Sources of error in the spike-triggered averaging method of motor unit number estimation (MUNE)

Motor unit number estimation (MUNE) is an electrophysical technique to estimate the number of motor units innervating a muscle or muscle group. MUNE may be useful as a measure of progression of lower motor neuron loss in amyotrophic lateral sclerosis (ALS). Several methods of MUNE have been developed. The spike-triggered averaging method can be readily performed on EMG machines with signal averaging capabilities and is suitable for estimating the number of motor neurons innervating proximal muscles. We have used MUNE as a measure of disease state in a drug efficacy trial for ALS. From our experience with this method we have identified sources of error which can affect MUNE accuracy. We have investigated these sources and report their effect on MUNE.© 1995 John Wiley & Sons, Inc.     

Assessment of axonal loss in Charcot-Marie-Tooth neuropathies

Sensory loss and weakness in Charcot-Marie-Tooth (CMT) neuropathy is due to axonal loss. However, the pattern and degree of axonal loss cannot be accurately determined from routine electrodiagnostic or strength testing due to collateral reinnervation. We sought to quantify axonal loss in two upper extremity muscles in CMT1A and CMT2 subjects using the electrophysiologic endpoint measure of motor unit number estimation (MUNE). Hypothenar and biceps-brachialis muscle groups were studied in 9 CMT1A, 9 CMT2, and 10 control subjects. The spike-triggered averaging (STA) technique was used to collect surface motor unit potentials for MUNE calculations, and a needle electrode was used to collect corresponding intramuscular data. Maximal voluntary hypothenar and handgrip strength was measured quantitatively, while biceps-brachialis strength was measured qualitatively. Compared to normal subjects, CMT1A and CMT2 subjects had significantly lower MUNE values in hypothenar muscles. Biceps-brachialis MUNE values were reduced in CMT2 but not in CMT1A subjects. In support of proximal axonal loss in CMT2 subjects, surface motor unit and intramuscular potential amplitudes were higher in biceps-brachialis muscles compared to controls. Correlations between quantitative strength and MUNE were significant for hypothenar but not for grip muscle groups. Axonal loss is demonstrated in distal muscles in CMT1A and CMT2 supporting a length-dependent axonopathy. Despite clinical findings of normal or near-normal strength and small reductions in compound muscle action potential (CMAP) amplitude, MUNE values were significantly lower in CMT2 subjects in proximal muscles, consistent with more diffuse denervation. These data indicate that subclinical axonal loss is present that cannot be appreciated using clinical examination or routine electrodiagnostic techniques.     

Motor unit number estimates in idiopathic Parkinson's disease

We previously reported changes in motor unit morphology in patients with Parkinson's disease (PD) using subjective and computerized quantitative electromyography. Now, we present data on motor unit number estimates (MUNE) to address the hypothesis of motor neuron dropout in PD. Twenty patients with PD and 20 age-matched control subjects were screened by clinical criteria and nerve conduction studies to exclude those with neuropathy. Motor unit number estimates in the extensor digitorum brevis and hypothenar group were assessed by three different MUNE techniques. The MUNE technique types included (1) the statistical method developed by Daube, (2) a threshold method, and (3) an F-wave method. The overall multivariate comparison for the six MUNE measurements was significantly lower for the patients than the controls (P=0.02). The only significant difference in the individual measures was found in the threshold MUNE method of the hypothenar group (P<0.05). These results are consistent with those of our previous work, and both support the hypothesis that mild motor neuron dropout occurs in idiopathic PD. However, MUNE methods characteristically have large standard deviations which make it difficult to detect small changes. Progress in decreasing the variance of MUNEs will facilitate their use in detecting small motor unit number changes in neurodegenerative disease.     

