Blood flow restriction (BFR) is used with low-load exercise to mimic high-load work and achieve similar benefits toward improving strength, power, and hypertrophy, which are not typically produced at lower loads (1–3). The technique involves applying a cuff over a limb to partially occlude arterial blood flow at a selected percentage during exercise. The occlusion of the target muscle group creates a hypoxic environment that promotes increased strength, power, and hypertrophy as normally seen after high-intensity exercise (1–3). BFR training 2–3 d·wk−1 at a load of 30%–50% one-repetition maximum (1-RM) and 15–30 repetitions can result in muscle adaptations in as little as 3–5 wk (1,3–7). It is believed that low-load, high-repetition training without BFR does not produce these improvements because the hypoxic environment is necessary to promote appropriate metabolic stress for muscle hypertrophy and strength gains (6,8). BFR training provides a safe alternative for clinical populations that cannot meet exercise guidelines from the American College of Sports Medicine but would still benefit from achieving therapeutic goals to increase muscle strength, power, and hypertrophy (2,9).
BFR has been demonstrated to increase muscle strength and hypertrophy compared with control interventions in healthy participants, athletes, postsurgery and clinical rehabilitation individuals, and older adults (2,9–12). Studies with varying protocols in training volume, percent blood flow occlusion, and exercises selected have indicated that BFR can be useful across different training protocols and populations (2,9–13). Much of the previous literature involves single-joint exercises, with little focus on functional movements or aerobic exercise (7). Hernandez et al. (14) assessed studies with functional movements and BFR and concluded that 80% of the studies analyzed demonstrated strength gains in functional BFR resistance training. These studies included a range of participants who were uninjured to those recovering from surgery or a chronic condition (14). Most often, BFR with aerobic exercise is conducted using elastic wraps, which makes the percent of blood occlusion unknown (6,15,16). BFR in combination with aerobic exercise has been shown to improve aerobic capacity in young adults (16), but there is no knowledge about an effective threshold for occlusion of blood flow. The purpose of this study is to provide a specific percent occlusion of blood flow with aerobic exercise in combination with functional resistance exercise.
Current BFR literature consists of substantive variance regarding control group comparisons and lacks definitive recommendations for volume and intensity of exercises performed with BFR to achieve optimal results in resistance training and aerobic exercise. The control groups’ uses of the same repetitions, high-intensity loads, low-intensity loads, training volume, or BFR vary. Further, the threshold of blood occlusion in aerobic training is unknown. This study aimed to investigate if 4 wk of BFR training with a combined approach of rowing and deadlifts increased measures of strength, power, and muscle size compared with an isovolumetric control intervention in apparently healthy participants. We hypothesized that during 4 wk of training with or without BFR, participants would increase strength, power, and muscle cross-sectional area (CSA) with no change in maximal aerobic capacity and no difference between groups. This study is unique in that it is intended to evaluate well-controlled BFR in a combined exercise approach with aerobic and strength training. This is particularly relevant because many clinical exercise programs include both types of training.
Participants and Study Design
This was a prospective, single-blinded, randomized longitudinal study. To achieve 80% power, two groups with n = 15 in each group (estimated effect size of 3.8) was determined a priori for primary outcome of changes in muscle size (17). Participant inclusion criteria were 1) between 18 and 60 yr of age at the initial testing session and 2) recreationally active, defined as engaging in 3 d or more of moderate-intensity physical activity per week, including a combination of aerobic and strength training. Exclusion criteria were failure to achieve clearance for exercise via the Physical Activity Readiness Questionnaire for Everyone (Par-Q+) screening tool, currently pregnant or within 3 months postpartum, presence of chronic low back pain, history of surgery to change the structure of joints within the lower extremity (e.g., joint replacement, osteotomy), known blood clotting disorder or taking blood thinner medications, presence of peripheral vascular disease, presence of systemic or widespread pain disorder (e.g., fibromyalgia, chronic regional pain syndrome), and/or currently under activity limitations. Data collection began January 2020 and was cut short because of the coronavirus disease 2019 (COVID-19) pandemic (March 2020). Thus, only 20 participants completed the study. A total of 24 participants were screened for study participation; three participants were ineligible for participation because of geographical location, and one participant did not qualify because of chronic pain in the lower extremity. The 20 participants were randomized and completed the study prior to the pandemic.
