Neuromuscular Function and Blood Flow Occlusion with Dynamic Arm Flexor Contractions

Abstract

Introduction 

Blood flow-restricted or occlusion exercise enhances muscle hypertrophy and strength during resistance training. The acute effects on voluntary and electrically evoked muscle contractile characteristics with impaired blood flow at low- and high-contraction forces have not been explored.

Methods 

On separate days, nine males completed two different protocols of concentric elbow flexor contractions. A repetitive low-force (~25% of isometric maximum voluntary contraction [MVC]) with blood flow occlusion (BFO) (300 mm Hg) protocol was compared with a high-force (~80% MVC) free blood flow protocol (HF), until range of motion (0°–90°) was impaired. Torque, velocity, and power were compared with baseline and between protocols. Maximum voluntary contraction and voluntary activation were assessed during and after each protocol. Muscle twitch, low (20 Hz) and high (50 Hz) tetanus, and compound muscle action potential (M_max) area were measured at 0, 2, 5, 10, and 20 min of recovery.

Results 

Repetitions to failure (FP) were lower for HF (~16) versus BFO (~21), and MVC at FP was reduced more during BFO (~77%) compared with HF (~23%), with no difference in voluntary activation (~10% loss) between protocols. At FP, velocities for BFO and HF were similarly reduced by approximately 63% and 56%, respectively; however, peak power decreased more during BFO (~90%) compared with HF (~67%). Total work for BFO was approximately 40% lower than for HF. Peak power for HF was recovered by 2 min, whereas BFO required 20 min. Low-frequency fatigue (20 Hz/50 Hz) was greater after BFO (~70% decrease vs ~29% decrease after HF), whereas M_max area was unaffected.

Conclusion 

Concentric elbow flexions at low-force with BFO cause greater impairments in strength and power than HF and, therefore, may be a replacement for high-force exercise used in chronic training.

It is generally accepted that resistance training with contractile forces of ≥70% of maximum voluntary contraction (MVC) are required to stimulate muscle hypertrophy and strength gains (^1,2). However, chronic training with blood flow-restricted (BFR) exercise, sometimes referred to as occlusion training, has shown strength gains can be accomplished at low-forces (^3–5). Although BFR exercise has been studied in relation to chronic adaptations of the muscle, the acute effects of blood flow occlusion (BFO) that may promote longer-term muscular adaptations are not well-understood. The benefit of low-force exercise is that it can be performed at high velocity, and the addition of BFR is unlikely to reduce the peak velocity of nonfatiguing contractions.

The number of repetitions completed during a training protocol is less important to induce chronic long-term changes in hypertrophy and strength than doing repetitions-to-failure which likely causes greater metabolic stress (^6,7). Furthermore, repetitions-to-failure (fatigue) induce significant decreases in contractile velocity (⁶). Because power output of a muscle fiber is determined by the product of force and velocity (⁸), the ability to maintain a high velocity during repetitions with low-force will cause greater impairments for power generating movements when completed to failure, compared with high-force low velocity movements. Indeed, velocity and power are improved to a greater amount after chronic exercise training for those that use a faster cadence compared with training at a slow cadence (^9,10). Low-force contractions with BFO performed at greater velocities than those possible with higher loads may therefore be more beneficial at increasing muscular adaptations.

Neuromuscular fatigue has both voluntary and involuntary components, the involuntary (peripheral) or muscle component of the neuromuscular system can be assessed from electrically evoked tetanic contractions of the muscle, which does not involve spinal or supraspinal efferent input. During prolonged exercise that involves a high number of repetitions of a moderate to high force, the muscle fibers generate less force and contract slower (fatigue), often due to muscle damage (¹¹) or excitation–contraction (E-C) failure (¹²). The lack of oxygenated blood flow caused by BFR, likely facilitates the recruitment of type II fibers that are not oxygen dependent (⁸), causing greater type II fatigue, damage, and E-C coupling failure. Excitation-contraction coupling failure is one of a number of possible causes of fatigue (^13,14) and can be evaluated indirectly after fatiguing contractions by comparing the response of the muscle at lower frequencies of tetanic stimulation (i.e., ≤20 Hz) to the response from maximal frequencies of excitation (i.e., 50 Hz), (^14,15). The main features of low-frequency fatigue (LFF) are that forces at low-frequency of stimulation are more impaired than those generated by higher stimulation frequencies, and impairment may last hours or days (¹³). A greater reduction in the 20/50 Hz ratio after fatiguing contractions is an indication of peripheral fatigue. We observed that low-force contractions with BFO caused greater peripheral fatigue (assessed by LFF) when compared with a low-force blood free flow protocol (¹⁶).

