Journal Logo


Mechanisms of Neuromuscular Fatigability in People with Cancer-Related Fatigue


Author Information
Medicine & Science in Sports & Exercise: August 2022 - Volume 54 - Issue 8 - p 1355-1363
doi: 10.1249/MSS.0000000000002919
  • Free


Cancer-related fatigue (CRF), defined as a distressing, persistent sense of physical, emotional, and/or cognitive tiredness or exhaustion (1), is the most common and debilitating symptom for people living with cancer (2,3). CRF affects almost every person during cancer treatment, and around one-third report persistent CRF for months or years after cancer treatment (4,5). This posttreatment CRF is of interest because it negatively affects health-related quality of life and the ability to return to work (6). Given its high prevalence as well as the burden of CRF and the potential economic impact, understanding the etiology of CRF is a pertinent issue. Although the precise underpinnings of CRF are unclear, it is thought that CRF is a multifactorial symptom that is associated with several biological and psychosocial factors (7–9). Impaired exercise tolerance is another potentially important contributor to CRF and likely contributes to reported difficulties in performing physical activities of daily living among people with CRF (10,11). As such, the physiological (e.g., cardiopulmonary, metabolic, and neuromuscular) alterations contributing to impaired exercise tolerance in people with CRF warrant further investigation, particularly given that these impairments could be reversible through exercise training.

One physiological alteration that could hinder the ability to perform physical activities of daily living is neuromuscular fatigability, defined as the reduction in neuromuscular function measured after exercise of a discrete time period (12). In a recent study (8), we assessed neuromuscular fatigability in response to a standardized cycling test (13), with incremental stages interspersed with the assessment of neuromuscular function of the knee extensors, including isometric maximal voluntary contractions (MVC), voluntary activation (VA), and potentiated twitch force (Qtw,pot). Using a clinical cut point to identify people with moderate to severe CRF (14), we found that the reduction in isometric muscle force-generating capacity at the final common stage (the final stage completed by all participants) was higher in people with CRF compared with a nonfatigued group (the latter group being people after cancer treatment who did not have clinically meaningful CRF based on the clinical cut point). Moreover, we reported that alongside reduced peak oxygen uptake (V̇O2peak), CRF severity was associated with neuromuscular fatigability measured at the final common stage of cycling exercise (8). Although it remains unclear whether neuromuscular fatigability is causally related to CRF, greater fatigability could impede the ability to perform everyday physical tasks, cause an avoidance of physical activity and thus further deconditioning, and/or cause fatigue to accumulate throughout the day. Accordingly, gaining an understanding of the mechanisms underpinning the greater fatigability in people with CRF could help provide targets for future interventions to reduce this symptom.

In our previous study (8), the primary aim was to identify the physiological correlates of CRF from a comprehensive group of physiological measures, including neuromuscular fatigability. Although this analysis identified neuromuscular fatigability as being independently associated with CRF severity, such an analysis could not provide insight into the mechanisms of fatigability or how fatigability manifests during incremental cycling. Additional analyses of our data set can help to provide such insight. First, analysis of the kinetics of fatigability can reveal the temporal manifestation of impaired neuromuscular function during exercise. This is of interest as it can shed light on the potential contribution of fatigability toward impairments in the ability to perform physical activities of daily living, and it can also facilitate understanding of the mechanisms contributing to impaired neuromuscular function. Second, the contribution of reductions in the capacity of the nervous system to activate muscle and/or impairments within the contractile machinery, as well as their kinetics of change during exercise, can be assessed using neurostimulation methods to determine the site(s) of impaired neuromuscular function (15,16). Third, we set power outputs relative to body mass during our incremental cycling test. This approach was used because body mass influences the power requirements of daily activities such as walking and climbing stairs and, thus, provides an ecologically valid approach to assess neuromuscular fatigability. Understanding the relative intensity of these power outputs with respect to the gas exchange threshold (GET) and respiratory compensation point (RCP) is of interest, particularly because metabolic and neuromuscular disturbances are exacerbated during exercise above these thresholds (17). Such an analysis can provide insight into whether the greater fatigability in people with CRF is a result of surpassing metabolic thresholds earlier during the neuromuscular fatigability test or due to other mechanisms. Finally, the assessment of EMG during cycling can be used to determine exercise-induced alterations in neuromuscular activity and how these might relate to the greater fatigability in people with CRF. Together, these analyses can provide a more comprehensive understanding of the etiology of fatigability in individuals with CRF. Accordingly, the overarching aim of the present study was to examine the etiology of greater neuromuscular fatigability in people with posttreatment CRF. We hypothesized that (i) in comparison to a nonfatigued group, impairments in neuromuscular variables would be higher during a neuromuscular fatigability test in people with CRF, and (ii) given that our previous study (8) showed lower V̇O2peak in the fatigued than nonfatigued group with GET and RCP occurring at similar relative intensities, the greater impairment in neuromuscular variables would be related to people with CRF exceeding metabolic thresholds.



The data presented herein are a secondary analysis, using data collected as part of a larger study that investigated the physiological and psychosocial correlates of CRF (8). The study received ethical approval from the Conjoint Health Research Ethics Board and the Health Research Ethics Board of Alberta Cancer Committee (REB14-0398 and HREBA.CC-16-10-10, respectively) and was conducted in accordance with all aspects of the Declaration of Helsinki, apart from registration in a database. All participants provided written informed consent to take part in the study. Participants were eligible if they were adults who had received a cancer diagnosis of any type and had completed any type of active treatment (surgery, chemotherapy, or radiation therapy; people on long-term hormonal therapy were eligible to participate). Participant recruitment is described in detail in Brownstein et al. (8). A health screening was conducted to assess for contraindications to exercise, including arrhythmias, uncontrolled hypertension, and physical activity readiness (8). The data from 96 of the 97 participants recruited for our previous study were included in the present analysis, with one participant excluded due to missing cardiopulmonary exercise test (CPET) and neuromuscular fatigability data.

