Effects of Blood Flow Restriction Training in Patients before and after Anterior Cruciate Ligament Reconstruction: A Systematic Review and Meta-Analysis

(1) Objective: To examine the effects of blood flow restriction (BFR) training on muscle strength, cross-sectional area and knee-related function in patients selected for anterior cruciate ligament reconstruction (ACLR). (2) Methods: A literature search was conducted in PubMed, PEDro, Cochrane Library, Web of Science, SCOPUS, and ProQuest databases until 20 May 2024. Controlled clinical trials comparing the effects of BFR training with unrestricted training in patients before or after ACLR were selected. The GRADE approach was used to assess the degree of certainty for each meta-analysis. (3) Results: Ten studies were included (n = 287 participants). Standardized mean differences in favor of BFR training applied postoperatively were observed in knee extensor (SMD = 0.79; 95% CI = 0.06 to 1.52; I2: 68%) and flexor isokinetic strength (SMD = 0.53; 95% CI = 0.04 to 1.01; I2: 0%), and quadriceps cross-sectional area (SMD = 0.76; 95% CI = 0.27 to 1.26; I2: 0%). No changes were found in knee extensor isometric strength and knee-related function. The degree of certainty according to the GRADE was very low. (4) Conclusions: Very low degree of certainty suggests that BFR training provides additional benefits to unrestricted training on isokinetic strength and quadriceps cross-sectional area in patients undergoing ACLR.


Introduction
Anterior cruciate ligament (ACL) tears are one of the most common injuries among recreational and elite athletes [1].The annual incidence has been estimated at 68.6 ACL tears per 100,000 person-years, affecting mainly young people between the ages of 14 and 25 [2].
ACL reconstruction is the most used option to restore joint stability in ACL tears [2].The strength deficit and the muscle size loss in the affected leg occur after the ACL tear and are accentuated after surgery, regardless of graft type or concomitant meniscal procedures [3][4][5][6].A strength loss of 17% preoperatively and close to 40% one month postoperatively has been identified [3,7].Atrophy of the quadriceps, biceps femoris, and semitendinosus muscles begins after injury and undergoes a fourfold volume loss one month after surgery [3,8].Similarly, the cross-sectional area of the thigh muscles was also reduced compared to the uninjured leg, both before and after ACL [9].Some gains occur in the following weeks, but only half of the patients fully recover quadriceps and hamstring muscle strength and size after a year from the reconstruction [3,4].
Muscle strength and size appear to be important contributors to patient recovery.Preoperative and postoperative strength can predict knee-related function 6 months after ACL reconstruction, or upon return to sport [10][11][12].Furthermore, quadriceps muscle size was positively correlated to ACL volume in a recent study [13].Thus, treatment should focus on preventing strength loss and atrophy in patients undergoing or scheduled for ACL reconstruction.
The American College of Sports Medicine suggests loads of 70% of the individual's one-repetition maximum (1RM) to optimize muscle strength and hypertrophy gains, or even higher than 80% 1RM in experienced athletes [14].Patients with ACL injuries are often unable to tolerate training due to pain intensity, swelling, and impaired knee function, especially near the perioperative period [15].Moreover, an early return to vigorous training could compromise graft laxity [16], so it is important to develop adequate training procedures for these patients.Blood flow restriction (BFR) therapy is applied to elicit greater muscle activation, and consists of generating partial occlusion through external pressure on a body region.The pressure is often applied with pneumatic cuffs on the proximal limb, while the patient exercises [17,18].
The potential effects of BFR training on muscle fitness have been investigated in different populations with muscle weakness and knee atrophy.BFR training appears to promote larger strength and hypertrophy gains than isolated exercise in patients with knee osteoarthritis, rheumatoid arthritis, and older adults [19,20].The primary mechanisms underlying the clinical improvements through BFR training are under debate.The hypoxic muscular environment generated by BFR could induce metabolic stress that promotes muscle growth.Furthermore, BFR seems to increase neuromuscular demand during exercise, eliciting the recruitment of fast-twitch muscle fibers [21].
Previous studies have conducted systematic reviews and meta-analyses on the effects of BFR, grouping different musculoskeletal conditions [22,23], but its estimated effect on a clinically homogeneous population remains poorly understood.In recent years, there is emerging evidence proposing BFR training in patients scheduled for ACL reconstruction, given its clinical applicability and its promising effects on muscle strength and hypertrophy.This emerging evidence should be systematically and critically reviewed and reported according to evidence-based guidelines.Therefore, the aim of this systematic review with meta-analysis is to evaluate the effects of BFR training compared to the same unrestricted training on muscle strength, cross-sectional area, and knee-related function in patients before and after ACL reconstruction.

