Fitness · Research
Blood flow restriction training vs high-load resistance training: high loads built more strength overall, and training status changed the picture
A 2024 systematic review and meta-analysis pooled 53 studies that compared low-load blood flow restriction training (BFR-RT) with high-load resistance training (HL-RT) in healthy people. Overall, HL-RT led to larger strength gains (effect difference −0.335) and muscle hypertrophy was similar (−0.067). In trained people the results leaned toward BFR-RT, but those studies were short and the muscle-size estimate rests on only 3 comparisons.
The short version
This 2024 review in Sports Medicine – Open compared two ways of training for muscle strength and muscle hypertrophy in healthy people. One was blood flow restriction training (BFR-RT): lifting light loads, below 50% of your one-repetition maximum (1RM), while blood flow to the working limb is partly restricted. The other was high-load resistance training (HL-RT), above 65% of 1RM. The authors pooled 53 studies that tested the two head to head.
Across everyone, HL-RT produced bigger strength gains. The effect difference was −0.335 (95% CI −0.515 to −0.156). Muscle growth came out about the same: −0.067 (95% CI −0.205 to 0.071, P = 0.34).
The paper reports these as Hedges' g, an effect size. For each study, it turns each group's before-to-after change into a standardized score, using the means, standard deviations, group sizes and the correlation between pre- and post-test results, with a correction for small samples. The effect difference is the BFR-RT score minus the HL-RT score. Below zero favours HL-RT, above zero favours BFR-RT. When the 95% confidence interval (CI) doesn't include zero, the difference is statistically significant.
Then the authors sorted people by training status. In untrained people, HL-RT won on strength (−0.552) and muscle growth was similar. In trained people the direction flipped. Strength leaned toward BFR-RT (0.491, from 14 comparisons), and so did muscle growth (0.695). That second number rests on only 3 comparisons from three studies, and every trained study lasted 9 weeks or less.
“The main finding of the present study was that training status was an important influencing factor in the effects of BFR-RT.”
Geng et al. · Abstract · full text
What the researchers did
Earlier studies disagreed. Some found bigger strength gains with HL-RT, others found similar gains, and some reported more with BFR-RT. A previous meta-analysis couldn't test moderators such as training frequency because there weren't enough studies yet. This team wanted to see which factors change the comparison.
They searched three English-language databases and two Chinese-language ones. A study qualified if it included healthy people, compared BFR-RT (below 50% 1RM) with HL-RT (above 65% 1RM), measured muscle strength or muscle size before and after training, and scored 4 or more on the PEDro quality scale. Fifty-three studies made it in: 51 for strength (1,164 participants, 63 comparisons) and 28 for muscle size (703 participants, 34 comparisons). One study can add several comparisons when it tested more than one protocol.
“Trained” meant athletes or people who already did regular resistance training before the study. “Untrained” meant sedentary people or people without regular resistance training. Strength was measured with dynamic tests such as 1RM, or with isometric or isokinetic tests. Muscle size was measured with MRI, CT or ultrasound.
The authors used a random-effects model, because the studies differed in people, protocols and measurements. Then they ran subgroup analyses by training status, sex, age, upper or lower limbs, training duration and frequency, type of strength test, and, for muscle size, where on the limb it was measured.
Key findings
| Measure | Effect difference | 95% CI | Location |
|---|---|---|---|
| Strength, everyone (63 comparisons; 51 studies, 1,164 participants) | −0.335 | −0.515 to −0.156 | Results; Table 3; Fig. 2 |
| Strength heterogeneity (I²) | 60.65% | – | Table 3 |
| Strength, trained (14 comparisons) | 0.491 | 0.154 to 0.827 | Results; Table 3; Fig. 3 |
| Strength, untrained (49 comparisons) | −0.552 | −0.722 to −0.382 | Results; Table 3; Fig. 3 |
| Strength, trained vs untrained gap | Q = 29.39, P < 0.01 | – | Results; Table 3 |
| Strength, untrained, 1RM test (37 comparisons) * | −0.715 | −0.909 to −0.522 | Table 3 |
| Strength, untrained, isometric or isokinetic test (24 comparisons) * | −0.422 | −0.657 to −0.188 | Table 3 |
| Hypertrophy, everyone (34 comparisons; 28 studies, 703 participants) | −0.067 (P = 0.34) | −0.205 to 0.071 | Results; Table 4; Fig. 4 |
| Hypertrophy, trained (3 comparisons) † | 0.695 | 0.189 to 1.200 | Results; Table 4; Fig. 5 |
| Hypertrophy, untrained (31 comparisons) | −0.128 (P = 0.08) | −0.272 to 0.015 | Results; Table 4; Fig. 5 |
| Hypertrophy, trained vs untrained gap | Q = 9.41, P < 0.01 | – | Results; Table 4 |
| Publication bias tests | strength: Begg P = 0.84, Egger P = 0.41; hypertrophy: Begg P = 0.16, Egger P = 0.65 | – | Results |
| Source: Geng et al. 2024, Results, Tables 3 and 4. * The gap between the two test types had P = 0.06, not significant. † Only three studies; the authors say this result should be interpreted with caution. The trained and untrained rows come from subgroup analyses, which compare groups of studies. Nobody was randomized by training status. CI = confidence interval · 1RM = one-repetition maximum · Full text (PMC) | |||
How to read the result
Training status is where the results clearly split. The gap between trained and untrained people was significant for strength (Q = 29.39) and for muscle size (Q = 9.41), both P < 0.01. Within the trained and untrained groups, none of the other moderators the paper tested, such as sex, age, limb, training duration or frequency, showed a significant difference.
