Neural Activation after Maximal Isometric Contractions at Different Muscle Lengths


Medicine & Science in Sports & Exercise: May 2006 - Volume 38 - Issue 5 - pp 937-944
doi: 10.1249/01.mss.0000218136.58899.46

Purpose: To investigate i) whether neural activation dependence on muscle length is preserved with neuromuscular fatigue and ii) whether fatigue induced by a maximal isometric exercise is muscle length dependent.

Methods: Twelve male subjects performed two fatiguing quadriceps muscle exercises: FS is the fatigue carried out at short muscle length (S) (S = 40° of knee flexion) and FL is the fatigue at long muscle length (L) (L = 100°). Before and after each fatiguing exercise (i.e., three maximal isometric contractions maintained until 80, 60, and 40% of the initial maximal torque, respectively), activation level (AL, assessed by means of twitch interpolation technique), EMG activity (RMS), and peak doublet torque (Pd) were measured at the two lengths (S and L).

Results: First, AL was greater (P < 0.05) at L compared with S before and after both exercises. Second, despite a similar decrease in maximal voluntary torque (~21% of the initial value) after the two exercises, AL and RMS were significantly reduced after FS (P < 0.05) but remained unchanged after FL, whereas the Pd decrease was more pronounced after FL than FS (P < 0.05). Nevertheless, after a given fatiguing exercise (i.e., FS or FL), AL, RMS, and Pd changes were similar at both postexercise test lengths (S and L).

Conclusion: These results clearly demonstrate that i) the neural activation dependence on quadriceps muscle length is maintained with fatigue, and ii) neuromuscular fatigue after maximal isometric contractions is dependent on the muscle length at which the exercise is performed: short length preferentially induces neural activation impairment, whereas long length leads to higher contractile failure.

1 INSERM/ERIT-M Motricity-Plasticity, Faculty of Sport Science, University of Burgundy, Dijon, FRANCE; 2INRS, Occupational Physiology Laboratory, Nancy, FRANCE; and 3Motor Performance and Health, Faculty of Sport Science, Strasbourg, FRANCE

Address for correspondence: Kévin Desbrosses, INRS, Occupational Physiology Laboratory, Avenue de Bourgogne, BP 24, 54501 Vandoeuvre Cedex, France; E-mail:

Submitted for publication June 2005.

Accepted for publication December 2005.

Article Outline

The maximal isometric torque that human muscle can produce is dependent on lever arm and muscle length (28). On both sides of the optimal muscle length (i.e., the length at which the force production is maximal), the torque can be decreased, reflecting a reduction in actin-myosin cross-bridge number and tension (12). During a maximal voluntary contraction (MVC), this disadvantage in torque production capacity may be partially compensated for by increased neural activation (2,19), corresponding to the recruitment of additional motor units and/or the increase in their firing frequency. So, in fresh muscles, it appears that neural activation is muscle length dependent. However, with the aim to assess neuromuscular adaptations, we can wonder whether this dependence is preserved despite alterations like those induced by fatigue.

Neuromuscular fatigue is generally defined as a decrease in torque production capacity (10) and is traditionally divided into two distinct components (8): i) central fatigue, corresponding to a deficiency in neural excitation from the motor cortex to the neuromuscular junction, and ii) peripheral fatigue, related to failure from the neuromuscular junction to cross-bridges. Several studies observed greater fatigue after submaximal (13,17,21,23,29) or maximal (6) isometric knee extensions performed at long (L) compared with short (S) muscle length. As reported by these authors, peripheral phenomena (i.e., excitation-contraction coupling, energy consumption) could account for fatigue dependence on muscle length because no difference in neural activation modifications was found between the different quadriceps muscle lengths (17,23). However, for the quadriceps femoris muscle, these measurements were carried just out after submaximal exercises, which obviously does not involve the same fatigue processes as those in contractions performed at maximal intensity. Indeed fatigue is task dependent; Linnamo et al. (18) showed that central fatigue was increased with a higher contraction intensity. In addition, it should be noted that fatigue is generally evaluated at the length at which the fatiguing exercise is performed, and not at another length. As a result, activation and fatigue dependence on muscle length still remain undocumented.

In light of these considerations, the purposes of the present study were to investigate i) whether the neural activation dependence on muscle length is persistent with fatigue and ii) whether the effects of muscle length on neuromuscular fatigue after maximal isometric contractions. We assessed activation level by means of the twitch interpolation technique (1), EMG activity, and maximal voluntary and electrically evoked doublet torque at two different muscle lengths (short and long) before and after a fatiguing exercise performed at each of these lengths in two different sessions. To compare a similar level of fatigue for both sessions, the fatiguing exercises were maintained until equivalent torque decreases were obtained.

