The effects of 6 weeks of high load or low-load blood flow restriction resistance exercise training on motor unit firing rates in males and females

Abstract

The purpose of the present investigation is to examine the changes in isometric strength and motor unit (MU) firing rates of the vastus lateralis (VL) and vastus medialis (VM) in response to 6-weeks of high-load (HL; n = 10, 4 female) and low-load blood flow restriction resistance exercise training (RET) (LL-BFR; n = 10, 5 females). MU firing rates and recruitment threshold relationships were analyzed from surface electromyographic (sEMG) recordings of the leg extensors during submaximal isometric contractions performed at 10%, 40%, and 70% maximal voluntary contraction (MVC) pre- and post-training. Linear mixed-effects models were used to assess the influence of recruitment threshold, contraction intensity, muscle, sex, and training group on MU firing rates. Isometric strength increased in response to training (p = 0.010, d = 0.49), with no significant interaction for training groups. Firing rates remained unchanged in response to interventions. For the VL, females displayed greater firing rates regardless of recruitment position during the 10% MVC, whereas later-recruited MUs during the 40% and 70% MVC were higher, with no difference in earlier-recruited MU. For the VM, females exhibited greater firing rates of the higher-threshold MUs across all submaximal intensities. Interestingly, males displayed greater firing rates of the earlier-recruited MUs of the VM across all submaximal intensities. Thus, there were non-uniform sex-related differences in MU firing rates for the VL and VM. In conclusion, similar increases occurred for isometric strength following 6-weeks of LL-BFR or HL RET in absence of an increase in firing rates of the leg extensors.

NEW & NOTEWORTHY

MU firing rates did not differ in response to HL or LL-BFR training when accounting for individual variability, contraction intensities, and recruitment position. Contributing to the divergent literature on sex-related differences in firing rates, highlighting non-uniform differences across muscle and contraction intensity. In addition, firing rates differ between VL and VM across parameters. To our knowledge, no prior investigation has compared firing rates of both leg extensor muscles in response to RET.

Graphical Abstract

INTRODUCTION

It is well-established that traditional heavy resistance exercise training (RET) leads to an increase in muscular strength and hypertrophy (Jenkins, 2017; Schoenfeld, 2015, 2017; Vann, 2020). Early strength adaptations in response to such training are primarily attributed to the modulation of neural mechanisms in the absence of muscular hypertrophy (Carroll, 2001; Del Vecchio, 2019; Moritani & deVries, 1979; Škarabot, 2021). Speculation of such contributions often involve elevated motor unit (MU) firing rates. However, conflicting evidence exists due to the musculature measured, electromyography (EMG) methodology (intramuscular vs. surface EMG), and procedural analysis (De Souza, 2018; Del Vecchio, 2019; Guo, 2022; Herda, 2022; Inglis & Gabriel, 2020; Peng, 2018).

Low-load blood flow restriction (LL-BFR) has gained traction as a RET modality due to its ability to elicit muscular strength and hypertrophic responses in absence of high-loads (Abe, 2006; Cook, 2013; Hill, 2020; Kambič, 2019; Lauver, 2020; Lowery, 2014; Perera, 2022). There is, however, limited information regarding neural adaptations following chronic LL-BFR. Acute research demonstrates changes in neural activation (Davis, 2025; Fatela, 2019; Iskarevskii, 2024; Lowe, 2023; Olmos, 2024; Taleshi, 2025), such as, elevated muscle activation of the vastus lateralis (VL) following three bouts of submaximal isometric LL-BFR as measured by electromyographic (EMG) amplitude (Davis et al., 2025). EMG amplitude is a useful indicator of neural activation, however, the decomposition of EMG signals into individual MU action potential (AP) trains will yield deeper insight on the potential neural adaptations from LL-BFR RET. For example, Iskarevskii et al. (2025) reported a shift toward earlier recruitment of high-threshold MUs, rather than lower-threshold, of the wrist flexors during a 20% isometric maximal voluntary contraction (MVC), when BFR was applied. Others report increases in firing rates at 40% MVC during or following BFR of the VL and biceps brachii (Fratela, 2019; Olmos, 2024). Overall, acute BFR alters MU recruitment and firing rate patterns as measured with action potential trains. It remains, however, unclear if chronic adaptations occur at the level of the MU.

