Delinquency, Victimization, and the Developing Brain: Results from the ABCD-Social Development Study

BARBARA TATEM KELLEY: Welcome to the National Institute of Justice webinar on Delinquency, Victimization, and the Developing Brain: Results from the ABCD, that's the Adolescent Brain and Cognitive Development, Social Development Study. My name is Barbara Tatem Kelley and I'm privileged to serve as the National Institute of Justice Social Science Research Analyst on the Adolescent Brain and Cognitive Development-Social Development Study, better known ABCD-SD. Before bringing on our three panelists to discuss the study, I would like to provide you with a very brief overview of other longitudinal research efforts supported either in the past or currently by the National Institute of Justice, NIJ, and our sister agency, the Office of Juvenile Justice and Delinquency Prevention, OJJDP.

We have a rather remarkable portfolio of research studies that focus on delinquency and crime throughout the life course of the individual which may include childhood, adolescence, and young adulthood. The first one which comes to mind is Marvin Wolfgang's seminal 1945 and 1958 Philadelphia Birth Cohort which included examination of juvenile court records to increase our understanding of who comes to the attention of the court, what is the nature, frequency, and duration of their offending, and what action is taken in response of the juvenile justice system. This study helps show the value of longitudinal research on delinquency, as it illustrated that a small proportion of all juvenile offenders were responsible for the majority of delinquency cases. OJJDP then moved beyond the juvenile court and into the school system to gather longitudinal data on students in the Seattle Social Development Project, which was led by Joe Weis and David Hawkins. And this further helped influence our understanding of how to better prevent delinquency in the community.

OJJDP then launched in the late 1980s, the program research on the causes in correlative delinquency, which involved three distinct research efforts with collaborative planning to develop and gather common core measures across three projects. These three projects included the Denver Youth Survey, which was directed by David Huizinga and Delbert Elliott, the Pittsburgh Youth Study, which looked at boys in Pittsburgh and was directed by Rolf Loeber, Magda Stouthamer-Loeber, David Farrington, and Justin Pardini. I should further note that we later funded the Pittsburgh Girl Study to also look at the girls in Pittsburgh and their delinquency courses. And the Rochester Youth Development Study was directed by Terry Thornberry, Marvin Krohn, and Alan Lizotte. These three seminal projects provide a more expansive examination of risk and protective factors influencing delinquency onset, persistence, and desistance among community samples of youth.

OJJDP then directed its attention to juvenile justice system involved youth in the following three longitudinal efforts: Pathways to Desistance, examining more serious juvenile offenders across several jurisdictions under the leadership of Ed Mulvey; the Northwestern Juvenile Project, which sampled juvenile detainees in Cook County Chicago led by Linda Teplin; and the Crossroads Study examining youth at first intake to the juvenile courts across three jurisdictions under the leadership of Elizabeth Cauffman. These three studies were particularly helpful in terms of tracking the youth engaging in delinquent behavior into their young adulthood and better understanding their strengths and needs, and the ramifications of juvenile justice system involvement and sanctions on key life outcomes.

Currently, NIJ is supporting several longitudinal initiatives including the following. We're looking at the intergenerational influences under the Northwestern Juvenile project and the Rochester Youth Development Study. So, starting with our earlier samples and now looking at their kids. We're also extending follow-up into young adulthood of the Crossroads sample study of youth offenders, and we've extended collection of individual criminal histories for the multi-cohort samples children, first studied in 1995 under the Project on Human Development in Chicago Neighborhoods. We are looking at trajectories and turning points related to delinquency and crime among juvenile justice involved females in Oregon. And finally, we're looking at the root causes of school violence with examination of interrelationships among factors at school, the family, the individual, and community levels under the Comprehensive School Safety Initiative.

NIJ recognizes it is essential to identify the risk and protective factors for delinquency involvement in order to inform and advance effective prevention and treatment efforts. This is best accomplished through a holistic assessment of the youth development which includes cognitive development. As we learned from working with experts convened by the National Academies of Science, adolescence and young adulthood is a time of key brain maturation which can influence risk of both delinquent behavior and victimization experiences. It can also influence what the juvenile justice system needs to do to better approach these youth and advance their pathways to successful adulthood.

This brings me to today's presentation on the adolescent brain and cognitive development study, social development substudy. As will be discussed in much more detail by our presenters today, the ABCD study was launched by the National Institute on Drug Abuse, NIDA, and it is a rather remarkable study in terms of five key characteristics I'm just going to touch on here. It is advancing shared data collection protocols across 21 sites, the 21 research teams. It is incorporating a broad expansive research measures, including the conduct of brain scans across the period of maturation for these youths. It is building a very large sample of youth nationwide with more than 11,000 youths involved in the study. It's archiving data annually to advance secondary analysis by the research community, and NIDA's commitment has been made for 10 years of data collection from the onset. This is an unprecedented assessment, cognitive development relationships to delinquency victimization and substance use. It offers this potential to us and its potential is recognized at the outset by Dr. Rolf Loeber who championed the implementation of the social development substudy. NIJ's support of the social development substudy is co-funded by the Centers for Disease Control and Prevention. And I am very fortunate that Chris Harper--Chris is my colleague at CDC in this collaborative effort. NIJ has also, in the past, been supportive in this effort with funding from the Office for Victims of Crime. I also had the privilege of working with the staff of NIDA, and particularly Gaya Dowling who is guiding this effort with exceptional leadership. We also work with the very capable staff of the ABCD Coordination Center at the University of California in San Diego and the 21 participating sites. Without a doubt, the most critical contributors to the social development substudy are the investigators at the University of Pittsburgh, Duncan Clark and Lia Ahonen, who have coordinated this undertaking across investigators at five of the ABCD sites. Together, Duncan and Lia have extensive research experience spanning medical, mental health, neuroscience, adolescence development, and criminology. Duncan's presentation today will provide an overview of the ABCD broader study, and Lia then will be covering the focus on delinquency and victimization under the social development study. They are joined today in this presentation by Ashley Parr who will provide more insight on why brain development matters and how neuroscience can advance our understanding of delinquency risk. Before moving to Duncan's presentation, I would also like to recognize each and every child involved in this study and their families for sharing their lives with us so that we can gather such valuable information to inform policy and practice. Thank you again for joining us today. Now to Duncan Clark.

DR. DUNCAN B. CLARK: Thank you. In today's presentation, before we discuss the social development study, I will provide an introduction to the ABCD Study. I'd like to express my appreciation for my colleagues who contributed to this presentation, including members of our Pittsburgh team, Lia Ahonen, Doug Fitzgerald, Kaylee Klingensmith, and colleagues from University of Michigan, Brian Hicks, David Clark, and Sarah Brislin. Barbara Tatem Kelley from NIJ and Chris Harper from CDC collaborate on the ABCD Social Development Project and also contribute to the ABCD Study. The efforts of the ABCD organization and funding from NIJ, CDC, NIH and other federal agencies are the foundation for this work. At the conclusion of the presentation, I will describe our ABCD Study website where you will find more information on the organization assessment protocols and publications, along with data sharing instructions.

The aims of the ABCD Study are wide-ranging, encompassing brain, cognitive, emotional, academic, and physical health development from late childhood through late adolescence. The 21 sites across the United States, recruitment was centrally directed and monitored and was primarily conducted through school systems to provide a large and diverse sample. As an illustration of this diversity, this figure depicts the household education characteristics of the ABCD sample by race and ethnicity categories. While not ideally representative of the U.S. population, the representation of some groups in ABCD supports the application of statistical methods to approximate results with a representative U.S. sample. Which is shown here by household educational characteristics from the American Community Survey. The youth participation timeline begins at age nine or ten years through the collection of an extensive array of brain imaging, cognitive testing, and psychosocial measures. One parent also participates. In addition to short interim assessments, annual assessments are conducted with the full assessment including MRI and cognitive testing conducted every two years. ABCD workgroups in specialty areas make editions and changes to the protocol to account for participants advancing development. The data I will focus on today includes information collected at the initial, or baseline, assessment, which I will label ABCD Visit 1, and the second assessment which we might call the first annual follow-up and I will label ABCD Visit 2.