This study was approved by the Campbell University Institutional Review Board (CUIRB-491) in accordance with the Declaration of Helsinki. All participants completed an informed consent and were provided with the opportunity to ask questions prior to study participation. This study was registered on www.clinicaltrials.gov (NCT03983070). Data collection took place in the Physical Therapy Department at Campbell University. Participants were randomly allocated to either the BFR intervention group or the control group after they were enrolled in the study. Figure 1 depicts the study timeline for measurements and training.
All testing sessions were conducted by blinded research personnel at baseline and after training completion. Testing sessions included completion of surveys and measurements of resting hemodynamics, muscle CSA, maximal aerobic capacity (V̇O2max), and estimated 1-RM of deadlift. Resting blood pressure and heart rate (HR) were measured after the participant sat quietly for approximately 5 min. Height (cm) and weight (kg) were measured using a calibrated Health-o-meter scale (Pelstar LLC, McCook, IL).
Thigh circumference (cm) was measured bilaterally in supine at the midpoint between markings made at the lateral femoral condyle and proximal greater trochanter. The mean of the three measurements taken was recorded for each leg. CSA measurements were taken of the vastus lateralis (VL) and biceps femoris (BF) following the protocol established by Ruas et al. (18) using the GE LOGIC e with 4C-RS convex transducer (GE Healthcare, Waukesha, WI). Standardized settings for gain (52 dB), depth (9 cm), and frequency (12 MHz) were used for all measures to maintain consistency. Using a water-gel coupling agent, the ultrasound transducer was applied perpendicularly to the skin interface over the muscle of interest at the line that marks 50% of the distance between lateral femoral condyle and proximal greater trochanter. Participants laid supine for VL and prone for BF measurement. Once the target muscle was visualized via ultrasound, the image was saved for CSA measurement. Each image was a single snapshot of a given muscle, attempting to capture the entire CSA within the screen. Three images were recorded each for VL and BF bilaterally. Saved ultrasound images were exported for CSA measurement using ImageJ software (National Institutes of Health, Bethesda, MD), and mean values were used for data analysis.
V̇O2max was assessed during a maximal graded exercise test on a stationary rower (Concept2, Morrisville, VT) using a metabolic analyzer (Parvo-medics, Provo, UT) until volitional exhaustion via a previously validated exercise protocol (19). HR was measured using a Polar chest strap (Polar Electro Oy, Kempele, Finland) and ratings of perceived exertion (RPE), from 1 to 10 (20), were recorded within the final 10 s of each stage. The test was terminated if the participant requested to stop or failed to maintain desired watts for that stage.
Lower extremity strength was assessed through an estimated 1-RM deadlift with a dumbbell in each hand. Participants completed a warm-up of eight repetitions with approximately 50% effort. Testing weight was selected with the aim for participants to complete between two and nine repetitions (21). Participants completed as many deadlifts as possible until volitional exhaustion. If they performed at least two and no more than nine repetitions of deadlifts, they were finished with the estimated 1-RM test. If they performed more than nine repetitions, a second bout was completed after 2 min of rest with increased weight. Periods of testing and rest were alternated until the participant completed between two and nine repetitions. The final deadlift weight and number of successful repetitions were entered into the estimated 1-RM equation established by Brzycki (21); this equation was not developed specially for deadlift but for general application to estimated 1-RM.
Participants trained twice a week for 4 wk. All training began with a warm-up of jumping jacks, squats, knee hugs, and straight leg kicks, followed by a standardized protocol of rowing and deadlifts, with or without BFR depending on group allocation. Training was prescribed to result in similar training volume between groups for both power and strength; thus, any differences between groups were related to method of training (BFR or control). For the BFR group, cuffs were donned on each thigh as proximally as possible after warm-up completion. Occlusion was established by use of Delphi Personalized Tourniquet System (Vancouver, BC, Canada) with the subject lying supine and relaxed. Both groups performed one bout of rowing for 2 min and three bouts of rowing for 1 min, with 30 s of rest between bouts. Rowing wattage was determined based on the max power attained during baseline testing. The BFR group rowed at 40% maximum power, and the control group rowed at 80%. The BFR cuffs stayed inflated during the 30-s breaks. After a 3-min rest, with cuffs deflated for the BFR group, participants completed four sets of deadlifts with 30 s of rest between sets. The BFR group lifted 30% 1-RM for one set of 20 reps and three sets of 10 reps, and the cuffs stayed inflated for the entire deadlift protocol. The control group lifted 60% 1-RM for one set of 10 reps and three sets of 5 reps. If a participant did not complete the protocol for either rowing or deadlift, the resistance was retained for the next session. RPE was recorded after completion of the final stage of rowing and deadlifts. If a participant reported less than 6/10 RPE, the load for that exercise was increased by 10% at the subsequent session.
