Blood Flow Restriction Training Applied

With High-Intensity Exercise Does Not

Improve Quadriceps Muscle Function After
Anterior Cruciate Ligament Reconstruction
A Randomized Controlled Trial
Michael T. Curran,* MEd, ATC, Asheesh Bedi,y MD, Christopher L. Mendias,yz§ PhD, ATC,
Edward M. Wojtys,y MD, Megan V. Kujawa,* and Riann M. Palmieri-Smith,*y|| PhD, ATC
Investigation performed at the University of Michigan, Ann Arbor, Michigan, USA

Background: A major goal of rehabilitation after anterior cruciate ligament reconstruction (ACLR) is restoring quadriceps muscle
strength. Unfortunately, current rehabilitation paradigms fall short of this goal, such that substantial quadriceps muscle strength
deficits can limit return to play and increase the risk of recurrent injuries. Blood flow restriction training (BFRT) involves the
obstruction of venous return to working muscles during exercise and may lead to better recovery of quadriceps muscle strength
after ACLR.
Purpose: To examine the efficacy of BFRT with high-intensity exercise on the recovery of quadriceps muscle function in patients
undergoing ACLR.
Study Design: Randomized controlled trial; Level of evidence, 2.
Methods: A total of 34 patients (19 female, 15 male; mean age, 16.5 6 2.7 years; mean height, 169.0 6 19.7 cm; mean weight,
73.2 6 17.7 kg) scheduled to undergo ACLR were randomly assigned to 1 of 4 groups: concentric (n = 8), eccentric (n = 8), con-
centric with BFRT (n = 9), and eccentric with BFRT (n = 9). The exercise component of the intervention consisted of patients per-
forming a single-leg isokinetic leg press, at an intensity of 70% of the patients’ 1-repetition maximum during either the concentric
or eccentric action, for 4 sets of 10 repetitions 2 times per week for 8 weeks beginning at 10 weeks postoperatively. Patients
randomized to the BFRT groups performed the leg-press exercise with a cuff applied to the thigh, set to a limb occlusion pressure
of 80%. Isometric and isokinetic (60 deg/s) quadriceps peak torque, quadriceps muscle activation, and rectus femoris muscle
volume were assessed before ACLR, after BFRT, and at the time that patients returned to activity and were converted to the
change in values from baseline for analysis. Also, 1-way analyses of covariance were used to compare the change in values
for each dependent variable between groups after BFRT and at return to activity (P  .05).
Results: No significant differences were found between groups for any outcome measures at either time point (P . .05).
Conclusion: An 8-week BFRT plus high-intensity exercise intervention did not significantly improve quadriceps muscle strength,
activation, or volume. On the basis of our findings, the use of BFRT in conjunction with high-intensity resistance exercise in pa-
tients undergoing ACLR to improve quadriceps muscle function may not be warranted.
Registration: NCT03141801 (ClinicalTrials.gov identifier)
Keywords: KAATSU; occlusion training; knee; vascular occlusion; blood flow restriction

Anterior cruciate ligament (ACL) reconstruction (ACLR) is and associated surgical intervention result in a significant
a common orthopaedic procedure performed in the United loss of quadriceps muscle strength that often persists for
States.20 After an ACL rupture, ACLR is the recommended a long time after patients are released from formal postop-
path of care to restore knee joint stability, but the injury erative rehabilitation protocols.24,26,40,41 Lingering defi-
ciencies in quadriceps muscle strength are problematic,
as they are associated with poor functional performance,7
The American Journal of Sports Medicine biomechanical asymmetry,8,30 lower patient-reported out-
2020;48(4):825–837
comes,31 and the early onset of posttraumatic osteoarthri-
DOI: 10.1177/0363546520904008
Ó 2020 The Author(s) tis.39 Participation in arduous physical activity with

825
826 Curran et al The American Journal of Sports Medicine

quadriceps muscle weakness increases the risk of ACL greater improvements in muscle volume, cross-sectional
reinjuries; however, if quadriceps strength symmetry is area, and strength are noted.5 A study by Lepley and col-
recovered to 90%, a reduction in ACL reinjury rates has leagues15 found that eccentric exercise improved quadriceps
been observed.7 muscle activation and isometric strength in patients under-
Blood flow restriction training (BFRT) is a muscle going ACLR in conjunction with standard-of-care rehabilita-
strengthening technique in which a tourniquet is applied to tion when compared with those in the ACLR population who
the proximal aspect of the muscle to be trained to compress undergo standard-of-care rehabilitation alone. However,
the underlying vascular structures during exercise.18,35 Com- despite the improvements found with eccentric exercise,
pression of the vascular structures allows arterial blood flow patients did not achieve clinically recommended (90%) lev-
into the limb but restricts venous return.28 BFRT is thought els of quadriceps strength symmetry.15 BFRT may be able to
to stimulate muscle growth and improvements in strength by enhance the efficacy of high-intensity resistance exercise that
several mechanisms, with some of the primary being (1) met- is traditionally used in ACL rehabilitation. A study in
abolic accumulation that encourages increases in anabolic healthy adults found improvements in squat strength when
growth factors, (2) enhanced fast-twitch fiber recruitment, BFRT was utilized alongside high-intensity exercise.4 If
and (3) increased protein synthesis through the rapamycin BFRT can enhance strength and muscle growth throughout
(mTOR) pathway or decreases in myostatin expression.13,18 ACL rehabilitation, it could prove invaluable to restoring
Quadriceps muscle weakness and atrophy after ACLR are the muscle function necessary to help patients safely return
associated with decreased fast-twitch fiber frequency and to sport and physical activity and thus deserves further
recruitment21,22 as well as increased myostatin expression,27 study.
and thus, BFRT may be an effective intervention to target The purpose of the current study was to incorporate
the muscular deficits apparent after ACLR. BFRT alongside high-intensity concentric or eccentric exer-
Previous studies have shown promising effects of BFRT cise into standard-of-care ACL rehabilitation and examine
when applied to healthy participants, including increased its effects on the recovery of quadriceps muscle strength,
muscle volume,1,17,35 increased muscle strength,19,35 and activation, and atrophy. We hypothesized that those partic-
improved muscle activation.36 Studies to date on BFRT ipants assigned to the BFRT groups would have greater
in patients undergoing ACLR have reported mixed results, quadriceps muscle strength and activation as well as less
with some showing an improved recovery of strength23 and muscle atrophy than patients not receiving BFRT.
muscle size23,37 and another showing no effect on muscle
size recovery.9 Additionally, all of the studies examining
BFRT in patients undergoing ACLR have applied the METHODS
intervention while the patients were exercising at a low
intensity (exercises performed with no weight added or Study Design
exercises performed under body weight). One of the pro-
posed benefits of BFRT is the ability to exercise at a low This study was a randomized clinical trial (NCT03141801)
intensity but achieve strength/hypertrophy improvements in which patients scheduled to undergo ACLR were ran-
similar to those in patients undergoing high-intensity domly assigned, using block randomization, into 1 of 4
exercise. As such, BFRT is an attractive intervention for intervention groups: (1) concentric exercise only, (2) eccen-
patients early after knee injuries and surgery when high- tric exercise only, (3) concentric exercise with BFRT, and
intensity exercise cannot be completed because of tissue (4) eccentric exercise with BFRT. The study intervention
healing and pain. However, quadriceps muscle weakness was delivered for 8 weeks beginning at 10 weeks postoper-
is persistent long after patients who have undergone atively. Neither participants nor outcome assessors were
ACLR return to play, and therefore, the application of blinded to treatment assignment.
BFRT may not only be beneficial early after surgery but Participants completed study testing at 4 time points:
may be useful to apply alongside high-intensity resistance (1) preoperative (within 2 weeks before surgery), (2) prein-
exercise later in the rehabilitation process. tervention (before starting the study intervention), (3)
High-intensity eccentric exercise has been shown to postintervention (within 2 weeks after completing the
improve muscle atrophy and force production.12 When eccen- study intervention), and (4) return to activity (RTA; after
tric exercise is added to standard-of-care ACL rehabilitation, the participant’s clearance to return to activity by the

