Abstract:Arginine vasopressin (AVP), a neuropeptide with widespread receptors in brain regions important for socioemotional processing, is critical in regulating various mammalian social behavior and emotion. Although a growing body of task-based brain imaging studies have revealed the effects of AVP on brain activity associated with emotion processing, social cognition and behaviors, the potential modulations of AVP on resting-state brain activity remain largely unknown. Here, the current study addressed this issue by adopting a machine learning approach to distinguish administration of AVP and placebo, employing the amplitude of low-frequency fluctuation (ALFF) as a measure of resting-state brain activity. The brain regions contributing to the classification were then subjected to functional connectivity and decoding analyses, allowing for a data-driven quantitative inference on psychophysiological functions. Our results indicated that ALFF across multiple neural systems were sufficient to distinguish between AVP and placebo at individual level, with the contributing regions distributed across the social cognition network, sensorimotor regions and emotional processing network. These findings suggest that the role of AVP in socioemotional functioning recruits multiple brain networks distributed across the whole brain rather than specific localized neural pathways. Beyond these findings, the current data-driven approach also opens a novel avenue to delineate neural underpinnings of various neuropeptides or hormones.

Intranasal Vasopressin Modulates Resting State Brain Activity Across Multiple Neural Systems: Evidence from a Brain Imaging Machine Learning Study.