Comparison of multiple point and statistical motor unit number estimation

This study compares two common techniques for motor unit number estimation, multiple point stimulation and statistical method, to determine which is m ore reproducible. Surface recorded motor unit action potentials (SMUPs) of the left hypothenar muscle group were measured on 20 controls and 10 ALS pa tients. For multiple point, 10 different threshold SMUPs were recorded. For statistical method, mean SMUP amplitude was measured at several stimulus l evels, typically spanning >40% of CMAP amplitude range. Both techniques wer e performed twice, results averaged, electrodes changed, and all recording repeated. For controls, mean of two motor unit number estimation (MUNE) (+/ - standard deviation) was 60 (+/-5) for statistical method, and 108 (+/-38) for multiple point. For ALS patients, these values were 21 (+/-16) for sta tistical method and 55 (+/-39) for multiple point. Test-retest correlation coefficients and coefficients of variation for mean of two MUNE were 0.98 a nd 7% for statistical method, and 0.90 and 12% for multiple point, respecti vely. Statistical method was more reproducible and faster than multiple poi nt, supporting its utility in monitoring rates of MUNE change. (C) 2000 Joh n Wiley & Sons, Inc.