||
Address correspondence to Riann M. Palmieri-Smith, PhD, ATC, School of Kinesiology, University of Michigan, 4745G CCRB, 401 Washtenaw Avenue,
Ann Arbor, MI 48109-2214, USA (email: riannp@umich.edu) (Twitter: @riannps).
*School of Kinesiology, University of Michigan, Ann Arbor, Michigan, USA.
y
Department of Orthopaedic Surgery, University of Michigan, Ann Arbor, Michigan, USA.
z
Hospital for Special Surgery, New York, New York, USA.
§
Department of Physiology and Biophysics, Weill Cornell Medicine, Cornell University, New York, New York, USA.
Submitted May 31, 2019; accepted December 4, 2019.
Presented at the annual meeting of the AOSSM, Boston, Massachusetts, July 2019.
One or more of the authors has declared the following potential conflict of interest or source of funding: Funding for this study was provided by the
University of Michigan Mcubed Program. A.B. has received consulting fees from Flexion Therapeutics, Smith & Nephew, and Stryker and royalties from
Smith & Nephew. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investiga-
tion on the OPD and disclaims any liability or responsibility relating thereto.
AJSM Vol. 48, No. 4, 2020 Blood Flow Restriction Training and ACLR 827

Figure 1. Study design illustrating the flow of patients through the study protocol and outcome measurements recorded at
respective time points. BFR, blood flow restriction; PRO, patient-reported outcome.

supervising physician) (Figure 1). Not all outcome meas- severe hypertension. Additionally, any potential female
ures were assessed at the preintervention time point patient who was pregnant or planning to become pregnant
because of concerns surrounding knee pain. Our coprimary over the next year was excluded from participation. ACLR
outcome measures were the change in isometric and isoki- for all participants was performed by 1 of 4 orthopaedic
netic quadriceps muscle strength from preoperative to surgeons from our local sports medicine clinic. All patients
postintervention and from preoperative to RTA. Our sec- and their parents/guardians (if applicable) were required
ondary outcome measures were the change in rectus femo- to read and sign the informed consent form before study
ris muscle volume and central activation ratio (CAR) from participation. This study was approved by the University
preoperative to postintervention and from preoperative to of Michigan Medical Institutional Review Board.
RTA. Tertiary outcome measures included the (1) change
in the rectus femoris muscle volume from preintervention Participants
to postintervention, (2) change in the International Knee
Documentation Committee (IKDC) score from preoperative There were 48 patients who were scheduled to undergo
to postintervention, (3) change in the IKDC score from ACLR at Michigan Medicine with either a bone–patellar
preintervention to postintervention, (4) change in the 1- tendon–bone autograft, semitendinosus tendon autograft,
repetition maximum (1RM) leg press from preintervention or quadriceps tendon autograft and completed their post-
to postintervention, and (5) change in the IKDC score from operative rehabilitation at our local sports medicine clinic;
preoperative to RTA. The randomization sequence was they were recruited, were enrolled, and underwent base-
generated by the principal investigator (R.M.P.-S.), and line testing. After ACLR, 10 patients ceased participation
the assignment was placed in sealed envelopes that were in or dropped out of the study (Figure 2). An additional 4
given to the study team member providing the treatments participants were excluded between postintervention and
(M.T.C.). RTA testing: 2 because of additional injuries and 2 because
of failure to return for follow-up. A total of 34 patients (19
Eligibility Criteria female, 15 male; mean age, 16.50 6 2.69 years; mean
height, 168.95 6 19.68 cm; mean weight, 73.22 6 17.70
Patients were eligible for inclusion in the study if they kg) participated in the 3 testing sessions and the study
were aged between 14 and 30 years and were willing to intervention as outlined in the experimental protocol
participate in intervention and follow-up testing as out- (Table 1). There was 1 patient who declined to undergo
lined in the protocol. Participants were excluded if they muscle activation testing at the postintervention and
had (1) undergone prior knee surgery, (2) a previous ACL RTA time points because of discomfort with the procedure.
injury, (3) received surgical interventions on a knee liga- At the time that this investigation began, little data
ment other than the ACL, (4) a cardiac demand-type pace- existed on the effects of BFRT on muscle strength in
maker, (5) a history of blood clots or deep vein thrombosis, injured populations; thus, we powered our investigation
(6) a history of cerebrovascular disease, (7) been using to find differences between standard-of-care ACL rehabili-
estrogen or progestin contraceptives, (8) a history of sickle tation and eccentric exercise.15 Given our hypothesis that
cell anemia, (9) a history of diabetes, and (10) a history of the effect of eccentric exercise with BFRT would be larger
828 Curran et al The American Journal of Sports Medicine

Figure 2. CONSORT (Consolidated Standards of Reporting Trials) flow diagram illustrating the flow of participants throughout the
trial. ACL, anterior cruciate ligament; BFRT, Blood flow restriction training; MCL, medial collateral ligament.

than that for eccentric exercise alone, we considered these underwent a leg-press test in which they were expected to
data to be adequate. To detect differences with an a level of perform a minimum of 15 repetitions of a single-leg leg
.05 and 1 – b of 0.80, it was determined that at least 9 press with a load equal to 100% of their body weight.
patients per group would be needed for this investigation Patients must have successfully completed this task before
(mean of 2.9 for eccentric exercise group and 2.1 for starting straight-line running. The final phase of the proto-
standard-of-care group, with a standard deviation of 0.6). col targeted strength maximization, functional activity,
and an agility program that consisted of various movement
Standard-of-Care ACL Rehabilitation activities (running, hopping, cutting, shuttle runs, etc).
The minimal functional criteria for participants to be med-
All patients in our study underwent standard-of-care ACL ically cleared to return to activity were as follows: (1) leg
rehabilitation. The first 4 weeks of the rehabilitation proto- press of 15 single-leg repetitions with a load equal to
col was focused on minimizing pain and swelling, recover- 100% body weight, (2) single-leg forward hop for 95%
ing full range of motion, improving quadriceps muscle maximal distance symmetry, (3) 2-leg deep squat for qual-
control, and achieving full weightbearing. During months itative symmetry, and (4) single-leg deep squat for qualita-
2 through 4, the protocol was focused on increasing tive symmetry. These functional criteria are in addition to
strength of the quadriceps, hamstring, and hip muscula- the physician’s examination for (1) appropriate joint laxity,
ture as well as improving single-leg balance and transition- (2) full knee joint range of motion, and (3) no knee joint
ing to functional exercise. During this time frame, patients effusion.
AJSM Vol. 48, No. 4, 2020 Blood Flow Restriction Training and ACLR 829

TABLE 1
Participant Characteristics for All Intervention Groupsa
Time From Time From
Concomitant Time to Surgery to Surgery to
Group n Age, y Sex Height, cm Weight, kg Graft Type Procedure Surgery, d Postintervention, d RTA, d

Concentric 8 16.1 6 2.6 3 M; 5 F 158.7 6 12.1 69.9 6 11.7 7 BPTB; 0 0 MMR; 1 LMR; 7.00 6 5.56 158.63 6 17.78 295.25 6 52.20
STG; 1 QT 0 MMX; 1 LMX;
0 CMB; 1 other
Eccentric 8 18.8 6 3.9b 2 M; 6 F 169.3 6 8.9 75.7 6 24.3 5 BPTB; 1 MMR; 2 LMR; 4.25 6 2.32 151.13 6 13.44 278.13 6 65.74
3 STG; 0 QT 0 MMX; 0 LMX;
0 CMB; 0 other
Concentric 9 15.3 6 0.9 5 M; 4 F 173.8 6 9.4 79.0 6 20.4 6 BPTB; 2 MMR; 1 LMR; 7.11 6 3.14 147.78 6 17.39 281.00 6 53.46
with BFRT 1 STG; 2 QT 0 MMX; 0 LMX;
1 CMB; 1 other
Eccentric 9 16.0 6 1.7 5 M; 4 F 172.9 6 11.9 68.2 6 12.5 7 BPTB; 0 MMR; 3 LMR; 5.44 6 4.45 147.78 6 16.95 279.67 6 47.23
with BFRT 2 STG; 0 QT 0 MMX; 2 LMX;
0 CMB; 1 other

a
Data are reported as mean 6 SD unless otherwise indicated. BFRT, blood flow restriction training; BPTB, bone–patellar tendon–bone; CMB, multiple menis-
cal intervention; F, female; LMR, lateral meniscal repair; LMX, lateral meniscectomy; M, male; MMR, medial meniscal repair; MMX, medial meniscectomy;
other, meniscal abrasion or debridement; QT, quadriceps tendon; RTA, return to activity; STG, semitendinosus-gracilis tendon.
b
Denotes a significance: P \ .05.

