Child Maltreatment: A New Interdisciplinary Model for Identifying Hormonal and Contextual Mediators and Assessing Interventions

This study will further develop an innovative model for studying biopsychosocial pathways to child maltreatment and identify points for testing preventive interventions.

Principal Investigator: 
Sari van Anders
Assistant Professor
School of Social Work, University of Michigan
2014 - 2015
Abstract Text: 
A caregiver's perception of aversive child behavior is a key factor in child maltreatment. We sought to further develop a biopsychosocial model for pathways to child maltreatment and identify points for preventative intervention in child abuse. To do so, we recruited 353 participants (women, n = 200; men, n = 152), of whom 40% were parents (mothers, n = 96; fathers, n = 50), to complete the Michigan Infant Nurturant Simulation Paradigm (MINSP). In the MINSP, a previously validated protocol for testing responses to infant interactions, participants were exposed to a responsive simulated infant (SI), a non-responsive SI, audio sounds of the SI, or a neutral book (van Anders, Volling and Tolman, 2012; van Anders, Tolman, and Jainagaraj, 2014). Pre- and post-manipulation, participants completed saliva samples to measure testosterone and demographic and mood measures. After completing the MINSP, participants completed a grip task, as a proxy of aggression. We aim to assess aggressive and neuroendocrine responses to the MINSP in order to identify points for testing preventative interventions of child abuse. We conducted univariate analyses of variance, using condition (responsive nurturing, non-responsive nurturing, cries only, and neutral), gender, parent status as fixed factors to examine whether there were significant differences in the percent of T-change prior to and after participation in the experimental condition. Age, exercise, season, nicotine use, and BMI were included as covariates. Contrary to our hypothesis that the nurturing conditions would have lower T change than the cries only condition, we found no significant effects for condition on T-change. We also conducted univariate analyses of variance on our proxy aggression measure—the strength of hand grip measured by dynamometer in response to an imagined infant touching a soiled diaper. In that analysis, we included age and maximum handgrip strength as covariates, and condition, gender, and parent status as fixed factors. We found that gender and percent T-change did significantly predict the strength of the hand grip. These results provide some support for the role of T in mediating aggression, though the role of the nurturance in response to infant cries must be further explored. Our research has the potential to provide profound implications for the intervention into child abuse. Specifically research about the pathway from infant crying to child abuse through endocrine mechanisms provides a base for future research to test how to ameliorate negative psychological and behavioral responses to negative infant stimuli (e.g. crying).
PI Title: 
Assistant Professor