Data were analyzed using SPSS version 25 (IBM Statistics, Chicago, IL), and the alpha level was set at 0.05 to determine differences. Dependent variables were muscle CSA of the VL and BF, V̇O2max, estimated 1-RM of deadlift, maximal power, training volume, and RPE. Independent variables were time and groups. Bilateral measurements for thigh circumference, VL, and BF were averaged to create one measure. Intraclass correlation coefficients (ICC) and 95% confidence intervals (CI) were calculated for right and left VL and BF. The data were shown to be normally distributed via a Shapiro–Wilks test.
Mean and standard deviation (SD) were calculated for each group at each time point. A 2 (group) × 2 (time points) repeated-measures analysis of covariance (RM-ANCOVA) was used to evaluate group and time differences for estimated 1-RM, power, muscle CSA, and V̇O2max. Sex was included as a covariate because the random assignment of groups did not distribute the sexes evenly between groups. If there was a main effect difference, univariate analyses were used to determine where these differences occurred. Effect sizes were calculated using partial eta squared (ηp2) and interpreted as small (0.01–0.05), moderate (0.06–0.13), and large (≥0.14) (22). Percent change from pre- to postassessment was also calculated.
A second RM-ANCOVA was conducted to evaluate group–time differences during the training sessions for mean rowing power, rowing RPE, mean weight deadlifted, and mean deadlift RPE. Sex was the covariate. If a main effect difference occurred, univariate analyses were used to decipher these differences. Effect sizes were calculated using ηp2 and interpreted as indicated above.
Twenty participants completed the study (BFR n = 9, 6 males, 3 females; control n = 11, 4 males, 7 females). The anthropometric and hemodynamic means ± SD for each group are shown in Table 1. The reliability of the CSA assessments indicated excellent reliability for right BF (ICC = 0.987, 95% CI = 0.978–0.993), left BF (ICC = 0.982, 95% CI = 0.971–0.990), right VL (ICC = 0.987, 95% CI = 0.977–0.992), and left VL (ICC = 0.987, 95% CI = 0.977–0.992).
TABLE 1 -
Descriptive Baseline Data (Mean ± SD) for the BFR and Control Groups.
||29.2 ± 8.0
||28.8 ± 6.3
||172.4 ± 8.2
||165.1 ± 7.2
||80.4 ± 22.6
||67.0 ± 8.4
|Resting heart rate (bpm)
||75.0 ± 9.1
||68.8 ± 8.9
|SBP (mm Hg)
||126.6 ± 8.8
||121.1 ± 9.0
|DBP (mm Hg)
||78.0 ± 10.4
||72.9 ± 6.5
|Activity level (MET min·wk−1)
||3188 ± 1233
||2756 ± 1664
DBP, diastolic blood pressure; MET, metabolic equivalent task; SBP, systolic blood pressure.
The 2 × 2 RM-ANCOVA for 1-RM, power, muscle CSA, and V̇O2max showed no main effect differences by group (λ7,11 = 0.756, P = 0.63, ηp2 = 0.325) but did show a main effect of sex (λ7,11 = 17.317, P < 0.001, ηp2 = 0.917). There was a main effect of time (λ7,11 = 7.955, P = 0.001, ηp2 = 0.835), but not an interaction between sex and time (λ7,11 = 1.537, P = 0.25, ηp2 = 0.494) or group and time (λ7,11 = 0.501, P = 0.82, ηp2 = 0.242). All effect sizes are interpreted as large. Univariate analyses for time revealed that posttraining measures for BF CSA (P = 0.001), VL CSA (P = 0.001), V̇O2max (P = 0.003), and maximum power (P = 0.001) were larger than the pretraining measurements in both groups. The control group showed a larger percent change than the BFR group for both CSA measurements and V̇O2max, but the BFR group had a larger percent change for maximum power output. These data are shown in Table 2.
