This multicenter randomized controlled trial was conducted from January 2021 to June 2022 in two trauma centers (Shanghai Jiao Tong University affiliated Sixth People’s Hospital and The People’s Hospital of Mengla County). A total of 120 eligible outpatients were recruited for this two-arm single-assessor blinded randomized control trial.

Inclusion criteria included:

Exclusion criteria included:

MASLD was confirmed based on the diagnostic criteria (18): Hepatic steatosis identified by imaging and the presence of at least one of the five cardiometabolic risk factors and there are no other causes of hepatic steatosis: (1) BMI ≥ 23 kg/m² or waist circumference > 94 cm in men or > 80 cm in women; (2) fasting serum glucose ≥ 5.6 mmol/L or 2h plasma glucose levels ≥ 7.8 mmol/L or glycated hemoglobin A1c ≥ 5.7% or type 2 diabetes or treatment for type 2 diabetes; (3) blood pressure ≥ 130/85 mmHg or antihypertensive drug treatment; (4) triglycerides ≥ 1.70 mmol/L or lipid lowering treatment; or (5) high-density lipoprotein cholesterol ≤ 1.0 mmol/L in men or ≤ 1.3 mmol/L in women or lipid lowering treatment.

All patients provided written informed consent in compliance with the declaration of Helsinki before joining the study. All study protocols were approved by the Research Ethics Committee of Shanghai Jiao Tong University affiliated Sixth People’s Hospital (REC reference no.: 2016-110). This clinical trial was registered at chichr.org.cn (ID: ChiCTR2100042872).

The primary outcome measure of pain (Knee Injury and Osteoarthritis Outcome Score [KOOS] pain) was used to calculate the required sample size using G*power version 3.1 (Heinrich-Heine-Universität Düsseldorf Universitätsstr, Germany) (19). The minimal detectable change in KOOS pain was based on the values in a recent report (20). To achieve a power of 95% at an alpha level of 0.05, a total of 96 patients (48 per group) was required. Thus, to account for up to a 20% drop-out rate, a total of 120 patients were recruited.

Two trauma centers joined this two-arm, single-assessor blinded randomized controlled trial. All recruited patients were block randomized to the WB (n=60) or BFR (n=60) group by an independent research team member. The blinding procedure was performed using opaque envelopes (n=60 for each group). Each participant was asked by an independent research team member to choose one of the 120 envelopes. The groups were coded by an independent member of the research team, while the principal assessor of the results and data analysis was blinded to the group allocations.

The patients’ demographic data, including age, sex, BMI, smoking, diabetes, and Kellgren and Lawrence system grade, were recorded, as were their baseline data, including ROM, muscle strength (scaled maximal isotonic strength [10RM]), KOOS subscales, and 30-s chair sit-to-stand test results. All participants were then instructed to complete 12 weeks of training and undergo assessments at 4 and 12 weeks ( Figure 1 ).

Each WB training intervention session included three parts: stretching exercise, ROM exercise, and strength exercise. The stretching exercises included triceps stretching in the standing position, hamstring stretching in the supine position, and quadriceps stretching in the prone position. The ROM exercises included knee extension at the end of the long sitting position, knee flexion at the end of the long sitting position, and stationary bike or trainer use. The strength exercise included quadriceps activation in the knee extension position and advanced closed-chain training of the lower limb. All training sessions were supervised by a physical therapist from the research team. Both groups were designed according to standard recommended protocols for each exercise type ( Table 1 ).

An automatically personalized tourniquet (PT) system (Delfi Medical, Vancouver, BC, Canada) was used in the BFR group. The system, set to automatically calculate occlusion pressure of the lower limb (LOP) defined as the minimum pressure required for full arterial occlusion (21), has clinically acceptable accuracy and high reliability (22–24). The PT device increases the cuff pressure in a step-up fashion, analyzes the pressure pulsation in the cuff bladder from the arterial pressure pulsation at each increase, and uses these properties to determine the LOP. The system consists of an easy-to-install dual-purpose variable profile nylon cuff (11.5 cm × 86 cm; 5 mm thick) that connects to the PT system via a sealing hose and automatically adjusts the pressure to within acceptable limits. Before the exercise, the cuff was placed at the proximal end of the limb and LOP was calculated in the position stimulated by the BFR. The BFR pressure was set at 80% LOP to maximize the recruitment of fast-twitch fibers and maximize muscle adaptation (25, 26). The LOP of each limb was computed separately for each session.