Back to Top | Article Outline



Twelve male subjects (23.6 ± 1.7 yr, 178.2 ± 7.6 cm, 73.6 ± 9.8 kg; mean ± SD) without a history of knee joint pathology took part in this investigation. They were physically active (i.e., taking part in sport training for at least 4 h·wk−1). All subjects gave written informed consent to participate in the study after being informed in detail about the investigation's purposes and procedures. The experimental protocol was conducted according to the Declaration of Helsinki and approved by the University of Burgundy committee on human research.

Back to Top | Article Outline

Experimental Protocol

Two testing sessions were randomly conducted 2 wk apart. Each session consisted of a fatiguing exercise performed at a given quadriceps muscle length. The short length fatiguing session (FS) was carried out at 40° of knee flexion (S = short length), with 0° being the full extension of the knee joint. The long length fatiguing session (FL) was conducted at 100° of knee flexion (L = long length). Knee joint angles were determined during an MVC. Pre- and postfatigue measurements were executed at the two lengths, S and L, in a randomized order for both sessions (Fig. 1). As shown by Pincivero et al. (22), the maximal quadriceps torque production is close to 70° of knee flexion; thus, 40 and 100° were chosen in order to place the quadriceps muscle aside from the optimal length, that is, at a "short" and a "long" length. Each session began with a standardized warm-up, which consisted of 20 submaximal voluntary isometric contractions (with a 15-s rest period) of increasing intensity (starting at approximately 10% MVC) performed at both muscle lengths until the MVC was obtained.

Back to Top | Article Outline

Fatiguing exercise.

The fatiguing exercise was randomly completed at one muscle length (S or L), the other length being retained for the second session (Fig. 1). The exercise was composed of three MVC maintained until a torque target value, determined as a percentage of the prefatigue maximal torque. The first fatiguing contraction was maintained until 20% reduction of the prefatigue MVC (−20%C), the second (−40%C), and third contractions (−60%C) until reductions of 40 and 60%, respectively. There was a 1-min rest period between each fatiguing contraction. This protocol therefore induced identical relative torque reductions for both sessions in order to compare a similar fatigue level. We verified that each fatiguing contraction started at similar relative torque levels between the two sessions to avoid a possible recovery effect depending on muscle length. The three fatiguing contractions allowed the evaluation of the neuromuscular fatigue development. During each contraction, subjects were verbally encouraged to deliver maximal performance for as long as possible.

Back to Top | Article Outline

Pre- and postfatigue measurements.

Pre- and postfatigue measurements consisted of two 5-s maximal voluntary isometric knee extensions completed for each length (S and L) and accompanied by an electrical stimulation procedure (Fig. 2). Stimulations consisted of i) two paired (doublets) and one single transcutaneous electrical impulses, delivered 2 s apart on relaxed muscles, before the MVC; ii) two superimposed doublets at a 3-s interval applied during the maximal voluntary torque plateau; and iii) two doublets delivered 2 s apart on relaxed muscles, 1 s after the MVC. To determine the amount of coactivation, a 5-s maximal isometric knee flexion was performed at both lengths. Contractions were performed successively without any rest period. Postfatigue measurements were completed after the −60% C, with a 15-s transition period corresponding to changes of joint angle.

Back to Top | Article Outline

Fatigue development.

During the three fatiguing contractions (−20, −40, and −60%C), two superimposed doublets were delivered (at a 2-s interval) using a trigger connected to the dynamometer when subjects attained the target torque deficit (namely, a reduction of 20, 40, or 60% of the MVC). Immediately after, subjects were instructed to relax, and again two doublets were delivered 2 s apart. The duration of each fatiguing MVC was recorded.

Back to Top | Article Outline

Data Collection

Torque measurements.

Torque of the right quadriceps and hamstring muscles was measured by an isokinetic dynamometer (Biodex, Shirley, NY) using the isometric mode. Torque (150-Hz sampling rate) and EMG signals was passed through an AD converter (ITC-16, Heka Elektronik, Lambrecht, Germany), recorded with a 2.0-kHz sampling frequency, and analyzed offline using Tida software (Heka Elektronik, Lambrecht, Germany). Subjects were seated in a comfortable position, at a 70° hip flexion angle (0° = full extension) and were secured with Velcro straps across the thorax and pelvis to limit disturbing movements. The lower leg was secured to the dynamometer lever arm with the lateral femoral condyle (i.e., knee joint axis) aligned with the rotation axis of the dynamometer.

Back to Top | Article Outline

Electrical stimulations.

