Lithium for the prevention of suicide in US veterans: a target trial emulation.

Emotion and autonomic activity are coupled. The central autonomic network (CAN) coordinates interactions between emotional behavior and autonomic signals (, , , ). While anatomical studies have located nodes of the CAN in the human brain (), how the identified regions interact with each other to represent or control emotional behavior is as yet unclear (, , , , ). One uncertainty lies in whether previously identified regions are receiving incoming information from or sending out commands to the periphery (, , , , , ). Direct stimulation studies with implanted electrodes in the human brain may provide more straightforward findings. However, even with direct stimulation, inconsistencies still exist. For example, the insula is known as the main substrate for regulating autonomic signals (), but some studies do not agree on its subregional roles (, ) and a lesion study suggested that the insula might not be critical for heartbeat sensation (). In addition, those studies are based on a small number of neurosurgical patients with a priori sets of regions such as the insula, anterior cingulate cortex (ACC), and amygdala (, , ) and provide limited interpretations about their interactions. Functional magnetic resonance imaging (fMRI) studies have developed a variety of methods to capture the cortical and subcortical representation of autonomic regulation (, , , ). Studies employed cognitive, emotional, or physical tasks that could cause changes in peripheral autonomic activity and related the evoked changes to brain blood oxygenation level dependent (BOLD) activity (). To locate autonomic control regions, researchers developed tasks involving physical and mental arousal and attributed the task-induced shifts in physiological states (e.g. heart rate or blood pressure) to the changes in brain activity (, , , , , , , ). The brain representation of incoming autonomic signals was also examined by using task scans, during which participants attended to internal sensations such as their own heartbeats and arousal states (, , ) or experienced perturbations in the baroreflex which controls heart rate to maintain steady blood pressure (, ). Whereas multiple tasks were designed to locate autonomic brain regions, tasks investigating autonomic regions associated with emotional processes are scarce (, , , , ). The fMRI studies investigating brain areas associated with afferent or efferent autonomic signals converge on a few common regions (e.g. insula and ACC; , , , ). This raises questions as to (i) whether afferent and efferent activities are controlled by those shared brain regions and (ii) whether there are alternative methods to disentangle the two activities. Challenges come from anatomical arrangements where afferent and efferent activities interact at multiple levels involving reflexes, brainstem modulation, and higher cortical functioning (, ). In fact, tasks designed to increase afferent autonomic activity often demand mental effort, which can also activate multiple effort-induced arousal (efferent) pathways and obscure the identification of the afferent regions (, ).