Previous investigations provide conflicting results regarding sex-related differences in MU firing rates (Dimmick, 2022; Guo, 2022; Inglis & Gabriel, 2020; Lecce, 2024; Olmos, 2024; Parra, 2020; Peng, 2018). There were no sex-related differences in firing rates for the VL (Dimmick et al., 2022) or biceps brachii (Olmos et al., 2024) during 40% MVCs when data was analyzed separately for subjects and contractions while also accounting for recruitment threshold. In contrast, Leece et al. (2024) used similar data analyses procedures and reported greater firing rates in females during 35% MVCs with reverse relationships during 75% MVCs. Adding to the conflicting evidence, Parra et al. (2020) observed that firing rates were greater in the first dorsal interosseous for females during higher intensity contractions. Other findings suggest elevated firing rates in females across submaximal force levels from 20% to 80% MVC (Guo, 2022; Inglis & Gabriel, 2020; Peng, 2018) in comparison to males for the VL (Guo et al., 2022), tibialis anterior (Inglis & Gabriel, 2020), and the vastus medialis (VM) when not accounting for variability as a function of subject, contraction intensity, and recruitment threshold. Subsequently, potential non-uniform sex-related differences may be related to the muscle examined and contraction intensity.

Therefore, the primary purpose of the present investigation is to examine the changes in isometric strength and MU firing rates of the VL and VM in response to 6-weeks of high-load (HL) RET and LL-BFR in untrained males and females. MU firing rate and recruitment threshold relationships were analyzed from surface EMG (sEMG) recordings of the VL and VM during submaximal isometric contractions of the leg extensors performed at 10%, 40%, and 70% MVC before and after training. To evaluate these changes, linear mixed-effects models were used to assess the influence of recruitment threshold, contraction intensity, muscle, sex, and training group on MU firing rates. It is hypothesized that increases in isometric leg extensor strength will be similar between the HL and LL-BFR groups, as evidenced by previous investigations (Centner, 2022; Kambič, 2019; Perera, 2022). Additionally, it is hypothesized that MU firing rates will change in response to BFR training because of the reported acute changes. Finally, it is hypothesized that females will exhibit greater MU firing rates of the VL and VM during the 70% MVC with no sex-related differences during relatively lower contraction intensities (10% and 40% MVC) as observed in previous reported when analyzed with analytical procedures that account for variability among subjects, contraction intensities, and recruitment position (Trevino, 2022; Olmos, 2024; Parra, 2020).

METHODS

Participants

Twenty recreationally active males and females participated in this study. Subjects were randomly selected to participate in either a high-load resistance training group (HL; n = 10, 4 female) or a low-load resistance training with BFR (BFR; n = 10, 5 females) (Table 1). Exclusion criteria included any cardiovascular, metabolic, or musculoskeletal condition. Individuals with a history of severe lower extremity and/or pathological condition impairing motor control were also excluded. Male and female participants ranged from 18 to 30 years of age in this study. Participants were untrained as defined by having engaged in zero to six total structured resistance training sessions in the previous six months. A health history questionnaire administered on visit 1 determined each participant’s activity status. Individuals taking part in recreational activities (e.g., jogging, intramural sports, etc.) were eligible to participate only if the resistance training requirements were met. Participants were instructed to refrain from resistance training outside of the supervised training sessions for the duration of the current study. Participants of both sexes had minimal resistance training experience as indicated by the questionnaire. The institutional review board for human subjects research at the participating institution approved this study, and was conducted in accordance with the Declaration of Helsinki (IRB Study no. 00147374). This study was registered as a clinical trial prior to the recruitment of the first participant (June 25, 2021, NCT049389). All participants read and signed an informed consent form.

Table 1.

Participant Anthropometric Information

	Subject #	Age (yrs)	Height (cm)	Body Mass (kg)
Male
HL	5	21.2 ± 1.9	182.2 ± 6.3	73.7 ± 9.1
BFR	6	21.3 ± 2.8	179.9 ± 8.7	80.9 ± 23.2
Female
HL	4	23 ± 3.5	167.3 ± 7.1	79.6 ± 10.4
BFR	5	21.5 ± 3.8	167.2 ± 10.4	67.1 ± 7.9

Number (#) of males and females in the high-load (HL) resistance exercise training and low-load resistance exercise training with blood flow restriction (BFR) groups and the mean ± SD of age, height, and body mass.