Transcutaneous electrical stimulations were delivered using a high-voltage (maximal voltage = 400 V) constant-current stimulator (Digitimer DS7, Hertfordshire, UK). Stimuli were applied via bipolar surface electrodes. The ball probe cathode (~10 mm in diameter) was pressed onto the femoral triangle, over the femoral nerve, and moved to the position giving the greatest contractions of the whole quadriceps muscle group. The anode, a 10 × 5-cm self-adhesive electrode (Compex, Ecublens, Switzerland), was positioned midway between the superior aspect of the greater trochanter and the inferior border of the iliac crest. Single stimulations were used to determine the peak-to-peak amplitude of the compound muscle action potential (M-wave), whereas doublets were used to investigate contractile properties. Doublets were used to increase the signal-to-noise ratio and also to minimize any series elastic effects on torque production. Single stimulation was a 1-ms duration square-wave impulse, and the doublet consisted of two single stimulations at a 100-Hz frequency. Stimulations were delivered at an intensity ranging from 40 to 120 mA. After the warm-up, the maximal stimulus intensity was determined with single impulses, at L, by progressively increasing the current until there was no further increase in the evoked torque and in the M-wave. It was verified that this intensity was also maximal at S. Supramaximal stimuli (maximal intensity + 10%) were then applied for all measurements.

Back to Top | Article Outline

EMG recordings.

EMG recordings from the vastus lateralis (VL), rectus femoris (RF), and long head of the biceps femoris (BF) were obtained by means of pairs of silver chloride surface EMG electrodes (Contrôle Graphique Medical, Brie-Comte-Robert, France) positioned parallel to the longitudinal axis of the muscular fibers over the middle of the muscle belly. The interelectrode distance was 2.5 cm (center to center). The reference electrode was placed on the opposite wrist. Previous to the electrode placement, the skin was shaved, rubbed with sandpaper, and cleaned with alcohol to reduce interelectrode impedance (Z < 2 kΩ). The anatomical electrode positions were recorded, using indelible ink to ensure identical electrode placement for the subsequent session. EMG signals were amplified with a bandwidth frequency ranging from 1.5 Hz to 2.0 kHz (common mode rejection ratio = 90 dB; impedance = 100 MΩ; gain = 1000) and recorded with a sampling frequency of 2.0 kHz.

Back to Top | Article Outline

Data Analysis

Torque parameters.

The torque voluntarily produced and electrically evoked were considered. Maximal voluntary torque was measured during each 5-s MVC performed before the onset of the fatiguing procedure. Prefatigue values allowed the determination of torque reduction for the three fatiguing contractions. The torque amplitude evoked by doublets (Pd) was measured on relaxed muscle, before and after each MVC. Pre-MVC Pd was used to assess muscular contractile properties and postactivation potentiation (PAP) by comparison with post-MVC Pd. In accordance with the twitch interpolation technique, for each MVC, the torque increment elicited by the superimposed doublets (A in Fig. 2) was compared with post-MVC Pd amplitude (B in Fig. 2) to estimate activation level (AL) as follows (1): AL (%) = (1 − (A/B)) × 100.

Back to Top | Article Outline

EMG recordings.

M-wave peak-to-peak amplitude, evoked by a single electrical impulse delivered on relaxed muscles, was measured for VL and RF. To quantify the EMG activity, the root mean square values (RMS) were calculated and averaged over a 500-ms period before the two superimposed stimulations during the torque plateau. The RMS values for the two superficial knee extensors were normalized by the corresponding M-wave amplitude (RMS·M−1).

Because RMS·M−1 of both muscles similarly evolve (see prefatigue measurements results), a mean quadriceps femoris RMS·M−1 value (aRMS) was calculated by averaging the RMS·M−1 of VL and RF to present a unique representative value of the quadriceps femoris muscle group. The coactivation level was determined by normalizing the BF RMS value, when the muscle acted as an antagonist, with respect to the BF RMS recorded during maximal knee flexions at S and L.

Back to Top | Article Outline

Statistical Analysis

Data are reported as mean values ± SD, except figures where SE are presented for greater clarity. Normality of the data was verified by using the Kolmogorov-Smirnov test, and homeoscedasticity was assessed by applying the Bartlett test. A three-factor (length × session × fatigue) ANOVA with repeated measures was performed on each variable to assess the fatigue effects. Two-factor (length × session) ANOVA with repeated measures were used separately on pre- and postfatigue values to determine the dependence of neuromuscular parameters on muscle length. Another two-factor (session × intermediate fatiguing contraction; i.e., −20 and −40%C) ANOVA with repeated measures was performed on contraction duration, AL, and Pd for the first two fatiguing contractions. F ratios were considered significant at P < 0.05. A Newman-Keuls post hoc test was conducted if significant main effects or interactions were present. Statistical analyses were undertaken using Statistica software for Windows (StatSoft, Version 5, Tulsa, OK).