Experimental Design

Participants visited the laboratory on two occasions separated by ~2 wk. During the first visit, participants performed a CPET, followed by a familiarization with the neuromuscular assessment procedures. During the second visit, participants performed incremental cycling exercise interspersed with measurements of neuromuscular function.


Fatigue was measured using the Functional Assessment of Chronic Illness Therapy - Fatigue (FACIT-F) scale (18), which is widely recommended for the assessment of CRF (19). Participants were classified as fatigued if they scored ≤34, based on recommendations for the diagnosis of CRF derived from diagnostic interviews (14). All participants with scores ≥35 were allocated to the nonfatigued group.

Cardiopulmonary exercise test

After the measurement of stature (cm) and mass (kg), a CPET was conducted using a custom-built recumbent ergometer, using an electromagnetically braked Velotron system (RacerMate Inc., Seattle, WA). Breath-by-breath pulmonary gas exchange and ventilation were measured throughout the test (Quark CPET, COSMED, Rome, Italy). Before each visit, the Cosmed was calibrated, following manufacturer guidelines, to gases of known concentrations (oxygen, 15.15%; carbon dioxide, 5.03%). Ventilatory volumes were calibrated using a 3-L syringe. The starting power output (25–50 W) and increment (8–20 W) were estimated and adjusted on an individual basis for a desired test duration of 8–12 min. Participants were permitted to select their own cadence (≥60 rpm) and were instructed to maintain this cadence during cycling assessments. The power output was increased at 1-min intervals until task failure, defined as a reduction in cadence of ≥5 rpm for ≥5 s. Verbal encouragement was provided by the same experimenters throughout the assessment.

Neuromuscular fatigability test

The neuromuscular fatigability test was also performed on the recumbent cycle ergometer, which permits the immediate assessment of neuromuscular function after cycling (13). Each stage of the cycling test lasted 3 min, beginning with a power output of 0.3 W·kg−1, with an increment of 0.3 W·kg−1 for the next four stages and 0.4 W·kg−1 for the following five stages. Isometric force and force during pedal rotations were measured using a wireless PowerForce system (Model PF1.0.0; Radlabor GmbH, Freiburg, Germany). Preexercise, between each stage, and after task failure, a neuromuscular assessment was performed (described below). During cycling, participants received real-time feedback of cadence and were given verbal instruction to maintain their self-selected cadence when it drifted by ≥4 rpm.

Neuromuscular function

For the neuromuscular assessments during the neuromuscular fatigability test, the seat position was adjusted to ensure that knee and hip were at 90° flexion when the pedal was locked. While on the cycle ergometer, participants were secured at the hip and chest with noncompliant straps. The preexercise neuromuscular assessment began with two isometric MVC without stimulation to ensure potentiation of subsequent evoked twitches. Subsequently, two ~3-s MVC were performed, separated by 1 min. Supramaximal electrical stimulation of the femoral nerve was delivered at the plateau in force, with the same stimulation delivered 2 s after the MVC while at rest to measure VA and Qtw,pot. Between cycling stages and at task failure, the pedal was locked instantly, and participants immediately performed one MVC with stimulations delivered during and after the MVC. Isometric force was measured during voluntary and evoked contractions using a wireless PowerForce pedal force analysis system (Model PF1.0.0, Radlabor GmbH) (20) situated between the pedal and the crank. A detailed description of the settings associated with the innovative cycle ergometer can be found in Doyle-Baker et al. (13).


Electromyographic activity was recorded during neuromuscular assessments and throughout cycling. Self-adhesive surface electrodes (10-mm recording diameter; Meditrace 100, Covidien, Mansfield, MA) were placed on the vastus lateralis (VL) and rectus femoris (RF) of the right knee extensors using a bipolar configuration with a 30-mm interelectrode distance. As a high number of participants (13) either had missing data or displayed no discernible compound muscle action potential (M-wave) in the RF, EMG data from this muscle are not reported. A reference electrode was placed on the patella. Before applying the electrodes, the skin was shaved, gently abraded, and cleaned using isopropyl alcohol. The EMG signals were analog-to-digitally converted at a sampling rate of 2000 Hz using a PowerLab system (16/35; ADInstruments, Bella Vista, Australia) and octal bioamplifier (ML138, ADInstruments, gain = 500) with bandpass filter (5–500 Hz) and were analyzed offline using Labchart 8 software (ADInstruments).

Motor nerve stimulation

Single electrical stimuli (1-ms duration) were delivered to the right femoral nerve using a constant current stimulator (DS7A; Digitimer, Welwyn Garden City, Hertfordshire, UK). The cathode electrode (10-mm stimulating diameter; Meditrace 100, Covidien) was secured with tape and a gauze plug to apply pressure on the inguinal triangle, and a 50 × 90-mm rectangular anode electrode (Durastick Plus, DJO Global, Vista, CA) was placed on the gluteal fold. The optimal stimulus intensity was determined as the minimum current that elicited a maximum resting twitch response (Qtw) and maximal compound muscle action potential (Mmax) in both the VL and the RF, with the intensity subsequently multiplied by 1.3 to ensure the stimulus was supramaximal for all neuromuscular assessments.