For untrained people, HL-RT built more strength (−0.552, 95% CI −0.722 to −0.382, 49 comparisons). Muscle growth didn't differ significantly (−0.128, 95% CI −0.272 to 0.015, P = 0.08). A few smaller subgroups leaned toward HL-RT for size, such as the upper limbs (−0.354, 8 comparisons) and the proximal thigh, the part nearer the hip (−0.417, 4 comparisons). Still, the paper found no significant differences between limbs or measurement regions.
How strength was tested may matter too. In untrained people, HL-RT's edge was larger on 1RM tests (−0.715) than on isometric or isokinetic tests (−0.422), with P = 0.06 for that gap. The authors raise this in the paper's Background: a 1RM test is close to what the HL-RT groups practiced, while BFR-RT groups never train with high loads.
For trained people, BFR-RT came out ahead on strength (0.491, 95% CI 0.154 to 0.827, 14 comparisons) and on muscle size (0.695, 95% CI 0.189 to 1.200, 3 comparisons). The Key Points are careful about it.
“Trained individuals may obtain greater muscle strength and hypertrophy gains in BFR-RT compared to HL-RT.”
Geng et al. · Key Points · full text
The wording isn't consistent. The Key Points and the Abstract's Results say trained people “may” gain more, the Abstract's Conclusion and the main Conclusion say they “can obtain” greater gains, and the Discussion says they “will get” them. This article sticks with “may.”
The overall results held up in the checks the authors ran. Removing one study at a time didn't change the strength or the muscle-size result, and the publication bias tests found no sign of bias.
What this doesn't tell us
Whether BFR-RT beats heavy training for experienced lifters over months or years. Every trained study lasted 9 weeks or less. One of the three trained muscle-size studies ran for just 2 weeks, at 5 sessions a week. The trained group was also mixed: powerlifters, handball, tennis and soccer players, young swimmers, adolescent wrestlers and 16- to 17-year-olds. Six of the 12 trained strength studies appeared in journals published in mainland China, such as China Sport Science and Technology, so they were likely written in Chinese; the paper doesn't list each study's language.
“Because of the sparse number of studies, the results comparing muscle hypertrophy in trained individuals should be interpreted with caution.”
Geng et al. · Limitations · full text
Why the trained HL-RT groups gained so little. The Discussion gives an average relative strength change in trained people of 8.4% with BFR-RT and 3.96% with HL-RT. A heavy-load gain under 4% is small, and these trials were short. The formula is also printed as (pre-training − post-training)/pre-training. Taken literally, that gives a negative number for a strength gain, yet both percentages are positive, so the sign looks reversed. Treat both percentages with care.
Whether it applies to women. The authors write that “the lack of studies including females limited generalizability of the findings.” In the muscle-size analysis, only one study of untrained people included female participants.
Whether BFR-RT is safe for you. This review didn't look at safety or side effects.
How solid the method is. When a study measured an outcome in several ways, the results were averaged into one. Where a study didn't report the correlation between pre- and post-tests, it was estimated with a formula. The training-status findings come from subgroup analyses, which are observational. The full text doesn't mention a registered protocol. Data from one study were left out because its author didn't respond, though the sensitivity analysis suggests no single study changed the results.
How to apply it
The authors see BFR-RT as a possible add-on for trained people and as an option for people who can't train with heavy loads. They don't give a specific program.
“From a practical standpoint, BFR-RT could be a beneficial supplemental training protocol for trained population.”
Geng et al. · Conclusion · full text
“Finally, it is important to highlight that BFR-RT remains a valid and effective alternative for people who cannot perform high-load resistance training.”
Geng et al. · Conclusion · full text
Opinion, not a finding from the paper: If I were in your spot and could train heavy, I would keep HL-RT as the base, especially if you're new to lifting, because that's where this review found a clear strength advantage. If you already train, I would treat BFR-RT as a possible add-on to learn properly rather than a replacement, because the trained results rest on a few short studies. For more on how load relates to reps, see the site's article on load and repetition range.
A safety note: this review didn't assess safety. BFR training uses a cuff that puts pressure on the arm or leg, so learn it from a qualified professional, and if you have a health condition, check with a clinician first. This article isn't injury or rehab advice.
Sources
Source: Geng et al., Sports Medicine – Open 2024; 53 studies (strength: 51 studies, 1,164 participants; hypertrophy: 28 studies, 703 participants). DOI 10.1186/s40798-024-00719-3.
- StudyGeng Y, Wu X, Zhang Y, Zhang M. Potential Moderators of the Effects of Blood Flow Restriction Training on Muscle Strength and Hypertrophy: A Meta-analysis Based on a Comparison with High-Load Resistance Training. Sports Medicine – Open. 2024;10:58.
DOI: 10.1186/s40798-024-00719-3 · Full text (PMC, CC BY 4.0) · Publisher
This is a research explainer for general education, not medical advice.
