Blood flow restriction attenuates surface mechanomyography lateral and longitudinal, but not transverse oscillations during fatiguing exercise

Objective. Surface mechanomyography (sMMG) can measure oscillations of the activated muscle fibers in three axes (i.e. X, Y, and Z-axes) and has been used to describe motor unit activation patterns (X-axis). The application of blood flow restriction (BFR) is common in exercise studies, but the cuff may restrict muscle fiber oscillations. Therefore, the purpose of this investigation was to examine the acute effects of submaximal, fatiguing exercise with and without BFR on sMMG amplitude in the X, Y, and Z-axes among female participants. Approach. Sixteen females (21 ± 1 years) performed two separate exercise bouts to volitional exhaustion that consisted of unilateral, submaximal (50% maximal voluntary isometric contraction [MVIC]) intermittent, isometric, leg extensions with and without BFR. sMMG was recorded and examined across percent time to exhaustion (%TTE) in 20% increments. Separate 2-way repeated measures ANOVA models were constructed: (condition [BFR, non-BFR]) × (time [20, 40, 60, 80, and 100% TTE]) to examine absolute (m·s−2) and normalized (% of pretest MVIC) sMMG amplitude in the X-(sMMG-X), Y-(sMMG-Y), and Z-(sMMG-Z) axes. Main results. The absolute sMMG-X amplitude responses were attenuated with the application of BFR (mean ± SD = 0.236 ± 0.138 m·s−2) relative to non-BFR (0.366 ± 0.199 m·s−2, collapsed across time) and for sMMG-Y amplitude at 60%–100% of TTE (BFR range = 0.213–0.232 m·s−2 versus non-BFR = 0.313–0.445 m·s−2). Normalizing sMMG to pretest MVIC removed most, but not all the attenuation which was still evident for sMMG-Y amplitude at 100% of TTE between BFR (72.9 ± 47.2%) and non-BFR (98.9 ± 53.1%). Interestingly, sMMG-Z amplitude was not affected by the application of BFR and progressively decreased across %TTE (0.332 ± 0.167 m·s−2 to 0.219 ± 0.104 m·s−2, collapsed across condition.) Significance. The application of BFR attenuated sMMG-X and sMMG-Y amplitude, although normalizing sMMG removed most of this attenuation. Unlike the X and Y-axes, sMMG-Z amplitude was not affected by BFR and progressively decreased across each exercise bout potentially tracking the development of muscle fatigue.