TABLE 2 -
Pre- and Posttraining Mean ± SD of Muscular Measurements, Aerobic Capacity, Power, and Estimated Maximum Weight Lifted for BFR and Control Groups.
|Thigh circumference (cm)
||56.5 ± 6.3
||56.8 ± 6.3
||53.7 ± 4.1
||53.5 ± 4.1
|BF CSA (cm2)
||12.5 ± 3.9
||13.3 ± 3.3*
||11.0 ± 2.1
||11.8 ± 1.7*
|VL CSA (cm2)
||30.1 ± 7.0
||31.0 ± 7.1*
||24.5 ± 4.4
||25.9 ± 3.9*
||48.2 ± 14.0
||51.6 ± 16.7*
||50.1 ± 12.6
||54.6 ± 8.9*
|Maximum power (W)
||222.2 ± 72.3
||250.0 ± 66.1*
||209.1 ± 74.4
||229.6 ± 66.9*
|Estimated 1-RM (kg)
||41.1 ± 14.4
||49.3 ± 17.4
||35.2 ± 12.4
||41.7 ± 14.6
* There is a time effect (P < 0.05) indicating that the posttraining measurement is greater than the pretraining measurement.
Rowing RPE and power over the eight sessions are shown in Figure 2. Figure 3 depicts the weight lifted and RPE for deadlift over the eight sessions. The 2 × 8 RM-ANCOVA (evaluating rowing power, rowing RPE, mean weight deadlifted, and mean deadlift RPE) showed a main effect by group (λ4,13 = 7.365, P = 0.003, ηp2 = 0.694) and a group–sex interaction (λ4,13 = 5.069, P = 0.01, ηp2 = 0.609). There was also a main effect for time (λ28,394 = 4.000, P < 0.001, ηp2 = 0.202), but not for time–sex interaction (λ28,394 = 1.145, P = 0.28, ηp2 = 0.068), or for time–group interaction (λ28,394 = 1.457, P = 0.07, ηp2 = 0.085). All effect sizes are considered large. The univariate analyses indicated group differences for RPE during rowing (P = 0.04), power output (P < 0.001), and weight lifted (P = 0.001). In all cases, the control group had higher values. There was also a time difference for power output (P < 0.001) and weight lifted (P < 0.001).
The results displayed improvements in muscle CSA, estimated 1-RM, power, and V̇O2max in both groups—all with large effect sizes. There was an influence of sex because there was a mismatch in the number of males and females in the two groups. This influence was shown in the baseline differences in height and VL CSA with the BFR group (66% male) having greater group means than the control group (36% male). None of the training responses were sex dependent. Results also showed that both groups increased training loads for rowing and deadlift over the eight training sessions, with the BFR participants consistently recording lower RPE values than the controls.
The improvements in CSA, with no group differences, are in line with previous research. Giles et al. (2) showed an increase in muscle size over the longer training period with greater volume (30% and 70% 1-RM for BFR and control groups, respectively). The isovolumetric design of interventions between groups provided enough stimulus in the low load with BFR training group to match the high load without BFR group. Further, Giles et al. (2) and Segal et al. (11) included clinical populations, whereas the present study and Wells et al. (13) included only healthy individuals. Collectively, the studies show increased CSA with BFR in healthy and movement-limited populations, regardless of sex.
Although CSA increased in both groups in the present study, there was no change over time for thigh circumference. This contradicts previous literature showing increased muscle circumference with BFR in the upper and lower extremity (5,9,13). Interestingly, Credeur et al. (5) used the contralateral upper extremity of each participant as the control and showed increases in both arms. Tennent et al. (9) showed improvements in circumference after 6 wk of training three times per week with patients after knee arthroscopy. The design of Tennent et al. (9) only used BFR for three additional exercises and otherwise had both groups perform the same protocol.
