The American College of Sports Medicine (ACSM) recommends resistance training at moderate to high loads (70-85% of one-repetition maximum (RM)) using multiple sets of 8-12 repetitions with short rest periods (1-2 min) two to three times per week (3). Resistance training regimens that use the principles of progressive overload are likely to result in increased muscle size and, consequently, strength (3). Skeletal muscle responds to overload with enhanced rates of protein synthesis leading to increases in the size and amount of contractile proteins within each muscle fiber (20). Although the exact mechanisms that trigger protein synthesis remain elusive, muscle hypertrophy and increased strength as a result of high-load (HL) resistance training is generally thought to be associated with the recruitment of higher threshold motor units (14) resulting in increases in mechanical stress (10), endocrine responses (12), and metabolite accumulation (10). For example, large, acute increases in growth hormone (GH) immediately after HL exercise is theorized to stimulate the secretion of insulin-like growth factors (IGFs) leading to increased protein synthesis and ultimately muscle hypertrophy (3,12).
During periods of reduced muscle use, such as spaceflight and post-injury/surgery, substantial losses in muscle mass and strength are observed (18,24). Because performing HL resistance training during these times are not always feasible and at times actually contraindicated, it is plausible that the benefits of HL exercise can be achieved using considerably lower loads and blood flow restriction (BFR) (also known as vascular occlusion or ischemia). Exercising with BFR is routinely performed in Japan where it is referred to as Kaatsu. Researchers have reported similar muscular adaptations in Kaatsu as seen in HL exercise, and sometimes in a more rapid time course (1,25,27). For example, Takarada et al. (25) report similar gains in strength and muscle cross-sectional area (CSA) after 16 wk of either BFR or HL resistance training of the elbow flexors, and Abe et al. (1) describe an almost 9% increase in CSA accompanied by a 17% rise in squat strength after only 2 wk of training twice per day with a vascular occlusion at 20% 1RM. They also found 20% higher serum insulin-like growth factor-1 (IGF-1) at the end of the 2 wk of training, attributing the rise in CSA and strength to increased protein synthesis through the GH-IGF-1 axis (1). Comparable improvements in strength and IGF-1 levels have been shown after 13 wk of HL training at 70% 1RM (5). In addition to these chronic positive effects, our laboratory has reported an acute ninefold increase in serum GH from baseline to the cessation of an exercise bout consisting of knee extensions at 20% maximum voluntary contraction (MVC) with BFR (17), and others have reported GH increases of up to 290 times baseline values (21,23,26,28), with the GH responses following BFR exercise being similar to, or even high than those reported during HL exercise of intensities at about 70% 1RM (11,19).
When the long- and short-term effects of BFR exercise on muscle and endocrine function are considered collectively, the data suggest that BFR exercise may indeed be a potent stimulator of muscle growth. However, since training with a blood flow restriction is rather novel (since about 2000), little is known about the most effective BFR protocol regarding exercise intensity and occlusion pressure and occlusion duration. Currently, studies using BFR vary greatly in terms of the protocol employed. For instance, exercise loads have ranged from 20 to 50% of 1RM or MVC (1,6,16,17,21-23,25-29), occlusion pressures have been moderate (about 30% above systolic blood pressure, or about 160 mm Hg) (1,6,19,21,22,25) or high enough to completely restrict blood flow (~300 mm Hg) (8,16,17,26,27,30), and the occlusion duration is not often specified, thus making it unclear whether the pressure was constantly applied throughout rest periods or released at the completion of each exercise set (25,27). If BFR exercise is to be developed as a viable intervention to enhance muscle mass and strength, understanding the interplay among these factors is critical in the development of the most effective BFR exercise protocol. Therefore, the purpose of this study was to determine the acute effect of eight different BFR protocols on skeletal muscle fatigue (decrement in MVC immediately following the performance of exercise) and compare these changes with the currently recommended resistance exercise intensity (~80% MVC) by the ACSM (3). On the basis of our previous finding that subjects who experienced the greatest decrement in force after a bout of BFR exercise concomitantly experienced the highest plasma GH concentrations exercise (Spearman r = −0.69) (17), and on the basis of the findings of Häkkinen and Pakarinen (9), who report that the more fatiguing the exercise, the greater the GH response, we suggest that in the present study, the most fatiguing exercise protocol could reasonably be the most potent stimulator for muscle growth and should be further tested in training studies.
