Effect of blood-flow restricted vs heavy-load resistance training on strength, power, and speed for healthy volunteers: a systematic review and meta-analysis

Introduction

Traditional high-load resistance training (HLR; i.e., > 65% of one repetition maximum (1RM)) is widely regarded as an effective strategy for enhancing maximal strength, muscle power, and speed (Kraemer & Ratamess, 2004; Marshall et al., 2021). The American College of Sports Medicine (ACSM) also recommends high-load training using external loads of 60–90% 1RM to improve maximal muscle strength and power (Garber et al., 2011). However, high-load training may not be suitable for older adults, untrained individuals, and short-season athletes, as it may increase the risk of injury, discomfort, and negatively affect athletic performance during competitions (Liu & Latham, 2010; Slysz, Stultz & Burr, 2016). This is because high-load resistance training can increase the risk of injury and discomfort, thereby affecting the continuity and effectiveness of the training. Consequently, there has been a growing interest in exploring low-load alternatives to high-load exercise.

Low-load resistance training typically has minor or no positive effects on strength, power, and speed compared to traditional HLR training (Schoenfeld et al., 2016; Suetta et al., 2004; McBride et al., 2002). However, in recent years, low-load blood flow restriction resistance training (LL-BFR) has gained increasing attention as a promising alternative (Burton, 2022; Hughes et al., 2017). Low-load blood flow restriction training (LL-BFR), which combines low external loads (20–30% 1RM) with partial blood flow restriction, induces metabolic stress and muscle fatigue, leading to adaptations in muscle strength, hypertrophy, and power that are comparable to those from high-load training (Jessee et al., 2018; Pearson & Hussain, 2015; Lixandrão et al., 2018). Therefore, LL-BFR training may provide a viable alternative to HLR training.

LL-BFR training has been shown to significantly improve muscle strength, power, and speed, with promising effects not only in healthy individuals and athletes (Centner et al., 2019; Vergara et al., 2024; Wortman et al., 2021), but also in certain pathological populations (Ahmed, Mustafaoglu & Erhan, 2024; Nitzsche et al., 2021; Wen et al., 2023). Previous studies have found that there were comparable effects of LL-BFR training and HLR training on muscle strength, muscle power, and speed in healthy individuals (Castilla-López & Romero-Franco, 2023; Wang et al., 2022; Laurentino et al., 2012). However, some studies indicated that LL-BFR training was less effective compared to HLR training in terms of training gains (Clark et al., 2011; de Lemos Muller, Ramis & Ribeiro, 2019; Ramis et al., 2020). Besides the risk of Type II error (i.e., failing to detect a true effect due to factors like insufficient sample size or high variability), these discrepancies may be attributed to differences in training variables (e.g., training duration, training frequency, cuff pressure, and cuff width) and subject characteristics (Lixandrão et al., 2018; Xiaolin et al., 2023; Fabero-Garrido et al., 2022). Differences in these factors can lead to different physiological adaptations, ultimately impacting the effectiveness of the training outcomes. Therefore, it is necessary to conduct a meta-analysis to compare the effects of LL-BFR and HLR training on muscle strength, power, and speed in healthy people comprehensively. However, existing meta-analyses only compared the effects of LL-BFR and HLR training on muscle strength (Lixandrão et al., 2018; Grønfeldt et al., 2020), lacking systematic reviews and meta-analyses that thoroughly compared their impacts on muscle strength, power, and speed.

Therefore, the aim of this study was to conduct a systematic review and meta-analysis to compare the effects of LL-BFR training and HLR training on muscle strength, muscle power, and speed in healthy people, thereby providing scientific evidence for related training practices. Additionally, this study aims to explore the potential moderating effects of individual characteristics (i.e., age, gender, and athletic level) and training variables (i.e., training duration, frequency, cuff pressure, and cuff width) on the responses in muscle strength, muscle power, and speed.

Methods

Registration and literature search

This meta-analytical review was performed in accordance with the PRISMA statement guidelines for Systematic Reviews and Meta-Analyses (Page et al., 2021) (Prospero registration number: CRD42024536163).

The articles published by May 13, 2024, were identified through the SCOPUS, SPORTDiscus, PubMed, Web of Science, and CNKI databases. The search strategy combined the following keywords: “blood flow restriction”, “vascular occlusion”, “KAATSU”, “strength training”, “resistance training”, “high intensity training”, “weight training”, “high load training”, to ensure comprehensive coverage of relevant literature. The full search string is provided in Supplemental Material 1. After removing duplicates, the screening process was conducted in three phases: titles, abstracts, and full texts (Fig. 1). To further ensure the inclusion of all relevant studies, reference lists from included studies, previous reviews, and the lead authors’ personal libraries were also manually examined. To enhance methodological rigor, two assessors (Y.L. and Z.G.) independently retrieved articles and extracted data. Discrepancies were resolved by a third author (T.W.).

