Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Serotonin transporter (5-HTT) gene network moderates the impact of prenatal maternal adversity on orbitofrontal cortical thickness in middle childhood

  • Aleeza Sunderji,

    Roles Data curation, Formal analysis, Writing – original draft, Writing – review & editing

    Affiliation Department of Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, ON, Canada

    ⨯
  • Heather D. Gallant,

    Roles Data curation, Methodology, Writing – original draft

    Affiliation Department of Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, ON, Canada

    ⨯
  • Alexander Hall,

    Roles Data curation, Formal analysis

    Affiliation Department of Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, ON, Canada

    ⨯
  • Andrew D. Davis,

    Roles Formal analysis, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Department of Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, ON, Canada

    ⨯
  • Irina Pokhvisneva,

    Roles Data curation, Investigation, Methodology, Writing – original draft

    Affiliations Department of Psychiatry, Faculty of Medicine and Health Sciences, McGill University, Montreal, QC, Canada, Ludmer Centre for Neuroinformatics and Mental Health, Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada

    ⨯
  • Michael J. Meaney,

    Roles Conceptualization, Investigation, Methodology, Writing – review & editing

    Affiliations Department of Psychiatry, Faculty of Medicine and Health Sciences, McGill University, Montreal, QC, Canada, Ludmer Centre for Neuroinformatics and Mental Health, Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada, Translational Neuroscience Program, Singapore Institute for Clinical Sciences and Brain–Body Initiative, Agency for Science, Technology and Research (A*STAR), Singapore Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore

    ⨯
  • Patricia P. Silveira,

    Roles Conceptualization, Data curation, Investigation, Methodology, Writing – review & editing

    Affiliations Department of Psychiatry, Faculty of Medicine and Health Sciences, McGill University, Montreal, QC, Canada, Ludmer Centre for Neuroinformatics and Mental Health, Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada

    ⨯
  • Roberto B. Sassi,

    Roles Conceptualization, Investigation, Writing – original draft, Writing – review & editing

    Affiliation Department of Psychiatry, University of British Columbia, Vancouver, BC, Canada

    ⨯
  • Geoffrey B. Hall

    Roles Conceptualization, Formal analysis, Funding acquisition, Methodology, Supervision, Writing – original draft, Writing – review & editing

    hallg@mcmaster.ca

    Affiliation Department of Psychology, Neuroscience & Behaviour, McMaster University, Hamilton, ON, Canada

    ⨯

Abstract

In utero, the developing brain is highly susceptible to the environment. For example, adverse maternal experiences during the prenatal period are associated with outcomes such as altered neurodevelopment and emotion dysregulation. Yet, the underlying biological mechanisms remain unclear. Here, we investigate whether the function of a network of genes co-expressed with the serotonin transporter in the amygdala moderates the impact of prenatal maternal adversity on the structure of the orbitofrontal cortex (OFC) in middle childhood and/or the degree of temperamental inhibition exhibited in toddlerhood. T1-weighted structural MRI scans were acquired from children aged 6–12 years. A cumulative maternal adversity score was used to conceptualize prenatal adversity and a co-expression based polygenic risk score (ePRS) was generated. Behavioural inhibition at 18 months was assessed using the Early Childhood Behaviour Questionnaire (ECBQ). Our results indicate that in the presence of a low functioning serotonin transporter gene network in the amygdala, higher levels of prenatal adversity are associated with greater right OFC thickness at 6–12 years old. The interaction also predicts temperamental inhibition at 18 months. Ultimately, we identified important biological processes and structural modifications that may underlie the link between early adversity and future deviations in cognitive, behavioural, and emotional development.

Citation: Sunderji A, Gallant HD, Hall A, Davis AD, Pokhvisneva I, Meaney MJ, et al. (2023) Serotonin transporter (5-HTT) gene network moderates the impact of prenatal maternal adversity on orbitofrontal cortical thickness in middle childhood. PLoS ONE 18(6): e0287289. https://doi.org/10.1371/journal.pone.0287289

Editor: Emma Duerden, Western University, CANADA

Received: October 14, 2022; Accepted: June 3, 2023; Published: June 15, 2023

Copyright: © 2023 Sunderji et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The identification of study group membership is based on clinical data (SES, Mood Disorders, Anxiety) collected from the mothers of participants as well as assessment, imaging and developmental data collected at predefined intervals on the children. This is sensitive information and the study group size and the longitudinal collection of data weigh on participant confidentiality and respect. Inquiries to arrange data sharing can be directed to the corresponding author or to the Hamilton Integrated Research Ethics Board (https://hireb.ca/) at telephone # 905-521-2100 ext 42680 or by mail; 293 Wellington Street North, Suite 120, Hamilton ON L8L 8E7.

Funding: This research was made possible by a grant from the Natural Sciences and Engineering Research Council of Canada, RGPIN-2019-06860, awarded to Geoffrey B. Hall. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The prenatal period is one of the most rapid and formative stages of neurodevelopment. Foundational processes such as neurogenesis, synaptogenesis, migration, and differentiation are occurring to establish the baseline architecture for future development [1]. Importantly, such sensitive and dynamic periods of growth are typically characterized by a high degree of plasticity, rendering developing systems to be highly responsive and vulnerable to the environment [2, 3]. Therefore, it follows that prenatal development is strongly contingent upon environmental input provided in utero. These cues are taken to reflect the child’s expected postnatal conditions [4] and thereby function to tailor trajectories in ways that would be most conducive to survival after birth [3]. This idea of preparedness is captured in the long-standing fetal programming hypothesis, which posits that environmental insults during sensitive periods can exert lasting organizational effects on development [5]. In support, there is a wealth of research reporting alterations in structural, functional, cognitive, behavioral, and emotional development amongst children whose mothers experienced adverse events prenatally [6–16]. However, studies investigating the underlying mechanisms that give rise to such outcomes are needed.

There is a growing body of research which explores factors that alter an individual’s susceptibility to their environment. For instance, genetic variation can influence development so that individuals are better positioned to navigate within the conditions of their postnatal environment. In this way, functional allelic variations may be viewed as facilitating plasticity rather than predicting vulnerabilities [17]. Currently, there is a paucity of research which explores how genotype and one’s early environment interact to affect development during the instrumental middle childhood years. Accordingly, in the present study we aimed to address whether genetic variation in the serotonin transporter (5-HTT) gene network can moderate an individual’s vulnerability to the environment as early as the prenatal period during which environmental perturbations begin to affect long-term development.

Serotonin is one of the first neurotransmitters to be expressed during fetal development with serotonergic neurons detectable as early as 5 weeks gestation [18]. This neurotransmitter performs dual functions during development. In early periods, serotonin acts as a trophic factor regulating critical developmental processes such as cell division, differentiation, migration, myelination, synaptogenesis, and dendritic pruning [19]. In this role, 5-HT shapes the development of its own system and other closely related neural networks. During maturation, 5-HT functions as a modulatory neurotransmitter impacting arousal, stress responsivity as well as higher-order functions including cognition, attention, emotion, and learning [19]. It is important to note that while serotonin’s potential for neurotransmission does exist during early development, the mere presence of serotonergic synapses does not suggest that 5-HT is actively modulating neural activity in the fetal brain as it does in the mature/adult brain [20]. Furthermore, some 5-HT modulated developmental processes are not typically observed in the adult brain and occur exclusively during fetal and early postnatal periods. These processes include neuronal progenitor cell proliferation, neuronal migration, and axonal wiring [20]. Therefore, it is plausible that changes in serotonin signaling during these early critical periods of development can have substantial implications for long-term brain organization and function.

The serotonin transporter is the primary modulator of serotonergic activity (5-HT) in the brain. Alterations in its expression and availability influences the structural and functional development of neural systems, thereby subsequently affecting behaviour, emotion, and cognition [21]. Serotonergic activity has been broadly implicated in the etiology of a number of mood, anxiety, and neurodevelopmental disorders [22]. As such, it is important to understand the role of the serotonin transporter in regulating one’s vulnerability to environmental experiences. Caspi et al. (2003) found that a functional polymorphism in the promoter region of the serotonin transporter gene (SLC6A4) moderates the effect of stressful life events on the development of depression [23]. This seminal research sparked interest in exploring the moderating role the serotonin transporter gene polymorphism (5-HTTLPR) on the relation between early adversity and an increased risk for psychopathology, with a particular focus on a potential risk allele–the short allele. Carriers of the short allele show reduced transporter expression and function (e.g. reduced binding in limbic regions, decreased 5-HT reuptake into presynaptic neurons, etc), thereby suggesting a higher level of extracellular 5-HT [20]. Accumulating evidence has revealed that the short allele is associated with an increased risk for mood disorders, anxiety related temperaments, and structural and functional alterations in regions implicated in affective disorders [24–26].

