Functional dissociation of lower and upper frequency mu rhythms in relation to voluntary limb movement

Abstract

Objective: The goal of this study is to investigate the reactivity of central rhythms in the alpha band during self-paced voluntary finger and foot movement and to give an answer to the question, whether different types of mu rhythms exist.

Methods: The effect of self-paced, voluntary finger and foot movement was studied in a group of 12 right-handed healthy volunteers. The EEG was recorded from a grid of 34 electrodes placed over sensorimotor areas with inter-electrode distances of approximately 2.5 cm. The event-related desynchronization (ERD) was quantified in the 8–10 and 10–12 Hz bands.

Results: Both frequency components are blocked prior to and during movement and therefore, they have to be considered as mu rhythms. The lower frequency component results in a widespread movement-type non-specific ERD pattern, whereas the upper frequency component shows a more focused and movement-type specific pattern, clearly different with finger and foot movement.

Conclusions: The distinct reactivity patterns provide evidence for the existence of two types of mu rhythms, a somatotopically non-specific lower frequency mu rhythm and a somatotopically specific mu rhythm characteristically found in the upper alpha frequency band.

Introduction

Desynchronization in the alpha band is not a unitary phenomenon. Klimesch et al. (1992) reported that in a cognitive task, the lower alpha band (8–10 Hz) displays a topographically wide-spread desynchronization, whereas the upper alpha band (10–12 Hz) desynchronization is topographically restricted. This different behavior of lower and upper alpha band components in a cognitive task can be interpreted that the lower alpha desynchronization probably reflects general task demands and attentional processes and is non-task specific, the upper alpha desynchronization, in contrast, develops during processing of sensory-semantic information and is therefore task-specific.

Factor analyses of the power spectra obtained from EEGs of 243 patients with eyes closed revealed one alpha band between 9 and 10.5 Hz and another between 10.5 and 12.5 Hz (Lopes da Silva, 1993). A division in upper and lower alpha bands was also reported by Matousek and Petersén (1973) and Hermann et al. (1978). It is of interest that analyses of ongoing EEG during rest with eyes closed and dynamic EEG changes in a cognitive task differentiate both between lower and upper alpha band components. It can therefore be speculated that in a motor task, distinct reactivity patterns can be found with different alpha frequency components and lower and upper frequency mu rhythms, respectively.

In the literature, contrary reports on the mu frequency can be found. On the one side, Stancák and Pfurtscheller (1996) reported a mean mu frequency of 9.8±1.6 Hz (mean±SD) in right-handed subjects. In a similar study with voluntary movement, Guieu et al. (1999) found a mu peak frequency of 9.81±1.01 Hz and Kuhlman (1978) described a mu rhythm in the range from 7.75 to 13.75 Hz with a mean of 10.1 Hz. From these studies performed in 3 different laboratories, the conclusion can be drawn that the mu rhythm can cover a broad frequency range between 7 and 14 Hz with a mean frequency around 10 Hz or even slightly below. This is also in line with ECoG studies discussing a cortical mu rhythm in the 7–11 Hz band (Arroyo et al., 1993), in the 8–12 Hz band (Toro et al., 1994) and the 8–13 Hz band (Crone et al., 1998). While Arroyo et al. (1993) found that the presence and blocking of mu rhythms are specific to somatic representation areas of the cortex, Crone et al. (1998) reported on a relative diffuse event-related desynchronization with voluntary movement. Similar somatotopically discrete distributions of the event-related desynchronization (ERD) in the 8–10 and 10–12 Hz band in subdural recordings during contralateral finger movement were investigated by Toro et al. (1994).

On the other side, there are some indications that the mu rhythm is present dominantly in the upper alpha band between 10 and 13 Hz. So for example, in a voluntary hand movement study, a bicoherence between mu rhythm and its first harmonic was only found in subjects with a mu rhythm frequency between 11 and 13 Hz (Pfurtscheller and Stancák, 1997). In other words, the characteristically arch-shaped mu rhythm had a frequency of clearly above 10 Hz. Lüders et al. (1987) reported on two patients who had an 11–13 Hz rhythm over the sensorimotor cortex and Pfurtscheller (1981) investigated a normal subject with a mu rhythm of 13.5 Hz. Based on interhemispheric coherence analysis, Storm van Leeuwen et al. (1978) reported on a frequency of 10.5±0.9 Hz for the mu rhythm and a lower frequency for the alpha rhythm.

An interesting observation is that voluntary foot movement can result in an enhancement of the hand area mu rhythm dominant in the 10–12 Hz band (Pfurtscheller and Neuper, 1994). This synchronization of frequency components over the hand area was accompanied in one out of 3 subjects by a circumscribed suppression of the foot area mu rhythm in the 10–12 Hz band.

It can be summarized that the movement-related desynchronization in EEG recordings over sensorimotor areas is found with frequency components between 7 and 14 Hz. Furthermore, there is some evidence that the upper frequency components within the alpha band display a less diffuse and more somatotopically-specific spatial ERD pattern as compared to the lower frequency components. The goal of this study is to investigate the EEG reactivity (alpha band ERD) during self-paced voluntary finger and foot movement and to give an answer to the following questions:

• Is it possible to differentiate between lower and higher frequency mu rhythms?

• Do somatotopically specific and non-specific mu rhythms exist?