Research Design

Participants visited the laboratory on two occasions prior to (visits 1 and 2) and following (visits 3 and 4) a 6-week resistance training intervention of leg extensions that included either a high load or low load with BFR. The first and third visit included anthropometric measures and a familiarization session that included practicing maximal and submaximal isometric contractions of the knee extensors that was performed during the experimental trial (visits 2 and 4). Maximal lower body strength is not impaired by one or more phases of a female’s menstrual cycle provided by consistent evidence (Blagrove, 2020; De Jonge, 2001; Fridén, 2003; Haun, 2019; Lebrun, 1995; Ruple, 2022). However, potential strength fluctuations due to the menstrual cycle were considered via a questionnaire in the initial first visit. Hormonal conditions during the pre- and post-training tests were accounted for by delaying the training intervention for female participants for 2 – 3 weeks after visit 2 in attempt to provide testing within the same phase of each female’s menstrual cycle. The average cycle length ranges 28 – 35 days (Creinin et al., 2004). Thus, the start of the intervention was delayed (14 to 21 days) to allow post-testing to occur during the same phase of each participant’s menstrual cycle.

Resistance Exercise Training Intervention and Blood Flow Restriction

Participants in both interventions completed 18 resistance training sessions (3x/week for 6 weeks). Training sessions were limited to two consecutive days per week. Participants performed three sets of bilateral leg extensions to failure in each session. Failure was determined if the concentric muscle action could not be completed through the whole range of motion as determined by a team member. Participants were instructed to perform the concentric portion of each repetition as fast as possible, while controlling the eccentric portion of each repetition. Participants were instructed to refrain from additional resistance exercise training outside of the supervised sessions. Throughout each resistance exercise training session, strong instruction and encouragement were provided. All resistance exercise training sessions were conducted under the direct supervision of the research staff. The training interventions selected for the HL (Jenkins, 2017; Martín-Hernández, 2013; Mitchell, 2012) and BFR (Ellefsen, 2015; Laurentino, 2012; Martín-Hernández, 2013) groups have demonstrated significant gains in muscular size in untrained participants.

Participants in the HL group performed 8 – 12 repetitions for each set. The load increased (~ 2.2 kg) if 12 repetitions were completed to achieve progressive overload as well as ensuring failure was reached. Participants unable to perform the minimum repetitions within the range of motion led to a decrease in load (~ 2.2 kg). Participants performed 80% of their 1RM in the initial set in the first training session. For each training session, the last set was used to determine the initial load for the following session. Two-minutes of rest was provided between each set.

Participants in the BFR group performed sets to volitional failure utilizing 30% of their 1RM with 1 minute of rest between each set (Clark et al., 2011). The load performed by participants was held constant throughout the entirety of the intervention due to the high variability in response to BFR (Dankel, 2017; Madarame, 2008; Takarada, 2004; Wang, 2025). LL-BFR training to volitional failure ensures standardization across participants, particularly given the variability in repetition performance at fixed loads (Dankel et al., 2017). Mean repetitions across sessions and total training volume data for both HL and LL-BFR can be viewed in Figure 1. The 10-cm wide BFR cuffs (SmartTools, Strongsville, OH) were applied to the most proximal region of both legs. The pressure of each cuff was set to 50% arterial occlusion pressure (AOP) as estimated from each participant’s thigh circumference measured in the proximal region (100 – 180 mmHg) (Loenneke et al., 2012, 2015). Participant discomfort was minimized with 50% AOP, as increases in muscle activity have not been demonstrated with higher pressures while performing leg extensions (Loenneke et al., 2015). The pressure applied was in accordance with the instructions in the SmartTools User’s Manual (SmartTools). Deflated cuffs were securely positioned proximally on the thighs of participants while standing, cuffs were gradually inflated to the desired pressure. Upon achievement of pressure, resistance training sessions would last no more than 10 minutes (Patterson et al., 2019). The inflated cuffs remained until the completion of the final set of the training session.

Figure 1.

Mean (± SD) number of repetitions completed per training session across weeks for the 6-week intervention for the blood flow restriction (BFR; red, n = 10) and high-load (HL; blue, n = 10) groups. Repetitions in the BFR group were performed to volitional failure, resulting in progressive increases in repetitions across weeks whereas the HL group remained stable.

