Effect of blood flow restriction on knee joint position sense in healthy recreationally active adults: A randomized pilot trial

INTRODUCTION

Blood flow restriction (BFR), a controlled form of vascular occlusion using an external cuff, has gained increasing attention in sports and rehabilitation settings.^[1-3] It is widely used in combination with resistance training, aerobic exercise, and sport-specific training, as well as during rest periods.^[4-8] It has also been combined with cross education.^[9] The latest recommendations for knee surgery rehabilitation also consider exercises with the application of BFR.^[10,11] Still,the evidence of its effectiveness remains controversial.^[1,12]

Despite the growing body of research focusing on strength-related outcomes of BFR, relatively few studies have examined its effects during occlusion, particularly on sensorimotor function.^[13,14] Some evidence suggests that BFR training may improve proprioception. For example, a randomized controlled trial (RCT) demonstrated that 6 weeks of low-load training with BFR improved joint position sense (JPS) and functional performance in the upper limb.^[15] However, these findings reflect training-induced adaptations and do not provide insight into the immediate effects of BFR applied during task performance.

Another randomized, double-blind, placebo-controlled study in healthy, recreationally active individuals provides evidence that BFR applied to the arm impairs wrist JPS, as shown by an increase in absolute angular error (AAE) during occlusion.^[14] This is particularly important because impaired JPS may compromise neuromuscular control and increase the risk of injury during dynamic tasks.

However, it remains unclear whether similar effects occur in the lower limb, particularly at the knee joint, which plays a key role in functional activities and is frequently involved in sports-related injuries. To date, only limited evidence is available regarding knee JPS under BFR conditions, and existing studies have primarily assessed proprioception after, rather than during, occlusion.^[13]

Consequently, investigating the effects of BFR on JPS, particularly during occlusion, represents an important and clinically relevant gap in the literature, especially given that BFR is widely used in both healthy individuals and clinical populations.

Conducting an appropriately designed pilot randomized trial before a future definitive RCT is considered good practice that improves research standards in orthopedics, sports medicine, and rehabilitation.^[16] It should be emphasized that the aims and objectives of pilot and feasibility studies differ from those of RCTs. They are rather designed to support the development of a future definitive RCT, where “definitive” means an appropriately powered study focusing on effectiveness or efficiency. Feasibility studies aim to assess whether the future trial can be conducted, whether it should be conducted, and if so, how it should be carried out. Pilot studies are a subset of feasibility studies; in addition to assessing feasibility, they also involve conducting a future definitive RCT, or part of it, on a smaller scale.^[17] The recommendation for pilot studies mainly focuses on three key aspects: They must be randomized, conducted before the definitive RCT, and primarily evaluate feasibility.^[18]

Therefore, the aim of this pilot randomized trial was to assess the feasibility of conducting a future definitive RCT investigating the effect of BFR on knee JPS in healthy recreationally active adults and to obtain preliminary estimates of the potential effect to inform the design of such a trial.

MATERIALS AND METHODS

The study was reported in accordance with an appropriate checklist, precisely the extension of the consolidated standards of reporting trials for randomized pilot and feasibility trials conducted in advance of a future definitive RCT.^[18,19]

Trial design

The study was a prospective, randomized, parallel-group, placebo-controlled pilot trial with three arms and an allocation ratio of 1:1:1. Participants were assigned to the group that received intervention (BFR Group), the group that received sham intervention with no treatment value (Placebo Group), or the group that received no intervention (Control Group). All participants underwent two assessments of knee JPS. The sequence of assessments and group-specific conditions is shown in Figure 1. The assessments were carried out separately for the right and left lower limbs. There were no changes to the methods after the pilot trial commenced.

Close

Figure 1:

Schematic illustration of the study protocol.. BFR: Blood flow restriction.

Export to PPT

Interventions

Intervention (BFR group)

BFR was applied using wireless pneumatic BFR cuffs for the lower limbs (AirBands, VALD Health, Brisbane, Queensland, Australia) as presented in Figure 2. The cuffs were connected through Bluetooth to a dedicated mobile app (AirBand App) installed on a tablet (Galaxy Tab S7 SM-T870, Samsung Electronics Co., Republic of Korea), which allowed for device calibration and pressure control.

Close

Figure 2:

(a) The front view and (b) the top-downside view of the lower limb blood flow restriction cuff used in the present study.

