Treatments and gadgets. 13 minute read.
Blood flow restriction training: what it is for, and who should not do it
By the PocketPhysio editorial team. Published 2026-10-01.
What is blood flow restriction training?
Blood flow restriction training is exercise done with a light load while a cuff around the top of the arm or leg restricts the blood leaving the limb. An expert position paper in Frontiers in Physiology describes what the cuff does as partly restricting arterial flow into the working muscle and fully restricting venous flow out of it, which leaves the muscle short of oxygen with blood pooling in its capillaries (Patterson and colleagues, 2019). Why that lets a light load produce part of what a heavy load produces is not settled, so the outcomes are the part of the literature worth reading. The pressure is measured on the person wearing the cuff, not taken from a chart.
Two things follow from that description, and they shape the whole page. The first is that this is a way of getting strength work done when heavy work is off the table. The second is that a cuff that changes blood flow has a list of people it is not safe for, which is longer and more specific than the equipment suggests.
Why a light load with a cuff can do the work of a heavy one
The honest version of the claim is narrower than the marketing version. An expert position paper in Frontiers in Physiology summarizes the comparisons this way: low load exercise with the cuff produces increases in muscle mass comparable with heavy load training, and it improves strength more than the same light load without a cuff, while generally producing less strength gain than heavy load training does (Patterson and colleagues, 2019).
That is a trade, not a free lunch. You get size and some strength at a load a sore or recently operated limb can handle. You do not get everything heavy lifting gives, and the same paper places its value in populations where heavy loading is inadvisable: after an operation, in cardiac rehabilitation, in inflammatory diseases, and in frail older adults.
So if you can already load a limb heavily without pain, this is the wrong tool. The plain progression of load is simpler and better understood.
Where it is actually used
The knee after surgery is where the evidence concentrates. A 2026 systematic review and meta-analysis pooled 9 randomized trials with 372 participants after ACL reconstruction, 185 of them training with the cuff and 187 without. Quadriceps peak torque favored the cuff groups, rated moderate certainty on GRADE, pooled from 6 trials and 284 participants, with a weighted mean difference of 8.34 (95% confidence interval 7.47 to 9.22) in the units the included trials reported. The Lysholm knee score also favored them, by 5.75 points (0.60 to 10.90) in 230 participants, at low certainty (Hu and colleagues, 2026). One caution about reading that review: its results text and its certainty table do not agree on how many trials sit behind the Lysholm score, the IKDC score and the pain estimates, counting 6 then 5 then 3 in one place against 5 then 4 then 4 in the other, so the participant totals are the figures to trust.
What it did not change is as useful. In the same review there was no difference in IKDC score, pooled from 5 trials and 158 participants at low certainty, none in pain, from 3 trials and 164 participants at low certainty, and none in knee range of motion, from 3 trials and 188 participants at very low certainty. Follow-up in every included trial ran between 6 and 24 weeks, so nothing in this body of work describes a year later. An evidence summary published the same year adds a qualifier of its own, at evidence level 1b with a grade B recommendation: in the first 3 weeks after the operation this method does not appear to improve quadriceps strength or muscle size, and the knee function gains show up at midterm follow-up instead (Wang and colleagues, 2026).
The second place it earns its keep is pain. That same summary, built from 18 documents including 2 clinical guidelines, 13 systematic reviews, 1 evidence summary and 2 expert consensus statements, names the group it suits at evidence level 1b with a grade A recommendation on the JBI system: people in early rehabilitation after ACL reconstruction, and particularly those who cannot tolerate high-intensity training because of significant pain after the operation (Wang and colleagues, 2026).
On this site it appears in one place for that reason, inside the program for ACL reconstruction rehab, where it is named as an option for the early phase when knee pain is what limits loading. The same evidence summary also names early rehabilitation after knee arthroscopy and after fracture fixation among the situations this method suits, at the same evidence level 1b and grade A, and adds that it may be used before an ACL reconstruction as well. Read that next to its own relative contraindications, which include a recent fracture, and you can see why the judgment belongs to the clinician in front of you.
Beyond those, this page has no pooled evidence to offer. No meta-analysis of this method after a total knee replacement or for patellar tendinopathy was found while the page was written, so if a clinician raises it for either of those, treat it as their clinical judgment about your limb rather than as a result with numbers behind it. The programs on those two pages are the part with guideline support.
