A comparison of EMG-based muscle fatigue assessments during dynamic contractions

Several EMG-based approaches to muscle fatigue assessment have recently been proposed in the literature. In this work, two multivariate fatigue indices developed by the authors: a generalized mapping index (GMI) and the first component of principal component analysis (PCA) were compared to three univariate indices: Dimitrov’s normalized spectral moments (NSM), Gonzalez-Izal’s waveletbased indices (WI), and Talebinejad’s fractal-based Hurst Exponent (HE). Nine healthy participants completed two repetitions of fatigue tests during isometric, cyclic and random fatiguing contractions of the biceps brachii. The fatigue assessments were evaluated in terms of a modified sensitivity to variability ratio yielding the following scores (mean ± std.dev.): PCA: (12.6 ± 5.6), GMI: (11.5 ± 5.4), NSM: (10.3 ± 5.4), WI: (8.9 ± 4.6), HE: (8.0 ± 3.3). It was shown that PCA statistically outperformed WI and HE (p < 0.01) and that GMI outperformed HE (p < 0.02). There was no statistical difference among NSM, WI and HE (p > 0.2). It was found that taking the natural logarithm of NSM and WI, although reducing the parameters’ sensitivity to fatigue, increased SVR scores by reducing variability.     

Neuromuscular fatigue development during maximal concentric and isometric knee extensions.

This study aimed to investigate mechanisms of neuromuscular fatigue during maximal concentric and isometric leg extensions inducing similar torque decrements. Nine physically active men performed two separate fatiguing sessions maintained until similar torque decreases were obtained. The first session, only conducted under isokinetic concentric conditions (CON), consisted of three series of 30 maximal voluntary concentric knee extensions (60 degrees/s). The second session, exclusively isometric (ISO), mimicked the torque decreases registered during the CON session while performing three long-lasting ISO contractions. Maximal voluntary torque, activation level (twitch interpolation technique), electromyographic activity (root mean square and median frequency) of the vastus lateralis muscle, and electrically evoked doublet-twitch mechanical properties were measured before and at the end of each of the three series. After the three series, similar torque decrements were obtained for both fatiguing procedures. The total fatiguing contraction durations were not different among procedures. With equivalent voluntary torque decrements, the doublet-twitch amplitude reduction was significantly greater (P<0.01) during the two first series of the CON procedure compared with ISO. No difference was observed for the third series. Although no difference was recorded with fatigue for median frequency changes between CON and ISO, activation levels and root mean square values demonstrated greater reductions (P<0.05) for all three series during the ISO procedure compared with CON. Performing CON or ISO fatiguing exercises demonstrated different fatigue origins. With CON exercises, peripheral fatigue developed first, followed by central fatigue, whereas with ISO exercises the fatigue pattern was inverted.     

Central and peripheral contributions to fatigue in relation to level of activation during repeated maximal voluntary isometric plantar flexions.

This study aimed to investigate central and peripheral contributions to fatigue during repeated maximal voluntary isometric plantar flexions (MVCs). Changes in joint torque, level of activation (LOA), resting twitch amplitude (RT), electromyographic signals (EMG), and presynaptic inhibition of Ia afferents were investigated during 9 bouts of 10 MVCs. MVCs lasted for 2 s and were separated by 1 s. The interval between bouts was 10 s. Electrical stimulation was applied to the tibial nerve; at rest to evoke RTs, M waves, and two (1.5-s interval) H reflexes; with the soleus EMG at 30% of that during MVC to evoke M waves and two H reflexes; and during MVCs to measure LOA. Over the nine bouts, LOA decreased by 12.6% and RT by 16.2%. EMG root mean square during MVCs remained unchanged for the soleus and tibialis anterior muscles, but it decreased for medial gastrocnemius. Peripheral fatigue (decrease in RT) was positively correlated to LOA, whereas central fatigue (decrease in LOA) was not. Depression of both H reflexes suggests that presynaptic inhibition after the first bout was partly induced by homosynaptic postactivation depression of the Ia terminal. The H-reflex-to-M-wave ratio increased with fatigue in both passive and active states, with no change in the ratio of the second H reflex to the first, thereby indicating a decrease of presynaptic inhibition during fatigue. The results indicate that both central and peripheral mechanisms contributed to the fatigue observed during repeated MVCs and that the development of peripheral fatigue was influenced by the level of voluntary activation and initial plantar flexor torque.     