Data Analysis

Relative exercise intensity during the neuromuscular fatigability test

To determine the relative intensity of the three common stages of the fatigability test for the present analysis, the power outputs associated with the GET and RCP were first determined. To do so, the power outputs were adjusted to take into account the mean rise time of V̇O2 during step-incremental exercise, which approximates two-thirds of the rate of step increment (21). Subsequently, the power outputs during the three common stages of the fatigability test were expressed as percentages of the power outputs associated with the GET and RCP.

Neuromuscular responses during the neuromuscular fatigability test

In our previous study (8), the relative change in MVC peak force, Qtw,pot, and VA between baseline and the final common stage of the test (stage 3), as well as at task failure, was assessed. For the present analysis, the kinetics of the changes in MVC peak force, Qtw,pot, VA, and the amplitude of the negative phase of Mmax (22) were assessed across all three common stages of exercise. Alterations in these variables were expressed in percentages relative to the baseline value. VA was determined using the twitch interpolation method (15), quantified by comparing the amplitude of the superimposed twitch (SIT) to the Qtw,pot using the following equation: VA (%) = [1 − (SIT/Qtw,pot) × 100]. In instances when superimposed stimuli were not delivered at peak force, a correction was applied using the force at stimulation (Fatstim), the peak force, the amplitude of the SIT, and the Qtw,pot, with the following equation applied: VA (%) = [1 − (SIT × (Fatstim/MVC)/(Qtw,pot) × 100)] (23). For the EMG measurements during cycling, the EMG onset and offset of the rectified signal during pedal rotations were visually determined. The root-mean-square EMG (EMGRMS) was recorded between EMG onset and offset, and the average EMGRMS was taken from 10 to 40 s of stage 1 and 1 min 50 s to 2 min 50 s of each stage. The first 10 s of stage 1 and the final 10 s of each stage were not analyzed so as not to include acceleration and deceleration phases, respectively, in the EMG analysis. The EMGRMS was normalized to the maximum EMGRMS obtained over a 0.5-s epoch during the plateau in the baseline MVC. The amplitude of the negative phase of Mmax amplitude was calculated from EMG responses to single femoral nerve stimulation in the relaxed muscle.

Statistical Analysis

Jamovi statistical software (jamovi, version 1.0, 2019, the jamovi project; retrieved from was used for all statistical analyses. All data are presented as mean ± SD. Statistical significance was set at an α of 0.05. The normality of the data was assessed by the Shapiro–Wilk test, with no data requiring transformation. To test our first hypothesis, group differences in neuromuscular changes from baseline after the three common stages of the neuromuscular fatigability test were assessed using a two-way mixed-design ANOVA (group–time). Similarly, to test our second hypothesis, a two-way mixed-design ANOVA was used to assess group differences in relative exercise intensity (with respect to GET and RCP) during the three common stages of the neuromuscular fatigability test. Assumptions of sphericity were explored using Mauchly’s test and controlled for using the Greenhouse–Geisser adjustment in instances where the α for Mauchly’s test was <0.05. In the event of a significant interaction or main effect, post hoc comparisons were performed with Bonferroni correction. Independent-sample t-tests, or a Mann–Whitney U-test if Levene’s test revealed unequal variance, were used to assess between-group differences in power outputs associated with GET and RCP (expressed in absolute units and relative to body mass) and baseline neuromuscular variables. Partial eta squared (ηp2; ANOVA) was calculated to estimate effect sizes, with values representing small (ηp2 < 0.13), medium (ηp2 ≥ 0.13, <0.26), and large (≥0.26) (24). Cohen’s d effect size (t-test) was calculated, with values interpreted as small (d ≥ 0.2, <0.6), moderate (d ≥ 0.6, <1.2), and large (≥1.2) (24).


Participant Characteristics

Participant characteristics are displayed in Table 1. Of the 96 included participants, 54 were in the fatigued group and 42 were in the nonfatigued group based on scores derived from the FACIT-F scale. There were no statistical differences in age (P = 0.097, d = 0.39) or body mass (P = 0.088, d = 0.36) between groups (Table 1).

TABLE 1 - Participant characteristics.
Variable Fatigued (n = 54) Nonfatigued (n = 42)
Age (yr)
 Mean (SD) 54 (9) 58 (12)
Stature (cm)
 Mean (SD) 170 (11) 169 (11)
Mass (kg)
 Mean (SD) 82 (21) 75 (18)
Sex, n (%)
 Male 21 (39) 19 (45)
 Female 33 (61) 23 (55)
Time since treatment (months)
 Mean (SD) 34 (34) 44 (28)*
Cancer type, n (%)
 Breast 24 (44) 19 (45)
 Prostate 5 (9) 12 (29)
 Head and neck 7 (13) 2 (5)
 Colon 5 (9) 3 (7)
 Hematological 1 (2) 0 (0)
 Other 14 (26) 7 (17)
 Multiple cancer types 2 (4) 1 (2)
Treatment received, n (%)
 Surgery 42 (78) 32 (76)
 Radiotherapy 23 (43) 13 (31)
 Chemotherapy 25 (46) 15 (36)
 Single modality 31 (59) 29 (69)
 Multimodality 22 (41) 13 (31)
Fatigue (FACIT-F score)
 Mean (SD) 26 (6) 44 (5)*
 Median 27 45
 Range 10–34 35–51
Single modality refers to chemotherapy or radiotherapy only; multimodality refers to chemotherapy and radiotherapy.
*Between-group difference (P < 0.05).