Who should not do this without clearance
This is the part of the page to read twice. That same 2026 evidence summary sets out the absolute contraindications after ACL reconstruction at evidence level 1b with a grade A recommendation (Wang and colleagues, 2026):
- Severe cardiovascular disease, which it lists as high blood pressure, a history of deep vein thrombosis or a high risk of thrombosis, and peripheral arterial disease.
- Skin irritation or infection.
- Open wounds.
- Pregnancy.
- Kidney insufficiency.
- Disorders of the lymphatic system.
The same summary has a second list, graded lower: relative contraindications at evidence level 2b with a grade B recommendation. It names severe swelling of the limb, fluid in the joint, and anyone at high risk of muscle injury, which it spells out as severe osteoporosis, a recent fracture, clotting disorders, diabetes with nerve damage in the feet, and people who are sensitive to pain or cannot tolerate the cuff being tight (Wang and colleagues, 2026). Lower grade does not mean ignore it. It means the evidence behind it is weaker, not that the concern is smaller, and several of those items describe people who turn up in rehabilitation every week.
A risk stratification paper in Frontiers in Physiology adds the conditions for which it says a physician's clearance, closer monitoring or a different exercise is needed. It names blood pressure at or above 160 mmHg systolic or 100 mmHg diastolic before exercise, and lower readings than that, from 130 systolic or 85 diastolic upward, where there is organ damage, chronic kidney disease or diabetes. It also names a heart attack within the last 3 months, class III or IV heart failure, stable angina as well as unstable angina, uncontrolled heart rhythm problems, acute inflammation of the heart muscle or its linings, an abdominal aortic aneurysm, aortic syndrome, a past stroke or transient ischemic attack, severe or symptomatic valve disease, peripheral artery disease, a 20 mmHg drop in systolic pressure on standing with dizziness or light-headedness, venous thromboembolism and severe pulmonary hypertension (Nascimento and colleagues, 2022).
For one group that paper goes further than clearance. Someone who has high blood pressure together with cardiovascular disease should probably avoid this method altogether, it says, with something else put in the plan instead. Its authors call that possibly overly cautious themselves, and then keep it.
That blood pressure figure is a threshold for ruling somebody out, not a test you can pass. A reading below it does not mean you have been screened, because the same paper sends everyone through a thrombosis risk assessment and a medical history that one reading says nothing about.
Cancer belongs on this page, and the way it belongs is worth being precise about. That same paper routes people through established thrombosis risk scores instead of a flat list, and on the Caprini score it reproduces, malignancy scores 2 points whether it is present or past, and chemotherapy 1 point. Other items on that score catch a lot of ordinary rehabilitation patients: age 41 to 59, swollen legs, varicose veins and smoking at 1 point each; age 60 to 74, a cast within the past month and major surgery lasting more than 45 minutes at 2 points; age over 75, a previous deep vein thrombosis or pulmonary embolism, a family history of thrombosis and inherited clotting disorders at 3 points; and a recent hip, pelvis or leg fracture, a major lower limb joint replacement, a recent stroke or multiple trauma at 5 points.
Read that list beside the timing of rehabilitation and the problem becomes obvious. The weeks when a cuff is most tempting, straight after an operation, are also the weeks the paper identifies as the period of greatest thrombosis risk. It also flags pregnancy and the period after birth as raised-risk, which is why pregnancy sits on the absolute list above.
With a heart condition or high blood pressure, check with your doctor before you start. That line is not a formality here, because the cuff raises blood pressure during the set and worsening of ischemic heart disease is on the reported adverse event list. For this method it is also the floor rather than the answer: high blood pressure sits on the absolute contraindication list above, at evidence level 1b with a grade A recommendation, so the conversation with your doctor is about whether a cuff belongs in your program at all, not about how to fit it in.
How often something goes wrong, on the numbers that exist
Rarely, and the numbers need reading slowly, because the same expert position paper uses 0.008% for two different things and that is how the figure gets misquoted.
