Blood flow restriction training in post-operative orthopedic rehabilitation: A systematic review and meta-analysis of randomized controlled trials

Study design

This review was designed as a systematic review and meta-analysis, focusing exclusively on RCTs. The entire process adhered to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 2020 guidelines for reporting.

Eligibility criteria

Studies were considered eligible if they were RCTs involving adults who had orthopedic surgery and took part in postoperative rehabilitation that included BFRT. To qualify, trials needed to compare BFRT with a non-occlusive approach, like standard physiotherapy or conventional resistance training, and report at least one outcome related to muscle strength, muscle size, pain, or balance. Studies were not included if they involved non-surgical groups, were not RCTs, were reviews, or did not have a comparison group.

Search strategy

We performed a literature search in PubMed, Web of Science (WoS), Embase, Cochrane Central, and Google Scholar, and we screened studies from the Emerald database and the Saudi digital library (SDL). Keywords and phrases were based on the population, intervention, comparison, and outcome framework and included (“blood flow restriction therapy” OR “BFRT”) AND (“orthopedic” OR “surgery”) AND (“post-operative rehabilitation”) AND (“strength” OR “range of motion” OR “pain” OR “Balance” OR “function”). Boolean operators were used to improve specificity. No language restrictions were applied. In addition, trial registries were checked, and reference lists and citation tracking of included studies were reviewed to identify any further relevant studies. All citations were managed using Zotero. The last search dates for PubMed, WoS, Google Scholar, Emerald, and SDL were July 20, 2025, whereas Embase and Cochrane Central were last searched on September 05, 2025.

Study selection

After duplicate entries were removed, titles and abstracts were screened independently by two reviewers (AAA and SAA) for relevance. Studies that met the eligibility criteria were reviewed in full for final inclusion. Any discrepancies were resolved through discussion, and a third reviewer (WSA) was consulted when needed. The selection process is shown in the PRISMA flow diagram [Figure 1].

Close

Figure 1:

Preferred Reporting Items for Systematic Reviews and Meta-analyses flow diagram.

Export to PPT

Data extraction

Data extraction was performed independently by two reviewers (AAA, SAA) using a preformatted spreadsheet. For each study, we noted important details, including the author, year, journal, study design, and participant characteristics, such as sample size, mean age, type of surgery, and the BFRT intervention protocol, including occlusion pressures. For comparison groups, we recorded rehabilitation protocols. We recorded outcomes such as muscle strength, size, pain, and balance, along with the measurement tools used. When possible, we gathered means, standard deviations, and sample sizes. Discrepancies were resolved through discussion, and if necessary, by a third reviewer (WSA).

For consistency, when multiple follow-up points were reported, we prespecified time windows for short-term outcomes: Early (≤6 weeks) and later (8–12 weeks). If more than 1 time point fell within a window, the assessment nearest to the midpoint was extracted.

RoB and certainty of evidence

The RoB was assessed independently by two reviewers (AAA, WSA) using the Cochrane RoB 2 tool.^[31] The certainty of evidence for each outcome was evaluated using the grading of recommendations assessment, development, and evaluation (GRADE) framework, considering RoB, inconsistency, indirectness, imprecision, and publication bias.

Data synthesis

Where at least two studies reported the same outcome using compatible measures, meta-analyses were conducted. Continuous outcomes were combined using standardized mean differences (SMDs) and 95% confidence intervals (CIs). We applied a random-effects model to address the differences between studies. We assessed heterogeneity using the Chi-square test and the I-squared statistic (I²). When multiple follow-up points were available, outcomes were prespecified into early (≤6 weeks) and later (8–12 weeks) windows, and forest plots were stratified by these subgroups. The assessment nearest to each window midpoint was used. All analyses were performed using RevMan Web.

Study overview and population

Eleven RCTs conducted between January 2015 and February 2025 included 293 participants (159 in the BFRT group and 134 in the control group). Most studies focused on lower limb surgeries, particularly ACLR and TKA, while one looked at the upper limb. Participants’ ages ranged from 19 to 61 years. BFRT protocols involved using cuffs or bands at 40–80% of arterial occlusion pressure (AOP) during low-load exercises, which were 20–30% of 1RM, for 2–12 weeks, 2–3 times a week. Participants received standard physiotherapy, proprioceptive neuromuscular facilitation (PNF), low-load resistance training, or high-load resistance training. Key study details are in Table 1.

BFRT intervention

BFRT protocols used cuffs or bands to restrict blood flow during low-load exercises. Occlusion pressures ranged from 40% to 80% AOP or fixed values like 130 mmHg [Table 2]. Exercises included knee extensions, squats, step-ups, and cycling. One study involved upper limb training. Interventions lasted 2–12 weeks, with 2–3 sessions per week. Loads were 20–30% 1RM, with three to four sets of 10–15 repetitions. Control groups received standard rehabilitation. Protocol details are in Table 3.

RoB and certainty of evidence

Of the 11 studies included, one was rated as low risk and 10 had some concerns [Figure 2 and Table 4]. Common issues included a lack of blinding for assessors, unclear allocation concealment, and protocol deviations. Based on GRADE, the overall certainty was moderate for muscle strength and muscle size and very low for pain and balance [Table 2].

