Professional Documents
Culture Documents
net/publication/26891895
CITATIONS READS
85 281
4 authors, including:
Some of the authors of this publication are also working on these related projects:
Effects of high-frequency blood flow restricted resistance exercise (BFRRE) on neuromuscular adaptations with special reference to the time course of changes View
project
All content following this page was uploaded by Jesper Augustsson on 11 October 2017.
I
to concentric torque failure at 30% of the 1 repetition maximum. n the last 8 to 10 years, low-load training with blood
One leg was randomized to exercise with cuff occlusion and the flow restriction has attracted a lot of attention, both as
other leg to exercise without occlusion. The muscle activity in the a possible alternative to heavy resistance exercise in
the rehabilitation setting and as a training method to
quadriceps was recorded with electromyography (EMG).
increase muscle strength and size in healthy individuals.
Ratings of perceived exertion (RPE) and acute pain were col-
Several studies have shown that blood flow restriction by
lected immediately, and delayed onset muscle soreness
pressure cuffs in combination with low-intensity resistance
(DOMS) was rated before and at 24, 48, and 72 hours after exercise induces muscle mass increases at rates comparable
exercise. The results demonstrated high EMG levels in both with those seen with conventional strength training (17,26–
experimental conditions, but there were no significant differ- 28) and sometimes at even higher rates (2,12). In the studies
ences regarding maximal muscle activity, except for a higher that have included control groups that have trained at the
EMG in the eccentric phase in set 3 for the nonoccluded same intensities and volumes but without cuff occlusion, the
condition (p = 0.005). Significantly more repetitions were per- nonoccluded groups have generally made little if any gains in
formed with the nonoccluded leg in every set (p , 0.05). The muscle size and strength (1,2,12,27,28).
RPE and acute pain ratings were similar, but DOMS was higher However, because vascular occlusion reduces the muscle
in the nonoccluded leg (p , 0.05). We conclude that blood flow endurance at the low-to-moderate loads typically used in
training with blood flow restriction (33), and because the
restriction during low-intensity dynamic knee extension de-
occlusion has usually been applied continuously (including
creases the endurance but does not increase the maximum
rest periods) throughout the training session, the effort
muscle activity compared with training without restriction when
required to complete a given amount of work in an exercise
both regimes are performed to failure. The high levels of muscle bout should be substantially less with free flow conditions.
activity suggest that performing low-load dynamic knee exten- Therefore, one might argue that previous studies have not
sions in a no-relaxation manner may be a useful method in knee controlled for the degree of effort when comparing training
with and without cuff occlusion. Furthermore, it has been
Address correspondence to Mathias Wernbom, mathias.wernbom@ reported that there was little difference in the acute pain
orthop.gu.se. ratings between the occluded and nonoccluded legs when
23(8)/2389–2395 dynamic knee extensions were performed in a continuous
Journal of Strength and Conditioning Research nonstop manner to fatigue, indicating that ischemic
Ó 2009 National Strength and Conditioning Association conditions were induced by the exercise itself (33). Therefore,
remained seated in the machine during the rest periods other connected devices. The EMG was normalized to
between the sets. a maximum voluntary isometric activation (MVIA) using the
Continuous partial occlusion rather than intermittent normalization function of the MuscleLab unit. After a
severe occlusion was chosen because most of the previous warm-up with gradually increasing force developments, 2
studies have used a partial occlusion training model and MVIA trials were performed at a knee angle of 70°. The
because partial blood flow restriction is more tolerable than subjects were instructed to smoothly increase force, and once
severe or complete occlusion (8). Hoeger et al. (15) the maximum was reached, the effort was sustained for at
demonstrated a test-retest reliability of r = 0.86 to 0.96 for least 3 to 4 seconds, during which time EMG was collected
knee extension endurance tests at loads of between 40% and for normalization. Alkner et al. (3) reported a coefficient of
60% of 1RM when performed in a similar continuous fashion variation between trials for EMG to be in the range of
as in our study. Tagesson and Kvist (23) reported intraclass 4.7–9.2% for the MuscleLab system.