Study Interventions 1RM. Sets 2 through 5 were set to specific intensities

depending on the group affiliation. Those participants who
All participants, regardless of group assignment, com- were randomly assigned to groups with concentric strength-
pleted their muscle strengthening exercise with or without ening (concentric and concentric with BFRT) performed sets
BFRT utilizing a single-leg leg press, with the surgical 2 through 5 of the leg press with the concentric intensity
limb only, on the BLAST! Leg Press system (version 1.2; (the push-out/knee extension phase of the leg press) set to
Bio Logic Engineering). Participants were seated on the 70% of their 1RM and the eccentric intensity (the knee flex-
BLAST! device with their back firmly contacting the seat ion phase of the leg press) maintained at 20%. Conversely,
back, surgical limb bent to 60° of knee flexion, and foot those participants who were randomly assigned to groups
placed on the foot plate so that the hip, knee, and ankle with eccentric strengthening (eccentric and eccentric with
were aligned. The range of motion for the single-leg BFRT) performed sets 2 through 5 of the leg press with
leg press was set for 60° to 20° of knee flexion. Once prop- the eccentric intensity set to 70% of the 1RM and the con-
erly positioned, participants were asked to perform 3 centric intensity maintained at 20%. The eccentric protocol
concentric-only, 1RM knee extension efforts with 2 described herein has previously been shown in our labora-
minutes of rest between each repetition. The best repeti- tory to increase muscle strength after ACLR when com-
tion (highest maximal leg-press value in foot-pounds) was pared with standard of care.14,15
recorded as the 1RM and saved as 100% maximal intensity Patients who were randomly assigned to receive the BFRT
for the participant. Participants performed the 1RM intervention underwent BFRT while they performed either
assessment on the first day of the intervention and then the aforementioned concentric- or eccentric-focused leg press.
once every week of the intervention, or every alternate Before initiating the exercise, participants had an Easi-Fit
intervention session. The 1RM was then used to adjust Tourniquet Cuff (Delfi Medical Innovations) applied to the
the intensity of the intervention each week so as to pro- proximal thigh of the surgical limb. Participants were asked
gressively increase the resistance of the exercise as the to lie supine on a treatment table, in a relaxed position, while
patients became stronger (ie, the 1RM assessed in week 1 the PTS Personalized Tourniquet System (Delfi Medical Inno-
was used to set the intensity of the exercises for the 2 treat- vations) was used to assess 80% of full limb occlusion pressure
ment sessions in week 1; in week 2, the 1RM was reas- (in millimeters of mercury). Once 80% limb occlusion pressure
sessed and used to set the intensity of the exercise for was assessed, it was saved for use during the leg press. For
the treatment sessions in week 2; etc). The 1RM gathered sets 2 through 5 of the leg press, patients in the BFRT groups
before the intervention started and after the intervention had the tourniquet cuffs inflated to their previously recorded
was completed was also used as a tertiary outcome mea- 80% limb occlusion pressure before performing each set.13
sure to determine if changes occurred between groups. The tourniquet cuffs were deflated at the completion of each
After completion of the 1RM, participants then per- set, remained deflated during the 2-minute rest period
formed 5 sets of 10 repetitions of the single-leg leg press between sets, and were re-inflated just before the commence-
with 2 minutes of rest between sets. For both the concentric ment of the next exercise set.
and the eccentric groups, the first of the 5 sets of the leg Participants in all intervention groups received 2 study
press were a warm-up, with the concentric and eccentric treatments per week for 8 weeks, for a total of 16 treatment
intensity set to 20% of the participants’ recorded voluntary sessions.
830 Curran et al The American Journal of Sports Medicine

Quadriceps Muscle Strength and Activation torque and then multiplying by 100 (equation 3).10 The
CAR, which represents the percentage of voluntary muscle
To measure quadriceps muscle activation with the super- activation, was also converted to the change in values from
imposed burst (SIB) technique, participants were seated preoperative to postintervention (equation 1) and RTA
on the Biodex System 3 dynamometer (Biodex Medical Sys- (equation 2) and used in statistical analysis.
tems) with the hip and knee of the testing limb flexed to
 
90°. Participants were secured in the proper position MVIC
with straps across the torso, waist, thigh, and shank. Elec- CAR5  100: ð3Þ
MVIC1Superimposed Burst
trodes (Dura-Stick II [7 3 13 cm]; Chattanooga Group)
were placed on the distal vastus medialis and proximal
vastus lateralis of the testing limb. Participants were
then instructed on how to perform the isometric knee Muscle Volume
extension exercise and completed 1 set of 3 practice repeti-
tions at a submaximal effort. After the practice set, partic- The rectus femoris muscle volume was assessed utilizing
ipants were given 2 minutes of rest and then instructed to musculoskeletal ultrasound (GE LOGIQ e; GE Healthcare).
perform maximal voluntary isometric contractions Ultrasound has shown good accuracy compared with mag-
(MVICs), with a minimum of 3 repetitions, until no further netic resonance imaging in detecting quadriceps muscle atro-
increase in torque was observed. Maximal knee extension phy and hypertrophy (sensitivity: 73.7%; specificity:
torque during the MVIC was recorded with a custom- 74.2%).34 Participants were positioned supine on a treatment
written program (LabVIEW version 8.5; National Instru- table, in a relaxed position, with a 5 inch–diameter bolster
ments). Participants were then given a 2-minute rest pad placed under the knee of the limb to be imaged. Bilateral
period between the MVIC and SIB trials. For the SIB tri- measurements were obtained from the superior patella to the
als, participants performed 3 repetitions of MVICs, during ipsilateral anterior superior iliac spine, and marks were
which a supramaximal stimulus (Model S88 Dual Output made on the thigh 10 cm and 15 cm superior to the patella.
Square Pulse Stimulator/Model SIU8T Stimulus Isolation Then, 1 ultrasound image was taken at each marked location
Unit; Grass Technologies) was delivered to the quadriceps on each leg, with a predetermined image depth of 3.5 cm. All
muscle. The supramaximal stimulus was delivered once images were processed with ImageJ 1.47r software (National
participants had matched the maximal knee extension Institutes of Health). In ImageJ, the polygon tool was used to
torque value from the MVIC and then had their torque mark the outermost margin of the rectus femoris muscle and
drop by 1 Nm. When these parameters were met, the Lab- calculate the cross-sectional area. This process was repeated
VIEW program automatically delivered the electrical stim- 3 times per image, and the average of the 3 cross-sectional
ulus (100 pulses/s, 600-ms pulse duration, 10 pulse train, area values was recorded in volume calculations. To optimize
130 V), as performed in previous studies.15,25 consistent identification of the rectus femoris muscle mar-
Bilateral isokinetic knee extension strength was mea- gins, all images were marked by the same research staff
sured at 60 deg/sec. Similar to isometric testing, partici- member. The rectus femoris muscle volume was calculated
pants were asked to perform 1 set of 3 submaximal using equation 4.
concentric repetitions as a warm-up. Participants were
then given 2 minutes of rest after the warm-up set and Rectus Femoris Volume5
instructed to perform 1 set of 5 maximal concentric knee   
ðCSA10cm1CSA15cmÞ
extension repetitions. The maximal knee extension torque  5cm : ð4Þ
2
of all 5 repetitions was recorded, and the highest value was
used for analysis.
The maximal torque values recorded during the MVIC The change in values from preoperative to postinterven-
and isokinetic knee extension trials were used to calculate tion (equation 1) and RTA (equation 2) for rectus femoris
the change from preoperative to postintervention (equation muscle volume was then calculated and recorded for use
1) and RTA (equation 2) and used in statistical analysis. in statistical analysis.
Post Intervention Change From Baseline Value5
Statistical Analysis
ðPost Intervention Value  Pre Operative ValueÞ: ð1Þ
To assess if the study groups differed on demographic varia-
bles, analyses of variance (ANOVAs; continuous outcomes)
or Kruskal-Wallis tests (categorical outcomes) were utilized.
RTA Change From Baseline Value5 Separate 1-way analyses of covariance (ANCOVAs) were
ðRTA  Pre Operative ValueÞ: ð2Þ used to examine the effects of group (concentric, eccentric, con-
centric with BFRT, and eccentric with BFRT) on our primary
outcomes (change from preoperative to postintervention and
To assess quadriceps muscle activation, we used the RTA for maximal isokinetic knee extension and maximal iso-
peak torque of the 3 MVIC and SIB trials and employed metric knee extension peak torque) and secondary outcomes
the CAR. The CAR was calculated by dividing the MVIC (change from preoperative to postintervention and RTA for
torque by the SIB maximal isometric knee extension muscle activation [CAR] and muscle atrophy [rectus femoris
AJSM Vol. 48, No. 4, 2020 Blood Flow Restriction Training and ACLR 831