Motor unit number estimation in infants and children with spinal muscular atrophy

Spinal muscular atrophy (SMA) is a disease of lower motor neurons. Motor unit number estimation (MUNE) is an electrophysiologic method to estimate the number of motor neurons innervating a muscle group. We applied the multiple point stimulation technique to the ulnar nerve--hypothenar muscle group to study lower motor neuron loss in 14 SMA subjects, including those presymptomatic, and varying from newborn through 45 years of age. Preliminary data support the value of MUNE to help understand the time course of motor neuron loss in SMA.     

Reproducibility of motor unit number estimation in individual subjects

Although the reproducibility of motor unit number estimation (MUNE) for groups of subjects has been studied, there is little such data for individuals. Prediction intervals represent a tool to study individual MUNE reproducibility and represent the range of values expected for a future MUNE if the true number of motor units remains unchanged. MUNE was performed using the statistical method on 48 normal individuals. The prediction interval was found to be a function of the intrasubject coefficient of variation. Using a commercial manufacturer's recommended technique and software, prediction intervals were found to be so broad as to be of uncertain value. We found that by averaging two MUNE observations for each determination, and using the method of weighted averages for calculating the size of an average single motor unit potential, the intrasubject coefficient of variation was reduced from 16.48% to 8.77%, and the 90% prediction interval became sufficiently narrow to be clinically useful. False-negative rates were also lowered substantially using these techniques. Thus, simple modifications of an existing MUNE program improved the clinical utility of this program for the longitudinal study of patients in whom changes in motor unit number over time are of importance, such as those with motor neuron diseases.     

Motor unit number estimation: A technology and literature review

Introduction: Numerous methods for motor unit number estimation (MUNE) have been developed. The objective of this article is to summarize and compare the major methods and the available data regarding their reproducibility, validity, application, refinement, and utility. Methods: Using specified search criteria, a systematic review of the literature was performed. Reproducibility, normative data, application to specific diseases and conditions, technical refinements, and practicality were compiled into a comprehensive database and analyzed. Results: The most commonly reported MUNE methods are the incremental, multiple‐point stimulation, spike‐triggered averaging, and statistical methods. All have established normative data sets and high reproducibility. MUNE provides quantitative assessments of motor neuron loss and has been applied successfully to the study of many clinical conditions, including amyotrophic lateral sclerosis and normal aging. Conclusions: MUNE is an important research technique in human subjects, providing important data regarding motor unit populations and motor unit loss over time. Muscle Nerve 50 : 884–893, 2014     

Motor unit number estimation by decomposition-enhanced spike-triggered averaging: control data, test-retest reliability, and contractile level effects

Decomposition-enhanced spike-triggered averaging (DE-STA) has been developed as a method for obtaining a motor unit number estimate (MUNE). We describe the method and report control data for the first dorsal interosseous/adductor pollicis and thenar muscles and reliability in the thenar muscles. Seventeen subjects (ages 20-50 years) took part in the study. The maximum M potential was elicited with supramaximal stimulation of the ulnar or median nerve at the wrist. Surface and intramuscularly detected electromyographic signals were then collected simultaneously during mild to moderate contractions. Decomposition algorithms were used to detect and sort the individual motor unit potential (MUP) occurrences of several concurrently active motor units in the needle-detected signals. The MUP occurrences were used as triggering sources to estimate their corresponding surface-detected MUPs (S-MUPs) using STA. The mean S-MUP size was calculated and divided into the maximum M-potential size to derive a MUNE. The MUNE values were consistent with those previously reported with other methods, and thenar MUNEs for the two trials were similar (249 +/- 78 and 246 +/- 90), with high test-retest reliability (r = 0.94, P < 0.05). DE-STA thus appears to be a valid and reliable method to obtain MUNEs.     