Central components of fatigue are more difficult to evaluate because they include assessing the degree of voluntary drive generated from spinal and supraspinal factors. Voluntary activation (VA) of the system can be indirectly assessed using the interpolated twitch technique (ITT) (¹⁷). However, very few studies with BFR exercise have explored central and peripheral factors. In one study, after five sets of 20 dynamic fatiguing submaximal contractions at 20% of one-repetition-max with BFR, VA was decreased by 13% (¹⁸) and maximal surface EMG, as a measure of voluntary neural activation, was reduced by 12%, compared with no change in VA and maximal EMG amplitude during an MVC for a control group that performed the same protocol without BFR. In this same study the submaximal EMG increased during each consecutive set (¹⁸) in agreement with other studies reported in both chronic and acute use of BFR (^5,19,20). Overall, these BFR findings were interpreted as an increase in central (VA) and peripheral (EMG) fatigue after exercise (¹⁸), however, these findings were based on protocols that did not reach failure. Failure would likely cause decreases in VA and increases in submaximal EMG.

The purpose of the current study was to use an acute dynamic isotonic fatiguing task to compare the effects on neuromuscular properties (velocity and power) of a low-force exercise with BFO with a high-force exercise bout without any imposed blood flow resistance or occlusion. A study by Yasuda et al. (2009) performed bouts of exercise with moderate and complete occlusion, and found that complete occlusion prevents arterial supply of metabolic substrates and venous return. Although the use of BFO is unlikely used chronically as an exercise intervention, it is a good model to explore BFR by enhancing fatigue during dynamic contractions to provide functional insight into the effects of restricted blood flow rather than with isometric contractions previously studied (^16,21). Thus, as a first comparison using an acute exercise bout and with specific peripheral and central measures, we chose to contrast free blood flow with an occluded rather than a restricted state. From this, insights may be gained about sites of adaptations that are affected during chronic BFR training and whether low-force BFR exercise is a viable training modality for muscular improvements, especially for those with functional limitations such as due to disease, rehabilitation or aging, as compared with the usual type of high-force exercise training.

Therefore we hypothesized, H1: that acute BFO exercise will induce greater decreases in velocity, and therefore power impairments to a similar or greater amount than HF exercise, for the same number of repetitions, and H2: that acute low-force BFO will cause greater peripheral fatigue compared with the unrestricted blood flow during high-force exercise. Maximal velocity contractions were used to assess peak velocity and power, whereas electrically evoked tetanic responses were used to specifically target intrinsic (peripheral) changes in the muscle. Voluntary activation in combination with repetitions-to-failure and recording surface EMG were used as measures to assess neural activation and muscle excitability.

METHODS

Subjects

Nine healthy male subjects (see Table 1 for characteristics) participated in two testing protocols (HF and BFO) administered in a random order, and each was separated by at least 48 h, with both protocols completed within 7 d. All procedures were approved by the local institutional ethics committee (REB 107212, at The University of Western Ontario) and conformed to the declaration of Helsinki. Written and verbal consent were obtained from each participant. All subjects were right-hand dominant, and therefore, to minimize limb dominance effects, the nondominant left arm was tested in each participant.

TABLE 1: Values are means ± SD.

Experimental setup

Participants were seated upright in a Humac Norm dynamometer (Computer Sports Medicine Inc., Stoughton, MA) with feet resting on an adjustable foot rest placing their hip and knee joints at approximately 90°. To record elbow flexion, the left arm was in the dependent position, and the elbow joint allowed to move freely between 0° and 90° with the supinated wrist holding a hand grip in a comfortable position (wrist pronated ~5°). The axis of rotation of the Humac norm was adjusted to the elbow joint center of rotation, and lever arm with hand grip was adjusted for all participants. A snug three-point harness was adjusted around the torso for each participant to eliminate extraneous body movement, as well as a large inelastic strap was fastened securely across the chest to secure the participant to the chair. Surface EMG of the elbow flexors was recorded via adhesive Ag-AgCl electrodes (Kendall, H59P cloth electrodes) arranged in a monopolar fashion. One electrode was placed on the skin over the mid belly of the muscle, and a reference electrode was secured over the ulna on the posterior forearm. To induce tetanic elbow flexor activation custom-made aluminum foil electrode pads (~2 × 5 cm) covered in damp paper towel were placed over the distal and proximal portions of the arm flexors. Electrical stimulation was delivered from a Digitimer (model DS7AH) stimulator.

