The hidden brain-state dynamics of tACS aftereffects.

Non-invasive techniques to electrically stimulate the brain such as transcranial direct and alternating current stimulation (tDCS/tACS) are increasingly used in human neuroscience and offer potential new avenues to treat brain disorders. Previous research has shown that stimulation effects may depend on brain-states. However, this work mostly focused on experimentally induced brain-states over the course of several minutes. Besides such global, long-term changes in brain-states, previous research suggests, that the brain is likely to spontaneously alternate between states in sub-second ranges, which is much closer to the time scale at which it is generally believed to operate. Here, we utilized Hidden Markov Models (HMM) to decompose magnetoencephalography data obtained before and after tACS into spontaneous, transient brain-states with distinct spatial, spectral and connectivity profiles. Only one out of four spontaneous brain-states, likely reflecting default mode network activity, showed evidence for an effect of tACS on the power of spontaneous α-oscillations. The identified state appears to disproportionally drive the overall (non-state resolved) tACS effect. No or only marginal effects were found in the remaining states. We found no evidence that tACS influenced the time spent in each state. Although stimulation was applied continuously, our results indicate that spontaneous brain-states and their underlying functional networks differ in their susceptibility to tACS. Global stimulation aftereffects may be disproportionally driven by distinct time periods during which the susceptible state is active. Our results may pave the ground for future work to understand which features make a specific brain-state susceptible to electrical stimulation.