Study Design
This systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guideline and Cochrane Recommendations [24].The study protocol was registered in the International Prospective Register of Systematic Reviews under registry code CRD42022345026.

Search Strategy
The bibliographical search was conducted in PubMed (MEDLINE), the Physiotherapy Evidence Database (PEDro), the Cochrane Library, Web of Science (WOS), and SCOPUS from inception to 20 May 2024.The search strategy was defined using the Population, Intervention, Comparison, Outcomes and Study (PICOs) framework.Search terms were selected based on Medical Subject Headings (MeSH): anterior cruciate ligament, anterior cruciate ligament reconstruction, anterior cruciate ligament injuries, and blood flow restriction therapy.A gray literature search was conducted using Scopus and ProQuest databases, and a hand search was carried out on the reference lists of the included studies.The search strategy for each database is detailed in Appendix A.

Study Selection
The PICOs scheme was employed to frame the research question of whether BFR training applied before or after ACL reconstruction has additional benefits to unrestricted training.The included studies met the following criteria: Population (P), patients scheduled for ACL reconstruction; Intervention (I), BFR techniques combined with exercise therapy; Comparison (C), sham-BFR techniques combined with exercise therapy or isolated exercise therapy; Outcomes (O), isometric strength, isokinetic strength, cross-sectional area, and/or knee-related function; Studies (s), randomized or nonrandomized controlled clinical trials.
Studies were excluded due to the following: (1) intervention comprised therapies other than BFR or exercise; (2) BFR was not applied in combination with exercise therapy; (3) restricted and unrestricted training was different between the comparison groups; (4) primary intervention was pharmacological or surgical treatment; (5) selection criteria were not specified; (6) the design was a conference abstract, dissertation, or book; (7) they were not published in English, Spanish, or French.
After retrieving searches, references were exported to Mendeley Desktop, and duplicates were removed.Two reviewers (E.F. and I.H.) independently assessed the titles and abstracts of each reference to determine potential eligibility.The same reviewers independently evaluated the full texts of potentially eligible studies.Discrepancies between the two reviewers were resolved by a third author (L.C.).

Data Extraction
Both reviewers independently gathered information from each study in a standardized manner adapted from the Cochrane Collaboration.Collected data included (1) participants characteristics; (2) description of interventions administered to experimental and control groups; (3) outcome measures; and (4) results.Qualitative synthesis was used for data analysis, and whenever feasible, quantitative synthesis (meta-analysis) was also performed.

Risk of Bias and Quality Assessment
The PEDro scale was applied to assess the methodological quality of each study.This is an 11-item scale based on the Delphi list developed by Verhagen and Colleagues [25].The first item (eligibility criteria) focuses on external validity and was excluded from the total score.A score of 7 or higher was categorized as "high" quality and 6 or lower was indicative of "poor quality" [26,27].The Cochrane Risk of Bias tool was used to assess the risk of bias of the included studies.This tool evaluates potential bias and internal validity across 7 criteria, classifying studies as "low", "unclear", or "high" risk [28].The PEDro scale and the Cochrane Risk of Bias tool have demonstrated reliability in assessing study quality and identifying bias [28][29][30].The methodological quality of the studies was evaluated by the same independent reviewers, with any discrepancies resolved by a third reviewer.
Two investigators assessed the degree of certainty of each meta-analysis using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.This classification categorizes the evidence as "high", "moderate", "low", or "very low".The downgrading of the certainty level was based on the presence of risk of bias, inconsistency of results, indirectness of evidence, and imprecision [28,31,32].Detailed criteria for grading the evidence in each domain are found in Table 1.