Twenty-one healthy, normotensive males and females were recruited from the university community for participation in this study. All subjects signed an informed consent approved by the Syracuse University Institutional Review Board. Subjects completed a health questionnaire and anyone with a family history (including themselves) of blood clotting disorders and individuals with orthopedic injuries or limitations associated with the legs were excluded from the study. Additionally, subjects who reported using drugs that are known to increase the risk of circulatory problems (hormonal contraceptive medicine) were excluded. The subjects were 27 ± 4.9 yr, had a mean height of 1.73 ± 0.07 m, and weighed 74.4 ± 14.4 kg. All subjects participated in some form of regular physical activity (i.e., running, cycling, resistance training), but were asked to refrain from strenuous bouts of exercise for 2 d before each testing session.
The study required subjects to visit the laboratory on five separate occasions, with at least 5 d between visits. Subjects completed nine different exercise protocols, and the order of the protocols was counterbalanced. Each testing session involved the completion of two exercise protocols and lasted approximately 1.5 h. Before testing, blood pressure was taken and used to determine occlusion pressure.
Isometric muscle strength was assessed before and immediately upon completion of each exercise protocol. To measure exerted force, subjects were seated in a knee extension dynamometer (MedX, Ocala, FL). The back rest was adjusted to create a hip joint angle of 100° from flexion, and a seat belt was secured across the subject's hips to reduce any movements of the hip joint, and to minimize assistance from other muscle groups. For MVC testing, the knee joint angle was set at 60° from extension. During testing the contralateral limb was extended and rested on a pad. To obtain an MVC, subjects were instructed to push as hard as possible against an immovable arm attached to a force transducer. Isometric force was measured by a force transducer (model U1T, HBM Inc., Marlborough, MA; sensitivity of 0.002 V·N-1), amplified and recorded at 1000 Hz using a 16-bit data acquisition card (MP150, BioPac Systems Inc., CA). The force exerted by the subject was displayed on a 43-cm computer monitor located 1 m directly in front of the subject. Subjects performed MVC before the exercise protocols until two consecutive trials were within 5% of each other, and the highest MVC obtained was used in the analysis. The day-to-day reliability of the MVC using coefficient of variation was less than 8%. Subjects performed one MVC at the completion of the exercise protocol, and this post strength was measured within 30 s after volitional fatigue. During all MVC, strong verbal encouragement was given.
At each testing session subjects performed two different exercise protocols (one with each leg) that involved three sets of dynamic knee extensions to volitional task failure. Before each protocol, the subject performed a brief warm-up of 10-15 repetitions at a very light (less than 20% MVC) workload. Each set of knee extensions was separated by a 90-s rest period, and completion of the two protocols was separated by approximately 30 min of rest. The range of motion of the exercise was between 12 and 108° of knee extension (0° = full extension), and the subjects performed the leg extensions at a fixed cadence (approximately a 2-s concentric contraction and a 2-s eccentric contraction). The BFR protocols included two different exercise intensities (20 or 40% MVC), two occlusion pressures (partial (~160 mm Hg) or complete (~300 mm Hg)) and two occlusion durations (no pressure during the rest between sets or continuously applied). Each of these variables was independently manipulated resulting in a total of eight different protocols with an occlusion stimulus. An additional protocol was performed at 80% MVC with no BFR. The order of testing was counterbalanced and the leg tested on each protocol was randomly selected. On the fifth testing session after all nine protocols were completed, each subject repeated the first protocol they performed at the first testing session to ensure there was no training effect.
The occlusion pressure was applied via a 6 × 83-cm tourniquet cuff (D.E. Hokanson, Inc., Bellevue, WA), placed at the most proximal position on the subject's thigh, and inflated (E20 Rapid Cuff Inflator, D.E. Hokanson, Inc., Bellevue, WA) to the designated pressure. In the partial occlusion, the cuff was inflated to a pressure 30% higher than the subject's resting systolic blood pressure, and during the complete occlusion, pressure was set at 300 mm Hg. It has been suggested that a pressure 1.3 times systolic blood pressure (partial occlusion) impedes venous blood flow causing blood to pool in the capacitance vessels distal to the cuff while restricting some arterial blood flow (25). Using Doppler ultrasound technology, our laboratory has previously confirmed that a pressure of 300 mm Hg results in blood flow cessation to the limb (7).