The extensive ABCD assessment protocol has been organized into seven categories which I will briefly review. The mental health in this youth and parent protocols for ABCD Visit 1 are outlined here. In a few minutes, I will discuss some of these measures in more detail. Also, note that some of these measured areas are complemented by the ABCD-SD Assessment, which will be presented by Dr. Ahonen. ABCD collects information about physical health and development monitored by weight, evaluate vision, collect medical and developmental histories, conduct a traumatic brain injury screen, assess sleep and ask about activities. The culture and environment assessment collects information on parenting, family functioning, perceived school characteristics, acculturation, and identity. Biospecimen collection includes breath, oral fluid, hair, blood in some participants, and baby teeth which are requested from the parent. Other data sources include residential history and related census data, school records, and teacher report. As the ABCD Study may indicate, C is for cognition, which is thoroughly assessed. The protocol emphasizes tests of working memory and the executive function. B is for brain. And the foundation of the ABCD Study is a neuroimaging protocol. The processed MRI data provide macro and microstructural brain variables, resting state functional MRI data for connectivity analyses, and information on the functional brain responses to tasks on executive function and reward.

So, I think you can imagine that we're not going to be able to review analyses using all these ABCD measures. I will focus on just a few that I outlined here. I will describe the Brief Delinquency Measure. The data I present were collected at ABCD Visit 2 from the youth. Well, most of the analyses I will present were focused on this measure of BDM or delinquency relationships with mental health indicators. I will also illustrate other ABCD protocol areas by presenting analyses involving physical health, neurocognition, and environment characteristics. I here indicate the visit and source of these measures. Since youth self-report measures of delinquent behaviors were not collected at ABCD Visit 1, we recommended 10 items be added to Visit 2 and subsequent visits. These items represent a range of delinquency areas and severities, including hitting someone with the idea of hurting them, been unruly resulting in complaints, throwing objects at people, damaging others' property, stealing, carrying a weapon, participating in gang fights, and being arrested or picked up by police. We have examined the rates of item endorsement represented for Visit 2, which is ages 10 or 11 years old and conducted analyses on BDM scale properties, exploratory factor analyses supported a one-factor solution. The item response theory analyses provided additional information on item characteristics. The difficulty indicator, beta, corresponds to severity in this context with hitting someone or assault being the least severe and being arrested or picked up by police being the most severe. The discrimination indicator, or alpha, quantifies the degree to which the item discriminates between persons in different regions on latent continuum. These statistics indicate that the items do have a substantial range and severity and each contributes to scale variation.

A qualitative review of the BDM identified two items that may have different implications by race or ethnicity, the disorderly and police side of these involved responses by others, and these responses may be influenced by youth characteristics. To examine whether findings would be changed by deleting these items, we examine ten-item and eight-item versions. The ten-item version shown was highly correlated with the eight-item version. Correlations with risk and outcomes were also similar after deleting these two items. We conducted a differential item functioning analysis which statistically identifies items that perform differently according to participant characteristics. These DIF statistics have differences in difficulty. Overall, the analyses indicated that there were some item functioning differences between white youth and Black youth. The disorderly item did not show differential functioning by race. The police side of analysis indicated a significant difference with lower difficulty in Black youth compared to white youth. In other words, for youth at a particular level of delinquency, Black youths were more likely to endorse this item. While this data did not indicate the cause of this difference, a possible interpretation would be that for Black youth, delinquent acts are more likely to result in being detained or arrested by police. In these measured project areas, we will be further exploring approaches to identifying and minimizing item bias.

We examine BDM predictors with risk characteristics from ABCD Visit 1 youth to predict BDM at Visit 2. At the initial assessment, the ABCD Program assigns each participant to a risk category, HR referring to higher risk, indicating global risks for subsequent substance use disorders or other adverse mental health outcomes. Several items are derived from analyses for ABCD using five datasets who identified predictors of early marijuana use. Of course, the risk conscript measured by these items as well as the conscript measured by disruptive behavior disorder diagnoses overlaps with BDM. Confirming our expectations, each of these risk items, the HR categorization and disruptive behavior disorder diagnoses at Visit 1 predicted Visit 2 BDM score. Compared to disruptive behavior disorders, treat impulsivity is more distinct from delinquency. UPPS, or U-P-P-S, refers to the impulsivity traits of urgency including negative urgency and positive urgency, lack of perseverance, lack of premeditation or planning, and sensation seeking. Visit 1 youths' UPPS responses significantly predicted Visit 2 BDM. Two analysis methods are presented here. The analyst statistics used the entire Visit 2 dataset, which has over 11,000 participants. That dataset recently became available through the NIMH Data Archive as ABCD NDA, or NDA 3.0. However, these statistics do not take into consideration the fact that many participants are twins or siblings. And the results do not adjust for the ABCD sample biases when compared to a nationally representative sample.

The ABCD NDA study also provides the Data Exploration and Analysis Portal applying our base statistical modeling in a simple user interface deep provides a multilevel modeling approach, or generalized attitude mixture model, accounting for family model and site clustering. Also included is an option to utilize propensity rating, get this, that calibrates ABCD weighted distributions to nationally representative controls from the American Community Survey. At this time, deep data from the ABCD 2.0 at least includes about 4,500 participants. Similar findings resulted from the ANOVA and deep approaches.

At Visit 2, ABCD collects information about stressful and traumatic life events from parent and youth. Examples include the youth being the victim of crime or violence, a family member being arrested, and a caregiver being incarcerated where some of the adverse life events were significantly related to BDM. Trauma history is collected as part of the KSADS PTSD assessment. The items range from relatively common experiences, such as the unexpected death of a loved one, to uncommon experiences that are critical for understanding individual development, including youth threats, physical abuse, and sexual abuse. Youth PTSD was rare at Visit 1. And scale analyses to create summary variables from the trauma history are needed to advance study of the effects of these traumatic experiences. The participants' residential history is integrated with census data to drive neighborhood characteristics. ABCD generated the Area Deprivation Index with the weighted sum of area housing income and demographic variables. ADI derives from the primary residence at ABCD Visit 1 predicted BDM, although this relationship had a small effect size.

A major focus of the ABCD study and assessment protocol is a one-hour neurocognitive battery conducted with the youth at Visit 1 and then every other year. An example test is the Flanker, a test executive function measuring inhibitory control and attention. We did not find performance on this task to be related to BDM. As part of the physical health assessment, ABCD collects information on the youth's history of traumatic brain injury. It's shown here few participants less than one percent have a history of moderate or severe traumatic brain injury and TBI was not related to BDM. It is important to note that most potential participants with significant brain injury were excluded from the ABCD sample. The ABCD sample was not intended to represent heavy health problem, and despite the last sample size, is not optimal for studying some youth characteristics. I hope I have at least stimulated your interest in the ABCD study. The brief delinquency measure was shown to have adequate imperfect psychometric properties, was correlated with conceptually overlapping measures, was related to impulsivity trait that mentions adverse life events, and the Area Deprivation Index, and was not significantly related to a neurocognitive test of inhibitory control or to traumatic brain injury in this sample. Even limiting our focus to Visit 1 and Visit 2, this leaves much work to be done, developing scales and investigating important hypotheses including relationships among brain characteristics, delinquency, and victimization.

