Multicenter Retrospective Study of Blood Flow Restriction Training After Anterior Cruciate Ligament Reconstruction

Anterior cruciate ligament (ACL) injuries are among the most common sports-related knee injuries. Anterior cruciate ligament reconstruction (ACLR) is the primary surgical treatment, but it is frequently associated with postoperative quadriceps atrophy, muscle weakness, and delayed functional recovery¹. Epidemiological data suggest an incidence of approximately 68.6 cases per 100,000 individuals annually, with nearly 75% of patients requiring ACLR in 1 year². Postoperative muscle atrophy and strength deficits prolong recovery time and increase the risk of reinjury. Patients fail to return to sport due to persistent muscle weakness^3,4,5. Studies have shown that quadriceps strength loss can occur within the first 6 months after surgery, and this deficit may persist for more than 12 months, becoming a key factor contributing to increased risk of reinjury and poor functional recovery⁶.

Conventional rehabilitation protocols, such as high-load resistance training, are often challenging in the early postoperative stage due to surgical wound sensitivity and compromised tissue integrity. In adolescents, high-load strength training is not recommended due to concerns regarding skeletal development. Therefore, there is an urgent need to explore safe and effective alternatives. Blood flow restriction training (BFRT), which combines low-load resistance exercise with peripheral vascular occlusion, has recently gained considerable attention in ACLR rehabilitation. Its mechanism is thought to involve ischemia-reperfusion effects that activate anabolic signaling pathways, thereby mitigating muscle atrophy and promoting strength recovery^7,8.

Current clinical studies have shown that BFRT, even under relatively low loads, can significantly reduce early postoperative quadriceps atrophy, enhance knee extensor torque, and accelerate the course of functional recovery. However, multicenter data remains heterogeneous, and some studies have failed to demonstrate additional long-term strength gains with BFRT during extended follow-up. Therefore, in light of the current epidemiological burden of ACLR and the clinical need for postoperative rehabilitation, further investigation is warranted to clarify the efficacy and applicability of BFRT^9,10. In this context, this multicenter retrospective analysis aims to systematically evaluate the effect of BFRT on strength recovery, functional outcomes, and return to sport rates in ACLR patients. The methodology and results are detailed below.

Ethical approval was granted by the Ethics Committee of Suining County Hospital of Traditional Chinese Medicine (Approval No.20240116001). Written informed consent was obtained from all participants prior to their enrollment in the study. Details of the instruments and software used in this study can be found in the Table of Materials.

Participant demographic and general information

A total of 428 patients who underwent ACLR between December 2021 and December 2024 at Suining County Hospital of Traditional Chinese Medicine, The First Affiliated Hospital of Soochow University, and Xuzhou Rehabilitation Hospital were retrospectively reviewed. After excluding 112 patients and applying 1:1 propensity score matching (PSM), 316 patients were included, with 158 in each group: the BFRT group and the conventional rehabilitation group (Figure 1). Patient allocation to either the BFRT or conventional rehabilitation group was based on the established clinical pathway for postoperative care¹¹ and shared decision-making between the physician and the patient¹².

The control group included 85 males and 73 females, aged 18 to 55 years, with a mean age of 35.85 ±5.41 years. The affected side was the left in 92 cases and the right in 66 cases. The BFRT group included 82 males and 76 females, aged 19 to 53 years, with a mean age of 36.15 ±4.72 years. The affected side was the left in 88 cases and the right in 70 cases. There were no statistically significant differences in baseline characteristics between the two groups (p > 0.05), indicating comparability (Table 1).

The study adopted the following inclusion and exclusion criteria. Inclusion criteria comprised ACLR performed in accordance with the 2022 Clinical Practice Guidelines for ACL injury, primary single-bundle ACL reconstruction, postoperative adherence to rehabilitation, and an Outerbridge classification of no more than grade II. Exclusion criteria included other significant knee pathologies, severe osteoporosis, major organ dysfunction (cardiac, hepatic, or renal), coagulopathy, and neurovascular injuries of the knee.

Rehabilitation interventions

To ensure consistency across the multicenter study, a standardized conventional rehabilitation protocol was implemented in all three participating centers. This protocol was developed based on established ACL postoperative rehabilitation guidelines^11,13 and formally endorsed through expert consensus among the participating institutions. The key components are as follows. Standardized protocol goals and principles: all centers adhered to identical phased objectives-early phase focused on reducing swelling and pain, restoring joint range of motion; later phase emphasized progressive muscle strengthening-guided by the principle of progressive loading. Permitted variability and control measures for equipment use: While centers were permitted to use locally available but functionally equivalent rehabilitation devices (e.g., different brands of quadriceps trainers or balance boards), the prescribed exercises, targeted muscle groups, and intensity progression criteria were strictly standardized across sites. Quality assurance: To minimize implementation variability, a centralized supervision team was established. Monthly virtual meetings were conducted to review case report forms, deliver standardized training, and address clinical questions, ensuring consistent delivery of the intervention across centers. Sensitivity analysis for center effects: To evaluate potential site-specific influences on outcomes, the center was included as a random effect in a linear mixed model. Sensitivity analyses revealed no statistically significant variance attributable to center effects (p > 0.05) for both primary and secondary outcomes, confirming that minor operational differences did not compromise the consistency or validity of the overall results.