Motor unit number estimate of distal and proximal muscles in Charcot-Marie-Tooth disease

In order to determine the utility of motor unit number estimation (MUNE) in assessing axonal loss in chronic inherited neuropathies, we determined MUNEs in 54 patients with Charcot-Marie-Tooth (CMT) disease (29 patients with CMT-1A, 13 with CMT-X, and 12 with CMT-2) by using spike-triggered averaging (STA) of the ulnar-innervated abductor digiti minimi/hypothenar muscles (ADM) and the musculo-cutaneous innervated biceps/brachialis (BB) muscles. MUNEs were analyzed in relationship to the corresponding compound muscle action potential (CMAP) amplitudes as well as to clinical strength. Proximal muscles, which appeared strong clinically, had evidence of chronic denervation/reinnervation, although to a lesser extent than weak distal hand muscles, supporting the concept that axonal loss in CMT occurs in a length-dependent fashion. The reduction in ADM-MUNE strongly correlated with clinical weakness in the hand. Both the ADM-MUNE and BB-MUNE were abnormal more often than CMAP amplitude, probably reflecting extensive motor unit reconfiguration and enlargement that maintains CMAP amplitude despite severe motor unit loss. This study suggests that MUNE can assess motor unit loss in CMT and may better reflect axonal loss than CMAP amplitude. The STA technique of MUNE may be useful in longitudinal studies of proximal and distal motor unit changes in CMT.     

Age-related reductions in the estimated numbers of motor units are minimal in the human soleus

The documented impact of contractile level on decomposition-enhanced spike-triggered averaging motor unit number estimates (MUNEs) in young adults demonstrates the importance of selecting an objective contraction intensity that yields the most representative MUNE for a given muscle. Whether the same contraction intensity would be ideal in an altered system (e.g., by aging or disease) has yet to be examined. Thus, the main purpose of this study was to compare the effects of contraction intensity on MUNEs from the soleus muscle in young ( approximately 27 years) and old ( approximately 75 years) men. Using decomposition-enhanced spike-triggered averaging, surface and intramuscular electromyographic signals were collected from the soleus during a range of submaximal isometric plantar-flexion contractions (threshold, 10%, 20%, and 30% of maximum voluntary contraction; MVC). Five MUNEs were calculated, one for each of the four contraction intensities and an ensemble MUNE was derived from all MUs collected. Although MUNE decreased similarly with increased effort in both groups, MUNEs were not significantly reduced in the old men compared to the young men. Consequently, the ensemble MUNE was extrapolated to an intensity of approximately 15% MVC in both young and old. The results suggest that, in the soleus, the use of the same contraction intensity across age groups is a valid comparison.     

Monitoring disease progression using high‐density motor unit number estimation in amyotrophic lateral sclerosis

In amyotrophic lateral sclerosis (ALS), progressive motor neuron loss causes severe weakness. Functional measurements tend to underestimate the underlying pathology because of collateral reinnervation. A more direct marker of lower motor neuron loss is of significant importance. We evaluated high‐density motor unit number estimation (MUNE), as compared with the ALS Functional Rating Scale (ALSFRS) and maximal compound muscle action potential (CMAP) amplitude, for monitoring and classifying disease progression. MUNE showed good reproducibility (intraclass correlation coefficient = 0.86). MUNE showed a significantly greater decrease than the ALSFRS, the Medical Research Council (MRC) scale, and CMAP amplitude. Patients could be stratified into groups with rapidly or slowly progressive disease based on a decrement in MUNE at 4 months from baseline; ALSFRS score at 8 months was significantly lower in the rapidly progressive group. MUNE was sensitive to motor neuron loss early in the disease course when compared to other clinical measures. Stratification of patients based on a decrease in MUNE seems feasible. Muscle Nerve, 2010     

Motor unit number estimates in the tibialis anterior muscle of young, old, and very old men

The rate of motor unit (MU) loss and its influence on the progression of sarcopenia is not well understood. Therefore, the main purpose of this study was to estimate and compare numbers of MUs in the tibialis anterior (TA) of young men ( approximately 25 years) and two groups of older men ( approximately 65 years and >/=80 years). Decomposition-enhanced spike-triggered averaging was used to collect surface and intramuscular electromyographic signals during isometric dorsiflexions at 25% of maximum voluntary contraction. The mean surface-MU potential size was divided into the maximum M wave to calculate the motor unit number estimate (MUNE). The MUNE was significantly reduced in the old (91) compared to young (150) men, and further reduced in the very old men (59). Despite the smaller MUNE at age 65, strength was not reduced until beyond 80 years. This suggests that age-related MU loss in the TA does not limit function until a critical threshold is reached.     

