Examination of surface-recorded motor units and a new approach to motor unit number estimates

A motor unit comprises an anterior horn cell, a motor axon and its terminal nerve branches, and the muscle fibres supplied by that axon. The number of motor units in each muscle is not known and has been a research area for 30 years. An ability to measure the number of motor units is needed for the assessment of progression of diseases of the motor unit such as motor neuron disease (MND). Tests to determine the number of motor units by electrophysiological means are termed “motor unit number estimates” (MUNE). A number of MUNE methods have been devised. All of the MUNE methods have limitations, the most important being the need to allow for all of the variables affecting motor unit responses to electrical stimulation. The first aim of this project was to assess motor unit responses by evaluating the data collection parameters used in the Poisson statistical MUNE method in a normal and MND population. Different methods of collecting the compound muscle action potential (CMAP) responses following a fixed stimulus were studied. The problems associated with sampling of a portion of the motor units and the increasing influence of large motor units with progression of MND was found to be important limitations of the Poisson statistical method. This study identified the importance of the stimulus-activation curve, and showed that a large number of data points can be collected. Technical issues with data collection were found, and importantly the presence of large steps in MND subjects. This study also identified small motor units with low variability. The second aim was to assess the features of the whole stimulusactivation curve obtained by stimulating motor nerves with increasing stimulus intensity and recording the CMAP in normal and MND populations. The size of motor units within the stimulus-activation curve was determined. In normal subjects, there is a pattern of electrical activation from smallest to largest motor unit. A third aim was to use the stimulus-activation curve and the collection of a large number of data points to assess the variability in the size of surface-recorded motor unit responses following electrical stimulation in normal and MND subjects. MUNE methods can be misled by the inherent variability in single motor units. This is particularly true in disorders with unstable motor units, such as MND. A value of 100 µVms was an approximate size for the variability arising from surfacerecordings of individual motor units. The fourth aim was to examine the reduction in size of the CMAP that occurs with repetitive electrical stimulation. This needs to be understood as a cause of variability of motor unit firing that could be important in the development of MUNE techniques. The study showed that a maximal CMAP decrement of 10% with train-of-four repetitive stimulation occurs in a small subset of MND subjects and is probably related to disease severity. From studies of individual motor units, the decrement in MND arose from a reduction in the size of components of individual motor units rather than loss of firing of entire motor units. The reduction in size was probably due to conduction block in terminal axons of individual motor units. The final aim was to use a new MUNE method, developed by uniting the stimulus-activation curve with understanding of the causes of variability in a Bayesian statistical model to estimate motor unit numbers in normal subjects and subjects with MND. Using this Bayesian statistical method, the number of motor units was reduced in MND compared to normal subjects and showed progressive loss over time in MND subjects. The method had the advantages of applicability to all stages of MND and simplicity of the data collection