correlation coefficient values of 0.90 to 0.96 for the unilateral A display with the EMG RMS curve was placed in front
knee-extension 1RM test. of the leg extension machine for monitoring purposes and
to provide real-time feedback to the subject. Because the
Electromyography and Goniometry. The skin was shaved with normalization value was chosen automatically by the
a hand razor and carefully cleaned with ethanol before MuscleLab, the first of the 2 MVIA attempts was saved and
electrode placement. Bipolar surface self-adhesive Ag/AgCl used as the reference value (100%) if the 2 attempts did not
monitoring electrodes (model 2223, 3M Medical, Neuss, differ by more than approximately 5% in RMS amplitude, in
Germany) with a 4.0 cm interelectrode distance were placed which case further attempts were made until a plateau in
on the medial portion of the vastus lateralis (VL) and the amplitude was noted. During the 30% of 1RM endurance tests
vastus medialis (VM) muscles of the quadriceps femoris with and without occlusion, as in the MVIA trials, the EMG
muscle group approximately 15 and 10 cm above the RMS curve was displayed in real-time in front of the subject
proximal border of the patella. These placings were for motivation, and loud verbal encouragements were
somewhat similar to those used by Andersen et al. (4), and also used to motivate the subjects to continue to the point
they were chosen to avoid having the electrodes passing over of concentric torque failure. The average RMS values for
the innervation zones of the muscles during the exercise complete concentric and eccentric phases, respectively, were
movement while still leaving enough room for the used for analysis. The first repetition of each set was discarded
tourniquet. The approximate locations of the innervation because the subjects generally had not found the proper
zones were based on the anatomic guidelines provided by tempo. During the final 2 or 3 repetitions in each set, most
Rainoldi et al. (20). Knee joint angle was measured with an subjects also had difficulty following the correct tempo
electrogoniometer, which was positioned laterally to the because of the accumulating fatigue, but these repetitions
knee joint. Calibration of the goniometer signal was were considered for analysis as long as they were in the full
performed at anatomic knee joint angles of 0° and 90° using range of motion.
a geometric retractor. A linear encoder was used to measure
the position and displacement of the weight stack. The Ratings of Pain, Perceived Exertion, and Delayed Onset Muscle
goniometer and encoder data were used to identify the Soreness. The subjects were asked at the end of both the
concentric and eccentric
phases of each repetition.
The EMG and goniometer/
encoder data were collected
using the MuscleLab 3010e
recording and acquisition sys-
tem (Ergotest Technology,
Langesund, Norway), using
the accompanying software
program on a laptop computer
(3). The MuscleLab EMG leads
had built-in preamplifiers, and
the pre-amplified signal was
filtered through a 8 to 1,200
Hz band pass filter before it
was root mean square (RMS)
converted. The RMS converted Figure 1. Number of repetitions with and without partial occlusion. * = p , 0.05 between the nonoccluded and
signal was then sampled at 100 occluded conditions.
Hz and synchronized with the
TABLE 2. Lowest and highest electromyographic (EMG) activation (% of maximum voluntary isometric activation [MVIA])
for complete concentric and eccentric phases for vastus medialis (VM) and vastus lateralis (VL) during occluded and
nonoccluded conditions, set 1 to 3 (means 6 SD).
significantly different between the conditions (Table 3). For occluded leg but also the nonoccluded leg were submitted to
DOMS, significantly greater ratings on the VAS scale were relative ischemia. We therefore suggest that because of the
noted for the nonoccluded leg (p , 0.05) (Table 3). ischemic conditions, the type I fibers became highly fatigued
during the first set, thus necessitating the recruitment of an
DISCUSSION increasing number of motor units containing type II fibers as
The major finding of the current study was that although the set progressed, as indicated by the increase in EMG
partial restriction of blood flow with a thigh tourniquet did (Table 2). In addition, the EMG data indicate that the type II
influence the endurance in the dynamic knee extension fibers were recruited earlier in the second and third sets
exercise at an intensity of 30% of 1RM, it did not increase the than in the first set, presumably because of residual fatigue in
maximum level of muscle activity in the quadriceps either in type I fibers.