Back to Top | Article Outline


Prefatigue Measurements

The neuromuscular properties examined in prefatigue conditions are presented in Table 1 as averages of FS and FL because no difference was found between sessions for a given length (S or L). Maximal voluntary torque, Pd, and PAP, as well as M-wave amplitudes showed significantly higher values (P < 0.05) at S than at L. In contrast, the AL, aRMS, and BF coactivation level were lower (P < 0.05) at S than at L. RMS·M−1 values of the two knee extensors were not reported because both muscles produced similar results.

Back to Top | Article Outline

Fatigue Development

Contraction durations needed to achieve the required torque reductions (−20, −40, and −60% of the initial MVC) were significantly longer for FS as compared with FL (Fig. 3), pointing to a belated apparition of fatigue at short lengths. Consequently, the total duration of the three contractions was 68.4 ± 19.6% longer (P < 0.05) for FS (140.1 ± 44.6 s) than FL (89.6 ± 16.9 s). At the end of the three fatiguing contractions (−20, −40, and −60%C), AL and Pd (presented in Table 2) were only assessed at the length solicited during the fatiguing contractions (S for FS and L for FL). At FS, for both fatiguing contractions, AL decreased significantly (P < 0.05) in comparison with prefatigue values. At FL, AL decreased significantly (P < 0.05) for −60%C in comparison with prefatigue values, but remained unchanged for −20 and −40%C. When comparing FS and FL, AL changes were significantly different at −40 and −60%C. During FS, Pd decreased (P < 0.05) after −40 and −60%C, whereas it declined significantly for both fatiguing contractions during FL. Looking at the two sessions, different changes in Pd (P < 0.05) were found for the three fatiguing contractions, with a significantly greater decrease at FL.

Back to Top | Article Outline

Postfatigue Measurements

Torque parameters.

After the third fatiguing contraction (−60%C), two 5-s MVC were randomly performed at each length. The MVC torque values significantly declined for the four postfatigue conditions (i.e., S after FS, L after FS, S after FL, and L after FL; Fig. 1), with no difference observed among conditions (Table 3) as required by the protocol. The average torque reduction was 20.8 ± 7.6%. So, in a state of fatigue, MVC torque values were i) 176.7 ± 42.4 and 145.1 ± 27.2 N·m at S and L, respectively, after FS; and ii) 162.5 ± 25.4 and 148.9 ± 34.9 N·m at S and L, respectively, after FL. Pd significantly decreased after both sessions with no difference between S and L (Fig. 4a). However, the Pd decrease was more pronounced (P < 0.05) after FL than after FS. Thus, Pd amplitude values were i) 88.8 ± 25.3 and 73.3 ± 24.9 N·m at S and L, respectively, after FS; and ii) 75.5 ± 19.1 and 67.4 ± 15.3 N·m at S and L, respectively, after FL.

Back to Top | Article Outline

Neural activation.

A similar decline (P < 0.05) in AL was observed at S and L after FS, but no modifications were recorded after FL (Fig. 4?B). So, AL values were i) 79.8 ± 13.1 and 87.1 ± 10.6% at S and L, respectively, after FS; and ii) 84.9 ± 9.9 and 93.4 ± 4.9% at S and L, respectively, after FL. aRMS evolved similarly to AL (Fig. 4C). No difference in prefatigue values was found after FL. In contrast, after FS, a similar aRMS decrease (P < 0.05) was observed at both lengths. aRMS values were 0.046 ± 0.023 and 0.061 ± 0.015 at S and L, respectively, after FS, and 0.055 ± 0.032 and 0.073 ± 0.037 at S and L, respectively, after FL. Concerning the antagonistic muscle, a coactivation reduction (P < 0.05) was observed, but ANOVA did not provide any evidence of a session effect.

Back to Top | Article Outline

M-wave and PAP.

The M-wave peak-to-peak amplitudes of the two superficial agonistic muscles were not altered (P < 0.05) by fatigue. On the contrary, a significant decline in PAP (P < 0.05) was observed at both lengths after FS and at S after FL (Table 3).

Back to Top | Article Outline


The results of the present study demonstrate a preservation of neural activation dependence on quadriceps muscle length with neuromuscular fatigue. Additionally, this investigation demonstrated that neuromuscular fatigue origins were related to the quadriceps muscle length at which the fatiguing exercise was performed: short length preferentially induced a central fatigue, whereas long length essentially produced peripheral failures.