For muscle breakdown it reports isolated case reports of rhabdomyolysis with this method, an analysis of incidence in the published literature that put the risk at 0.07 to 0.2% without giving the number of people behind it, and survey data from Japan, where this style of training has been practiced by a large number of people, suggesting 0.008%. That survey figure comes from the same questionnaire of over 12,000 people that produced the clot numbers in the next paragraph, which is worth knowing before either figure is quoted on its own. The paper's own reading is that exertional rhabdomyolysis during this kind of exercise is possible, but that the evidence does not currently suggest the risk is higher than with ordinary exercise (Patterson and colleagues, 2019).
For clots the same paper reports a large epidemiological questionnaire in Japan of over 12,000 people, in which venous thrombus came out at 0.055% and pulmonary embolism at 0.008%, and it adds that a true medical diagnosis for the pulmonary embolism cases was not confirmed. It also points out that the deep vein thrombosis rate in that survey is lower than the rate reported for the general population in Asia, 0.2 to 0.26% (Patterson and colleagues, 2019). Read that the careful way round. A questionnaire that finds fewer clots than the background rate is more likely to be undercounting them than to be showing protection, so it is a reason to treat self-reported safety data cautiously rather than a reassurance.
None of those figures comes from a study designed to measure harm. The 2026 meta-analysis of the trials after ACL reconstruction found adverse event reporting inconsistent, with only 3 of its 9 trials reporting any safety outcome at all (Hu and colleagues, 2026). Supervised use looks well tolerated. That is a different statement from safe, and the screening above is why.
Why there are no pressures on this page
You will have noticed what is missing. There is no cuff pressure here, no percentage of anything, and no protocol, and all of that is deliberate.
The pressure that slows venous return without closing off arterial flow depends on the circumference of your limb, the width of the cuff and your own blood pressure. Get it wrong one way and nothing happens except discomfort; get it wrong the other way and you have occluded an artery. The 2026 meta-analysis found that the published trials differed substantially in how they calibrated pressure, in the width and type of cuff, in the exercise chosen, in how often people trained and in how long for, which is part of why most of its outcomes sat at low or very low certainty (Hu and colleagues, 2026).
The same reasoning rules out a frequency, a set count and a rep count. All of them belong to a prescription, and a prescription needs someone who has read your history and measured your limb. An article that handed you numbers would be pretending to have done both.
What to ask the clinician who offers it
Five questions tell you most of what you need, and a clinician who uses this routinely will have quick answers to all of them.
- What training have you had in this, and how do you set the cuff pressure for my limb?
- Which parts of my medical history did you check against the contraindications, and what did you find?
- What am I meant to feel, and what would make you stop the session?
- How will we know in 4 to 6 weeks whether it is working, and what do we measure?
- When do we stop using it and go back to ordinary loading?
The last one matters more than it sounds. This is a bridge across a period when a limb cannot take load, and the plan should include the far bank. An evidence summary at evidence level 1b, grade A, also describes the monitoring it expects during a session, including checks of the pulse below the cuff, oxygen saturation and the skin (Wang and colleagues, 2026), so it is fair to ask what will be watched while you train.
Warning signs during and after a session
Some discomfort is expected with this method, and a few things are not. Numbness and tingling in the limb below the cuff point to pressure that is too high: the position paper reports that the wide range of pressures practitioners were using produced a large amount of numbness, and recommends setting the pressure from a measurement taken on the limb instead (Patterson and colleagues, 2019). What to do about it is a published rule, not just practice habit. The 2026 evidence summary puts loosening the cuff immediately for numbness, tingling or severe pain at evidence level 1b with a grade A recommendation, and gives the criteria for stopping the session as numbness, dizziness, abnormal pain or a limb that goes pale, at the same level and grade (Wang and colleagues, 2026). So stop the set and have the pressure rechecked rather than pushing through.
A limb that goes cold or blue, or a cuff you cannot tolerate at all, ends the session too. The same summary lists nerve paralysis, injury from lack of blood supply, and clots among the serious risks, which is what those small signs are there to catch early.
The serious events are rare and specific, and all of them are in the warning list below: a possible clot in a calf or thigh, a possible clot in an arm, a possible clot in the lungs, chest pain, then the dark urine and badly swollen or very weak muscle that point to muscle breakdown. Learn them before your first session.