the concentric or the eccentric phase compared with no cuff The gradual increase in the EMG during the eccentric
occlusion when both regimes were performed to the point of phase also deserves comment. During the first few repetitions
concentric torque failure. If anything, there was a tendency for of each set, the activity was low, as expected (approximately
greater activity in both of the vastus muscles during the 25–30% of MVIA), but as the point of concentric torque
nonoccluded condition for both the concentric and the failure drew nearer, the eccentric activity showed a marked
eccentric phases, which became significant for the eccentric increase, ending at levels of approximately 50–60% of MVIA,
phase in the third set. Generally, the lowest activity was seen which suggests the possibility that motor units containing fast
in the beginning of the sets and the highest during the last few fibers were increasingly recruited also during the eccentric
repetitions of each set, and this was true for both conditions phase. To the best of our knowledge, this increase in
and for both the concentric and the eccentric phases. Also, the activation during eccentric muscle actions has not been
muscle activity during the first repetitions, especially during reported before in the literature on training combined with
the concentric phases, was clearly greater during the second blood flow restriction. Therefore, we propose that fatiguing
and the third sets compared with the first set. low-intensity dynamic training during ischemic conditions
At first glance, our findings may appear to contradict includes a significant eccentric component and that this could
those of previous studies that have compared training with potentially contribute to the training effects observed with
and without cuff occlusion in terms of muscle activity and this type of training.
in which occlusion was demonstrated to yield higher levels of Not surprisingly, fatiguing low-load strength training was
muscle activation. However, in these studies (25,26,34,35), the associated with a significant degree of pain because of the
degree of effort most likely differed between the conditions, ischemia and accumulation of metabolites in the quadriceps.
in contrast with the current study where the effort was The Borg RPE values were also generally high, as expected,
maximal in both conditions. Furthermore, in the present considering that the subjects performed all-out sets. In-
study, high levels of activation (86–107% of MVIA) were seen terestingly, the pain ratings were similar between conditions.
in both conditions in the last few repetitions during the In contrast, in our previous study (33), occlusion was
concentric phases, and shorter bursts (lasting a few tenths of associated with greater pain levels. However, in that study
a second) of approximately 120–160% of MVIA were not (33), the pain levels were maximal in the moments
uncommon. Therefore, our results confirm the findings of immediately after the exercise, when the cuff had not yet
Takarada et al. (25,27) of high EMG levels during low- been deflated. The high pressure (200 mm Hg) used in
intensity (20–40% of 1RM) training with blood flow combination with the relatively wide cuff probably imposed
restriction. The novel finding in our study is that a high a nearly complete restriction of blood flow. In the current
level of activation is possible also without cuff occlusion if the study, the lower pressure used (100 mm Hg) most likely
repetitions are performed in a no-relaxation manner. allowed for some washout of metabolites, and therefore the
The high EMG levels suggest, but do not directly prove, pain did not increase in the rest periods between the sets.
type II fiber recruitment. However, Krustrup and colleagues Nevertheless, it is possible that metabolites accumulated to
(16) recently demonstrated that blood flow restriction during such a degree during and after the exercise with cuff
pure concentric knee extensor exercise at a low workload (29 occlusion that it may have impacted on the activation of
Watts) resulted in a lowered creatine phosphate content in the muscle, thus explaining the differences observed in the
93% of the fast-twitch fibers, which strongly supports that the third set.
majority of type II fibers can be recruited at low loads during Although we did not measure blood flow in this study, the
ischemic conditions. endurance data, along with feedback from the subjects,
It is well known that muscle blood flow during both static suggested that the occlusion was indeed affecting the muscle.
and dynamic exercise occurs mainly during the relaxation In fact, several of the subjects experienced difficulty in starting
period between contractions (5,18). By eliminating the the first repetition in the second and third sets with the
relaxation of the muscle between repetitions in our protocol, occluded leg, but once they had managed to do the first
the muscle blood flow likely became insufficient to meet the repetition, they were usually able to do several more.
metabolic demands, and consequently not only the cuff- Furthermore, we determined in separate pilot experiments
that 100 mm Hg in combination with our relatively wide cuff factors and hormones in the blood flow restricted condition.
reduced blood flow in the femoral artery at rest by On the other hand, the greater amount of work and the
approximately 50–60% in the seated position, as measured slightly higher muscle activity during the eccentric phase
by Doppler ultrasound. Takano et al. (24) used a narrow could in theory lead to greater adaptations for the non-
tourniquet (33 mm in width) and pressures of 160 to 180 mm occluded regime. Finally, the findings of the present study
Hg and reported an approximately 70% reduction of the concern the dynamic knee extension exercise with variable
femoral blood flow at rest in the seated position. Wider cuffs resistance and may not necessarily be generalizable to other
are generally more effective in occluding the blood flow than exercises or other muscle groups.