This acknowledgment statement emphasizes the investment that has been made in this research program. While we certainly look forward to exploring these reporting results ourselves, I want to encourage you all to take advantage of this resource. An excellent introduction to the project is provided by a special issue of the journal Developmental Cognitive Neuroscience. The abcdstudy.org website provides publications including 70 so far in 2020. Protocol updates, publications, news, and instructions for accessing these data are available at the project website. Over 2,000 scholars are authorized to use ABCD data. If you are interested in testing your hypotheses, I encourage you to apply for access to the ABCD data. Also feel free to contact me if I can support your efforts. Next, you will hear from Dr. Lia Ahonen. Dr. Ahonen is a Faculty Member at the University of Pittsburgh. She is a leader of the ABCD Social Development Project and she will be describing the project and reviewing some results. Dr. Ahonen, I will pass the virtual microphone to you.

DR. LIA AHONEN: So, thank you for that introduction, Duncan. And I will talk a little bit more about the Social Development Study that is a substudy of what we call the parent study, the ABCD Study. First of all, there are so many people that we need to acknowledge for helping us out starting this study and running this study. And for this particular presentation, I'd like to give a shout out to Dr. Clark, Douglas Fitzgerald, and Kaylee Klingensmith who are important team members are the coordinating site in Pittsburgh. Of course, all the investigators and staff groups at the different data collection sites. And I would especially like to acknowledge Dr. Rolf Loeber who designed the Social Development Study already in 2014 together with me. And we have a large group of experts in longitudinal studies that have advised us over the years and are still active in helping us developing the assessment protocols.

So let me first introduce the participating Social Development Sites & Faculty. Pittsburgh acts as the coordinating center, and we are collaborating with University of Michigan; University of Maryland, Baltimore; University of Florida; and Yale University. We are of course closely collaborating with the Scientific Officers at both NIJ and CDC, Barbara Tatem Kelley and Chis Harper. The Social Development Study is made up of 2,500 participants. The main areas of assessment are delinquency, victimization, and protective factors. The participants were between 10 and 12 years of age at baseline and will be followed up annually ideally for 10 years. And the specific aims of the Social Development Study were to examine the extent to which premorbid brain structure and function deficits, identify vulnerabilities for delinquency and substance use. The second aim was to investigate the adverse effects of substance use on adolescent brain development, and whether those newer developmental deficits related to both the onset and persistence of delinquency, as well as victimization. And finally, we investigate brain maturation and its associations with early forms of desistance from delinquency and cessation of victimization.

And the Social Development Study can basically be summarized in a conceptual model, shown on this slide, that consists of three overarching phases. The first phase, childhood, will include early brain anomalies and early victimization and trauma. The second phase, late childhood and adolescence, includes the main objectives of the Social Development Study. So this is where we do most of our assessments, specifically, such as for example emerging delinquency and victimization, other disruptive behaviors, and substance use. Phase three includes desistance from delinquency and cessation of victimization. And both in the second and the third phase, brain maturation is going to be a key concept. Through the three phases, we look closely at social environmental factors, such as personality disposition, emotion regulation, and upbringing strategies that may interact with brain development and externalizing behaviors. The assessment protocol for the Social Development Study was carefully developed together with the National Institute of Justice and Centers for Disease Control and Prevention, and all the site faculty, and investigators. The assessment protocol of the Social Development Study is revised annually to collect high priority data. And the purpose is to complement the delinquency and victimization data that is already collected in the ABCD Study, and also to facilitate cross-disciplinary analysis including brain scans. So the assessment protocol that we are using was carefully developed in close collaboration with the National Institute of Justice and Centers for Disease Control and Prevention, and the faculty site investigators, and the advisory group of criminological experts. The protocol is dynamic, which means that it is revised annually, and the reason is that we want to collect the highest priority data that complement the parent study, the ABCD Study, without being redundant and without adding burden to the participants. So here is a table with the main construct that the Social Development Study focuses on. And the assessment protocol consists of delinquency, information about firearms, victimization, psychopathy, prosociality, callous unemotional traits, fearlessness, pro and reactive aggression, peer behaviors, emotion regulation, and parenting strategies, and finally we also measure parents' perception about their neighborhood.

So today, I will present a few selected descriptive data of half of the baseline sample, just to give you an idea of the richness of the data and to inspire you to collaborate with us, and to use this data for your own secondary analysis. So let me first give you some of the demographics of the Social Development Study sample. The average age of assessment using the first year of baseline data was 10.8 years. This will likely increase slightly with a completed sample. Half of the sample is male and approximately four out of ten participants are white. About 40% were identified by the ABCD screener as high-risk for marijuana use. The caregivers in the Social Development Study were on average above 38 years of age, with a wide range, though, ranging from 23 to 73 years of age. Almost a fourth of the caregivers had a bachelor's degree or higher, and 7.5% did not have a high school diploma. Moving down to household income, we see that the median household income was in the category of $50,000 to $75,000 a year. And finally, at the bottom, we can see that more than a quarter of the families reported income below the national poverty limit.

And looking a little bit closer at the Social Development Study demographics and comparing them to the National Statistics from the year of 2019, when we started collecting most of the data for the SD Study, we see that the gender distribution corresponds perfectly, 51% male participants. Slightly fewer children are white than the national average, 40.6% compared to 50%. And the total median household income in the study is similar to the national average which was in 2019, $68,703. But it's still lower than the national average for households with children which national was $78,000. We also note that more than a quarter of the SD families reported incomes under the poverty line which is much higher than the national number which was 10.5%. And this is based on a four-person household, two adults and two children basically. And that number was $25,926.

So moving to some actual descriptive data of Social Development Study, we start with delinquency. And we do see that the prevalence of delinquency is similar to previous longitudinal studies. Approximately a third of the participants have engaged in vandalism or theft at some point. The prevalence of violence is--it looks remarkably high at 76%, but it's important to recognize that this construct for the presentation today includes all forms of violence, and that also includes physical conflicts with siblings, and this will of course be disentangled in future analysis. But already at this early age, on average 10.8 years, almost 6% had already had contact with the police. For carrying a hidden weapon, though, the base rate is lower, and we expect this to increase with age and based on experience from other longitudinal studies. And it's important to note that there are some significant differences between the sites, looking at the base rate of different behaviors. And this is also important to take into consideration in upcoming analysis.

So moving to victimization, the prevalence of lifetime conventional crime victimization is high, 67% across the sites. And here we don't see any significant differences between the sites. Peer and sibling victimization is also a complement, 83%, but future analysis will disentangle peers and siblings to look more closely at what types of victimization is more common. And over half of the participants have been subject to social intimidation by peers. And more than half have also been exposed to victimization indirectly or by witnessing a crime situation. And this also includes domestic violence. A small proportion of children have also experienced gun violence at this early age, and we do observe some interesting differences between data collection sites. If we look at exposure to gun violence, we can see that using half of the baseline sample, there is one site, in this case Pittsburgh, that’s--sort of sticks out from the crowd and that's something to look into more closely. A fourth of the participants, more than a quarter have experienced violence or threats to their schools and 10% have also been the subject to internet and online victimization in the form of harassment or posting pictures or other types of harassment.

I'm also going to give a few examples of things that we do measure in the social development studies. And that can be very interesting for future analysis, also for other researchers. We do know, for example, that callous unemotional traits of psychopathic personality features are associated with delinquency. We don't know, however, if and to what extent individual emotion regulation is associated with delinquency victimization. And if emotion regulation is quantitatively distinct from psychopathic traits and other personality features. So in the social development study, we are measuring both of those aspects of personality. And we start by looking at the associations between delinquency and emotional regulation. And first I need to explain that for today's purpose of this presentation I have included participants who scored in the top 25% of dysfunctional emotional regulation, calling it the worst quartile. And I'm doing this because some behavioral measures have a very skewed distribution including emotional regulation. So it makes sense to look more closely at the participants who displayed greater difficulties with emotion regulation and in practice than dichotomizing those variables. The full spectrum of emotion regulation, however, is significantly associated with all levels of theft and violence and if we break this down into the subscales of the instruments that we are using, most factors are still associated with some delinquency. And interestingly enough all the subscales, as well as the total emotion regulation score, is significantly correlated to more serious violence. Emotion regulation can also potentially be used for further analysis as a potential protective factor in presence of risk.

