B3110 - Computational Models for the Prediction and Prevention of Child Traumatic Stress - 29/05/2018

B number: 
Principal applicant name: 
Glenn Saxe | NYU Langone (USA)
Constantin Aliferis, MD, PhD, FACMI
Title of project: 
Computational Models for the Prediction and Prevention of Child Traumatic Stress
Proposal summary: 

More than 20% of children will experience a traumatic event before they are 16 years old. Of those who experienced a trauma, a sizable minority will develop Posttraumatic Stress Disorder (PTSD), and other deleterious developmental, health, and psychiatric consequences (herein called Child Traumatic Stress). To diminish the considerable burden of traumatic stress on children and their families, the capacity to predict a child’s risk and to intervene to diminish this risk is extremely important. The literature on prediction of child traumatic stress from risk factors has yielded only modest results and - of those risk factors found to be predictive – it is difficult to determine which represent processes would lead to a diminution of risk, if effective intervention were applied. Almost certainly, traumatic stress results from a complex set of interacting bio-behavioral and social environmental processes, unfolding in specific ways over the course of development, and related to specific aspects of the traumatic exposure. Our project aims to apply state-of-the-art Machine Learning predictive modeling methods with a wide array of risk variables from the ALSPAC data set to generate reliable and accurate predictive models of PTSD and other child traumatic stress outcomes. We also aim to apply advanced non-experimental causal discovery algorithms to discover potentially remediable processes leading to traumatic stress outcomes that may reveal new opportunities for preventative intervention.

Impact of research: 
To create reliable, accurate and interpretable predictive models of child traumatic stress that can guide future research and clinical care. To discover remediable processes that influence traumatic stress and can inform the development of new and promising preventative interventions.
Date proposal received: 
Friday, 4 May, 2018
Statistics/methodology, Behaviour - e.g. antisocial behaviour, risk behaviour, etc., Computer simulations/modelling/algorithms, Statistical methods, Childhood - childcare, childhood adversity, Genetics, Methods - e.g. cross cohort analysis, data mining, mendelian randomisation, etc., Psychology - personality