Comparison of incremental with multipoint MUNE methods in transgenic ALS mice.

Several methods of motor unit number estimation (MUNE) are in current use. Uncertainty still exists about which is preferable and how results obtained from one method compare to another. We studied changes with MUNE over time in the SOD1(G93A) transgenic mouse model of amyotrophic lateral sclerosis (ALS), using both incremental and multipoint methods. This mouse model of motor neuron degeneration is highly consistent, with a monotonic decline in motor neuron number starting at approximately 60 days of life. Five mice were studied four times each, starting at day 60 of life and approximately every 20 days thereafter, using both methods. Results were quite comparable for both methods, with the incremental method yielding slightly higher estimates of motor unit size, and hence smaller MUNEs. Correlations between the two methods were 0.71 for single motor unit action potential (SMUAP) amplitude and 0.95 for MUNE. In this model, therefore, both MUNE methods yield similar estimates and are equally effective at documenting progression of a lower motor neuron disorder.     

Changes in neuromuscular function in elders: Novel techniques for assessment of motor unit loss and motor unit remodeling with aging

Functional muscle fiber denervation is a major contributor to the decline in physical function observed with aging and is now a recognized cause of sarcopenia, a muscle disorder characterized by progressive and generalized degenerative loss of skeletal muscle mass, quality, and strength. There is an interrelationship between muscle strength, motor unit (MU) number, and aging, which suggests that a portion of muscle weakness in seniors may be attributable to the loss of functional MUs. During normal aging, there is a time-related progression of MU loss, an adaptive sprouting followed by a maladaptive sprouting, and continuing recession of terminal Schwann cells leading to a reduced capacity for compensatory reinnervation in elders. In amyotrophic lateral sclerosis, increasing age at onset predicts worse survival ALS and it is possible that age-related depletion of the motor neuron pool may worsen motor neuron disease. MUNE methods are used to estimate the number of functional MU, data from MUNIX arguing for motor neuron loss with aging will be reviewed. Recently, a new MRI technique MU-MRI could be used to assess the MU recruitment or explore the activity of a single MU. This review presents published studies on the changes of neuromuscular function with aging, then focusing on these two novel techniques for assessment of MU loss and MU remodeling.     

The effect of contraction intensity on motor unit number estimates of the tibialis anterior.

OBJECTIVE: To examine the effect of contractile level on motor unit number estimates (MUNEs) and establish the contraction intensity that will yield the most representative MUNE for the tibialis anterior (TA) muscle. METHODS: Surface and intramuscular electromyographic (EMG) signals were collected during a range of submaximal (threshold, 10, 20, 30 and 40% MVC) isometric dorsiflexion contractions using decomposition-enhanced spike-triggered averaging (DE-STA). Six MUNEs were calculated, one for each of the five intensities, and an ensemble sixth MUNE that had equal MU contributions from all intensities. RESULTS: Mean surface-motor unit potential sizes increased significantly (26-69 microV) and MUNEs decreased accordingly (226-91) as contraction intensity increased from threshold to 40% MVC, respectively (P<0.05). The ensemble MUNE was 153, and extrapolated to approximately 25% MVC. CONCLUSIONS: There was a significant and progressive decline in the MUNE as contraction intensity increased, confirming the importance of monitoring torque during data collection. The ensemble MUNE suggests that collecting EMG signals at a contraction intensity of approximately 25% MVC provides the most representative sample of the actual number and sizes of MUs in the TA. SIGNIFICANCE: Establishing appropriate contraction intensities improves the utility of DE-STA as a useful method for tracking changes to the MU pool in disease states and healthy aging.     