The arm one is easy to miss, because most people picture a clot as a leg problem. If the cuff goes on an arm, the signs to know are a forearm, hand or upper arm that becomes swollen, warm, tender, red or darker than usual, has swollen veins that are sore to touch, or has a throbbing or cramping pain that feels different from normal muscle ache, and they need medical advice the same day. The case reports in the risk stratification paper include a clot in the vein under the collarbone (subclavian vein thrombosis), which is an upper limb event (Nascimento and colleagues, 2022).
When to see a physio (physical therapist)
Before you buy or borrow a cuff, if nobody has assessed the limb. What decides whether this method is right is a short conversation about your history and a measurement of what the limb can currently do, and both of those change the plan more than the equipment does.
Come in sooner if a knee is still giving way, if a thigh has visibly wasted after surgery or a long spell in a cast, or if pain is the reason you cannot load a limb at all. Bring the operation notes if you have had surgery, and a list of your medical conditions and medicines, because that list is what the screening above is done against. The warning list below overrides everything else on this page, and each of those lines carries its own timing.
See a doctor promptly if
- Emergency: if chest pain, pressure or tightness does not ease quickly when you rest, spreads to your arm, neck, jaw, stomach or back, or comes with sweating, feeling sick, feeling light-headed or being short of breath, call emergency services straight away, as this can be a heart attack. If it eases quickly and none of these happen, get medical advice the same day, and do not exercise again until you have been checked.
- Emergency: you suddenly become short of breath, get chest pain or pain in your upper back that may be worse when you breathe in, cough up blood, your heart beats very fast, or you faint. Call emergency services straight away, even if your leg looks normal, as this can be a blood clot in the lungs (pulmonary embolism).
- Emergency: dark pee (brown, red or tea-colored) after exercise, or muscles that become badly swollen, extremely painful or too weak to use, are not normal soreness. Go to an emergency department straight away, as these can be signs of serious muscle damage (rhabdomyolysis).
- Same day: a forearm, hand or upper arm that becomes swollen, warm, tender, red or darker than usual, has swollen veins that are sore to touch, or has a throbbing or cramping pain that feels different from normal muscle ache. Stop and get medical advice the same day, as this can be a blood clot. If you are also short of breath or have chest pain, call emergency services.
- Same day: a calf or thigh that is newly swollen, warm, tender, red or darker than usual, has swollen veins that are sore to touch, or has a throbbing or cramping pain that does not fit your injury or condition, or that feels different from normal muscle soreness after exercise. This can be a sign of a blood clot (deep vein thrombosis), especially after surgery, an injury, time in a cast or boot, a long trip or a spell of being much less mobile than usual. Get medical advice the same day. If you are also short of breath or have chest pain, call emergency services.
Common questions
What are the contraindications for blood flow restriction training?
An evidence summary published in 2026 drew on 18 documents, including 2 clinical guidelines, 13 systematic reviews, 1 evidence summary and 2 expert consensus statements, and set out the absolute contraindications after ACL reconstruction at evidence level 1b with a grade A recommendation on the JBI system: severe cardiovascular disease, which it lists as high blood pressure, a history of deep vein thrombosis or a high risk of thrombosis, and peripheral arterial disease; skin irritation or infection; open wounds; pregnancy; kidney insufficiency; and disorders of the lymphatic system (Wang and colleagues, 2026). The same summary carries a second, weaker list, its relative contraindications, at evidence level 2b with a grade B recommendation: severe swelling of the limb, fluid in the joint, and anyone at high risk of muscle injury, which it spells out as severe osteoporosis, a recent fracture, clotting disorders, diabetes with nerve damage in the feet, and people who are sensitive to pain or cannot tolerate the cuff (Wang and colleagues, 2026). A separate risk stratification paper adds the conditions for which it says a physician's clearance, closer monitoring or a different exercise choice is needed, among them blood pressure at or above 160 mmHg systolic or 100 mmHg diastolic before exercise, lower readings than that when there is organ damage, chronic kidney disease or diabetes, a heart attack in the last 3 months, class III or IV heart failure, stable or unstable angina, uncontrolled heart rhythm problems, acute inflammation of the heart or its lining, an abdominal aortic aneurysm, aortic syndrome, a past stroke or transient ischemic attack, severe or symptomatic valve disease, peripheral artery disease, blood pressure that drops 20 mmHg on standing with dizziness, venous thromboembolism and severe pulmonary hypertension; for someone who has both high blood pressure and cardiovascular disease, that paper says the method should likely be avoided and another one used instead (Nascimento and colleagues, 2022). Cancer appears in the same paper inside the thrombosis risk scores it recommends rather than on a flat list, with malignancy, present or previous, scored as 2 points and chemotherapy as 1 point on the Caprini score. None of this is a self-screening tool, and the paper itself says the judgment belongs to a clinician who knows your history.