Multichannel EMG-based estimation of fiber conduction velocity during isometric contraction of patients with different stages of diabetic neuropathy

This study compares muscle fiber conduction velocities estimated using surface electromyography during isometric maximal voluntary contraction in different stages of diabetic neuropathy. Eighty-five adults were studied: 16 non-diabetic individuals and 69 diabetic patients classified into four neuropathy stages, defined by a fuzzy expert system: absent (n = 26), mild (n = 21), moderate (n = 11) and severe (n = 11). Average muscle fiber conduction velocities of gastrocnemius medialis, tibialis anterior, vastus lateralis and biceps femoris were assessed using linear array electrodes, and were compared by ANOVA. Conduction velocities were significantly decreased in the moderate neuropathy group for the vastus lateralis compared to other groups (from 18% to 21% decrease), and were also decreased in all diabetic groups for the tibialis anterior (from 15% to 20% from control group). Not only the distal anatomical localization of the muscle affects the conduction velocity, but also the proportion of muscle fiber type, where the tibialis anterior with greater type I fiber proportion is affected earlier while the vastus lateralis with greater type II fiber proportion is affected in later stages of the disease. Generally, the muscles of the lower limb have different responsiveness to the effects of diabetes mellitus and show a reduction in the conduction velocity as neuropathy progresses.     

Constituents of human muscle in isometric fatigue.

Three subjects performed five successive isometric contractions to fatigue; the tension in any one experiment was constant at tensions varying from 20 to 80% of the maximal voluntary contraction (MVC). The interval between contractions was held constant at 11 min. Muscle biopsy specimens were obtained at the start of the experiment, after the first, fourth, and fifth, and before the second and fifth of the successive contractions. The concentrations of ATP, CP, glycogen, and lactate were measured in each sample of muscle. Changes in ATP and glycogen were insufficient to be held accountable for the development of isometric fatigue. Changes in CP and lactate were large after fatigue at intermediate tensions, but those of CP were considered unlikely to be responsible for the fatigue. At tensions of 30-50% MVC the increase in lactate could be responsible for fatigue either directly or by indirect changes in pH; at higher and lower tensions the possibility that lactate is directly implicated in the development of fatigue seems remote.     

Age-related enhancement of fatigue resistance is evident in men during both isometric and dynamic tasks.

It has been suggested that the effects of old age on the ability to resist fatigue may be task dependent. To test one aspect of this hypothesis, we compared the neuromuscular responses of nine young (26 +/- 4 yr, mean +/- SD) and nine older (72 +/- 4 yr) healthy, relatively sedentary men to intermittent isometric (3 min, 5 s contract/5 s rest) and dynamic (90 at 90 degrees /s) maximum voluntary contractions (MVC) of the ankle dorsiflexor muscles. To assess the mechanisms of fatigue (defined as the ratio of postexercise MVC to preexercise MVC), we also measured isometric central activation ratios (CAR), tetanic torque, contractile properties, and compound muscle action potentials before and immediately after exercise. Because dynamic contractions are more neurally complex and metabolically demanding than isometric contractions, we expected an age-related fatigue resistance observed during isometric exercise to be absent during dynamic exercise. In contrast, older men (O) fatigued less than young (Y) during both isometric (O = 0.77 +/- 0.07, Y = 0.66 +/- 0.02, mean +/- SE; P < 0.01) and dynamic (O = 0.45 +/- 0.07, Y = 0.27 +/- 0.02; P = 0.04) contractions (ratio of postexercise to preexercise MVC), with no evidence of peripheral activation failure in either group. We observed no obvious limitations in central activation in either group, as assessed using isometric CAR methods, after both isometric and dynamic contractions. Preexercise half-time of tetanic torque relaxation, which was longer in O compared with Y, was linearly associated with fatigue resistance during both protocols (r = 0.62 and 0.66, P < or = 0.004, n = 18). These results suggest that relative fatigue resistance is enhanced in older adults during both isometric and isokinetic contractions and that age-related changes in fatigue may be due largely to differences within the muscle itself.     