Statistics are reported as means ± standard deviations. Boxplots were used to display specific percentile points and the overall shape of the data distribution. The central horizontal line within each box is the median (50th percentile) of the data. The top and bottom of each box displays the 25th and 75th percentiles while the ends of the whiskers are the 10th and 90th percentiles (29). One-way repeated-measures analysis of variance were used (within-subject main effect: protocol) on the following variables: exercise load, repetitions, protocol volume (load × repetitions), force decrements, and force decrements per protocol volume. Post hoc comparisons were performed using the Bonferroni adjustment for multiple comparisons. Paired t-tests comparing MVC, force decrement, and number of repetitions on the protocol that each subject repeated were also done. Power and eta squared (η2) were reported as appropriate, and significance was set at an alpha level of P< 0.05. Statistics were computed using SPSS version 15.0 (Chicago, IL).
Although males had higher initial MVC than females (868.8 ± 168.3 vs 550.8 ± 61.0 N, P < 0.01, power = 1.0, η2 = 0.55), the amount of fatigue (percent decrement in force) was similar across all protocols (P = 0.31, power = 0.54, η2 = 0.06). Further gender comparisons were not made because the fewer number of female subjects in the sample substantially decreased the power of this analysis.
When the subjects repeated their first protocol on their last visit, there were no differences in MVC (P = 0.89), force decrement (P = 0.82) and number of repetitions performed (P = 0.17). Furthermore, there were no significant differences between the subjects' MVC over the nine protocols (P = 0.91). This provides evidence that there was no training effect of the repeated testing.
The average exercise load in the HL was 609.3 ± 170.8 N, whereas the LLO protocols at 20 and 40% MVC exhibited loads of 161.1 ± 50.4 and 304.2 ± 85.0 N, respectively. All three exercise loads were significantly different from one another (P < 0.01).
Subjects performed more repetitions during all BFR protocols when compared with HL (Table 1; P < 0.01, power = 1.0, η2 = 0.44). The number of repetitions completed in the protocols at 20% MVC were not different, regardless of occlusion pressure and duration (mean = 165.0 ± 108.9; P > 0.05). The number of repetitions performed in the protocols at 40% MVC were similar (mean = 83.8 ± 43.4; P > 0.05) but were significantly less than the 20% MVC protocols (P < 0.05).
Protocol volume (load × repetitions) was lowest in the HL protocol, and this was significantly lower in comparison with the 40% continuous partial occlusion (40%ConPar) and 40% intermittent venous occlusion (40%IntPar) only.
On the average, subjects experienced the least amount of fatigue during the HL protocol (~19 ± 12%) and the highest during the 20% continuous complete (20%ConCom) and 20% continuous partial (20%ConPar) occlusion protocols (33 ± 18 and 32 ± 12%, respectively). However, the only significant difference in fatigue among the nine protocols was between HL and 20%ConPar (P=0.01, power = 0.91, η2 = 0.15). Although there was a slightly greater decrement in force observed during 20%ConCom, this was not significantly different from HL (P = 0.29), because of the greater variability associated with the protocol. The boxplot (Fig. 1) reveals the data distribution associated with muscle fatigue. The 20%ConPar protocol has the largest minimum fatigue value of all the protocols, indicating that subjects exhibited at least a 12% decrement in force, whereas other protocols resulted in small force decrements (20%IntPar = 2%, HL = 3%, and 20%ConCom = 4%). Also, the 20%ConPar protocol has a small spread (low variability) of values among the 25th and 75th percentiles. The boxplots with the widest spread (most variability) are the 20%IntPar, 20%ConCom, 40%ConCom, and 40%IntPar protocols, respectively.
To evaluate the efficiency (whether more fatigue was induced using lower volumes of exercise) of the protocols, fatigue was normalized to each protocol volume (decrement in force/[load × repetitions]). Although significant differences between protocols were found (P = 0.01), no significant pairwise comparisons were evident using the Bonferroni adjustment for multiple comparisons. The boxplot (not shown) exhibits similar median values for force decrement per load in all protocols, and all have similar spreads of values.