The pain was evaluated using the pain subscale of the KOOS (9-question, with a score of 0 indicating extreme pain to 36 indicating no pain).

Knee flexion was measured using a goniometer with each patient positioned supine and moving the heel as close as possible to the buttocks. Knee extension was measured with the patient maximally extending the knee joint and recorded as the difference from 0° of extension. The ROM of the knee was calculated as maximum flexion degree – maximum extension degree.

Scaled 10RM strength was assessed using a MED leg press (Technogym, Bracknell, UK) following a warm-up of 10 min light cycling. Beginning at 80% estimated 10RM the maximum load that could be lifted ten times through complete, the full ROM performed using the correct form was recorded as the concentric 10RM. All 10RM values were achieved within five attempts at 5-kg increments in external load at each attempt and a 3-min rest between attempts to allow full muscle recovery. The 10RM load for each patient was established at a level where they were capable of completing the 10th repetition but unable to perform an 11th repetition. Leg press exercise techniques followed the recommendation of the National Strength and Conditioning Association (27).

The KOOS is a self-reported tool used to assess patients’ opinions of their knee function and associated problems. The KOOS has five subscales: pain, symptoms, function in daily living, function in sport and recreation, and knee-related quality of life. Each subscale includes questions with standardized answer options across five Likert boxes scored 0–4. Each subscale was scored independently, with 0 indicating maximum symptoms.

The 30-s chair sit-to-stand test (30s-CST) test is the American College of Sports Medicine (ACSM) recommended function and strength measurement test for elderly individuals. Participants were asked to stand from the chair and sit down as many times as possible in 30 s. The MDC value of the 30s-CST is 1.13 s (28).

All patients’ data were coded and stored on the electronic data capture system for a specific disease. The system was operated within the hospital’s local servers. All statistical analyses were performed using SPSS Statistics version 24.0 (IBM Corp, Chicago, IL, USA). Data are presented as mean ± SD with 95% confidence intervals unless stated otherwise. Intergroup differences in patients’ baseline characteristics were assessed using independent-sample t-tests for continuous dependent variables and Fisher’s exact test for categorical data. The Shapiro-Wilks test (p>0.05) was used to examine normally distributed data, while Levene’s test of homogeneity was used to examine the homogeneity of the variances. The analysis of variance test was used to assess pain, 10RM strength, self-reported function, and 30s-CST with group allocations as the inter-subject independent factor and time as the intra-subject dependent factor.

Eight participants (5 in the BFR group, 3 in the WB group) failed to complete the study due to unplanned surgery and injury (n=5) and unrelated reasons (n=3). The remaining 112 participants (93%) completed the study and the follow-up survey. There were no significant intergroup differences in baseline demographic data ( Table 2 ), including adherence or protocol changes. No adverse events were reported.

Statistically significant improvement in KOOS pain scores were detected at the 4- and 12-week follow-up assessments, with significant intergroup differences at 12 weeks ( Table 3 ). The change in KOOS pain score differ between groups from baseline to 4- and 12- weeks ( Table 4 ).

A statistically significant increase in knee ROM was noted at 4 and 12 weeks, with a significant intergroup difference ( Table 3 ). The intergroup change in knee ROM from 0 to 4 and 0 to 12 weeks was significant ( Table 4 ).

Scaled 10RM strength increased significantly from baseline to 12-week follow-up with significant intergroup differences ( Tables 3 , 4 ).

Statistically significant difference was noted for all KOOS subscale scores. The scores of subscales of symptoms, ability in daily living, function in sports and recreation, and knee-related quality of life showed significant intergroup differences at the 12-week follow-up ( Table 3 ). The change of scores in subscales of daily living and knee-related quality of life differed significantly between groups at 0-4 weeks, 4-12 weeks, and 0-12 weeks ( Table 4 ).

The mean 30s-CST increased significantly from baseline to 12-week follow-up with significant intergroup differences ( Tables 3 , 4 ).

Different studies have evaluated the effects of training programs on pain, symptoms, physical function, and quality of life among patients with OA (29–31). Almost all evidence emphasized improvements in various aspects of patients’ daily lives following a regular exercise program. BFR is a novel auxiliary training method that reportedly improves the postoperative rehabilitation of anterior cruciate ligament reconstruction patients (32).