The conventional rehabilitation protocol for both groups was as follows. At postoperative weeks 0-2, begin pain-free active knee range-of-motion (ROM) exercises 3x-5x per day. While seated or lying supine, gently move the knee within a comfortable, non-painful range. For edema control, instruct the patient to maintain a supine position with the ankle elevated above the knee and the knee above the heart. Have the patient perform ankle pumps for 30-50 repetitions per set, 3 sets per session, 2x a day. Instruct the patient to perform straight leg raises by maintaining full knee extension and raising the leg to a 15°-30° angle. Hold for 5-10 s. Perform 10 sets of 10 repetitions per day. Apply cryotherapy for 10-15 min per session after exercise or when pain occurs. ADL training encompassed bed-to-chair transfers, dressing, toileting, and grooming. At postoperative week 2 onward (outpatient/inpatient rehab), initiate progressive resistance training using knee extension/flexion machines for 15-30 repetitions x 3-5 sets, 2x-3x per week. Implement balance training, progressing from double-leg to single-leg stance. Hold for 30 s for 10 repetitions per day. Start advanced ADL training (walking, stair climbing). Then proceed to progressive power training as appropriate.

Blood flow restriction training (BFRT) group

Patients in the BFRT group received the same standard protocol as the control group, with the addition of BFR cuff application during resistance training.

BFRT setup and pressure determination: Provide pre-measurement instructions. Inform the patient about the purpose and procedure of the blood flow restriction training pressure assessment (LOP measurement). Confirm that there are no contraindications such as vascular disease, severe diabetes, or coagulation disorders. Instruct the patient to lie in a comfortable supine or seated position; supine is preferred for better palpation or visualization of the dorsalis pedis artery. The affected limb should be extended and relaxed, without crossing or tension. Assess skin integrity to avoid placing the cuff over any open wounds or compromised areas. Roll up the trouser leg to the groin area to expose the thigh. Locate the proximal one-third of the femur; this can be measured by using a tape measure from the anterior superior iliac spine (ASIS) to the upper border of the patella. Select a cuff of appropriate width (commonly 12-15 cm) and wrap it around the designated area. Ensure moderate tightness without compressing the skin. The cuff should be laid flat, without folds, twisting, or risk of slipping. The outlet (connector) should face outward for convenient connection to the pressure system. Connect the cuff to the pressure pump device. Manual inflation bulbs may be used. Ensure the device is functioning properly, the gauge is calibrated to zero, and pressure values are clearly visible. Check that all connections are sealed. Deflate the cuff temporarily to allow for relaxation. Palpate the dorsalis pedis artery on the lateral side of the foot, between the first and second metatarsals. If the pulse is not easily palpable, use one of the following: a stethoscope placed over the pulse site or a portable Doppler flow detector. Once identified, mark or fix the probe at the correct position for later use. Inflate the cuff at a slow rate (approximately 10 mmHg/s) while continuously monitoring the pulse. Record the pressure at which the pulse completely disappears. This value is the LOP for one trial. Deflate the cuff completely and wait 5 min. Repeat the measurement 2x for a total of three trials. Calculate the average of the three LOP values to establish the final LOP for the patient as given below.
Example: LOP Trial 1 = 130 mmHg , Trial 2 = 128 mmHg , Trial 3 = 132 mmHg
then Final LOP = (130 + 128 + 132)/ 3 =130 mmHg.
Calculate the target training pressure. Training pressure should be set between 60% and 80% of the LOP. Formula: Target Pressure = LOP x (0.6-0.8)
Example: LOP = 130 mmHg, then target pressure = 78-104 mmHg. Recommended safe range: 50-150 mmHg (to avoid complete occlusion or risk).
Special Considerations: For patients with high blood pressure or large limb circumference, adjustments to inflation pressure may be needed. If any two LOP readings differ by more than 10 mmHg, perform an additional measurement for accuracy. Prior to exercise, inflate the cuff to the target pressure and monitor the patient for 1-2 min for any adverse reactions (e.g., severe pain, numbness).

