Prenatal Polydrug Exposure: Effects of Timing on Functional Connectivity at Birth

386 

Abstract Submission 

Janelle Liu¹, Rina Eiden², Karen Grewen³, Wei Gao⁴

¹Cedars-Sinai Medical Center, Los Angeles, CA, ²Pennsylvania State University, University Park, PA, ³University of North Carolina at Chapel Hill, Chapel Hill, NC, ⁴Cedars Sinai Medical Center, Los Angeles, CA

Janelle Liu  
Cedars-Sinai Medical Center
Los Angeles, CA

Rina Eiden  
Pennsylvania State University
University Park, PA

Karen Grewen  
University of North Carolina at Chapel Hill
Chapel Hill, NC

Wei Gao  
Cedars Sinai Medical Center
Los Angeles, CA

Prenatal drug exposure (PDE) impacts infant brain development with documented long-term consequences (Ross 2015). Functional magnetic resonance imaging (fMRI) studies of infants and youth with PDE reveal aberrant brain functional connectivity (Salzwedel 2020). Animal models demonstrate that PDE timing significantly impacts offspring outcome (Byrnes 2018), but most human fMRI studies use a binary categorization to assess drug exposure, limiting the ability to detect timing effects associated with PDE. Here, we use resting-state fMRI (rsfMRI) to characterize timing-related effects of PDE on the neonatal functional connectome. For the first time, we examine neural mechanisms associated with full PDE across all three trimesters (PDE-F) and partial PDE during only the first and/or second trimester (PDE-T1T2).

Subjects included drug-free controls (CTR; n=24) and neonates with PDE (n=85). The PDE group was separated into subgroups based on whether infants experienced full exposure to any drugs across all three trimesters (PDE-F; n=49) or partial exposure to any drugs during only the first and/or second trimester (PDE-T1T2; n=21). The Timeline Follow Back (TLFB; Robinson 2014) calendar/interview was conducted to assess prenatal frequency of drug use in each trimester across 9 drug categories (alcohol, nicotine, marijuana, cocaine, opioids, medication for opioid use disorder, stimulants, depressants, other). RsfMRI scans were acquired during natural sleep at 2 weeks of age. Groups were matched on sex, race, gestational age at birth, gestational age at scan, area deprivation index, and motion. Birthweight, maternal education, and maternal depression were included as covariates of no interest in all analyses to control for group differences in these variables. Functional connectivity measures were derived using a neonate functional parcellation-based atlas (Shi 2018). For each seed region of interest (ROI; n=223), the average time series was extracted and correlated with every other ROI in the brain. Next, linear regression was conducted to detect significant functional connectivity differences between the groups. Two summary measures were used: 1) the percentage of connections showing differences at p<.05, and 2) the percentage of connections with at least a medium effect size (bias-corrected Hedge's g>.5; Nakagawa 2007). These were calculated and assigned to the seed ROI. These processes were repeated for all ROIs to generate heatmaps indicating pairwise between-group differences in functional connectivity between each ROI and the whole brain.

Compared to CTR, PDE infants showed differences in connectivity localized to visual, subcortical, parietal, and temporal areas. Compared with CTR, PDE-F infants showed differences mainly in parietal, subcortical, frontal, and sensorimotor regions, whereas PDE-T1T2 infants showed differences in visual frontal, temporal, and parietal areas. Direct comparison between PDE-F and PDE-T1T2 revealed widespread differences in temporal, subcortical, frontal, sensorimotor, and parietal regions. Consistent results were observed using the Hedge's g effect size as a threshold, larger differences were observed between PDE-T1T2 and CTR than between PDE-F and CTR.

At birth, the timing of PDE is associated with distinct effects, with PDE during the first and/or second trimester having the largest impact on early brain development. This result is surprising and may be related to the PDE-T1T2 mothers being more homogenous in their drug usage (predominantly alcohol and nicotine). By contrast, the PDE-F mothers had more heterogeneity in their drug usage, which may have resulted in smaller effects due to possible cancellation effects of different drug exposures. However, widespread functional connectivity differences between PDE-F and PDE-T1T2 point toward potential interactions between differential timing of PDE and specific neurodevelopmental processes during the prenatal period.

Neurodevelopmental/ Early Life (eg. ADHD, autism) ¹

Higher Cognitive Functions Other

Early life, Adolescence, Aging

fMRI Connectivity and Network Modeling ²

Cognition

Development

FUNCTIONAL MRI

PEDIATRIC

Other - Connectivity; Infant; Prenatal drug exposure