Journal Logo


Central Factors Contribute to Knee Extensor Strength Loss after 2000-m Rowing in Elite Male and Female Rowers


Author Information
Medicine & Science in Sports & Exercise: March 2017 - Volume 49 - Issue 3 - p 440-449
doi: 10.1249/MSS.0000000000001133
  • Free


Muscle fatigue is commonly characterized as any exercise-associated reduction in maximal voluntary force produced by a muscle or muscle group (21). During exercise, the decline in the force-generating capacity of the neuromuscular system can originate from multiple physiological alterations along the motor pathway (21). Typically, a distinction is drawn between peripheral and central sites that contribute to muscle fatigue. Peripheral fatigue refers to exercise-related impairments that occur at or distal to the neuromuscular junction and can be measured by reductions in twitch force/torque produced by a muscle in response to transcutaneous stimulation (7). The underlying mechanisms encompass potential alterations in neuromuscular transmission and/or excitation–contraction coupling mainly caused by the accumulation of intramuscular metabolites and/or muscular damage (1). Central fatigue refers to the exercise-induced reduction in the voluntary activation of the tested muscle group and can be assessed by using the interpolated twitch technique (21). The impairment in voluntary activation can originate from several sites of the central nervous system (CNS). Thus, central fatigue involves any spinal and supraspinal alterations, which lead to impairments in motoneuron excitation (21).

The development of fatigue is generally considered to be task dependent. Specifically, the relative importance of central and peripheral mechanisms differs depending on the tested muscle group, the contraction mode, and the exercise intensity and duration (17). The task dependency of muscle fatigue has been extensively investigated during single-joint exercise. For instance, there is substantial evidence that long-duration, low-force voluntary contractions are predominantly limited by central factors, whereas near maximal voluntary contractions of short duration are rather limited by peripheral factors of fatigue (for a review, see Taylor and Gandevia [50]). In addition, further research interest focuses on task-specific alterations of central and peripheral fatigue after locomotor endurance exercise (43). Several studies have investigated the effect of locomotor endurance exercise (i.e., cycling and running) on central and peripheral aspects of fatigue after short (6–8 min [3,4,23]), medium (20–90 min [35,46]), and long exercise durations (3–5 h [37]). Together, findings from locomotor exercise studies suggest a greater contribution of peripheral impairments to reductions in voluntary force production after shorter, more intense locomotor exercise compared with longer, less intense exercise. Higher exercise intensities are shown to induce stronger metabolic disturbances within the muscle (31), which are likely responsible for the greater extent of peripheral fatigue (51,52). By contrast, a greater portion of central fatigue was typically demonstrated after a low-intensity locomotor exercise of longer durations. There is evidence suggesting that increased firing rates of group III/IV muscle afferents in response to intramuscular metabolic disturbances facilitate central fatigue (21). However, the distinct mechanisms of fatigue after locomotor endurance exercise are still controversially discussed (36).

Little is known about the origins of muscle fatigue after whole-body exercise such as rowing. In rowing, the leg muscles serve as important contributors to the propulsion of the rowing boat by pushing against the foot stretcher (47). Several studies have emphasized the major role of the quadriceps muscle to produce power during the rowing stroke (24,54). Furthermore, approximately 70% of total muscle mass is involved because of the fact that upper and lower body muscles work synchronized during the rowing stroke (48). There is growing evidence that the motor performance of the lower limb muscles can be affected when it is preceded by fatiguing contractions of the upper limbs (25,30). Therefore, rowing offers a unique possibility to assess the effect of simultaneously working upper body muscles on quadriceps muscle fatigue. To our knowledge, no previous study has investigated the effect of a simulated rowing race on quadriceps muscle fatigue. Therefore, the present study was designed to assess the effect of a 2000-m rowing time trial (~6–7 min) on indices of central and peripheral fatigue. We hypothesized that after termination of the 2000-m rowing time trial, both central and peripheral mechanisms would contribute to knee extensor strength loss.



Eight German heavyweight rowers, four males (mean ± SD; age = 19 ± 4 yr, height = 190 ± 4 cm, weight = 91 ± 6 kg) and four females (age = 20 ± 4 yr, height = 183 ± 3 cm, weight = 78 ± 5 kg), were recruited to participate in this study. At the time of the study, all subjects regularly participated in national and international-level championships. The average competitive rowing experience of the sample was 8.3 ± 3.5 yr. Subjects were asked to avoid caffeine and alcohol consumption for 24 h before the investigations. The study was approved by the university ethics committee and was conducted according to the declaration of Helsinki. All subjects were informed about possible risks and discomfort associated with the investigations before giving their written consent to participate.

Experimental procedure

All subjects participated in three experimental sessions at the laboratory. In the first session, subjects were familiarized with the procedures used to assess the neuromuscular function of the knee extensors. The second and the third sessions were separated by at least 1 wk and took place at the same time of the day. To control for potential effects of time and rowing warm-up on neuromuscular function, subjects randomly performed two experimental conditions: 1) a 2000-m time trial on an indoor rower and 2) a control intervention, in which the subjects rested on the indoor rower for a period similar to the time trial duration (Fig. 1). The period for the control condition was set according to the subjects' prediction of the race time. Before each experimental condition, participants performed their individual and standardized warm-up routine on the indoor rower (warm-up II) within a time window of 20 min. The warm-up period comprised low-intensity intervals of rowing with expanded resting periods in between. Between the warm-up period and each experimental condition, subjects rested for 5 min on the indoor rower. Before and 3 min after each experimental condition, neuromuscular tests were performed. The time lag of 3 min between exercise termination and neuromuscular testing was due to the time required for the preparation of postexercise measurements.