Muscle blood flow during isometric activity and its relation to muscle fatigue

The effect of isometric exercise on blood flow, blood pressure, intramuscular pressure as well as lactate and potassium efflux from exercising muscle was examined. The contractions performed were continuous or intermittent (5 s on, 5 s off) and varied between 5% and 50% maximal voluntary contraction (MVC). A knee-extensor and a hand-grip protocol were used. Evidence is presented that blood flow through the muscle is sufficient during low-level sustained contractions (less than 10% MVC). Despite this muscle fatigue occurs during prolonged contractions. One mechanism for this fatigue may be the disturbance of the potassium homeostasis. Such changes may also play a role in the development of fatigue during intermittent isometric contractions and even more so in the recovery from such exercise. In addition the role of impaired transport of substances within the muscle, due to long-lasting daily oedema formation, is discussed in relation to fatigue in highly repetitive, monotonous jobs.     

Human skeletal muscle responses vary with age and gender during fatigue due to incremental isometric exercise.

The purpose of this study was to compare the magnitude and mechanisms of ankle dorsiflexor muscle fatigue in 20 young (33 +/- 6 yr, mean +/- SD) and 21 older (75 +/- 6 yr) healthy men and women of similar physical activity status. Noninvasive measures of central and peripheral (neuromuscular junction, sarcolemma) muscle activation, muscle contractile function, and intramuscular energy metabolism were made before, during, and after incremental isometric exercise. Older subjects fatigued less than young (P < 0.01); there was no effect of gender on fatigue (P = 0.24). For all subjects combined, fatigue was modestly related to preexercise strength (r = 0.49, P < 0.01). Neither central (central activation ratio) nor peripheral (compound muscle action potential) activation played a significant role in fatigue in any group. During exercise, intracellular concentrations of P(i) and H(2)PO increased more and pH fell more in young compared with older subjects (P < 0.01) and in men compared with women (P < 0.01). These varied metabolic responses to exercise suggest a greater reliance on nonoxidative sources of ATP in young compared with older subjects and in men compared with women. These results suggest that the mechanisms of fatigue vary with age and gender, regardless of whether differences in the magnitude of fatigue are observed.     

Mechanical and metabolic determination of VO2 and fatigue during repetitive isometric contractions in situ

Repetitiveisometric tetanic contractions (1/s) of the caninegastrocnemius-plantaris muscle were studied either at optimal length(L_o) or shortlength (L_s;~0.9 · L_o),to determine the effects of initial length on mechanical and metabolicperformance in situ. Respective averages of mechanical and metabolicvariables were(L_o vs.L_s, allP < 0.05) passive tension (preload) = 55 vs. 6 g/g, maximal active tetanic tension(P_o) = 544 vs. 174 (0.38 · P_o)g/g, maximal blood flow () = 2.0 vs. 1.4 ml · min¹ · g¹,and maximal oxygen uptake(O₂) = 12 vs. 9 µmol · min¹ · g¹.Tension at L_odecreased to0.64 · P_o over20 min of repetitive contractions, demonstrating fatigue; there were nosignificant changes in tension atL_s. In separatemuscles contracting atL_o, was set to that measured atL_s (1.1 ml · min¹ · g¹),resulting in decreased O₂(7 µmol · min¹ · g¹),and rapid fatigue, to0.44 · P_o. Thesedata demonstrate that 1)muscles at L_ohave higher andO₂ values than those at L_s;2) fatigue occurs atL_o with highO₂, adjusting metabolic demand (tension output) to match supply; and3) the lack of fatigue atL_s with lowertension, , andO₂ suggestsadequate matching of metabolic demand, set low by shortmuscle length, with supply optimized by low preload. Thesedifferences in tension andO₂ betweenL_o andL_s groupsindicate that muscles contracting isometrically at initial lengthsshorter than L_oare working under submaximal conditions.

