Automated Assessment of Maternal Sensitivity to Infant Distress: Leveraging Wearable Sensors for Substance Use Disorder Prevention and Research

点击翻译按钮获取中文摘要

“This study is part of the NIH’s Helping to End Addiction Long-term (HEAL) initiative to speed scientific solutions to the national opioid public health crisis. The NIH HEAL Initiative bolsters research across NIH to improve treatment for opioid misuse and addiction.” Decades of research have established the importance of early mother-infant interactions for lifelong adaptive social emotional functioning, with implications for the development of later problem behaviors including substance use disorder. A history of maternal sensitivity to infant distress – that is, responses that are consistently contingent, nurturing, and appropriate – is thought to allow infants to internalize socioemotional competencies for maintaining states of emotional security. This is reflected in the development of secure mother-infant attachment and self-regulation behaviors. In turn, mother-infant attachment security is predictive of a constellation of behaviors in childhood and adolescence – including parent-child and parent-adolescent relationship quality, childhood internalizing and externalizing behaviors, competence with peers, and school success – each of which independently predict substance abuse and substance use disorder. The objective of this proposal is to advance opportunities for research and preventative interventions for the intergenerational transmission of substance use disorders by developing and validating mobile sensor algorithms that can be used to remotely assess maternal sensitivity in both standardized protocols and everyday ecologically valid home interactions. Using audio recorders worn by infants and mother-infant motion and proximity data, we will develop models that can distinguish sensitive from insensitive maternal responses to infant distress. Critically, we develop our models with “training data” from a diverse sample of families, including families at both high- and low- risk for substance use disorders and who speak both English and Spanish. This will ensure that our tools will benefit the families who need them the most. We envision that future efforts could leverage our algorithms to identify families at greatest need for existing evidence-based interventions to improve maternal sensitivity and child outcomes. Once trained, our algorithms could be integrated into “just in time” interventions to provide real-time feedback and progress reports for mothers participating in interventions. Additionally, these algorithms will be invaluable to research examining the development of challenges in early caregiving and how such challenges can become amplified over time. For example, these tools could be used to observe the role of difficult infant characteristics, like aversive or excessive crying, maternal stressors or substance use cravings, and maternal support systems, including paternal involvement. As such, the innovative tools produced in the present proposal will both contribute to real-world public health efforts and expand research on the dynamics of early child development.