Illustration of the experimental procedure. For further information, see Experimental procedure section. [La]b, blood lactate concentration; [NH3]b, blood ammonia concentration; SS, single electrical stimuli; PS100, paired electrical stimuli at 100 Hz; MVCiso, maximal voluntary isometric contraction; MVCcon, maximal voluntary concentric contraction; M max, maximum compound muscle action potential; TWPS100, paired stimuli twitch torque.

Before baseline measurements, subjects performed a short warm-up on a stationary bicycle (warm-up I; 5 min, 100 W, 80 rpm). Neuromuscular tests comprised supramaximal electrical stimulations of the femoral nerve during isometric as well as concentric maximal voluntary contractions (MVC) and at rest (Fig. 1). Contraction sequences were performed in a randomized order. All measurements were conducted on the quadriceps muscle of the right leg. During neuromuscular testing, subjects were comfortably seated and secured on a CYBEX NORM dynamometer (Computer Sports Medicine®, Inc., Stoughton, MA). The seating position was adjusted for each subject, and settings were documented for the subsequent sessions.

The 2000-m time trial

Time trial and rowing warm-up were performed on an instrumented Concept II rowing ergometer (Model E, Morrisville). The 2000-m time trial protocol was chosen according to the standard race distance of the Olympics and the World Rowing Championships. All subjects were familiar with both the ergometer and the test protocol used in this study. The 2000-m time trial started in the catch position. Drag factor was set at 140 for males and 130 for females. Mechanical data (i.e., power per rowing stroke) were recorded by using a force sensor fitted on the oar handle and a position sensor fitted on the flywheel (FES, Berlin, Germany). After finishing the time trial, subjects rated their perceived exertion using the Borg scale (9).

Torque recordings

A CYBEX NORM dynamometer was used to record instantaneous torques. Subjects were positioned on an adjustable chair with the hip fixed at 80° (0° = full extension). To avoid excessive movements during contractions, straps were applied tightly across the subject's waist. The dynamometer rotation axis was aligned with the knee joint rotation axis, and the lever arm was fixed to the lower leg just above the lateral malleolus. Isometric MVC were performed at 80° knee flexion (0° = full extension). Concentric isokinetic contractions were performed at an angular velocity of 25°·s−1 (5,6). After each contraction, participants were instructed to relax their knee extensor muscles, and the knee joint was passively extended via movement of the lever arm. During active and passive knee extension, the superimposed stimuli were automatically delivered at 80° of knee flexion. The range of motion was set at 85°, from 5° to 90° knee flexion. For each trial, subjects were instructed to cross their arms in front of their chest and to push as fast and as hard as possible against the lever arm of the dynamometer. Strong verbal encouragement was given to the participants by the investigator. Visual feedback of the torque–time curve was provided on a digital oscilloscope (HM1508, HAMEG Instruments, Mainhausen, Germany). Between each contraction, subjects recovered passively for 60 s. On average, three maximal attempts were performed for each contraction mode until the coefficient of variation of the maximal torque values was less than 5%.

EMG recordings

For a detailed description of the EMG recordings, we refer to a previously published study from our laboratory (6). Briefly, EMG signals from the vastus medialis (VM), rectus femoris (RF), and vastus lateralis (VL) were recorded using surface electrodes (EMG Ambu Blue Sensor N, Ballerup, Denmark). Myoelectrical signals were amplified (×2500), band-pass filtered (10–450 Hz), and digitized with a sampling frequency of 3 kHz by an analog-to-digital converter (NI PCI-6229, National Instruments, Austin, TX). EMG data were stored on a hard drive for further analysis using a custom-built LABVIEW based program (Imago, Pfitec, Endingen, Germany).

Electrical nerve stimulation

The femoral nerve was stimulated percutaneously using electrical stimulation to assess the neuromuscular function of the quadriceps muscle. Square wave pulses of 1-ms duration with maximal voltage of 400 V were delivered using a constant-current stimulator (Digitimer DS7A, Herfordshire, UK). A ball probe cathode (10 mm diameter) was pressed manually in the femoral triangle. The anode, a self-adhesive electrode (35 × 45 mm, Spes Medica, Genova, Italy), was affixed over the greater trochanter. After determining the optimal site for stimulation, the position was marked onto the subjects' skin to ensure repeatable measurements within each session. Individual stimulation intensity was progressively increased until the maximum compound muscle action potential (Mmax) of VM and a plateau in knee extensor twitch torque was achieved. During the subsequent testing procedures, the stimulation intensity was increased by additional 40% to guarantee supramaximal stimulation. To ensure consistency of the effective stimulus intensity, VM Mmax amplitude was redetermined after each experimental condition by using M-wave “mini-recruitment curves” (42). Three electrical single stimuli were used to elicit Mmax amplitudes. As previously recommended for quantification of peripheral fatigue (40), potentiated peak twitch torques were evoked after isometric MVC using supramaximal paired stimuli at 100 Hz (10 ms interstimulus interval). To determine the level of voluntary activation during isometric and concentric MVC, the interpolated twitch technique was used. Electrical paired stimuli were applied to the femoral nerve at 80° knee flexion. For the isometric condition, paired stimuli were delivered 2 s after torque onset (during the plateau phase) and 2 s after MVC. During concentric contractions, paired stimuli during MVC and at rest were automatically delivered at a knee angle of 80°. A LABVIEW based program (Imago, Pfitec, Erdingen, Germany) was used to trigger the electrical stimuli.

