Communication between brain areas and how they are influenced by changes in consciousness are not fully understood. One hypothesis is that brain areas communicate via oscillatory processes, utilizing network-specific frequency bands, that can be measured with metrics that reflect between-region interactions, such as coherence and phase amplitude coupling (PAC). To evaluate this hypothesis, and understand how these interactions are modulated by state changes, we analyzed electrophysiological recordings in humans at different nodes of one well-studied brain network: the basal ganglia-thalamo-cortical (BGTC) loops of the motor system, during loss of consciousness induced by anesthesia. We recorded simultaneous electrocorticography (ECoG) over primary motor cortex (M1) with local field potentials (LFPs) from subcortical motor regions (either basal ganglia or thalamus) in 15 movement disorder patients during anesthesia (propofol) induction as a part of their surgery for deep brain stimulation. We observed reduced coherence and PAC between M1 and the subcortical nuclei, which was specific to the beta band (~18-24 Hz). The fact that this pattern occurs selectively in beta underscores the importance of this frequency band in the motor system, and supports the idea that oscillatory interactions at specific frequencies are related to the capacity for normal brain function and behavior.