Data Synthesis and Analysis
The quantitative synthesis of the results was conducted based on the considered outcomes: isometric and isokinetic strength; cross-sectional area; and knee-related function.When studies used different tools or measurement scales to assess the same outcome, the authors performed inverse variance methods.Six different meta-analyses were performed for the included outcomes, and subgroup analyses of the studies were performed to compare BFR training versus unrestricted training before surgery or after surgery.Mean and standard deviation (SD) of post-intervention and sample size from each group were extracted.Standard mean differences (SMD) and 95% confidence intervals (95% CI) were estimated based on the post-intervention means and SDs.An SMD between 0.2 and 0.5 indicated a small effect size, between 0.5 and 0.8 a moderate effect size, and >0.8 a large effect size [34].
Data were pooled for meta-analysis using a minimum of two trials deemed clinically homogeneous.Statistical significance was set at p < 0.05.The 95% CI of the SMD were displayed in the forest plots.A fixed effects model was considered if there was a common intervention and outcome measures.A random effects model was used when the combination of intervention effects could incorporate an assumption that the studies are not all estimating the same intervention effect [33].In addition, a leave-one-out analysis was carried out to assess the presence of any potential outlying study that affects the estimate of each meta-analysis.All meta-analyses were performed using RevMan 5.4 software.

Literature Search and Screening
Ten studies were selected for inclusion in the qualitative and quantitative synthesis of 427 potentially eligible records retrieved from all databases [35][36][37][38][39][40][41][42][43].The kappa value of agreement between reviewers was 0.96.The third reviewer resolved discrepancies regarding one study.The selection process is detailed in the PRISMA flowchart diagram (Figure 1).

Characteristics of the Eligible Studies
The included studies comprised 287 participants, the sample sizes ranged from 12 to 45 patients, and the mean ages ranged from 15 to 38 years.Most studies selected patients that underwent ACL reconstruction via semitendinosus-gracilis tendon graft [35,36,38,40,41].A bone-patellar tendon-bone was predominantly used for ACL reconstruction in three studies [37,39,44].The type of graft was not specified in two studies [42,43].Surgical procedures concomitant to ACL reconstruction were only detailed in four studies, all of which involved meniscus surgeries [37,40,41,44].Sociodemographic and clinical characteristics of the participants are shown in Table 2.

Characteristics of the Eligible Studies
The included studies comprised 287 participants, the sample sizes ranged from 12 to 45 patients, and the mean ages ranged from 15 to 38 years.Most studies selected patients that underwent ACL reconstruction via semitendinosus-gracilis tendon graft [35,36,38,40,41].A bone-patellar tendon-bone was predominantly used for ACL reconstruction in three studies [37,39,44].The type of graft was not specified in two studies [42,43].Surgical procedures concomitant to ACL reconstruction were only detailed in four studies, all of which involved meniscus surgeries [37,40,41,44].Sociodemographic and clinical characteristics of the participants are shown in Table 2.    BFR was applied in combination with exercise in all of the studies.Most exercise protocols included low-load weight-bearing exercises involving knee extensions and straight leg raises [35,36,[38][39][40][41][42][43][44], with sets ranging from 4 to 6 and repetitions per set ranging between 10 and 40.Only Curran et al. [37] used high-load leg press exercises.

Study Quality and Risk of Bias
According to risk of bias Cochrane tool (Appendix B), most of the studies included in this systematic review with meta-analysis described a high risk of selection, performance, and detection bias.According to the PEDro scale (Appendix C), four studies presented high methodological quality [35,36,39,41] and six showed low quality [37,38,40,[42][43][44].All studies provided central and dispersion values, and between-group comparisons for the main outcomes.Only one study provided concealed allocation [37].Six studies were described as randomized [36][37][38][39]43,44] and four as nonrandomized [35,[40][41][42].Any disagreement between the two reviewers was resolved by the third author.
The overall degree of certainty according to GRADE was rated as very low for all the outcomes.The certainty assessment according to GRADE is shown in Table 4.

⊕
Very low CRITICAL CI: confidence interval; SMD: standardized mean difference; a: more than 50% of participants were from studies without blinding of participants or assessors; b: more than 50% of participants were from studies with lack of allocation concealment; c: more than 50% of participants were from studies with low methodological quality; d: more than 50% of participants were from studies without sample size estimation; e: heterogeneity among interventions; f: heterogeneity among results of studies (I 2 ≥ 50%); h: 95% CI of SMD > 0.2; i: 95% CI of SMD > 0.5; j: 95% CI of the risk ratio crosses the null value; k: sample sizes < 50; l: sample sizes < 30.

