Karg, K., Burmeister, M., Shedden, K., & Sen, S. (2011). The serotonin transporter promoter variant (5-HT-TLPR), stress, and depression meta-analysis revisited: Evidence of genetic moderation. Archives of
General Psychiatry, 68(5), 444-454.
Kendler, K.S., Gardner, C.O., & Prescott, C.A. (2003). Personality and the experience of environmental adversity. Psychological Medicine, 33(7), 1193-1202.
Kendler, K.S., Kessler, R.C., Walters, E.E., MacLean, C., Neal, M.C., Heath, A.C., ... Eaves, L.J. (1995). Stressful life events, genetic liability, and onset of an episode of major depression in women. American
Journal of Psychiatry, 152(6), 833-842.
Kim, K.J., Conger, R.D., Elder Jr., G.H., & Lorenz, F.O. (2003). Reciprocal influences between stressful life events and adolescent internalizing and externalizing problems. Child Development, 74(1), 127-143.
Kline, R.B. (2005). Principles and Practice of Structural Equation Modeling (2nd ed.). New York: The Guilford Press.
Landrø, N.I., Jonassen, R., Clark, L., Foss Haug, K.B., Aker, M., Bø, R., ... Stiles, T.C. (2015). Serotonin trans-porter polymorphisms predict response inhibition in healthy volunteers. Neuroscience Letters,
584(1), 109-112.
Laurens, B., Coy, K.C., & Collins, W.A. (1998). Reconsidering changes in parent-child conflict across adoles-cence: A meta-analysis. Child Development, 69(3), 817-832.
Lerner, R.M. (1991). Changing organism-context relations as the basic process of development: A devel-opmental contextual perspective. Develdevel-opmental Psychology, 27, 27-32.
Lesch, K.P., & Merschdorf, U. (2000). Impulsivity, aggression, and serotonin: A molecular psychobiological perspective. Behavioral Science and the Law, 18(5), 581-604.
Levy, F., Hay, D.A., McStephen, M., Wood, C., & Waldman, I. (1997). Attention-deficit hyperactivity disorder: A category or a continuum? Genetic analysis of a large-scale twin study. Journal of the American
Academy of Adolescent Psychiatry, 36(3), 737-744.
Lifford, K. J., Harold, G. T., & Thapar, A. (2009). Parent–child hostility and child ADHD symptoms: A ge-netically sensitive and longitudinal analysis. Journal of Child Psychology and Psychiatry, 50(12), 1468-1476.
4
Matthews, T., Danese, A., Wertz, J., Ambler, A., Kelly, M., Diver, A., ... Arseneault, L. (2015). Social isolation and mental health at primary and secondary school entry: A longitudinal cohort study. Journal of
the American Academy of Child and Adolescent Psychiatry, 54(3), 225-232.
Nakamura, B.J., Ebesutani, C., Bernstein, A., & Chorpita, B.F. (2009). A psychometric analysis of the child behavior checklist DSM-oriented scales. Journal of Psychopathology and Behavioral Assessment,
31(3), 178-189.
Miller, S.A., Dykes, D.D., & Polesky, H.F. (1988). A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Research, 16, 1215.
Müller, D.J., Mandelli, L., Serretti, A., DeYoung, C.G., De Luca, V., Sicard, T., Kennedy, J.L. (2008). Serotonin transporter gene and adverse life events in adult ADHD. American Journal of Medical Genetics. Part
B, Neuropsychiatric. Genetics, 147B(8), 1461-1469.
Muthén, L.K. & Muthén, B.O. (1998–2012). Mplus user’s guide (6th ed.). Los Angeles: Muthén & Muthén. Nijmeijer, J.S., Minderaa, R.B., Buitelaar, J.K., Mulligan, A., Hartman, C.A., & Hoekstra, P.J. (2008).
Attention-deficit/hyperactivity disorder and social dysfunctioning. Clinical Psychology Review, 28(4), 692-708.
Nugent, N.R., Tyrka, A.R., Carpenter, L.L., & Price, L.H. (2011). Gene-environment interactions: Early life stress and risk for depressive anxiety disorders. Psychopharmacology, 214(1), 175-196.
Oades, R.D., Lasky-Su, J., Christiansen, H., Faraone, S.V., Sonuga-Barke, E.J., Banaschewski, T., ... Asherson, P. (2008) The influence of serotonin- and other genes on impulsive behavioral aggression and cognitive impulsivity in children with attention-deficit/hyperactivity disorder (ADHD): Findings from a family-based association test (FBAT) analysis. Behavioral and Brain Functions, 84(4), 48. Oldehinkel, A.J., Rosmalen, J.G.M., Buitelaar, J.K., Hoek, H.W., Ormel, J., Raven, D., ... Hartman, C.A.(2015).
Cohort profile update: The Tracking Adolescents’ Individual Live Survey (TRAILS). International
Journal of Epidemiology, 44(1), 76-76n.
Oldehinkel, A.J., Verhulst, F.C., & Ormel, J. (2008). Low heart rate: A marker of stress resilience. the TRAILS study. Biological Psychiatry, 63(12), 1141-1146.