TABLE 2
Change in Primary and Secondary Outcome Measures for All Intervention Groupsa

Preoperative to Postintervention Preoperative to RTA

Isokinetic Rectus Isokinetic Rectus

Strength Femoris Muscle Strength Femoris Muscle
Group n D, Nm MVIC D, Nm CAR D Volume D, cm3 D, Nm MVIC D, Nm CAR D Volume D, cm3

Concentric 8 –19.2 6 35.9 –13.7 6 42.6 –3.8 6 11.6 –3.1 6 3.5 –1.5 6 26.9 –9.6 6 24.2 –0.2 6 10.3 –1.5 6 4.0
Eccentric 8 –10.8 6 34.7 –10.0 6 36.3 –3.0 6 9.2 –2.3 6 4.3 2.4 6 28.3 3.1 6 35.4 0.3 6 4.7 –0.2 6 3.2
Concentric with BFRT 9 –16.7 6 21.4 –18.0 6 34.5 –5.7 6 10.0 –1.8 6 2.6 9.6 6 43.2 –0.7 6 52.2 –6.9 6 9.5 –0.6 6 3.8
Eccentric with BFRT 9 –8.1 6 16.9 –14.6 6 29.3 0.2 6 7.0 –1.5 6 2.4 6.9 6 22.0 –4.9 6 40.7 2.3 6 9.6 –0.5 6 4.1

a
Data are reported as mean 6 SD. Negative values indicate that the dependent variable decreased from the preoperative time point. No
significant differences were noted between groups for any dependent variable at any time point. BFRT, blood flow restriction training; CAR,
central activation ratio; MVIC, maximal voluntary isometric contraction; RTA, return to activity.

muscle volume]). We also utilized ANCOVAs to examine the deviation for all study variables for the 4 study groups can
effects of group on our tertiary outcomes: (1) change in the rec- be found in Appendix Tables A1 and A2 (available in the
tus femoris muscle volume from preintervention to postinter- online version of this article).
vention, (2) change in the IKDC score from preoperative to We found no significant differences between groups in
postintervention, (3) change in the IKDC score from preinter- the change from preoperative to postintervention for max-
vention to postintervention, (4) change in the 1RM leg press imal isokinetic knee extension (F3,33 = 0.57; P = .64), max-
from preintervention to postintervention, and (5) change in imal isometric knee extension (F3,33 = 0.05; P = .98), CAR
the IKDC score from preoperative to RTA. A secondary anal- (F3,32 = 0.53; P = .67), and rectus femoris muscle volume
ysis was also performed in which exercise type (eccentric and (F3,33 = 0.28; P = .84) (Table 2). Additionally, there were
concentric) was no longer considered as an independent factor, no significant between-group differences in the change
and patient data were collapsed into groups, which only con- from preoperative to RTA for maximal isokinetic knee
sidered whether they underwent BFRT. For this analysis, 1- extension (F3,33 = 0.58; P = .63), maximal isometric knee
way ANCOVAs were again utilized, but BFRT was considered extension (F3,33 = 0.38; P = .77), CAR (F3,32 = 1.14; P =
as our independent variable (BFRT or non-BFRT group). The .35), and rectus femoris muscle volume (F3,33 = 0.11; P =
covariate for the analysis was the preoperative/baseline value .95) (Table 2). No significant differences were noted in
of the dependent variable included in the ANCOVA model (ie, the change from preintervention to postintervention for
in models where isokinetic strength values were the depen- the IKDC score (F3,33 = 0.96; P = .42), rectus femoris mus-
dent variable, the covariate was preoperative/baseline isoki- cle volume (F3,32 = 0.36; P = .78), or 1RM (F3,33 = 1.45; P =
netic strength). In addition to the ANCOVAs, the effect size .25) (Table 3). No group differences were found in the
with 95% confidence intervals (CIs) for each variable was change from preoperative to postintervention (F3,33 =
assessed with Cohen d, with thresholds defined as follows: 0.30; P = .83) or the change from preoperative to RTA
small = 0.20, medium = 0.50, and large = 0.80. To ensure (F3,33 = 0.16; P = .93) for the IKDC score (Table 3).
the appropriateness of the secondary analysis, 2-way mixed A moderate effect size was noted for BFRT on the change
ANOVAs were run with BFRT group affiliation and exercise in the CAR from preoperative to RTA (d = 20.68 [95% CI,
group affiliation (concentric or eccentric) as the 2 independent –1.65 to 0.37]), whereas all other effect sizes were small
variables to confirm that there was no interaction between the (Table 4). All effect sizes, despite the strength of the effect,
factors that would affect the findings of the secondary analysis had 95% CIs that crossed 0, suggesting that the effect is
(ie, to confirm that concentric versus eccentric exercise did not not significant or that an effect size of 0 is a possible outcome.
affect the findings between BFRT and non-BFRT usage). The
a priori a level was set at P  .05 for all tests. Statistical anal-
yses were performed with SPSS (version 22; IBM). Secondary Analysis: Effects of BFRT
With High-Intensity Resistance Exercise

RESULTS Given the lack of differences across the 4 intervention

groups, we collapsed participants into 2 groups based on
Primary Analysis: Effects of BFRT those who were randomized to BFRT (concentric with
With High-Intensity Eccentric or Concentric Exercise BFRT and eccentric with BFRT) and no BFRT (concentric
and eccentric). The characteristics for these 2 groups can be
Participants in the 4 study groups (concentric, eccentric, con- found in Table 2. No significant between-group (BFRT and
centric with BFRT, eccentric with BFRT) did not differ in non-BFRT) demographic differences were noted (Table 5).
terms of demographics, with the exception of age (Table 1). Before disregarding exercise type in our analysis, 2-way
Patients in the eccentric group were older than participants ANOVAs considering BFRT status (BFRT vs non-BFRT)
in the other 3 groups (P \ .05). Means and standard and exercise type (concentric vs eccentric) for all outcomes
832 Curran et al The American Journal of Sports Medicine

TABLE 3
Change in Values for Tertiary Study Outcomes for All Intervention Groupsa

IKDC Score D 1RM D Rectus Femoris Muscle Volume D

Preintervention to Preoperative to Preoperative Preintervention Preintervention to

Group n Postintervention Postintervention to RTA to Postintervention Postintervention

Concentric 8 18.54 6 10.54 19.98 6 17.30 33.48 6 13.27 2.32 6 0.90 2.53 6 3.84
Eccentric 8 22.41 6 14.25 15.81 6 18.02 34.06 6 15.98 1.87 6 1.00 2.02 6 2.56
Concentric with BFRT 9 13.81 6 11.43 9.97 6 15.96 28.23 6 7.64 2.94 6 0.96 1.12 6 2.22
Eccentric with BFRT 9 19.17 6 8.02 13.69 6 18.12 27.11 6 19.15 1.94 6 1.28 1.14 6 3.86

a
Data are reported as mean 6 SD. 1RM, 1-repetition maximum; BFRT, blood flow restriction training; IKDC, International Knee Docu-
mentation Committee; RTA, return to activity.

