Chronic pain is a pervasive condition that involves complex neurobiological mechanisms and frequently co-occurs with mental health challenges such as anxiety, depression, and posttraumatic stress disorder (PTSD), interacting significantly with psychiatric outcomes. Despite its prevalence, the neural mechanisms of chronic pain and its interactions with mental health remain poorly understood - however, functional brain imaging studies suggest the "pain connectome", a collection of regions amongst numerous brain networks, plays a role in mediating the subjective perception of pain, and serves aspects of psychopathology. Here, we tested whether neural synchrony, a key mechanism in brain network communication, was predictive of self-reported chronic pain severity in a group of Canadian Armed Forces (CAF) service members and Veterans (n = 99) and examined if brain network functioning within the pain connectome interacts with symptoms of mental health challenges. We applied a 5-minute eyes-open resting-state magnetoencephalography (MEG) paradigm combined with multivariate modelling using partial least squares regression (PLSR). MEG is a powerful electrophysiological technique for imaging neural activity, including synchrony and network interactions, and PLSR allows exploration of complex multivariate data with high dimensionality to extract primary and interacting effects of interest in brain data. Pain severity was moderately correlated with the continuum of depression, anxiety, and PTSD symptoms, with mental health outcomes highly correlated between each other. We identified significant positive and negative associations between neural synchrony in the pain connectome and chronic pain severity. Beta oscillations were primarily related to pain severity and showed little-to-no interaction with mental health outcomes. Similarly, theta band synchrony exhibited a more specific and prominent association with the pain-anxiety interaction. Conversely, high frequency gamma synchrony was associated with chronic pain severity and showed significant effects for pain-anxiety, pain-depression, and pain-PTSD interactions. Additionally, predictive modeling using machine learning revealed that cross-spectral synchrony could reliably predict chronic pain severity in individual cases, suggesting it as a robust neurobiological marker for chronic pain. It also demonstrated moderate predictive accuracy in indexing comorbid interactions between pain and mental health. Our findings show that complex, multidimensional patterns of neural synchrony in the pain connectome mediate both chronic pain intensity and its interaction with mental health, offering new avenues for targeted and personalized therapeutic interventions and objectively tracking treatment efficacy.
