Autonomic responses to emotional linguistic stimuli and amplitude of lowfrequency fluctuations predict outcome after severe brain injury

An accurate prognosis on the outcome of brain-injured patients with disorders of consciousness (DOC) remains a
significant challenge, especially in the acute stage. In this study, we applied a multiple-technique approach to
provide accurate predictions on functional outcome after 6 months in 15 acute DOC patients.
Electrophysiological correlates of implicit cognitive processing of verbal stimuli and data-driven voxel-wise
resting-state fMRI signals, such as the fractional amplitude of low-frequency fluctuations (fALFF), were employed.
Event-related electrodermal activity, an index of autonomic activation, was recorded in response to
emotional words and pseudo-words at baseline (T0). On the same day, patients also underwent a resting-state
fMRI scan. Six months later (T1), patients were classified as outcome-negative and outcome-positive using a
standard functional outcome scale. We then revisited the baseline measures to test their predictive power for the
functional outcome measured at T1. We found that only outcome-positive patients had an earlier, higher autonomic
response for words compared to pseudo-words, a pattern similar to that of healthy awake controls.
Furthermore, DOC patients showed reduced fALFF in the posterior cingulate cortex (PCC), a brain region that
contributes to autonomic regulation and awareness. The event-related electrodermal marker of residual cognitive
functioning was found to have a significant correlation with residual local neuronal activity in the PCC. We
propose that a residual autonomic response to cognitively salient stimuli, together with a preserved resting-state
activity in the PCC, can provide a useful prognostic index in acute DOC.