A Bayesian approach to analysing cortico-cortical associative stimulation induced increases in the excitability of corticospinal projections in humans

Abstract

Repeated pairing of transcranial magnetic stimulation (TMS) over left and right primary motor cortex (M1), at intensities sufficient to generate descending volleys, produces sustained increases in corticospinal excitability. In other paired associative stimulation (PAS) protocols, in which peripheral afferent stimulation is the first element, changes in corticospinal excitability achieved when the second stimulus consists of brief bursts of transcranial alternating current stimulation (tACS), are comparable to those obtained if TMS is used instead (McNickle and Carson 2015). The present aim was to determine whether associative effects are induced when the first stimulus of a cortico-cortical pair is tACS, or alternatively subthreshold TMS. Bursts of tACS (500 ms; 140 Hz; 1 mA) were associated (180 stimulus pairs) with single magnetic stimuli (120% resting motor threshold rMT) delivered over the opposite (left) M1. The tACS ended 6 ms prior to the TMS. In a separate condition, TMS (55% rMT) was delivered to right M1 6 ms before (120% rMT) TMS was applied over left M1. In a sham condition, TMS (120% rMT) was delivered to left M1 only. The limitations of null hypothesis significance testing are well documented. We therefore employed Bayes factors to assess evidence in support of experimental hypotheses—defined precisely in terms of predicted effect sizes, that these two novel variants of PAS increase corticospinal excitability. Although both interventions induced sustained (~ 20–30 min) increases in corticospinal excitability, the evidence in support of the experimental hypotheses (over specified alternatives) was generally greater for the paired TMS-TMS than the tACS-TMS conditions.