So if we look at the association between victimization and emotion regulation, we find that emotion regulation certainly plays a role in victimization as well as delinquency. This area however is so far unknown territory, and we will have a unique opportunity to look into this topic, also by adding other forms of measures from the ABCD dataset such as brain scans. Finally, I'd like to show you a slide about the associations between emotion regulation and one of the psychopathy scores that we use. It's called the youth psychopathy inventory and it consists of different dimensions, the impulsive and irresponsible dimension, the grandiose manipulative dimension, and the callous-unemotional dimension. And we do see here that there are associations between the youth psychopathy inventory and emotional regulation. This is really important information because there is very little knowledge about if and to what extent emotion regulation can interact with or moderate other personality features.

So with this little presentation, I'd like to emphasize a few things. And--while the ABCD-Social Development Study is a very unique development of the merging of life course and developmental criminology and newer development. The overall sample will be a recently good representation of a natural group of children in the overall population with a slight skewness toward high-risk and low socioeconomic status. Personality features having commonly investigated in many studies and one example is psychopathic traits. And we find that also emotional regulation plays an important role in explaining personality dispositions. We also found that there are important differences between data collection sites that certainly needs to be taken into consideration in future analysis. There's also word of caution when working with a new dataset of this magnitude, and, for example, there are significant differences in race, in prevalence, and frequency of delinquency, as well as victimization. And previous studies have shown that race is not necessarily a significant and independent precursor, for example, violence, when other environmental factors are taken into account. And what we are doing is that we are revising the assessment protocol annually. And we're doing this to better reflect social justice issues and inclusiveness in this study. And this is something that is very important and that the parent study, the ABCD study, is also making a high priority.

So let's talk about the future for a second. The social development study has great potential to significantly contribute to our understanding of organic brain development and how this relates to delinquent or other norm-breaking behaviors over time. We will also have the opportunity to learn more about if and how the early onset of marijuana use affects both brain development and behavior problems. And, of course, this decade, it's more important than ever to investigate those topics because of the legalization of marijuana across states and nations. And the data from this study will be deposited with the NIJ depository in Michigan and the ABCD consortium for public access.

And I'd like to conclude this presentation by saying that this is really the first study of this magnitude on brain development, delinquency victimization, and protective factors, especially with a large enough sample to differentiate race, gender, other demographic factors. And, importantly, as being one of the Dr. Rolf Loeber's student a long time ago, this is an opportunity for us to carry Rolf Loeber's legacy into the future. And to celebrate all the knowledge that he and Magda Stouthamer-Loeber have contributed to the long-term collaborations and organization of a number of longitudinal studies. And with that, I'm going to introduce Dr. Ashley Parr, who has done the very first analyses using MRI scanned data from the ABCD study and behavioral data from the social development study. Thank you.

DR. ASHLEY PARR: Good afternoon, everybody. I would just like to thank the organizers for having me. And thank Lia and Duncan for such insightful talks and for putting this symposium all together. My name is Ashley Parr. I'm a post-doc in the Laboratory of Neurocognitive Development at the University of Pittsburgh working under the mentorship of Dr. Bea Luna. And typically in our lab, we study the brain mechanisms that underly healthy adolescents and the transition to adulthood. But today, I'm going to sort of switch gears and I'm going to share some exciting new translational work that I've done in collaboration with Lia and Duncan with the ABCD-SD data. And this work gradually takes a neurodevelopmental approach to understanding the emergence and the persistence of delinquency behaviors in late childhood and adolescence.

So just to get a sense of what I'm going to cover today, I'll first discuss adolescent brain development in a broad sense. And I'll show you data that suggest that adolescence is a period of heightened brain plasticity. But it's also the period where there's increased vulnerability for the emergence of delinquency and victimization experiences. But the changeable nature of the adolescent brain also provides a critical window of opportunity for the implementation of restorative and rehabilitative practices. So next I'm going to give you some background on adolescent brain development that will help interpret some of the findings that I'm going to show on delinquency and victimization. And in particular, I'll be describing adolescence as a period of normative, healthy peaks in sensation seeking that could lead to the behavior set or characteristics of delinquency and victimization, if exacerbated. And then finally, I'm going to present some really exciting new results where we looked at differences in brain function in relation to delinquency and victimization, and in particular, we looked at dopamine and dopamine function which is the main neurotransmitter that underlies reward and as thought to sensation seeking. And the way that we measure this dopamine is using tissue-iron imaging and tissue-iron importantly here I'm talking about brain tissue-iron, and it can be measured noninvasively using MRI and I'll get into it a little bit more about the details about that in the next few slides. And then finally in our results, we'll describe some sex differences that are emerging that might point to the different risk pathways processes. Okay.

So as I've mentioned, adolescence is the period marked by increased sensation seeking which have been observed across species and societies which gives us the indication that it's a critical adolescent behavior that might be required for obtaining the skills and the experiences necessary to survive as an independent adult. However, sensation seeking can lead to increased risk taking, characteristic of delinquency behavior, and despite a--despite a peak in physical health during this period, there's a two-fold increase in mortality. And here I'm just showing three examples here on the bottom that contribute to increased risk, so on the left, we have substance use initiation which here if you just look at the center bar, you can see that it's increased in adolescence, and in the middle here we have fatal car accidents, and if you just focus on the top bar again you can see that this increases during the adolescent period, and then finally here we have the incidence of robberies and arrests during adolescence, which is particularly relevant for this talk here. Okay.

So underlying this peak in sensation seeking is thought to be a unique balance between brain systems in adolescence. And in particular the two brain systems that are the most important here for the purposes of this talk, are the prefrontal cortex which is shown here in blue, which is responsible for high level executive functions such as planning and impulse control, and then in red here I'm showing the striatum which is an area of the brain, in the middle of the brain, that is rich in dopamine and is critical for reward processing. Now during adolescence it's thought that the striatum and its activity predominates over the prefrontal cortex undermining aspects of cognition. Creating an impulse--or sorry, an increase in these striatal-driven, sensation-seeking behavior. Okay. So here I'm just showing a striatum in a different way, it's the same area that I showed in the last slide, but here I'm showing an image of MRI and as I've mentioned this area has been shown to be elevated in adolescence. So underlying function of this area is the main neurotransmitter dopamine.

So here what I'm showing is an image of a dopamine neuron, and at the top you can see one neuron that is in connection with the second neuron on the bottom of the figure. Now the gap between these two neurons is known as the synapse and these neurons communicate by passing neurotransmitters between them including dopamine, which is shown here in pink. Now MRI is fantastic in a lot of different ways but unfortunately it only allows for an indirect proxy of dopamine. So we can see increases in activity within this region, and we can infer that they're related to dopamine function, however, we just don't have that direct evidence with the MRI. Now there are methods to image dopamine directly in humans, including PET or Positron Emission Tomography, but these methods involve injecting a radioactive tracer, and there are therefore restrictions on their use in pediatric populations such as adolescence.

Now those are the problems for us as developmental neuroscientist because we know from animal studies that dopamine is changing across development, and essentially if we can't image this, we're missing out on these changes in humans. So to illustrate one aspect of these changes here, I'm showing one component of dopamine function, dopamine availability, here on the Y-axis and on the X-axis I'm showing different periods of cross development and what you can see across development is that dopamine is increasing throughout childhood into adolescence, stabilizing into early adulthood and again these findings have come largely from animal studies. So how do we understand and capture the development of the dopamine system in humans, if we can't use PET? So for this end, our laboratory uses brain tissue-iron, a Ferritin, which importantly is different from the tissue-iron you might typically think about, which is found in blood. So brain tissue-iron is critical for dopamine process and dopamine production; it's highly concentrated in the striatum, which is the area that we care about; and, most importantly, it can be measured noninvasively using common MRI sequences across the entire developmental sample. In order to ascertain whether this is a viable method for imaging dopamine in adolescence, we needed to establish two things first, first we need to know that tissue-iron is sensitive to the developmental effects across adolescence, and second we needed to know the tissue-iron corresponds to dopamine itself. So now I’ll show some data that came out of our lab, it's not ABCD data, that sort of established this as a marker. Okay.