Participants in both groups of the present study improved strength, with no difference between groups. Segal et al. (12) also found no difference between groups but did not describe any increase of load over the study with older men, in contrast to the present study where resistance was increased by 10% after a participant reported less than 6/10 RPE. The study showed an effect of exercise when combined with BFR, but without knowledge of progressed overload, it is unclear how the results may have differed (12). Mattocks et al. (23) did include a progressive increase in volume coupled with arterial occlusion during exercise, showing increased RPE with increased volume, but there was no report of changes in strength after training. Additionally, Tennent et al. (9) and Araújo et al. (10) both reported increases in strength of the BFR group compared with controls with longer training periods—6 and 8 wk, respectively. Although it does not appear that exercise volumes and intensities were selected to induce overload for muscular adaptations in the previous literature, it is possible that observed changes are a result of neural adaptations and the addition of BFR was able to provide more of a hypoxic muscular environment with low-load exercise training to allow greater increase in strength than low-load exercise without BFR, even in as little as 4 wk, which is a common therapeutic time frame for clinical populations.
A recent meta-analysis concluded that aerobic exercise training with BFR induced larger improvements in aerobic capacity than training without BFR, despite the method of training (i.e., treadmill, cycle) (16). In the present study, both groups demonstrated an increase in rowing power and V̇O2max over time, with a greater increase in maximum rowing power for the BFR group, which is consistent with the results of the meta-analysis (16). The low-intensity exercise studies included loads from 30% to 40% V̇O2max, and the current study used 40% max power during rowing for the BFR group and 80% for the control group. It is possible that the low-intensity BFR training was not enough to induce as much of a change in V̇O2max (7.1%) as the high-intensity training performed by the control group (9%). The literature for BFR and aerobic training lacks definitive information for BFR occlusion prescription and exercise intensity and time recommendations that would parallel high-intensity exercise (16,24). Continued research is warranted to determine if aerobic training with BFR requires higher than 40% intensity to mimic that of high-intensity aerobic exercise.
This study demonstrated consistently higher RPE values from control participants compared with the BFR participants. Wells et al. (13) reported an overall decrease in RPE for both groups, and Abe et al. (24) reported higher RPE in the BFR group, but neither reported an increase in exercise load over the duration of training. The positive results seen in the present study for the BFR group suggest that low-load BFR training can achieve desirable results with less perceived exertion. This reduced load coupled with reduced perceived exertion may be an acceptable mode of strength training for groups that have low tolerance for resistance training. It would be useful to assess if there is a similar response in clinical populations that may have comorbidities or precautions that restrict them from training at typical levels per American College of Sports Medicine guidelines.
A limitation to this study was a small sample size because of interruption from the COVID-19 pandemic, which resulted in restricted access to the lab for health and safety reasons in the spring of 2020. For this same reason, groups were unbalanced with respect to sample size and sex distribution. Including sex as a covariate helped to account for this issue. Training was designed to mimic a therapeutic schedule of 2 d·wk−1 for 4 wk, which may not have been enough to stimulate a significant response between groups. There is also error in using estimated 1-RM calculations, but the equation chosen has been shown to provide comparable results to a traditional 1-RM in bench press (25). This method was chosen to provide a clinically relevant strength measure that would also reduce the risk of injury in the participants completing the lift. Several participants were relative novices to the deadlift strength assessment; thus, completing a standard 1-RM test with this movement may have resulted in pain or injury for the participant.
This study further established that BFR training at low loads with multijoint exercises can produce similar physiological effects as high-load exercise in just 4 wk. Both groups increased muscle CSA, estimated 1-RM, power, and V̇O2max, but there was lower perceived exertion for the exercise in the BFR group. These data add to the body of literature supporting the role of BFR in strength training and help to clarify the literature for the role of BFR in aerobic training. Further, the data support that performing these two modes of exercise with BFR results in improvements that have performance and clinical implications. Despite the use of healthy participants, this information may be theoretically applied to populations that cannot train at 60%–80% of maximum because of risk of injury or pain. Specifically, improvements in power are important for translation to clinical populations where power is vital for activities like stair climbing or jumping.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
The authors have no conflicts of interest to report. This study did not receive any funding.