Use of the collision technique to improve the accuracy of motor unit number estimation.

OBJECTIVE: To evaluate the risk of stimulating the same motor axon at different points along the median nerve when using the multiple point stimulation (MPS) technique and how this affects the accuracy of the motor unit number estimate (MUNE). METHODS: Using the MPS technique, MUNE in the median innervated intrinsic hand muscles was done on two normal subjects, a patient with carpal tunnel syndrome and one with prior poliomyelitis. The collision technique was then used to confirm whether two motor unit action potentials (MUAPs) with similar configurations and sizes were generated by the same motor unit. A new MUNE was recalculated after the repeated inclusion of the same motor unit had been excluded. RESULTS: While the risk of stimulating the same motor axon at widely separate locations of the median nerve was negligible in normal subjects, this risk was much higher in patients with a depleted motor neuron pool. This resulted in marked distortion of the calculated size of the average single MUAP and, consequently, inaccuracy of the MUNE. CONCLUSIONS: The inadvertent inclusion of the same motor units, if not recognized, can markedly affect the accuracy of the MUNE. The collision technique can be useful in minimizing this risk.     

Decomposition-based quantitative electromyography: effect of force on motor unit potentials and motor unit number estimates

Decomposition-based quantitative electromyography (DQEMG) allows for the collection of motor unit potentials (MUPs) over a broad range of force levels. Given the size principle of motor unit recruitment, it may be necessary to control for force when using DQEMG for the purpose of deriving a motor unit number estimate (MUNE). Therefore, this study was performed to examine the effect of force on the physiological characteristics of concentric needle- and surface-detected MUPs and the subsequent impact on MUNEs obtained from the first dorsal interosseous (FDI) muscle sampled using DQEMG. Maximum M waves were elicited in 10 subjects with supramaximal stimulation of the ulnar nerve at the wrist. Intramuscular and surface-detected EMG signals were collected simultaneously during 30-s voluntary isometric contractions performed at specific percentages of maximal voluntary contraction (MVC). Decomposition algorithms were used to identify needle-detected MUPs and their individual MU firing times. These MU firing times were used as triggers to extract their corresponding surface-detected MUPs (S-MUPs) using spike-triggered averaging. A mean S-MUP was then calculated, the size of which was divided into the maximum M-wave size to derive a MUNE. Increased levels of contraction had a significant effect on needle- and surface-detected MUP size, firing rate, and MUNE. These results suggest that force level is an important factor to consider when performing quantitative EMG, including MUNEs with this method.     

Motor unit number estimation in the rat tail using a modified multipoint stimulation technique

Motor unit number estimation (MUNE) of the rodent hindlimb has been used mainly for following the progression of motor neuron disorders. By performing MUNE in the tail, however, progression of axonal neuropathy could also be assessed, as both proximal and distal regions would be available for study. In this investigation, three raters performed a modified multipoint stimulation MUNE technique in the tails of 14 healthy adult rats. The technique was straightforward to perform, with a relatively narrow range of motor unit number estimates of 40 ± 16 (standard deviation) for the proximal tail and 21 ± 11 for the distal tail. Intrarater reliability coefficients were 0.31 (P = 0.033) and 0.32 (P = 0.028) for the proximal and distal tail, respectively. Interrater reliability coefficients were 0.22 (P = 0.086) and 0.44 (P = 0.004). These reliability assessments, along with the relatively low motor unit estimates and narrow range of values, support the idea that rat tail MUNE may have utility in the evaluation of rodent models of neuromuscular disease, including length‐dependent neuropathy. Muscle Nerve 40: 115–121, 2009     

Age‐related changes in motor unit number estimates in adult patients with Charcot–Marie–Tooth type 1A