In all sessions, torque and EMG data were recorded using an A/D converter (CED 1401; Cambridge Electronic Design Ltd, Cambridge, UK) in conjunction with Spike2 software (v. 7.02; Cambridge Electronic Design). The torque and EMG data were sampled at 500 and 5000 Hz, respectively. EMG data were amplified (×1000) and bandpass filtered (10 Hz to 5 kHz, with a 60-Hz notch filter) using Neurolog; NL844, Digitimer, Welwyn Garden City, UK.

Experimental protocol

Participants were randomly assigned to complete either first the low-force (25% of MVC) with BFO to repetition failure protocol, or the high-force (80% of MVC) unrestricted blood flow (HF) to repetition failure protocol. Except for the addition of the blood pressure cuff to restrict blood flow all procedures were identical in both protocols. The blood pressure cuff (adult sized, sphygmomanometer) was applied over the proximal portion of the biceps brachii, and pressure was maintained at 300 mm Hg during the BFO protocol for all participants. This pressure was sufficient to abolish the forearm radial pulse during the resting state for all participants. To elicit an M-wave in the biceps brachii, a single electrical stimuli (200 μs pulse width at ≤400 V) was delivered to the brachial plexus at Erb’s point. The cathode and anode were placed in the supraclavicular fossa and over the acromion, respectively. Stimulus current was increased incrementally for successive stimuli until the peak-to-peak amplitude of the resting M-wave reached a plateau (M_max). Stimulus intensity was then set to 120% of the current required to produce M_max. The stimulation parameters remained constant pre- and postintervention protocol. To determine VA during the MVC of the elbow flexors maximal electrical stimulation doublets (200 μs pulse width; 400 V; 100 Hz doublet; range 90–168 mA) by increasing current intensity until the torque response no longer increased with an increase in current intensity, or coactivation of other muscles impeded the elbow flexor torque. For the interpolated twitch technique (ITT) doublets were used to assess VA (¹⁷) in the elbow flexors. Participants were instructed to perform two to three brief (~3–5 s) elbow flexor MVC, which included a superimposed doublet at the peak torque, and a doublet was applied at rest immediately after the MVC. If variability in maximal torque was 5% or more between MVC, then a third MVC was performed. Two minutes of rest was given between each MVC. Strong verbal encouragement and visual feedback were provided during all voluntary contractions. The greatest elbow flexion MVC was selected as the baseline value. Tetanic 50-Hz (200 μs pulse width; 400 V; 1 s duration; range, 35–90 mA) stimulation was applied using custom made aluminum foil stimulation pads (as previously described), for 1 s to elicit 30% of elbow flexor MVC torque which did not activate antagonist muscles of the arm. Evoked tetanic responses at 30% of MVC torque were selected based upon subject tolerance determined by pilot testing, and standardized at 30% for all participants in the experiment. This method is not uncommon (²²) and is based on activating a representative large fraction of the muscle and that it is unlikely that this fraction if equated to a percent of MVC is changed substantially between and during the protocols. Stimulations at 1 Hz (twitch) and 20 Hz were applied using this same intensity as for the 50 Hz. Three dynamic maximum velocity contractions were performed to calculate peak velocity. For the BFO protocol a 25% MVC load was used to measure velocity, whereas during the HF protocols, both a 25% load and an 80% load were used to calculate velocity reductions during HF repetitions. For the HF protocol, 80% MVC load was used to normalize the changes in velocity throughout the protocol; however, following both protocols (HF and BFO) during recovery, a 25% MVC load was used to compare the recovery of velocity and power.