Neuromuscular Fatigability Test

One participant from the nonfatigued group did not complete neuromuscular fatigability testing. Because of the discomfort associated with motor nerve stimuli, seven participants (four fatigued and three nonfatigued) did not receive stimuli between stages and instead only performed MVC, whereas the stimulation data from two participants (both from the fatigued group) were not included because of a lack of discernible M-waves. Thus, for the analysis of Qtw,pot and VA at stages 1–3, 48 and 38 participants from the fatigued and nonfatigued groups were included. For baseline Mmax data, seven participants (five and two from fatigued and nonfatigued groups, respectively) demonstrated either no discernible M-wave in the VL or had missing data. For EMG during cycling, 13 participants (seven and six from fatigued and nonfatigued groups, respectively) had missing data.

Baseline measures

At baseline, no differences were found between the fatigued and nonfatigued groups for knee extensor MVC peak force (239 ± 160 vs 205 ± 67 N, respectively, U = 1050, P = 0.822, d = 0.28), Qtw,pot (90 ± 69 vs 78 ± 27 N, respectively, U = 957, P = 0.426, d = 0.19), or VA (94% ± 6% vs 95% ± 5%, respectively, U = 982, P = 0.544, d = 0.17).

Neuromuscular fatigability

Raw values for neuromuscular variables are presented in Table 2. For the percentage reduction in MVC peak force during the neuromuscular fatigability test (Fig. 1A), there was a main effect of time (F1.5,128.7 = 34.6, P < 0.001, ηp2 = 0.29) and group (F1,93 = 17.9, P < 0.001, ηp2 = 0.18) and no interaction effect (F2,186 = 0.05, P = 0.953, ηp2 < 0.01). In other words, the reduction in MVC peak force was more pronounced in the fatigued group throughout the fatigability test, with MVC reduced by 10%, 12%, and 17% in the fatigued group and 2%, 4%, and 10% in the nonfatigued group at stages 1, 2, and 3, respectively. For the decrease in Qtw,pot from baseline during the neuromuscular fatigability test (Fig. 1B), there was a main effect of time (F1.7,138.4 = 42.35, P < 0.001, ηp2 = 0.34) and group (F1,85 = 4.92, P = 0.029, ηp2 = 0.06) and no interaction effect (F2,170 = 0.93, P = 0.397, ηp2 = 0.01). In line with maximal force, the reduction in Qtw,pot was more pronounced in the fatigued group, being reduced by 12%, 19%, and 28% in the fatigued group and 8%, 12%, and 20% in the nonfatigued group at stages 1, 2, and 3, respectively. For the decrease in VA from baseline during the neuromuscular fatigability test (Fig. 1C), there was a main effect of time (F1.8,138.9 = 3.72, P = 0.045, ηp2 = 0.05) with no effect of group (F1,85 = 1.22, P = 0.273, ηp2 = 0.02) and no group–time interaction (F2,170 = 2.27, P = 0.114, ηp2 = 0.03). For the amplitude of the negative phase of Mmax, there was a main effect of time (F1.8,139.0 = 6.42, P = 0.002, ηp2 = 0.08) and no group (F2,164 = 1.02, P = 0.316, ηp2 = 0.012) or interaction (F2,164 = 2.47, P = 0.088, ηp2 = 0.03) effects.

TABLE 2 - Raw neuromuscular values, including isometric MVC (fatigued n = 54, nonfatigued n = 41) peak force, Qtw,pot (fatigued n = 48, nonfatigued n = 38), VA (fatigued n = 48, nonfatigued n = 38), and the amplitude of the negative phase of the maximum compound muscle action potential (Mmax, fatigued n = 42, nonfatigued n = 38) during step-incremental cycling exercise in fatigued and nonfatigued groups.
Preexercise Stage 1 Stage 2 Stage 3
MVC (N) Fatigued 239 ± 159 218 ± 144 214 ± 145 204 ± 145
Nonfatigued 205 ± 67 200 ± 65 192 ± 66 185 ± 67
Q tw,pot (N) Fatigued 88 ± 68 78 ± 60 73 ± 58 66 ± 58
Nonfatigued 78 ± 27 71 ± 24 66 ± 22 60 ± 20
VA (%) Fatigued 94 ± 6 92 ± 8 90 ± 14 90 ± 10
Nonfatigued 95 ± 5 95 ± 4 91 ± 11 93 ± 6
Mmax (mV) Fatigued 3.3 ± 2.1 3.2 ± 2.1 3.3 ± 2.1 3.4 ± 2.1
Nonfatigued 4.4 ± 1.6 4.3 ± 1.5 4.4 ± 1.5 4.4 ± 1.5

Changes in isometric MVC peak force (panel A; fatigued n = 54, nonfatigued n = 41), Q tw,pot (panel B; fatigued n = 48, nonfatigued n = 38), and VA (panel C; fatigued n = 48, nonfatigued n = 38) during an incremental performance fatigability test in participants with and without cancer-related fatigue. *P = 0.029, **P < 0.001 significant main effect of group. Data are expressed as mean ± SD.

EMG during cycling

For EMGRMS expressed as a percentage of maximum EMGRMS (Fig. 2), there was a main effect of time (F1.5,109.8 = 168.12, P < 0.001, ηp2 = 0.69) and group (F1,78 = 5.47, P = 0.022, ηp2 = 0.7), with no group–time interaction (F1.5,109.8 = 2.48, P = 0.062, ηp2 = 0.03). Thus, EMGRMS increased during cycling and was, overall, higher in the fatigued group, but the rate of increase did not differ between groups.

EMGRMS measured in the VL normalized to baseline maximum EMGRMS during an incremental performance fatigability test in participants with and without cancer-related fatigue (fatigued n = 40, nonfatigued n = 32). Measurements were taken from 10 to 40 s of stage 1, and 1 min 50 s to 2 min 50 s of stages 1, 2, and 3. *P < 0.05 significant main effect of group. Data are expressed as mean ± SD.