Pingault, J., Viding, E., Galéra, C., Greven, C.U., Zhen, Y., Plomin, R., ... Rijsdijk, F. (2015). Genetic and en-vironmental influences on the developmental course of attention-deficit/hyperactivity disorder symptoms from childhood to adolescence. JAMA Psychiatry, 72(7), 651-658.
Reiss, D. (2010). Genetic thinking in the study of social relationships: Five points of entry. Perspectives of
Psychological Science, 5(5), 502-515.
Retz, W., Freitag, C.M., Retz-Junginger, P., Wenzler, D., Schneider, M., Kissling, C., ... Rösler, M. (2008). A functional serotonin transporter promoter gene polymorphism increases ADHD symptoms in de-linquents: Interaction with adverse childhood environment. Psychiatry Research, 158(2), 123–131. Rudolph, K.D., Hammen, C., Burge, D., Lindberg, N., Herzberg, & Daley, S.E. (2000). Toward an interpersonal
life-stress model of depression: The developmental context of stress generation. Development
and Psychopathology, 12(2). 215-234.
Sasser, T.R., Kalvin, C.B., & Bierman, K.L. (2016). Developmental trajectories of clinically significant attention-deficit/hyperactivity disorder (ADHD) symptoms from grade 3 through 12 in a high-risk sample: Predictors and outcomes. Journal of Abnormal Psychology, 125(2), 207-219.
Satorra, A., & Bentler, P.M. (2001). A scaled difference chi-square test statistic for moment structural analysis. Psychometrika, 66(4), 507-514.
Schneiders, J., Drukker, M., Van der Ende, J., Verhulst, F.C., Van Os, J., & Nicoloson, N.A. (2003). Neighbour-hood socioeconomic disadvantage and behavioural problems from late childNeighbour-hood into early adolescence. Journal of Epidemiology and Community Health, 57(9), 699-703.
Schriber, R.A., & Gruyer, A.E. (2016). Adolescent neurobiological susceptibility to social context.
Develop-mental Cognitive Neuroscience, 19(1), 1-18.
Shanahan, M.J., & Boardman, J.D. (2009). Genetics and behavior in the life course: A promising frontier. In
The Craft of Life Course Research (ed. G. H. Elder Jr. and J. Z. Giele), pp. 215-235. The Guilford Press:
New York.
Shaw, P. (2015). Gene and environment interactions in the brain: A pathway to ADHD? American Journal
of Psychiatry, 172(8), 702-703.
Shoval, G., Bar-Shira, O., Zalsman, G., John Mann, J., & Chechik, G. (2014). Transitions in the transcriptome of the serotonergic and dopaminergic systems in the human brain during adolescence. European
Neuropsychopharmacology, 24(7), 1123-1132.
Thapar, A., Langley, K., Asherson, P., & Gill, M. (2007). Gene-environment interplay in attention-defi cit hyperactivity disorder and the importance of a developmental perspective. The British Journal of
Psychiatry, 190, 1-3.
Thissen, A.J.A.M., Bralten, J., Rommelse, N.N.J., Arias-Vasques, A., Greven, C.U., Heslenfeld, D., ... Buitelaar, J.K. (2015). The role of age in association analyses of ADHD and related neurocognitive function-ing: A proof of concept for dopaminergic and serotonergic genes. American Journal of Medical
Genetics. Part B, Neuropsychiatric. Genetics, 168(6), 471-479.
Uher, R., & McGuffi n, P. (2008). The moderation by the serotonin transporter gene of environmental adversity in the aetiology of mental illness: Review and methodological analysis. Molecular
Psychiatry, 13(2), 131-146.
Van der Meer, D., Hartman, C.A., Richards, J., Bralten, J.B., Franke, B., Oosterlaan, J., ... Hoekstra, P.J. (2014). The serotonin transporter gene polymorphism 5-HTTLPR moderates the eff ects of stress on attentiondefi cit/hyperactivity disorder. Journal of Child Psychology and Psychiatry, 55(12), 1363-1371.
Verhulst, F.C., & Achenbach, T.M. (1995). Empirically based assessment and taxonomy of psychopathol-ogy: Cross-cultural applications. A review. European Child and Adolescent Psychiatry, 4(2), 61-76. Verlinden, M., Jansen, P.W., Veenstra, R., Jaddoe, V.W.V., Hofman, A., Verhulst, F.C., Tiemeier, H. (2015).
Preschool attention-defi cit/hyperactivity and oppositional defi ant problems as antecedents of school bullying. Journal of the American Academy of Child and Adolescent Psychiatry, 54(7), 571-579.
Zandstra, A.R.E., Hartman, C.A., Nederhof, E., van den Heuvel, E.R., Dietrich, A., Hoekstra, P.J., & Ormel, J. (2015). Chronic stress and adolescents’ mental health: Modifying eff ects of basal cortisol and pa-rental psychiatric history. The trails study. Journal of Abnormal Child Psychology, 43(6), 1119-1130.