After baseline measures were acquired (at the beginning of each testing day), participants completed one of the two intervention protocols. Participants were instructed to complete as many dynamic repetitions as possible for both protocols, whereas elbow flexor MVC (~3–5 s) were assessed at the completion of every two repetitions, and the ITT was assessed at the estimated (from pilot testing) 50% of repetition-to-failure point. Failure was defined as the point at which the participants were unable to complete a repetition through the full range of motion (0°–90°). During a 20-min recovery period at each time point (0, 2, 5, 10, and 20 min) participants completed in order three maximal velocity contractions (25% load for both protocols), the isometric M_max, 1-, 20-, and 50-Hz responses followed by an MVC with ITT. Parameters were assessed immediately after the fatiguing task (at failure point, FP; and ~2–3 s after failure [R0]), which accounted for the time needed to remove the blood pressure cuff and during the recovery period.

Data and statistical analyses

Off-line quantification of measures consisted of peak MVC torque, M_max area, RMS EMG, peak dynamic velocity at 25% MVC load (for HF protocol velocity at 80% MVC load was used to normalize reductions in velocity and power), peak power (peak dynamic torque × peak velocity), peak twitch torque (PT), time-to-peak torque (TPT), and half-relaxation time (HRT) for the 1-Hz stimulation, and PT and HRT for the 50- and 20-Hz stimulations. Voluntary activation was calculated using the interpolated twitch equation: 1 − (superimposed/potentiated twitch) × 100 (¹¹). The superimposed twitch refers to the doublet stimulation applied at the MVC peak, and the potentiated twitch refers to the doublet stimulation applied after the MVC at rest. The average amount of work performed in each protocol was calculated as W = TD, where T is the torque (N·m) performed in each protocol, and D is the distance the moment arm travelled degrees. Data are described as mean ± SD in all figures. A two-way repeated-measures ANOVA (within-within) design was used to establish differences resulting from between-protocol comparisons of time and protocol for MVC torque, VA, M_max area, peak dynamic velocity at 25% and 80% MVC load, peak power, PT, TPT, and HRT for the 1-Hz stimulation, and PT and HRT for the 50- and 20-Hz stimulations. When an interaction between protocols was observed a post hoc paired t test for pairwise comparison with a Bonferroni correction was used to establish protocol differences at each time point. When only a main effect of time was observed, paired sample t tests were used in conjunction with a Dunnet’s table test for multiple comparisons. Paired t tests were used to compare protocol differences in repetitions-to-failure, work, and MVC. A post hoc power analysis was completed for the peak power results (as one novel and main outcome measure) and indicated that the sample size of 9 was sufficient. All statistical analyses were performed using SPSS version 25. Statistical significance accepted at α < 0.05.

RESULTS

Voluntary characteristics

The number of repetitions required to reach failure (see Table 1.) during BFO (~21 ± 3) was greater than the number during HF (~16 ± 3) (P < 0.01). Baseline MVC was not different between protocols (P > 0.05). There was a main effect of time (P < 0.01) and protocol for MVC (P < 0.01) which decreased to a greater amount during BFO by approximately 77% at FP compared with a reduction of approximately 23% during HF (see Fig. 1). BFO MVC was significantly reduced compared with HF throughout the normalized repetitions-to-failure, however, immediately after the termination of the BFO protocol, the MVC between the two protocols was not statistically different (R0) (see Fig. 1). Despite the reduction of MVC (~77%) at the end of the BFO protocol, when the blood pressure cuff was removed (~3 s) and blood flow restored, MVC recovered to values that were no longer different between protocols (refer to RO in Fig. 1) However, BFO MVC still remained reduced from baseline for the recovery period (20 min), whereas HF MVC recovered by R2 compared with baseline values. Voluntary activation at baseline was not statistically different (P > 0.05) but was reduced at FP equally in both protocols to approximately 88% of baseline for the BFO protocol and approximately 92% for the HF protocol. At the completion of recovery (20 min), VA for both BFO and HF was not different from baseline at approximately 99% and 98%, respectively, or to each other (data not displayed).

FIGURE 1: Maximal voluntary contraction normalized as repetitions-to-failure displayed as 25% FP for both BFO and HF protocols. Recovery time points from R0 to R20 are in real time. Values represented as percent change from baseline and displayed as means ± SD. †Significant difference between BFO and HF.