Power outputs at GET and RCP

For peak power output and power outputs associated with the GET and RCP, no differences were found when expressed in absolute units (W; Table 3). However, when expressed relative to body mass (W·kg−1), the power output associated with the GET was lower in the fatigued versus nonfatigued group (U = 701, P = 0.046, d = 0.44), with no difference between the peak power output and the power output associated with the RCP.

TABLE 3 - Power output variables derived from step-incremental cycling exercise and performance fatigability test in fatigued and nonfatigued groups.
Variable P Value d Effect Size
CPET PO Fatigued (n = 53) Nonfatigued (n = 38)
 Absolute PO (W)
  PPO 158 ± 54 161 ± 53 0.757 0.06
  PO at RCP 112 ± 44 115 ± 40 0.524 0.07
  PO at GET 65 ± 26 70 ± 27 0.226 0.23
 Relative PO (W·kg−1)
  PPO 2.0 ± 0.6 2.2 ± 0.6 0.090 0.33
  PO at RCP 1.4 ± 0.4 1.5 ± 0.5 0.076 0.22
  PO at GET 0.8 ± 0.3 0.9 ± 0.4 0.046 0.28
Fatigability PO (% GET) Fatigued (n = 52) Nonfatigued (n = 35)
 Stage 1 47 ± 21 38 ± 14 0.100 0.50
 Stage 2 94 ± 15 75 ± 30 0.642 0.80
 Stage 3 141 ± 64 113 ± 45 0.039 0.51
Fatigability PO (% RCP) Fatigued (n = 52) Nonfatigued (n = 35)
 Stage 1 25 ± 8 21 ± 6 1.00 0.56
 Stage 2 50 ± 16 43 ± 13 0.641 0.48
 Stage 3 75 ± 24 64 ± 19 0.046 0.51
PO, power output; PPO, peak power output.

Relative exercise intensity during the neuromuscular fatigability test

During the three common stages of the neuromuscular fatigability test, at which time the power outputs were 0.3, 0.6, and 0.9 W·kg−1, the absolute power outputs did not differ between the fatigued and the nonfatigued groups (no main effect of group: F1,90 = 2.67, P = 0.106, ηp2 = 0.03). For exercise intensity relative to the GET, there were main effects of time (F1,85 = 416.8, P < 0.001, ηp2 = 0.83) and group (F1,85 = 4.9, P = 0.029, ηp2 = 0.06; Fig. 3) and a group–time interaction (F1,85 = 4.9, P = 0.029, ηp2 = 0.06). In post hoc analysis, the relative intensity of stage 3 (the last common stage) with respect to the GET was significantly higher in the fatigued compared with nonfatigued group (P = 0.039, d = 0.49, Fig. 3). Similarly, for exercise intensity relative to RCP, there were main effects of time (F1.0,85.2 = 817.0, P < 0.001, ηp2 = 0.91) and group (F1,85 = 4.8, P = 0.031, ηp2 = 0.05) and a group–time interaction (F1.0,85.2 = 4.6, P = 0.034, ηp2 = 0.05). Post hoc analysis showed that the relative intensity of stage 3 with respect to the RCP was significantly higher in the fatigued compared with the nonfatigued group (P = 0.046, d = 0.49). When expressed relative to the RCP, the power output at the final common stage was 75% ± 24% and 64% ± 18% in the fatigued and nonfatigued groups, respectively.

Power outputs during the three common stages of the performance fatigability test expressed relative to the power output at GET. Black horizontal lines represent the means, whereas blue and red circles represent individual data points for the fatigued and nonfatigued groups, respectively (fatigued n = 52, nonfatigued n = 35). *P = 0.039, significant between-group difference. Data are expressed as mean ± SD.


A number of key and novel findings from the present study help to shed light on the etiology of fatigability in people with posttreatment CRF. First, and in line with our first hypothesis, people with posttreatment CRF had a more pronounced reduction in maximal and evoked force in the knee extensors during an incremental cycling test compared with a nonfatigued group. Second, the concurrently higher reduction in Qtw,pot in the fatigued group, together with the lack of differences between groups in maximal VA, suggests that the greater fatigability in people with CRF is primarily due to greater disturbances at the muscle level. Third, although the fatigued group was exercising at a higher relative intensity with respect to the GET during stage 3 of the fatigability test, this was not the case during the first two stages, nor were there any differences in absolute power outputs. Thus, differences in the relative intensity of exercise at a given power output relative to body mass are unlikely to explain the greater fatigability in people with CRF during the early stages of exercise. Finally, the higher EMGRMS in the fatigued group throughout neuromuscular fatigability assessment indicates that a higher level of muscle activation during cycling was required to maintain power output compared with the nonfatigued group. The results from this study provide important insight into the etiology of fatigability in people with posttreatment CRF.

Impairments in contractile function are responsible for greater neuromuscular fatigability in people with posttreatment CRF

During the incremental neuromuscular fatigability assessment, the fatigued group demonstrated greater impairments in neuromuscular function compared with the nonfatigued group. Following stage 3, the reduction in MVC peak force relative to baseline was 17% ± 9% and 10% ± 10% in the fatigued and nonfatigued groups, respectively. To allow comparison to other populations, using the same ergometer and protocol, we have previously observed that the magnitude of reduced peak force in the fatigued group after stage 3 is higher than that observed in healthy young participants (~−5%) (13), similar to that observed in highly fatigued people with multiple sclerosis (~−18%) (24), but lower than that observed in people with head and neck cancer who had recently completed radiation and chemotherapy (−29%) (25). The overall reduction in Qtw,pot was greater in the fatigued group (20% ± 14% and 13% ± 11% across stages 1–3 in the fatigued and nonfatigued groups, respectively), with no between-group difference in the reduction in VA. By using dynamic, whole-body exercise, the present study improves on previous designs by using a more ecologically valid exercise mode compared with previous literature on neuromuscular fatigability in people with CRF, which have used isometric exercise protocols (26,27). Thus, during exercise with greater relevance to activities of daily living, the results from the present study indicate that greater fatigability in people with CRF can primarily be attributed to perturbations occurring within the contractile machinery rather than deficits in muscle activation.