Results

Primary meta‑analysis results

The meta-analysis on maximum strength included 39 studies, with a total of 51 treatment outcome measures involving 804 participants. The effect size was −0.19 (95% Confidence Interval, CI [−0.31 to −0.06]), p = 0.003 (see Fig. 2). For muscle power, eight studies were included, with nine outcome measures involving 175 participants. The effect size was −0.04 (95% CI [−0.33 to 0.24]), p = 0.76 (see Fig. 3). In terms of jump performance, 10 studies were included, with a total of 18 outcome measures involving 187 participants. The effect size was −0.08 (95% CI [−0.30 to 0.15]), p = 0.50 (see Fig. 4). For sprint performance, six studies were included, with six outcome measures involving 110 participants. The effect size was −0.28 (95% CI [−0.71 to 0.15]), p = 0.20 (see Fig. 5).

Discussion

The present meta-analysis compared the effects of LL-BFR training (20–30% 1RM) and conventional HLR training (60–90% 1RM) on muscle strength, power, and speed. The main findings indicated that, despite significant differences in maximal strength gains between LL-BFR and HLR training, the effect size was trivial (ES = −0.19), suggesting that both methods have similar practical effects on maximal strength. Both LL-BFR and HLR training also promoted comparable gains in muscle power and speed. Furthermore, individual characteristics (i.e., age, gender, training status) and training parameters (i.e., training duration, frequency, cuff pressure, and cuff width) did not influence the comparative training gains between the two methods. Based on the PEDro scores and GRADE assessment, the overall quality of the included studies was rated as moderate. These findings highlight the potential of LL-BFR training as a viable and effective low-load alternative to HLR training for improving neuromuscular function in healthy populations from young to older adults.

The present meta-analysis found that LL-BFR and HLR training produced similar gains in maximal strength, consistent with the meta-analysis by Grønfeldt et al. (2020) but contrasting with two other meta-analyses by Lixandrão et al. (2018) and Hughes et al. (2017). The disparate conclusions are likely due to variations in study inclusion criteria, leading to differences in study methods and sample characteristics, ultimately resulting in differing outcomes. The meta-analysis by Lixandrão et al. (2018) included studies with within-subject controls (i.e., right leg with HLR training and left leg with BFR training, potentially leading to crossover training effects) and non-randomized designs, and did not directly compare LL-BFR and HLR. Similarly, the meta-analysis by Hughes et al. (2017) included non-randomized studies and populations with osteoarthritis. In contrast, the present meta-analysis and that by Grønfeldt et al. (2020) focused solely on healthy populations and excluded studies with within-subject controls or non-randomized designs, which was expected to enhance methodological robustness. Notably, the present meta-analysis and that by Grønfeldt et al. (2020) showed higher effect sizes compared to Hughes et al. (2017) (−0.19 to −0.17 vs −0.63).

Besides the similar gains in maximal strength, our meta-analysis also found comparable improvements in muscle power and speed between LL-BFR training and HLR training. The unique effectiveness of this low-load resistance training method in enhancing neuromuscular function makes it particularly attractive in settings where high-load exercise may be contraindicated, such as basic functional training to prevent falls in frail elderly individuals, rehabilitation exercises for athletes recovering from injuries, or training for beginners. For athletes in rehabilitation, LL-BFR may be an effective method for quickly restoring competitive performance, as it not only improves maximal strength similarly to high-load resistance training but also enhances power and speed to the same extent. Additionally, even very low loading intensities (e.g., walking training) combined with blood flow restriction resulted in substantial improvements in muscle function (Ozaki et al., 2011; Abe et al., 2010). Therefore, LL-BFR training could potentially expand the options for personalized training and rehabilitation programs, particularly in situations where high-load intensities are not appropriate.

