Heterogeneity and Harmonization

Understanding Risk Heterogeneity Following Child Maltreatment:
An Integrative Data Analysis Approach


Principal Investigator

  • Justin Russotti, Ph.D.


  • National Institute of Child Health and Human Development

Project Mentors

  • Elizabeth Handley, Ph.D.
  • Jennie Noll, Ph.D.
  • Andrea Hussong, Ph.D.

Project Collaborators

  • Erin Dunn, ScD, MPH.
  • Scott Hofer, Ph.D.
  • Dena Swanson, Ph.D.


Project Summary 

The overall objective of this project is to apply Integrative Data Analysis (IDA)—a principled set of methodologies and statistical techniques used to conduct simultaneous analysis of raw data pooled from multiple datasets—as a method to address questions about risk heterogeneity following child maltreatment. This project will use IDA and other data harmonization techniques to pool data from 9 NIH-funded child maltreatment cohorts that used gold-standard methods to examine the development of long-term maltreatment sequelae across biopsychosocial domains. Pooling original data from multiple studies stretches the developmental period under observation, generates a more heterogenous sample, and increases statistical power to examine important sources of risk heterogeneity. The harmonized sample will include over 3,000 individuals assessed on an array of biopsychosocial processes from ages 4 through 50. The IDA dataset will be used to address three aims:

  • Determine how heterogeneity in child maltreatment exposure (i.e., variation in types, developmental timing, and chronicity of exposure) differentially influences developmental sequelae.
  • Identify heterogeneity in the developmental outcome trajectories of maltreatment survivors and examine which features of childhood maltreatment exposure are associated with specific trajectories.
  • Explore how child maltreatment exposure and subsequent developmental processes differ based on racial/ethnic heterogeneity.

This project will leverage $30 million of NIH investment in child maltreatment research to unlock the constraints of isolated studies, creating a pooled source of maltreatment data that is more powerful and diverse than any individual cohort, maximizing the value of complementary efforts in the field. This contribution will be significant because it will help to parse risk heterogeneity in maltreatment survivors, which is necessary to improve the precision of our interventions..