Can the causal role of brain oscillations be studied through rhythmic brain stimulation?

Many studies have investigated the causal relevance of brain-oscillations using rhythmic stimulation, either through direct-brain or sensory stimulation. Yet, how intrinsic rhythms interact with the externally generated rhythm is largely unknown. We either presented a flickered (60 Hz) visual grating or its correspondent unflickered stimulus in a psychophysical change-detection-task during simultaneous MEG-recordings to humans, to test the effect of visual entrainment on induced gamma-oscillations. Notably, we generally observed the co-existence of the broadband induced gamma-rhythm with the entrained flicker-rhythm (reliably measured in each participant), with the peak frequency of the induced response remaining unaltered in approximately half of participants -relatively independently of their native frequency. However, flicker increased broadband induced-gamma-power, and this was stronger in participants with a native frequency closer to the flicker-frequency (‘resonance’), and led to strong phase-entrainment. Presence of flicker did not change behaviour itself, but profoundly altered brain-behaviour correlates across the sample: whilst broadband induced gamma-oscillations correlated with reaction-times for unflickered stimuli (as known previously), for the flicker, the amplitude of the entrained flicker-rhythm (but no more the induced oscillation) correlated with reaction-times. This, however, strongly depended on whether a participant’s peak frequency shifted to the entrained rhythm. Our results suggests that rhythmic brain-stimulation leads to a coexistence of two partially independent oscillations with heterogeneous effects across participants on the ‘downstream relevance’ of these rhythms for behaviour. This may explain the inconsistency of findings related to external entrainment of brain-oscillations and poses further questions towards causal manipulations of brain-oscillations in general.