An interesting finding of the present study was that although the frequency- and pulse-duration-modulation protocols showed no differences in the amount of muscle fatigue or low-frequency fatigue (), the frequency-modulation protocol produced better isometric muscle performance in response to the fatiguing trains than the pulse-duration-modulation protocol ( and ). The better isometric performance demonstrated during frequency-compared to pulse-duration-modulation for similar levels of muscle fatigue may have important implications for clinical applications of FES because most current FES systems deliver a constant frequency and only increase the stimulation intensity to maintain muscle force as the muscle fatigues (Petrofsky and Stacy, 1992
; Raymond et al., 1999
). Thus, these results suggest that there is a need to further investigate strategies of control of force during FES in patient populations.
We compared the peak forces generated in response to the last fatiguing trains and the post-fatigue 60 Hz testing trains during the frequency- and pulse-duration-modulation protocols to further analyze differences between peak force responses of the two modulation protocols (). We are not sure how to explain the finding that an additional 1 s of recovery time between the last fatiguing trains and the post-fatigue testing trains (see Section 2.3 and for details) allowed for greater recovery of peak forces for the pulse-duration-modulation than the frequency-modulation protocol (see ). After additional recovery of peak forces in response to the 60 Hz testing trains for the pulse-duration-modulation protocol, the differences in the amount of muscle fatigue between the frequency- and pulse-duration-modulation protocols were not significant. These results could be due to differences in the level of metabolic energy expenditure, the extent of activation failure, or differences in recovery of force-generating ability during the frequency- and pulse-duration-modulation protocols. However, we did not have adequate data to further investigate the mechanisms underlying muscle fatigue and performance during the two modulation protocols. Future studies that measure the metabolic energy expenditure and/or electrophysiological activity of motor units during different modulation strategies may help to gain a better understanding of the mechanisms underlying muscle performance and fatigue during repetitive stimulation.
Fig. 7 Average peak forces generated in response to the last fatiguing train (left) and the post-fatigue 60 Hz testing train (right) during the frequency-modulation and the pulse-duration (PD)-modulation protocol. Both the last fatiguing train and the post-fatigue (more ...)
Although the no-modulation protocol produced the least muscle fatigue () and low-frequency fatigue (), both modulation protocols showed better performance in response to the fatiguing trains than the no-modulation protocol ( and ). The no-modulation protocol maintained a lower stimulation frequency (11.3 ± 1.6 Hz) than either the pulse-duration- (60 Hz throughout) or the frequency-modulation protocol (stepwise increase from 11.6 ± 1.5 to 60 Hz), which may have contributed to the lower muscle fatigue (Binder-Macleod et al., 1995
; Frank et al., 1998
; Marsden et al., 1983
) and low-frequency fatigue (Chin and Allen, 1996
; Chin et al., 1997
; Westerblad et al., 1993
) (). Low-frequency fatigue is characterized by the selective loss of force at low- versus high-frequencies during fatigue or recovery from fatigue (Edwards et al., 1977
; Stokes et al., 1989
). Low-frequency fatigue causes a rightward shift in the force–frequency curve, which requires higher frequencies to produce comparable forces to the pre-fatigued state (Binder-Macleod and McDermond, 1992
; Binder-Macleod et al., 1998
; Fuglevand et al., 1999
; Thomas et al., 1991
). Thus, low-frequency fatigue, as well as overall muscle fatigue, would cause an attenuation of performance in response to the low-frequency trains (11.3 ± 1.6 Hz) used during the no-modulation protocol. However, during the latter half of the frequency-modulation protocol, and throughout the pulse-duration-modulation-protocol, the muscle was stimulated with frequencies in the high-frequency range of the force–frequency curve (). This high-frequency stimulation would overcome the effects of low-frequency fatigue (Edwards et al., 1977
; Russ and Binder-Macleod, 1999
; Stokes et al., 1989
) and perhaps contribute to the better performance of the two modulation protocols compared to the no-modulation protocol.
Previously, Graupe and colleagues showed that the stochastic modulation of the inter-pulse intervals within stimulation trains decreased the rate of muscle fatigue of the quadriceps femoris muscle compared to stimulation at a constant frequency on a single subject (Graupe et al., 2000
). In contrast, Thrasher and colleagues recently showed that random modulation of frequency (mean: 40 Hz), amplitude (mean: 75% maximal tetanic force), and pulse-duration (mean: 250 μs) by ±15% of their mean values every 100-ms did not effect the rate of fatigue during isometric contractions of the tibialis anterior and quadriceps femoris muscles of seven spinal cord injured subjects (Thrasher et al., 2005
). Our present study is the first to show improvement in isometric muscle performance using stepwise increases in stimulation frequency during repetitive stimulation. In addition, Kebaetse and Binder-Macleod recently showed that for healthy subjects and subjects with spinal cord injury, starting at a low- and later switching to a high-frequency produced better performance during repetitive non-isometric contractions than stimulation using either a low- or high-frequency alone (Kebaetse and Binder-Macleod, 2004
; Kebaetse et al., 2005
). However, unlike the present study, Kebaetse and colleagues only increased frequency once in their study (i.e., used one frequency step) and did not modulate the stimulation intensity (Kebaetse et al., 2005
). The present study was the first to compare muscle fatigue and performance during systematic stepwise increases of frequency or pulse-duration during repetitive stimulation.
During our study, surface electrical stimulation was used to test isometric muscle fatigue and performance. Surface electrodes are commonly used for FES (Popovic et al., 2001
; Prochazka et al., 1997
; Snoek et al., 2000
) and for the rehabilitation training of individuals with acute paralysis (Chae et al., 1998
) because they are easily placed by the FES-users and are non-invasive. However, implanted nerve cuff electrodes are also used during FES, especially for individuals with chronic paralysis following spinal cord injury or stroke (Chae et al., 2001
; Daly and Ruff, 2000
; Peckham and Knutson, 2005
; Peckham et al., 2002
). Further studies are therefore needed to assess the validity of our current findings during stimulation using implanted electrodes.
During FES, muscle force must repetitively reach a targeted level to enable efficient task performance. The stepwise increases in frequency and pulse-duration in our study caused the peak force to rise above the 20% MVIC targeted force level (). This overshoot of force may have caused greater metabolic energy expenditure (Boska, 1994
; Potma et al., 1994
; Stienen et al., 1995
), and produced greater muscle fatigue (Cooke et al., 1988
; Sahlin et al., 1998
; Westerblad et al., 1998
) than would have been produced if the targeted force was not exceeded. A better strategy would have been to only increase the stimulation frequency or pulse-duration to the level needed to produce the targeted force with minimal overshoot. The combined modulation of frequency and pulse-duration may be a better strategy to improve muscle performance during FES. It is important to note that changes in skeletal muscle fiber-type composition and atrophy following paralysis (Gerrits et al., 2003
) may cause differences in the responses of paralyzed muscles versus muscles of able-bodied individuals. Future studies will use predictive mathematical models (Ding et al., 1998
) that account for changes in the force–frequency curve due to fatigue to determine the appropriate frequency and pulse-duration steps required to generate the targeted force (Chou et al., 2005
; Ding et al., 1998
) in muscles of paralyzed individuals.