However, attempts to replicate the interaction found by Caspi et al (2003) have not always been successful and some research has identified that not all of the outcomes associated with the short allele are negative [27]. It has been suggested that the inconsistent results stem from the single candidate gene approach used in these analyses. This approach considers single genes to be independently responsible for particular pathologies. However, complex, polygenic traits such as mental illnesses cannot solely be explained by a single gene and instead involve the interplay between multiple genes. It is more likely that weak effects from multiple genomic variations converge to influence a common biological system which in turn gives rise to instances of psychopathology [27, 28]. As such, polygenic risk scores (PRS) have shown some potential for clinical use in recent years. However, PRS fail to account for the biological events that underlie such complex states. Accordingly, a better approach should capture the biological processes involved in a particular disorder and provide us with a mechanistic link of how early adversity can lead to aberrant developmental outcomes. In the research reported here, we use a co-expression based polygenic risk score (ePRS) methodology developed by Silveira and colleagues to study how genetics interact with the prenatal environment to affect subsequent child development [28–30]. The ePRS is generated by identifying genes that are co-expressed with a particular gene of interest. The approach is based on the principle that genes which are expressed together, work together, and thereby comprise a gene network [28]. Here, we generate co-expression based polygenic risk scores that represent the function of the 5-HTT gene network in the amygdala expressed during prenatal development.

Cortical thickness (CT) has a well-defined developmental trajectory in middle childhood [31]. Specifically, the normative trajectory of CT development is characterized by a linear decline from early childhood to early adulthood [31, 32]. Therefore, gradual thinning reflects progressive maturation of brain organization, while reduced thinning (I.e. increased thickness) reflects a delay in cortical development [33, 34]. Factors such as synaptic pruning, dendritic development, changes in glial cell density, and myelination of white matter tracts can contribute to differences in CT [35–38]. Variations in CT are observed in various mental illnesses, thereby prompting investigation into potential risk factors that give rise to such alterations [39]. Some research has also reported regional differences in CT trajectories, however there is consistent evidence that OFC thickness shows a monotonic linear decrease with age [31, 32]. As such, our first aim is to explore whether adverse prenatal experiences and the 5-HTT gene network interact to predict cortical thickness of the orbitofrontal cortex (OFC) in middle childhood.

The OFC is the region of interest for several reasons. First, structural and functional alterations in regions of the Prefrontal Cortex have been observed following exposure to early life adversity, especially in regions actively involved in emotion regulatory processes [30, 40, 41]. The OFC, in particular, is highly interconnected with other brain regions responsible for sensory processing, emotion generation, emotion regulation, inhibitory control, executive control, and reward processing [42, 43]. It plays a critical role in integrating information from a variety of sources to guide behaviour. Therefore, in uncertain contexts, the OFC region is pivotal for generating adaptive and flexible behavioural responses [44]. Secondly, the rudimentary structure and connectivity of the prefrontal cortex is largely determined prenatally [45]. As such, environmental perturbations at this time can impact its gross architecture later in life [45, 46]. Moreover, during middle childhood, the development of the OFC undergoes continual refinement. As a pivotal region for the regulation of affect, cognition, and behaviour, its maturation also influences a child’s social and emotional development. Previous research has shown that structural changes in the OFC have been associated with poor behavioural and emotional outcomes [47, 48]. Therefore, it becomes important to understand the conditions under which alterations in OFC structure can ensue. In this study, we explore whether aberrant input (i.e. via altered 5-HTT-ePRS function) from a stress susceptible subcortical region influences the structural development of the OFC. Since the 5-HTT gene network is expressed in the amygdala, it would be reasonable to explore its influence on a closely connected region. Not only are both the amygdala and OFC highly regulated by serotonergic activity but animal research and a few human studies have revealed that they demonstrate strong bidirectional anatomical connectivity [49, 50]. fMRI research also indicates that the amygdala and both medial and lateral divisions of the OFC demonstrate functional connectivity [50, 51]. Furthermore, the degree of functional coupling between the OFC and amygdala is related to individual differences in emotion regulation and self-control [50, 52–54]. As such, we are interested in determining whether alterations in 5-HT activity within one region can moderate the structural developmental of another during middle childhood following exposure to prenatal maternal adversity.

Zhou and colleagues (2013) reported that cortical thickness asymmetry in frontal regions increases with age [55]. In the orbitofrontal region, rightward asymmetry (thicker right hemisphere relative to left) is detectable during middle childhood [55]. It is suggested that the maturation of cognitive functions may relate to the emerging hemispheric asymmetry from childhood into adulthood [55, 56]. Relatedly, a lack of asymmetry may be indicative of aberrant cognitive functioning or pathology [55, 56]. As such, given that a degree of asymmetry should be expected during middle childhood, we decided to explore hemispheric differences in the moderating role of the 5-HTT gene network by conducting separate analyses on the right and left OFC.

Given that the amygdala is a small, subcortical structure which is separated from other nearby regions by fine boundaries, manual tracing is considered to be the “gold-standard” volumetric segmentation technique. Commonly used automated segmentation tools are found to inflate estimates of amygdala volume, potentially reflecting poor boundary identification [57]. Particularly in pediatric samples, the agreement between manual and automated segmentation of the amygdala is said to be questionable [57]. Our concern regarding the accuracy of amygdala volume estimation in our sample of children guided our choice away from selecting the amygdala as the primary region of interest for this study. Further, when opting for an automated segmentation approach to conduct volumetric analyses of the amygdala, it is generally recommended that researchers acquire large sample sizes when expecting small effect sizes or exploring complex associations such as those involving the role of genetics [58].

The second aim of this study is to determine whether the 5-HTT gene network moderates the impact of prenatal adversity on early temperament. Behavioural inhibition (BI) can be defined as a "child’s initial behavioural reaction to unfamiliar people, objects, and contexts, or challenging situations’’ [59, 60]. This temperament can be detected as early as toddlerhood [26, 61]. Children that exhibit an inhibited temperament display withdrawn, fearful behaviour and increased vigilance in the face of novelty [62]. BI is an antecedent to shyness, which emerges in early childhood as social awareness begins to develop [63]. It has been suggested that a behaviourally inhibited temperament is also a precursor to anxiety and depression in late childhood and adolescence [62, 64–66]. Neurobiologically, BI is potentially attributed to high amygdala reactivity and poor emotion regulation ability by frontal regions [21, 60, 67]. As such, investigations into early indicators of poor regulatory ability are merited.

Ultimately, early temperament reflects a difference in reactivity to the environment which can be rooted in genetics and neurobiology [63]. However, much of the literature investigating differences in temperamental inhibition with relation to the 5HTTLPR polymorphism has led to inconsistent results [64, 68, 69]. Furthermore, prenatal stress is shown to impact the development of subcortical and cortical regions, and stress regulatory systems which underlie a behaviourally inhibited temperament [9, 70]. Interactions between an individual’s environment and the short 5-HTT allele have shown to increase the risk for BI in childhood [26]. For all these reasons, we wished to explore whether there is a 5-HTT gene network x prenatal environment interaction that predicts temperamental inhibition at 18 months.

Ultimately, this study aims to gain a better understanding of the biological mechanisms by which prenatal adversity can exert its downstream effects on temperament in toddlerhood (18 months) and neurodevelopment in middle childhood (6–12 years). We hypothesized that the impact of prenatal maternal adversity on OFC thickness at 6–12 years old and temperamental inhibition at 18 months will vary based on the function of the 5-HTT gene network in the amygdala.

Materials and methods

Sample

The Maternal Adversity, Vulnerability and Neurodevelopment (MAVAN) cohort is a community sample of Canadian mothers and their children who have been studied prospectively since the prenatal period. Dyads were recruited from two sites–either Hamilton, Ontario or Montreal, Quebec. The eligibility criteria entail mothers who are 18 years or older with singleton pregnancies and fluency in English or French. Severe maternal chronic illness, placental previa, a history of incompetent cervix, impending delivery, a fetus/infant affected by a major anomaly or born at a gestational age of <37 weeks or with birthweight less than 2000g comprise the exclusion criteria. The MAVAN project received approval from the ethics committee and affiliated university institutions (i.e. McMaster University, St. Joseph’s Healthcare Hamilton, McGill University, Université de Montréal, Royal Victoria Hospital, Jewish General Hospital, Centre Hospitalier de l’Université de Montréal, and Hôspital Maisonneuve-Rosemount). Informed written consent was obtained from all participants.

In order to study the presence of a gene network x environment interaction that moderates neurodevelopmental outcomes in middle childhood, MRI image acquisition was performed when MAVAN children reached 6–12 years of age (N = 90). Twenty-two structural scans were deemed unusable due to artifacts or missing files, and prenatal adversity data were not available for 11 subjects. Thus, the final sample consisted of 57 children aged 6–12 years old (boys: 30; girls: 27). The mean age of the sample was 9.28 years (SD = 1.46 years). Please refer to Table 1 for demographic information including maternal age at birth, gestational age, birth weight, duration of breastfeeding, smoking during pregnancy, maternal education, household income, maternal ethnicity and child ethnicity. For analyses that consider temperament at 18 months, an additional 5 participants were excluded (boys: 1, girls: 4) due to an incomplete data set (i.e. missing ECBQ scores at 18 months), thereby amounting to a total sample size of 52 children with a mean age of 9.28 years at scan (SD = 1.45, boys: 29, girls: 23).

thumbnail
Download:
Table 1. Sample demographics.

https://doi.org/10.1371/journal.pone.0287289.t001

Early Childhood Behaviour Questionnaire (ECBQ)

The Early Childhood Behaviour Questionnaire (ECBQ) is a 201-item parental report measure that assesses 18 dimensions of temperament between ages 18 to 36 months [71]. It was developed to address the lack of detailed and comprehensive instruments available to study temperament during the period between infancy and childhood [71]. The ECBQ was administered to mothers when their child reached 18 months. Mothers were provided with frequently occurring situations and asked to indicate how often their child demonstrates a particular behaviour using a 7-point Likert scale (1 = never,7 = always, NA = child not observed in the given situation). Total scores on the 12-item shyness subscale were used as a proxy of behavioural inhibition during toddlerhood [72]. The ECBQ subscales are internally consistent, demonstrate good inter-rater reliability, and have moderate stability across time [71].