TABLE 4
Effect Sizes for Change in Strength, Activation, and Muscle Volume for All Intervention Groupsa

Time Point Isokinetic Strength MVIC CAR Rectus Femoris Muscle Volume

Concentric vs concentric with BFRT

Preoperative to postintervention 0.09 (–0.87 to 1.03) –0.11 (–1.06 to 0.85) –0.18 (–1.16 to 0.82) 0.41 (–0.58 to 1.35)
Preoperative to RTA 0.30 (–0.67 to 1.25) 0.21 (–0.75 to 1.16) –0.68 (–1.65 to 0.37) 0.22 (–0.74 to 1.17)
Eccentric vs eccentric with BFRT
Preoperative to postintervention 0.10 (–0.86 to 1.05) –0.14 (–1.09 to 0.82) 0.39 (–0.59 to 1.33) 0.24 (–0.73 to 1.18)
Preoperative to RTA 0.18 (–0.79 to 1.12) –0.21 (–1.15 to 0.76) 0.27 (–0.70 to 1.21) –0.11 (–1.05 to 0.85)

a
Data are reported as d (95% CI). BFRT, blood flow restriction training; CAR, central activation ratio; MVIC, maximal voluntary isometric
contraction; RTA, return to activity.

TABLE 5
Participant Characteristics for BFRT and Non-BFRT Groupsa
Time From Time From
Concomitant Time to Surgery to Surgery
Group n Age, y Sex Height, cm Weight, kg Graft Type Procedure Surgery, d Postintervention, d to RTA, d

Non-BFRT 16 17.4 6 3.5 5 M b; 164.0 6 26.1 72.8 6 18.7 12 BPTB; 1 MMR; 3 LMR; 5.63 6 4.35 154.88 6 15.71 286.69 6 58.02
11 F 3 STG; 0 MMX; 1 LMX;
1 QT 0 CMB; 1 other
BFRT 18 15.7 6 1.3 10 M; 173.4 6 10.4 73.6 6 17.3 13 BPTB; 2 MMR; 4 LMR; 6.28 6 3.83 147.78 6 16.66 280.33 6 48.94
8F 3 STG; 0 MMX; 2 LMX;
2 QT 1 CMB; 2 other

a
Data are reported as mean 6 SD unless otherwise indicated. BFRT, blood flow restriction training; BPTB, bone–patellar tendon–bone; CMB, multiple menis-
cal intervention; F, female; LMR, lateral meniscal repair; LMX, lateral meniscectomy; M, male; MMR, medial meniscal repair; MMX, medial meniscectomy;
other, meniscal abrasion or debridement; QT, quadriceps tendon; RTA, return to activity; STG, semitendinosus-gracilis tendon.
b
There were fewer male patients in the non-BFRT group compared with the BFRT group (P \ .05).

were run. The results showed no interaction between (F1,33 = 0.50; P = .49), MVIC (F1,33 = 0.02; P = .88), CAR
BFRT status and exercise type on any of our study varia- (F1,32 = 0.46; P = .50), and rectus femoris muscle volume
bles (P = .12-.99). Given that no interaction effects were (F1,33 = 0.72; P = .40) (Table 6). Similarly, no differences
found in this analysis, we were confident in collapsing were detected in the change from preoperative to RTA for
the groups. This secondary analysis allowed us to deter- maximal isokinetic knee extension (F1,33 = 1.55; P = .22),
mine if BFRT regardless of exercise type (concentric or MVIC (F1,33 = 0.14; P = .71), CAR (F1,32 = 0.27; P = .61),
eccentric) affected our outcome measures. Means and stan- and rectus femoris muscle volume (F1,33 = 0.01; P = .92)
dard deviations for all study variables for the BFRT and (Table 6). No significant differences were found in the
non-BFRT study groups can be found in Appendix Tables change from preintervention to postintervention for the
A3 and A4 (available online). IKDC score (F1,33 = 1.52; P = .23), rectus femoris muscle
There were no differences in the change from preopera- volume (F1,32 = 1.09; P = .31), or 1RM (F1,33 = 1.45; P =
tive to postintervention when comparing the BFRT and .24) between groups (Table 7). Also, no group differences
non-BFRT groups for maximal isokinetic knee extension were identified in the change from preoperative to
AJSM Vol. 48, No. 4, 2020 Blood Flow Restriction Training and ACLR 833

TABLE 6
Change in Primary and Secondary Study Outcomes for BFRT and Non-BFRT Groupsa

Preoperative to Postintervention Preoperative to RTA

Isokinetic Rectus Isokinetic Rectus

Strength Femoris Muscle Strength Femoris Muscle
Group n D, Nm MVIC D, Nm CAR D Volume D, cm3 D, Nm MVIC D, Nm CAR D Volume D, cm3

Non-BFRT 16 –15.0 6 34.4 –11.8 6 38.3 –3.4 6 10.0 –2.7 6 3.8 0.4 6 26.8 –3.2 6 30.0 0.1 6 7.5 –0.8 6 3.6
BFRT 18 –12.4 6 19.2 –16.3 6 31.1 –2.8 6 8.9 –1.7 6 2.5 8.2 6 33.3 –2.8 6 45.5 –2.3 6 10.4 –0.6 6 3.8

a
Data are reported as mean 6 SD. Negative values indicate that the dependent variable decreased from the preoperative time point. No
significant differences were noted between groups for any dependent variable at any time point. BFRT, blood flow restriction training; CAR,
central activation ratio; MVIC, maximal voluntary isometric contraction; RTA, return to activity.

TABLE 7
Change in Values for Tertiary Study Outcomes for BFRT and Non-BFRT Groupsa

IKDC Score D 1RM D Rectus Femoris Muscle Volume D

Preintervention to Preoperative to Preoperative Preintervention to Preintervention to

Group n Postintervention Postintervention to RTA Postintervention Postintervention

Non-BFRT 16 20.48 6 12.27 17.89 6 17.20 33.77 6 14.20 2.10 6 0.95 2.26 6 3.11
BFRT 18 16.49 6 9.97 11.83 6 16.67 27.67 6 14.15 2.44 6 1.21 1.13 6 3.05

a
Data are reported as mean 6 SD. 1RM, 1-repetition maximum; BFRT, blood flow restriction training; IKDC, International Knee Docu-
mentation Committee; RTA, return to activity.

TABLE 8
Effect Sizes for Change in Strength, Activation, and Muscle Volume for BFRT and Non-BFRT Groupsa

Time Point Isokinetic Strength MVIC CAR Rectus Femoris Muscle Volume

Preoperative to postintervention 0.09 (–0.58 to 0.77) –0.13 (–0.80 to 0.55) 0.06 (–0.62 to 0.75) 0.33 (–0.36 to 1.00)
Preoperative to RTA 0.26 (–0.43 to 0.93) 0.01 (–0.66 to 0.68) –0.25 (–0.93 to 0.44) 0.06 (–0.61 to 0.74)

a
Data are reported as d (95% CI). BFRT, blood flow restriction training; CAR, central activation ratio; MVIC, maximal voluntary isometric
contraction; RTA, return to activity.

postintervention (F1,33 = 0.28; P = .60) or change from pre- aerobic or resistance exercise results in increased muscle
operative to RTA (F1,33 = 0.37; P = .55) for the IKDC score strength,19,35 activation,36 and volume1,17 in healthy par-
(Table 7). ticipants. We hypothesized that the ACLR population
A small effect size was noted for BFRT on the change in would see similar results to the healthy participants of pre-
the rectus femoris muscle volume from preoperative to post- vious studies and that patients who received BFRT would
intervention (d = 0.33 [95% CI, –0.36 to 1.00]) and the change see less reduction in quadriceps muscle strength, activa-
in maximal isokinetic knee extension strength (d = 0.26 [95% tion, and volume than the non-BFRT group directly after
CI, –0.43 to 0.93]) and CAR (d = 20.25 [95% CI, –0.99 to completion of the treatment protocol and at RTA. The
0.39]) from preoperative to RTA. No notable effect size was results of our study, however, demonstrate no functional
seen for any other change in variables from preoperative to improvements in quadriceps muscle strength or volume
postintervention or RTA (Table 8). with the addition of BFRT in this group.