     

Increased fatigue of isovelocity vs. isometric contractions of canine diaphragm

A comparison of fatigue as a loss of force with repeated contractions over time was performed in canine respiratory muscle by isometric (nonshortening) and isovelocity (shortening) contractions. In situ diaphragm muscle strips were attached to a linear ergometer and electrically stimulated (30 or 40 Hz) via the left phrenic nerve to produce either isometric (n = 12) or isovelocity (n = 12) contractions (1.5 s) from optimal muscle length (Lo = 8.8 cm). Similar velocities of shortening between isovelocity experiments [0.19 +/- 0.02 (SD) Lo/S] were produced by maximizing the mean power output (Wmax = 210 +/- 27 mW/cm2) that could be developed over 1.5 s when displacement was approximately 0.30 Lo. Initial peak isometric tension was 1.98 kg/cm2, whereas initial peak isovelocity tension was 1.84 kg/mc2 (P less than 0.01) or 93% of initial isometric tension. Fatigue trials of 5 min were conducted on muscles contracting at a constant duty cycle (0.43). At the end of the trials, peak isovelocity tension had fallen to 50% of initial isometric tension (P less than 0.01), whereas peak isometric tension had only fallen by 27%. These results indicate that muscle shortening during force production has a significant influence on diaphragm muscle fatigue. We conclude that the effects of shortening on fatigue must be considered in models of respiratory muscle function, because these muscles typically shorten during breathing.     

Repeatability of surface EMG parameters at various isometric contraction levels and during fatigue using bipolar and Laplacian electrode configurations.

The objective of this work was to assess the repeatability of two surface electromyographic (sEMG) recording techniques, the classical bipolar configuration and a Laplacian configuration to document their ability to provide reliable information during follow-up studies. The signals were recorded on 10 healthy subjects during voluntary isometric contractions of the biceps brachii muscle at different constant contraction levels. Slopes, area ratios (at 60% of the maximal voluntary contraction (MVC)) and initial values (at 20%, 40%, 60%, 80% and 100% MVC) of the root mean square (RMS), the mean power frequency (MPF) and the muscle fibre conduction velocity (CV) were estimated. Experimental sessions were repeated on three different days with both electrode sets to evaluate the repeatability of sEMG parameter estimates. Classical results were observed, such as an increase in the RMS and the CV with the contraction level. Only initial values of RMS and MPF were shown to be dependent on electrode type. These two parameters presented intra-class correlation coefficient values higher than .80 for high contraction levels. On the whole, the repeatability of the measures was good; however it was better for all sEMG parameter estimates with bipolar electrodes than Laplacian electrodes. Because a bipolar configuration is less selective than a Laplacian one, it provides a global view of muscular activity, which is more repeatable, hence more suitable for follow-up studies.     

The use of correlation integrals in the study of localized muscle fatigue of elbow flexors during maximal efforts.

Innovative applications of non-linear time series analysis have recently been used to investigate physiological phenomena. In this study, we investigated the feasibility of using the correlation integral to monitor the localized muscle fatigue process in the biceps brachii during sustained maximal efforts. The subjects performed isometric maximum contractions until failure in elbow flexion (90 degrees from neutral). The median and the 70th percentile frequency of the Surface electromyography (SEMG) power spectrum, the integrated SEMG, and the Correlation Integral (CI) were evaluated during the trials. The linear correlation between these variables and the elbow torque production was used to quantify the ability of a parameter to follow the fatiguing process. The CI had the highest linear correlation with torque (0.77 (0.12SD)), while the spectral indices correlations with torque were much lower. The decreasing trend of the torque production was followed by the spectral indices for only the beginning part of the contraction, while the CI increased sharply after the torque production fell to about 0.60 of the MVC. This suggests that the CI is sensitive to different changes of the SEMG signal during fatigue than the spectral variables.     