are more narrow ones (9,14). Consequently, we were able to
use a relatively low pressure to minimize compressive and PRACTICAL APPLICATIONS
shear forces on soft tissues, in particular nerves and blood Cuff occlusion decreases the muscle endurance at low loads
vessels (9,14). but does not necessarily result in greater maximal muscle
As in our previous study (33), multiple sets of low-intensity activity in the quadriceps in the knee extension exercise if
training to fatigue induced DOMS. However, the soreness performed to the point of concentric torque failure. The knee
in the nonoccluded leg was significantly greater than for the extension exercise in a variable resistance machine appears to
occluded leg, which may be explained by the greater number be sufficient in itself to induce relative ischemia already at low
of repetitions completed for the leg that trained without loads if performed in a continuous manner without relaxation
blood flow restriction. An important underlying factor of the muscle between the repetitions. The high degree of
behind the DOMS may be the relatively high increase in muscle activation suggests that these methods may be useful
activation observed during the eccentric phases from the in rehabilitation (e.g., after sports injuries), in which large
beginning to the end of each set, as discussed previously. The loads sometimes may be contraindicated. However, because
ischemia-reperfusion and the formation of reactive oxygen it is rather painful to train close to concentric torque failure at
species (ROS) occurring during exercise could also play these low loads regardless of whether a pressure cuff is used or
a role (29–31). Intriguingly though, Goldfarb and colleagues not, the applicability of these methods may be limited to
(13) reported that plasma protein carbonyl levels, a marker of highly motivated individuals. Furthermore, because multiple
ROS-mediated damage, were significantly less elevated after sets of fatiguing dynamic knee extensions at these low loads
blood flow restricted exercise compared with conventional often lead to DOMS, the volume or level of effort should
resistance training. probably be limited initially.
Because studies have demonstrated hypertrophy
(1,2,12,17,28), and increases in muscle protein synthesis ACKNOWLEDGMENTS
(MPS) (11), with low-intensity (20% of 1RM) occlusion This study was supported by a grant from the Swedish
training, and because the femoral blood flow may become National Centre for Research in Sports. The authors thank
insufficient already at low loads during continuous dynamic the subjects for their time and effort and Zimmer Sweden for
knee extensions (21), we previously speculated that it is providing the automatic tourniquet system. The authors
possible to induce hypertrophy also without cuff occlusion have no conflicts of interest that are directly relevant to the
at low intensities (approximately 20–30% of 1RM) in this content of this study. The results of the present study do not
exercise if performed in a no-relaxation manner (32). In line constitute endorsement of the product by the authors or the
with this hypothesis, Burd et al. (7) recently reported in NSCA.
a preliminary study that 4 sets to failure at 30% of 1RM
induced increases in mixed MPS to a similar extent as 4 sets REFERENCES
at 90% of 1RM, whereas performing 4 sets at 30% of 1RM 1. Abe, T, Kawamoto, K, Yasuda, T, Kearns, CF, Midorikawa, T, and
with submaximal effort resulted in markedly less stimulation Sato, Y. Eight days Kaatsu resistance training improved sprint but
of MPS. Future studies should investigate whether the acute not jump performance in collegiate male track and field athletes.
Int J Kaatsu Training Res 1: 23–28, 2005. Available at: http://
MPS increases induced by fatiguing low-intensity exercise www.jstage.jst.go.jp/browse/ijktr/_vols. Accessed April 2005.
without occlusion will translate into increases in muscle mass
2. Abe, T, Yasuda, T, Midorikawa, T, Sato, Y, Kearns, CF, Inoue, K,
and strength after a training period. Koizumi, K, and Ishii, N. Skeletal muscle size and circulating
The major limitation of the present study is that this was an IGF-1 are increased after two weeks of twice daily kaatsu
acute experiment and not a training study. Therefore, some resistance training. Int J Kaatsu Training Res 1: 7–14, 2005.