So first one I'm showing here answers that question number one, which is, you know, is tissue-iron sensitive to development? So here on the Y-axis, I'm showing tissue-iron in the striatum, and on the X-axis I'm showing age. And what you can hopefully see is that tissue-iron concentration increases throughout adolescence, stabilizing into adulthood, and if you notice this pattern is reminiscent of what I showed you on the previous slide with dopamine availability. So here we know it's developmentally sensitive, but what about its relationship to dopamine? So to answer that, a colleague in the lab before I got there took the adult sample from this study and compared tissue-iron with our best available indices of dopamine function, which came from PET, the technique I talked to you about before. And what he showed, here I'm showing on the Y-axis, dopamine availability measured by PET and on the X-axis, I'm showing tissue-iron. And what you can see isthat as dopamine availability increases so too those tissue-irons and this relationship was significant. So here now, armed with this knowledge, we're confident that tissue-iron reflects important aspects of dopamine availability, and we can therefore then go on to use it to measure changes through adolescence that might reflect changes in dopamine.

And using this metric, we've also been able to answer the questions about the role of dopamine and tissue-iron in the development of other brain systems, like the frontal and striatal systems I talked about at the beginning. In particular, I recently shown that tissue-iron mediates developmental changes in brain connectivity specifically between the prefrontal cortex and the striatum, which is elevated in adolescence.

So based on these studies, you know, this study provides a mechanistic understanding for how sensation seeking might decrease into adulthood, so it's possible that increased dopamine function coupled with increased cross-talk between the prefrontal cortex and the striatum may sort of provide that mechanism underlying the enhanced reward function. And then decreases across the system into adulthood, may reflect diminished influence of the striatal system into adulthood facilitating the [INDISTINCT]. So this basically leads to our hypothesis for what might be happening in delinquency and related behavioral disorders. So it's possible that if the system were to develop earlier, we might see earlier peaks in striatal function that predominate over cognitive control and this might lead to exacerbated seeking and impulsivity, manifested as delinquency behaviors. And this is what we're trying to assess using the ABCD-SD. Okay.

So very briefly, I'll go over our goals for the study and what we hope to obtain, and then I'll show you some data. So the main goal, the overarching goal of this study, is to identify pre-disclosing factors early on and the neurodevelopmental patterns throughout adolescence that predict the emergence and persistence of delinquency behaviors. So measures that are obtained at timepoint one which I'm showing here, will reflect brain function either before the onset of delinquency or, you know, at the early stages of delinquency. And I can tell you already and based on, you know, what Lia what saying is that we know that some parts has been through their sample are already endorsing these behaviors, and here I've shown that visually just by coloring in some of these individuals. Now the next question that we wish to answer is whether brain maturation processes through adolescence are predicted by delinquency or desistance from delinquency after adolescence. So to do this, we'll measure brain activity and all of these, you know, SD characteristics and delinquency at several time points throughout adolescence. And then finally, we measure brain function in early adulthood which will at this point, we will sort of have a sense of who was an adolescent limited offender and who went on to continue with these maladaptive tendencies later on and so at this point we'll be able to sort of look backwards in our data to get a clear picture of what might've contributed to that. Okay. So what I'm going to be focusing on today is of course timepoint one because that's available to us and what that is going to allow us to do is look at any pre-disposing factors that might interact throughout development to increase risk. Okay.

So just to give very brief details for our participants, today I'm going to be showing data from about 600 participants, and these are participants that had both ABCD-SD data in addition to tissue-iron data. And we looked at the relationship between tissue-iron and several ABCD-SD measures, including delinquency and victimization and pretty well almost everything that Lia talked about in the previous talk. And all of the models that we conducted that I'm going to show you here included age, and sex, and parental education as co-variants in the first level analyses, and then also tested for different effects across sex. Okay.

So now, I'll show you some initial findings just in terms of the tissue-iron distributions and how they differ across sex. Okay. So here, what I'm showing is tissue-iron distributions by sex and just to walk you through this figure, here on the Y-axis, we have tissue-iron and on the X-axis, we have sex. And these dots and triangles represent the individual participants. And right next to that, I'm showing you the distributions across the entire sample. So what you can see here, and what might be more clear in this--in this bar graph with the same data here, is that males in fact have higher tissue-iron as compared to females. Now, this isn't necessarily a novel results to the study, but it might help contextualize some of the upcoming findings that I'm going to show. Okay.

So next, let us begin to ask questions about whether these tissue-iron measures contribute specific aspects of delinquency and victimization, and related characteristics. So first here, what I'm showing is the relationship between tissue-iron and delinquency, separated out by sex, so males are in blue always, females in red. And here, the X-axis reflects striatal tissue-iron, and I just simplified it here going from low to high. And the Y-axis in this plot reflects delinquency. And I'll just note that in all the plots I'm showing you higher values on a vertical axis will reflect higher scores on the instrument. And in addition here today, I'm showing you total scores across each one of these assessments. In the future, we do plan to look at different subscales but that’s beyond the scope of our talk today. So what you can see in males here is that lower indices of striatal tissue-iron corresponds to more delinquent behaviors and we don't see this a fact in females, which, you know, could be due to the relatively low number of females that are endorsing these behaviors at this stage and it might come later on in development.

Factors showing psychopathy or indices of psychopathy obtained using the psychopathy inventory as a function of striatal tissue-iron. And again here, we can see the same pattern whereby lower tissue-iron is associated with higher risk. And then finally, we also saw a similar association with our assessment measure in fearlessness, so low tissue-iron and higher on this scale. So these were the associations that we've seen in males, and we didn't see any significant associations here with females. But now, I will pivot to show you some of the effects in females that were not evident in males. Okay.

So first here, what I'm showing is the relationship between tissue-iron and aggression across the males and females. And what you can see here is a little bit of a different pattern starts to emerge, whereby in females, it would appear that higher tissue-iron in fact confers more risk for aggression. We see a similar relationship with victimization, again where higher tissue-iron is associated with more victimization, and then, finally, we see the same effect in terms of difficulties in emotional regulation. And then were no effects in these three measures in males at this current time.

So just to summarize some of the initial findings I've shown you today, what I've shown is that variability in tissue-iron which reflects dopamine function might be associated with specific aspects of delinquency and victimization in late childhood, regardless of gender differences there. So overall, males endorse more high-risk behaviors including delinquency. And in males, we see that less tissue-iron corresponds to higher risks for delinquency, fearlessness, and psychopathy. And in females, more tissue-iron confers greater risk for victimization, aggression, and difficulties with emotion regulation. So therefore, you know, at wave one already, tissue-iron might be providing a predisposing neurobiological marker that can help us understand the emergence of these behaviors and their continuation across adolescence. Although it's very speculative at this point in time, it's possible that tissue-iron has is playing an important role in male delinquency and in female victimization, although much more future work will need to interrogate the specificity of these effects across the sexes.