Export to PPT

Just before the second assessment of knee JPS, the cuff was placed on the thigh of the assessed limb at the level of the largest circumference, just below the inguinal fold. Then, the device’s calibration procedure was used to determine the individual limb occlusion pressure (LOP), defined as the minimal pressure required to fully occlude arterial blood flow in the limb. Calibration was performed with the participant in a supine position with the lower limb relaxed. During the second assessment of knee JPS, cuff pressure was maintained at 80% of the individual LOP. The cuff placement, calibration, and pressure control were performed by an experienced examiner trained in the use of the device.

Inactive intervention (placebo group)

Just before the second assessment of the knee JPS, in the Placebo group, the cuff was positioned as in the BFR group. The device calibration procedure was also performed in the same manner as in the BFR Group. However, during the second assessment of the knee JPS, the cuff was deflated and remained uninflated throughout the procedure, resulting in no BFR. This approach ensured that participants in the BFR and Placebo groups underwent an identical preparation procedure. The cuff placement and calibration were carried out by an experienced examiner trained in the use of the device.

No intervention (control group)

Participants allocated to the Control Group underwent the same assessment protocol without cuff application or any intervention.

Assessment of knee JPS

All participants completed two knee JPS assessments with a 90-min break in between. Apart from the group-specific intervention applied during the second assessment, both assessments were conducted in exactly the same manner.

Knee JPS was assessed using an active joint position reproduction test performed with the use of an isokinetic dynamometer (Biodex System 4 Pro, Biodex Medical Systems, Shirley, New York, USA). The assessment was performed bilaterally, starting with the dominant limb.

During the assessment, the participant wore comfortable sports clothing, including shorts, and sports shoes. The eyes of the examined person were covered with a black impermeable mask. The assessment was performed in a seated position. The axis of rotation of the dynamometer was aligned with the lateral femoral epicondyle. The trunk and the unexamined limb were stabilized with stabilizing belts. The assessment included two consecutive tests differing in starting position, target position, and movement direction, as presented in Table 1 and illustrated in Figures 3 and 4.

Table 1: Characteristics of the knee joint position sense assessment protocol.

Feature	Test 1	Test 2
Starting position	0°	90°
Target position	60°	30°
Movement direction	Knee flexion	Knee extension
Mode	Active	Active
Time to remember the target position	10 s	10 s
Time of break	3 s	3 s
Number of repetitions	2	2

Close

Figure 3:

Knee joint position reproduction test showing the (a) starting position at 0° and (b) the target position at 60° during knee flexion.

Export to PPT

Close

Figure 4:

Knee joint position reproduction test showing the (a) starting position at 90° and (b) the target position at 30° during knee extension.

Export to PPT

Each position was first passively demonstrated to the participant and held for a few seconds to allow memorization. The limb was then returned to the starting position, and the participant was instructed to actively reproduce the target angle.

The recorded parameter during each repetition was the AAE, expressed in degrees (°), defined as the absolute difference between the target position and the position reproduced by the participant. For each condition, two repetitions were performed. Based on these, the minimal AAE (the lowest value of the two repetitions) and the mean AAE (the arithmetic mean of the two repetitions) were calculated. For further analysis, the outcome variables were defined as the change in AAE between the first (baseline) and second assessment. Two outcome measures were analyzed: Change in mean AAE and change in minimal AAE.

RESULTS

A total of 30 individuals were screened for eligibility, all of whom were included and randomized. All participants completed both assessments and were analyzed, with no missing data or protocol deviations. Recruitment, randomization, and adherence to the study protocol were achieved as planned, and all pre-defined feasibility criteria were fulfilled. No adverse events were noted.

The characteristics of the study groups are presented in Table 2. The groups were comparable in age, body mass, body height, and BMI (p = 0.511–0.763). In each group during both assessments, there were 10 participants (5 females and 5 males).

The analyses of exploratory outcomes are presented below to provide preliminary estimates and should be interpreted as hypothesis-generating for a future definitive RCT rather than confirmatory.

For the exploratory outcome of change in mean AAE, a significant between-group effect was observed for the right limb at the 30° target position (p = 0.004, η² = 0.335), the left limb at 60° (p = 0.034, η² = 0.222), and the left limb at 30° (p = 0.043, η² = 0.208). No significant differences were found for the right limb at 60° (p = 0.278, η² = 0.091). The details of the between-group comparison of the change in mean AAE are presented in Table 3.