Cumulative prenatal maternal adversity score

To obtain a measure of prenatal maternal adversity exposure, a cumulative score developed by Silveira et al (2017) was used [28]. This adversity score has been used across different cohorts and is consistently associated with poor physical and mental health outcomes [28, 73–77]. Each adverse event was coded by either its presence (1 point) or absence (0 points). For items represented with continuous scores, the 15th or 85th percentile was used as the cut-off for adding a point to the score. The total prenatal adversity score is the summation of all points. Instruments used to acquire information about various forms of adversity exposure were administered to women at 24–36 weeks of gestation and include the following:

The health and well-being questionnaire–a composite of several validated measures (short versions) [78]:

  1. A subscale from the daily hassles scale which measures the severity level (how often and to what degree) at which the woman lacked money for basic necessities such as food and electricity since the beginning of pregnancy. Responses were provided on a 3-point Likert scale (1 = somewhat severe, 2 = moderately severe, 3 = extremely severe) [79]. A score above 9 indicates a lack of money.
  2. The 9-item Marital Strain Scale of Pearlin and Schooler assesses chronic stress with a romantic partner. Women indicate how strongly they agree or disagree with various statements about non-acceptance, non-reciprocity, and frustration about role expectation using a 5-point Likert scale. Examples include (1) "I cannot completely be myself around my spouse." (2) "Generally, I give in more to my spouse’s wishes than he/she gives in to mine." [80]. A marital strain score greater than 2.9 suggests the presence of chronic stress with a romantic partner.
  3. The Abuse Assessment Screen consists of 5 items that assesses conjugal violence during pregnancy. Questions investigate the frequency, severity, perpetrator, and body sites of injury [81]. Indicating "yes" on questions 2, 3, and 4, indicate the presence of domestic violence or sexual abuse.
  4. The state version of Spielberger’s State-Trait Anxiety Inventory (STAI) assesses transient anxiety during pregnancy [82, 83]. Women are provided with 20 adjectives (e.g. "I feel tense") which are evaluated using a 4-point Likert scale (1 = not at all, 2 = somewhat, 3 = moderately, 4 = very much). An average score greater than 1.95 is suggestive of anxiety during pregnancy.
  5. Examination of the presence of current/resolved chronic disease during pregnancy (diabetes, hypertension, asthma) or severe acute conditions including vomiting, vaginal spotting or bleeding during the past 4–6 weeks, current anemia/constipation/blood in stool, and current vaginal/cervical/urinary tract infection.

Note the presence or absence of each of the above events were coded on the single item-level.

Maternal depressive symptomatology during pregnancy was evaluated using the Centre of Epidemiological Studies Depression Scale (CES-D) [84]. This 20-item self-report scale inquiries about the frequency of depressive symptoms in the past seven days allowing responses on a 3-point Likert scale (0 = never, 3 = most of the time). A score of 22 or above indicates the presence of prenatal depression. Smoking during pregnancy was coded as a binary outcome. Total household gross income below $30,000/year was scored as an adverse event. Information about birth weight and gestational age was attained from birthing units and the Canadian reference was used to calculate birth weight percentiles [85]. Birth weight percentile below the 10th percentile or above the 90th percentile in addition to a gestational age of 37 weeks or less were two additional instances of adversity.

Cortical thickness

Structural MRI image acquisition took place when children were 6–12 years old. High-resolution T1-weighted MRI images were obtained using a GE MR750 Discovery 3T MRI scanner at the Imaging Research Centre at St. Joseph’s Healthcare (Hamilton, Canada) and a 3T trio Siemens scanner in the Cerebral Imaging Center at Douglas Mental Health Institute (Montreal, Canada). The following acquisition parameters were used: Hamilton– 3D inversion recovery-prepped, FSPGR (fast spoiled gradient-echo), axial acquisition, 512 x 512 matrix with 1mm slice thickness, FOV = 24cm, TE = 3200ms, TR = 10.308ms, flip angle = 9 degrees; Montreal– 1mm isotropic 3D MPRAGE (magnetization-prepared rapid acquisition with gradient-echo), sagittal acquisition, 256 x 256 matrix, TR = 2300ms, TE = 4ms, flip angle = 9 degrees. T1-weighted images were preprocessed using the recon-all processing pipeline in FreeSurfer V6.0 (https://surfer.nmr.mgh.harvard.edu). Steps included in this pipeline involve cortical surface reconstruction, motion correction, normalization, skull stripping, registration, subcortical segmentation, and cortical parcellation. Parcellations were based on the Desikan-Killiany atlas, and a manual quality check was performed on all images using VisualQC (https://raamana.github.io/visualqc) [86]. to ensure accuracy. Cortical thickness values were extracted using the aparcstats2table command and is defined as the distance between the pial surface and white matter boundary, measured in millimeters (mm).

Genotyping

Buccal epithelial cells of children were genotyped for autosomal SNPs using genome-wide platforms PsychArray/PsychChip, Illumina. Quality control procedure was carried out using PLINK 1.951 [87, 88]. Samples with a call rate less than 90% were removed. SNPs that had a low call rate (<95%), low p-values on the Hardy-Weinberg Equilibrium (HWE) exact test (p < 1e-40), and minor allele frequency (MAF) < 5% were removed. After imputation using the Sanger Imputation Service [89] and the Haplotype Reference Consortium (HRC) as the reference panel (release 1.1), SNPs with an info score >0.80 were kept for analysis corresponding to 20,790,893 autosomal SNPs.

The population structure of the MAVAN cohort was described using principal component analysis, which was conducted on the genotyped autosomal SNPs with MAF > 5% with the following pruning parameters for linkage disequilibrium: not in high linkage disequilibrium (r2 > 0.2) across 50 kb regions and a sliding window of 5 SNPs [90, 91]. To account for population stratification, the first three PCs were included in all subsequent analyses.

Serotonin transporter co-expression based polygenic risk score (5-HTT-ePRS)

The serotonin transporter co-expression based polygenic risk score (5-HTT-ePRS) was generated by identifying genes that are co-expressed with the serotonin transporter in the amygdala. Full details of the protocol can be found in Silveira et al. (2017), Miguel et al. (2019), and Dass et al (2019) [28–30]. The 5-HTT-ePRS was created using: GeneNetwork (http://genenetwork.org), BrainSpan (http://www.brainspan.org), NCBI Variation Viewer (https://www.ncbi.nlm.nih.gov/variation/view), and The Genotype-Tissue Expression (GTEx) (https://gtexportal.org/ home/). Using GeneNework a list of genes that are co-expressed with the 5-HTT within the amygdala in mice was prepared with a threshold of absolute value of r ≥ 0.5. Further, the list of mouse genes was converted to human orthologs by matching mouse EnsemblID with human EnsemblID in Mouse Genomic Informatics (MGI) database. The list was narrowed by selecting genes that are expressed in the human brain and which are 1.5 times more present during the prenatal period as compared to adulthood. SNPs that were functionally associated with gene expression were gathered using the National Center for Biotechnology Information, U.S. National Library of Medicine (NCBI Variation Viewer)—GRCh37.p13. Based on SNP GTEx data in the human amygdala, a final list of common SNPs was reached. Linkage disequilibrium clumping (r2 < 0.25) was performed to account for correlations between SNPs and resulted in 463 SNPs included in the 5-HTT-ePRS score. To calculate 5-HTT-ePRS score, the number of alleles (genotyping data) at each cis-SNP was weighted by the estimated effect from a regression model predicting gene expression from genotype. The direction of the correlation between genes and 5-HTT gene expression was also accounted for. Summation of these values over all SNPs comprises the 5-HTT-ePRS score.