Greater impairments in neuromuscular function are evident early during whole-body exercise in people with CRF

For the present study, the kinetics of altered neuromuscular function were assessed across the three common stages of the fatigability test. After just 3 min of exercise, during which the power output was 0.3 W·kg−1 (i.e., ~20–25 W), the fatigued group demonstrated a reduction in MVC peak force, which was fivefold greater than the nonfatigued group (10% ± 10% and 2% ± 11% reduction in MVC peak force, respectively). Such a rapid decline in neuromuscular function in response to exercise of low power output, as expressed either in absolute terms, relative to body mass or relative to the GET (see below), has potentially important implications for the physiological and perceptual effect of typical daily physical activities. For example, the low intensity during the initial stages of the fatigability test is likely to correspond with low-intensity activities of daily living, such as walking, housework, gardening, or slowly climbing stairs. In turn, higher impairments in contractile function during such activities might necessitate a greater compensatory increase in muscle activity (28), as indicated by the elevated EMGRMS in the fatigued group (see below), and thus an increased sense of effort (29) and perception of fatigue (30).

During cycling exercise, impairments in contractile function are determined by perturbations in metabolic homeostasis (17). Specifically, increases in the concentrations of metabolites that inhibit the excitation–contraction coupling and/or cross-bridge force, such as inorganic phosphate (Pi) and hydrogen (H+) (31,32), induce impairments in the capacity of muscle to produce force in response to neural input. In turn, these metabolic perturbations are exacerbated when exercising above the GET (i.e., in the heavy domain) and are further exacerbated when exercising above critical power or the RCP (i.e., in the severe domain) (17). To gain insight into the relative intensities of exercise during the fatigability test, the power outputs were expressed relative to that associated with the GET and RCP. This analysis revealed that the fatigued group was exercising at a higher intensity relative to the GET in comparison with the nonfatigued group at stage 3 (141% ± 64% and 113% ± 45%, respectively). Although the intensity relative to the RCP was also higher in the fatigued group at stage 3, the vast majority of participants were exercising well below the RCP (75% ± 24% and 64% ± 19% of RCP, respectively). Accordingly, the greater neuromuscular fatigability in the fatigued group at the final common stage might have been due, at least in part, to the higher relative exercise intensity being performed and the greater proportion of participants exercising within the heavy domain compared with the nonfatigued group.

Although the higher exercise intensity relative to the GET presents a conceivable explanation for the exacerbated fatigability in the fatigued group at the final common stage, there were no statistical differences in the relative exercise intensity during the first or second stage of the fatigability test, despite fatigability being exacerbated across all stages (main effect of group). In fact, both the fatigued and the nonfatigued groups were exercising firmly within the moderate-intensity domain (i.e., below the GET) during the first stage of the task (47% ± 21% and 38% ± 14% of GET in the fatigued and nonfatigued groups, respectively). These results could point toward slower V̇O2 on-kinetics as a plausible explanation for the exacerbated deficits in neuromuscular function during the initial stages of the fatigability assessment. Specifically, slower V̇O2 on-kinetics are associated with a greater reliance on substrate-level phosphorylation and metabolites, which impair contractile function (33). Indeed, our group has previously demonstrated that the speed of V̇O2 on-kinetics is negatively associated with reductions in twitch force (34). Although V̇O2 on-kinetics were not measured in the present study, it is known that the time constant of the V̇O2 on-response is related to physical activity levels (35) and V̇O2peak (36), both of which were found to be lower in the fatigued group in our previous study on the same group of participants (8). Accordingly, slower V̇O2 on-kinetics represent a plausible mechanism contributing to the higher impairment in neuromuscular function in the fatigued group during the initial stages of the fatigability assessment in the present study. Given that transitions between different steady-state energetic levels are commonplace during everyday activities, a greater consideration for the role of V̇O2 on-kinetics in fatigability is warranted in individuals with CRF to provide insight into factors which could limit functional capacity.

EMG during whole-body exercise is greater in people with posttreatment CRF

Concurrent with the higher impairments in MVC peak force and Qtw,pot, the fatigued group also demonstrated higher EMGRMS/maximum EMGRMS of the VL throughout the neuromuscular fatigability test, likely due to higher muscle activation. A higher muscle activation might be a consequence of the greater impairments in contractile function in the fatigued group. Indeed, the greater EMGRMS/maximum EMGRMS concurrent with the higher reduction in Qtw,pot is indicative of a compensatory increase in motoneuron output owing to an impaired capacity of the muscle to respond to neural input in people with CRF. In turn, the activation of higher threshold motor units of a lower fatigue resistance could have further compounded these impairments because of their low oxidative, high glycolytic metabolic profile and their slower V̇O2 on-kinetics (37). Thus, the higher EMGRMS activity in the fatigued group might have been both a cause and a consequence of their higher fatigability. Moreover, the higher level of activation coupled with greater metabolic disturbances, even when exercising at low intensities, has potential implications for the perception of effort associated with performing activities of daily living.


