Literature search

Two researchers independently conducted a systematic literature search across seven electronic databases: Web of Science, PubMed, EBSCO, Cochrane Library, Embase, Scopus, and China National Knowledge Infrastructure (CNKI). Additionally, ClinicalTrials.gov was searched for relevant registered trials. The search covered articles published from January 2000 to the final search date, November 16, 2024. The search strategy included the following English terms: Kaatsu, Blood Flow, BFR, ACL, Occlusion, Osteoarthritis, Arthritis, KOA, OA, Knee, Exercise, and Training. The Boolean search formula applied was: (“Kaatsu” OR “Blood Flow Restriction” OR “BFR”) AND (“ACL” OR “Knee” OR “Osteoarthritis” OR “Arthritis” OR “KOA”) AND (“Exercise” OR “Training”). During the search process, two researchers independently screened article titles and abstracts for relevance. If any information was unclear, the corresponding authors were contacted via email for clarification. In cases of disagreement, a third researcher was consulted to facilitate discussion and reach a consensus.

This systematic review and meta-analysis was prospectively registered in the PROSPERO database with the registration number CRD42046892642. The review protocol adheres to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.

Inclusion and exclusion criteria

Exclusion criteria.

Exclusion criteria include: duplicate studies; review articles, systematic reviews, meta-analyses, or animal studies; non-English or non-Chinese literature without a reliable translation; non-full-text articles that could not be obtained via email contact; studies with inadequate experimental design, lacking randomized control, or with outcome measures that do not meet the inclusion criteria; studies with subjects under 45 years of age or with severe comorbidities or confounding factors; studies with a small sample size, significant data loss, or poor quality (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Flowchart of Literature Screening.

https://doi.org/10.1371/journal.pone.0323388.g001

Quality assessment of studies

The quality of the included studies was assessed using the Cochrane Risk of Bias Tool [22]. The evaluation focused on the following aspects: random sequence generation, allocation concealment, blinding of participants and researchers, blinding of outcome assessment, incomplete outcome data, selective reporting, and other potential biases. Each item was rated as “Yes” (low risk of bias), “No” (high risk of bias), or “Unclear” (unable to determine risk of bias). Fig 2 presents the distribution of risk of bias across studies for each item, and Fig 3 summarizes the overall risk of bias.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Risk of Bias Plot.

https://doi.org/10.1371/journal.pone.0323388.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Summary of Risk of Bias.

https://doi.org/10.1371/journal.pone.0323388.g003

Assessment of publication bias

Funnel plots were used to visually assess potential publication bias among the included studies. To quantitatively evaluate the symmetry of the funnel plot, Egger’s regression test was performed. Egger’s test assesses whether there is a statistically significant asymmetry in the distribution of effect sizes, which may indicate publication bias. If publication bias was detected, sensitivity analyses were conducted to assess its potential impact on the robustness of the findings.

Data analysis

Review Manager 5.3.5 software was used for the analysis. The standardized mean difference (SMD) was chosen as the effect size due to variations in measurement across studies [23]. Heterogeneity was assessed using the I² statistic: a fixed-effects model was applied if I² < 50%, and a random-effects model if I² ≥ 50% [24]. Subgroup analysis explored sources of heterogeneity, and funnel plots assessed potential publication bias.

Basic characteristics of included studies

A total of 12 randomized controlled trials (RCTs) were included in this study[25–36], involving 642 middle-aged and elderly patients, comprising 193 men and 378 women. Among them, 120 participants were assigned to the preoperative low-load blood flow restriction (L-BFR) group, 192 to the postoperative L-BFR group, 285 to the low-load resistance training (L-CON) group, and 45 to the high-load resistance training (H-RT) group.Eight studies reported lower limb extensor strength as an outcome measure [24–30,32], while 10 studies included physical function tests as outcome variables [25–27,29,30,32–36], and all 12 studies measured pain scores as an outcome [25–36]. The primary condition was osteoarthritis in eight studies [25–30,32,33], and total knee arthroplasty (TKA) in five studies [31,33–36].

Variability in intervention protocols

The included studies exhibited notable variability in training protocols, particularly in training duration, occlusion pressure, and exercise intensity. Preoperative interventions ranged from 3 days to 8 weeks, while postoperative interventions lasted 6–14 weeks. Training was typically performed 2 to 5 days per week, with 4–6 sets per session and 10–30 repetitions per set, although some studies implemented progressive overload, while others maintained a fixed workload.