Statistical analysis

All statistical analyses were performed in R Studio (https://www.R-project.org/) [92]. Descriptive statistics for variables of interest can be found in Table 2. Student’s t-test for independent samples were conducted to detect group differences (sex, site) and Pearson’s correlational analyses allowed us to explore associations between four primary study variables–prenatal adversity, 5-HTT-ePRS score, cortical thickness values, and ECBQ shyness scores. Stepwise multiple linear regression analysis was performed to study the presence of a gene network x environment interaction that predicts an inhibited temperament at 18 months and cortical thickness at 6–12 years old. Predictor variables were the prenatal adversity score, genetic score (5-HTT-ePRS), and the interaction term. Response variables were either right mOFC thickness, left mOFC thickness, right lOFC thickness, left lOFC thickness, or the ECBQ shyness subscale scores. Models were adjusted for covariates including sex, age, and/or scanning site. The Akaike Information Criterion (AIC) was used to determine the best fitting linear model as it discourages overfitting by introducing a penalty as the number of parameters increases.

thumbnail
Download:
Table 2. Descriptive statistics.

https://doi.org/10.1371/journal.pone.0287289.t002

Results

When comparing if measures of cortical thickness differed depending on acquisition site, a borderline significant difference was found for right mOFC thickness values (t = -2.090, p = 0.047). However, after controlling for age and sex, the relationship was no longer significant (β = 0.109375, p = 0.149). No significant difference was detected in the level of maternal adversity exposure between Hamilton and Montreal samples. See Table 3.

thumbnail
Download:
Table 3. Sex and site differences in age, maternal adversity, 5-HTT-ePRS, behavioural inhibition and cortical thickness.

https://doi.org/10.1371/journal.pone.0287289.t003

A moderate, negative correlation was observed between the left lOFC and age (r = -0.320, p = 0.015). BI during toddlerhood was moderately correlated with right lOFC (r = 0.306), and left medial (r = 0.298) and lateral (r = 0.315) OFC thickness values at 6–12 years old. Increased behavioral inhibition displayed at 18 months was associated with greater cortical thickness in regions of the OFC during middle childhood. Prenatal maternal adversity was not associated with 5-HTT-ePRS score, ECBQ subscale scores at 18 months, nor OFC thickness at 6–12 years old. The 5-HTT-ePRS was not correlated with a BI at 18 months. See Table 4.

thumbnail
Download:
Table 4. Correlations between variables related to maternal adversity.

https://doi.org/10.1371/journal.pone.0287289.t004

Multiple linear regression analysis

G x E interaction predicting right mOFC and lOFC thickness in middle childhood.

For our multiple linear regression analysis with right mOFC thickness as the outcome variable, the best fitting model included the main effect of prenatal adversity, the main effect of the 5-HTT-ePRS, their interaction, and MRI scanning site (either Hamilton or Montreal) as a covariate (see Table 5). This model explained nearly 15% of the variance in right mOFC thickness (adj R2 = 0.1499, p = 0.014). Similarly, right lOFC thickness was best modelled as a function of prenatal adversity, 5-HTT-ePRS, their interaction, after adjusting for age and sex (see Table 6). Approximately 15% of the variation in thickness was explained by the aforementioned linear model (adj R2 = 0.1476; p = 0.021). The proportion of variance accounted for by the main effect of prenatal adversity, 5-HTT-ePRS, and their interaction–beyond the effects of covariates, please see Table 7.

thumbnail
Download:
Table 5. Multiple linear regression analysis results–prenatal adversity x 5-HTT-ePRS interaction model predicts right medial orbitofrontal cortex thickness.

https://doi.org/10.1371/journal.pone.0287289.t005

thumbnail
Download:
Table 6. Multiple linear regression analysis results–prenatal adversity x 5-HTT-ePRS interaction model predicts right lateral orbitofrontal cortex thickness.

https://doi.org/10.1371/journal.pone.0287289.t006

thumbnail
Download:
Table 7. Proportion of variance accounted for by the main effect of prenatal adversity, 5-HTT-ePRS, and their interaction (no covariates).

https://doi.org/10.1371/journal.pone.0287289.t007

Ultimately, results from the multiple linear regression analyses revealed a significant interaction between prenatal adversity and the 5-HTT gene network that predicts both right medial and lateral orbitofrontal cortex thickness in middle childhood (medial: β = -0.334; p = 0.041; lateral: β = -0.424, p = 0.013). To probe the exact directionality of these interactions, a post-hoc simple slope analysis was conducted. 5-HTT-ePRS scores were held at the mean, 1 SD above the mean, and 1 SD below the mean. We found that at low 5-HTT-ePRS scores (mean -1SD), higher levels of prenatal adversity exposure are associated with greater right medial (p = 0.014) and lateral (p = 0.007) orbital frontal cortices at 6–12 years old (See Figs 1 and 2).

thumbnail
Download:
Fig 1. Prenatal maternal adversity and 5-HTT-ePRS prediction of right medial OFC thickness.

In the presence of a low functioning co-expression based serotonin transporter gene network (mean -SD), there is a positive association between prenatal adversity exposure and right medial orbitofrontal cortex thickness (i.e. higher levels of prenatal adversity predict higher right medial orbitofrontal cortex thickness values).

https://doi.org/10.1371/journal.pone.0287289.g001

thumbnail
Download:
Fig 2. Prenatal maternal adversity and 5-HTT-ePRS prediction of right lateral OFC thickness.

In the presence of a low functioning co-expression based serotonin transporter gene network (mean -SD), there is a positive association between prenatal adversity exposure and right lateral orbitofrontal cortex thickness (i.e. higher levels of prenatal adversity predict higher right lateral orbitofrontal cortex thickness values).

https://doi.org/10.1371/journal.pone.0287289.g002

G x E interaction predicting left mOFC and lOFC thickness in middle childhood

No significant interaction between adversity exposure and the 5-HTT gene network that could predict left mOFC (covariates = age) or left lOFC (covariates = age and site) thickness at 6–12 years old (medial: β = -0.145, p = 0.415; lateral: β = -0.228, p = 0.162) was detected.

G x E interaction predicting BI at 18 months

The interaction between prenatal adversity and the 5-HTT gene network was significantly associated with BI at 18 months (β = 0.35, p = 0.034). However, our full model including the main effect of prenatal adversity, the 5-HTT-ePRS score, and their interaction did not reach statistical significance (adjusted R2 = 0.071, p = 0.089).

Exploratory analysis: BI at 18 months and G x E interaction predicting CT at 6–12 years

As an exploratory analysis, we decided to investigate whether the degree of BI displayed at 18 months and the gene network x environment interaction can predict cortical thickness of the OFC in middle childhood. Our multiple linear regression analysis showed that after controlling for sex and scanning site, there was a main effect of temperament (β = 0.050, p = 0.009) on right lOFC thickness at 6–12 years old and a significant interaction (β = -7.505, p = 0.005). Together, both the gene (5-HTT-ePRS) x environment (prenatal adversity) interaction and BI at 18 months predict around 24% of the variation in right lOFC thickness in middle childhood (adj R2 = 0.244, p = 0.004).

Discussion

We report that the function of the 5-HTT gene network in the amygdala moderates the impact of prenatal adversity on structural neurodevelopment in middle childhood. Specifically, in the presence of a low functioning 5-HTT gene network, there is a positive correlation between prenatal maternal adversity exposure and cortical thickness in both the right medial and lateral orbitofrontal cortices at ages 6–12 years old. The 5-HTT gene network x prenatal adversity interaction, however, did not effectively capture variation in the degree of BI exhibited during toddlerhood.

Our findings build upon previous research which has shown that the expression of 5-HTT, a primary modulator of 5-HT activity, impacts the structure and connectivity of neural circuitry (i.e. amygdala, anterior cingulate, and prefrontal regions) involved in the pathophysiology of affective disorders [18, 21]. Similarly, another line of evidence identifies that atypical neural migration and cortical development is observed upon manipulating 5-HT activity pharmacologically. Specifically, in utero exposure to SSRIs (i.e. increased 5-HT activity) has been associated with low levels of reelin, a protein important for correct neural migration [93]. Since reelin typically decreases with gestational age [93], low levels may be indicative of accelerated neurodevelopment. The above findings suggest that distinct developmental stages/processes can be disrupted depending on the period at which 5-HT availability is altered. It is thus important to attend to the developmental period at which 5-HT availability is modified, how and what factors contribute to such alterations (e.g. genetics, environment, GxE interactions, epigenetics), as well as the downstream outcomes of early signaling interruptions. Here, we report that alterations in 5-HT signaling in the amygdala during the prenatal period can affect right orbitofrontal cortical thickness development in middle childhood.

Given the recognized association between psychopathology and both 5-HT signaling and prenatal adversity exposure, it is interesting to observe an interaction between the 5-HTT gene network and prenatal maternal adversity that predicts atypical development in a region central to emotion processing and regulation. In particular, we found that a low functioning 5-HTT gene network in the amygdala moderates the impact of adverse prenatal experiences on the development of the right medial and lateral OFC such that these regions present greater thickness at higher levels of adversity 6–12 years after exposure. Our results may suggest that the OFC development may be one neural target of prenatal adversity, but the effect can be influenced by the development and communication with nearby regions. From a neurobiological standpoint, the amygdala and OFC exhibit bidirectional structural connectivity, functional connectivity, and are profoundly modulated by 5-HT signalling as they receive dense serotonergic projections [94, 95]. Both regions are also implicated in a network associated with generating and regulating emotional responses [94]. The amygdala signals information about emotional salience of environmental stimuli to which the OFC exerts top-down modulation to regulate one’s behavioural response. Decision-making, behavioural flexibility, reward processing, inhibition, sensory integration, and attention all require OFC involvement [42, 96]. Thus, it is not surprising that morphometric and functional alterations in the amygdala and OFC are putative neural correlates of many emotional and behavioural problems. For example, thinning in the OFC is linked to a variety of psychiatric conditions such as depression, anxiety, ADHD, as well as internalizing and externalizing symptoms [8, 34, 47, 97–99]. Some research has also found that cortical thickness in the vmPFC is positively associated with hemodynamic response in subcortical regions during adulthood [35], suggesting that a thinner vmPFC may be indicative of a decreased ability to regulate negative affective states. Conversely, Whittle et al., (2020) found that increases in internalizing symptoms across ages 8–10 years was associated with reduced thinning in the bilateral OFC [34]. Behaviourally, fear extinction is positively related to the thickness of the right vmPFC [100]. Together, these research findings suggest that regional differences in cortical thickness are linked to variations in certain cognitive processes and emotion regulation ability, yet there still exists discrepancies in the directionality of change. Our results add to this pool of knowledge by demonstrating that the impact of higher levels of prenatal maternal adversity on OFC thickness is moderated by functional alterations of the 5-HTT gene network within a close connected region (i.e. amygdala). Based on previous literature, such developmental change may ultimately have implications for future behavioural and emotional development.