Back to Top | Article Outline

Contractile Properties of Fresh Muscle

As previously reported in fresh muscles (14,24), MVC torque was lower at L compared with S. The higher coactivation level recorded at L could slightly contribute to the MVC torque variation, but such a difference in torque would probably be the consequence of mechanical properties related to the joint angle as muscle length and lever arm (28). Quadriceps muscle length variation, corresponding to approximately 48% between a 40 and 100° knee angle (28), could be the main origin of these torque differences. Indeed, the quadriceps moment arm was found maximal at approximately a 30-50° knee angle (15), whereas maximal torque appears near 70° (22). So, in our study, the disadvantage in MVC torque at L mainly results from impairments induced by muscle length variations. Both muscle lengths (S and L) were aside from the maximal torque production angle (close to 70° (22)). Consequently, the actin-myosin overlap and, subsequently, the cross-bridge number were surely suboptimal. However, Pd (allowing the assessment of cross-bridge formation) was significantly smaller at L compared with S, as previously reported (11). Therefore, the muscle torque disadvantage evaluated at L could probably be due to a weakness in the contractile apparatus. Moreover, at L, the decrease in myofilament lattice spacing (due to sarcomere lengthening) could modify the regulatory myosin light chain phosphorylation (30), possibly resulting in impairments of the sensitivity of the myofilaments for intracellular Ca2+ concentration (24). This may explain the decline in PAP (related to Ca2+ sensitivity) with increasing muscle length, as previously reported (24). Muscle length could also affect neuromuscular transmission efficiency (19), as shown in the present study, by the reduction in muscle compound action potential at L. However, changes in M-waves, with respect to knee joint angle, should be interpreted with great care because changes in the relative electrode position with the joint angle could influence surface EMG signal. In summary, several peripheral impairments point to the presence of a greater contractile disadvantage at the long quadriceps muscle length compared with the short one.

Back to Top | Article Outline

Neural Activation Dependence on Muscle Length

To maintain a sufficient torque production, some authors suggest that the contractile disadvantage depending on muscle length could be partially compensated for by an increase in neural activation (2,19). The present study confirmed this assumption, showing greater neural activation (AL and RMS) at the length (L) presenting the greater contractile disadvantage. Earlier investigations similarly obtained a higher level of neural activation at long quadriceps muscle lengths (4,7,22), whereas others reported constant neural activation whatever the muscle length (23) or greater neural activation at short quadriceps muscle lengths (2,17,19). Conflicting results can be attributed to the different muscle lengths (i.e., knee angles) used in these protocols and thus to the relative meaning attributed to the short or long length notion. In the present study, we chose to position quadriceps femoris muscle at equal angular intervals on both sides of the optimal angle, generally located close to 70° of knee flexion (22). Although the majority of previous studies demonstrated a neural activation dependence on muscle length, none of them have addressed its evolution with fatigue. Some authors (17,23) evaluated activation level changes with fatigue, but only at one knee angle (the angle at which the fatiguing effort was carried out). Consequently, they assessed fatigue dependence on muscle length, but not the activation dependence on muscle length in fatigue conditions.

For that reason, we analyzed neuromuscular properties, and specifically neural activation, at two muscle lengths after two fatiguing sessions performed at these lengths. Our results showed that in fatigue conditions and at whatever muscle length the fatiguing exercise was performed, AL and aRMS were higher at L than at S. On the contrary, MVC torque and Pd were lower at L than at S. Thus, it appears that neural activation dependence on muscle length is preserved in fatigue conditions, and the higher activation level at L may be viewed as a neurophysiological mechanism partially compensating for the greater contractile disadvantage, as already described for fresh muscles. As reported by Becker and Awiszus (4), activation dependence on knee joint angle may be related to muscle spindle length. As a matter of fact, a longer spindle length with increasing knee flexion leads to a greater Ia input to the motor neuron pool and consequently to a greater excitatory drive. Another factor that could play a role is the variation in ligament tension with the joint angle. This variation could influence the gamma innervation of muscle spindles and thus contribute to the voluntary activation modulation (4).