Outcome measures

Evaluate functional outcomes at baseline (T0) and at 4 weeks (T1), 8 weeks (T2), and 12 weeks (T3) postoperatively. Use four primary measures.
30s sit-to-stand test¹⁴: Assess lower limb muscular function with the 30 s sit-to-stand test. Instruct participants to complete as many full stand-to-sit cycles as possible in 30 s from a standard chair with their arms crossed over their chest.
1 min single-leg heel raise test¹⁵: Evaluate calf muscle endurance of the affected limb was evaluated using the 1 min single-leg heel raise test. Record the total number of correct repetitions, defined as a full-range heel raise without postural compensation, completed in 1 min.
6-min walk test (6MWD)¹⁶: Measure functional walking capacity with the 6MWD. Record the total distance in m that a participant could walk on a flat surface in 6 min, with safety monitoring.
International Knee Documentation Committee (IKDC) Subjective Knee Evaluation^17,18: Assess patient-reported knee function using the IKDC Subjective Knee Evaluation Form. This questionnaire yields a standardized score from 0 to 100, where higher scores indicate better knee function.

Statistical analysis

Statistical analysis: Perform statistical analyses using SPSS software (version 27.0). Assess the normality of continuous variables using the Shapiro-Wilk test. Present normally distributed data as mean ± standard deviation (±s) and report non-normally distributed data as median with interquartile range [M(IQR)]. Summarize categorical variables as frequency and percentage [n(%)]. Prior to propensity score matching, conduct between-group comparisons of continuous variables using independent samples t-tests for normally distributed data or Mann-Whitney U tests for non-normally distributed data; compare categorical variables using χ² tests or Fisher's exact tests, as appropriate. Given the repeated measurements at four time points (T0, T1, T2, T3), employ generalized estimating equations (GEE) to evaluate the effects of time, group, and the time x group interaction on each outcome measure. If a statistically significant interaction effect is observed, perform further within-group and between-group pairwise comparisons. For post hoc analyses, apply Bonferroni correction to adjust for multiple comparisons, setting the significance threshold at p < 0.0083 (adjusted α = 0.05/6 comparisons) to control the familywise error rate. Handle missing values using multiple imputations: generate five imputed datasets, analyze them separately, and then pool the results to produce robust estimates. Use two-sided statistical tests, with a nominal significance level of α = 0.05; consider results statistically significant if p < 0.05.

As determined by the chi-square (χ2) test, the baseline demographic characteristics of the two groups, including sex (p = 0.822), age (p = 0.053), and affected side (p = 0.733), were comparable with no statistically significant differences. At the start of the study (T0), independent t-tests revealed that all functional outcome measures were also similar between the BFRT group and the control group.

30 s sit-to-stand test

For the 30 s sit-to-stand test, the baseline performance was similar (Control: 8.31 ±1.81 versus BFRT: 8.17 ±2.01 repetitions; t = 0.651, p = 0.516). However, from week 4 (T1) onward, the BFRT group consistently outperformed the control group, with clear advantages at T1, T2, and T3 (all p < 0.001; Figure 2).

1 min heel raise test

A similar pattern was observed in the single-leg heel raise test. After a comparable baseline (Control: 11.55 ±2.77 versus BFRT: 11.27 ±2.63 repetitions; t = 0.921, p= 0.358), the BFRT group performed significantly more repetitions at T1 (p = 0.038), T2 (p = 0.030), and T3 (p = 0.019; Figure 3).

6 min walk distance

In the 6 min walk distance test, both groups started at a similar level (Control: 304.74 ±29.34 m versus BFRT: 308.32 ±35.28 m; t = 0.981, p = 0.327), but the BFRT group walked significantly farther at all subsequent follow-up points: T1 (p = 0.037), T2 (p = 0.026), and T3 (p = 0.015; Figure 4).

IKDC scores

In terms of subjective knee function, the IKDC scores also showed no significant difference at baseline (Control: 39.69 ±6.54 versus BFRT: 41.00 ±6.41; t = 1.798, p= 0.073). Subsequently, the BFRT group reported significantly higher scores compared to the control group at T1 (p= 0.039), T2 (p = 0.032), and T3 (p = 0.020; Figure 5).

In summary, while both the BFRT group and the control group demonstrated functional improvements throughout the 12-week postoperative period, the BFRT group consistently achieved statistically superior outcomes. At every follow-up assessment (4 weeks, 8 weeks, and 12 weeks), the BFRT intervention led to significantly better performance in tests of muscular endurance (30 s sit-to-stand, single-leg heel raise), functional walking capacity (6 min walk distance), and patient-reported knee function (IKDC Scores) compared to the standard rehabilitation protocol.

Data availability:

The raw data are uploaded in Supplementary File 1.

Figure 1: Flowchart of patient screening and enrollment. Please click here to view a larger version of this figure.