Introduction
The use of non-invasive devices to evaluate human motor function has garnered substantial interest in both research and clinical settings.For example, our laboratory and others have leveraged surface electromyography (sEMG) and surface mechanomyography (sMMG) to investigate the motor control strategies that underlie human movement under a variety of conditions including fatiguing exercise, supplementation, and to delineate neural contributions facilitating strength changes (Cramer et al 2004, Beck et al 2005, Zak et al 2015, Hill et al 2018a, 2022).Specifically, the amplitude (i.e.time domain) of sEMG reflects an algebraic summation of all the excited motor units in the recording area of the electrodes and can be used to make inferences regarding changes in motor unit recruitment, firing rate, and synchronization (De Luca 1979, Farina et al 2014).Further, the frequency domain of sEMG, which is typically derived as the mean or median power frequency of the power density spectrum, reflects action potential conduction velocity, or the speed at which neural signals propagate along the muscle fibers (Mills 1982).sMMG, however, has been described as the mechanical counterpart of sEMG and captures the resultant oscillations of the activated muscle fibers, a key distinction from sEMG which denotes muscle excitation (Gordon and Holbourn 1948, Orizio et al 2003, Vigotsky et al 2018).Thus, the combined use of sEMG and sMMG provides unique and more complete information regarding excitationcontraction dynamics of human skeletal muscle (Smith et al 2017a(Smith et al , 2017b)).
The devices used to measure sMMG have improved incrementally over the decades (e.g.acoustic myography, piezoelectric sensors, etc), and currently, accelerometers are commonly implemented and exhibit high utility (Orizio et al 2003, Beck et al 2005).These small and lightweight accelerometers can be placed over the muscle bellies of diverse muscle groups (Beck et al 2005).Additionally, some accelerometers are 'triaxial' such that they can measure muscular oscillations in 3 axes (i.e.X, Y, and Z).The sMMG literature, however, has primarily focused on the X-axis (sMMG-X) which quantifies lateral oscillations of the activated muscle fibers (i.e.perpendicular movements of the muscle fibers relative to the skin's surface).The omission of the other two axes, however, may lead to an incomplete understanding of muscle activation.Specifically, sMMG-X amplitude is sensitive to changes in motor unit recruitment and typically increases from 0 to 60%-80% of maximal strength, and the subsequent increases (e.g.80%-100%) in maximal strength are thought to be mediated by changes in motor unit firing rate (Orizio 1993, Orizio et al 2003).During fatiguing exercise, however, sMMG-X amplitude is regularly reported to be affected by changes in muscle compliance whereby limiting its ability to discern changes in motor unit recruitment (Sjøgaard and Saltin 1982, Beck et al 2004a, 2005), especially near task failure.Furthermore, sMMG can be affected by a multitude of other physiological and non-physiological factors including individual differences in recruitment patterns (superficial to deep), pennation angle, muscle thickness, muscle fiber type composition, velocity of movement, and the modality of contraction (e.g.eccentric, concentric, and isometric) (Cramer et al 2004, Beck et al 2005, Hill et al 2016, Trevino et al 2023).Thus, these inherent limitations must be considered when interpreting sMMG-based responses as they relate to motor unit activation strategies apart from other physiological and non-physiological changes.
There is current rationale to suggest that the Y (longitudinal) and Z (transverse) axes may provide complementary and/or additional information regarding physiological and non-physiological responses assessed via sMMG.For example, other triaxial accelerometers (e.g.physical activity accelerometers) have been applied as a diagnostic tool to track daily movement among clinical populations (Steele et al 2000, Hurd et al 2013).Like physical activity tracking of human movement, muscle fiber perturbations oscillate in three axes.For example, the amplitude of sMMG-X quantifies lateral oscillations and is proportional to power output of the muscle (Bodor 1999, Cramer et al 2004, Hill et al 2017).Specifically, it was theorized that sMMG amplitude is analogous to a violin string whereby the loudness of the vibrating string was a function of the power applied by the bow (Bodor 1999).In theory, shortening the violin string (e.g. using a capo) while maintaining the power applied to the bow would decrease the amplitude (i.e.loudness), but increase the frequency (i.e.shortened wavelength, increased pitch).In vivo, this may manifest in muscle with the application of blood flow restriction (BFR) which is applied to the proximal portion of an exercising limb or nearest the origin of the muscle(s).The application of BFR is common in the scientific literature as a potential adjunct to exercise which stimulates robust physiological adaptations when paired with resistance exercise (Hill et al 2018a).The cuffs used to induce BFR, which vary in width, may theoretically act like a capo does to a violin or guitar string and dampen the amplitude of the activated muscle fibers by 'shortening' the distance between the origin and insertion.To clarify, during BFR exercise, the required cuff may indeed limit the lateral oscillations of the activated muscle fibers directly beneath it, while also attenuating oscillations downstream which are perpetuated from a 'shorter' muscle fiber.The resultant muscular swelling response (e.g.blood pooling) augmented by BFR, however, will likely not affect sMMG amplitude as intramuscular fluid pressure has previously been disassociated from sMMG amplitude (Søgaard et al 2006).Therefore, applying BFR provides a non-invasive way to determine the sensitivity of sMMG derived information in the X, Y, and Z-axes to examine musculoskeletal changes.Additionally, to-date, there are only a few investigations (Hill et al 2018a(Hill et al , 2019) ) that have reported sMMG responses during BFR exercise, and notably, these investigations only examined the X-axis among female populations.These investigations are encouraging as relative to males, females are underrepresented in exercise science and BFR research (Counts et al 2018).This is a critical current gap in the relevant scientific literature.Therefore, the purpose of this investigation was to examine the acute effects of submaximal, fatiguing exercise with and without BFR on sMMG amplitude in the X, Y, and Z-axes among female participants.Addressing this purpose will uncover potentially valuable metrics previously ignored during X-axis exclusive analyses.