Back to Top | Article Outline

Fatigue Dependence on Muscle Length

Maximal voluntary torque reductions were similar across the four postfatigue conditions (i.e., S after FS, L after FS, S after FL, and L after FL). Torque decline, about 21% of prefatigue values, is consistent with the findings of Chan et al. (6) involving maximal quadriceps fatiguing exercise. Because of a transition period after the effort to change knee angles (~15 s), torque decline was lower than that induced by the last fatiguing contraction (−60%C), certainly owing to a substantial torque recovery. With the identical torque decline for all conditions, fatiguing sessions can be compared for a similar amount of fatigue. After a specific fatiguing session (i.e., FS or FL), neuromuscular parameters changed in a similar manner for both lengths (i.e., S and L). However, these parameters (excepted voluntary torque) were differently modified when comparing the two fatiguing sessions: FS induced a greater deficit in AL and aRMS, whereas FL resulted in a higher Pd decline. It follows that neuromuscular alterations observed in fatigue conditions were related to the length at which the fatiguing exercise was performed and not to the length at which measurements were carried out. We can emphasize that this muscle length specificity appeared quickly during fatigue development, with different Pd modifications between FS and FL from 20% torque reduction and different AL changes from 40%.

Back to Top | Article Outline

Central fatigue after FS.

To reach the same relative torque reductions (−20, −40, and −60% of prefatigue MVC), fatiguing contraction durations were longer for FS than for FL. Such a finding is in keeping with several studies that reported greater fatigue after exercises performed at long muscle lengths (6,13,17,21,23,29). Several hypotheses, such as lower energy consumption (7,9,13,17) during fatiguing exercise performed at a short length, were advanced to explain the endurance time dependence on muscle length. Underlying causes for this lower energy consumption were not clearly identified. A possible explanation could be related to the lower cross-bridge number (9). In our study, S certainly allowed a larger number of cross-bridges than L (as shown by the greater Pd at S). According to the assumption of Fitch and McComas (9), the energy consumption should be greater at S than at L and thus involve reduced contraction durations at S. This assumption was not corroborated by our results since durations were longer at S. Other authors (3,26) recorded similar metabolism at different muscle lengths and thus refuted that fatigue dependence on muscle length could be related to energy consumption. These authors assumed that the relationship between activation level and muscle length could explain such a difference in endurance times. It was possible that an activation deficit at a short length could induce a submaximal use of contractile elements permitting a greater endurance (26). Additionally, the greater Ca2+ sensitivity at S could facilitate torque production and make fatiguing exercise relatively easier at S than at L. On the contrary, a greater AL at L could certainly induce greater utilization of contractile elements and consequently a more exhaustive exercise. So, AL could at least partly account for fatigue dependence on muscle length.

Considering that exercise performed at a short muscle length could be less exhaustive for contractile elements and that AL and aRMS notably decreased after FS, it is possible that contraction was limited by central factors. In the present study, we hypothesized that prolonged contraction duration obtained for FS would likely result in increased metabolite accumulation (16) and subsequently in a pH decrease. This metabolite accumulation could have some neural repercussions, thus explaining the AL and aRMS decline recorded after FS. Indeed, the increased activity of group III and IV muscle afferents due to metabolic modifications may reduce neural excitation at a spinal level, with an inhibitory effect on the motor neuron pool (27), and at a supraspinal level by affecting the central neural drive (5). So, a possible explanation for central fatigue at a short length could be related to reduced activation, which would induce a submaximal use of the contractile elements (facilitated by the greater Ca2+ sensitivity) and subsequently a lengthening of the contraction duration. A consequence of increasing time could be the accumulation of metabolites (16), resulting in central failure via motor neuron pool inhibition by group III and IV muscle afferents (5,27). During FS, a decrease in AL was recorded from 20% of torque reduction, suggesting that central failure was rapidly responsible for fatigue development. Furthermore, central failure became progressively greater with the exercise duration, up to an AL decline of more than 40% observed at the end of the last fatiguing MVC. However, it should be mentioned that some studies did not find any difference in activation level comparing different quadriceps muscle lengths (17,23). The submaximal intensity protocol used in these investigations could account for the disagreement with our results.

In addition to central failures, FS presented some peripheral alterations, as demonstrated by Pd and PAP decreases. This result could also be a consequence of metabolic accumulation. Indeed, pH decrease may participate in contractile failures at the cross-bridge level (20) and may alter myofilament sensitivity for Ca2+ (25).

Back to Top | Article Outline

Peripheral fatigue after FL.