Our results revealed a positive relationship between prenatal adversity and right OFC thickness in the presence of a low functioning 5-HTT gene network. It may be plausible that a low functioning 5-HTT gene network reflects an underlying network of genes involved in biological processes that interfere/delay typical CT maturation in middle childhood with heightened exposure to prenatal maternal adversity. Differences in maturational trajectories have been noted in the literature. For instance, the stress acceleration hypothesis posits that exposure to early life adversity promotes accelerated maturation of emotion circuits [40]. However, much evidence in support this theory stems from a body of research reporting that caregiver deprivation is associated with accelerated maturation in mPFC-amygdala circuitry [101, 102]. Specifically, negative functional connectivity between these regions is reached at an earlier timepoint [102]. It is also suggested that accelerated maturation of some circuits may come at the cost of delayed maturation in others [103]. Further studies are required to confirm whether a pattern of accelerated development upon ELA exposure extends to various structural metrics as well, if there is regional specificity, and whether acceleration is observed across various types of early adversity exposure. Multimodal imaging techniques can help elucidate whether there is congruency between structural and functional development. The findings from our study offer evidence of divergent structural maturation in the right OFC with higher levels of prenatal maternal adversity exposure and lower functioning of the 5-HTT gene network. Importantly, the unilateral hemispheric findings do corroborate with other research findings that suggest a pattern of right laterality in structural brain development upon early life stress exposure [104, 105]. Furthermore, the right prefrontal cortex is typically associated with inhibitory control [106]. Greater involvement of the right PFC is observed during tasks involving response selection and inhibition [107]. Likewise, suboptimal response inhibition in childhood is related to insufficient recruitment of the right VLPFC [107]. Evidence of PFC lateralization is also observed in psychopathologies. In particular, the right medial OFC is thinner in individuals with MDD compared to controls [108]. Similarity, a slower rate of cortical thinning in right frontal regions is associated with higher levels of inattention and hyperactivity/impulsivity during childhood [109]. Taken together, atypical right OFC development may represent a neurodevelopmental outcome of prenatal adversity exposure that could underpin aberrant inhibitory functioning. The 5-HTT-ePRS may moderate this effect of prenatal experiences on OFC thickness by influencing the progression of myelination and synaptic pruning.

The current study moves past the single-gene reductionist approach and acknowledges that genes do not code for disease states but rather, critical biological processes. Ultimately, while the 5-HTT has received much research attention as the primary regulator of 5-HT activity, it does work among a complex network of genes, transporters, receptors, their subtypes, and metabolic enzymes which collectively determine overall 5-HT signaling efficiency [19, 20, 28, 73]. It is possible that weak effects from multiple genes as opposed to a single gene polymorphism underlie atypical cognitive, emotional, behavioral, and neurodevelopmental outcomes [28]. Therefore, the gene network approach may serve utility in rectifying the discordant findings from methylation patterns and outcome associations with single genes. Instead, the gene network models multiple, critical biological processes that contribute long-term programming of early adverse experiences. Specifically, the 5-HTT gene network consists of genes involved in epigenetic modifications (e.g. DNA methylation, histone phosphorylation and methylation, chromatic remodeling) and other cellular and metabolic processes [73]. Ten of 35 genes in the network are identified as epifactors which could play a role in epigenetic processes that program the trajectory of OFC thickness development [73]. An enrichment analysis also revealed that the 5-HTT-ePRS network is highly involved in biological processes underlying nervous system development, and muscarinic acetylcholine and nicotine receptor signaling pathways [73]. Ultimately, the 5-HTT-gene network in the amygdala may be involved in biological processes that facilitate the biological embedding of adversity.

We also found that the interaction between prenatal adversity x 5-HTT-ePRS is significantly related to temperamental inhibition at 18 months. This finding indicates that there may be value in considering the moderating role of co-expression-based gene networks when studying the relationship between prenatal maternal adversity and behavioural/developmental outcomes [28]. However, our overall model including the main effects of prenatal adversity, the 5-HTT-ePRS, and the 5-HTT-ePRS x prenatal adversity interaction was not able to effectively capture differences in BI at 18 months given by a low and insignificant amount of variance explained by the predictor variables. Further, due to our small sample size, it is unclear whether the relationship between the interaction and BI will generalize to other samples. Future work should increase sample size and consider additional variables that are hypothesized to influence the degree of BI exhibited in toddlerhood. Such analyses may be better able to detect a stronger and more stable relationship between the interaction and BI at 18 months and better capture the variation. Nonetheless, our results suggest that it could be useful to consider the effect of gene network x environment interactions when studying early infant behaviour.

Our exploratory analysis revealed that the interaction between prenatal adversity and the 5-HTT gene network in addition to a main effect of BI at 18 months predicts almost a quarter (24%) of the variation in right lOFC thickness in middle childhood. The amygdala and OFC can be thought of as a network of structures subserving a behavioural spectrum ranging from disinhibition and impulsiveness to an inhibited and cautious temperament [70]. It is estimated that approximately 15–20% of children are born with extreme BI and this temperament exhibits moderate stability to adulthood [110–113]. Previous research efforts have found evidence of early temperament affecting structural development of the OFC. For instance, Hill et al (2010) note that 5-year-old children who demonstrated a heightened reaction when encountering a new child had greater right to left OFC volume ratios [70]. Such results suggest that the anatomical characteristics of the OFC may indicate a disposition towards either an inhibited or impulsive temperament. The present study extends upon this research by offering evidence that the degree of BI displayed in toddlerhood (18 months) can predict right lOFC thickness in childhood in conjunction with a 5-HTT-gene network x prenatal environment interaction. Similar results were found by Schwartz et al (2010) who observed that temperamental differences at 4 months predicted differences in left orbital and right ventromedial PFC thickness at 18 years old [114]. Hill et al (2010) also reported that BI measured at 5 years old was associated with greater right OFC volume at 15 years old [70]. It is therefore possible that OFC morphology may be one neural substrate that facilitates long-term maintenance of this temperament.

Limitations and future directions

Our research does not go without limitations and an opportunity for future work. Firstly, the results of our study are limited by small sample sizes (power ranging from 0.62–0.67) and the lack of correction for multiple comparisons. It is possible that the reported effect sizes may be inflated due to sampling bias. For these reasons, we encourage replication of this research in an independent cohort with a larger sample size.

Future studies should also account for continued exposure to adversity after birth. Understanding the balance between shared and independent variance explained by prenatal and postnatal environments would further our understanding of the conditions under which particular long-term developmental consequences ensue.

Furthermore, assessing the role of OFC thickness in mental health is needed to elucidate whether delayed maturation is protective against or poses risk for poor cognitive, behavioural, and emotional development. Currently, there is a dearth of research that explores relations between CT and pediatric and adolescent mental health [37]. Since it is reported that 50% of all mental illnesses are present by age 15 years yet remain undetected [115], identifying early indicators of psychopathology can bring us one step closer towards designing appropriately targeted and timed interventions.

Multimodal, longitudinal analyses capturing development into adulthood are also required to confirm whether the observed developmental trajectory in middle childhood leads to atypical outcomes at a later time point or whether we simply identified a temporary deviation along a normative path. In the latter case, we may expect to see differences in the rate of maturation during one developmental period followed by a return to average development further into adulthood.

Finally, since it is possible to observe accelerated maturation in some regions/networks at the cost of delayed maturation in others [103], whole brain multimodal investigations would assist in furthering our knowledge about whether ELA is associated with patterns of compensatory development.

Conclusion

Our research revealed that an interaction between the 5-HTT gene network expressed in the amygdala and prenatal maternal adversity predicts structural neurodevelopment in middle childhood. The gene network approach offers insight into potential biological processes that could lead to long-term programming of early adversity. We found that higher levels of prenatal maternal adversity exposure are associated with greater right OFC thickness values during middle childhood in the presence of a low functioning 5-HTT gene network. Future work should explore whether these cross-sectional results translate into longitudinal changes in OFC maturational trajectories. Interestingly, 5-HT signalling in the amygdala was associated with the structural development of its emotion regulatory counterpart, which helps us flexibly adapt our behaviour within changing situations. Through an exploratory analysis we also identified that temperamental inhibition at 18 months predicts right lateral OFC thickness at age 6–12 years. Together, both early temperament as well as the biological embedding of prenatal maternal adversity via the 5-HTT gene network may establish a developmental profile that programs an individual’s susceptibility to their environment.