So what I've shared with you today are initial univariate associations between tissue-iron and specific, you know, behaviors and traits, so tissue-iron and delinquency or tissue-iron and victimization. So next, we hope to implement predictive models that will allow us to identify patterns among brain markers that correspond to different trajectories of high-risk behavior. And this might give us a sense of, you know, who might be an adolescent limited offender and who might go on to develop for a life course persistent pattern. And in future analysis once we've amassed more timepoints from the study, we'll apply longitudinal models that will identify patterns among predisposing brain markers, such as those that I showed today, and deviations from normative development across adolescence that might lead to the persistence of these high-risk behaviors into adulthood. And this very rich ABCD dataset is critical for these types of studies because if we held assessments, and the large sample size, and the longitudinal analysis will allow us to identify robust and reproducible associations between the brain, development, and the emergence and persistence of these behaviors through adolescence. And it's our ultimate goal that the results from this study will be used to inform how practitioners intervene in the lives of these youth who might be at high-risk for problematic behaviors and how to better protect individuals from the deleterious pathways such as victimization that might create dysfunction into adulthood. And, you know, the enhanced plasticity during adolescence means that this is a period where the brain is amenable to change and, you know, [INDISTINCT] neurodevelopmental approach to rehabilitative and restorative practices will really enhance the services that we're providing to young people. And I will leave you with that optimistic note and I would like to thank our funding sources and the main ABCD for all the technical support that they've provided me, in addition to my lab and my mentor, Bea Luna, who's been an invaluable resource chair. Thank you.

MARY JO GIOVACCHINI: Thank you so much, Ashley. Barbara, we can go ahead and open it up, and address the questions if you like. I would like to remind the attendees if they do have questions for the panelists to please submit them to the Q&A, which is located at the bottom right-hand portion of your screen under the little button with the three dots. And please address it to all panelists in the dropdown screen, but as you submit your question, if there is a particular panelist that you would like to have address it, if you could please list their name along with the question, that would be appreciated. The other thing too, that I'm going to leave up here is information on the panelists, their email addresses if we are unable to get to your questions in the amount of time that we have remaining, then you can go ahead and reach out to them. I will also periodically bounce back and forth between the slide with the panelists' email addresses and a “Stay Connected” slide so that you can reach out to the National Institute of Justice. So Barbara, if you want to go ahead and start the questions, you may.

BARBARA TATEM KELLEY: All right. I'll be happy to. First, I want to thank Duncan, Lia, and Ashley for their presentations and for the participants that are hanging in with us. The first question that was asked was whether or not I could provide the references for the studies that I was discussing in my introduction. I would suggest the best place to get a preliminary look at those studies, particularly the ones that have been published on is to go to the National Institute of Justice fiscal year 2020 solicitation that was titled “Longitudinal Research on Delinquency and Crime,” and most of those studies are referenced in that report. Should you have any other questions, feel free to reach out to me. The next question asked is for Lia. Can you please describe your methods for identifying your sample of interest, that being an N of 989. What were your specific inclusion, exclusion criteria? So Lia, if you could answer that question that would be great.

DR. LIA AHONEN: Absolutely. So as you could see the total sample, it's much larger. We will reach very close to 2,500 participants. We have already collected more data at this point. We are working together with the ABCD data center in San Diego, and for this presentation we could only finalize the cleaning and organizing of the first year of collecting baseline data. Baseline data and every follow-up data collection wave takes place over two years, so we have several data collection waves that are ongoing parallel. So the 989 participants presented today is the sort of the first half of the baseline time period. I hope that answers the question.

BARBARA TATEM KELLEY: Thank you, Lia. Appreciate that. It's an interesting screen to try to navigate because messages keep popping on top of questions. Oh, let's see. Sorry. All right. This question I think I'm going to ask Duncan to respond to first, and then perhaps some of the other presenters would want to jump in. Is there a plan to compare and contrast the brain development and the social development study with youth that have family incomes at national average and families below the poverty level?

DR. DUNCAN B. CLARK: Well, yeah, so we collect not only income information, but other demographic information that may indicate adversity, both on an individual family level as well as on a neighborhood level--it's, you know, that's an area that we're certainly interested in, in relationship to delinquency and victimization, but of course there are many other investigators who are examining the characteristics that influence brain development. And I would encourage the participants who are interested in that attribute to have a look at recent publications that are posted on abcdstudy.org, because there have been several publications already on diversity and brain development.

BARBARA TATEM KELLEY: Thank you, Duncan. Lia, I'd like you to address the next question, if you could. Are there discussions ongoing that using this data to determine how many--another message is covering it, I'm sorry. How early intervention efforts should be implemented in communities? How do you use this data to determine how early intervention efforts should be implemented?

DR. LIA AHONEN: Well, that is always one of the more important questions we face with longitudinal research. And, of course, there will be great opportunities to--we don't have a finalized plan for that, obviously, because we are just about preparing the start of data collection wave number three. But we would have great opportunities to look at early intervention and we hope to be able to inform intervention and science about our findings, and be a part of new developments in that area, absolutely. And the very rich data that we can also utilize from the ABCD, together with our own data in the social development study, is going to set the stage for working with prevention and intervention scientists.

BARBARA TATEM KELLEY: Thank you, Lia. Ashley, you have, kind of, I think a series about three questions coming up. And I think some of them may relate to each other, but I'll just start in piece. So the questions on the screen keep jumping around. All right. All right. The first question is, was it not true that more tissue-iron corresponded to higher risk of fearlessness in females also?

DR. ASHLEY PARR: I believe that, at that one--oh, sorry, one second here I just [INDISTINCT] I think that we saw a trend toward increase in females, but it was not statistically significant. Let me just pull that back up here and check. Yeah. So in the fearlessness one, we saw that in males the lower tissue-iron was associated with higher fearlessness. But again, in the females, we saw a trend level affect in the opposite direction, where, again, in following the pattern that we're seeing with females, higher indices of tissue-iron were actually associated with trend level increases in the fearlessness trait. Yeah.

BARBARA TATEM KELLEY: Thank you, Ashley. Now, Ashley, the next part is: the same brain marker might be responsible for opposite behavior, victimization, or delinquency. Do you have any comments on that statement?

DR. ASHLEY PARR: I think it absolutely might be at this point, that's what I can say. The way that I would like to comment on it is, you know, there are distinct in sort of interrelated relationships among these factors and their development and often their overlapping characteristics as well. So that's one of the things that we're hoping to try to understand a little bit better by using the predicted modeling. So we're hoping to basically feed in multiple brain variables and multiple behavioral sort of phenotypes or behavioral types. And, hopefully, what we can do is identify patterns among each. And that'll sort of give us a sense about what's distinct among these factors. Like, what's specific to delinquency? And what might be a shared sort of neurobiological marker that cuts across a number of these disorders.

BARBARA TATEM-KELLEY: Okay. And, Ashley, the next one coming up. Why does lower brain iron tissues as a marker for the dopamine level indicate, higher risk for delinquency in males? Is the risk-seeking behavior a result of the lower dopamine level?

DR. ASHLEY PARR: Yeah. It's actually interesting. I am still working my way through the interpretation of, you know, what the gender differences might mean. So, you know, to me, I was a little bit just surprised to see the different effects across the two sectors, you know? Yeah, at this point, that's something that, hopefully, we're going to be able to get to, but I'm not exactly sure yet.

BARBARA TATEM-KELLEY: Oh, I also wonder whether it has anything to do with the relative maturation rate of girls versus boys? Because girls may be already entering puberty, maybe boys aren't there yet. And so there could be some hormonal influences here too that you haven't had a chance to assess yet because you're at wave one. But it's an interesting question.

DR. ASHLEY PARR: Absolutely. Absolutely.

BARBARA TATEM-KELLEY: All right. Can the panelists say more about the role of protective factors? I think we should give Ashley a break. I don't know if Lia or Duncan want to take this one on.