DISCUSSION Lack of Differences Between

BFRT and Non-BFRT Groups
The purpose of this study was to examine the effect of
BFRT with high-intensity resistance exercise on quadri- Patients in the BFRT (when concentric or eccentric exercise
ceps muscle strength, activation, and atrophy when uti- type was considered or when collapsed across exercise type)
lized as part of rehabilitation protocols after ACLR. and non-BFRT groups had similar postoperative muscle
Previous studies have shown that BFRT when added to atrophy, strength, and activation at the postintervention
834 Curran et al The American Journal of Sports Medicine

and RTA testing time points. These results suggest that exercise 2 to 3 times to induce strength gains.2 Further,
BFRT delivered alongside high-intensity resistance exercise a meta-analysis showed that BFRT delivered alongside
is not beneficial to quadriceps muscle function after ACLR. low-intensity exercise 2 to 3 times per week produced
The available literature examining the effect of BFRT on larger effects compared with BFRT delivered 4 to 5 days
the recovery of quadriceps muscle function in patients per week.19 This meta-analysis also showed no difference
undergoing ACLR is limited and presents mixed results. in strength gains between BFRT interventions for less
Ohta et al23 noted a significantly smaller postoperative than 4 weeks and 5 to 8 weeks. It is important to note, how-
decline in isokinetic quadriceps muscle strength and cross- ever, that the optimal dosage for BFRT is still to be deter-
sectional area in patients undergoing 15 weeks of low- mined and may differ for different patient populations,
intensity resistance exercise with BFRT compared with could vary depending on the targeted muscle, or could
a group only receiving exercise. A decline in strength from vary depending on the deficits in muscle strength. Further,
preoperative values of 44% was noted in the exercise group while 2 to 3 times per week for 5 to 8 weeks may be effec-
compared with a decline of only 10% in the BFRT group at tive for BFRT with low-intensity exercise, the same may
16 weeks after ACLR. The cross-sectional area was also not be true when delivered along with high-intensity
improved, with an increase of 10% from preoperative values exercise.
at 16 weeks after ACLR noted in the BFRT group, while no Our study is the only one, to our knowledge, to examine
change in the cross-sectional area was found for the BFRT with high-intensity resistance exercise in the ACLR
exercise-only group. Takarada et al37 noted similar benefits population, which likely is the major reason behind the dif-
of BFRT after ACLR, with declines in quadriceps muscle ferences between our work and the previous literature.
atrophy between days 3 and 14 postoperatively being less Our participants completed a leg-press exercise at 70% of
(11% vs 22%) in patients who received BFRT during rest their 1RM, which was selected because of recommenda-
for 50 minutes per day for 10 days. A recent study by Kilgas tions that training at a 1RM of 60% to 70% is necessary
et al11 found increases in muscle hypertrophy and strength to induce muscle hypertrophy.2 Further, we had hypothe-
in 9 patients at an average of 5 years after ACLR who com- sized that the benefits noted with BFRT and low-intensity
pleted BFRT with low-load exercise 5 times per week for 4 resistance exercise would be magnified with high-intensity
weeks. In agreement with our work, Iversen et al9 found resistance exercise, resulting in superior outcomes for
no difference in the change in the quadriceps cross-sectional patients undergoing ACLR. In the abovementioned studies
area in patients undergoing ACLR who received BFRT for noting the beneficial effects of BFRT in patients undergo-
50 minutes per day for 12 days alongside quadriceps-focused ing ACLR, participants underwent BFRT while performing
low-intensity resistance exercise when compared with those a variety of low-load exercises (straight-leg raises, quadri-
not receiving BFRT. ceps sets, etc)23 or at rest with no exercise at all.37 It is pos-
The difference in findings between studies could be sible that BFRT is not effective when delivered alongside
because of a number of study-specific differences, one being high-intensity resistance exercise, as the mechanisms
the limb occlusion pressure utilized. In our study, BFRT leading to improvements in muscle function with BFRT
was delivered at each participant’s 80% limb occlusion are not additive. Muscle hypertrophy induced by resis-
pressure (range of 110-186 mm Hg, with a mean low of tance exercise is considered to result from 2 main mecha-
138 mm Hg and mean high of 167 mm Hg), whereas the nisms: (1) increased mechanical tension and/or (2)
other studies23,37 used predetermined occlusive pressures metabolic stress.32,33 Work by Biazon et al3 showed that
(180 mm Hg or 180-260 mm Hg) that are comparatively while metabolic stress (assessed with blood deoxyhemoglo-
high versus our study, and Iversen et al9 used a similar bin concentration) was greatest during high-intensity
occlusive pressure to that in our investigation (range, resistance exercise with BFRT, muscle hypertrophy and
130-180 mm Hg). It is possible that a limb occlusion pres- strength gains were similar to training protocols with
sure higher than 80% of a patient’s limb occlusion pressure lower levels of metabolic stress (high-intensity resistance
is necessary for a therapeutic effect in patients after exercise only) or mechanical tension (BFRT with low-
ACLR. However, we contend that this is unlikely, as occlu- intensity exercise). The authors suggested that their
sion pressures identical to that in our study have led to results may support that muscle protein synthesis reaches
improvements in quadriceps muscle strength and thigh its peak with high-intensity exercise and that adding met-
girth in patients after knee arthroscopic surgery.38 Fur- abolic stress to the muscle may not lead to further improve-
thermore, a study in healthy participants showed that ments. This tenet was supported by a weak correlation (r =
increasing occlusion pressure only seemed to have an effect 0.14) between deoxyhemoglobin concentrations and
on the cross-sectional area and muscle strength when increased muscle cross-sectional areas in participants com-
resistance exercise was completed at very low intensities pleting high-intensity resistance exercise alongside BFRT
(20% of the 1RM).16 More research needs to be conducted compared with a strong correlation (r = 0.71) between
to identify the optimal limb occlusion pressure for maxi- these variables in participants who completed low-
mizing quadriceps muscle function in patients after ACLR. intensity resistance exercise with BFRT. Therefore, we
The duration of BFRT application during treatments as contend that adding metabolic stress to the muscle of
well as the number of treatment sessions in which BFRT is patients undergoing ACLR with BFRT did not result in
utilized is also important to consider when comparing improvements in quadriceps muscle function, as protein
results. In our study, we based our BFRT treatment dosage synthesis was already high because of the large amount
on exercise recommendations that support the use of of mechanical tension on the quadriceps muscle during
AJSM Vol. 48, No. 4, 2020 Blood Flow Restriction Training and ACLR 835