Is blood flow restriction training safe?
In supervised rehabilitation it has a reassuring track record and a short list of serious events that are worth knowing by name, and a risk stratification paper names what has been reported: numbness, nausea, raised blood pressure, headache, venous thrombus, worsening of ischemic heart disease, fainting, tingling, excessive pain, blockage of the central vein of the retina and rhabdomyolysis, and it describes such occurrences as very infrequent but documented, with individual case reports including a subclavian vein thrombosis, a retinal vein occlusion in a person with diabetes and exercise-related fainting in a person on dialysis (Nascimento and colleagues, 2022). An expert position paper reports isolated case reports of rhabdomyolysis with this method, alongside an analysis of incidence in the published literature that put that risk at 0.07 to 0.2% with no denominator given, and a Japanese questionnaire of over 12,000 people suggesting 0.008%; its own reading is that the evidence does not currently suggest the risk of rhabdomyolysis is higher than with ordinary exercise. For clots the same paper reports a questionnaire of over 12,000 people in Japan in which venous thrombus came out at 0.055% and pulmonary embolism at 0.008%, with no confirmed medical diagnosis for the pulmonary embolism cases, and it notes that the deep vein thrombosis rate in that survey is lower than the rate reported for the general population in Asia, which is a reason to treat self-reported safety data carefully rather than a reassurance; it also notes that the wide range of pressures practitioners were using produced a large amount of numbness (Patterson and colleagues, 2019). A 2026 meta-analysis of 9 trials after ACL reconstruction found adverse event reporting inconsistent, with only 3 of the 9 trials reporting safety outcomes at all, so the honest summary is that it looks well tolerated in the trials and that the trials were not designed to settle safety (Hu and colleagues, 2026). Safety here depends on screening: putting the lists together, and keeping in mind that they were written for different purposes, anyone with a history of blood clots, pregnancy, a heart condition, high blood pressure, kidney problems or lymphatic problems needs medical clearance before starting, and so does anyone with severe osteoporosis, a recent fracture, a clotting disorder or diabetes with nerve damage in the feet (Wang and colleagues, 2026; Nascimento and colleagues, 2022).
Can you do blood flow restriction training at home?
Not as something you set up for yourself from an article, and this page gives no pressures for that reason. The pressure that restricts venous return without cutting off arterial flow depends on the size of your limb, the width of the cuff and your own blood pressure, which is why trials measure it on the person rather than using a single number. A 2026 meta-analysis noted that the published trials varied substantially in how they calibrated pressure, in the width and type of cuff, in the exercise used, in how often people trained and in how long the program ran (Hu and colleagues, 2026). A 2026 evidence summary puts it plainly at evidence level 2b with a grade A recommendation: training should be supervised by a professional, so that the cuff is used properly and the exercises are done correctly (Wang and colleagues, 2026). Home use does exist, and it exists as a clinician-prescribed program with a measured cuff pressure, a reviewed medical history and a plan for what to do if something goes wrong; the practical version of this question is which clinician near you is trained in it, and whether you are one of the people who should not be doing it at all.
Does blood flow restriction training build muscle?
Muscle size, yes, by about as much as heavy lifting in the comparisons that have been made, and strength by less than heavy lifting. An expert position paper summarizes it as comparable increases in muscle mass against heavy load training, with strength gains that are better than the same light load without a cuff but generally smaller than heavy load training achieves (Patterson and colleagues, 2019). That trade is the whole point of it. It is offered when heavy loading is not available, which the same paper gives as the period after an operation, cardiac rehabilitation and frail older adults, and not as a shortcut for people who could simply lift heavier. If you can load a limb normally and without pain, the plain progressive program is the simpler option and it needs no cuff.
Is blood flow restriction training used in physical therapy?