DR. DUNCAN B. CLARK: Well, no, I think we have a pretty broad array of protective factors assessed, and they range from individual characteristics, such as personality characteristics that may provide the advantages for individuals, family characteristics like parental monitoring. You know, one of the factors that we found to be associated with substance use development is low parental monitoring. And I think the other side of the coin of effective parenting likely has a protective influence. And, as another participant raised, we have enough information and a large enough sample to look at where thresholds might be for income or other kinds of environmental adversity. And while I can't really speak specifically to these relationships in the context of the social development at this point, we do have an opportunity to look at these interactions. Another thought I had was about the--and there are interesting data--this context, we have such a large sample. That combined with the open science philosophy, I think, is going to help litigate against a publication bias. You know, often times our emerging studies have been small in size and there's been a tendency to publish results that confirm our expectations. In this context, really, other scholars are going to be checking our work, and the fact that the [INDISTINCT] to really any scholar who's interested, I think, is going to be helpful in straightening out these relationships.

BARBARA TATEM-KELLEY: Thank you, Duncan. I didn't know if Lia wanted to comment further or move to the next question. Do you have any more to add, Lia?

DR. LIA AHONEN: I can just say that many of us know that protective factors is an area that is extremely important. We know quite a lot about risk factors or problem behaviors. We know some about protective factors, and they are sometimes referred to as promoted factors. We still have a lot of work to do in that area. And we have many protective factors in terms of low or high scoring on different instruments that can represent, worst-case, risk factors and, in best-case, protective factors. We are revising the assessment protocol annually, and we are always taking into consideration adding different risk and protective factors in discussion with all our advisory experts. So that is always an area for development, and we are working actively to do just that.

BARBARA TATEM-KELLEY: Thank you, Lia. And I'd like to make one comment on Duncan's statement about publication bias and how--by making these data available for secondary analysis, there is a real emphasis on trying to avoid that scenario and really fully utilize the fruits of the labors of all the people that are doing the data collection. And this particular ABCD study is uniquely positioned to do that by making it a requirement for all the investigators not to present or publish on any data that is not accessible to the archive's other researchers for duplication, replication, expansion, and further exploration. So it is truly a unique way in which the data is being made available to the broader research community. Now, I'm going to jump back to Ashley. The question is: is it possible that by introducing or increasing a dopamine-rich diet, some of these high risk behaviors would decrease? Now, Ashley, before you answer, I just want to say, I didn't know that you could intake food in your diet that would increase your dopamine level, but is that the case? And could that change behavior?

DR. ASHLEY PARR: I would have to look into it a little bit more carefully myself to find out which foods and everything like that. And I--you know, some dopamine manipulation, in terms of, you know, to provide a therapeutic benefit, perhaps. I think that, for me, we'll have to figure out the direction, I guess, of, you know, the dopamine here, whether it's too much that's bad or too little that's bad, and, you know, go from there, I guess, in terms of trying to figure out whether we would increase it or, you know, try to decrease it in order to reduce some of these behaviors.

BARBARA TATEM-KELLEY: Thank you.

DR. ASHLEY PARR: I see another one, and it says, "Are you familiar with eye tracking/pupil dilation literature measuring dopamine levels? And if so, do you think eye tracking would be a valid and less expensive measure to use in answering your research questions?" And I'm also really excited about this talk because a lot of my backgrounds from my PhD and in my postdoc work now does use eye tracking as sort of a readout of the behavior of a decision process. And I actually am leading a line of inquiry right now to see whether blink rates and certain other eye movement metrics that we have can serve as sensitive markers for dopamine function. And so I'm playing a similar game with those as we did with the tissue iron. You know, we're comparing our blink rate and pupil levels and eye movement metrics to the PET data that we do have in the lab to try to establish those as biomarker--or, sorry, as markers for dopamine. Because, yes, it is cheaper, you know, easier often to implement eye tracking. But the good news is this ABCD data's available to me at this point, and we have this really nice measured tissue iron, so perhaps I could use, you know, both in the long run.

BARBARA TATEM KELLY: And I think in the meantime the ABCD approach to the MRI provides a lot of other information for other kind of factors to be analyzed. So there's a lot more being looked at the MRIs than just the dopamine functions, right?

DR. ASHLEY PARR: Absolutely. And I didn't get into it today because I just haven't gotten there yet, but there's a wealth of test data that would allow me to look at cognitive function and decision-making and inhibitory control and all of those really important behavioral markers for cognitive function. So I will certainly get there.

BARBARA TATEM KELLY: Thank you. So there's been ongoing discussion in the field with regards to the application of psychopathy construct to adolescents. So I think I'll ask Duncan or Lia to address that one.

DR. LIA AHONEN: So the psychopathy construct--yes, it is true. It's always a topic of some controversy, but it's also a measure that is commonly used in criminal, again, psychology. And if I understand the question correctly, it's specifically about applying this concept to children and adolescents, and that is, of course, something that shouldn't be done lightly and without very careful consideration. We have decided in this study to use this construct. We have an expert group in our study. Some of the investigators, especially in Michigan, are very, very careful in implementing the instruments and items used to sort of look at personality disposition in general. And we don't rely solely on psychopathy measures. We have many other aspects of personality dispositions. So the terms that we are using are carefully considered. And this is also an ongoing discussion within criminology, as we know: is it appropriate to use those terms and assessment protocols on very young people? Our investigators, the experts on psychopathy measures, don't just look at psychopathy as a whole construct. It's broken down carefully in dimensions and different scales, and then combined with other personality features as well.

BARBARA TATEM KELLY: Thank you, Lia. The next question is also for you. What are the inclusion-exclusion criteria for the entire sample of interest? How are you deciding who to include from the parent sample?

DR. LIA AHONEN: So we have first selected which sites from the parent study that we are going to collaborate with, and the selection criteria has basically been--I mean, the ABCD by itself is a very large logistically-challenging study, so we had to work with sites that had the infrastructure to be able to take on another demanding substudy. Once those sites were chosen and had expressed interest to participate, we made the decision to make every active ABCD participants and their families eligible for participating in the social development study. So, basically, anyone from the ABCD sites that are included in the project is eligible as long as they are active--ABCD study.

DR. DUNCAN B. CLARK: I think I'll comment on that as well. From the standpoint of the overall ABCD sample, our attempt was to include any children who would be able to complete the protocol. And, of course, it's a fairly demanding process to participate in the study. We made great efforts to help children get through it. And I'm, on a daily basis, make these steps so that people complete all of this. But it is the case that some children would be able to complete the process, and one example is from the slide I've shown about having so few participants with severe traumatic brain injury. It would typically be the case that not only the brain imaging procedures but some of the tasks wouldn't be possible for a participant with severe traumatic brain injury. So there are some, you might say, gaps in the sample that do influence the areas that we can study, but we did try to be as inclusive as possible.

BARBARA TATEM KELLY: Thank you, Duncan. Ashley, next is for you. Question asking for confirmation. Were you using gender and sex synonymously?

DR. ASHLEY PARR: I should've used sex. I will say that we have both, and I have the ability to look at both. I think that, at this point, when I originally looked, there was good correspondence between the two, but, specifically, for the purpose of this talk, I was using sex, biological sex. So I apologize if I said gender.

BARBARA TATEM KELLY: Thank you. And then, Ashley, the next question. Do you find the results of tissue iron and delinquency associations counterintuitive? I was thinking that high dopamine activity would indicate higher sensation seeking and resulting delinquency. You know, I had that same question, Ashley, but I didn't write that.

DR. ASHLEY PARR: They do seem counterintuitive, but, you know, the differences that we're seeing across males and females as we sort of--as you mentioned, Barbara, before, might reflect a maturation story. And one of--you know, they might be at different stages. So, you know, typically, I might think that, yes, higher indices would be related to more sensation-seeking behavior, but there could be some nuances there, you know, and interrelationships that I haven't quite considered yet. So, you know, yeah, interpreting that in relation to, you know, dopamine function and too much or too little is definitely an area that I'm working through.