In this study, we sought to determine the most effective combination of exercise load, occlusion intensity, and occlusion duration to produce muscle fatigue. Our data show that all acute bouts of exercise with BFR are as fatiguing or more fatiguing than a bout of HL exercise. Whereas HL exercise resulted in a 19% decrement in isometric force after a bout of exercise, BFR exercise showed strength reductions of 24-33%. Häkkinen and Parkarinen (9) report similar amounts of fatigue in that muscle force was reduced by 10.3 ± 4.7 and 24.6 ± 18.9% after two different high-load protocols (20 × 1RM and 10 sets of 10 repetitions at 70% 1RM, respectively). Using a BFR protocol similar to 20%ConCom but with a higher volume of exercise (five vs three sets) being performed, Pierce et al. (17) have reported a 55.8% strength decrement. The magnitude of muscle fatigue after resistance exercise is known to be associated with the intensity of the exercise load, type of load (concentric or eccentric muscle contractions), the rest intervals between sets, and individual characteristics such as muscle morphology and training status (9,13). In the present study, the intensity of the load was the only factor that was manipulated, and we have demonstrated that a bout of BFR exercise can be as effective as HL in generating muscle fatigue.
Large increases in serum GH levels have been correlated with fatigue after single bouts of HL exercise (2,9), and GH is well-known to influence protein regulation and muscle growth during HL resistance training (11,12). We recently reported a relationship between acute fatigue and GH secretion after nine subjects were observed under both BFR with resistance exercise and BFR without exercise (muscle ischemia without contraction) (17). Here, the ischemic condition alone resulted in a 16% decrement in muscle force independent of a significant rise in GH, and low-load exercise coupled with BFR resulted in a 55.8% reduction in muscle force accompanied by a ninefold increase in GH levels. When the relationship between fatigue and GH response was evaluated using these data, a significant correlation of −0.69 was evident, indicating that subjects who experienced the largest acute muscle fatigue also had the highest plasma GH secretion (17). Therefore, on the basis of these findings (17), we speculate that the most fatiguing BRF exercise protocol in this study is the most potent stimulator for muscle growth, as it would likely elicit the greatest GH response. Although current studies reveal that BFR exercise results in muscle hypertrophy and increased strength in as little as 2 wk (1,25,27,28), the muscular adaptations leading to hypertrophy remain unknown. Additional training studies are warranted to further establish the role of GH as a hypertrophic factor in response to BFR exercise.
Although there were no differences in the amounts of fatigue exhibited among the BFR protocols, it is important to determine which protocol would be best suited for a resistance training regimen. We evaluated several factors in the BFR protocols, including degree of muscle fatigue, the range of fatigue within each protocol, fatigue variability, and protocol efficiency (force decrement per protocol volume). On the basis of these criteria, at the present time we recommend using the 20%ConPar protocol for several reasons. First, this protocol was the only one that resulted in a significantly greater decrement in force than HL (a decrease of 240 vs 140 N, or a 32% decrease in MVC vs a 19% decrease in MVC). Although 20%ConCom resulted in similar decrements as 20%ConPar, the subject variability in fatigue was much higher, making the impact of this protocol less predictable. Second, all subjects tended to respond to 20%ConPar. For example, the minimum force decrement in the 20%ConPar protocol was 64 N (or 12% MVC), whereas the minimum force decrements in the HL and 20%ConCom protocols were 10 N (or 2% MVC) and 15 N (or 2.5% MVC), respectively. Therefore, we expect that individuals would exhibit fatigue after 20%ConPar. Third, the variability (as displayed by the spread of the boxes and the ends of the whiskers in Fig. 1) is considerably uniform in 20%ConPar when compared with the other protocols. HL exercise is rather consistent, but, as stated previously, it exhibits much lower force decrements than 20%ConPar. Fourth, although more repetitions are performed during BFR, the exercise volume (load × repetitions) is consistent in the 20%ConPar, 20%ConCom, and HL protocols. Furthermore, the amount of force decrement per exercise volume is the same in all protocols, indicating that although one may perform fewer repetitions with HL, it is not more efficient in creating fatigue than an occlusion protocol.