Velocity, power, and work

Baseline velocity at a 25% load was not different between protocols (P > 0.05). Velocity showed a main effect of time (P < 0.01), with both protocols becoming significantly reduced by 25% of normalized repetitions-to-failure, and recovered by 5 min (see Fig. 2). Baseline absolute power at low force (25% of MVC load) was not statistically different between protocols (P > 0.05). There was a main effect for time (P < 0.01) and protocol (P < 0.01) for power and an interaction of protocol and time (P < 0.01), such that by 25% of normalized repetitions-to-failure until FP, BFO was reduced by 90% compared with 67% for HF (see Fig. 3). Power in each protocol recovered at a similar rate but HF recovered by 2 min, whereas BFO was not recovered until 20 min. The amount of work (W = T D) completed during the BFO protocol (181 ± 57 J per repetition) was significantly lower than the HF protocol (528 ± 130 J per repetition), with an overall amount of work of 3947 ± 1637 J for BFO and 8199 ± 2276 J for HF (P < 0.01) (see Table 1).

FIGURE 2: Peak velocity values normalized as repetitions-to-failure displayed as 25% FP for both BFO and HF protocols. Recovery time points from R0 to R20 are in real time. Values represented as percent change from baseline and displayed as means ± SD.

FIGURE 3: Relative peak power values normalized as repetitions-to-failure displayed as 25% FP for both BFO and HF protocols. Recovery time points from R0 to R20 in real time. Values represented as percent change from baseline and displayed as means ± SD. †Significant difference between BFO and HF.

Twitch and tetanic properties

Baseline M_max area was not different between protocols (P > 0.05) and remained unchanged throughout both protocols and recovery compared with baseline, M_max at R0 for BFO and HF was nonsignificantly lower at 97.1 ± 25.4; and 97.9% ± 13.2%, respectively. Baseline twitch torque was not different between protocols (BFO, 5.8 ± 2.4 N·m; HF, 3.9 ± 2.5 N·m) (P > 0.05), but showed a main effect during each protocol of time (P < 0.01) and protocol (P = 0.023). The twitch amplitude was significantly decreased at R0 after BFO at approximately 88% from baseline, compared with HI at approximately 51% from baseline (P < 0.01). However, no statistical difference was observed between protocols at R2 or for the remainder of the recovery (recovery data not displayed). Twitch torque at 20 min of recovery remained depressed for BFO at approximately 57% of baseline, but HF was recovered by. The 20-Hz tetanic stimulation had a greater decrease at R0 after BFO (~78% of baseline) compared with HF (~48% of baseline) (P < 0.01). Both protocols had a main effect of time (P < 0.01) and protocol (P = 0.01) and 20 Hz remained depressed from baseline after 20 min of recovery (see Fig. 4A). The 20-Hz HRT showed a main effect of time (P < 0.01) and protocol (P = 0.02), with BFO having a greater increase in HRT at R0 (P = 0.03), compared with HF. The 20-Hz HRT recovered by 5 min (see Fig. 4B). The 50-Hz peak torque was not significantly different between protocols (at RO BFO was reduced to 61.2% ± 11.2%, compared with 69.2% ± 19.5% after HF) but did show a main effect of time (P < 0.01) remaining significantly reduced from baseline by approximately 20% until 10 min of recovery (data not displayed). The 20- to 50-Hz peak torque ratio had a main effect of time (P < 0.01) and protocol (P < 0.01) with the ratio after BFO reduced to approximately 70% of baseline at R0, compared with a reduction in HF of approximately 9% from baseline (P < 0.01) (see Fig. 5).

FIGURE 4: A, Values of 20 Hz torque percent changes during R0 to R20 minutes of recovery compared with baseline. B, Values of 20 Hz half-relaxation time percent change during R0 to R20 minutes of recovery compared with baseline. †Significant difference between BFO and HF protocols. Values displayed as mean ± SD.

FIGURE 5: Represents values of the 20/50 Hz torque ratio as percent changes during R0 to R20 minutes of recovery compared with baseline. †Significant difference between BFO and HF protocols. Values displayed as mean ± SD.

EMG

Maximal EMG for BFO was reduced from baseline beginning at 25% of normalized repetitions-to-failure (P < 0.01), whereas during the HF protocol EMG remained unchanged from baseline values (P > 0.05). At the FP, maximal EMG from the BFO protocol was significantly reduced to approximately 50%, whereas HF remained unchanged at approximately 104% (P < 0.003) (data not displayed). However, by R0 and throughout recovery, maximal EMG was not statistically different from baseline values in both protocols.