Post hoc analysis (Tukey’s HSD) revealed that, for the right limb at the 30° target position, the BFR Group demonstrated significantly greater increases in AAE compared to both the Placebo Group (mean difference = 4.39°, p = 0.014) and the Control Group (mean difference = 4.78°, p = 0.007).

For the left limb at the 60° target position, a significant difference was observed between the BFR and Placebo groups (mean difference = 4.46°, p = 0.038), whereas no significant differences were found between the BFR and Control groups (p = 0.101) or between the Placebo and Control groups (p = 0.889).

For the left limb at the 30° target position, although the overall ANOVA indicated a significant effect, post hoc comparisons did not reveal statistically significant differences among the groups (all p > 0.05).

The between-group comparison of the change in minimal AAE is presented in Table 4.

For the change in minimal AAE, a significant between-group effect was observed for the left limb at the 60° target position (p = 0.005, ε² = 0.312), the right limb at 30° (p = 0.010, ε² = 0.269), and the left limb at 30° (p = 0.041, ε² = 0.163). No significant differences were found for the right limb at 60° (p = 0.167, ε² = 0.059).

Post hoc pairwise comparisons with Bonferroni correction revealed that, for the left limb at the 60° target position, the BFR Group differed significantly from the Control Group (p = 0.005), whereas no significant differences were found between the BFR and Placebo groups (p = 0.076) or between the Placebo and Control groups (p = 1.000).

For the right limb at the 30° target position, significant differences were observed between the BFR Group and both the Control Group (p = 0.019) and the Placebo Group (p = 0.034), with no difference between the Placebo and Control groups (p = 1.000).

For the left limb at the 30° target position, post hoc comparisons did not reveal statistically significant differences between groups after correction for multiple comparisons (all p > 0.05).

Across all tested conditions, the mean change scores in the BFR Group were positive, indicating worsening of JPS during the second assessment when compared to the first assessment, whereas the Placebo and Control groups showed negative mean change scores, suggesting improvement or no worsening.

DISCUSSION

The primary findings of the present study demonstrated the feasibility of conducting a future definitive RCT to investigate the effect of BFR on knee JPS in healthy, recreationally active adults. Recruitment, randomization, protocol completion, and tolerance of the study procedures were satisfactory, and no adverse events were reported.

Importantly, the present study was designed as a pilot trial and was not powered to detect statistically significant differences between groups. Therefore, any statistically significant findings from the exploratory analyses should be interpreted with caution and not considered evidence of efficacy. Instead, these findings are intended to provide preliminary estimates of effect size and variability to inform the design of a future adequately powered trial. In this context, p-values in the Results section are reported for completeness but should not be interpreted as evidence of true between-group differences.

The exploratory findings indicate a potential tendency for BFR to negatively affect knee JPS. Across all tested conditions, the BFR Group showed positive mean change scores, indicating worsening of JPS. In contrast, the Placebo and Control groups generally showed negative change scores, suggesting improvement or no worsening. This improvement or lack of worsening may reflect a learning or familiarization effect across repeated assessments; however, this effect was not observed in the BFR Group, where performance deteriorated. The most pronounced differences between the BFR Group and the Placebo and Control groups were observed in the right limb at the 30° target position and in the left limb at the 60° target position.

Interestingly, the findings of the present study align with previous research on the upper limb. In a randomized, double-blind, placebo-controlled study by Królikowska et al., conducted in healthy, recreationally active individuals using an active joint position reproduction test, BFR applied to the arm was shown to impair wrist JPS, as evidenced by increased AAE during occlusion.^[14] Similarly, in the present study, BFR application was associated with a consistent tendency toward an increase in AAE, which may reflect reduced accuracy of joint position reproduction. Importantly, both studies assessed JPS at the time of occlusion, which appears to be a key methodological aspect when considering the safety of BFR application. This is particularly relevant given that proprioception seems critical for movement control.^[21] Still,given the exploratory nature of the present study and the limited number of available studies in this area, these findings should be interpreted with caution. They nevertheless provide important preliminary evidence supporting the need for a future adequately powered, definitive RCT to confirm these effects and to further investigate their clinical relevance.