The subjects of this study were obese individuals with KOA who experience greater joint stress involving a cycle of pain and loss of strength and knee function due to damaged cartilage and altered joint mechanics. Heavy load training can be risky for such patients due to further cartilage degeneration. Brky et al. reported that patients experienced significantly less joint pain during BFR than heavy load training (33). The results of our study demonstrated the superiority of BFR over WB training alone in terms of long-term pain reduction (12 weeks) with no adverse events.

Muscle strength and volume increases have been reported after all exercise protocols regardless of duration (4–12 weeks). Recent studies reported that 8–13% of knee extensor muscle strength could be recovered with the aid of BFR (34) (35), primarily due to the neuromuscular adaptations induced by the restricted blood flow. The hypoxic environment created by BFR induces greater recruitment of type II fibers and interleukin-6 and growth hormone concentrations (36, 37). Our results proved that BFR exercise for 12 weeks significantly increased muscle strength.

Arthrogenic inhibition is associated with joint cartilage impairments, effusion, and pain (38); therefore, the reduced KOOS pain scores observed in our study may have contributed to strength adaptations of the OA limb.

The significant and clinically important improvements in all patient’s self-reported function measures and 30s-CST performance results observed in both groups are in line with those of recent studies of knee OA patients (14, 33). The improvements in physical function are attributable to improved strength and reduced pain. According to the KOOS subscale results, the pain, symptom, function in daily living, function in sports and recreation, and knee-related quality of life scores decreased significantly in both groups and differed significantly between them at the 12-week follow-up survey. Thus, the greater pain reduction and strength improvement may have contributed to improved ROM. A study examining a 3-week exercise program after TKA reported increased knee ROM and decreased knee pain (39). Thus, muscle strength training of the knee is suitable for the postoperative exercise of young patients undergoing anterior cruciate ligament reconstruction and the rehabilitation of obese patients with KOA. The American Academy of Orthopedic Surgeons, Osteoarthritis Research Society International, and American College of Rheumatology explain that strength increment promotes pain reduction, attenuates symptomatology, and reduces joint damage progression in individuals with KOA (40–42).

Recent research also suggested that BFR may have a hypoalgesia effect, as knee pain was significantly reduced during, immediately after, and 24 h after BFR versus heavy load resistance training (43, 44). Although the mechanism of this effect is unclear, there are several possibilities. Muscle pain due to ischemia and stress is often used as conditioned stimuli to modulate pain and can alter pain sensitivity in healthy individuals. Thus, structured pain modulation by BFR cuff pressure and exercise-induced muscle pain due to high ischemia levels and BFR resistance training may contribute to the antivaccination response. Other possible mechanisms include the release of endogenous opioids and endocannabinoids during exercise (45–47).

The application of BFR passively or in combination with aerobic exercise during early anterior cruciate ligament rehabilitation was discussed previously (13). The present study showed that BFR is also suitable for the rehabilitation of obese patients with KOA comorbid with MASLD, with the advantages that it affords a greater pain reduction and better improves physical function than WB exercise alone.

Pain extent is a major factor affecting knee OA rehabilitation results, as the pain has a detrimental effect on motor control and muscle function that results in modified movement patterns (48). The effects of pain reduction by BFR can improve the benefits of strength training. Therefore, BFR may be a superior tool during early KOA rehabilitation, particularly among patients reporting a high degree of pain.

During knee OA rehabilitation, applying progressively heavier loads is important to preventing muscular adaptations to the exercise (9). Our study did not use progressively heavier loads during the exercise protocol, which could have reduced the exercise’s effects. Further studies are needed to reveal the relationship between load progression and functional results.

Muscle morphology is another parameter used to evaluate functional outcomes of training. Still, it was not used in our study due to the 4.2–13.0 MHz wide-band linear array scanning transducer head and the image analysis software available to our study group.

BFR resistance exercise is employed by individuals with limited or incapacity of performing high-intensity resistance exercise, such as elderly (49) and orthopedic rehabilitation (13). Several studies offered caution to BFR resistance exercise (50–52), particularly regarding its cardiovascular effects. It’s suggested that individuals with conditions like heart failure, hypertension, or peripheral artery disease may lead to exaggerated metaboreflex activation during exercise (50), making BFR training potentially risky for them. There were no adverse events noted during the trial and follow-up. But still, careful prescription and monitoring are essential for safe practice, particularly in clinical populations.

This study included a specific subgroup of KOA patients, which limits the generalization of our findings to other patient populations. Moreover, it focused on a specific degree of KOA (Kellgren and Lawrence grade II–III), which may also limit the generalizability of our results.