Physiological data recordings

All blood samples were drawn from the earlobe, with the site standardized for each subject. A portable analyzer (Lactate Scout, SensLab GmbH, Leipzig, Germany) was used for determination of blood lactate concentrations ([La]b). Blood samples were taken before and 1, 5, and 10 min after finishing each experimental condition (Fig. 1). Blood ammonia concentrations ([NH3]b) were measured using an Ammonia Test Kit II for a PocketChem BA device (Arkay, Inc., Kyoto, Japan). Measurements for ammonia blood sampling were conducted before and 10 min after each experimental condition. Heart rate data were continuously recorded during the time trial using a Polar RS 800 heart rate monitor (Kempele, Finland).

Data analysis

All torque signals were corrected for the effect of gravity. Potentiated peak twitch torques (i.e., highest value of the torque–time curve) in response to paired electrical stimuli were determined and averaged for each trial. The three best isometric and concentric MVC trials were selected for further analysis. Isometric maximal voluntary torque (MVT), i.e., the highest torque value before the superimposed twitch, and concentric MVT, i.e., the torque value immediately before the superimposed twitch, were calculated. Mmax amplitudes elicited by single stimuli were measured peak-to-peak and averaged over the three trials. Muscle activation during isometric MVC was estimated by calculating the root-mean-square of the EMG signal (RMS-EMG) over a time interval of 200 ms at MVT, i.e., 200 ms around the MVT for the isometric contraction and 200 ms before the superimposed stimuli for the concentric contraction. RMS-EMG of VM, RF, and VL were normalized to the corresponding Mmax values (RMS·M−1). In addition, RMS·M−1 was averaged across VM, RF, and VL to estimate quadriceps muscle activation during isometric MVC (Q RMS·M−1). The level of voluntary activation for isometric MVC was calculated using a corrected formula: (1 − superimposed twitch [Tb × MVT−1] × control twitch−1) × 100 (49). MVT is the maximal torque level, and Tb is the torque value immediately before the superimposed twitch. The corrected formula is used to avoid the potential problem that the superimposed stimuli are not applied during the maximum torque level. For concentric MVC, voluntary activation was computed using the common formula: (1 − superimposed twitch × control twitch−1) × 100 (2).

Statistical analysis

All data were screened for normal distribution using the Shapiro–Wilk test. A two-way repeated-measures ANOVA, on time (before and after) and condition (time trial and control), was conducted for all neuromuscular parameters. Metabolic parameters were analyzed in separate repeated-measures ANOVA, on time (four measurement time points) and condition. Post hoc analyses were performed using Holm–Sidak tests. The effect size was determined by calculating partial eta squared (η2). Values of 0.10 show small, 0.25 medium, and 0.40 large effects (11). Data were analyzed using the Statistical Package for the Social Sciences version 22.0 (SPSS Inc., Chicago, IL), and statistical significance was accepted at P ≤ 0.050.


Exercise responses

Subjects completed the 2000-m rowing time trial in a mean time of 6:46 ± 0:32 min (range, 6:11 to 7:09 min). Mechanical, cardiovascular, and perceptual recordings during the time trial are provided in Table 1. Significant time–condition interactions were found for [La]b (F1,7 = 53.72, P < 0.001, η2 = 0.89) and [NH3]b (F1,7 = 11.50, P = 0.012, η2 = 0.62). [La]b increased from 1.0 ± 0.4 mmol·L−1 at rest to 8.8 ± 2.3, 10.6 ± 2.5, and 10.3 ± 3.0 mmol·L−1 measured 1, 5, and 10 min after termination of the time trial, respectively (all P < 0.001). Compared with resting values (31.9 ± 22.4 μmol·L−1), [NH3]b increased more than two times postexercise (77.4 ± 49.0 μmol·L−1, P < 0.05). As expected for the control condition, [La]b and [NH3]b were not significantly different from baseline values for all times of measurement.

Exercise measures during the 2000-m rowing time trial.

Maximal voluntary torque

Postexercise measurements were conducted with a mean time lag of 3:15 ± 0:17 min. There were significant time–condition interactions for isometric (F1,7 = 22.52, P = 0.002, η2 = 0.76) and concentric (F1,7 = 12.98, P = 0.002, η2 = 0.65) MVT. After rowing, isometric (−20.4% ± 8.9%, P = 0.002) and concentric (−18.1% ± 7.3%, P = 0.002) knee extensor MVT declined significantly from preexercise values (Figs. 2A and 2B). As expected for the control condition, no significant changes from baseline MVT levels were observed for isometric (P = 0.477) or concentric (P = 0.872) MVC (Table 2).