Close

Figure 2:

Risk of bias assessment, using the Cochrane risk of bias 2 tool.

Export to PPT

Outcome measurement methods

Muscle strength outcomes

Nine studies assessed strength^[32-40] and eight were included in the meta-analysis.^{[32-34,36-40]} Most studies examined knee flexor and extensor strength using isokinetic dynamometry at speeds ranging from 60°/s to 300°/s, measured in kg or Nm. One study assessed upper limb strength, specifically grip, pinch, and wrist flexors and extensors, in patients with distal radius fracture surgery.^[36]

Control treatments included standard physical therapy (n = 4),^{[32,36,37,40]} heavy-load resistance training (n = 2),^[35,41] low-intensity resistance training (n = 4),^{[33,34,38,42]} and PNF (n = 1).^[39]

Two studies found no notable differences in strength gains,^[37,40] while seven studies reported significant improvements with BFRT.^{[32-36,38,39]} Two trials compared BFRT at 80% and 40% AOP, showing greater strength gains with higher occlusion pressures.^[34,38]

The meta-analysis demonstrated significant improvements in muscle strength favoring BFRT over controls (SMD = 0.90; 95% CI: 0.44–1.35; I² = 77%, χ² = 68.37, df = 16, Z = 3.84, P = 0.0001) [Figure 3].

Close

Figure 3:

Forest plot illustrating the differences in muscle strength between the blood flow restriction training (BFRT) and control groups. Numbers (1–3) denote distinct outcome measures reported within the same study. Positive standardized mean difference (SMD) indicates greater improvement with BFRT; negative SMD indicates greater improvement with control. DL: DerSimonian and laird, Z: Z-value, df: Degrees of freedom.

Export to PPT

Muscle size outcomes

Six studies examined muscle size using circumference, magnetic resonance imaging, ultrasound, or volume,^{[33.34,37,38,40,42]} and four of these studies found larger increases in muscle size with BFRT.^{[33,34,38,40]} Two of those studies also showed extra benefits from using higher occlusion pressures.^[34,38] The meta-analysis confirmed significant improvements in muscle size favoring BFRT (SMD = 0.74; 95% CI: 0.34–1.14; I² = 46%, χ² = 22.26, df = 12, Z = 3.63, P = 0.0003) [Figure 4].

Close

Figure 4:

Forest plot illustrating the differences in muscle size between the blood flow restriction training (BFRT) and control groups. Numbers (1– 4) denote distinct outcome measures reported within the same study. Positive standardized mean difference (SMD) indicates greater improvement with BFRT; negative SMD indicates greater improvement with control. DL: DerSimonian and laird, Z: Z-value, df: Degrees of freedom.

Export to PPT

Post-operative pain

Six studies examined post-operative pain using VAS, KOOS, or NPRS.^{[33,35,36,38,40,41]}

Four studies reported less pain with BFRT,^{[33,35,36,40]} one found less joint pain but more muscle soreness,^[41] and one found no difference.^[38] The combined results indicated no significant difference (SMD = 0.33; 95% CI: −1.16–1.82; I² = 94%, χ² = 79.93, df = 5, Z = 0.43, P = 0.67).

Exploratory subgroup analyses by comparator type, occlusion pressure, and outcome measurement were performed for pain outcomes; these did not meaningfully reduce heterogeneity, which may be influenced by other factors not tested, such as patient age, baseline pain, surgical procedure, or sample size, so results are presented according to follow-up duration subgroups only [Figure 5].

Close

Figure 5:

Forest plot illustrating the differences in post-operative pain between the blood flow restriction training (BFRT) and control groups. Negative standardized mean difference (SMD) indicates greater improvement with BFRT; positive SMD indicates greater improvement with control.

Export to PPT

Balance outcomes

Three studies used YBT or postural stability measures to assess balance.^[32-34] Two studies reported better balance with BFRT.^[32,33] However, one study found no difference.^[34] The combined analysis showed no significant effect of BFRT on balance (SMD = −0.07; 95% CI: −0.77–0.63; I² = 71%, χ² = 13.75, df = 4, Z = 0.19, P = 0.85) [Figure 6].

Close

Figure 6:

Forest plot illustrating the differences in balance (8–12-week follow-up) between the blood flow restriction training (BFRT) and control groups. Numbers (1–2) denote distinct outcome measures reported within the same study. Positive standardized mean difference (SMD) indicates greater improvement with BFRT; negative SMD indicates greater improvement with control.

Export to PPT

Safety and adverse events

Adverse events were systematically extracted. Seven studies explicitly reported no adverse events of clinical concern, often specifying the absence of deep vein thrombosis, neurovascular compromise, wound complications, or excessive pain.^{[33-36,38-40,41]} Three studies did not report on safety outcomes.^[32,39,42] One study noted knee stiffness and lateral knee pain 16 weeks after BFRT rehabilitation.^[37] No major thromboembolic or wound-related events were observed.