Isometric Testing

All isometric knee extensor strength assessments were performed on the right leg at a flexion of 90° on a Biodex System 3 isokinetic dynamometer (Biodex Medical Systems, Shirely, NY). The input axis of the Biodex was aligned with the lateral condyle of the femur of each participant seated with restraining straps over the pelvis, trunk, and contralateral thigh. Participants performed two isometric maximal voluntary contractions (MVC), followed by submaximal trapezoidal contractions at 10%, 40%, and 70% MVC during both pre- and post-assessments. For all isometric trapezoidal contractions, torque increased at 10% MVC/s to targeted torque levels (10%, 40%, and 70% MVC), torque was sustained for 40 sec, 20 sec, and 10 sec, respectively, and then decreased at 10% MVC/s to baseline (Figure 2A–C). Contraction intensities were performed in a randomized order. Three minutes of rest was provided between each muscle action.

Figure 2.

A-C: Submaximal trapezoidal contractions for the isometric knee extensors at 10% (A), 40% (B), and 70% (C) maximal voluntary contraction (MVC) for one participant. Each row of vertical tick marks represents an individual motor unit’s (MU) firing instances, black line represents force output. D-F: The firing rates versus the recruitment threshold relationships for the vastus lateralis (VL) from each of the 10% (D), 40% (E), and 70% (F) MVCs. The MUs depicted corresponds to the tick marks in A-C.

Electromyography

sEMG signals were collected from the VL and VM using a 5-pin surface sensor array (Delsys, Inc, Natick, MA) during all isometric trapezoidal contractions pre- and post-intervention. The pins have a diameter of 0.5 mm and are arranged in a 5 x 5 mm square with the fifth pin positioned in the center. Prior to sensor placement, the skin was prepared by shaving, repeated removal of superficial dead skin and sterilized with an alcohol swab. The sensor was positioned over the VL at approximately 50% of the distance between the greater trochanter and the lateral condyle of the femur with hypoallergenic tape. For the VM, sensor was positioned 2/3 on the line from the anterior spina iliac spina superior to the lateral side of the patella. The reference electrode was positioned over the left patella. The sEMG signals were sampled at 20kHz and stored on a computer for offline analysis. The four channels collected from sEMG signals with the 5-pin array sensor were decomposed with the Precision Decomposition III (PDIII) algorithm into firing events of single MUs (De Luca et al., 2006). Further detailed information regarding the signal processing of EMG signals is described by De Luca et al. (2006) and Nawab et al. (2010). A reconstruct-and-test procedure examined the accuracy of decomposed firing times, only MUs with > 90% accuracies were included for further analysis (Nawab et al., 2010). Additionally, a secondary spike trigger average (STA) validation procedure was applied to the PDIII algorithm-derived firing times and action potential waveforms (Herda, 2020; Hu, 2013a). Inclusion of a MU for further analysis required a high correlation between the STA- and PDIII-derived waveforms (r > 0.7) across the four sEMG channels and the coefficient of variation of the STA-derived action potential peak-to-peak amplitude remained < 0.3 across time selected (Hu et al., 2013a). MU action potential waveforms could potentially appear despite not corresponding with discharge times (Farina et al., 2014). Therefore, Gaussian noise set at 1% standard deviation for interspike interval was applied to the discharge times for each MU (Hu, 2013b; Thompson, 2018). Correlations were performed on the MU action potential waveforms created from the raw EMG signals with the small amount of noise (< 2 ms shift) added to the firing times with the action potential waveforms that were triggered with the PDIII algorithm derived firing times. Furthermore, the peak-to-peak amplitudes of the STA action potential waveforms were compared to the action potential waveforms derived from the firing times with the addition of Gaussian noise. The small additional noise provided to the firing times should lower the correlation between action potential waveforms derived form the PDIII algorithm and the STA procedures (Hu, 2013b; Thompson, 2018). If no true action potential waveform is present, the peak-to-peak amplitudes derived from the firing times with Gaussian noise should decrease. MU action potential amplitudes were quantified as the average peak-to-peak amplitude from the four unique waveforms created from the four channels of the sEMG sensor (Hu et al., 2014) (Figure 1D–F) while its recruitment threshold (RT, expressed as %MVC) was calculated as the average force firing the 0.10 sec following its initial firing instance (Sterczala et al., 2020).