References

  1. 1. Stiles J, Jernigan TL. The basics of brain development. Neuropsychology review. 2010 Dec;20(4):327–48. pmid:21042938
  2. 2. Glynn LM, Sandman CA. Prenatal origins of neurological development: a critical period for fetus and mother. Current Directions in Psychological Science. 2011 Dec;20(6):384–9.
  3. 3. Nelson CA, Gabard-Durnam LJ. Early adversity and critical periods: neurodevelopmental consequences of violating the expectable environment. Trends in neurosciences. 2020 Mar 1;43(3):133–43. pmid:32101708
  4. 4. Pluess M, Belsky J. Prenatal programming of postnatal plasticity?. Development and psychopathology. 2011 Feb;23(1):29–38. pmid:21262037
  5. 5. Barker DJ. Mothers, babies and health in later life. Elsevier Health Sciences; 1998.
  6. 6. Bick J, Nelson CA. Early adverse experiences and the developing brain. Neuropsychopharmacology. 2016 Jan;41(1):177–96. pmid:26334107
  7. 7. Toso K, De Cock P, Leavey G. Maternal exposure to violence and offspring neurodevelopment: a systematic review. Paediatric and perinatal epidemiology. 2020 Mar;34(2):190–203. pmid:32026500
  8. 8. Davis EP, Glynn LM, Schetter CD, Hobel C, Chicz-Demet A, Sandman CA. Prenatal exposure to maternal depression and cortisol influences infant temperament. Journal of the American Academy of Child & Adolescent Psychiatry. 2007 Jun 1;46(6):737–46. pmid:17513986
  9. 9. Van den Bergh BR, Mulder EJ, Mennes M, Glover V. Antenatal maternal anxiety and stress and the neurobehavioural development of the fetus and child: links and possible mechanisms. A review. Neuroscience & Biobehavioral Reviews. 2005 Apr 1;29(2):237–58. pmid:15811496
  10. 10. Herba CM, Glover V, Ramchandani PG, Rondon MB. Maternal depression and mental health in early childhood: an examination of underlying mechanisms in low-income and middle-income countries. The Lancet Psychiatry. 2016 Oct 1;3(10):983–92. pmid:27650772
  11. 11. Van den Bergh BR, Marcoen A. High antenatal maternal anxiety is related to ADHD symptoms, externalizing problems, and anxiety in 8‐and 9‐year‐olds. Child development. 2004 Jul;75(4):1085–97. pmid:15260866
  12. 12. Buss C, Davis EP, Muftuler LT, Head K, Sandman CA. High pregnancy anxiety during mid-gestation is associated with decreased gray matter density in 6–9-year-old children. Psychoneuroendocrinology. 2010 Jan 1;35(1):141–53. pmid:19674845
  13. 13. Sandman CA, Buss C, Head K, Davis EP. Fetal exposure to maternal depressive symptoms is associated with cortical thickness in late childhood. Biological psychiatry. 2015 Feb 15;77(4):324–34. pmid:25129235
  14. 14. Davis EP, Hankin BL, Glynn LM, Head K, Kim DJ, Sandman CA. Prenatal maternal stress, child cortical thickness, and adolescent depressive symptoms. Child development. 2020 Mar;91(2):e432–50. pmid:31073997
  15. 15. Farah MJ. The neuroscience of socioeconomic status: Correlates, causes, and consequences. Neuron. 2017 Sep 27;96(1):56–71. pmid:28957676
  16. 16. Yu Q, Daugherty AM, Anderson DM, Nishimura M, Brush D, Hardwick A, et al. Socioeconomic status and hippocampal volume in children and young adults. Developmental science. 2018 May;21(3):e12561. pmid:28464381
  17. 17. Belsky J, Jonassaint C, Pluess M, Stanton M, Brummett B, Williams R. Vulnerability genes or plasticity genes?. Molecular psychiatry. 2009 Aug;14(8):746–54. pmid:19455150
  18. 18. Nordquist N, Oreland L. Serotonin, genetic variability, behaviour, and psychiatric disorders-a review. Upsala journal of medical sciences. 2010 Feb 1;115(1):2–10. pmid:20187845
  19. 19. Oberlander TF. Fetal serotonin signaling: setting pathways for early childhood development and behavior. Journal of Adolescent Health. 2012 Aug 1;51(2):S9–16. pmid:22794534
  20. 20. Brummelte S, Mc Glanaghy E, Bonnin A, Oberlander TF. Developmental changes in serotonin signaling: Implications for early brain function, behavior and adaptation. Neuroscience. 2017 Feb 7;342:212–31. pmid:26905950
  21. 21. Homberg JR, Lesch KP. Looking on the bright side of serotonin transporter gene variation. Biological psychiatry. 2011 Mar 15;69(6):513–9. pmid:21047622
  22. 22. Booij L, Tremblay RE, Szyf M, Benkelfat C. Genetic and early environmental influences on the serotonin system: consequences for brain development and risk for psychopathology. Journal of psychiatry and neuroscience. 2015 Jan 1;40(1):5–18. pmid:25285876
  23. 23. Caspi A, Sugden K, Moffitt TE, Taylor A, Craig IW, Harrington H, et al. Influence of life stress on depression: moderation by a polymorphism in the 5-HTT gene. Science. 2003 Jul 18;301(5631):386–9. pmid:12869766
  24. 24. Pluess M, Velders FP, Belsky J, Van IJzendoorn MH, Bakermans-Kranenburg MJ, Jaddoe VW, et al. Serotonin transporter polymorphism moderates effects of prenatal maternal anxiety on infant negative emotionality. Biological psychiatry. 2011 Mar 15;69(6):520–5. pmid:21126730
  25. 25. Green CG, Babineau V, Jolicoeur-Martineau A, Bouvette-Turcot AA, Minde K, Sassi R, et al. Prenatal maternal depression and child serotonin transporter linked polymorphic region (5-HTTLPR) and dopamine receptor D4 (DRD4) genotype predict negative emotionality from 3 to 36 months. Development and psychopathology. 2017 Aug;29(3):901–17. pmid:27427178
  26. 26. Fox NA, Nichols KE, Henderson HA, Rubin K, Schmidt L, Hamer D, et al. Evidence for a gene-environment interaction in predicting behavioral inhibition in middle childhood. Psychological Science. 2005 Dec;16(12):921–6. pmid:16313653
  27. 27. Risch N, Herrell R, Lehner T, Liang KY, Eaves L, Hoh J, et al. Interaction between the serotonin transporter gene (5-HTTLPR), stressful life events, and risk of depression: a meta-analysis. Jama. 2009 Jun 17;301(23):2462–71. pmid:19531786
  28. 28. Silveira PP, Pokhvisneva I, Parent C, Cai S, Rema AS, Broekman BF, et al. Cumulative prenatal exposure to adversity reveals associations with a broad range of neurodevelopmental outcomes that are moderated by a novel, biologically informed polygenetic score based on the serotonin transporter solute carrier family C6, member 4 (SLC6A4) gene expression. Development and psychopathology. 2017 Dec;29(5):1601–17. pmid:29162172
  29. 29. Dass SA, McCracken K, Pokhvisneva I, Chen LM, Garg E, Nguyen TT, et al. A biologically-informed polygenic score identifies endophenotypes and clinical conditions associated with the insulin receptor function on specific brain regions. EBioMedicine. 2019 Apr 1;42:188–202. pmid:30922963
  30. 30. Miguel PM, Pereira LO, Silveira PP, Meaney MJ. Early environmental influences on the development of children’s brain structure and function. Developmental Medicine & Child Neurology. 2019 Oct;61(10):1127–33. pmid:30740660
  31. 31. Ducharme S, Albaugh MD, Nguyen TV, Hudziak JJ, Mateos-Pérez JM, Labbe A, et al. Trajectories of cortical thickness maturation in normal brain development—The importance of quality control procedures. Neuroimage. 2016 Jan 15;125:267–79. pmid:26463175
  32. 32. Wierenga LM, Langen M, Oranje B, Durston S. Unique developmental trajectories of cortical thickness and surface area. Neuroimage. 2014 Feb 15;87:120–6. pmid:24246495
  33. 33. Whittle S, Vijayakumar N, Dennison M, Schwartz O, Simmons JG, Sheeber L, et al. Observed measures of negative parenting predict brain development during adolescence. PloS one. 2016 Jan 29;11(1):e0147774. pmid:26824348
  34. 34. Whittle S, Vijayakumar N, Simmons JG, Allen NB. Internalizing and externalizing symptoms are associated with different trajectories of cortical development during late childhood. Journal of the American Academy of Child & Adolescent Psychiatry. 2020 Jan 1;59(1):177–85. pmid:31047992
  35. 35. Foland-Ross LC, Altshuler LL, Bookheimer SY, Lieberman MD, Townsend J, Penfold C, et al. Amygdala reactivity in healthy adults is correlated with prefrontal cortical thickness. Journal of Neuroscience. 2010 Dec 8;30(49):16673–8. pmid:21148006
  36. 36. Sowell ER, Peterson BS, Thompson PM, Welcome SE, Henkenius AL, Toga AW. Mapping cortical change across the human life span. Nature neuroscience. 2003 Mar 1;6(3):309–15. pmid:12548289
  37. 37. Jaworska N, MacMaster FP, Foster J, Ramasubbu R. The influence of 5-HTTLPR and Val66Met polymorphisms on cortical thickness and volume in limbic and paralimbic regions in depression: a preliminary study. BMC psychiatry. 2016 Dec;16:1–0.