Yes, and the knee after surgery is where most of the evidence sits. A 2026 systematic review and meta-analysis of 9 randomized trials with 372 participants, 185 with the cuff and 187 without, found quadriceps peak torque favored the blood flow restriction groups after ACL reconstruction at moderate certainty on GRADE, pooled from 6 trials and 284 participants, and the Lysholm knee score favored them by 5.75 points (95% confidence interval 0.60 to 10.90) from 6 trials and 230 participants at low certainty. The same review found no difference in IKDC score, pain or knee range of motion, the last of those rated very low certainty, and the follow-up in every trial ran between 6 and 24 weeks with nothing at a year or beyond (Hu and colleagues, 2026). An evidence summary published in 2026 names the group it suits at evidence level 1b and grade A: people in early rehabilitation after ACL reconstruction, particularly those who cannot tolerate high-intensity training because of significant pain after the operation (Wang and colleagues, 2026). It is a tool inside a program, not a program.
How often can you do blood flow restriction training?
There is no frequency this page can give you, because frequency is part of a prescription that also covers cuff pressure and load, the exercise itself, and the medical screening that comes before any of it. The published trials disagree with each other on all of those: a 2026 meta-analysis of 9 trials after ACL reconstruction reported substantial variation in how pressure was calibrated, in the width and type of cuff, in the exercise used, in how often people trained and in how long the program ran, which is one reason it rated most of its outcomes at low or very low certainty (Hu and colleagues, 2026). What a clinician decides it from is your stage of recovery, what else is in your week and how the limb responded last time, and the questions worth asking them are listed on this page. Anyone with a history of blood clots or a high thrombosis risk, pregnancy, high blood pressure, a recent heart event, peripheral arterial disease, kidney insufficiency, a lymphatic disorder, or broken or infected skin under the cuff should not start at all without medical clearance, and the same goes for severe osteoporosis, a recent fracture, a clotting disorder and diabetes with nerve damage in the feet, which sit on the same summary's lower graded list. That combined list draws on two sources written for different purposes, one specific to rehabilitation after ACL reconstruction and one a general risk stratification paper, so each item keeps the strength of the source it came from (Wang and colleagues, 2026; Nascimento and colleagues, 2022).
References
- Hu Y, Guo L, Zeng X, Yuan R. Blood flow restriction training after anterior cruciate ligament reconstruction: a systematic review and meta-analysis of randomized controlled trials. Orthopaedic Journal of Sports Medicine. 2026;14(7):23259671261440196. https://doi.org/10.1177/23259671261440196
- Nascimento DC, Rolnick N, Neto IVS, Severin R, Beal FLR. A useful blood flow restriction training risk stratification for exercise and rehabilitation. Frontiers in Physiology. 2022;13:808622. https://doi.org/10.3389/fphys.2022.808622
- Patterson SD, Hughes L, Warmington S, Burr J, Scott BR, Owens J, Abe T, Nielsen JL, Libardi CA, Laurentino G, Neto GR, Brandner C, Martin-Hernandez J, Loenneke J. Blood flow restriction exercise: considerations of methodology, application, and safety. Frontiers in Physiology. 2019;10:533. https://doi.org/10.3389/fphys.2019.00533
- Wang Y, Wang F, Xu B, Yin X. Summary of the best evidence on blood flow restriction training for patients after anterior cruciate ligament reconstruction surgery. Frontiers in Rehabilitation Sciences. 2026;7:1806660. https://doi.org/10.3389/fresc.2026.1806660
- NHS. Chest pain. Page last reviewed 8 August 2023. https://www.nhs.uk/symptoms/chest-pain/
- NHS. DVT (deep vein thrombosis). Page last reviewed 30 April 2026. https://www.nhs.uk/conditions/deep-vein-thrombosis-dvt/
- NHS. Pulmonary embolism. Page last reviewed 25 May 2023. https://www.nhs.uk/conditions/pulmonary-embolism/
- Cleveland Clinic. Rhabdomyolysis. Last updated 24 February 2023. https://my.clevelandclinic.org/health/diseases/21184-rhabdomyolysis
Written and checked by the PocketPhysio editorial team. Last updated 2026-10-01.
This page is general education, not a diagnosis or a personal treatment plan. See a physiotherapist or doctor for advice about your own situation.