Discussion
The aim of this study was to assess the effects of BFR training on strength, crosssectional area, and knee-related function before and after ACL reconstruction.This systematic review with meta-analysis found very low evidence suggesting that BFR training applied after surgery was more effective than sham BFR training or unrestricted exercise in increasing isokinetic strength at 60 • /s and quadriceps cross-sectional area in patients scheduled for ACL reconstruction.There were no statistically significant differences in knee extensor isometric strength and knee-related function.
The results for isokinetic knee extensor strength should be interpreted with caution because when the study by Ohta et al. [38] was treated as an outlier, no statistically significant differences were observed.The sources of bias observed in this study (blinding and allocation concealment) are prevalent across other included studies, and the number of studies in the subgroup meta-analysis is relatively small, as their results could be similarly biased.Nonetheless, clinicians and researchers must remain attentive to the possibility that individual data points may be influencing the stability of the results of the pooled analysis.
The methodological quality of the included controlled trials ranged from 3 to 7 points according to the PEDro scale.Only three studies included in the meta-analyses showed high methodological quality.According to PEDro and Cochrane Risk of Bias tool, common methodological flaws were lack of allocation concealment and blinding of participants and personnel.Four studies were nonrandomized, but three of them were described as quasirandomized [35,40,41].In these studies, the authors assigned the first two participants to the BFR group and the next two to the sham group, repeating the process until fulfilled groups.Despite the lack of adequate randomization, these studies reported similar clinical and sociodemographic characteristics between groups at baseline.The overall degree of certainty in the GRADE approach was downgraded mainly due to the low methodological quality of the included studies, heterogeneity among interventions, and small sample sizes.The measurement procedures and the clinical and sociodemographic characteristics of the included participants seem to be sufficiently consistent among studies to conduct a subgroup meta-analysis, but readers should consider the low level of certainty provided by the available evidence.
The pooled meta-analyses revealed moderate effects in favor of BFR training on knee isokinetic strength (extensor: SMD = 0.79; 95% CI = 0.06 to 1.52; flexor: SMD = 0.53; 95% CI = 0.04 to 1.01) and quadriceps cross-sectional area (SMD = 0.74; 95% CI = 0.33 to 1.15).Other meta-analyses also found that BFR could promote benefits in strength and quadriceps cross-sectional area in patients with various knee conditions [22,23].As main differences compared to these previous meta-analyses, the present study analyzed isometric and isokinetic strength separately, examined the effects of BFR added to exercise therapy, and included exclusively ACL-injured patients.Other systematic reviews suggested that low-load BFR training achieved a greater increase in isokinetic strength at 60 • /s than unrestricted low-load training, and similar to high-load training [45,46].According to this study, BFR training promoted an increase in isokinetic, but not in isometric, knee extensor strength.These results could reflect specific training adaptations, as the training interventions were based on dynamic exercises with concentric and eccentric muscle contractions.
BFR training led to gains in isokinetic strength and cross-sectional area for knee extensors, and strength gains for knee flexors were observed.Two studies analyzed the effects on hamstring cross-sectional area, but these were not included in the metaanalysis due to heterogeneity in the patients' clinical status and the interventions (before vs. after ACL reconstruction) [35,38].From a qualitative synthesis, none of these studies demonstrated additional benefits of BFR training in increasing knee flexors cross-sectional area.In this way, changes in muscle size could predict changes in strength by less than 50%, and other factors may be involved, such as musculotendinous stiffness, motor unit recruitment and synchronization, or neuromuscular inhibition [3,47].The knee-related function, as evaluated by the IKDC, showed no differences between BFR training and unrestricted training.However, for enhanced clinical significance, it would be important if the added advantages in terms of increased strength and cross-sectional area were linked to improvements in perceived function.
Regarding the subgroup analysis, there were also moderate benefits on isokinetic strength and quadriceps cross-sectional area favoring BFR training after ACL reconstruction.Nevertheless, there were no statistically significant differences in any outcome when BFR training was applied before surgery.This may be due to longer interventions when BFR was applied after surgery.Most of these interventions lasted from 8 to 14 weeks [37,38,42,44], whereas they lasted from 1 to 3 weeks when BFR was applied presurgery [35,[39][40][41].Additionally, weakness and atrophy occur after ACL rupture, but are particularly noticeable after surgical repair [3,48].In this sense, BFR could achieve additional benefits to exercise in highly deconditioned muscles, which is consistent with the first weeks after the reconstruction.Curran et al. [37] was the only study that started BFR intervention late after surgery, and found no benefit through BFR for any outcome.This supports that BFR training does not appear to be more effective than isolated training before or a few months after surgery, when muscle deconditioning is not maximal.
The study samples included young adult individuals with a higher proportion of males in most studies.Regarding age, only Kacin et al. [35] and Curran et al. [37] recruited patients with average ages above 35 or below 25 years old, respectively.Therefore, the sociodemographic characteristics appear consistent across studies, and the clinical differences could depend on the timing of BFR training interventions relative to ACLR.BFR training interventions applied close to the immediate perioperative period used a high number of repetitions  and low intensity (10-30% 1RM or 40 RM).Curran et al. [37] used a lower number of repetitions (10) and concentric or eccentric phases at high intensity (70% 1RM), probably because patients were able to work at high intensities 10 weeks after surgery.Most studies used three to five sets, and the most commonly used pressure for restriction was 150 mmHg or 80% of limb occlusion pressure.However, there was variability in the way BFR was applied in terms of cuff pressure, inflation, and deflation times; this added to the existing heterogeneity among exercise interventions.The characteristics of the BFR training in patients scheduled for ACL reconstruction are quite similar to in other pathologies, such as knee osteoarthritis, patellofemoral pain and rheumatoid arthritis [19,46].There were some important limitations.Firstly, the heterogeneity between interventions and measurement time points.Despite performing a subgroup analysis, the pooled analysis of isometric strength and cross-sectional area includes both operated and nonoperated patients who had varying levels of strength, muscle size, and functional status.Furthermore, the BFR and exercise interventions varied among studies.Secondly, the sample sizes were small or insufficient, which may overestimate the results.Thirdly, nonrandomized clinical trials were included due to the small number of available studies, which may have compromised the overall quality of the meta-analyses.Finally, studies comparing low-load BFR training with high-load training were excluded, so relevant results for BFR clinical application may have been omitted.
The main practical applications of the study suggest that BFR may add moderate benefits when applied during exercise in isokinetic strength and quadriceps size in patients undergoing ACL reconstruction.Most interventions applied BFR in the early perioperative period at 80% of LOP or 150 mmHg, ensuring cuff deflation between sets, and combined with low-load training and high repetitions per set (>15).
Future randomized clinical trials with good methodological quality should examine the effects of BFR training in patients with ACL injury.Further research is needed to compare low-load BFR training with high-load training in patients undergoing ACL reconstruction.In addition, future meta-analyses should take into account the baseline muscle fitness and start time of the intervention relative to surgery.