A reduction in AL was recorded at the end of the third fatiguing contraction performed at L. However, the transition period (after the effort) to change knee angles (~15 s) enabled this central failure to recover since AL and aRMS were unchanged after the exercise performed at L. Regarding postfatigue MVC, results allowed exclusion of any central fatigue phenomenon after FL. In accordance with previous investigations (9,14,29), greater peripheral alterations were detected after L exercise since Pd reductions were greater after FL as compared with FS. These peripheral alterations strongly affected torque production during FL fatigue development; as a matter of fact, Pd declined from the first fatiguing contraction (−20%C) and reached a 40% decrease at the end of the third MVC. After FL, the absence of M-wave impairments, as previously reported (9,17), possibly suggests preservation of neuromuscular transmission. A similar decrease in PAP after FL and FS reveals that Ca2+ sensitivity cannot account for the greater Pd decrease after FL. Four days after isometric fatiguing contractions, Jones et al. (14) found greater tenderness (related to muscle damage) when contractions were performed at L. Accordingly, it was hypothesized that isometric contractions performed at L could probably induce muscle damage (14,29). At L, the reduced proportion of active muscle fibers (due to less actin-myosin overlap) and the higher level of neural activation could induce major tension per fiber, thereby generating an important stress on the contractile structure. Moreover, some sarcomeres could be overstretched when muscle was in a lengthened position ("creep" phenomenon), inducing mechanical cross-bridge disruption and consequently muscle damage. It was suggested that such damage could affect elements in series with contractile proteins (14). Additionally, torque decreases recorded after FL could be the result of muscle damage. However, this explanation remains hypothetical because markers of muscle damage were not evaluated.

Back to Top | Article Outline


Our study confirms the dependence of neural activation on fresh muscle length and shows preservation of this dependence in fatigue conditions. Indeed, whatever the muscle length at which fatigue was induced, the neuromuscular system maintains greater neural activation at muscle lengths presenting the greater contractile disadvantages. In addition, the present study points to fatigue dependence on muscle length. It appears that whatever the muscle length at which fatigue was evaluated, i) maximal isometric fatiguing contractions performed at S mainly affect central nervous system efficiency, as shown by the significant deficit in neural activation, whereas ii) maximal isometric fatiguing efforts carried out at L principally alter contractile structures. In conclusion, our results demonstrate that the angle at which exercise is performed is a significant parameter to take into account for rehabilitation, sport training, or physiological research because it strongly influences the neuromuscular system response. For example, in rehabilitation, the choice of the exercise joint angle could be made in relation to the pathology and, consequently, to the neuromuscular component that has to be mainly solicited (i.e., central or peripheral).

The authors gratefully acknowledge Dr. Nicola A. Maffiuletti for his useful suggestions and Mrs. Mary Bouley for carefully reviewing the English text.