Neuromuscular function of the knee extensor muscles before and after the 2000-m rowing time trial. Mean values and individual data for MVT during isometric (MVTiso) (A) and concentric contractions (MVTcon) (B), voluntary activation during isometric (VAiso) (C) and concentric MVC (VAcon) (D), normalized RMS-EMG averaged across the superficial quadriceps muscles (Q RMS·M−1) (E), and paired stimuli twitch torque (TwPS100) (F). Significant before and after change: *P < 0.050, **P < 0.010, ***P < 0.001.
Neuromuscular function of the knee extensors before and after each experimental condition.

Voluntary activation and RMS-EMG

The statistical analysis indicated significant time–condition interactions for voluntary activation during isometric (F1,7 = 11.79, P = 0.011, η2 = 0.63) and concentric (F1,7 = 10.03, P = 0.016, η2 = 0.59) MVC. Compared with preexercise values, there was a significant decay in isometric (−18.2% ± 15.4%, P = 0.012) and concentric (−17.6% ± 14.5%, P = 0.013) voluntary activation, respectively (Figs. 2C and 2D). As expected for the control condition, both isometric and concentric voluntary activation was not significantly different from baseline values (P = 0.508 and P = 0.830, respectively; Table 2).

Data analysis also revealed significant time–condition interactions for Q RMS·M−1 (F1,7 = 9.88, P = 0.016, η2 = 0.59), RF RMS (F1,7 = 6.43, P = 0.039, η2 = 0.48), and RF RMS·M−1 (F1,7 = 21.33, P = 0.002, η2 = 0.75) during isometric MVC. After rowing, Q RMS·M−1 recorded during isometric MVC decreased significantly from preexercise values (P = 0.001; Fig. 2E). There were also significant reductions in RF RMS and RMS·M−1 during isometric MVC after rowing (P = 0.049 and P = 0.007, respectively; Fig. 3B). Conversely, there were no time–condition interactions for VM and VL RMS·M−1. For the control condition, RMS·M−1 values for all muscles were not statistically different from baseline values (Table 2).

Normalized RMS-EMG (RMS·M−1) recorded during isometric MVC. Mean values and individual data for VM (A), RF (B), and VL (C) before and after the 2000-m rowing time trial. Significant before and after change: **P < 0.010.

Electrically evoked twitch torque

No significant time–condition interaction was found for paired stimuli twitch torque (F1,7 = 4.91, P = 0.062, η2 = 0.41) (Fig. 2F and Table 2).

Electrically evoked potentials

There was a significant time–condition interaction for RF Mmax (F1,7 = 12.75, P = 0.009, η2 = 0.65). However, there was no such interaction for VM and VL Mmax amplitudes. After rowing, RF Mmax amplitude increased significantly from preexercise values (P = 0.012; Table 2). For the control condition, Mmax values for all muscles were not statistically different from baseline values (Table 2).


The present study investigated the effect of a 2000-m rowing time trial on indices of central and peripheral fatigue in elite male and female rowers. The main findings were as follows: (i) the torque-generating capacity of the quadriceps muscle significantly decreased after rowing as indicated by isometric (−20.4% ± 8.9%) and concentric (−18.1% ± 7.3%) MVT reductions; (ii) the rowing time trial induced significant declines in voluntary activation during isometric (−18.2% ± 15.4%) and concentric (−17.6% ± 14.5%) contractions; (iii) Q RMS·M−1 significantly decreased by 13.1% ± 4.1%; (iv) significant reductions in RF RMS·M−1 were found after rowing, whereas no significant changes were observed for the normalized muscle activity of VM and VL; and (v) no significant changes were found in paired stimuli twitch torque after rowing compared with preexercise values. Furthermore, because indices of central and peripheral fatigue remained unchanged after the control condition, we suggest an absence of any quadriceps muscle fatigue over time and due to the rowing warm-up.

Maximal voluntary torque

The loss in knee extensor strength, indicated by significant reductions in isometric and concentric MVT, demonstrates that quadriceps muscle fatigue occurred after the 2000-m rowing time trial. To the best of our knowledge, to date no study has shown quadriceps muscle fatigue after rowing exercise. The closest rowing can be compared with is cycling, as it involves predominantly concentric contractions of the knee extensor muscles. A similar decay in isometric knee extensor MVT (−18%) has been reported after high-intensity cycling time trials lasting approximately 6 min (52). On the contrary, several studies observed only minor reductions (8%–14%) in isometric knee extensor MVT after cycling time trials lasting 7–8 min (3,4).

Indices of central fatigue

Evidence for central fatigue after the 2000-m rowing is provided by reductions in voluntary activation during isometric (−18.2%) and concentric (−17.6%) contractions as well as decrements in Q RMS·M−1 during isometric MVC (−13.1%). Accordingly, both methods attested that the impaired voluntary drive to the quadriceps muscle contributed to the knee extensor strength loss after rowing.

