Cortical and subcortical brain networks predict prevailing heart rate

Amy Isabella Sentis,Javier Rasero,Peter J. Gianaros,Timothy D. Verstynen

DOI: https://doi.org/10.1101/2023.09.23.559114

2024-01-07

Abstract:Resting heart rate may confer risk for cardiovascular disease (CVD) and other adverse cardiovascular events. While the brainstem’s autonomic control over heart rate is well established, less is known about the regulatory role of higher-level cortical and subcortical brain regions, especially in humans. The present study sought to characterize the brain networks that predict variation in prevailing heart rate in otherwise healthy adults. We used machine learning approaches designed for complex, high-dimensional datasets, to predict variation in instantaneous heart period (the inter-heartbeat-interval) from whole brain hemodynamic signals measured by fMRI. Task-based and resting-state fMRI, as well as peripheral physiological recordings, were taken from two datasets that included extensive repeated measurements within individuals. Our models reliably predicted instantaneous heart period from whole brain fMRI data both within and across individuals, with prediction accuracies being highest when measured within-participants. We found that a network of cortical and subcortical brain regions, many linked to psychological stress, were reliable predictors of variation in heart period. This adds to evidence on brain-heart interactions and constitutes an incremental step towards developing clinically-applicable biomarkers of brain contributions to CVD risk.

Neuroscience