Although LL-BFR training produces less mechanical tension compared to traditional HLR training, the metabolic stress generated by blood flow restriction can compensate for this disadvantage. Studies have shown that metabolic stress is equally important as mechanical tension in promoting neuromuscular function (Moritani et al., 1992; Duchateau et al., 2021). Blood flow restriction leads to local hypoxia, increasing lactate production, which subsequently promotes the secretion of hormones (e.g., growth hormone, insulin-like growth factor-1, and vascular endothelial growth factor), thereby stimulating muscle protein synthesis (Manini et al., 2012; Yinghao et al., 2021; Ferguson et al., 2018). In a hypoxic environment, early recruitment of fast-twitch fibers occurs, and the activation of high-threshold motor units enhances the involvement of these fibers, thereby increasing muscle strength, power, and speed (May et al., 2022; Yasuda et al., 2011). Additionally, metabolic stress induces neuromuscular adaptations that increase motor unit activation, further improving muscle strength and endurance (Castilla-López & Romero-Franco, 2023; Moritani et al., 1992; Duchateau et al., 2021). Some evidence suggests that LL-BFR training may be less effective in neuromuscular activation compared to HLR training on surface electromyography (sEMG) parameters (Manini & Clark, 2009; Cook, Murphy & Labarbera, 2013). This could explain the minor differences in training benefits between LL-BFR and HLR training (trivial effect, ES = −0.19). However, higher sEMG amplitudes do not necessarily indicate greater motor unit recruitment. Therefore, more detailed and comprehensive research methods are required to accurately assess the impact of LL-BFR and HLRT on neuromuscular activation. In summary, despite the lower mechanical tension in LL-BFR training, the metabolic stress-induced physiological and neural adaptations compensate for this limitation. This makes LL-BFR training an effective training method with benefits in strength, power, and speed comparable to HLR training, particularly suitable for older adults and individuals undergoing rehabilitation.

This meta-analysis of moderator variables revealed no significant difference in training effects between durations of <8 weeks and ≥8 weeks. This underscores the advantages of LL-BFR training in terms of time efficiency and adaptability. While 3–4 week regimens in the present review can achieve results comparable to HLR training, durations longer than 4 weeks are recommended for sustained improvements. Additionally, training frequency did not have a significant moderating effect on training outcomes. However, based on practical considerations and the balance of the training load, it is recommended to perform moderate-frequency training 2 to 3 times per week. This helps to better integrate into regular training routines while avoiding overtraining or undertraining.

The relative safety of LL-BFR training has not been thoroughly investigated; however compared to other strength training methods, its safety concerns primarily stem from potential risks induced by blood flow restriction, such as blood coagulation, cardiovascular responses, and oxidative stress (Loenneke et al., 2011). It is important to note that the data presented in this meta-analysis is based on healthy participants, and the safety profile may differ for individuals with other comorbidities or disease states. This meta-analysis examined the moderating effects of occlusion pressure (70–220 mmHg or 40–80% arterial occlusion pressure, AOP) and cuff width (5–18 cm) on training outcomes and found no significant impact, which is consistent with a previous meta-analysis (Lixandrão et al., 2018). Therefore, to ensure safety while maintaining training effectiveness, it is advisable to use lower occlusion pressure. The studies included in this review utilized varied pressure prescriptions, with most employing fixed occlusion pressure values and a few using AOP percentages. Given individual differences, it is recommended to adopt personalized occlusion pressure prescriptions (i.e., AOP percentages) and account for cuff width to prevent underestimation or overestimation of occlusion pressure, which could lead to suboptimal training effects or safety issues.

Limitations

Several potential limitations of this meta-analysis warrant cautious interpretation of the findings. Firstly, the number of studies comparing the effects of BFR and HL-RT training on muscle power and speed is insufficient. This necessitates cautious interpretation of the comparative results of these two training methods and also limits our analysis and understanding of related moderator variables. Future research should focus more on comparative studies in the areas of muscle power and speed to provide more comprehensive insights. Secondly, the included studies did not report adverse reactions or injuries related to LL-BFR training. However, this does not mean that BFR training is without potential safety issues, which may be due to limitations in study design or reporting. Therefore, LL-BFR training should be applied with caution, with comprehensive risk assessment and monitoring. Thirdly, the studies included varied in their occlusion pressure prescriptions, adding complexity to the analysis of moderator variables. Since individual responses to the same pressure can differ, personalized blood flow restriction protocols are crucial. Future BFR training research should adopt occlusion pressure prescriptions based on arterial occlusion pressure (AOP percentages) to better accommodate individual differences and ensure the effectiveness and safety of the training.

Conclusions

The studies included in this meta-analysis were of moderate quality. This analysis indicates that LL-BFR and HLR training produce similar effects on muscle power, jump performance, and speed, with LL-BFR showing slightly less improvement in maximal strength compared to HLR. Additionally, individual characteristics (i.e., age, gender, and training status) and training parameters (i.e., training duration, frequency, cuff pressure, and cuff width) do not significantly moderate the training effects of either method. Therefore, LL-BFR training appears to provide a viable and effective alternative to traditional HLR training, suitable for healthy individuals of all ages.

Supplemental Information