What’s more, research by Clark et al. in individuals with chronic ankle instability demonstrated that adding BFR to dynamic balance exercises reduced balance performance and was accompanied by increased perceived instability and exertion.^[22] This is consistent with the results of an RCT by Malmir et al. in individuals with chronic ankle instability, which reported a significant increase in joint repositioning error and perceived instability following BFR application.^[23]

Taken together, the findings presented above may indicate that the effects of BFR on JPS are not limited to a specific joint or population but may reflect a more general response of the sensorimotor system to occlusive conditions. One possible explanation is that BFR introduces a fatigue-like constraint and alters afferent input from muscle mechanoreceptors, which may transiently impair proprioceptive accuracy and postural control. This is supported by previous research showing that BFR induces local deoxygenation and earlier fatigue without necessarily improving postural control performance.^[24]

However, not all studies report a negative effect of BFR on proprioception. For example, an RCT demonstrated that 6 weeks of low-load exercise with BFR improved JPS and functional performance in the upper limb.^[15] These findings may seem inconsistent with the results of the present study; however, they probably reflect differences between effects during BFR use and those that occur over time with BFR use. While the present study and other similar investigations assessed JPS during occlusion, the aforementioned study by Yeşilyaprak and Dere^[15] evaluated the effects of a training intervention, where neuromuscular adaptations may have occurred over time.

Similarly, previous research in healthy individuals has shown that although BFR induces muscular deoxygenation and fatigue, it does not necessarily impair postural control performance.^[24]

When it comes to other studies contrasting with the findings of the present study, a recent study in healthy male adults by Edgington et al. found that BFR did not impair ankle JPS, as no significant differences were observed between BFR and sham conditions.^[25] These findings appear to contrast with the results of the present study. This discrepancy may be explained by methodological differences between studies. In particular, assessing JPS under passive conditions, as used in the aforementioned study, may place lower demands on the sensorimotor system than active joint position reproduction tasks. In addition, differences in the tested joint, outcome measures, and BFR protocols may contribute to the observed variability. It should also be noted that the relatively small sample sizes in both studies may limit the ability to detect subtle effects; therefore, the findings should be interpreted with caution. Consequently, the effects of BFR on proprioception may be task-specific and context-dependent.

Of course, the findings of the study performed by Królikowska et al. and the present study do not suggest that BFR should not be used in clinical or sports settings.^[14] Rather, they indicate that, during application, BFR may be associated with transient impairments in JPS. This may be particularly relevant when exercises requiring precise motor control are performed and should be considered when designing and supervising training or rehabilitation protocols.

Importantly, the findings of the present study should be interpreted in the context of the broader literature on BFR in knee conditions. A recent review of a clinical trials registry performed by Reese et al. of interventional clinical trials found that most studies focus primarily on the benefits of BFR, such as improvements in muscle strength, endurance, and hypertrophy, while potential adverse effects remain largely underexplored.^[2] This imbalance highlights a significant gap in the current evidence base concerning potentially negative effects of BFR.^[26]

It should also be highlighted that the observed effects were not fully consistent across all target positions/movement directions and limbs, and not all significant so-called omnibus tests (ANOVA and Kruskal–Wallis tests) were followed by significant pairwise differences. This likely reflects the exploratory nature of the study, the limited sample size, and the variability of individual responses. Interestingly, the analyses based on mean AAE and minimal AAE showed a similar overall direction of effect, although the exact pattern of statistical significance differed. This may suggest that BFR influences not only the best achieved performance but also the consistency of joint position reproduction. Still, it should be remembered that this study is exploratory and hypothesis-generating rather than confirmatory. Therefore, these findings should be interpreted as preliminary and should not be used to draw definitive conclusions regarding the effect of BFR on JPS.

Strengths of the study include the randomized placebo-controlled design, standardized assessment protocol, blinded outcome assessment between the BFR and Placebo groups, and the analysis of JPS during rather than after occlusion.

CONCLUSION

This randomized pilot study demonstrated the feasibility of conducting an RCT to investigate the effect of BFR on knee JPS in healthy, recreationally active adults.

Exploratory findings indicate a tendency toward increased AAE in the BFR Group, suggesting reduced accuracy in reproducing the target position under BFR conditions, while the Placebo and Control groups generally did not show this. These effects were most clearly observed for the right limb at the 30° target position and for the left limb at the 60° target position.

Given the pilot and exploratory nature of the study, these findings should be interpreted with caution and regarded as hypothesis-generating rather than confirmatory. They provide preliminary estimates that may inform the design and sample size calculation of a future adequately powered definitive RCT.