Occlusion pressure settings varied, with some studies using 40–60% limb occlusion pressure (LOP), while others applied absolute pressures ranging from 97.4 to 200 mmHg. Most studies used continuous occlusion, while some incorporated intermittent occlusion release. Sham BFR conditions were inconsistent, with some studies applying 20 mmHg, while others did not specify sham settings.

Training intensity also differed, with most studies prescribing low-load resistance training (20 ～ 30% 1RM), whereas high-load resistance training groups used 60–80% 1RM. Exercise selection included leg press, knee extension, and calf flexion, with variations in single-joint vs. multi-joint exercises. Rest intervals ranged from 30 to 180 seconds, influencing metabolic stress and neuromuscular recovery.

These differences in training parameters may have contributed to the heterogeneity in study outcomes, highlighting the need for standardized intervention protocols in future research (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of Basic Characteristics of Included Studies.

https://doi.org/10.1371/journal.pone.0323388.t001

Sensitivity analysis

Sensitivity analysis was performed by altering the analytical model, selecting different effect sizes, and excluding individual studies. The results of the meta-analysis remained consistent after sensitivity testing, confirming the robustness of the findings.

Publication bias

This analysis included 12 studies, meeting the minimum sample size required for funnel plot construction, as supported by Lu et al. [37]. Funnel plots of the included studies are presented in Fig 4A–4F. Notably, Fig 4C and 4D display apparent asymmetry, suggesting the presence of potential publication bias. This may indicate that studies with significant findings are more likely to be published, thereby inflating the overall effect size. Contributing factors may include the inclusion of low-quality studies or variations in BFR training protocols (e.g., frequency, intensity, duration).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Funnel Plot.

https://doi.org/10.1371/journal.pone.0323388.g004

Meta-analysis results

L-BFR vs. L-CON on pain scores.

A meta-analysis was conducted on 5 studies, including 224 participants, to evaluate the overall effect of L-BFR training compared to L-CON on pain scores (Fig 5A). The results showed an effect size of SMD = -0.10 (95% CI: -0.36 to 0.16; Z = 0.75; P = 0.46), indicating no significant difference in pain scores between L-BFR (n = 110) and L-CON (n = 114) groups (P = 0.46). As there was homogeneity among the studies (I² = 0, P = 0.72), a fixed-effects model was used.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. A-F Forest Plot.

https://doi.org/10.1371/journal.pone.0323388.g005

L-BFR vs. L-CON on muscle strength.

A meta-analysis of 5 studies involving 324 participants was performed to evaluate the impact of L-BFR compared to L-CON on muscle strength (Fig 5B). The results revealed an effect size of SMD = 1.11 (95% CI: 0.67 to 1.55; Z = 4.92; P < 0.00001), demonstrating that L-BFR training significantly improved muscle strength compared to L-CON. Due to heterogeneity among the studies (I² = 68%, P = 0.01), a random-effects model was applied. Subgroup analysis by gender revealed homogeneity within subgroups (I² = 0), indicating that gender was a source of heterogeneity (Table 2). The effect size for the female group was SMD = 0.77 (95% CI: 0.07 to 1.47; P = 0.002), while for the mixed group, it was SMD = 1.68 (95% CI: 0.73 to 2.64; P < 0.00001), suggesting that L-BFR improved muscle strength across different genders.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Subgroup analysis.

https://doi.org/10.1371/journal.pone.0323388.t002

The effects of BFR training on muscle strength in patients with knee injuries and its physiological mechanisms

L-BFR training is widely used to enhance muscle strength in athletes and older adults, and its application in rehabilitation has expanded [38,39]. Studies consistently show that BFR training significantly improves strength [40–42]. Hughes et al. found that L-BFR training outperformed traditional low-load training in musculoskeletal rehabilitation but was less effective than high-load training [40]. Rodrigo-Mallorca et al. reported significant strength gains (SMD = 0.61) in older adults, comparable to high-intensity training [41]. Wengle et al. found postoperative BFR training enhanced quadriceps strength and reduced muscle atrophy, but preoperative BFR showed no effect [42].