Participants
Sixteen females (n = 16; mean age ± SD = 21 ± 1 years; body mass = 65.4 ± 9.2 kg; height = 164.8± 7.4 cm) who were recreationally active (at least 150 min of exercise per week) volunteered to participate in this study.Participants had no known cardiovascular, pulmonary, metabolic, muscular, and/or coronary heart disease and were excluded if they were using prescription medication or ergogenic aids.This study was approved by the University of Central Florida Institutional Review Board (STUDY00004665).The research was conducted in accordance with the principles embodied in the Declaration of Helsinki and in accordance with local statutory requirements.All participants provided their written informed consent to participate in the study and consent was given for publication by all participants.Additionally, all participants completed a medical and activity history questionnaire form prior to testing.

Experimental approach
A randomized, counterbalanced design was used for this study.The female participants visited the laboratory on two separate visits (familiarization & experimental visit), which were separated by at least 48 h.On the second laboratory visit, participants performed three MVICs (separated by 90 s each) to determine their submaximal exercise load.Then participants performed two experimental protocols, which were separated by 15 min of rest.Each protocol consisted of fatiguing unilateral, submaximal (50% of maximal voluntary isometric contraction [MVIC]), isometric muscle actions of the leg extensors with their dominant or non-dominant leg.The submaximal fatiguing protocol was performed with or without BFR to volitional exhaustion which was defined as three consecutive failed attempts to sustain at least 50% of MVIC torque for five seconds.After a 15 min rest period, the protocol was repeated using the opposite leg (dominant or non-dominant) and the other condition (with or without BFR).All muscle actions were performed at a knee-joint angle of 90°(where 180°represents full knee extension) and all testing was performed on a calibrated isokinetic dynamometer (Biodex 3, Biodex Medical Systems, Inc., Shirley, NY).

Procedures Familiarization (visit 1)
The first laboratory visit consisted of an orientation session to familiarize participants with the testing protocols.During familiarization, participants' height and weight were measured using a scale and stadiometer (Health o meter professional, Sunbeam Products, Boca Raton, FL).For the application of BFR, total arterial occlusion pressure was calculated as the lowest pressure needed to fully occlude venous and arterial blood flow as observed by a lack of downstream blood flow via ultrasound.Participants then performed a self-paced five-minute warmup on a stationary cycle ergometer (Corival | cpet, Lode, Groningen, The Netherlands).Following a brief rest period, participants practiced maximal and submaximal muscle actions with their dominant and non-dominant legs until they felt comfortable with the testing procedures.

Experimental visit (visit 2)
Maximal strength testing and submaximal fatiguing protocol During the second laboratory visit, participants performed a five-minute warm-up on a stationary cycle ergometer at a self-selected pace and resistance.Following a brief rest period, participants performed three, fivesecond, unilateral MVICs of the leg extensors that were separated by 90 s of rest, and each were performed at a knee joint angle of 90°on a calibrated isokinetic dynamometer.For each unilateral MVIC muscle action, the participants were instructed to reach maximal force as rapidly as possible and maintain maximal force until instructed to relax.Each participant reached maximal force within 1 s, sustained maximal force for approximately 4 s, and returned to baseline force output within 1-2 s.The highest force produced among the three trials was used to determine the submaximal exercise load for the fatiguing protocol.After a rest period of two minutes, participants completed a submaximal fatiguing protocol that consisted of unilateral, intermittent, isometric muscle actions performed at 50% of MVIC which was displayed on a computer to provide real-time visual feedback.Each submaximal muscle action was sustained for a period of five seconds followed by three seconds of rest until volitional exhaustion was achieved (three consecutive failed attempts to sustain at least 50% of MVIC torque for five seconds).Following a 15 min rest period, the unilateral MVICs and submaximal fatiguing protocol were repeated on the other leg (dominant or non-dominant) with the other condition (i.e. with or without BFR).Both leg and condition were randomized for each participant.All maximal and submaximal muscle actions were completed on the calibrated isokinetic dynamometer described previously.