DISCUSSION

The results of the current study indicate that a bout of low-force dynamic concentric arm flexor contractions to failure with BFO causes a greater amount of peripheral fatigue, as well as a greater reduction of power compared with a bout of high-force contractions with free (unrestricted) blood flow. Although the number of repetitions required to reach failure was greater for the BFO compared with HF (Table 1), the amount of total Work done was less in BFO compared with the HF (Table 1). Although a greater amount of work was performed for the HF protocol, the relative decrease in peak power was greater for the BFO protocol (~90% from baseline) compared with the HF protocols (~67% from baseline), due to the reduction in velocity combined with the greater reduction in MVC torque for the BFO protocol. Despite the greater amount of work performed for the HF protocol, the BFO protocol caused more neuromuscular impairments as indicated by a greater decline in power and greater LFF. Thus, blood flow occlusion results in greater peripheral fatigue with lower submaximal force contractions than with higher-force contractions when blood flow is not impeded.

Numerous studies have explored the chronic effects of blood flow restriction during isometric contractions, and observed increases in strength after a training program (²¹). However, the acute effects of BFO have not been explored comprehensively to understand the underlying factors. One study by Copithorne and Rice (2019) comparing low-force contractions with and without BFO, observed that a sustained low-force (20% of MVC) isometric contraction of the arm flexors with blood flow occlusion produced greater peripheral fatigue, compared with a low-force sustained contraction with no obstruction to blood flow. Only a few studies have explored the acute effects of BFO or BFR during dynamic contractions (²¹). The current study contrasted the acute effects of BFO at low force in comparison to a high-force protocol with no blood flow restriction. We observed a greater decrease in peak power after BFO (90% reduction from baseline) than the HF protocol (67% reduction from baseline). This larger reduction in peak power with BFO could indicate that fast type II fibers that generate greater velocity and torque (^8,23) are affected more than the slow type I fibers. Conversely, type II fibers during the BFO protocol could be activated for a longer amount of time, as recruitment of type II fibers is enhanced with greater force demand (²⁴). Maximum voluntary contraction during the BFO protocol was also reduced to a greater extent than during HF (~77% and ~23%, respectively); however, this result is expected due to the failure criteria. That is, the HF protocol terminated when participants were unable to produce full range of motion at 80% of MVC, and the BFO protocol terminated when the participants were unable to produce full range of motion at 25% of MVC. Type I fibers that contract more slowly and generate less mechanical power, therefore, require less ATP, whereas type II fibers require more ATP to generate force quickly and to reach the “optimal velocity” for the required movement (²⁴). Furthermore, fibers are not only recruited in order for force development but also in relation to the speed of contraction (^8,23). Therefore, during the BFO condition, the reduction of peak power is likely due to expedited fatigue of type I fibers in a nonoxygenated environment, leading to a greater activation of type II fibers that consume a greater amount of ATP, as compared with the HF protocol which may be able to use type I fibers to produce force for a longer duration. These results are in agreement with single muscle fiber glycogenolysis experiments in which it has been observed that with BFO, glycogenolysis is markedly accelerated in type I muscle fibers, thus leading to increased recruitment of type II muscle fibers (²⁵). This is likely caused by an afferent response to BFO (²⁶). This increased demand placed on type II fibers during BFO causes greater reductions in torque and velocity compared with HF.

For both protocols, VA was reduced from baseline at FP, but was not different between protocols (data not displayed) and returned to baseline in both protocols by the end of the recovery period. Karabulut et al. found that low-force dynamic leg extensions with blood flow restriction resulted in greater voluntary inactivation compared with non–blood flow-restricted control participants. It was proposed that the greater decrease in VA with blood flow restriction may indicate a greater inhibition of central drive to motor units. However, that study was not performed to failure, but rather to a predetermined termination criteria of 20 repetitions. Despite the differences in repetitions to FP between the two protocols in the current study, the criterion for task termination was failure, and thus, VA was similarly affected. The loss of VA of approximately 10% indicates a greater amount of peripheral fatigue than central (supraspinal) when protocols are terminated at voluntary failure. It has been shown previously in studies using a similar fatiguing protocol that in well-motivated participants, the central nervous system remains capable of fully activating (VA) muscle and that the reduction in force results mainly from the failure of the muscle contractile apparatus (²⁷). Although the current study had no measure of pain or relative exertion, a recent study by Martín-Hernández and colleagues (²⁸) reported that RPE was greater after a high-force (85%) exercise model, compared with a low-force (20%) model with blood flow restriction, and with no differences in pain between protocols. They concluded that perceptual response of pain and ratings of exertion likely will not limit the application of BFR in highly motivated individuals. However, the current study did use BFO rather than BFR, but we did not assess pain systematically and, therefore, cannot rule out that it was a factor. However, we had highly motivated subjects familiar with these protocols, and pain was not reported as a factor.