Statistical Analysis

Data collected from pre- and post-testing sessions were used to access MU firing rates and action potential amplitudes. To understand the impact of muscle, sex, and training intervention on MFR, linear mixed effects models were fit using restricted maximal likelihood with the lme4 package in R (Bates et al., 2015). Linear mixed effect modeling is well-suited for smaller sample sizes as this technique accounts for individual variability by including both fixed and random effects. These benefits reduce the noise and provide more precise estimates, thus limiting inflation of type I errors. MU recruitment threshold (RT) and contraction intensity were used as covariates. For all models, contraction intensity was implemented as a continuous covariate within the fixed-effects and as categorical within the random effect grouping factors. This was done to improve statistical power and prevent information loss when drawing inferences while still being able to account for within-intensity correlations within the data. To aid model convergence and interpretation, the RT variable was contraction-centered by subtracting the mean RT for each ramp contraction from the raw RT variable. This centered metric (RT_c) provides the RT relative to the mean value for each combination of subject, time, intensity, and muscle. An initial, maximal, model was fit with a 6-way RT_c x Time x Muscle x Sex x Group x Intensity interaction and all lower-order terms as fixed effects. To fully account for all of the dependencies within the design, random intercepts were fit for each Subject and all of the combinations of Subject with the other categorical variables (e.g. Subject:Time, Subject:Intensity, Subject:Intensity:Muscle) with random slopes for centered-RT fit for each grouping factor. Singularity warnings led to the removal of the Subject and Subject:Muscle random intercepts as well as the random RT_c slopes within the Subject:Time factor and Subject:Time:Muscle factors. Prior to removal of the random slope terms, the intercept-slope correlation within Subject:Time and Subject:Time:Muscle factors was removed in an attempt to aid model fitting, but when singularity warnings were still encountered, the slope terms were removed altogether for these grouping factors. The significance of fixed effects was assessed using Type II Wald Chi-square tests using the Anova function from the car package (Fox et al., 2001). Non-significant fixed effects were removed from the model such that the only significant interaction terms and their lower-order constituent terms remained. Our final model had 3 significant 3-way interaction (Muscle:Time:RT_c, Muscle:Sex:RT_c, Muscle:Sex:Group), 2 significant 2-way interactions (Muscle:Percent_MVC, Percent_MVC:RT_c) and all of the lower order terms involved in these interactions. A likelihood-ratio test revealed no significant difference between the maximal model and the final, more parsimonious, model. Univariate and multivariate normality for level 1 and level-2 residuals, respectively, were confirmed using quantile-quantile plots. A two-way mixed factorial repeated measures ANOVAs [group (BFR vs. HL) x time (PRE vs. POST)] was used to analyze maximal isometric peak torque of the knee extensors. If appropriate, independent sample t-tests were performed with holm corrections as a follow-up analysis.

RESULTS

Isometric Strength

There was no significant two-way interaction (p = 0.239, ɳ² = 0.060) or a main effect for group (p = 0.689, ɳ² = 0.007), however, there was a significant main effect for time (p = 0.022, ɳ² = 0.208). Isometric strength was greater at post- (257.5 ± 98.3 Nm; p = 0.010; d = 0.49) in comparison to pre- testing (224.4 ± 93.3 Nm) when collapsed across groups. Percentage of strength gains between intervention groups were 8. 9 ± 12.6% and 14.0 ± 27.0% for BFR and HL, respectively.

Firing Rates

The final linear mixed effects model included three significant 3-way interactions (Muscle x Time x RT_c, Muscle x Sex x RT_c, and Muscle x Sex x Group), two significant 2-way interactions (Muscle x Percent_MVC and Percent_MVC x RT_c) and all lower-order terms involved in these significant interactions. Decomposition of the interactions revealed that there were no time-related changes in firing rates as a function of muscle (VL and VM) or training (HL and BFR) (Figure 3). Overall, the firing rates did not change as a result of the interventions. Time-related changes in slope for each group are located in Table 2.

Figure 3.

The predicted mean firing rates (MFR) versus recruitment threshold relationships for the vastus lateralis (VL) and medialis (VM) from the 10%, 40%, and 70% maximal voluntary contractions (MVC) pre- (red) and post-intervention (blue). Recruitment thresholds were centered by subtracting the mean recruitment threshold for each trapezoidal contraction from the raw recruitment threshold variable.

Table 2.