Our study aligns with these findings, confirming that L-BFR training is more effective than traditional L-CON in improving muscle strength in patients with knee injuries. Both L-BFR and H-RT yielded significant strength improvements, yet L-BFR provided superior outcomes compared to L-CON. Interestingly, we observed that preoperative L-BFR training significantly improved postoperative muscle strength, contrary to Wengle et al.‘s results [42]. The discrepancies between studies could be attributed to variations in training protocols (such as cuff pressure, training intensity, and duration), patient characteristics (such as baseline muscle strength or preoperative atrophy), and the methods used to assess muscle strength (subjective vs. objective measures). Additionally, differences in the interventions provided to control groups may contribute to the differences observed in outcomes, with our control group potentially receiving less structured rehabilitation compared to that of Wengle et al.’s study, which could influence the results.

BFR training enhances muscle strength through several key physiological mechanisms, including metabolic stress, neuromuscular adaptation, endocrine responses, and cellular swelling. The cumulative effects of these mechanisms help stimulate muscle hypertrophy and strength development, even under low external loads.

Metabolic stress and lactate accumulation

BFR training creates a hypoxic and ischemic environment in the target muscles by partially restricting venous return while maintaining arterial inflow. This condition reduces metabolite clearance, leading to the accumulation of lactate, hydrogen ions, and inorganic phosphate [43,44]. The buildup of these metabolic by-products increases intramuscular acidosis, which plays a critical role in initiating anabolic responses.

The metabolic stress induced by BFR training triggers a hormonal cascade, stimulating the secretion of growth hormone (GH), insulin-like growth factor-1 (IGF-1), and testosterone—all of which are essential for muscle protein synthesis and hypertrophy [45,46]. Additionally, elevated lactate levels enhance motor unit recruitment, particularly in fast-twitch (Type II) muscle fibers, which are responsible for high force output [47]. This preferential activation of Type II fibers under low-load conditions is a key factor distinguishing BFR training from conventional low-load resistance training.

Neuromuscular adaptation and fast-twitch fiber recruitment

Under normal conditions, low-load resistance training primarily activates slow-twitch (Type I) muscle fibers, as these fibers rely on oxidative metabolism. However, during BFR training, the restricted oxygen supply forces the recruitment of fast-twitch (Type II) fibers to compensate for the increased metabolic demand. This shift in fiber recruitment is crucial for enhancing neuromuscular adaptations and promoting greater force generation.

The increased activation of high-threshold motor units during BFR training leads to significant neuromuscular adaptations, even at low external loads [48]. Electromyographic (EMG) studies have demonstrated that muscle activation during low-load BFR training is comparable to that observed in high-resistance training (H-RT), reinforcing the effectiveness of BFR for strength development [49]. This finding supports the notion that BFR training can elicit substantial neural drive and motor unit engagement, facilitating muscle strength improvements without the need for high mechanical loads.

Endocrine and molecular signaling pathways

The combination of hypoxia and metabolic stress induced by BFR training stimulates multiple anabolic signaling pathways, which promote muscle hypertrophy and protein synthesis. One of the most critical pathways is the mechanistic target of rapamycin (mTOR) signaling cascade, which regulates muscle protein synthesis and adaptation. Studies have shown that BFR training significantly upregulates mTOR activity, thereby enhancing muscle growth and strength [50].Additionally, systemic increases in GH and testosterone following BFR training further amplify muscle hypertrophy and strength gains [51]. These hormonal responses are comparable to those observed in high-intensity resistance training, suggesting that BFR training can mimic the endocrine effects of traditional strength training despite using lower external loads.

Cellular swelling and anabolic response

BFR training induces cellular swelling due to venous pooling in the restricted limb, which creates an osmotic shift that draws fluid into muscle cells. This increase in intracellular pressure serves as a potent anabolic stimulus, promoting protein synthesis while inhibiting protein degradation [52]. The mechanosensitive pathways activated by this swelling response further contribute to muscle hypertrophy and structural adaptations [53]. The combination of cellular swelling, metabolic stress, and endocrine signaling reinforces BFR training’s ability to induce muscle hypertrophy with minimal mechanical load, making it an effective rehabilitation tool.