  38. 38. Zatorre RJ, Fields RD, Johansen-Berg H. Plasticity in gray and white: neuroimaging changes in brain structure during learning. Nature neuroscience. 2012 Apr;15(4):528–36. pmid:22426254
  39. 39. Korom M, Tottenham N, Valadez EA, Dozier M. Associations between cortical thickness and anxious/depressive symptoms differ by the quality of early care. Development and Psychopathology. 2021 Oct 22:1–2. pmid:35045914
  40. 40. Callaghan BL, Tottenham N. The stress acceleration hypothesis: Effects of early-life adversity on emotion circuits and behavior. Current opinion in behavioral sciences. 2016 Feb 1;7:76–81. pmid:29644262
  41. 41. Tomalski P, Johnson MH. The effects of early adversity on the adult and developing brain. Current opinion in psychiatry. 2010 May 1;23(3):233–8. pmid:20308900
  42. 42. Hooker CI, Knight RT. The role of lateral orbitofrontal cortex in the inhibitory control of emotion. The orbitofrontal cortex. 2006 Oct 12;307:1–8.
  43. 43. Monninger M, Kraaijenvanger EJ, Pollok TM, Boecker-Schlier R, Jennen-Steinmetz C, Baumeister S, et al. The long-term impact of early life stress on orbitofrontal cortical thickness. Cerebral Cortex. 2020 Mar 14;30(3):1307–17. pmid:31504259
  44. 44. Elliott R, Dolan RJ, Frith CD. Dissociable functions in the medial and lateral orbitofrontal cortex: evidence from human neuroimaging studies. Cerebral cortex. 2000 Mar 1;10(3):308–17. pmid:10731225
  45. 45. Hodel AS. Rapid infant prefrontal cortex development and sensitivity to early environmental experience. Developmental Review. 2018 Jun 1;48:113–44. pmid:30270962
  46. 46. Lotfipour S, Ferguson E, Leonard G, Perron M, Pike B, Richer L, et al. Orbitofrontal cortex and drug use during adolescence: role of prenatal exposure to maternal smoking and BDNF genotype. Archives of general psychiatry. 2009 Nov 1;66(11):1244–52. pmid:19884612
  47. 47. Ameis SH, Ducharme S, Albaugh MD, Hudziak JJ, Botteron KN, Lepage C, et al. Cortical thickness, cortico-amygdalar networks, and externalizing behaviors in healthy children. Biological psychiatry. 2014 Jan 1;75(1):65–72. pmid:23890738
  48. 48. Peterson BS, Warner V, Bansal R, Zhu H, Hao X, Liu J, et al. Cortical thinning in persons at increased familial risk for major depression. Proceedings of the National Academy of Sciences. 2009 Apr 14;106(15):6273–8. pmid:19329490
  49. 49. Carmichael ST, Price JL. Limbic connections of the orbital and medial prefrontal cortex in macaque monkeys. Journal of comparative neurology. 1995 Dec 25;363(4):615–41. pmid:8847421
  50. 50. Matyi MA, Spielberg JM. Differential spatial patterns of structural connectivity of amygdala nuclei with orbitofrontal cortex. Human Brain Mapping. 2021 Apr 1;42(5):1391–405. pmid:33270320
  51. 51. Ray RD, Zald DH. Anatomical insights into the interaction of emotion and cognition in the prefrontal cortex. Neuroscience & Biobehavioral Reviews. 2012 Jan 1;36(1):479–501. pmid:21889953
  52. 52. Hahn A, Stein P, Windischberger C, Weissenbacher A, Spindelegger C, Moser E, et al. Reduced resting-state functional connectivity between amygdala and orbitofrontal cortex in social anxiety disorder. Neuroimage. 2011 Jun 1;56(3):881–9. pmid:21356318
  53. 53. Milad MR, Rauch SL. The role of the orbitofrontal cortex in anxiety disorders. Annals of the new York Academy of Sciences. 2007 Dec;1121(1):546–61. pmid:17698998
  54. 54. Xie C, Jia T, Rolls ET, Robbins TW, Sahakian BJ, Zhang J, et al. Reward versus nonreward sensitivity of the medial versus lateral orbitofrontal cortex relates to the severity of depressive symptoms. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging. 2021 Mar 1;6(3):259–69. pmid:33221327
  55. 55. Zhou D, Lebel C, Evans A, Beaulieu C. Cortical thickness asymmetry from childhood to older adulthood. Neuroimage. 2013 Dec 1;83:66–74. pmid:23827331
  56. 56. Shaw P, Lalonde F, Lepage C, Rabin C, Eckstrand K, Sharp W, et al. Development of cortical asymmetry in typically developing children and its disruption in attention-deficit/hyperactivity disorder. Archives of general psychiatry. 2009 Aug 1;66(8):888–96. pmid:19652128
  57. 57. Schoemaker D, Buss C, Head K, Sandman CA, Davis EP, Chakravarty MM, et al. Hippocampus and amygdala volumes from magnetic resonance images in children: Assessing accuracy of FreeSurfer and FSL against manual segmentation. Neuroimage. 2016 Apr 1;129:1–4. pmid:26824403
  58. 58. Morey RA, Petty CM, Xu Y, Hayes JP, Wagner HR II, Lewis DV, et al. A comparison of automated segmentation and manual tracing for quantifying hippocampal and amygdala volumes. Neuroimage. 2009 Apr 15;45(3):855–66. pmid:19162198
  59. 59. Kagan J, Reznick JS, Snidman N. Temperamental inhibition in early childhood. InThe study of temperament 2013 Aug 21 (pp. 53–65). Psychology Press.
  60. 60. Fox NA, Henderson HA, Marshall PJ, Nichols KE, Ghera MM. Behavioral inhibition: Linking biology and behavior within a developmental framework. Annu. Rev. Psychol. 2005 Feb 4;56:235–62. pmid:15709935
  61. 61. Morneau-Vaillancourt G, Dionne G, Brendgen M, Vitaro F, Feng B, Henry J, et al. The genetic and environmental etiology of shyness through childhood. Behavior Genetics. 2019 Jul 15;49:376–85. pmid:30877413
  62. 62. Lahat A, Hong M, Fox NA. Behavioural inhibition: Is it a risk factor for anxiety?. International Review of Psychiatry. 2011 Jun 1;23(3):248–57. pmid:21923226
  63. 63. Tang A, Van Lieshout RJ, Lahat A, Duku E, Boyle MH, Saigal S, et al. Shyness trajectories across the first four decades predict mental health outcomes. Journal of abnormal child psychology. 2017 Nov;45:1621–33. pmid:28120251
  64. 64. Arbelle S, Benjamin J, Golin M, Kremer I, Belmaker RH, Ebstein RP. Relation of shyness in grade school children to the genotype for the long form of the serotonin transporter promoter region polymorphism. American Journal of Psychiatry. 2003 Apr 1;160(4):671–6. pmid:12668354
  65. 65. Korja R, Nolvi S, Grant KA, McMahon C. The relations between maternal prenatal anxiety or stress and child’s early negative reactivity or self-regulation: a systematic review. Child Psychiatry & Human Development. 2017 Dec;48:851–69. pmid:28124273
  66. 66. Rapee RM. Preschool environment and temperament as predictors of social and nonsocial anxiety disorders in middle adolescence. Journal of the American Academy of Child & Adolescent Psychiatry. 2014 Mar 1;53(3):320–8. pmid:24565359
  67. 67. Posner MI, Rothbart MK. Developing mechanisms of self-regulation. Development and psychopathology. 2000 Sep;12(3):427–41. pmid:11014746
  68. 68. Battaglia M, Ogliari A, Zanoni A, Citterio A, Pozzoli U, Giorda R, et al. Influence of the serotonin transporter promoter gene and shyness on children’s cerebral responses to facial expressions. Archives of general psychiatry. 2005 Jan 1;62(1):85–94. pmid:15630076
  69. 69. Schmidt LA, Fox NA, Rubin KH, Hu S, Hamer DH. Molecular genetics of shyness and aggression in preschoolers. Personality and Individual Differences. 2002 Jul 19;33(2):227–38.
  70. 70. Hill SY, Tessner K, Wang S, Carter H, McDermott M. Temperament at 5 years of age predicts amygdala and orbitofrontal volume in the right hemisphere in adolescence. Psychiatry Research: Neuroimaging. 2010 Apr 30;182(1):14–21. pmid:20236805
  71. 71. Putnam SP, Gartstein MA, Rothbart MK. Measurement of fine-grained aspects of toddler temperament: The Early Childhood Behavior Questionnaire. Infant behavior and development. 2006 Jul 1;29(3):386–401. pmid:17138293
  72. 72. Holzman JB. Examining cross-lagged relations between behavioral inhibition and inhibitory control during early childhood: Predicting subsequent internalizing and externalizing problems. Northern Illinois University; 2018.
    • 73. de Lima RM, Barth B, Arcego DM, de Mendonça Filho EJ, Clappison A, Patel S, et al. Amygdala 5-HTT gene network moderates the effects of postnatal adversity on attention problems: anatomo-functional correlation and epigenetic changes. Frontiers in Neuroscience. 2020 Mar 17;14:198. pmid:32256307