Percent Change (Δ%) in Firing Rate vs Recruitment Threshold Slopes for each Group

Group	Intensity	Muscle	Slope (Δ%)
BFR	10	VL	−6.53 ± 11.22
BFR	10	VM	0.02 ± 17.12
BFR	40	VL	−0.22 ± 16.6
BFR	40	VM	5.51 ± 21.52
BFR	70	VL	−2.58 ± 21.3
BFR	70	VM	11.27 ± 23.95
HL	10	VL	−0.07 ± 13.8
HL	10	VM	−6.79 ± 12.45
HL	40	VL	9.75 ± 25.83
HL	40	VM	10.87 ± 27.44
HL	70	VL	10.17 ± 32.46
HL	70	VM	−8.46 ± 19.95

Decomposition of the interactions did indicate sex-, muscle-, and contraction intensity-related differences in firing rates. The most significant findings were that females possessed greater firing rates for the higher-threshold MUs of the VL and VM when analyzed separately for 10%, 40% and 70% MVCs (Table 3 and Figure 4). For the main effect of muscle, there were greater firing rates for the VL in comparison to the VM regardless of group, intensity, and sex (Table 4 and Figure 3). Finally, as expected, the firing rates were highest during the 70% MVC in comparison to 40% and 10% MVCs while the firings rate from the 40% MVC were greater than the 10% MVC.

Table 3.

Sex-related differences in MU firing rates as a function of contraction intensity, muscle, and recruitment threshold. Early, mid, and late represent the estimated firing rates for the lowest, median, and highest recorded MUs.

Intensity	Recruitment Positions	Muscle	Sex	Estimate	p-value
10%	Early	VL	M - F	−1.6167	0.027*
	Mid	VL	M - F	−1.9649	0.003*
	Late	VL	M - F	−2.3241	0.001*
40%	Early	VL	M - F	−1.0086	0.472
	Mid	VL	M - F	−1.9649	0.003*
	Late	VL	M - F	−2.9782	0.003*
70%	Early	VL	M - F	−0.5042	0.652
	Mid	VL	M - F	−1.9649	0.003*
	Late	VL	M - F	−3.4881	0.007*
10%	Early	VM	M - F	0.683	0.534
	Mid	VM	M - F	−0.5678	0.534
	Late	VM	M - F	−1.8578	0.010*
40%	Early	VM	M - F	2.8669	0.003*
	Mid	VM	M - F	−0.5678	0.473
	Late	VM	M - F	−4.2072	< 0.001 *
70%	Early	VM	M - F	4.6783	0.001*
	Mid	VM	M - F	−0.5678	0.625
	Late	VM	M - F	−6.0385	< 0.001 *

^*Represents significant differences

Figure 4.

The predicted mean firing rates (MFR) versus recruitment threshold relationships for the vastus lateralis (VL) and medialis (VM) from the 10%, 40%, and 70% maximal voluntary contractions (MVC) for males (M; blue, n = 11) and females (F; red, n = 9). * Indicates significant differences between males and females for the given recruitment threshold, (p < 0.05).

Table 4.

Differences in firing rates across muscles at average recruitment threshold (collapsed across sex, group, and time).

MVC Intensity	Muscle_pairwise	estimate	p.value
10%	VL - VM	3.4223	<0.001*
40%	VL - VM	2.8209	<0.001*
70%	VL - VM	2.2196	<0.001*

^*Represents significant differences.

DISCUSSION

Increases in isometric strength were similar following HL or LL-BFR training with no corresponding changes in firing rates. A novel finding was that there were non-uniform sex-related differences in firing rates as a function of muscle and contraction intensity (Table 3 and Figure 4). Additionally, VL firing rates were greater than the VM at each intensity when averaged across intervention (HL and BFR), time and sex (Table 4 and Figure 3).

In the present study, firing rates of the VL and VM remained unchanged following 6-weeks (18 sessions) of HL-RET in a similar manner to other investigations that analyzed data separately for subject and contraction (Sterczala et al., 2020). A few studies do report alterations in firing rates following RET in healthy young adults when not accounting contraction intensity (Del Vecchio, 2019; Kamen & Knight, 2004; Kidgell, 2006; Vila-Chã, 2010), recruitment threshold (Kamen & Knight, 2004; Kidgell, 2006; Vila-Chã, 2010) and/or disregard variability among subjects in statistical models (Del Vecchio, 2019; Kamen & Knight, 2004; Kidgell, 2006; Vila-Chã, 2010). Comparable increases in strength and muscle fiber cross-sectional area (CSA) have been observed following 4- and 9-week LL-BFR or HL resistance training of the VL and VM (Davids, 2021; Reece, 2023; Libardi, 2024). Indirect assessments via the ultrasound are in support of the limited direct evidence for the comparable increases in muscle fiber CSA in response to LL-BFR and traditional resistance training (Farup, 2015; Freitas, 2017). Similar signaling pathways have been observed at the fiber level (Davids et al., 2021) potentially contributing to the comparable gains in hypertrophy, therefore increasing force output. In summary, improvements in isometric strength following RET in relatively young healthy adults might primarily be a function of mechanisms at the level of the muscle fibers (Davids, 2021; Reece, 2023; Libardi, 2024).