In contrast to previous findings, which suggest that central fatigue is greater after longer compared with shorter exercise durations (18,52), we demonstrate a contribution of central factors to knee extensor strength loss already after short-duration endurance exercise (06:46 min). For instance, few comparable studies have shown no (3,4) or only minor reductions (5%−7% [22,52]) in voluntary activation after high-intensity cycling lasting 6–8 min. However, it should be noted that magnetic (3,4) and electrical (22,52) single stimuli were used to assess voluntary activation via the interpolated twitch technique. Because single stimuli have been suggested to produce less extra force and more variable twitch responses during MVC (13), these studies may underestimate the extent to which central fatigue is documented after exercise. Nonetheless, our findings lead us to the assumption that task-specific characteristics of a 2000-m rowing time trial might be responsible for the strong contribution of central factors to knee extensor strength loss after short-duration, high-intensity rowing exercise.

By using RMS·M−1, we were able to the estimate neural activation of individual knee extensor muscles. Interestingly, our data revealed significant reductions in RF RMS·M−1 during isometric MVC after rowing, whereas no significant changes in normalized muscle activity were observed for VM and VL. These findings suggest that the superficial muscles of the quadriceps were differentially affected by the rowing time trial. Such an observation is not unusual because previous studies have shown that the knee extensor muscles are not homogeneously activated and fatigued during different types of exercise (14,15). Our data seem to be in line with findings from Gerževič et al. (22), who have shown that the median power frequency (MDF) of the RF EMG signal progressively declined throughout a 6-min all-out test on an indoor rower, whereas the MDF of the vastii muscles did not significantly change. Because declines in MDF are thought to be indicative of muscle fatigue (13), these results might support the idea that RF is more fatigued after high-intensity rowing efforts than the vastii muscles. The increased RF muscle fatigue might be the consequence of its biarticular function and its biphasic activity pattern during the rowing stroke (18). RF has been shown to act as knee extensor during the early drive phase and as hip flexor during the early recovery phase (18), whereas the vastii muscles are only active in the former (24). Although single-joint studies have also suggested that RF is more susceptible to fatigue compared with the vastii (15,34), the results of investigations on multijoint leg extension exercise are rather contradictory. Several studies have shown that RF is less activated and does not increase its activity during fatiguing multijoint leg extension exercise, unlike the vastii muscles (10,15). Similar observations were also made by Guével et al. (24) during on-water rowing. However, the reasons for an attenuated RF activation during multijoint leg extension exercise are unclear. It has been suggested that the addition of hip extension leads to increased inhibitory and/or decreased excitatory input to the RF motoneuron pool (16). Finally, the exact causes for reductions in RF RMS·M−1 during isometric MVC after rowing remain unclear in the present study. Increased fatigability of the biarticular RF muscle during rowing and/or persistently altered neural input to the RF motoneuron pool might have played a role. Nevertheless, it should be emphasized that EMG data should be interpreted with caution, especially because amplitude cancellation of the EMG signal has been shown to increase during fatiguing exercise (32).

However, the attenuated RF muscle activation might not fully explain the greater reductions in voluntary activation after rowing compared with high-intensity, short-duration cycling exercise. One explanation refers to the additionally active muscle mass during rowing exercise. We propose, in this regard, that impairments in voluntary drive develop to a greater extent during rowing due to a spread of central fatigue from simultaneously working upper body muscles to the knee extensor muscles. This assumption is supported by recent findings from Johnson et al. (30), who found that central fatigue occurred more quickly during cycling when it is preceded by arm-cranking exercise. This effect was attributed to the inhibitory feedback of metabosensitive group III/IV afferents. Regarding to our results, it is conceivable that inhibitory muscle afferent feedback from simultaneously working upper body muscles led to further impairments in the voluntary activation of the remote quadriceps muscle. Evidence for metabosensitive muscle afferents as a contributor to a “spillover” of central fatigue was recently provided by Sidhu et al. (44), who reported a significant reduction in the cortical voluntary activation of the unexercised elbow flexors after high-intensity cycling exercise. This “spillover” of supraspinal fatigue was abolished after attenuating the group III/IV afferent feedback from the locomotor muscles using intrathecal fentanyl injections. Therefore, in addition to factors such as contraction mode, exercise duration, and intensity, the amount of active muscle mass involved in the task might affect the development of central fatigue during endurance exercise.

Besides inhibitory effects of metabosensitive afferents and reductions in RF muscle activation, several other mechanisms might be responsible for impairments in voluntary drive after rowing. In the present study, we found more a than twofold rise in [NH3]b after rowing. Despite a lack of direct evidence, exercise-induced cerebral ammonia uptake is thought to be associated with central fatigue (38); hence, it cannot be ruled out that cerebral ammonia uptake has contributed to impairments in voluntary drive after rowing. Furthermore, brain neurotransmitter concentrations (33) and/or cerebral deoxygenation (53) might also affect the development of central fatigue. Nonetheless, by using motor nerve stimulation to assess the voluntary activation of the quadriceps muscle, we were unable to determine whether spinal and/or supraspinal sites contribute to impairments in motoneuron excitation after rowing.