1. Cook CJ, Kilduff LP, Beaven CM. Improving strength and power in trained athletes with 3 weeks of occlusion training. Int J Sports Physiol Perform
2. Giles L, Webster KE, Mcclelland J, Cook JL. Quadriceps strengthening with and without blood flow restriction in the treatment of patellofemoral pain: a double-blind randomised trial. Br J Sports Med
3. Takarada Y, Tsuruta T, Ishii N. Cooperative effects of exercise and occlusive stimuli on muscular function in low-intensity resistance exercise with moderate vascular occlusion. Jpn J Physiol
4. Garber CE, Blissmer B, Deschenes MR, et al. American College of Sports Medicine Position Stand: quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: guidance for prescribing exercise. Med Sci Sports Exerc
5. Credeur DP, Hollis BC, Welsch MA. Effects of handgrip training with venous restriction on brachial artery vasodilation. Med Sci Sports Exerc
6. Loenneke JP, Wilson JM, Marín PJ, et al. Low intensity blood flow restriction training: a meta-analysis. Eur J Appl Physiol
7. Takarada Y, Takazawa H, Sato Y, et al. Effects of resistance exercise combined with moderate vascular occlusion on muscular function in humans. J Appl Physiol
8. Hughes L, Paton B, Haddad F, et al. Comparison of the acute perceptual and blood pressure response to heavy load and light load blood flow restriction resistance exercise in anterior cruciate ligament reconstruction patients and non-injured populations. Phys Ther Sport
9. Tennent DJ, Hylden CM, Johnson AE, et al. Blood flow restriction training after knee arthroscopy: a randomized controlled pilot study. Clin J Sport Med
10. Araújo JP, Neto GR, Loenneke JP, et al. The effects of water-based exercise in combination with blood flow restriction on strength and functional capacity in post-menopausal women. Age (Dordr)
11. Segal NA, Williams GN, Davis MC, et al. Efficacy of blood flow-restricted, low-load resistance training in women with risk factors for symptomatic knee osteoarthritis. PM R
12. Segal N, Davis MD, Mikesky AE. Efficacy of blood flow-restricted low-load resistance training for quadriceps strengthening in men at risk of symptomatic knee osteoarthritis. Geriatr Orthop Surg Rehabil
13. Wells E, Eustace D, Gupton CS, et al. Eccentric and blood flow restriction exercises in women induce hypertrophy. J Sports Med Phys Fitness
14. Hernandez NJ, Myers BJ, Feito Y, Bunn JA. Functional strength exercise and blood-flow restriction protocols: a systematic review. Cent Eur J Sport Sci Med
15. Paton CD, Addis SM, Taylor L-A. The effects of muscle blood flow restriction during running training on measures of aerobic capacity and run time to exhaustion. Eur J Appl Physiol
16. Formiga MF, Fay R, Hutchinson S, et al. Effect of aerobic exercise training with and without blood flow restriction on aerobic capacity in healthy young adults: a systematic review with META-analysis. Int J Sports Phys Ther
17. Madarame H, Ochi E, Tomioka Y, et al. Blood flow-restricted training does not improve jump performance in untrained young men. Acta Physiol Hung
18. Ruas CV, Pinto RS, Lima CD, et al. Test-retest reliability of muscle thickness, echo-intensity and cross sectional area of quadriceps and hamstrings muscle groups using B-mode ultrasound. Int J Kinesiol Sports Sci
19. Wiener SP, Garber CE, Manfredi TG. A comparison of exercise performance on bicycle and rowing ergometers in female master recreational rowers. J Sports Med Phys Fitness
20. Borg G. Borg’s Perceived Exertion and Pain Scales
. Champagne (IL): Human Kinetics Publishers Inc.; 1998. p. 104.
21. Brzycki M. Strength testing: predicting a one-rep max from repetitions to fatigue. J Phys Educ Recreat Dance
22. Cohen J. Statistical Power Analysis for the Behavioural Sciences
. 2nd ed. Hillsdale (NJ): Lawrence Erlbaum Associates; 1988. p. 590.
23. Mattocks KT, Mouser JG, Jessee MB, et al. Perceptual changes to progressive resistance training with and without blood flow restriction. J Sports Sci
24. Abe T, Fujita S, Nakajima T, et al. Effects of low-intensity cycle training with restricted leg blood flow on thigh muscle volume and VO2max
in young men. J Sports Sci Med
25. Amarante do Nascimento M, Serpeloni Cyrino E, Yuzo Nakamura F, et al. Validation of the Brzycki equation for the estimation of 1-RM in the bench press. Rev Bras Med Esporte