Maximal normalized EMG during the BFO protocol was reduced to a greater level at FP compared with the HF protocol. Greater reductions in normalized maximal EMG with BFR compared with free flow protocols at the same low-force have been previously reported (¹⁸). Furthermore, the ability of participants to maintain high levels of VA despite reductions of force and EMG is likely dependent on competing processes, such as intrinsic adaptation of motoneuron excitability (²⁹), inhibitory feedback from peripheral metabolite-sensitive muscle receptors (²⁷), and perhaps reductions in muscle spindle (³⁰) and Golgi tendon organs (³¹). The output as assessed by EMG is likely attributed to either reductions in motor unit firing rates or reductions in the number of active motor units, or both (³²). The greater reduction in maximal EMG after BFO therefore is likely due to reductions in the number of active motor units caused by a greater metabolic stress of the working muscle, and inactivation of the fibers, as a result of the nonoxygenated environment. After the termination of both protocols, M_max area was not significantly reduced, indicating that despite that greater amount of peripheral fatigue experienced after the BFO protocol, muscle fibers can still be activated fully with imposed stimulation on the system as previously noted above (^33,34). Thus, the greater impairments in MVC, velocity, and power with BFO compared with HF are due mainly to peripheral factors as indicated by the greater decrease in maximal EMG.

After a sustained muscular contraction of several minutes, recovery of voluntary force is largely completed in a few minutes; however, the long-lasting fatigue component observed at low frequencies (20 Hz) of excitation compared with those at high frequencies (50 Hz) can remain reduced for hours or days (¹²). This low-frequency fatigue indicates some damage to the structure of the muscle fiber and EC coupling mechanisms (³⁵). In the current study, this was evident by finding that MVC and 50-Hz recovery after both protocols as previously observed (¹²), whereas the 20-Hz and twitch responses were more depressed at R0 after BFO than HF (~78% and ~48%, respectively). By R2, there was no difference in low-frequency fatigue between protocols but each remained depressed from baseline for the 20-min recovery period. Furthermore, there was greater slowing in the 20-Hz HRT measures in the BFO protocol compared with HF, indicating that the BFO protocol likely caused a greater slowing of type II fibers which resulted in a reduction in velocity and impacting power (³⁶).

The results of the current study have shown that although less Work is performed during a low-force single bout of dynamic biceps contraction with BFO to failure, compared with high-force with no external impediment to blood flow, power, as a function of strength and velocity, is depressed approximately 23% more. Measures of VA, maximal EMG, and peripheral contractile properties, measured using tetanic involuntary contractions, have been combined to provide a comprehensive evaluation of the determining factors during the different fatiguing protocols. Results may indicate that central factors are not likely a major determining cause for the level of fatigability experienced during these protocols as reflected by the VA measure, but rather, peripheral fatigue is the major contributing factor. During fatigue, the greater loss of velocity, torque, and therefore, power during the BFO protocol may be due to structural damage and EC coupling failure of type II fibers, which are recruited for a longer duration. This is further indicated by the greater reductions in low-frequency involuntary tetanic contractions (20 Hz), and the slowing of the 20-Hz HRT after the BFO compared with HF. These results may indicate a greater challenge to the system when training with blood flow restriction compared with an alternative high-force bout without blood flow restriction. Because much of the literature on this topic has used BFR rather than BFO (as in the current study), extrapolation of results to chronic training with BFR, although sensible, needs to be carefully considered. Positive compensatory adaptations of strength and hypertrophy may be realized during chronic dynamic training using low-force exercise with blood flow restriction that may be beneficial in compromised situations, such as rehabilitation and aging.