Available at: http://www.jstage.jst.go.jp/browse/ijktr/_vols.
caution is obviously warranted when interpreting the Accessed April 2005.
findings. Furthermore, we recognize that we measured only 3. Alkner, BA, Tesch, PA, and Berg, HE. Quadriceps EMG/force
a few variables (muscle activity and work) of the many that relationship in knee extension and leg press. Med Sci Sports Exerc
may influence the hypertrophic response with this type of 32: 459–463, 2000.
training, as reviewed by Wernbom et al. (32) and Manini and 4. Andersen, LL, Magnusson, SP, Nielsen, M, Haleem, J, Poulsen, K,
and Aagaard, P. Neuromuscular activation in conventional
Clark (19). For example, there may have been a greater build- therapeutic exercises and heavy resistance exercises: implications for
up of local metabolites and a larger response of growth rehabilitation. Phys Ther 86: 683–697, 2006.
the TM
5. Bangsbo, J and Hellsten, Y. Muscle blood flow and oxygen uptake in 21. Shoemaker, JK, Hodge, L, and Hugson, RL. Cardiorespiratory
recovery from exercise. Acta Physiol Scand 162: 305–312, 1998. kinetics and femoral artery blood velocity during dynamic knee
6. Borg, G. Borg’s Perceived Exertion and Pain Scales. Champaign, IL: extension exercise. J Appl Physiol 77: 2625–2632, 1994.
Human Kinetics, 1998. 22. Staron, RS, Malicky, ES, Leonardi, MJ, Falkel, JE, Hagerman, FC,
7. Burd, NA, West, DW, Staples, AW, Holwerda, AM, Moore, DR, and Dudley, GA. Muscle hypertrophy and fast fiber type conversions
Tang, JE, Baker, SK, and Phillips, SM. Influence of muscle in heavy resistance-trained women. Eur J Appl Physiol 60: 71–79,
1990.
contraction intensity and fatigue on muscle protein synthesis
(MPS) following resistance exercise [Abstract]. Med Sci Sports Exerc 23. Tagesson, SK and Kvist, J. Intra- and interrater reliability of the
41(5 Suppl): 149, 2009. establishment of one repetition maximum on squat and seated knee
extension. J Strength Cond Res 21: 801–807, 2007.
8. Cook, SB, Clark, BC, and Ploutz-Snyder, LL. Effects of exercise load
and blood-flow restriction on skeletal muscle function. Med Sci Sports 24. Takano, H, Morita, T, Iida, H, Asada, K, Kato, M, Uno, K, Hirose, K,
Exerc 39: 1708–1713, 2007. Matsumoto, A, Takenaka, K, Hirata, Y, Eto, F, Nagai, R, Sato, Y, and
Nakajima, T. Hemodynamic and hormonal responses to a short-
9. Crenshaw, AG, Hargens, AR, Gershuni, DH, and Rydevik, B. Wide
term low-intensity resistance exercise with the reduction of muscle
tourniquet cuffs more effective at lower inflation pressures. Acta
blood flow. Eur J Appl Physiol 95: 65–73, 2005.
Orthop Scand 59: 447–451, 1988.
25. Takarada, Y, Nakamura, Y, Aruga, S, Onda, T, Miyazaki, S, and Ishii, N.
10. Deschenes, MR and Kraemer, WJ. Performance and physiologic
Rapid increase in plasma growth hormone after low-intensity
adaptations to resistance training. Am J Phys Med Rehabil resistance exercise with vascular occlusion. J Appl Physiol 88: 61–65,
81(11 Suppl): S3–S16, 2002. 2000.
11. Fujita, S, Abe, T, Drummond, MJ, Cadenas, JG, Dreyer, HC, Sato, Y, 26. Takarada, Y, Sato, Y, and Ishii, N. Effects of resistance exercise
Volpi, E, and Rasmussen, BB. Blood flow restriction during low- combined with vascular occlusion on muscle function in athletes.
intensity resistance exercise increases S6K1 phosphorylation and Eur J Appl Physiol 86: 308–314, 2002.
muscle protein synthesis. J Appl Physiol 103: 903–910, 2007.
27. Takarada, Y, Takazawa, H, Sato, Y, Takebayashi, S, Tanaka, Y, and
12. Fujita, T, Brechue, WF, Kurita, K, Sato, Y, and Abe, T. Increased Ishii, N. Effects of resistance exercise combined with moderate
muscle volume and strength following six days of low-intensity vascular occlusion on muscular function in humans. J Appl Physiol 88:
resistance training with restricted muscle blood flow. Int J Kaatsu 2097–2106, 2000.