Previous literature supports the potential chronic adaptation to MU firing rates in response to LL-BFR. For example, BFR applied to a low-load unilateral leg extensor exercise (20% 1RM) resulted in elevated firing rates of the VL during an isometric trapezoid at 40% MVC when accounting for recruitment threshold and analyzed separately for contraction and subject (Fatela et al., 2019). Therefore, it is plausible that chronic LL-BFR may lead to increases in MU firing rates. However, the current investigation examined 6-weeks (18 sessions) of LL-BFR on the VL and VM demonstrating no effect in firing rates across isometric contractions performed from low to relatively higher intensities. Therefore, the ability to perform the isometric trapezoidal contractions at higher torque levels post-LL-BFR was not dependent on greater firing rates of the leg extensors but rather might primarily be the result of improvements in mechanical output. Acute, 6-week, LL-BFR RET has resulted in hypertrophy (Reece, 2023; Libardi, 2024), muscle protein turnover, and ribosomal biogenesis (Sieljacks et al., 2019) of the muscle fibers of the VL. These mechanical adaptations of the muscle fibers has led to improvements in strength of the knee extensors (Reece, 2023; Libardi, 2024; Sieljacks et al., 2019).

The current investigation observed comparable increases in isometric strength following HL or LL-BFR training with no increases in MU firing rates. Alternative neural mechanisms may have contributed to the improvement in isometric strength. The synchronization of active MU firing at fixed time intervals may explain the increase in the isometric strength in response to RET commonly measured via EMG amplitudes (Aagaard, 2002; De Luca, 1993). Increases in isometric strength alongside corticospinal excitability, measured via transcranial magnetic stimulation (TMS), have been observed in the knee extensors in absence of MU firing rates (Wilson et al., 2023). Therefore, increased corticospinal excitability may serve as a potential explanation. The recruitment or increase in size of the higher threshold MUs may potentially explain the increase in isometric strength, however, this suggestion is mechanical in origin (Sterczala et al., 2020). The improvements in isometric strength may primarily be the result of mechanical adaptations to muscular fibers. In support, Reece et al. (2023) reported similar increases in muscle cross-sectional area of the VL between HL and LL-BFR conditions via ultrasound imaging. The authors, however, reported that HL did promote greater increases in muscle fiber cross-sectional area in comparison to LL-BFR (Reece et al., 2023). In addition, increases in strength, however, may be attributed to improvements in mechanical output via elevated evoked twitch torque (Colson, 2009). There is potential that neural adaptations at the level of the motoneuron other than peak firing rates could improve strength following HL and LL-BFR, such as, persistent inward currents (Orssatto et al., 2023), firing rate synchronization, etc. (Davis, 2025; Fatela, 2019; Gabriel, 2006; Iskarevskii, 2024; Lowe, 2023; Olmos, 2024; Taleshi, 2025).

There is conflicting evidence regarding potential sex-related differences in firing rates. For example, Dimmick et al. (2022) observed greater firing rates in females for the VL in comparison to males during a 40% MVC when not accounting for recruitment position and inter-individual variability. However, the authors report no differences between MU firing rates when analyzing the data on a subject-by-subject basis and regressed against recruitment threshold. Other studies report elevated firing rates at lower contraction intensities of the VL (Guo et al. 2022), higher contraction intensity for the VM (Peng et al., 2018), and for the anterior tibialis (Inglis and Gabriel, 2020) without accounting for recruitment position and variability among individuals.