Combined effects of fatigue and temperature manipulation on skeletal muscle electrical and mechanical characteristics during isometric contraction

Peripheral fatigue and muscle cooling induce similar effects on sarcolemmal propagation properties. The aim of the study was to assess the combined effects of muscle temperature (Tm) manipulation and fatigue on skeletal muscle electrical and mechanical characteristics during isometric contraction. After maximum voluntary contraction (MVC) assessment, 16 participants performed brief and sustained isometric tasks of different intensities in low (Tm(L)), high (Tm(H)) and neutral (Tm(N)) temperature conditions, before and after a fatiguing exercise (6s on/4s off at 50% MVC, to the point of fatigue). During contraction, the surface electromyogram (EMG) and force were recorded from the biceps brachii muscle. The root mean square (RMS) and conduction velocity (CV) were calculated off-line. After the fatiguing exercise: (i) MVC decreased similarly in all Tm conditions (P<0.05), while EMG RMS did not change; and (ii) CV decreased to a further extent in Tm(L) compared to Tm(N) and Tm(H) in all brief and sustained contractions (P<0.05). The larger CV drop in Tm(L) after fatigue suggests that Tm(L) and fatigue have a combined and additional effect on sarcolemmal propagation properties. Despite these changes, force generating capacity was not affected by Tm manipulation. A compensatory mechanism has been proposed to explain this phenomenon.     

Mechanomyogram and electromyogram responses of upper limb during sustained isometric fatigue with varying shoulder and elbow postures

To investigate the behavior of mechanomyogram (MMG) and electromyogram (EMG) signals in the time and frequency domains during sustained isometric contraction, MMG and surface EMG were obtained simultaneously from four muscles: upper trapezius (TP), anterior deltoid (DL), biceps brachii (BB), and brachioradialis (BR) of 10 healthy male subjects. Experimental conditions consisted of 27 combinations of 9 postures [3 shoulder angles (SA): 0 degree, 30 degrees, 60 degrees and 3 elbow angles (EA): 120 degrees, 90 degrees, 60 degrees] and 3 contraction levels: 20%, 40%, and 60% of maximum voluntary contraction (MVC). Subjective evaluations of fatigue were also assessed using the Borg scale at intervals of 60, 30, and 10 sec at 20%, 40%, and 60% MVC tests, respectively. The mean power frequency (MPF) and root mean square (RMS) of both signals were calculated. The current study found clear and significant relationships among physiological and psychological parameters on the one hand and SA and EA on the other. EA's effect on MVC was found to be significant. SA had a highly significant effect on both endurance time and Borg scale. In all experimental conditions, significant correlations were found between the changes in MPF and RMS of EMG in BB with SA and EA (or muscle length). In all four muscles, MMG frequency content was two or three times lower than EMG frequency content. During sustained isometric contraction, the EMG signal showed the well-known shift to lower frequencies (a continuous decrease from onset to completion of the contraction). In contrast, the MMG spectra did not show any shift, although its form changed (generally remaining about constant). Throughout the contraction, increased RMS of EMG was found for all tests, whereas in the MMG signal, a significant progressive increase in RMS was observed only at 20% MVC in all four muscles. This supports the hypothesis that the RMS amplitude of the MMG signal produced during contraction is highly correlated with force production. Possible explanations for this behavioral difference between the MMG and EMG signals are discussed.     

The reliability and validity of fatigue measures during short-duration maximal-intensity intermittent cycling

The purpose of the present study was to assess the reliability and validity of fatigue measures, as derived from 4 separate formulae, during tests of repeat sprint ability. On separate days over a 3-week period, 2 groups of 7 recreationally active men completed 6 trials of 1 of 2 maximal (20 x 5 seconds) intermittent cycling tests with contrasting recovery periods (10 or 30 seconds). All trials were conducted on a friction-braked cycle ergometer, and fatigue scores were derived from measures of mean power output for each sprint. Apart from formula 1, which calculated fatigue from the percentage difference in mean power output between the first and last sprint, all remaining formulae produced fatigue scores that showed a reasonably good level of test-retest reliability in both intermittent test protocols (intraclass correlation range: 0.78-0.86; 95% likely range of true values: 0.54-0.97). Although between-protocol differences in the magnitude of the fatigue scores suggested good construct validity, within-protocol differences highlighted limitations with each formula. Overall, the results support the use of the percentage decrement score as the most valid and reliable measure of fatigue during brief maximal intermittent work.     