Background: Charcot–Marie–Tooth disease type 1A (CMT1A) is known as a demyelinating hereditary neuropathy. Secondary axonal dysfunction is the most important determinant of disease severity. In adult patients, clinical progression may be because of further axonal deterioration as was shown with compound muscle action potential (CMAP) amplitude reductions over time. The motor unit number estimation (MUNE) technique may be more accurate to determine the number of axons as it is not disturbed by the effect of reinnervation. The purpose of this study was to investigate the number and size of motor units in relation to age in patients and controls. Methods: In a cross‐sectional design, we assessed arm and hand strength and performed electrophysiological examinations, including CMAP amplitudes and MUNE of the thenar muscles using high‐density surface EMG in 69 adult patients with CMT1A and 55 age‐matched healthy controls. Results: In patients, lower CMAP amplitudes and MUNE values were related to hand weakness. The CMAP amplitude and MUNE value of the thenar muscles were significantly lower in patients than in controls. CMAP amplitudes declined with age in controls, but not in patients. MUNE values declined with age in both patients and controls. Conclusions: The age‐dependent decrease in the number of motor units was not significantly different between patients with CMT1A and controls, indicating that loss of motor units in adult patients is limited.     

Reductions in motor unit number estimates (MUNE) precede motor neuron loss in the plasma membrane calcium ATPase 2 (PMCA2)-heterozygous mice

The potential of MUNE as a unique electrophysiological tool to detect early motor unit abnormalities during a clinically silent period was investigated in the plasma membrane calcium ATPase 2 (PMCA2)-heterozygous mice. There was a significant reduction in MUNE in the PMCA2-heterozygous mice as compared to the wild type littermates at two months of age. In contrast, the compound motor action potential (CMAP) was not altered. The conduction velocity (CV) of the sensory nerve and sensory nerve action potentials (SNAP) were not modified indicating lack of major sensory deficits. Interestingly, despite a decline in MUNE at this age, no changes were detected in choline acetyl transferase (ChAT) positive motor neuron number in the ventral horn of the lumbar spinal cord. Hindlimb grip strength, a test that evaluates clinical dysfunction, was also similar to that of the wild type controls. However, motor neuron number significantly decreased by five months suggesting that a drop in MUNE preceded motor neuron loss. In the two-month-old PMCA2-null mice, reduced MUNE measurements coincided with lower motor neuron number and decreased hindlimb grip strength. The fall in motor neuron number was already detectable at three weeks, the earliest time studied, and became more pronounced by five months. Our results show that even partial reductions in PMCA2 levels are sufficient to cause delayed death of motor neurons and that MUNE may be a reliable and sensitive approach to detect pathology prior to cell loss and in the absence of overt clinical signs.     

The relationship between Bayesian motor unit number estimation and histological measurements of motor neurons in wild-type and SOD1(G93A) mice

Objective

To assess the relationship between Bayesian MUNE and histological motor neuron counts in wild-type mice and in an animal model of ALS.

Methods

We performed Bayesian MUNE paired with histological counts of motor neurons in the lumbar spinal cord of wild-type mice and transgenic SOD1^G93A mice that show progressive weakness over time. We evaluated the number of acetylcholine endplates that were innervated by a presynaptic nerve.

Results

In wild-type mice, the motor unit number in the gastrocnemius muscle estimated by Bayesian MUNE was approximately half the number of motor neurons in the region of the spinal cord that contains the cell bodies of the motor neurons supplying the hindlimb crural flexor muscles. In SOD1^G93A mice, motor neuron numbers declined over time. This was associated with motor endplate denervation at the end-stage of disease.

Conclusion

The number of motor neurons in the spinal cord of wild-type mice is proportional to the number of motor units estimated by Bayesian MUNE. In SOD1^G93A mice, there is a lower number of estimated motor units compared to the number of spinal cord motor neurons at the end-stage of disease, and this is associated with disruption of the neuromuscular junction.

Significance

Our finding that the Bayesian MUNE method gives estimates of motor unit numbers that are proportional to the numbers of motor neurons in the spinal cord supports the clinical use of Bayesian MUNE in monitoring motor unit loss in ALS patients.