Training Res 4: 1–8, 2008. Available at: http://www.jstage.jst.go.jp/
browse/ijktr/_vols. Accessed December 2008. 28. Takarada, Y, Tsuruta, T, and Ishii, N. Cooperative effects of exercise
and occlusive stimuli on muscular function in low-intensity
13. Goldfarb, AH, Garten, RS, Chee, PD, Cho, C, Reeves, GV, resistance exercise with moderate vascular occlusion. Jpn J Physiol
Hollander, DB, Thomas, C, Aboudehen, KS, Francois, M, and 54: 585–592, 2004.
Kraemer, RR. Resistance exercise effects on blood glutathione status
and plasma protein carbonyls: Influence of partial vascular occlusion. 29. Tanimoto, M, Madarame, H, and Ishii, N. Muscle oxygenation and
Eur J Appl Physiol 104: 813–819, 2008. plasma growth hormone concentration during and after resistance
exercise: Comparison between ‘‘KAATSU’’ and other types of
14. Graham, B, Breault, MJ, McEwen, JA, and Mcgraw, RW. Occlusion regimen. Int J Kaatsu Training Res 1: 51–56, 2005. Available at:
of arterial flow in the extremities at subsystolic pressures through the http://www.jstage.jst.go.jp/browse/ijktr/_vols. Accessed February
use of wide tourniquet cuffs. Clin Orthop 286: 257–261, 1993. 2008.
15. Hoeger, WWK, Hopkins, DR, Barette, SL, and Hale, DF. Re- 30. Uchiyama, S, Tsukamoto, H, Yoshimura, S, and Tamaki, T.
lationship between repetitions and selected percentages of one Relationship between oxidative stress in muscle tissue and weight-
repetition maximum: A comparison between untrained and trained lifting-induced muscle damage. Pflu¨gers Arch 452: 109–116, 2006.
males and females. J Appl Sports Sci Res 4: 47–54, 1990.
31. Volek, JS, Kraemer, WJ, Rubin, MR, Gomez, AL, Ratamess, NA,
16. Krustrup, P, Söderlund, K, Relu, MU, Ferguson, RA, and Bangsbo, J. and Gaynor, P. L-carnitine L-tartrate supplementation affects
Heterogeneous recruitment of quadriceps muscle portions and fibre markers of recovery from exercise stress. Am J Physiol Endocrinol
types during moderate intensity knee-extensor exercise: effect of Metab 282: E474–E482, 2002.
thigh occlusion. Scand J Med Sci Sports 19: 576–584, 2009.
32. Wernbom, M, Augustsson, J, and Raastad, T. Ischemic strength
17. Kubo, K, Komuro, T, Ishiguro, N, Sato, Y, Ishii, N, Kanehisa, H, and training: a low-load alternative to heavy resistance exercise? Scand J
Fukunaga, T. Effects of low-load resistance training with vascular Med Sci Sports 18: 401–416, 2008.
occlusion on the mechanical properties of muscle and tendon. J Appl
33. Wernbom, M, Augustsson, J, and Thomeé, R. Effects of vascular
Biomech 22: 112–119, 2006.
occlusion on muscular endurance in dynamic knee extension
18. Lutjemeier, BJ, Miura, A, Scheuermann, BW, Koga, S, Townsend, DK, exercise at different submaximal loads. J Strength Cond Res
and Barstow, TJ. Muscle contraction-blood flow interactions during 20: 372–377, 2006.
upright knee extension exercise in humans. J Appl Physiol 98: 1575–
34. Yasuda, T, Brechue, WF, Fujita, T, Sato, Y, and Abe, T. Muscle
1583, 2005. activation during low-intensity muscle contractions with varying
19. Manini, TM and Clark, BC. Blood flow restricted exercise and levels of external limb compression. J Sports Sci Med 7: 467–474,
skeletal muscle health. Exerc Sport Sci Rev 37: 78–85, 2009. 2009.
20. Rainoldi, A, Melchiorri G, and Carusi, I. A method for positioning 35. Yasuda, T, Brechue, WF, Fujita, T, Shirakawa, J, Sato, Y, and Abe, T.
electrodes during surface EMG recordings in lower limb muscles. Muscle activation during low-intensity muscle contractions with
J Neurosci Methods 134: 37–43, 2004. restricted blood flow. J Sports Sci 27: 479–489, 2009.