our leg-press exercise. The application of BFRT after comparable concentric or eccentric intensities (ie, similar
ACLR thus might be most beneficial during early phases percentages of a 1RM), no differences in total strength
postoperatively when high-intensity exercise is not possi- improvement after training were observed. Therefore,
ble and the benefits of metabolic stress on protein synthe- eccentric exercise may not be superior to concentric exer-
sis can be ‘‘activated.’’ Further study of BFRT alongside cise when training at comparable loads, and to achieve
low-intensity exercise early after ACLR is necessary to the benefits of eccentric exercise, patients must train at
determine if early postoperative improvements in quadri- eccentric loads higher than those of concentric loads.
ceps muscle function can be achieved. Another difference between the current study and that of
Another factor that may help explain the differences in Lepley et al requiring discussion is the training intensity.
results among studies examining BFRT in patients under- While all participants in Lepley et al’s study were asked
going ACLR is the duration of BFRT treatment. For the to train at 60% of their 1RM, the average eccentric train-
intervention protocol by Ohta et al,23 participants per- ing intensity for their study was 108% of the 1RM, while
formed an exercise protocol that consisted of 7 exercises, our participants all trained at 70% of their 1RM. This dif-
for 2 to 3 sets, of 20 repetitions for 6 days a week for 15 ference in training intensity could undoubtedly affect the
weeks. The BFRT group performed all exercises with magnitude of strength gains noted. However, a training
BFRT applied, regardless of the duration of the exercise intensity of 60% to 70% of the 1RM is considered to induce
program.23 Takarada et al37 were more controlled with strength gains2 and therefore should lead to improvements
the duration of BFRT, with 10 repetitions of 5 minutes of in quadriceps muscle strength. While a decline in strength
BFRT daily for 10 consecutive days. However, Iversen from preoperative to postintervention was noted in all 4
et al9 had a similar BFRT treatment duration to Takarada groups, the decline would most likely have been larger if
et al,37 with a BFRT treatment time of 50 minutes per day patients were not participating in prescribed exercises
for almost 2 weeks. The amount of exercise and length of (study exercise and/or standard-of-care exercise). Without
blood flow restriction required to optimize quadriceps mus- a group that performed no study-related exercise and
cle function after ACLR should be systematically consid- only underwent standard of care, we cannot determine if
ered in future studies. our study exercise resulted in less strength loss than
patients who only completed standard-of-care ACL
rehabilitation.
Lack of Differences Between Eccentric Gerber et al5,6 also noted superior gains in muscle
and Concentric Exercises strength and size for patients undergoing ACLR who per-
formed eccentric muscle action using a recumbent stepper
Prior research has shown that eccentric exercise delivered compared with those undergoing standard-of-care therapy
alongside standard-of-care ACL rehabilitation results in with a concentric recumbent stepper, which also contra-
greater improvements in muscle strength,5,15 activation,15 dicts the findings of the current study. To standardize
and volume5 when compared with standard rehabilitation eccentric exercise intensity, Gerber et al5,6 utilized a rating
alone. We had hypothesized therefore that all groups of perceived exertion that participants reported while per-
receiving eccentric exercise would be superior to groups forming the eccentric cycling exercise and not a percentage
receiving concentric exercise, which was not supported by of force generated by the limb, as we did in our study. A
our results. self-perceived exertion level is not objective and also does
In a study by Lepley et al,15 patients undergoing ACLR not provide information about force generation for the
performed eccentric exercise on a BLAST! Leg Press sys- limbs between the groups. Therefore, it is possible that
tem at 60% of the 1RM and had improved recovery of iso- the perceived exertion led to different training intensities
metric strength and quadriceps muscle activation when for their study groups, allowing for greater improvements
compared with patients receiving standard of care only. in muscle strength and volume for the eccentric group.5
Our eccentric protocol employed the same leg-press device, Previous research has shown that when eccentric training
sets, and repetitions, with participants performing eccen- is completed at higher intensities than concentric training,
tric action at 70% of the 1RM, but participants in the greater improvements in muscle strength are realized.
eccentric groups had no difference in quadriceps muscle
strength or activation compared with the concentric Limitations
groups. Our study is different than Lepley et al’s in that
all participants performed the leg-press exercise at 70% For our primary analysis, the eccentric group was found to
of the 1RM (concentric or eccentric), whereas in Lepley be significantly older than the other 3 groups. However, the
et al’s study, participants who were not in the eccentric mean ages for all groups ranged from 15.3 to 18.8 years, and
group (ie, control or standard-of-care group) did not per- we are not aware of any difference in response to BFRT or
form any exercises on the BLAST! device. It is possible eccentric versus concentric exercise for this age range.
that the leg-press exercise, regardless of the concentric or The sample size for this study is relatively small. The
eccentric exercise type, leads to similar improvements in sample size, however, is comparable with that in other sim-
quadriceps muscle strength and activation, and thereby, ilar studies,9,37 and we contend that our lack of significant
differences between groups would not exist. Supporting differences is the result of a lack of effect of the interven-
this idea are results from a meta-analysis by Roig et al,29 tion, as illustrated by the very small effect sizes with
who found that when healthy participants trained at 95% CIs crossing 0, rather than a lack of statistical power.
836 Curran et al The American Journal of Sports Medicine

Further, we had participants drop out of the investigation, atrophy when compared with a high-intensity intervention
bringing the sample size of our concentric and eccentric without BFRT. Further studies are necessary to under-
groups to 8 participants per group, which is less than the stand the lack of efficacy in this population and the consid-
9 per group that we had powered for. Our secondary anal- eration of other intensities and frequencies of BFRT
ysis, however, had double the number of participants than protocols to improve quadriceps muscle function in
was required based on our power analysis, and still, no sig- patients after ACLR.
nificance was found. Future investigations should include
larger sample sizes.
The study sample included patients who underwent REFERENCES
ACLR using various graft types and who suffered different
meniscal injuries with surgical procedures employed to 1. Abe T, Kearns CF, Sato Y. Muscle size and strength are increased
after walk training with restricted venous blood flow from the leg
repair or remove the tear. Graft type and meniscal treat-
muscle, Kaatsu-walk training. J Appl Physiol (1985). 2006;100(5):
ment (removal vs repair) would increase the variability of 1460-1466.
the patient population and the rehabilitation that the 2. American College of Sports Medicine. American College of Sports
patients may have received. This variability in the popula- Medicine position stand: progression models in resistance training
tion may, in part, account for the large standard deviations for healthy adults. Med Sci Sports Exerc. 2009;41(3):687-708.
that we see in many of our outcome measures. However, 3. Biazon T, Ugrinowitsch C, Soligon SD, et al. The association between
we believe that the variability in our groups is representa- muscle deoxygenation and muscle hypertrophy to blood flow
restricted training performed at high and low loads. Front Physiol.
tive of the ACLR population as a whole, and as such, clini-
2019;10:446.
cians should be aware that patients may respond 4. Cook CJ, Kilduff LP, Beaven CM. Improving strength and power in
differently to BFRT. In some patients, it may be more or trained athletes with 3 weeks of occlusion training. Int J Sports Phys-
less effective than in others, and potentially, there may iol Perform. 2014;9(1):166-172.
even be responders and nonresponders. Future research 5. Gerber JP, Marcus RL, Dibble LE, Greis PE, Burks RT, LaStayo PC.
is necessary to better understand whether meniscal injury Effects of early progressive eccentric exercise on muscle size and
function after anterior cruciate ligament reconstruction: a 1-year
or graft type influences how patients respond to BFRT.
follow-up study of a randomized clinical trial. Phys Ther. 2009;89(1):
The high-intensity resistance exercise completed in this 51-59.
study was a closed kinetic chain task in which the quadriceps 6. Gerber JP, Marcus RL, Dibble LE, Greis PE, Burks RT, Lastayo PC.
muscles were not isolated. As such, it is possible that the par- Safety, feasibility, and efficacy of negative work exercise via eccen-
ticipants may have avoided using the quadriceps muscles tric muscle activity following anterior cruciate ligament reconstruc-
during this task and instead primarily used muscles such tion. J Orthop Sports Phys Ther. 2007;37(1):10-18.
7. Grindem H, Snyder-Mackler L, Moksnes H, Engebretsen L, Risberg
as the gluteal muscles to meet the exercise demands. We
MA. Simple decision rules can reduce reinjury risk by 84% after
carefully observed for compensations (lifting the buttock off ACL reconstruction: the Delaware-Oslo ACL cohort study. Br J
of the device, etc), and these were corrected when identified. Sports Med. 2016;50(13):804-808.
Participants also reported that their quadriceps muscles 8. Ithurburn MP, Paterno MV, Ford KR, Hewett TE, Schmitt LC. Young
were tired after exercise, suggesting that they were utilized athletes with quadriceps femoris strength asymmetry at return to
during the intervention. Further, we saw a strong correlation sport after anterior cruciate ligament reconstruction demonstrate
between the 1RM leg press and isokinetic (r = 0.888) and iso- asymmetric single-leg drop-landing mechanics. Am J Sports Med.
2015;43(11):2727-2737.
metric (r = 0.768) strength measures at postintervention,
9. Iversen E, Rostad V, Larmo A. Intermittent blood flow restriction does
suggesting that the 1RM was highly correlated with gold not reduce atrophy following anterior cruciate ligament reconstruc-
standard measures of muscle strength in our ACLR study tion. J Sport Health Sci. 2016;5(1):115-118.
cohort and that our participants were likely utilizing the 10. Kent-Braun JA, Le Blanc R. Quantitation of central activation failure
quadriceps muscles during exercise. during maximal voluntary contractions in humans. Muscle Nerve.
Neither patients nor outcome assessors were blinded to 1996;19(7):861-869.
11. Kilgas MA, Lytle LLM, Drum SN, Elmer SJ. Exercise with blood flow
group assignment, which may have introduced bias into
restriction to improve quadriceps function long after ACL reconstruc-
the study. tion. Int J Sports Med. 2019;40(10):650-656.
The intervention schedule for this study was 2 sessions 12. LaStayo PC, Woolf JM, Lewek MD, Snyder-Mackler L, Reich T, Lind-
per week for 8 weeks for a total of 16 sessions. All partici- stedt SL. Eccentric muscle contractions: their contribution to injury,
pants completed 16 sessions, but because of a number of prevention, rehabilitation, and sport. J Orthop Sports Phys Ther.
variables (eg, illness, personal trips, scheduling conflicts, 2003;33(10):557-571.
13. Laurentino GC, Ugrinowitsch C, Roschel H, et al. Strength training
inclement weather, transportation problems), they were
with blood flow restriction diminishes myostatin gene expression.
not always able to complete the intervention within the Med Sci Sports Exerc. 2012;44(3):406-412.
8-week time frame. The mean time for the participants to 14. Lepley LK, Palmieri-Smith R. Effect of eccentric strengthening after
complete the intervention was 68.3 days. anterior cruciate ligament reconstruction on quadriceps strength.
J Sport Rehabil. 2013;22(2):150-156.
15. Lepley LK, Wojtys EM, Palmieri-Smith RM. Combination of eccentric
exercise and neuromuscular electrical stimulation to improve quadri-
CONCLUSION ceps function post-ACL reconstruction. Knee. 2015;22(3):270-277.
16. Lixandrao ME, Ugrinowitsch C, Laurentino G, et al. Effects of exer-
BFRT with high-intensity resistance exercise added to an cise intensity and occlusion pressure after 12 weeks of resistance
ACL rehabilitation program did not significantly improve training with blood-flow restriction. Eur J Appl Physiol. 2015;
the recovery of quadriceps muscle strength, activation, or 115(12):2471-2480.
AJSM Vol. 48, No. 4, 2020 Blood Flow Restriction Training and ACLR 837