Indices of peripheral fatigue

Despite strong impairments in the torque-generating capacity of the quadriceps muscle after rowing, no significant changes were observed for quadriceps twitch torque in response to paired electrical stimuli. By contrast, previous studies have typically shown that quadriceps twitch torque in response to single stimuli was significantly decreased by 32%–40% after cycling time trials lasting 6–8 min (3,4,52), suggesting that considerable peripheral fatigue occurs after high-intensity, short-duration cycling exercise. Interestingly, although high levels of quadriceps contraction intensity were observed during the rowing stroke (24), the present data seem to reveal a lack of peripheral fatigue after the 2000-m rowing time trial. A physiological explanation for an absence of peripheral fatigue after rowing points toward the observation that the tolerance limit of peripheral fatigue seems to be regulated in a task-dependent manner (41,52). Particularly in regard to rowing exercise, the active muscle mass might play a role as a determining factor for the development of peripheral fatigue. Because Rossman et al. (41) have shown that increasing the amount of active muscle mass lowers the degree to which peripheral fatigue occurs after dynamic exercise, it might be conceivable that the great amount of active muscle mass during rowing limits the development of peripheral quadriceps fatigue. This assumption receives further support by recent findings from Johnson et al. (30), who reported less peripheral quadriceps fatigue after cycling when it is preceded by arm-cranking exercise. Rossman et al. (41) hypothesized, in this regard, that the CNS tolerates peripheral fatigue to a lesser extent when inhibitory feedback of group III/IV afferents originates from a greater amount of muscle mass.

However, it is important to emphasize that the rowing-induced changes in quadriceps twitch torque are characterized by large interindividual variability, which could also serve as an explanation for the lack of significance (Fig. 2F). Variability in quadriceps twitch response after rowing might be related to interindividual differences in the rowing technique. It is conceivable that subjects who have generated a large proportion of power during the rowing stroke by using their knee extensor muscles would have experienced more peripheral fatigue than participants who have predominantly used their upper body muscles as compensatory strategy to ensure adequate power output. Furthermore, because of a limited access to competitive rowers, our sample comprised both sexes. It has been shown that females are less fatigable compared with males when performing isometric fatiguing contractions. However, these sex-based differences are diminished during shortening contractions, suggesting that sex differences in fatigability are task dependent (28). Individual data from the present study reveal that three out of four female subjects have shown no changes in quadriceps twitch torque after rowing (Fig. 2F). This lack of peripheral fatigue would be in line with studies demonstrating a greater muscle perfusion during exercise (39) and a greater proportional area of type I fibers in females (45). However, present data on sex-specific differences in peripheral fatigue after the 2000-m rowing should be interpreted with caution because of the small sample size.

Interestingly, data revealed a significant rise in RF Mmax amplitude after rowing, whereas no significant changes in Mmax were observed for VM and VL. Because M-wave amplitude is commonly used as an index of changes in neuromuscular propagation, our data seem to indicate an exaggerated sarcolemmal excitability of the RF muscle. Similar to the twitch characteristics, the M-wave responses of a muscle depend on the contraction history (27) as well as the fiber type composition of a muscle (26). Because potentiation and fatigue-induced depression of M-wave amplitude have shown to be greater in muscles with a higher percentage of type II fibers (26), the higher proportion of type II fibers in RF (29) and a lower muscle activation during rowing compared with the vastii muscles (24) might be responsible for the increase in RF Mmax amplitude after the rowing time trial. Nevertheless, as stated previously (12), the validity of Mmax amplitude as an index of fatigue-related changes in neuromuscular propagation remains questionable.


In the present study, neuromuscular testing was performed with a mean time lag of 03:15 min after exercise termination. Although this is common in studies investigating muscle fatigue induced by whole-body endurance exercise, Froyd et al. (20) have shown that substantial recovery in twitch torque occurs within 2 min after single-joint exercise. By contrast, a recent study by Blain et al. (8) has demonstrated persistent reductions in quadriceps twitch torque for at least 15 min after a high-intensity cycling time trial. On the basis of these ambiguous findings, one might suggest that the recovery process of twitch amplitude depends on the contraction history of the muscle group. Therefore, it cannot be ruled out that the lack of change in paired stimuli twitch torque might result from the recovery process within the time lag between time trial termination and postexercise measurements.


The 2000-m rowing time trial resulted in significant quadriceps muscle fatigue. Our data further revealed large impairments in voluntary drive to the knee extensors after high-intensity rowing exercise for 6–7 min. In particular, the neural drive to the RF muscle was significantly decreased after rowing whereas no changes were observed for the vastii muscles. Furthermore, the present results showed that there was no significant change in indices of peripheral fatigue after intense rowing in elite rowers. The present results are in contrast to previous findings, which have typically shown that short-duration endurance exercise primarily induces peripheral and rather less central fatigue (52). Therefore, we conclude that muscle fatigue of the knee extensor muscles after high-intensity rowing exercise was explained primarily by central factors that led to large reductions in voluntary drive. The attenuated the neural activation of the biarticular RF muscle and/or the great amount of active muscle mass might be responsible for the pronounced contribution of central factors to quadriceps muscle fatigue after the 2000-m rowing in elite male and female rowers.

The authors thank all athletes of the Olympic Rowing Club Rostock (ORC) who participated in this study. They also show appreciation to Dr. Nicola Maffiuletti for his valuable comments on our manuscript. The authors did not receive any funding to conduct the present study. No conflicts of interest are directly relevant to this article. The present results do not constitute endorsement by the American College of Sports Medicine. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.


1. Allen DG, Lamb GD, Westerblad H. Skeletal muscle fatigue: cellular mechanisms. Physiol Rev. 2008;88(1):287–332.
2. Allen GM, Gandevia SC, McKenzie DK. Reliability of measurements of muscle strength and voluntary activation using twitch interpolation. Muscle Nerve. 1995;18(6):593–600.
3. Amann M, Dempsey JA. Locomotor muscle fatigue modifies central motor drive in healthy humans and imposes a limitation to exercise performance. J Physiol. 2008;586(1):161–73.
4. Amann M, Proctor LT, Sebranek JJ, Pegelow DF, Dempsey JA. Opioid-mediated muscle afferents inhibit central motor drive and limit peripheral muscle fatigue development in humans. J Physiol. 2009;587(Pt 1):271–83.
5. Babault N, Pousson M, Ballay Y, Van Hoecke J. Activation of human quadriceps femoris during isometric, concentric, and eccentric contractions. J Appl Physiol (1985). 2001;91(6):2628–34.
6. Behrens M, Mau-Moeller A, Weippert M, et al. Caffeine-induced increase in voluntary activation and strength of the quadriceps muscle during isometric, concentric and eccentric contractions. Sci Rep. 2015;5:10209.
7. Bigland-Ritchie B, Woods JJ. Changes in muscle contractile properties and neural control during human muscular fatigue. Muscle Nerve. 1984;7(9):691–9.
8. Blain GM, Mangum TS, Sidhu SK, et al. Group III/IV muscle afferents limit the intramuscular metabolic perturbation during whole body exercise in humans. J Physiol. 2016;594(18):5303–15.
9. Borg GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc. 1982;14(5):377–81.
10. Chin LM, Kowalchuk JM, Barstow TJ, et al. The relationship between muscle deoxygenation and activation in different muscles of the quadriceps during cycle ramp exercise. J Appl Physiol (1985). 2011;111(5):1259–65.
11. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Stat Power Anal Behav Sci. 1988; 2nd:567.
12. Crone C, Johnsen LL, Hultborn H, Orsnes GB. Amplitude of the maximum motor response (Mmax) in human muscles typically decreases during the course of an experiment. Exp Brain Res. 1999;124(2):265–70.
13. De Luca CJ. The use of surface electromyography in biomechanics. J Appl Biomech. 1997;13(2):135–63.
14. Duchateau J. Stimulation conditions can improve the validity of the interpolated twitch technique. J Appl Physiol (1985). 2009;107(1):361, discussion 367–8.
15. Ebenbichler GR, Kollmitzer J, Glöckler L, Bochdansky T, Kopf A, Fialka V. The role of the biarticular agonist and cocontracting antagonist pair in isometric muscle fatigue. Muscle Nerve. 1998;21(12):1706–13.
16. Ema R, Sakaguchi M, Akagi R, Kawakami Y. Unique activation of the quadriceps femoris during single- and multi-joint exercises. Eur J Appl Physiol. 2016;116(5):1031–41.
17. Enoka RM, Baudry S, Rudroff T, Farina D, Klass M, Duchateau J. Unraveling the neurophysiology of muscle fatigue. J Electromyogr Kinesiol. 2011;21(2):208–19.
18. Fleming N, Donne B, Mahony N. A comparison of electromyography and stroke kinematics during ergometer and on-water rowing. J Sports Sci. 2014;32:1127–38.
19. Froyd C, Beltrami FG, Millet GY, Noakes TD. Central regulation and neuromuscular fatigue during exercise of different durations. Med Sci Sports Exerc. 2016;48(6):1024–32.
20. Froyd C, Millet GY, Noakes TD. The development of peripheral fatigue and short-term recovery during self-paced high-intensity exercise. J Physiol. 2013;591(5):1339–46.
21. Gandevia SC. Spinal and supraspinal factors in human muscle fatigue. Physiol Rev. 2001;81(4):1725–89.
22. Gerževič M, Strojnik V, Jarm T. Differences in muscle activation between submaximal and maximal 6-minute rowing tests. J Strength Cond Res. 2011;25(9):2470–81.
23. Goodall S, González-Alonso J, Ali L, Ross EZ, Romer LM. Supraspinal fatigue after normoxic and hypoxic exercise in humans. J Physiol. 2012;590(11):2767–82.
24. Guével A, Boyas S, Guihard V, Cornu C, Hug F, Nordez A. Thigh muscle activities in elite rowers during on-water rowing. Int J Sports Med. 2011;32(2):109–16.
25. Halperin I, Copithorne D, Behm DG. Unilateral isometric muscle fatigue decreases force production and activation of contralateral knee extensors but not elbow flexors. Appl Physiol Nutr Metab. 2014;39(12):1338–44.
26. Hamada T, Sale DG, MacDougall JD, Tarnopolsky MA. Interaction of fibre type, potentiation and fatigue in human knee extensor muscles. Acta Physiol Scand. 2003;178(2):165–73.
27. Hicks A, Fenton J, Garner S, McComas AJ. M wave potentiation during and after muscle activity. J Appl Physiol (1985). 1989;66(6):2606–10.
28. Hunter SK. Sex differences in human fatigability: mechanisms and insight to physiological responses. Acta Physiol (Oxf). 2014;210(4):768–89.
29. Johnson MA, Polgar J, Weightman D, Appleton D. Data on the distribution of fibre types in thirty-six human muscles. An autopsy study. J Neurol Sci. 1973;18(1):111–29.
30. Johnson MA, Sharpe GR, Williams NC, Hannah R. Locomotor muscle fatigue is not critically regulated after prior upper body exercise. J Appl Physiol (1985). 2015;119(7):840–50.
31. Jones AM, Wilkerson DP, DiMenna F, Fulford J, Poole DC. Muscle metabolic responses to exercise above and below the “critical power” assessed using 31P-MRS. Am J Physiol Regul Integr Comp Physiol. 2008;294(2):585–93.
32. Keenan KG, Farina D, Maluf KS, Merletti R, Enoka RM. Influence of amplitude cancellation on the simulated surface electromyogram. J Appl Physiol (1985). 2005;98(1):120–31.
33. Klass M, Duchateau J, Rabec S, Meeusen R, Roelands B. Noradrenaline reuptake inhibition impairs cortical output and limits endurance time. Med Sci Sports Exerc. 2016;48(6):1014–23.
34. Kouzaki M, Shinohara M, Fukunaga T. Non-uniform mechanical activity of quadriceps muscle during fatigue by repeated maximal voluntary contraction in humans. Eur J Appl Physiol Occup Physiol. 1999;80(1):9–15.
35. Lepers R, Millet GY, Maffiuletti NA. Effect of cycling cadence on contractile and neural properties of knee extensors. Med Sci Sports Exerc. 2001;33(11):1882–8.
36. Marcora S. Is peripheral locomotor muscle fatigue during endurance exercise a variable carefully regulated by a negative feedback system? J Physiol. 2008;586(Pt 7):2027–8, author reply 2029–30.
37. Millet GY, Lepers R. Alterations of neuromuscular function after prolonged running, cycling and skiing exercises. Sports Med. 2004;34(2):105–16.
38. Nybo L, Dalsgaard MK, Steensberg A, Møller K, Secher NH. Cerebral ammonia uptake and accumulation during prolonged exercise in humans. J Physiol. 2005;563(Pt 1):285–90.
39. Parker BA, Smithmyer SL, Pelberg JA, Mishkin AD, Herr MD, Proctor DN. Sex differences in leg vasodilation during graded knee extensor exercise in young adults. J Appl Physiol (1985). 2007;103(5):1583–91.
40. Place N, Maffiuletti NA, Martin A, Lepers R. Assessment of the reliability of central and peripheral fatigue after sustained maximal voluntary contraction of the quadriceps muscle. Muscle Nerve. 2007;35(4):486–95.
41. Rossman MJ, Garten RS, Venturelli M, Amann M, Richardson RS. The role of active muscle mass in determining the magnitude of peripheral fatigue during dynamic exercise. Am J Physiol Regul Integr Comp Physiol. 2014;306(12):R934–40.
42. Rupp T, Girard O, Perrey S. Redetermination of the optimal stimulation intensity modifies resting H-reflex recovery after a sustained moderate-intensity muscle contraction. Muscle Nerve. 2010;41(5):642–50.
43. Sidhu SK, Cresswell AG, Carroll TJ. Corticospinal responses to sustained locomotor exercises: moving beyond single-joint studies of central fatigue. Sports Med. 2013;43(6):437–49.
44. Sidhu SK, Weavil JC, Venturelli M, et al. Spinal μ-opioid receptor-sensitive lower limb muscle afferents determine corticospinal responsiveness and promote central fatigue in upper limb muscle. J Physiol. 2014;592(Pt 22):5011–24.
45. Simoneau JA, Bouchard C. Human variation in skeletal muscle fiber-type proportion and enzyme activities. Am J Physiol. 1989;257(4 Pt 1):E567–72.
46. Skof B, Strojnik V. Neuromuscular fatigue and recovery dynamics following prolonged continuous run at anaerobic threshold. Br J Sports Med. 2006;40(3):219–22 discussion 219–22.
47. Soper C, Hume PA. Towards an ideal rowing technique for performance: the contributions from biomechanics. Sports Med. 2004;34(12):825–48.
48. Steinacker JM. Physiological aspects of training in rowing. Int J Sports Med. 1993;14:S3–10.
49. Strojnik V, Komi PV. Neuromuscular fatigue after maximal stretch-shortening cycle exercise. J Appl Physiol (1985). 1998;84(1):344–50.
50. Taylor JL, Gandevia SC. A comparison of central aspects of fatigue in submaximal and maximal voluntary contractions. J Appl Physiol (1985). 2008;104(2):542–50.
51. Thomas K, Elmeua M, Howatson G, Goodall S. Intensity-dependent contribution of neuromuscular fatigue after constant-load cycling. Med Sci Sports Exerc. 2016;48(9):1751–60.
52. Thomas K, Goodall S, Stone M, Howatson G, St Clair Gibson A, Ansley L. Central and peripheral fatigue in male cyclists after 4-, 20-, and 40-km time trials. Med Sci Sports Exerc. 2015;47(3):537–46.
53. Verges S, Rupp T, Jubeau M, et al. Cerebral perturbations during exercise in hypoxia. Am J Physiol Regul Integr Comp Physiol. 2012;302(8):R903–16.
54. Wilson J, Robertson DGE, Stothart JP. Analysis of Lower Limb Muscle Function in Ergometer Rowing. Int J Sport Biomech. 1988;4:315–25.


© 2017 American College of Sports Medicine