Conclusions
The results of this systematic review with meta-analysis found very low degree of certainty suggesting that BFR training provides additional benefits compared to the same unrestricted training on isokinetic knee extensor and flexor strength at 60 • /s, and quadriceps cross-sectional area in patients selected for ACL reconstruction.These additional benefits likely emerge when BFR is applied after ACL reconstruction rather than before it, but further research is needed.Similar benefits between BFR and unrestricted training were observed in isometric knee extensor strength and knee-related function.These results should be interpreted cautiously due to the heterogeneity and risk of bias among studies.abstract(("anterior cruciate ligament reconstruction" OR "anterior cruciate ligament injuries" OR "anterior cruciate ligament" OR "acl") AND ("blood flow restriction" OR "kaatsu" OR "vascular occlusion" OR "occlusive training" OR "occlusion training" OR " blood vessel restriction")) 42 #1 title(("anterior cruciate ligament reconstruction" OR "anterior cruciate ligament injuries" OR "anterior cruciate ligament" OR "acl") AND ("blood flow restriction" OR "kaatsu" OR "vascular occlusion" OR "occlusive training" OR "occlusion training" OR " blood vessel restriction"))

Table 1 .
[33]eria for downgrading the quality of evidence in each domain according to the GRADE approach[33].

Table 2 .
Sociodemographic and clinical characteristics of included studies.

Table 3 .
Intervention characteristics of included studies.

Table 4 .
GRADE assessment of blood flow restriction compared to control.