The present study performed a secondary analysis using data collected as part of a larger study that investigated the physiological and psychosocial correlates of CRF. The protocol for the cycling exercise in the present study was designed with the objectives of the previous study in mind. Thus, there were some limitations with the approaches used in the present study to better understand the etiology of fatigability. Specifically, to facilitate the interpretation of factors behind the greater fatigability in the fatigued group, the present study determined the power outputs associated with the GET and RCP during the step-incremental cardiorespiratory exercise test to assess the relative intensity of exercise during the neuromuscular fatigability test at the three common stages. However, differences were present between the fatigability and the cardiorespiratory exercise tests in terms of the starting power output (24 ± 7 W vs 36 ± 16 W, respectively), increment (24 ± 7 W vs 11 ± 3 W, respectively), and stage duration (3 vs 1 min, respectively). It has previously been demonstrated that a stage duration of 3 min is associated with a lower power output at the GET relative to 1 min (38). Conversely, steeper increments in power output are associated with a higher power output at GET (39). Thus, calculating the power output during the fatigability test relative to that at the GET during the cardiorespiratory exercise test might have resulted in imprecise estimates in relative exercise intensity. Using measurement of gas exchange during the neuromuscular fatigability test would have permitted more accurate determination of relative exercise intensity, while also allowing for V̇O2 on-kinetics to be assessed. Nevertheless, the results provide novel insight into the etiology and temporal manifestation of neuromuscular fatigability in response to locomotor exercise in people with CRF, which can be used to guide future research. Finally, because this was a secondary analysis and therefore a convenience sample, we may have been underpowered to detect small effects in variables such as VA.


The present findings provide important insight into the etiology of fatigability in people with posttreatment CRF. Specifically, our findings indicate that the greater fatigability in a fatigued versus nonfatigued group of people living beyond cancer can be attributed to exacerbated disturbances at the muscle level, rather than differences in the level of VA. The substantially greater level of fatigability after just 3 min of relatively low-intensity exercise in the fatigued versus nonfatigued group, when the relative intensity of exercise did not differ between groups, suggests a potential role of slower V̇O2 on-kinetics in people with CRF, which has potential implications on fatigability during daily physical activities. Similarly, the higher EMGRMS in the fatigued group during cycling, which might have occurred to compensate for impairments in contractile function, has potential implications for perceptions of effort during activities of daily living and may be a contributing factor to the difficulties in performing such activities in people with CRF.

The authors thank the participants for their commitment during the study.

Data are available upon request.

This study has no conflicts of interest, financial or otherwise. The results of the present study 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.

G. Y. M. and C. G. B. conceived and designed the study; R. T., J. T., T. M., and M. M. performed experiments; C. G. B., R. T., and J. T. analyzed the data; C.G.B. and G.Y.M. interpreted the results of the experiments; C. G. B. drafted the manuscript; R. T., J. T., T. M., M. M., N. C. R., and G. Y. M. edited and revised the manuscript; and C. G. B., R. T., J. T., T. M., M. M., N. C. R., and G. Y. M. approved final version of manuscript.

The study was funded by the Canadian Cancer Society (grant no. 704208-1).