17. Loenneke JP, Fahs CA, Thiebaud RS, et al. The acute muscle swell- 30. Schmitt LC, Paterno MV, Ford KR, Myer GD, Hewett TE. Strength
ing effects of blood flow restriction. Acta Physiol Hung. 2012;99(4): asymmetry and landing mechanics at return to sport after anterior
400-410. cruciate ligament reconstruction. Med Sci Sports Exerc. 2015;
18. Loenneke JP, Wilson GJ, Wilson JM. A mechanistic approach to 47(7):1426-1434.
blood flow occlusion. Int J Sports Med. 2010;31(1):1-4. 31. Schmitt LC, Paterno MV, Hewett TE. The impact of quadriceps fem-
19. Loenneke JP, Wilson JM, Marin PJ, Zourdos MC, Bemben MG. Low oris strength asymmetry on functional performance at return to sport
intensity blood flow restriction training: a meta-analysis. Eur J Appl following anterior cruciate ligament reconstruction. J Orthop Sports
Physiol. 2012;112(5):1849-1859. Phys Ther. 2012;42(9):750-759.
20. Mall NA, Chalmers PN, Moric M, et al. Incidence and trends of ante- 32. Schoenfeld BJ. The mechanisms of muscle hypertrophy and their
rior cruciate ligament reconstruction in the United States. Am J application to resistance training. J Strength Cond Res. 2010;
Sports Med. 2014;42(10):2363-2370. 24(10):2857-2872.
21. McHugh MP, Tyler TF, Nicholas SJ, Browne MG, Gleim GW. Electro- 33. Schoenfeld BJ. Potential mechanisms for a role of metabolic stress in
myographic analysis of quadriceps fatigue after anterior cruciate lig- hypertrophic adaptations to resistance training. Sports Med.
ament reconstruction. J Orthop Sports Phys Ther. 2001;31(1):25-32. 2013;43(3):179-194.
22. Noehren B, Andersen A, Hardy P, et al. Cellular and morphological 34. Scott JM, Martin DS, Ploutz-Snyder R, et al. Panoramic ultrasound:
alterations in the vastus lateralis muscle as the result of ACL injury a novel and valid tool for monitoring change in muscle mass.
and reconstruction. J Bone Joint Surg Am. 2016;98(18):1541-1547. J Cachexia Sarcopenia Muscle. 2017;8(3):475-481.
23. Ohta H, Kurosawa H, Ikeda H, Iwase Y, Satou N, Nakamura S. Low- 35. Slysz J, Stultz J, Burr JF. The efficacy of blood flow restricted exer-
load resistance muscular training with moderate restriction of blood cise: a systematic review & meta-analysis. J Sci Med Sport. 2016;
flow after anterior cruciate ligament reconstruction. Acta Orthop 19(8):669-675.
Scand. 2003;74(1):62-68. 36. Sousa J, Neto GR, Santos HH, Araujo JP, Silva HG, Cirilo-Sousa MS.
24. Palmieri-Smith RM, Lepley LK. Quadriceps strength asymmetry after Effects of strength training with blood flow restriction on torque, mus-
anterior cruciate ligament reconstruction alters knee joint biome- cle activation and local muscular endurance in healthy subjects. Biol
chanics and functional performance at time of return to activity. Sport. 2017;34(1):83-90.
Am J Sports Med. 2015;43(7):1662-1669. 37. Takarada Y, Takazawa H, Ishii N. Applications of vascular occlusion
25. Palmieri-Smith RM, Thomas AC, Karvonen-Gutierrez C, Sowers M. A diminish disuse atrophy of knee extensor muscles. Med Sci Sports
clinical trial of neuromuscular electrical stimulation in improving Exerc. 2000;32(12):2035-2039.
quadriceps muscle strength and activation among women with 38. Tennent DJ, Hylden CM, Johnson AE, Burns TC, Wilken JM, Owens
mild and moderate osteoarthritis. Phys Ther. 2010;90(10):1441-1452. JG. Blood flow restriction training after knee arthroscopy: a random-
26. Palmieri-Smith RM, Thomas AC, Wojtys EM. Maximizing quadriceps ized controlled pilot study. Clin J Sport Med. 2017;27(3):245-252.
strength after ACL reconstruction. Clin Sports Med. 2008;27(3):405-424. 39. Tourville TW, Jarrell KM, Naud S, Slauterbeck JR, Johnson RJ, Beyn-
27. Peck BD, Brightwell CR, Johnson DL, Ireland ML, Noehren B, Fry CS. non BD. Relationship between isokinetic strength and tibiofemoral
Anterior cruciate ligament tear promotes skeletal muscle myostatin joint space width changes after anterior cruciate ligament recon-
expression, fibrogenic cell expansion, and a decline in muscle qual- struction. Am J Sports Med. 2014;42(2):302-311.
ity. Am J Sports Med. 2019;47(6):1385-1395. 40. Williams GN, Snyder-Mackler L, Barrance PJ, Buchanan TS. Quadri-
28. Pope ZK, Willardson JM, Schoenfeld BJ. Exercise and blood flow ceps femoris muscle morphology and function after ACL injury: a dif-
restriction. J Strength Cond Res. 2013;27(10):2914-2926. ferential response in copers versus non-copers. J Biomech. 2005;
29. Roig M, O’Brien K, Kirk G, et al. The effects of eccentric versus con- 38(4):685-693.
centric resistance training on muscle strength and mass in healthy 41. Yasuda K, Ohkoshi Y, Tanabe Y, Kaneda K. Muscle weakness after
adults: a systematic review with meta-analysis. Br J Sports Med. anterior cruciate ligament reconstruction using patellar and quadri-
2009;43(8):556-568. ceps tendons. Bull Hosp Jt Dis Orthop Inst. 1991;51(2):175-185.

For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav.