Kent Kiehl - US grants

? DESCRIPTION (provided by applicant): Increasing public awareness has accompanied recent scientific progress understanding the relationship between mental illness and some forms of persistent antisocial behavior. This has incited calls for research into possible interventions and preventive measures. A critical barrier to research in this arena has been an outdated, descriptive taxonomy of psychiatric constructs with overlapping symptomatology and little integration of emerging knowledge from neuroscience research. An array of traits and symptoms characterized within the framework of internalizing and externalizing psychopathology are features of several psychiatric constructs common in forensic settings. It will be essential for continued progress to identify basic features of pathology that are closely aligned with specific neurobiological systems underlying domains of cognitive processing. Among these, systems governing social processing, including emotion-related cognition and perspective-taking are particularly relevant in antisocial outcomes due to psychopathology. Our research team has previously explored the domains of social-affective processing as they relate to psychopathic traits in a large, forensic male sample. Here we propose to extend this work in a female forensic sample. Further, we integrate a wider array of dimensional constructs of pathology in socio-affective processing by examining features of psychopathic traits as well borderline personality disorder. Our research strategy utilizes functional magnetic resonance imaging for the investigation of neural circuits involved in dynamic facial affective processing, inferring affective states from social situations, and emotional perspective-taking. These data will provide us with essential information about gender differences in these processes, and whether critical features of pathology are uniquely related to variation in these circuits. Furthermore, we will examine the utility of variation within these circuits to predict poor behavioral outcomes of interest including antisocial behavior, substance abuse, and suicide. Importantly, to determine key features predictive of poor outcomes, we plan to compare traditional hierarchical modeling procedures with more advanced data-driven approaches. Traditional approaches utilize regions of interest identified through prior neuroimaging work, combined with psychological traits of interest and other key demographic variables. Advanced data-driven approaches utilize Independent Component Analysis for determining key functional networks of brain activity, and utilize machine learning approaches for selecting features essential for building appropriate models. Comparing these approaches will inform our planned future efforts for developing remediation strategies and evaluating efficacy at both a neurological level as well as behavioral level. These are essential, incremental steps toward a larger translational goal to develop improved, targeted treatment strategies informed by emerging neuroscience.

The need to anticipate who will reoffend, relapse, or recover is an important responsibility of clinical and legal practitioners and a prerequisite for the provision of effective healthcare and social services to adjudicated individuals with different risk-needs. It is widely accepted that the brain plays an essential role in shaping antisocial behaviors, but the precise nature of these processes is poorly understood. Previous scholarship has raised the possibility that methods of assessing risk in adjudicated individuals could be improved by including measures of brain function along with traditional behavioral and social measures. This project examines whether the inclusion of noninvasive brain measures can enhance the ability to correctly distinguish between those inmates most and least likely to experience antisocial outcomes, such as rearrest. The project also supports the training and professional development of students in cognitive neuroscience. The results of this project are expected to support the development of clinical tools, procedures and treatments for assessing and remediating risk in forensic populations, thereby reducing the associated costs to society.

The objectives of this project are achieved by conducting the first large, out-of-sample longitudinal test of a neurocognitive model of persistent antisocial behavior. The model is developed in a sample of 600 adult criminal offenders who have undergone task-driven functional magnetic resonance imaging and been assessed for reoffense risk 12 months following release from prison (the training sample). This model is then used to classify a separate sample of 100 criminal offenders who are similarly scanned, released, and followed (the testing sample). The project employs both theory-driven (null hypothesis testing of a neurocognitive model of impulse control) and data-driven (e.g., whole-brain machine learning classification) analytical approaches. This project will advance basic and clinical knowledge of how neurobiological and behavioral risk factors interact to produce antisocial behavior by characterizing their combined and relative utility in assessing risk. This knowledge can, in turn, be used to inform the development of treatment approaches that are more sensitive to individual defendants' unique risk needs.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.