Blood flow restriction
Prior to the use of BFR, total arterial occlusion pressure of the femoral artery was determined for each participant using brightness mode (B-mode) on a portable ultrasound imaging device (GE logiq e NextGen R7, Chicago, IL) and an attached multi-frequency linear-array probe (GE L4-12T-RS; 4.2-13 MHz; 38.4 mm fieldof-view).Specifically, participants rested quietly on a cushioned table in the supine position for a period of 5 min prior to determining total arterial occlusion pressure.Thereafter, a deflated 11 cm wide cuff was placed most proximally on the participants' dominant or non-dominant leg.The leg used for BFR condition was randomized for each participant.Using a rapid cuff inflator (Hokanson E20 Rapid Inflator, Bellevue, WA), cuff pressure was initially applied at 25 mmHg and intermittently and progressively (1-25 mmHg) inflated (2-5 s) and deflated (5-10 s) until cessation of blood flow through the downstream posterior tibial artery was observed via ultrasonography.This value was recorded and 60% of this value was used for each cuff inflation during maximal and submaximal testing.During the submaximal fatiguing protocol performed with BFR, the BFR cuff was inflated immediately prior to the first repetition and deflated immediately following volitional exhaustion.

Mechanomyographic assessments
A triaxial accelerometer (356A32, PCB Piezotronics, Depew, NY) with a sensitivity of 10.2 mV/m•s −2 was placed on the rectus femoris muscles of the dominant and non-dominant legs away from the innervation zone at approximately 50% of the distance between the medial border of the patella and anterior superior iliac spine (Barbero et al 2012).Prior to placement, the skin was shaved and cleaned with alcohol to ensure the accelerometer would remain affixed to the leg with double-side adhesive tape (B-205-1, Factor II, Lakeside, AZ) The raw sMMG signals were digitized at 2000 Hz with a 32-bit analog-to-digital converter (Model MP150, Biopac Systems, Inc., Goleta, CA) and stored in a personal computer (ATIV Book 9 Intel Core i7 Samsung Inc., Dallas, TX) for subsequent analyses.The sMMG signals were digitally bandpass filtered (fourth-order Butterworth, zero-phase shift) at 5-100 Hz.All signal processing was performed in LabVIEW (v 22.0; National Instruments, Austin, TX) using custom written programs.The sMMG amplitudes in the X, Y, and Z-axes of MMG (m•s −2 ) were calculated over the middle third of each contraction.Thus, signal epochs of 1.67 s (approximately 3340 data points) were used to calculate the amplitude values of the sMMG signals during the submaximal and maximal isometric muscle actions.The subsequent sMMG amplitude voltage values were converted into m•s −2 (absolute values) and were also separately normalized to pretest MVIC (% of pretest MVIC).
Polynomial regression analyses (first, second, and third order) were used to examine the individual and composite patterns of responses for normalized (to pretest MVIC) sMMG amplitude in the X, Y, and Z-axes during the submaximal fatiguing protocols performed with BFR and without BFR (non-BFR).The F-test was used to assess if the increment in the proportion of variance accounted for by a higher-order polynomial was significant (i.e.R 2 change).These regression analyses were also performed across normalized time points expressed as %TTE, but in 10% increments.Rationale for the additional timepoints includes not restricting the probability of a 2nd or 3rd order polynomial relationship, which may have been attenuated by fewer data points (more likely to be linear).Conversely, for the ANOVA-based analyses, to maintain sufficient power and to apply conservative post-hoc correction factors (i.e.Bonferroni), %TTE was limited to five levels within the time factor.Further, the polynomial regression analyses were provided for descriptive purposes, but were not statistically evaluated or compared (e.g.slope or y-intercept comparisons).Therefore, 10 levels for the polynomial regression analyses enabled better resolution of the patterns of response, whereas five levels for the ANOVAbased analyses provided specificity.All statistical analyses were performed using IBM SPSS versus 29 (Armonk, NY) and an alpha of p 0.05 was considered statistically significant for all regression analyses.

Individual patterns of responses
For the individual BFR responses, there were 4 linear, 3 quadratic, and 9 non-significant relationships among the normalized (% of pretest MVIC) sMMG-X amplitude versus %TTE responses.For the individual non-BFR responses, there were 6 linear, 1 quadratic, and 9 non-significant relationships among the normalized sMMG-X amplitude versus %TTE responses.For the composite responses (average across participants), there was a nonsignificant sMMG-X amplitude versus %TTE relationship for BFR (r 2 = 0.367) and a cubic sMMG-X amplitude versus %TTE relationship for non-BFR (R 2 = 0.955) (table 1).