BARBARA TATEM KELLY: Thank you. And it just jumped again right when I wanted to read. Here we go. Are there studies related to brain development and the potential impact by and from juvenile detention and/or other institutionalization? I would like to just mention again that I did reference the National Academies of Sciences' work which focused on adolescent development, cognitive development during that timeframe, and how that might influence or impact the way the whole juvenile justice system operates, how we proposed delinquency prevention and intervention in ways to hurry on the system. So, that's one study that was done with a lot of different experts. And a report was published on that. And there's also quite a bit of discussion by Elizabeth Kaufman who is directing the Crossroads study that I mentioned--crossroads study--on looking at youth coming into the juvenile justice system, and she frequently in her presentations speaks to the issue of brain development and cognitive development in the adolescents areas. So that's another area that you might want to consider. But now I want to open up that question to the actual presenters. Are there any studies related to brain development and the potential impact by and from juvenile detention and/or other institutionalization? Did you have anything you'd like to add?

DR. DUNCAN B. CLARK: Well, I would say for the ABCD study, this is a topic we have discussed. The ages of the participants currently are a little young to yield to address this issue. But in the end, following these participants for another eight or ten years, this is a topic that we expect to be able to address.

BARBARA TATEM KELLY: Okay. Anything you wanted to add, Lia?

DR. LIA AHONEN: I think part of the question was also whether there are studies doing just that already. And I think--Duncan, you can jump in if you know of anything. But I am not aware that are any studies who have already addressed this issue, in terms of also using investigations of the actual organic brain development. So I would think that that's going to be a very unique feature in the ABCD study in the future when the children gets older, and we can actually look into out-of-home placements in different contexts.

BARBARA TATEM KELLY: Okay. Thank you. I just had a moment here where everything froze in my question and answers. So I've got to get back on track. If anyone else has the next question up and rolling, they could ask it and answer it.

MARY JO GIOVACCHINI: Barbara, I can ask the question for you.

BARBARA TATEM KELLEY: Thank you.

MARY JO GIOVACCHINI: I believe this is the next one in line. Were any of the participants on the psychotropic drugs have a psychiatric diagnosis?

DR. DUNCAN B. CLARK: Well, you know, I can speak to the overall ABCD sample. And the participants that are participating in the social development study I think are reasonably representative. We do collect information on psychiatric disorders. We have a fairly extensive assessment that utilizes what a measurement is often--or an interview that's often, just in clinical studies. The presentation of the interview is by--a computer presentation with assistance from an assessor. And we get confirmation of those psychiatric characteristics through other subreport measures. We also collect information on medications that children are taking. Now, we don't exclude participants based on psychiatric medication so we have quite a few participants who are taking, for example, different types of ADHD medications who are participating in the study, antidepressant medications, and so forth. That information is collected, and we're working on compiling it in a way that we'd be able to readily analyzed so even though that information doesn't yet appear in the dataset, it's something that we collect and are working on.

BARBARA TATEM KELLEY: I think the next question might be is there any comparable study of this kind in another country?

DR. DUNCAN B. CLARK: There is a European study that goes by the name IMAGEN, and I believe that study started at age 14 and has gone through to late adolescents. So one of our investigators in ABCD, Hugh Garavan, was one of the leaders of the Imaging study. And I think that provides information that can guide us in this study, as well as in some of the topics that we've been discussing.

BARBARA TATEM KELLEY: The next question is can we have the web link for this project-related all publications? I think that you've posted in your PowerPoint the link to the ABCD site that has a number of these publications.

DR. DUNCAN B. CLARK: Right. I'll just speak it out. It's abcdstudy.org. And there's about 150 publications posted.

BARBARA TATEM KELLEY: Are any studies related to brain development and the potential impact by from juvenile detention or other institutionalization? I think we kind of addressed part of this but now the question is longer. In other words, is there work on emotional depravations, the influence on cognition, and emotional regulation? So I think they've added on to an earlier question, in terms of emotional depravation, the influence on cognition, and emotional regulation.

DR. ASHLEY PARR: I can say--I think that there likely is. I'm not entirely familiar with the literature, but certainly not, you know, at this large scale that we have here.

BARBARA TATEM KELLEY: And then there was another question on whether or not any of these subjects were on psychopathic drugs or had psychiatric diagnoses.

DR. ASHLEY PARR: We covered that one.

BARBARA TATEM KELLEY: Okay. I think that's when I was scrolling. I think I had finished with the questions that have come in so far, which is pretty amazing, and we've got four minutes left. I do know that one of the things that I've discussed extensively with Duncan and Lia is the annual process they're going through to update the protocols on their data collection, particularly with the victimization and delinquency items, to make sure these kids mature, that we are addressing issues of concerns to the appropriate age group. So getting developmentally appropriate questions on these topics across the age range from, say, 10 or 11 to 20 or 21, need some adjustment from year to year. I don't know if Duncan and Lia would care to comment on that very briefly, what they're looking at for the future.

DR. DUNCAN B. CLARK: Well, we have an annual review of the protocol to identify areas that we believe are necessary to add or change in the protocol. And, as I think you alluded to, we're engaged in that process right now. It's not only one that involves the investigators from the participating sites of the social development study but also the ABCD Committee. So we do get a lot of input on the areas that we assess in the substudy. In the overall ABCD study, we have work groups focused on each of the seven areas that I've mentioned. So those work groups meet on an ongoing basis as do the SD sites. So we do scrutinize the assessments and work pretty hard to have each of the points be developmentally appropriate.

BARBARA TATEM KELLEY: Thank you. So I wanted to make a couple of comments. I didn't want to kind of derail the focus of this presentation at the onset by mentioning that we have a global pandemic underway, but this has been something that has been a focus on a lot of discussions among the council of investigators for the ABCD study, in trying to deal with issues that many of you in the audience are probably experiencing as well, in terms of restrictions or concerns about in-person data collection, in-person service delivery, in many of your cases. And so the ABCD Council has been working on a case by case basis about trying to have an overall strategy for the project to, as much as possible, continue to harvest the data they need while still maintaining the safety and security of all their staff, as well as the participants within, of course, the restrictions of their local universities.

And then the other thing I wanted to mention, if you're thinking about exploring the data for ABCD, the University of California and San Diego coordinating group has also worked with the council of investigators to provide training to people that are interested in utilizing these data, and I think they have archived different modules on how to access and how to utilize and how to conduct analysis of what is a very complicated and extensive dataset. And they are trying to build into the archive easier ways for scientists to access the data with different ways to approach analysis. So do take advantage of those kinds of guidance and services if you proceed with looking at the data. And with that, it is 4:00, and I need to thank the presenters. Duncan, Lia, and Ashley, as well as I want to thank Mary Jo, and Michelle, and Daryl who have been providing the support for this event. I am known as one of the most technologically non-advanced people in my office. Everybody else is younger and better at it. So I really appreciate all the support you have provided to us to present this, and I appreciate everyone's attendance, and hope to hear more about how this project is progressing as we move forward. Thank you so much.

MARY JO GIOVACCHINI: Barbara, thank you very much. And I know you ended this, there is one question that came up. So, Ashley, it is for you. Have you considered imagining the orbital cortex for future experiments as well? Perhaps this could be--excuse me. Part of a multiple investigation in tandem with your original paradigm?

DR. ASHLEY PARR: That's a great question. And actually--so that's our next step. So, you know, we first--step one with tissue iron. And then in my work, in healthy adolescents, I've shown that tissue iron is really involved in communication. So connectivity between the striatum and in particular the ventromedial prefrontal cortex which is where the orbitofrontal cortex is. And so now--you know, first stage was to characterize tissue iron, and the next stage is to look at connectivity between, like I said, the striatum in the orbitofrontal cortex, because absolutely that circuit is critical for reward and for value-guided decision making. And we know from our studies that it is maturing in adolescents. It's actually decreasing in healthy adolescents. So it's a great question because that's certainly our next target.

MARY JO GIOVACCHINI: Ashley, thank you so much. Now we can officially end the webinar. Thank you all for attending. Have a wonderful day.

DR. ASHLEY PARR: Thank you.