Figure 2: 30 s sit-to-stand test outcomes. A comparative analysis of the 30 s sit-to-stand test outcomes was conducted across multiple time points between the two patient groups. Data are presented as mean ± standard deviation (SD) for each group (n = 158 per group). Timepoints are T0 (baseline), T1 (4 weeks), T2 (8 weeks), and T3 (12 weeks) postoperatively. Please click here to view a larger version of this figure.

Figure 3: 1 min single-leg heel raise test outcomes. A comparative evaluation of the number of single-leg heel raises completed by the affected limb in the two patient groups within 1 min was performed at multiple time points. Data are presented as mean ± standard deviation (SD) for each group (n = 158 per group). Timepoints are T0 (baseline), T1 (4 weeks), T2 (8 weeks), and T3 (12 weeks) postoperatively. Please click here to view a larger version of this figure.

Figure 4: 6-min walk distance (6MWD) outcomes. A comparative analysis of the 6 min walk distance (6MWD) was conducted between the two patient groups at multiple time points. Data are presented as mean ± standard deviation (SD) for each group (n = 158 per group). Timepoints are T0 (baseline), T1 (4 weeks), T2 (8 weeks), and T3 (12 weeks) postoperatively. Please click here to view a larger version of this figure.

Figure 5: IKDC score comparison. A comparative assessment of the IKDC scores between the two patient groups was carried out at multiple time points. Data are presented as mean ± standard deviation (SD) for each group (n = 158 per group). Timepoints are T0 (baseline), T1 (4 weeks), T2 (8 weeks), and T3 (12 weeks) postoperatively. Please click here to view a larger version of this figure.

Table 1: Baseline characteristics of participants. Presents the baseline characteristics of the participants in both groups, including gender distribution (male/female), age (mean ± SD), and affected side (left/right). Abbreviations: BFRT = Blood Flow Restriction Training. Please click here to download this file.

Supplementary File 1: Study raw data. Please click here to download this file.

The core challenge of postoperative rehabilitation following ACL reconstruction lies not only in graft strength and fixation techniques, but also in muscle atrophy-induced strength loss-particularly quadriceps atrophy-which more directly impacts functional recovery and the timeline for return to sport¹⁹. High-load training is often not feasible early after surgery. BFRT has emerged as a low-load alternative, with its mechanism thought to involve the promotion of hypertrophy through hormonal and metabolic stress responses^5,7. This study proposes that incorporating BFRT into standard rehabilitation can significantly accelerate functional recovery in the crucial early postoperative phase.

This multicenter retrospective study investigated the effects of BFRT on quadriceps and hamstring strength, triceps surae endurance, walking ability, and joint function in patients following ACL reconstruction²⁰. The quadriceps and hamstrings are key muscle groups for maintaining dynamic knee stability²¹. Insufficient strength in these muscles can result in poor sit-to-stand coordination, abnormal gait patterns, difficulty with stair climbing, and increased risk of reinjury²². The 30 s sit-to-stand test is a practical tool for quickly assessing lower limb strength recovery, guiding rehabilitation progression, and preventing delayed strength gains from impairing overall functional outcomes²³. In this study, the BFRT group showed a clear upward trajectory from T0 to T2 and consistently outperformed the control group. This finding aligns with previous studies showing that BFRT, by inducing a localized hypoxic environment, stimulates growth hormone secretion and facilitates fast-twitch muscle fiber recruitment, thereby achieving muscle hypertrophy comparable to that of high-load resistance training under low-load conditions²⁴. The narrowing of group differences after T2 reflects compensatory effects of later-stage conventional rehab. 6MWD and IKDC scores also improved more rapidly with BFRT, indicating benefits in endurance, gait efficiency, and joint function. However, differences narrowed over time as the control group caught up to T3. By providing robust multicenter evidence, this research helps advance the scientific field by demonstrating that BFRT can shorten recovery timelines, potentially establishing a new standard for early-phase ACLR rehabilitation.

This study is limited by its retrospective, multicenter design, which introduces potential selection bias from factors like patient adherence and confounding variables from inter-center differences in protocol execution. Furthermore, the conclusions are limited by the absence of objective imaging data and the lack of long-term follow-up for outcomes such as reinjury risk.

For future directions, prospective randomized controlled trials (RCTs) with isokinetic strength testing and imaging should be conducted to confirm these findings. BFRT protocols must be standardized for pressure calibration and safety monitoring. Alternative methods for studying this hypothesis could include directly comparing different BFRT protocols (e.g., varying cuff pressures or training frequencies) or using advanced techniques like electromyography (EMG) to explore neuromuscular adaptations.

The primary importance of this method is its potential application as a safe and effective low-load training option in early ACLR rehabilitation. BFRT may also extend to other populations, such as elderly patients with sarcopenia or individuals recovering from meniscal surgery, where high-load training is contraindicated²⁵.