1. Berger AM, Mooney K, Alvarez-Perez A, et al. Cancer-related fatigue, version 2.2015. J Natl Compr Canc Netw. 2015;13(8):1012–39.
2. Williams LA, Bohac C, Hunter S, Cella D. Patient and health care provider perceptions of cancer-related fatigue and pain. Support Care Cancer. 2016;24(10):4357–63.
3. Lawrence DP, Kupelnick B, Miller K, Devine D, Lau J. Evidence report on the occurrence, assessment, and treatment of fatigue in cancer patients. J Natl Cancer Inst Monogr. 2004;32:40–50.
4. Jones JM, Olson K, Catton P, et al. Cancer-related fatigue and associated disability in post-treatment cancer survivors. J Cancer Surviv. 2016;10(1):51–61.
5. Goedendorp MM, Gielissen MFM, Verhagen CAHHVM, Bleijenberg G. Development of fatigue in cancer survivors: a prospective follow-up study from diagnosis into the year after treatment. J Pain Symptom Manage. 2013;45(2):213–22.
6. Islam T, Dahlui M, Majid HA, Nahar AM, Mohd Taib NA, Su TT. Factors associated with return to work of breast cancer survivors: a systematic review. BMC Public Health. 2014;14 Suppl 3(3 Suppl):S8.
7. Bower JE. Cancer-related fatigue—mechanisms, risk factors, and treatments. Nat Rev Clin Oncol. 2014;11(10):597–609.
8. Brownstein CG, Twomey R, Temesi J, et al. Physiological and psychosocial correlates of cancer-related fatigue. J Cancer Surviv. 2021. doi: 10.1007/s11764-021-01115-6.
9. Saligan LN, Olson K, Filler K, et al. The biology of cancer-related fatigue: a review of the literature. Support Care Cancer. 2015;23(8):2461–78.
10. Hofman M, Ryan JL, Figueroa-Moseley CD, Jean-Pierre P, Morrow GR. Cancer-related fatigue: the scale of the problem. Oncologist. 2007;12(S1):4–10.
11. Curt GA, Breitbart W, Cella D, et al. Impact of cancer-related fatigue on the lives of patients: new findings from the fatigue coalition. Oncologist. 2000;5(5):353–60.
12. Enoka RM, Duchateau J. Translating fatigue to human performance. Med Sci Sports Exerc. 2016;48(11):2228–38.
13. Doyle-Baker D, Temesi J, Medysky ME, Holash RJ, Millet GY. An innovative ergometer to measure neuromuscular fatigue immediately after cycling. Med Sci Sports Exerc. 2018;50(2):375–87.
14. Van Belle S, Paridaens R, Evers G, et al. Comparison of proposed diagnostic criteria with FACT-F and VAS for cancer-related fatigue: proposal for use as a screening tool. Support Care Cancer. 2005;13(4):246–54.
15. Merton PA. Voluntary strength and fatigue. J Physiol. 1954;123(3):553–64.
16. Millet GY, Martin V, Martin A, Vergès S. Electrical stimulation for testing neuromuscular function: from sport to pathology. Eur J Appl Physiol. 2011;111(10):2489–500.
17. Black MI, Jones AM, Blackwell JR, et al. Muscle metabolic and neuromuscular determinants of fatigue during cycling in different exercise intensity domains. J Appl Physiol (1985). 2017;122(3):446–59.
18. Yellen SB, Cella DF, Webster K, Blendowski C, Kaplan E. Measuring fatigue and other anemia-related symptoms with the functional assessment of Cancer therapy (FACT) measurement system. J Pain Symptom Manage. 1997;13(2):63–74.
19. Minton O, Stone P. A systematic review of the scales used for the measurement of cancer-related fatigue (CRF). Ann Oncol. 2009;20(1):17–25.
20. Stapelfeldt B, Mornieux G, Oberheim R, Belli A, Gollhofer A. Development and evaluation of a new bicycle instrument for measurements of pedal forces and power output in cycling. Int J Sports Med. 2007;28(4):326–32.
21. Wasserman K, Whipp BJ, Davis JA. Respiratory physiology of exercise: metabolism, gas exchange, and ventilatory control. Int Rev Physiol. 1981;23:149–211.
22. Rodriguez-Falces J, Place N. Determinants, analysis and interpretation of the muscle compound action potential (M wave) in humans: implications for the study of muscle fatigue. Eur J Appl Physiol. 2018;118(3):501–21.
23. Strojnik V, Komi PV. Neuromuscular fatigue after maximal stretch-shortening cycle exercise. J Appl Physiol (1985). 1998;84(1):344–50.
24. Coates KD, Aboodarda SJ, Krüger RL, et al. Multiple sclerosis-related fatigue: the role of impaired corticospinal responses and heightened exercise fatigability. J Neurophysiol. 2020;124(4):1131–43.
25. Lavigne C, Lau H, Francis G, Culos-Reed SN, Millet GY, Twomey R. Neuromuscular function and fatigability in people diagnosed with head and neck cancer before versus after treatment. Eur J Appl Physiol. 2020;120(6):1289–304.
26. Cai B, Allexandre D, Rajagopalan V, et al. Evidence of significant central fatigue in patients with cancer-related fatigue during repetitive elbow flexions till perceived exhaustion. PloS One. 2014;9(12):e115370.
27. Yavuzsen T, Davis MP, Ranganathan VK, et al. Cancer-related fatigue: central or peripheral?J Pain Symptom Manage. 2009;38(4):587–96.
28. Amann M, Dempsey JA. Locomotor muscle fatigue modifies central motor drive in healthy humans and imposes a limitation to exercise performance. JPhysiol. 2008;586(1):161–73.
29. Marcora SM, Bosio A, de Morree HM. Locomotor muscle fatigue increases cardiorespiratory responses and reduces performance during intense cycling exercise independently from metabolic stress. Am J Physiol Regul Integr Comp Physiol. 2008;294(3):R874–83.
30. Twomey R, Aboodarda SJ, Kruger R, Culos-Reed SN, Temesi J, Millet GY. Neuromuscular fatigue during exercise: methodological considerations, etiology and potential role in chronic fatigue. Neurophysiol Clin. 2017;47(2):95–110.
31. Fitts RH. The cross-bridge cycle and skeletal muscle fatigue. J Appl Physiol (1985). 2008;104(2):551–8.
32. Allen DG, Lamb GD, Westerblad H. Skeletal muscle fatigue: cellular mechanisms. Physiol Rev. 2008;88(1):287–332.
33. Grassi B. Skeletal muscle V̇O2 on-kinetics: set by O2 delivery or by O2 utilization? New insights into an old issue. Med Sci Sports Exerc. 2000;32(1):108–16.
34. Temesi J, Mattioni Maturana F, Peyrard A, Piucco T, Murias JM, Millet GY. The relationship between oxygen uptake kinetics and neuromuscular fatigue in high-intensity cycling exercise. Eur J Appl Physiol. 2017;117(5):969–78.
35. Cerretelli P, Pendergast D, Paganelli WC, Rennie DW. Effects of specific muscle training on V̇O2 on-response and early blood lactate. J Appl Physiol Respir Environ Exerc Physiol. 1979;47(4):761–9.
36. Inglis EC, Iannetta D, Murias JM. The relationship between the time constant of V̇O2 kinetics and V̇O2max in humans. Eur J Appl Physiol. 2021;121(9):2655–6.
37. Han YS, Proctor DN, Geiger PC, Sieck GC. Reserve capacity for ATP consumption during isometric contraction in human skeletal muscle fibers. J Appl Physiol (1985). 2001;90(2):657–64.
38. Roffey DM, Byrne NM, Hills AP. Effect of stage duration on physiological variables commonly used to determine maximum aerobic performance during cycle ergometry. J Sports Sci. 2007;25(12):1325–35.
39. Weston SB, Gray AB, Schneider DA, Gass GC. Effect of ramp slope on ventilation thresholds and V̇O2peak in male cyclists. Int J Sports Med. 2002;23(1):22–7.


Copyright © 2022 by the American College of Sports Medicine