Individual patterns of responses
For the individual BFR responses, there were 4 linear, 2 quadratic, and 10 non-significant relationships among the normalized (% of pretest MVIC) sMMG-Y amplitude versus %TTE responses.For the individual non-BFR responses, there were 5 linear, 2 quadratic, 1 cubic, and 8 non-significant relationships among the normalized sMMG-Y amplitude versus %TTE responses.For the composite responses, there was a cubic sMMG-Y amplitude versus %TTE relationship for BFR (R 2 = 0.929) and a non-significant sMMG-Y amplitude versus % TTE relationship for non-BFR (r 2 = 0.247) (table 1).

Discussion
The purpose of this investigation was to compare the acute effects of submaximal, fatiguing exercise (with and without BFR) on sMMG amplitude in the X, Y, and Z-axes.Indeed, here we demonstrated for the first time that the X, Y, and Z-axes provided unique information that was evidenced by different individual patterns of responses (i.e.linear, quadratic, cubic, and non-significant).This was further supported by the observed mean differences between the BFR and non-BFR conditions.Specifically, in both the X and Y-axes, sMMG amplitude was attenuated with the application of BFR, while the Z-axis remained unaffected.Thus, the application of BFR restricted lateral and longitudinal, but interestingly, did not affect the transverse oscillations.These differences, however, were not as prevalent when the sMMG amplitude responses were normalized to pretest MVIC.Collectively, there are differences in the patterns of responses and mean changes among axes which may underly physiological and/or non-physiological effects of fatiguing exercise.

sMMG-X amplitude (lateral oscillations)
The individual sMMG-X amplitude versus %TTE relationships were highly variable between the BFR and non-BFR fatiguing exercise conditions, and thus, limited our ability to generalize condition-specific differences among the patterns of responses.However, regardless of timepoint and consistent with our hypothesis, sMMG-X amplitude was attenuated with the application of BFR (0.236 ± 0.138 m•s −2 ) relative to the non-BFR (0.366 ± 0.199 m•s −2 ) condition.Normalizing the absolute (m•s −2 ) sMMG-X amplitude responses to pretest MVIC, however, eliminated the attenuation of sMMG between BFR (74.1 ± 43.6%) and non-BFR (92.4 ± 48.4%) conditions (figures 1(a), (b)).Perhaps, this is yet more rationale to normalize neuromuscular data (e.g.sMMG) to corresponding MVIC trials.
In the present study, both BFR and non-BFR conditions remained relatively unchanged during the initial 20%-60% of TTE, whereby both increased (although not significantly) thereafter.The sMMG-X amplitude increases after 60% of TTE were likely the result of a fatigue-induced increase in motor unit recruitment which occurred at 80% and 100% of TTE for both conditions (Rodriquez et al 1993).Typically, exercise performed with BFR induces muscle fatigue more rapidly than non-BFR exercise (Cook et al 2013), which was again demonstrated by the present study with fewer repetitions required to achieve volitional exhaustion for BFR than non-BFR (20 ± 7 versus 26 ± 14 repetitions, respectively).Our sMMG amplitude responses, however, were examined across normalized time (i.e.%TTE) removing potential differences in the rate of fatigue between conditions as our primary aim was to explore the effects of BFR on sMMG responses, not differences in the rate of fatigue.
During fatiguing exercise, sMMG amplitude may increase, decrease, or remain unchanged depending on the exercise load/intensity (Orizio et al 1989, Goldenberg et al 1991, Rodriquez et al 1993).For example, across maximal fatiguing exercise, sMMG amplitude progressively decreases possibly due to the dropout of fatigable fast-twitch motor units and/or reductions in muscle compliance (Sjøgaard and Saltin 1982, Beck et al 2004b, Camic et al 2013, Hill et al 2018b).During submaximal to moderate load fatiguing exercise, sMMG amplitude may increase or remained unchanged depending on the relative contributions between physiological (e.g.motor unit recruitment/de-recruitment) and non-physiological (e.g.muscle compliance, temperature) factors (Sjøgaard and Saltin 1982, Orizio et al 1989, Bodor 1999, Cramer et al 2004, Beck et al 2005, Trevino et al 2023).
In the present study, this was consistent with some of the individual sMMG amplitude versus %TTE relationships as well as the composite response for the non-BFR condition which exhibited an increase.Furthermore, the mean responses appear to increase at 80% and 100% TTE, although not significantly, so caution is warranted when interpreting a perceived increase.In contrast, the majority of the individual relationships, the composite BFR response, and the statistical outcomes comparing the mean responses were not different across %TTE.Additionally, the large variability among the patterns of responses may reflect individual differences in motor unit recruitment versus maximal strength relationships and/or the effects of motor control strategies (e.g.piper rhythm, muscle wisdom, motor unit synchronization) which have been theorized to affect sMMG amplitude within the X-axis (Orizio et al 1989, Bodor 1999, Beck et al 2005, Camic et al 2013).Collectively, the sMMG amplitude versus %TTE relationships were similar for BFR and non-BFR, while the non-normalized sMMG amplitude responses were attenuated with the application of BFR.