Our findings significantly expand upon previous results with an exhaustive examination of firing rate patterns across contraction intensities, recruitment thresholds, and the inclusion of multiple muscles. For the VL, firing rates were greater for females regardless of recruitment position during the 10% MVC. Whereas firing rates were higher for the mid- to -later recruited MUs for the females during the 40% and 70% MVCs with no differences for the earlier recruited MUs. For the VM, the later recruited MUs for the females possessed greater firing rates during the 10% MVC. Interestingly, the earlier recruited MUs for the males were greater than the females with the inverse being true for the later recruited MUs during the 40% and 70% MVCs. A novel finding is that there are non-uniform differences in firing rates between sexes for two leg extensor muscles.

Factors contributing to sex-related differences in MU firing rates typically include neural and mechanical mechanisms. Mechanical mechanisms related to cross-sectional area (CSA) and distribution of contractile properties (type I vs II fiber area) within skeletal muscle. For the leg extensors, males typically possess larger type II fiber area that explains larger CSAs than females for the leg extensors and forearm flexors (Guo, 2022; Haizlip, 2015; Herda, 2019; Lecce, 2024; Reece, 2023; Trevino, 2019). Furthermore, greater area of type II fibers often explains the larger muscle CSA for males and, subsequently, partially explains sex-related variability in the firing rate versus recruitment threshold relationships (Trevino et al., 2016). It is speculated that females gradually increase excitatory drive to compensate for the recruitment of relatively smaller and weaker muscle fibers in comparison to males that explains greater firing rates of later recruited MUs (Guo, 2021; Lecce, 2024). In addition, there were sex-related differences in the firing rates of lower-threshold MUs in the current study. It is theorized that the modulation in force production can also be influenced by intrinsic motoneuron properties (Del Vecchio et al., 2019). Potentially, the differences in firing rates of the lower-threshold MUs could be a function of intrinsic properties since type I fiber area is more similar across sexes unlike type II fiber area (Reece, 2023; Staron, 2000). Therefore, the differences in contractile (MHC content distribution) as well as the excitatory properties may account for the non-uniform firing rates during submaximal contraction intensities of the leg extensors.

Our results indicate that the firing rates of the VM were lower in comparison to the VL when analyzed on a contraction-by-contraction basis and regressed against recruitment thresholds. Interestingly, there is a scarcity of literature directly assessing disparities in MU firing rates among the leg extensors (De Souza et al., 2018), however, indirect evidence reports greater firing rates in the VM in comparison to the VL (De Luca & Kline, 2012). In contrast to the current findings, De Souza et al. (2018) reported elevated firing rates of the VM versus the VL during 20% MVC isometric contractions at 80-degree knee flexion. The difference in statistical analysis may provide insight into the conflicted findings. Inter-individual variability was not accounted for firing rates and recruitment thresholds, as MUs were pooled across participants rather than analyzed on a subject-by-subject basis (Herda, 2022). The authors reported no significant differences in pooled MU recruitment thresholds. However, the median and distribution range values provided by the authors (Figure 4) could potentially indicate that recruitment thresholds partially explain differences in pooled firing rates. The median and interquartile interval for VL recruitment thresholds presented were elevated in comparison to the VM and, thus, firing rates were greater for the VM as they quantified relatively lower-threshold MUs in comparison to the VL. Nonetheless, these discrepancies underscore the necessity of more granular analysis to reconcile the conflicted findings. Future research should address these methodological differences to provide a better understanding of the varying firing rates of MUs between the VL and VM. Furthermore, underlying mechanisms should be researched, such as proprioceptive feedback and motoneuron excitability.

In conclusion, similar increases occurred for isometric strength following 6-weeks of LL-BFR or HL RET in absence of an increase in firing rates of the leg extensors. Potentially, increases in mechanical output of fibers might be the primary contributor to increases in strength (Reece et al., 2023) in relatively healthy young untrained adults following 6-weeks of RET. A novel finding of the results revealed non-uniform sex-related differences in the firing rates as a function of muscle and contraction intensity. Differing distributions of contractile and/or excitatory properties during submaximal contractions among sexes may provide insight into the differing of firing rates, however, further investigation is required. Additionally, the VL regardless of sex or group displayed greater firing rates as compared to the VM. Further exploration is needed to determine underlying mechanisms for sex- and muscle-related differences in MU firing rates. The results of the current investigation are among young healthy untrained adults and therefore, may not apply to different populations (trained individuals, older populations, clinical groups, etc.). Future research should incorporate long-term RET programs, with the addition of a follow-up session, to examine chronic adaptations or delayed neural adaptations.

DATA AVAILABILITY

The data that support the findings of this study are available upon request.