Back to Top | Article Outline


1. Allen, G. M., S. C. Gandevia, and D. K. McKenzie. Reliability of measurements of muscle strength and voluntary activation using twitch interpolation. Muscle Nerve 18:593-600, 1995.
2. Babault, N., M. Pousson, A. Michaut, and J. Van Hoecke. Effect of quadriceps femoris muscle length on neural activation during isometric and concentric contractions. J. Appl. Physiol. 94:983-990, 2003.
3. Baker, A. J., P. J. Carson, A. T. Green, R. G. Miller, and M. W. Weiner. Influence of human muscle length on energy transduction studied by 31P-NMR. J. Appl. Physiol. 73:160-165, 1992.
4. Becker, R., and F. Awiszus. Physiological alterations of maximal voluntary quadriceps activation by changes of knee joint angle. Muscle Nerve 24:667-672, 2001.
5. Bigland-Ritchie, B. R., N. J. Dawson, R. S. Johansson, and O. C. Lippold. Reflex origin for the slowing of motoneurone firing rates in fatigue of human voluntary contractions. J. Physiol. 379:451-459, 1986.
6. Chan, A. Y., F. L. Lee, P. K. Wong, C. Y. Wong, and S. S. Yeung. Effects of knee joint angles and fatigue on the neuromuscular control of vastus medialis oblique and vastus lateralis muscle in humans. Eur. J. Appl. Physiol. 84:36-41, 2001.
7. de Ruiter, C. J., M. D. de Boer, M. Spanjaard, and A. de Haan. Knee angle dependent oxygen consumption during isometric contractions of the knee extensors determined with near infrared spectroscopy. J. Appl. Physiol. 99:579-586, 2005.
8. Edwards, R. H. T. Biochemical bases of fatigue in exercise performance: catastrophe theory of muscular fatigue. In: Biochemistry of Exercise, Boston: International meeting, 1983, pp. 3-28.
9. Fitch, S., and A. McComas. Influence of human muscle length on fatigue. J. Physiol. 362:205-213, 1985.
10. Fitts, R. H. Muscle fatigue: the cellular aspects. Am. J. Sports Med. 24:9-13, 1996.
11. Gandevia, S. C., and D. K. McKenzie. Activation of human muscles at short muscle lengths during maximal static efforts. J. Physiol. 407:599-613, 1988.
12. Gordon, A. M., A. F. Huxley, and F. J. Julian. The variation in isometric tension with sarcomere length in vertebrate muscle fibres. J. Physiol. 184:170-192, 1966.
13. Hisaeda, H. O., M. Shinohara, M. Kouzaki, and T. Fukunaga. Effect of local blood circulation and absolute torque on muscle endurance at two different knee-joint angles in humans. Eur. J. Appl. Physiol. 86:17-23, 2001.
14. Jones, D. A., D. J. Newham, and C. Torgan. Mechanical influences on long-lasting human muscle fatigue and delayed-onset pain. J. Physiol. 412:415-427, 1989.
15. Kellis, E., and V. Baltzopoulos. In vivo determination of the patella tendon and hamstrings moment arms in adult males using videofluoroscopy during submaximal knee extension and flexion. Clin. Biomech. 14:118-124, 1999.
16. Kent-Braun, J. A. Central and peripheral contributions to muscle fatigue in humans during sustained maximal effort. Eur. J. Appl. Physiol. Occup. Physiol. 80:57-63, 1999.
17. Kooistra, R. D., C. J. de Ruiter, and A. de Haan. Muscle activation and blood flow do not explain the muscle-length dependent variation in quadriceps isometric endurance. J. Appl. Physiol. 98:810-816, 2005.
18. Linnamo, V., K. Hakkinen, and P. V. Komi. Neuromuscular fatigue and recovery in maximal compared to explosive strength loading. Eur. J. Appl. Physiol. Occup. Physiol. 77:176-181, 1998.
19. Maffiuletti, N. A., and R. Lepers. Quadriceps femoris torque and EMG activity in seated versus supine position. Med. Sci. Sports Exerc. 35:1511-1516, 2003.
20. Metzger, J. M., and R. L. Moss. Effects of tension and stiffness due to reduced pH in mammalian fast- and slow-twitch skinned skeletal muscle fibres. J. Physiol. 428:737-750, 1990.
21. Ng, A. V., J. C. Agre, P. Hanson, M. S. Harrington, and F. J. Nagle. Influence of muscle length and force on endurance and pressor responses to isometric exercise. J. Appl. Physiol. 76:2561-2569, 1994.
22. Pincivero, D. M., Y. Salfetnikov, R. M. Campy, and A. J. Coelho. Angle- and gender-specific quadriceps femoris muscle recruitment and knee extensor torque. J. Biomech. 37:1689-1697, 2004.
23. Place, N., N. A. Maffiuletti, Y. Ballay, and R. Lepers. Twitch potentiation is greater after a fatiguing submaximal isometric contraction performed at short vs. long quadriceps muscle length. J. Appl. Physiol. 98:429-436, 2005.
24. Rassier, D. E. The effects of length on fatigue and twitch potentiation in human skeletal muscle. Clin. Physiol. 20:474-482, 2000.
25. Rassier, D. E., and W. Herzog. Effects of pH on the length-dependent twitch potentiation in skeletal muscle. J. Appl. Physiol. 92:1293-1299, 2002.
26. Sacco, P., D. B. McIntyre, and D. A. Jones. Effects of length and stimulation frequency on fatigue of the human tibialis anterior muscle. J. Appl. Physiol. 77:1148-1154, 1994.
27. Sacco, P., G. W. Thickbroom, M. L. Thompson, and F. L. Mastaglia. Changes in corticomotor excitation and inhibition during prolonged submaximal muscle contractions. Muscle Nerve 20:1158-1166, 1997.
28. Visser, J. J., J. E. Hoogkamer, M. F. Bobbert, and P. A. Huijing. Length and moment arm of human leg muscles as a function of knee and hip-joint angles. Eur. J. Appl. Physiol. Occup. Physiol. 61:453-460, 1990.
29. Weir, J. P., A. L. McDonough, and V. J. Hill. The effects of joint angle on electromyographic indices of fatigue. Eur. J. Appl. Physiol. Occup. Physiol. 73:387-392, 1996.
30. Yang, Z., J. T. Stull, R. J. Levine, and H. L. Sweeney. Changes in interfilament spacing mimic the effects of myosin regulatory light chain phosphorylation in rabbit psoas fibers. J. Struct. Biol. 122:139-148, 1998.

Cited By:

This article has been cited 1 time(s).

Medicine & Science in Sports & Exercise
Do Practical Durations of Stretching Alter Muscle Strength? A Dose-Response Study
Medicine & Science in Sports & Exercise, 40(8): 1529-1537.
PDF (348) | CrossRef
Back to Top | Article Outline


©2006The American College of Sports Medicine