Assessment of muscle fatigue in low level monotonous task performance during car driving.

Knowledge of the muscle activation and the development of muscle fatigue may provide more inside in the effects of long-term driving in the occurrence of health problems in the neck/shoulder/back area. The basic assumption behind fatigue detection with electromyography (EMG) is an increase in the EMG amplitude and a decrease of the mean frequency (MF). This study aimed at checking this assumption in monotonous task performance with low level activity during car driving. Surface electromyography was captured from left and right trapezius and deltoid muscles, during a repetitive, non-continuous, driving task (gearing and steering) and the active parts were separated from the non-active parts. Muscle stiffness was reported by more than half of the subjects after a 1 h drive. Only for the active parts a significant decrease of the MF was seen. But also the EMG amplitude decreased significantly. Two possible mechanisms are posted in literature for this finding: no extra recruitment of motor units (MU) and potentiation of muscle fibers. Literature also hypothesizes that low-force occupational work engages only a fraction of the MU available for recruitment and that these units are selectively type I muscle fibers (Cinderella fibers). Initiators of this phenomenon are probably the time lag between activations and the stress from driving and vibration exposure.     

The relative changes in isometric force and work during fatigue and recovery in isolated toad sartorius muscle

Toad sartorius muscle was subjected to sinusoidal varying length changes at 2 Hz to measure work. Both isometric tetanic force and work per cycle were measured before, during, and after a 3-min fatigue. Both isometric tetanic force and positive work, the work done by the muscle during the shortening part of the cycle, rapidly decreased in parallel in the first 40 s of fatigue. Thereafter, force continued to decrease, but at a slower rate, to about 10% of prefatigue values, whereas positive work levelled off at about 30% of prefatigue values. Negative work, the work done on the muscle during the lengthening part of the cycle, increased during fatigue to the extent that net work became negative. This was due to a prolonged relaxation, which resulted in active force still being generated while the muscle was being stretched. Work and force recovered at about the same rate. Isometric force measurements alone do not give any clear indication that net work will be negative under a particular set of experimental conditions.     

Correlations between short-time Fourier- and continuous wavelet transforms in the analysis of localized back and hip muscle fatigue during isometric contractions.

The aims of the current study were to examine the stationarities of surface electromyographic (EMG) signals obtained from eight bilateral back and hip muscles during a modified Biering-Sørensen test, and to investigate whether short-time Fourier (STFT) and continuous wavelet transforms (CWT) provided similar information with regard to EMG spectral parameters in the analysis of localized muscle fatigue. Twenty healthy subjects participated in the study after giving their informed consent. Reverse arrangement tests showed that 91.6% of the EMG signal epochs demonstrated no significant trends (all p > 0.05), meaning 91.6% of the EMG signal epochs could be considered as stationary signals. Pearson correlation coefficients showed that STFT and CWT in general provide similar information with respect to the EMG spectral variables during isometric back extensions, and as a consequence STFT can still be used.     

Mathematical model of muscular fatigue I. Metabolite level changes during exercises of different intensities

To quantify muscular fatigue we study by a mathematical modelization the changes of the levels of ATP and various metabolites, and of the oxygen uptake with the intensity of a requested muscular exercise. The proposed model is consistent with the main phenomena experimentally observed during exercises; it allows also one to extrapolate these experimental results to any condition and to simulate exercise-complete and/or incomplete recovery cycles. The lactate kinetics allows one particularly to justify the empirical optimization of rest rhythms during exercises of long duration. Equations are solved using a 4th order Runge-Kutta method; computations are performed on an IRIS 80 CH computer.