We also believe that 20%ConPar is more conservative (and possibly safer) than several of the other BFR protocols, because the occlusion pressure used is lower. Although severe and often irreversible damage to skeletal muscle is believed to occur after a prolonged exposure to ischemia (>3 h) (4), each BFR protocol in this study was completed in less than 20 min, and no adverse effects of the exercise occurred. The protocols using a partial occlusion likely do not completely restrict arterial flow, so the exercising muscle is still receiving some blood flow. In surgical situations in which a bloodless (via complete occlusion) operating field is needed, McEwen et al. (15) recommend using the minimum occlusion pressure possible to prevent nerve, muscle, and skin injury. Although the long-term effects of BFR resistance training are unknown, other authors have suggested that muscle and microvascular damage may result from ischemia, including thrombus and pulmonary embolism formation (4,25). However, Takarada et al. (26) found no evidence of muscle damage or oxidative stress when assessed by creatine phosphokinase and lipid peroxide, respectively, after a bout of exercise. Because there were no differences in fatigue among any of the BFR protocols tested in this study, we recommend using at most a partial occlusion of 1.3 times systolic blood pressure around the quadriceps femoris. With more research, it is possible that occlusion pressures less than that can effectively elicit muscle fatigue.
Although we did not take measurements of pain or discomfort associated with the exercise protocols, we anecdotally observed that 20%ConPar protocol was more comfortable and tolerable than protocols with higher exercise loads and occlusion pressures. Wernbom et al. (30) have reported pain ratings to be higher with BFR exercise compared with a nonoccluded exercise at loads higher than 20% 1RM. They used a pressure of about 200 mm Hg, which is greater than the pressure used on 20%ConPar. Therefore, although there were no differences between the eight BFR protocols in the amount of fatigue exhibited, 20%ConPar seems to be a suitable protocol in terms of comfort and effectiveness to induce muscle fatigue.
In conclusion, all BFR protocols elicited at least as much fatigue as HL, despite the fact that much lower loads were used. The 20%ConPar protocol was the only one that elicited significantly more fatigue than HL exercise and, therefore, could have the potential to stimulate muscle growth. Future studies should include a comprehensive analysis of the acute and long-term training adaptations and mechanisms for hypertrophy from BFR resistance training and compare them to HL exercise training. The overall safety of BFR exercise needs to be evaluated as the effect of repeated, regular bouts is unknown. Health risks associated with long-term BFR training should be determined, and populations in which this type of training is contraindicated should be established. Finally, the effectiveness of BFR training for individuals who cannot engage in HL resistance training (frail elderly, individuals postsurgery) should be further evaluated.
This investigation was supported by the National Aeronautics and Space Administration (NASA) (NNX06AG26H).
Present address of B.C. Clark as of August 2006: Dept. of Biomedical Sciences, College of Osteopathic Medicine, Ohio Univ., 211 Irvine Hall, Athens, OH 45701.
1. Abe, T., T. Yasuda, T. Midorikawa, et al. Skeletal muscle size and circulating IGF-1 are increased after two weeks of twice daily "KAATSU
" resistance training. Int. J. Kaatsu Train. Res.
2. Ahtiainen, J. P., A. Pakarinen, W. J. Kraemer, and K. Häkkinen. Acute hormonal and neuromuscular responses and recovery to forced vs. maximum repetitions multiple resistance exercises. Int. J. Sports Med.
3. American College of Sports Medicine position stand. Progression models in resistance training for healthy adults. Med. Sci. Sports Exerc.
4. Blaisdel, F. W. The pathophysiology of skeletal muscle ischemia
and the reperfusion syndrome: a review. Cardiovasc. Surg.
5. Borst, S. E., D. V. DeHoyos, L. Garzarella, et al. Effects of resistance training on insulin-like growth factor-1 and IGF binding proteins. Med. Sci. Sports Exerc.
6. Burgomaster, K. A., D. R. Moore, L. M. Schofield, S. M. Phillips, D. G. Sale, and M. J. Gibala. Resistance training with vascular occlusion: metabolic adaptations in human muscle. Med. Sci. Sports Exerc.
7. Clark, B. C., S. R. Collier, T. M. Manini, and L. L. Ploutz-Snyder. Sex differences in muscle fatigability and activation patterns of the human quadriceps
femoris. Eur. J. Appl.