    • 74. de Mendonça Filho EJ, Barth B, Bandeira DR, de Lima RM, Arcego DM, Dalmaz C, et al. Cognitive development and brain gray matter susceptibility to prenatal adversities: Moderation by the prefrontal cortex brain-derived neurotrophic factor gene co-expression network. Frontiers in Neuroscience. 2021 Nov 24;15:744743. pmid:34899157
    • 75. Batra A, Chen LM, Wang Z, Parent C, Pokhvisneva I, Patel S, et al. Early life adversity and polygenic risk for high fasting insulin are associated with childhood impulsivity. Frontiers in neuroscience. 2021 Sep 1;15:704785. pmid:34539334
    • 76. Dalmaz C, Barth B, Pokhvisneva I, Wang Z, Patel S, Quillfeldt JA, et al. Prefrontal cortex VAMP1 gene network moderates the effect of the early environment on cognitive flexibility in children. Neurobiology of Learning and Memory. 2021 Nov 1;185:107509. pmid:34454100
    • 77. de Lima RM, Barth B, Mar Arcego D, de Mendonça Filho EJ, Patel S, Wang Z, et al. Leptin receptor co-expression gene network moderates the effect of early life adversity on eating behavior in children. Communications Biology. 2022 Oct 14;5(1):1092. pmid:36241774
    • 78. Kramer MS, Goulet L, Lydon J, Séguin L, McNamara H, Dassa C, et al. Socio‐economic disparities in preterm birth: causal pathways and mechanisms. Paediatric and perinatal epidemiology. 2001 Jul;15:104–23. pmid:11520404
    • 79. Kanner AD, Coyne JC, Schaefer C, Lazarus RS. Comparison of two modes of stress measurement: Daily hassles and uplifts versus major life events. Journal of behavioral medicine. 1981 Mar;4:1–39. pmid:7288876
    • 80. Pearlin LI, Schooler C. The structure of coping. Journal of health and social behavior. 1978 Mar 1:2–1. pmid:649936
    • 81. Parker B, McFarlane J, Soeken K, Torres S, Campbell D. Physical and emotional abuse in pregnancy: A comparison of adult and teenage women. Nursing Research. 1993 May 1;42(3):173–8. pmid:8506167
    • 82. Lobel M, Dunkel-Schetter C. Conceptualizing stress to study effects on health: Environmental, perceptual, and emotional components. Anxiety Research. 1990 Nov 1;3(3):213–30.
    • 83. Spielberger CD. Manual for the state-trait anxietry, inventory. Consulting Psychologist. 1970.
    • 84. Radloff LS. The CES-D scale: A self-report depression scale for research in the general population. Applied psychological measurement. 1977 Jun;1(3):385–401.
    • 85. Kramer MS, Platt RW, Wen SW, Joseph KS, Allen A, Abrahamowicz M, et al. A new and improved population-based Canadian reference for birth weight for gestational age. Pediatrics. 2001 Aug 1;108(2):e35–. pmid:11483845
    • 86. Raamana PR, Theyers A, Selliah T. VisualQC: Assistive tools for easy and rigorous quality control of neuroimaging data. April. https://doi.org/10.5281/ZENODO. 2018 Apr;1211365.
    • 87. Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015 Dec 1;4(1):s13742–015.
    • 88. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. The American journal of human genetics. 2007 Sep 1;81(3):559–75. pmid:17701901
    • 89. McCarthy S, Das S, Kretzschmar W, Delaneau O, Wood AR, Teumer A, et al. A reference panel of 64,976 haplotypes for genotype imputation. Nature genetics. 2016. pmid:27548312
    • 90. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nature genetics. 2006 Aug 1;38(8):904–9. pmid:16862161
    • 91. Patterson N, Price AL, Reich D. Population structure and eigenanalysis. PLoS genetics. 2006 Dec;2(12):e190. pmid:17194218
    • 92. R Core Team R. R: A language and environment for statistical computing.
    • 93. Brummelte S, Galea LA, Devlin AM, Oberlander TF. Antidepressant use during pregnancy and serotonin transporter genotype (SLC6A4) affect newborn serum reelin levels. Developmental Psychobiology. 2013 Jul;55(5):518–29. pmid:22692766
    • 94. Rempel‐Clower NL. Role of orbitofrontal cortex connections in emotion. Annals of the New York Academy of Sciences. 2007 Dec;1121(1):72–86. pmid:17846152
    • 95. Roberts AC. The importance of serotonin for orbitofrontal function. Biological psychiatry. 2011 Jun 15;69(12):1185–91. pmid:21353665
    • 96. Roberts AC. Primate orbitofrontal cortex and adaptive behaviour. Trends in cognitive sciences. 2006 Feb 1;10(2):83–90. pmid:16380289
    • 97. McLaughlin KA, Sheridan MA, Lambert HK. Childhood adversity and neural development: Deprivation and threat as distinct dimensions of early experience. Neuroscience & Biobehavioral Reviews. 2014 Nov 1;47:578–91. pmid:25454359
    • 98. Luby JL, Agrawal A, Belden A, Whalen D, Tillman R, Barch DM. Developmental trajectories of the orbitofrontal cortex and anhedonia in middle childhood and risk for substance use in adolescence in a longitudinal sample of depressed and healthy preschoolers. American Journal of Psychiatry. 2018 Oct 1;175(10):1010–21. pmid:29558817
    • 99. Ducharme S, Albaugh MD, Hudziak JJ, Botteron KN, Nguyen TV, Truong C, et al. Anxious/depressed symptoms are linked to right ventromedial prefrontal cortical thickness maturation in healthy children and young adults. Cerebral cortex. 2014 Nov 1;24(11):2941–50. pmid:23749874
    • 100. Milad MR, Quinn BT, Pitman RK, Orr SP, Fischl B, Rauch SL. Thickness of ventromedial prefrontal cortex in humans is correlated with extinction memory. Proceedings of the National Academy of Sciences. 2005 Jul 26;102(30):13. pmid:16024728
    • 101. Gee DG, Gabard-Durnam LJ, Flannery J, Goff B, Humphreys KL, Telzer EH, et al. Early developmental emergence of human amygdala–prefrontal connectivity after maternal deprivation. Proceedings of the National Academy of Sciences. 2013 Sep 24;110(39):15638–43. pmid:24019460
    • 102. Tottenham N. Social scaffolding of human amygdala-mPFCcircuit development. Social neuroscience. 2015 Sep 3;10(5):489–99. pmid:26313424
    • 103. Herzberg MP, Gunnar MR. Early life stress and brain function: Activity and connectivity associated with processing emotion and reward. NeuroImage. 2020 Apr 1;209:116493. pmid:31884055
    • 104. Ocklenburg S, Korte SM, Peterburs J, Wolf OT, Güntürkün O. Stress and laterality–The comparative perspective. Physiology & behavior. 2016 Oct 1;164:321–9. pmid:27321757
    • 105. Zach P, Vales K, Stuchlík A, Cermakova P, Mrzílková J, Koutela A, et al. Effect of stress on structural brain asymmetry. Neuroendocrinol. Lett. 2016 Jan 1;37:253–64. pmid:27857040
    • 106. Horn NR, Dolan M, Elliott R, Deakin JF, Woodruff PW. Response inhibition and impulsivity: an fMRI study. Neuropsychologia. 2003 Jan 1;41(14):1959–66. pmid:14572528
    • 107. Rubia K, Taylor E, Smith AB, Oksannen H, Overmeyer S, Newman S. Neuropsychological analyses of impulsiveness in childhood hyperactivity. The British Journal of Psychiatry. 2001 Aug;179(2):138–43. pmid:11483475
    • 108. Zhang FF, Peng W, Sweeney JA, Jia ZY, Gong QY. Brain structure alterations in depression: Psychoradiological evidence. CNS neuroscience & therapeutics. 2018 Nov;24(11):994–1003. pmid:29508560
    • 109. Friedman LA, Rapoport JL. Brain development in ADHD. Current opinion in neurobiology. 2015 Feb 1;30:106–11. pmid:25500059
    • 110. Clauss JA, Blackford JU. Behavioral inhibition and risk for developing social anxiety disorder: a meta-analytic study. Journal of the American Academy of Child & Adolescent Psychiatry. 2012 Oct 1;51(10):1066–75. pmid:23021481
    • 111. Calkins SD, Fox NA, Marshall TR. Behavioral and physiological antecedents of inhibited and uninhibited behavior. Child development. 1996 Apr;67(2):523–40. pmid:8625726
    • 112. Clauss JA, Avery SN, Blackford JU. The nature of individual differences in inhibited temperament and risk for psychiatric disease: A review and meta-analysis. Progress in neurobiology. 2015 Apr 1;127:23–45. pmid:25784645
    • 113. Kagan J, Snidman N, Arcus D. Childhood derivatives of high and low reactivity in infancy. Child development. 1998 Dec;69(6):1483–93. pmid:9914634
    • 114. Schwartz CE, Kunwar PS, Greve DN, Moran LR, Viner JC, Covino JM, et al. Structural differences in adult orbital and ventromedial prefrontal cortex predicted by infant temperament at 4 months of age. Archives of general psychiatry. 2010 Jan 1;67(1):78–84. pmid:20048225
    • 115. Kessler RC, Angermeyer M, Anthony JC, De Graaf RO, Demyttenaere K, Gasquet I, et al. Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization’s World Mental Health Survey Initiative. World psychiatry. 2007 Oct;6(3):168. pmid:18188442