Within-train neuromuscular propagation varies with torque in paralyzed human muscle.

Electromyographic (EMG) recordings may serve an important role in predicting torque during repetitive activation of paralyzed muscle. We compared the initial M-wave to the subsequent M-waves of the same train under fatigued and recovered conditions in the paralyzed human soleus muscle. Sixteen individuals with chronic (n = 13) or acute paralysis (n = 3) had the tibial nerve activated before and after a repetitive supramaximal stimulation protocol. The mean within-train M-wave amplitude and median frequency increased approximately 20%, whereas the duration decreased approximately 15% compared with the initial M-wave of each train. During fatigue, there was a linear decrease in the difference between the initial M-wave amplitude and subsequent train ( approximately 20% to 8%). Following fatigue, this difference recovered to approximately 12%. The difference between the M-wave train average and the initial M-wave for amplitude, duration, and median frequency closely followed torque (Pearson correlations = 0.99, 0.94, and 0.98, respectively) during fatigue. We conclude that the difference between the later-occurring M-waves (average of the train) and initial M-wave is large when muscle torque is high and less when torque is low and, therefore, predicts torque during activation of paralyzed muscle. This difference in the within-train M-wave amplitude, duration, and median frequency may reflect a mechanical change, such as muscle shortening and increased muscle cross-sectional area during isometric contractions. Electromyographic feedback may assist in the optimization of neuromuscular electrical stimulation of paralyzed muscle. Copyright 2002 Wiley Periodicals, Inc. Muscle Nerve 26: 673-680, 2002