8. Greenhaff, P. L., K. Söderlund, J. M. Ren, and E. Hultman. Energy metabolism in single human muscle fiber during intermittent contraction with occluded circulation. J. Physiol.
9. Häkkinen, K., and A. Pakarinen. Acute hormonal responses to two different fatiguing heavy-resistance protocols in male athletes. J. Appl. Physiol.
10. Jones, D. A., and O. M. Rutherford. Human muscle strength training: the effects of three different regimens and the nature of the resultant changes. J. Physiol.
11. Kraemer, R. R., J. L. Kilgore, G. R. Kraemer, and V. D. Castracane. Growth hormone, IGF-1, and testosterone responses during resistive exercise. Med. Sci. Sports Med.
12. Kraemer, W. J., and N. A. Ratamess. Hormonal responses and adaptations to resistance exercise and training. Sports Med.
13. Linnamo, V., K. Häkkinen, and P. V. Komi. Neuromuscular fatigue
and recovery in maximal compared to explosive strength loading. Eur. J. Appl. Physiol.
14. MacIntosh, B. R., P. F. Gardiner, and A. J. McComas. Skeletal Muscle:Form and Function
. Champaign, IL: Human Kinetics, 2005.
15. McEwen, J. A., D. L. Kelly, T. Jardanowski, and K. Inkpen. Tourniquet safety in lower leg applications. Orthop. Nurs.
16. Moritani, T., M. Sherman, T. Shibata, M. Matsumoto, and M. Shinohara. Oxygen availability and motor unit activity in humans. Eur. J. Appl. Physiol.
17. Pierce, J. R., B. C. Clark, L. L. Ploutz-Snyder, and J. A. Kanaley. Growth hormone and muscle function responses to skeletal muscle ischemia
. J. Appl. Physiol.
18. Ploutz-Snyder, L. L., P. A. Tesch, D. J. Crittenden, and G. A. Dudley. Effect of unweighting on skeletal muscle use during exercise. J. Appl. Physiol.
19. Reeves, G. V., R. R. Kraemer, D. B. Hollander, et al. Comparison of hormone responses following light resistance exercise with partial vascular occlusion and moderately difficult resistance exercise without occlusion. J. Appl. Physiol.
20. Russell, B., D. Motlagh, and W. W. Ashley. Form follows function: how muscle shape is regulated by work. J. Appl. Physiol.
21. Sato, Y., A. Yoshitomi, and T. Abe. Acute growth hormone response to low-intensity KAATSU
resistance exercise: comparison between arm and leg. Int. J. KAATSU Train. Res.
22. Takano, H., T. Morito, H. Iida, et al. Hemodynamic and hormonal responses to a short-term low-intensity resistance exercise with the reduction of blood flow. Eur. J. Appl. Physiol.
23. Takano, H. T., T. Morito, H. Iida, et al. Effects of low-intensity "KAATSU
" resistance exercise on hemodynamic and growth hormone responses. Int. J. KAATSU Train. Res.
24. Takarada, Y., H. Takazawa, and N. Ishii. Applications of vascular occlusion diminish disue atrophy of knee extensor muscles. Med. Sci. Sports Exerc.
25. Takarada, Y., H. Takazawa, Y. Sato, S. Takebayashi, Y. Tanaka, and N. Ishii. Effects of resistance exercise combined with moderate vascular occlusion on muscular function in humans. J.Appl. Physiol.
26. Takarada, Y. Y., S. Nakamura, T. Aruga, S. Onda, and N. Miyazaki. Rapid increase in plasma growth hormone after low-intensity resistance exercise with vascular occlusion. J. Appl. Physiol.
27. Takarada, Y., Y. Sato, and N. Ishii. Effects of resistance exercise combined with vascular occlusion on muscle function in athletes. Eur. J. Appl. Physiol.
28. Takarada, Y., T. Tsuruta, and N. Ishii. Cooperative effects of exercise and occlusive stimuli on muscular function in low-intensity resistance exercise with moderate vascular occlusion. Jpn. J. Physiol.
29. Tukey, J. W. Exploratory Data Analysis
. Reading, MA: Addison-Wesley, 1977.
30. Wernbom, M., J. Augustsson, and R. Thomeé. Effects of vascular occlusion on muscular endurance in dynamic knee extension exercise at different submaximal loads. J. Strength Cond. Res.