sMMG-Y amplitude (longitudinal oscillations)
The Y-axis is unique from the X-axis as it measures the longitudinal oscillations of the activated skeletal muscle fibers or muscle fiber movement that occurs in-line with the pull of the muscle (i.e.parallel to the muscle fibers).Like sMMG-X, there was considerable inter-subject variability among the individual sMMG-Y amplitude versus %TTE relationships that, in general, were non-significant for both the BFR and non-BFR conditions.Additionally, like sMMG-X, from 20%-60% of TTE sMMG-Y amplitude remained relatively unchanged and increased (although not significantly) at 80% and 100% of TTE (figure 2(a)).Similarly, like sMMG-X, absolute (m•s −2 ) sMMG-Y amplitude was attenuated at 60, 80, and 100% of TTE for BFR (0.213 ± 0.068, 0.212 ± 0.079, 0.232 ± 0.115 m•s −2 , respectively) relative to non-BFR (0.313 ± 0.118, 0.373 ± 0.208, 0.445 ± 0.353 m•s −2 , respectively) conditions.Unlike sMMG-X, for BFR, absolute sMMG-Y amplitude was lower at 60% (0.213 ± 0.068 m•s −2 ) of TTE than 40% (0.256 ± 0.084 m•s −2 ) TTE (figure 2(b)).There were no changes across time for the non-BFR condition which may be due, in part, to greater variability among mean responses which were not attenuated by the application of BFR (figure 2(b)).
Normalizing sMMG-Y amplitude removed some but not all of the differences between BFR and non-BFR conditions.Specifically, after normalizing the sMMG-Y amplitude, there were no differences at 60% or 80% of TTE.For BFR, both sMMG-Y amplitude at 20% (88.4 ± 51.8%) and 40% (82.7 ± 58.7%) of TTE were greater than 60% (68.3 ± 44.0%) of TTE.Therefore, like sMMG-X, normalizing the signal eliminated some of the attenuation of sMMG-Y amplitude, while differences in sMMG-Y amplitude across %TTE persisted for BFR.Collectively, in the present study, sMMG amplitude in the Y-axis is similar to the X-axis with marginal differences.

