Structural Brain Correlates of Childhood Inhibited Temperament: An ENIGMA-Anxiety Mega-analysis

Temperament involves stable behavioral and emotional tendencies that differ between individuals, which can be first observed in infancy or early childhood and relate to behavior in many contexts and over many years.1 One of the most rigorously characterized temperament classifications relates to the tendency of individuals to avoid the unfamiliar and to withdraw from unfamiliar people, objects, and unexpected events. This temperament is referred to as behavioral inhibition or inhibited temperament (IT).2 IT is a moderately heritable trait1 that can be measured in multiple species.3 In humans, levels of IT can be quantified from the first year of life through direct behavioral observations or reports by caregivers or teachers. Similar approaches as well as self-report questionnaires on current and/or retrospective levels of IT1 can be used later in life.


STUDY SYNOPSIS Introduction Summary
Temperament involves stable behavioral and emotional tendencies that differ between individuals, which can be first observed in infancy or early childhood and relate to behavior in many contexts and over many years. 1 One of the most rigorously characterized temperament classifications relates to the tendency of individuals to avoid the unfamiliar and to withdraw from unfamiliar people, objects, and unexpected events. This temperament is referred to as behavioral inhibition or inhibited temperament (IT). 2 IT is a moderately heritable trait 1 that can be measured in multiple species. 3 In humans, levels of IT can be quantified from the first year of life through direct behavioral observations or reports by caregivers or teachers. Similar approaches as well as self-report questionnaires on current and/or retrospective levels of IT 1 can be used later in life.
Variations in IT are present on a continuous scale within the population, and research suggests that about 20% of young children are characterized by high IT, 4 which is in general stable over time. 5 Considerable data suggest that this high childhood IT (cIT) has adverse long-term consequences: infants with cIT often become more reserved adults, and, on average, such infants exhibit poorer outcomes than noninhibited infants with respect to social relationships and internalizing psychopathology. 6 More specifically, almost half of all children with elevated and stable cIT will develop social anxiety disorder later in life compared with only 12% of noninhibited children. 7 Thus, cIT predicts risk for later psychopathology, especially social anxiety disorder. 8,9 Several neuroimaging studies have examined neurobiological correlates of cIT. Such research is important, as brain characteristics-including brain structure, function, and connectivity-may mediate the cIT-related risk for poor outcomes. 10 Previous studies have linked cIT to the structure and function of brain networks involved in emotion perception, experience, and regulation. 1 These brain networks involve the dorsal (caudal) and ventral (rostral) anterior cingulate cortex, insula, amygdala, dorsolateral and medial prefrontal cortex, orbitofrontal cortex, and striatum (cf. 1,10 ), all of which have also been implicated in the familial risk for social anxiety disorder. 11 In addition, translational work on anxious temperament has indicated involvement of the hippocampus. 3,12 Despite this progress, the few available studies on the neural structural correlates of cIT are often restricted to specific regions of interest, while, to the best of our knowledge, cortical surface area and cortical thickness have been examined in only one study with an exploratory approach. 13 Furthermore, most findings with respect to brain structure are unique to a specific sample, and crossstudy comparisons are limited by relatively small sample sizes and failure to consider potential modifying variables such as age and biological sex.
In this ENIGMA-Anxiety project, 14 we aim to extend previous work by examining brain structure associated with cIT in a large dataset, assembling data acquired at 12 research centers worldwide (17 samples, N = 4,681) ( Table 1). Compared with the individual studies, this new study is better powered owing to the larger number of research participants available for analysis. Moreover, by combining data through a mega-analytic approach, the present study facilitates the differentiation of consistent, generalizable findings from falsepositive findings that could emerge from studies with smaller samples. Such work has the potential to establish reproducible anatomical correlates and could inform the development of mechanistic studies and intervention research with clinical relevance. 15 We expect to corroborate findings in brain circuits found previously (involved in processing fear, reward, and emotion regulation), 1,10 with small-to-medium effect sizes. We hypothesize that structural alterations in brain regions involved in these processes, in particular gray matter volumes of multiple subcortical structures (amygdala, hippocampus, striatum including caudate and putamen), and characteristics of several frontal and temporal cortical areas (orbitofrontal cortex, anterior cingulate cortex, insula superior temporal gyrus, transverse gyrus, fusiform gyrus) are neural substrates of cIT.

Method Summary
This ENIGMA-Anxiety Working Group project 14 will include individual participant data assembled from studies in which participants underwent magnetic resonance imaging scanning (T1-weighted anatomical magnetic resonance imaging scans) between 6 and 25 years of age. Regardless of age at the time of scanning, all participants will be phenotyped for cIT (defined as age ≤12 years). These temperament assessments could be behavioral observations in childhood, parental reports, or self-report questionnaires on current or retrospective temperament. We will perform a mega-analysis with a whole-brain approach (regional and vertex-wise; familywise error rate-corrected) 16 and investigate the relation between cIT (continuous) and 3 distinct neuroanatomical metrics (determined using FreeSurfer software [https://surfer.nmr.mgh.harvard.edu/]), namely, volumes of subcortical structures, cortical thickness, and cortical surface area. Additionally, analyses will be performed in 3 subsets, based on the method and thus age at which cIT was determined: first (early-life) behavioral observations, second parental/teacher reports during childhood, and third self-report measures acquired during late childhood/adolescence. A fourth sensitivity analysis will exclude samples with retrospective measures of cIT.

Significance Summary
This initiative is the first mega-analysis of brain structure associated with the temperamental risk for developing internalizing psychopathology. This provides the possibility of detecting novel cIT-related brain alterations and clarifying inconsistent findings of prior work. 10 Mega-analyses combine existing datasets to increase the overall sample size. This is particularly valuable for data acquired in vulnerable participants, who are often difficult to recruit. Such studies exemplify next-generation science: previous studies within the ENIGMA Consortium have resulted in important insights in the neurobiology of psychiatric conditions. 17 These discoveries reflect the advantages of large-scale data analyses for testing the reproducibility and robustness of neuroimaging findings. 17 We expect the current project to provide similar insights, increasing our understanding of the development of psychopathology in youth at risk. In addition, by preregistering the study in advance of performing the analyses, we hope to contribute to a reduction of the potential publication bias in the field and to advance a more complete scientific record on this topic (cf. 18 ).  a With respect to time point temperament assessment and MRI scan for data used in this study: C = cross-sectional; L = longitudinal.