Transcranial magnetic and electrical stimulation compared: Does TES activate intracortical neuronal circuits?

Abstract

Objective

To determine whether, and under which conditions, transcranial electrical stimulation (TES) and transcranial magnetic stimulation (TMS) can activate similar neuronal structures of the human motor cortex, as indicated by electromyographic recordings.

Methods

Focal TMS was performed on three subjects inducing a postero-anterior directed current (p-a), TES with postero-anteriorly (p-a) and latero-medially (l-m) oriented electrodes. We analyzed the onset latencies and amplitudes (single-pulse) and intracortical inhibition and excitation (paired-pulse).

Results

TMS p-a and TES p-a produced muscle responses with the same onset latency, while TES l-m led to 1.4–1.9 ms shorter latencies. Paired-pulse TMS p-a and TES p-a induced inhibition at short inter-stimulus intervals (ISI) (maximum: 2–3 ms) and facilitation at longer ISIs (maximum: 10 ms). No inhibition but a strong facilitation was obtained from paired-pulse TES l-m (ISIs 1–5 ms).

Conclusions

Our findings support the hypothesis, that current direction is the most relevant factor in determining the mode of activation for both TMS and TES: TMS p-a and TES p-a are likely to activate the corticospinal neurons indirectly. In contrast, TES l-m may preferentially activate the corticospinal fibres directly, distant of the neuronal body.

Significance

TES is a suitable tool to induce intracortical inhibition and excitation.

Introduction

Since the pioneering studies of Merton and Morton (1980) several different forms of transcranial electrical (TES, e.g. anodal and cathodal unifocal or bifocal) and magnetic cortex stimulation (TMS, e.g. focal or round coil) have been developed.

As the induced electric field must have a component that is parallel to the neuronal fibre in order to stimulate the neuron (Rushton, 1927), the site of excitation is likely to be dependent on the direction of the induced electric field. The electric fields produced in the brain by TES and TMS have been theoretically modelled and compared by a number of studies (Grandori and Rossini, 1988, Nathan et al., 1993, Saypol et al., 1991). While TMS induced electrical fields are highly localised and parallel to the skull, near the head surface, electric stimulation produced electrical fields with a similar parallel component but also an additional component perpendicular to the surface (Saypol et al., 1991). As these induced electric field components depend on electrode arrangements, distances, polarity (anodal or cathodal) or coil orientation, previous studies have compared a variety of kinds of cortex stimulation. They demonstrated a strong relationship between coil orientation (TMS) or electrode arrangement (TES) and the induced stimulation effects (e.g. Amassian et al., 1987, Amassian et al., 1990, Cracco et al., 1990, Day et al., 1989, Mills et al., 1987, Mills et al., 1992, Rossini, 1988, Rossini et al., 1985, Werhahn et al., 1994, Zarola et al., 1989).

We directly compared bifocal TES and TMS on the motor hand area with the same postero-anterior orientation (p-a), following the hypothesis that the site of excitation of pyramidal neurons is predominantly dependent on current direction and not on stimulation methods.

Previous comparisons of bifocal TES with a latero-medial electrode arrangement (l-m, anodal TES) and TMS with an induced postero-anterior directed current (TMS p-a) revealed that TES induces motor responses with latencies of 1–2 ms shorter than that of TMS (Day et al., 1989, Werhahn et al., 1994). This difference in response latencies is thought to result from the activation at different sites in motor cortex, corresponding to the ‘D- and I-wave’ hypothesis. While transcranial electrical stimulation is supposed to activate cortico-spinal fibres directly inducing D-waves, TMS activates the pyramidal cells indirectly via excitatory interneurons inducing I-waves and leading to longer latencies (Day et al., 1989; for a review Ziemann and Rothwell, 2000). Several studies with direct recording of descending spinal cord volleys in anaesthetised humans (Berardelli et al., 1990, Thompson et al., 1991) and in patients with chronically implanted spinal electrodes (Kaneko et al., 1996, Nakamura et al., 1996) confirm this hypothesis.

The preferential axonal activation induced by TES l-m can be used as a control tool in TMS studies on human motor cortex physiology: To confirm the hypothesis of an intracortical origin of paired-pulse induced inhibition and facilitation (ICI and ICF) by TMS p-a, Kujirai et al. (1993) used an anodal electrical test stimulus (latero-medial oriented electrodes, l-m) as a control condition. They demonstrated that the magnetic conditioning stimuli failed to suppress the electrically induced test responses. This finding was interpreted as evidence for a direct activation of corticospinal fibres by TES, probably distal to the neuronal cell body where inhibitory and excitatory effects on a cortical level have little or no influence.

Until now TMS and bifocal TES on the motor hand area were usually compared using TMS in a p-a direction but l-m oriented TES. It is still unknown whether TES with an electric field component in the p-a direction (TES p-a) could activate similar neuronal structures as TMS p-a, inducing comparable stimulation effects. We, therefore, specifically asked whether current direction induced by TES on the motor hand area determines the mode of corticospinal activation.

To further clarify this question, we compared the stimulation effects of TMS and TES with two different electrode arrangements (p-a and l-m; Fig. 3) in three subjects. First, single pulse TES and TMS were used to compare latency and amplitude of the compound muscle potential. Then intracortical inhibitory and excitatory mechanisms were investigated by applying pairs of either electrical or magnetic pulses.