BFR training’s effects on pain in patients with knee injuries

While BFR training is highly effective in enhancing muscle strength, its impact on pain relief is less clear. Several studies have shown that BFR training can reduce pain in certain patient populations, but the findings remain inconsistent. Li et al. [38] found that in rehabilitation training for patients with knee injuries, the L-BFR group exhibited significant reductions in pain scores, indicating an advantage in alleviating pain. Conversely, Cuyul-Vásquez et al conducted a systematic review focusing on patients with knee pain and concluded that incorporating BFR into resistance training did not significantly reduce pain scores, suggesting limited short-term effects [41]. Similarly, Nitzsche et al analyzed patients with lower-limb musculoskeletal disorders and found no significant difference between L-BFR and traditional resistance training in pain relief [42].

The present study compared results from three perspectives: L-BFR versus L-CON, L-BFR versus H-RT, and preoperative L-BFR intervention versus conventional care. Across these comparisons, L-BFR showed limited effects, further demonstrating that its use does not significantly alleviate patient pain. This finding contrasts with previous studies that suggested BFR training has analgesic effects, potentially reducing knee pain before surgery, immediately after surgery, and up to 24 hours postoperatively [38]. Although BFR demonstrates significant efficacy in improving muscle strength and promoting functional recovery, its effectiveness in pain relief may be limited or even counterproductive in certain cases.

From a mechanistic perspective, BFR’s analgesic effects primarily rely on metabolic pressure-induced endogenous opioid release and inflammation modulation [51]. However, these mechanisms may only apply to inflammatory pain [52]. For mechanical pain (e.g., pain caused by joint lesions or soft tissue injuries), neuropathic pain, or pain associated with central sensitization, BFR’s effects may be limited. For instance, in chronic pain patients, central sensitization can cause persistent pain signaling even after peripheral pain sources have been addressed, potentially hindering pain relief [53].

Additionally, BFR application may have adverse effects. The accumulation of lactate and local hypoxia may induce metabolic pain, while cuff compression could directly stimulate the skin and nerve endings, causing mechanical discomfort [54]. Such adverse effects may obscure or counteract BFR’s potential analgesic benefits. Pain is also a complex, multidimensional phenomenon involving physiological, psychological, and emotional factors. For some patients, BFR might induce anxiety or exacerbate pain perception, weakening its analgesic efficacy.

Design flaws in interventions may further limit BFR’s effectiveness. For example, setting cuff pressure too high or too low could impact results [55]. Short intervention durations or inappropriate timing of pain measurement might miss the peak manifestation of analgesic effects. Existing literature also supports this view, as studies by Cuyul-Vásquez et al. [41] and Nitzsche et al. [42] failed to find significant pain relief effects of BFR for knee or other lower-limb pain, contradicting traditional assumptions about BFR’s analgesic mechanisms. These findings suggest that BFR may not be suitable for all types of pain or patient populations. Its application in pain management should be carefully tailored to individual cases.

Overall, the variability in BFR’s effects on pain may be attributed to differences in pain types, patient characteristics, training parameters, and psychological factors. These limitations necessitate further research into the underlying mechanisms and optimization of intervention strategies to clarify its scope of application and best practices.

Clinical implications and heterogeneity in BFR training application

While BFR training has demonstrated effectiveness in muscle strength enhancement, its impact on pain relief is variable and influenced by a variety of factors. In clinical practice, the heterogeneity of patient populations, training protocols, and equipment used can significantly affect the outcomes of BFR training. It is essential to consider individual patient characteristics—such as baseline muscle strength, injury severity, rehabilitation phase, and pain type—when designing BFR interventions. Patient-specific modifications to BFR protocols, such as adjusting cuff pressures, training intensity, or session frequency, may be necessary to optimize the benefits while minimizing potential risks. For example, older patients or those with severe muscle atrophy may benefit from lower cuff pressures and less frequent sessions, while athletes or younger patients may require more intense protocols to maximize strength gains. Additionally, variations in the equipment used (e.g., cuff design, inflation systems) and inconsistencies in training protocols may contribute to the heterogeneous effects of BFR training. To ensure safe and effective treatment, clinicians should regularly monitor patients and adjust BFR parameters as needed. Personalized assessments of muscle strength and pain should guide the modification of training protocols, with careful consideration of the patient’s progress, subjective reports, and objective outcomes. Further research on patient-tailored BFR applications will be invaluable in establishing more specific and universally applicable guidelines for its use in clinical settings.