sMMG-Z amplitude (transverse oscillations)
The Z-axis measures muscle oscillations in the transverse axis and produced similar sMMG amplitude versus % TTE relationships with no between condition differences.Specifically, across the fatiguing exercise bouts, regardless of condition, sMMG-Z amplitude decreased across %TTE reaching a nadir around 80% and 100% of TTE (figures 3(a), (b)).Moreover, the absolute (m•s −2 ) and normalized (% of pretest MVIC) sMMG-Z amplitude responses were similar suggesting sMMG-Z amplitude responses may be less influenced by factors known to adversely affect sMMG-X interpretation (Cramer et al 2004, Beck et al 2005, Hill et al 2016, Trevino et al 2023).Future studies may consider reporting Z-axis results alongside the X-axis derived values.To be clear, however, the physiological and non-physiological effects of exercise on sMMG-Z amplitude have not been identified, but it is likely dissociated from motor unit recruitment.Instead, the individual decreases in sMMG-Z amplitude versus %TTE relationships and mean responses decreases suggested there was a reduction in transverse movement of the muscle fiber(s).There are numerous factors which have been attributed to a reduction of sMMG amplitude in the X-axis (e.g.muscle stiffness, changes in elasticity, dropout of motor units) but the underlying mechanism mediating the observed decreases in sMMG-Z amplitude are purely speculative.Regardless, the current findings suggested that the Z-axis may provide a non-invasive way to track muscular fatigue similar to the application of sEMG frequency which tracks fatigue-induced reductions in action potential conduction velocity (Lindstrom et al 1970).For example, it is possible that reductions in transverse movement of the muscle fibers track the progression of muscle fatigue during exercise bouts.Future studies are needed to verify the utility of sMMG-Z amplitude (and sMMG-Y) as tool(s) to delineate physiological and/or nonphysiological outcomes which may be accomplished through incremental muscle contractions performed at various percentages of maximal strength.

Summary
The application of BFR attenuated sMMG amplitude in the X and Y-axes, but not in the Z-axis.Normalizing the sMMG amplitude responses removed some, but not all of the attenuation associated with BFR which was still partially evident in the Y-axis.A strength of our current investigation includes that our sMMG amplitude responses in the X-axis were consistent with the effects of fatiguing submaximal exercise performed at a moderate load.The sMMG amplitude responses of the Y-axis were largely similar to those of the X-axis perhaps providing additional informative information.The Z-axis was robust to the application of BFR and exhibited similar absolute and normalized patterns of responses.Specifically, for both BFR and non-BFR, sMMG-Z amplitude decreased across %TTE.Thus, sMMG-Z amplitude is likely dissociated from motor unit recruitment but may track other physiologically relevant phenomenon such as the progression of muscle fatigue during exercise bouts.It is plausible these axis-specific responses persist among other muscle(s) or muscle groups and between biological sex.Practically, our findings are similar to other applications of sMMG whereby normalizing the signal to a pretest maximal contraction may control or attenuate the influence that some non-physiological factors (e.g.applying BFR) exhibit on sMMG responses across axes.Collectively, further studies are needed to verify the utility of sMMG amplitude in the Y and Z-axes to delineate physiological and/or non-physiological outcomes which may be accomplished through incremental muscle contractions performed at various percentages of maximal strength.

Figure 1 .
Figure 1.(a), (b) displays the mean ± SE absolute (m•s −2 ) and normalized (% of pretest maximal voluntary isometric contraction [MVIC]) surface mechanomyographic responses in the X-axis (sMMG-X).sMMG-X amplitude was assessed across percent time to exhaustion (TTE) during intermittent, isometric, leg extension muscle actions performed at 50% of MVIC with blood flow restriction (BFR-filled squares) and without BFR (non-BFR-filled circles).Marginal mean responses for condition, collapsed across time, are depicted as hollow square and hollow circle for BFR and non-BFR, respectively.* p < 0.05 main effect for condition (non-BFR > BFR).

Figure 2 .
Figure 2. (a), (b) displays the mean ± SE absolute (m•s −2 ) and normalized (% of pretest maximal voluntary isometric contraction [MVIC]) surface mechanomyographic responses in the Y-axis (sMMG-Y).sMMG-Y amplitude was assessed across percent time to exhaustion (TTE) during intermittent, isometric, leg extension muscle actions performed at 50% of MVIC with blood flow restriction (BFR-filled squares) and without BFR (non-BFR-filled circles).†p < 0.05 simple main effect among time at a level of condition * p < 0.05 simple main effect between condition at a level of time.

Table 1 .
The individual and composite results for the polynomial regression analyses, slope coefficients, coefficients of determination (r 2 /R 2 ) of the normalized (to maximal voluntary isometric contraction[MVIC]) surface mechanomyographic (sMMG) amplitude (X, Y, Zaxes) versus percent time to exhaustion (%TTE) relationships.sMMG amplitude was assessed during the submaximal (50% of MVIC), isometric leg extension muscle actions performed with blood flow restriction (BFR) and without BFR.The sMMG amplitude regression analyses were examined across %TTE in 10% increments.