Can low‐load blood flow restriction training elicit muscle hypertrophy with modest inflammation and cellular stress, but minimal muscle damage?

Blood flow restriction (BFR) training is characterised by the application of a pneumatic cuff to the working limb during resistance or aerobic exercise. When performed at low loads (20–50% one repetition maximum (1‐RM)), BFR training can greatly improve skeletal muscle strength and size (Takarada et al. 2000). Therefore, improvements in skeletal muscle mass and function may be achieved with low‐load BFR training in patients who are contraindicated from performing traditional heavy‐load resistance exercise. Although the mechanisms mediating these adaptations are somewhat unclear, the localised hypoxia and metabolic stress during BFR exercise are thought to stimulate various cellular events. These cellular events include inflammation and oxidative stress, and are potentially mediated by stress‐induced proteins such as heat shock chaperone proteins. However, these responses are poorly characterised and their effect on muscle damage and hypertrophy are unclear in BFR training.

Regeneration and hypertrophy of skeletal muscle after damaging exercise requires a co‐ordinated cellular response, including the infiltration of pro‐ and anti‐inflammatory macrophages. A potential mediator of this regenerative process in BFR exercise is the expression of heat shock proteins (HSPs). During stress, HSPs function as molecular chaperones in the cell and may play a role in protein synthesis, folding, transport and turnover. An acute bout of BFR exercise can upregulate HSP expression within myofibres and induce translocation of HSPs to cytoskeletal structures (Cumming et al. 2014). This is perhaps related to metabolic stress and structural myofibre damage.

The effect of low‐load BFR training on skeletal muscle damage is somewhat unclear due to inconsistencies in current literature. This poses the question; is BFR training achievable in low‐physical‐functioning clinical populations? Nielsen and colleagues therefore aimed to investigate the effect of a longitudinal BFR training programme on markers of muscle damage, cellular stress and inflammation, including pro‐ and anti‐inflammatory macrophage activation and infiltration. This investigation is described in a recent article published in The Journal of Physiology (Nielsen et al. 2017).

Nielsen et al. (2017) performed two randomised controlled trials to achieve their aim. In a 3‐week study, the myocellular damage and inflammatory response to low‐load resistance BFR exercise was compared to non‐BFR control exercise performed at the same intensity. A separate 1‐week study compared the acute responses between low‐load BFR training and traditional non‐BFR heavy‐load resistance exercise. The authors hypothesised that 3 weeks of low‐load BFR training would increase myocellular damage and inflammation early in the training programme, while non‐BFR low‐load resistance training would not. Additionally, low‐load BFR training and traditional heavy‐load resistance training were expected to induce similar muscle damage and inflammation.

In the 3‐week study, 20 recreationally active 18‐ to 35‐year‐old men performed 23 sessions of unilateral knee extensor exercise at 20% 1‐RM. Twelve participants were randomised to BFR training (BFRE), whilst eight participants were randomised to work‐matched training without BFR (LLE). A pressure‐adjustable pneumatic cuff (13.5 cm width) was applied proximally to the upper leg, and inflated throughout BFRE training sessions at 100 mmHg. Dynamic contractions (knee extension) were performed to failure. Four muscle biopsies of the vastus lateralis were obtained; baseline, before session eight, and 3 and 10 days after training ceased. Immunofluorescent staining of CD68 and CD206 was conducted to quantify macrophages, and the ratio of CD68 to CD206 was used to distinguish between M1 and M2 macrophage populations. Immunofluorescence and immunoblotting was used to quantify and localise HSP27 and HSP70.

In the 1‐week study, 20 recreationally active males (also 18–35 years) were randomised to BFRE or traditional heavy‐load resistance training without restriction (HLE). The BFRE training group performed seven sessions of unilateral knee extensor exercise to failure at 20% 1‐RM, under the same restriction conditions as the 3‐week study. HLE participants performed the same exercise to failure over three sessions, at 70% 1‐RM. Venous blood samples were collected on the first and last exercise sessions; pre‐, 5, 15, 60 and 180 min post‐session, then again at 24 h. Creatine kinase (CK) was quantified as a marker of muscle damage and subjects scored their delayed onset of muscle soreness (DOMS). Inflammatory markers monocyte chemoattractant protein 1 (MCP‐1), interleukin‐6 (IL‐6) and tumour necrosis factor α (TNF‐α) were quantified in the blood samples. Total antioxidant capacity and glutathione were also quantified as oxidative stress markers.

After assessing the homogeneity of variance and Gaussian distribution of the data collected, the authors appropriately used mixed linear model analysis, unpaired t testing and Wilcoxon signed‐rank testing to elucidate any differences between experimental groups.

In the 3‐week study, modest increases in pro‐inflammatory M1 macrophages were present in skeletal muscle samples between pre‐training to 3 days post‐training in both groups, although no differences were detected between groups. In line with the authors’ hypothesis, there were no changes in anti‐inflammatory M2 macrophages following exercise in the LLE group, yet the BFRE group showed modest increases up to 10 days post‐training. This finding is interesting, given that the anti‐inflammatory profile is related to skeletal muscle regeneration, remodelling and myogenic satellite cell differentiation. The authors quantified the number of centrally located nuclei in the muscle samples as a marker of recent regeneration, finding modest increases in type II fibres between baseline and 10 days post‐training in the BFR group, and tendencies of increased central nuclei in type I fibres. As hypothesised, no changes were detected in the LLE control group, suggesting that minor muscle damage and regeneration occurred exclusively in the BFRE group. Using immunofluorescence and western blotting to quantify HSPs in skeletal muscle, no changes in HSP70 were observed in the BFRE group, despite minor elevations in the LLE group 3 days post‐training. Small but significant elevations in HSP27 were noted in the BFRE group between baseline and day 8 of training, whilst the LLE group showed increased HSP27 between baseline and 3 days post‐training. Immunofluorescence staining localised HSP27 to the skeletal muscle membrane of the BFRE participants, which is not surprising given that HSP27 is a molecular chaperone during cellular stress, preventing the denaturation and aggregation of proteins.

In the 1‐week study, authors quantified CK as a marker of muscle damage, finding no changes in the BFRE group, despite predictable increases in CK 24 h after the first training session in the HLE group. Minor elevations in DOMS were observed in both groups 24 h following the first and last training sessions. When assessing inflammatory markers in venous blood samples, small but significant changes in MCP‐1, IL‐6 and TNF‐α were observed in both groups. The modest inflammatory response was unexpected given that heavy‐load resistance exercise often elicits marked inflammation, and participants were unaccustomed to the exercise. However, the inflammatory response complemented the muscle damage results in both the 3‐ and 1‐week studies, with only minor indications of myocellular damage observed. Furthermore, markers of oxidative stress (total antioxidant capacity and glutathione) were unchanged in response to BFRE training, despite increasing in the HLE group following the first and last session.

The authors should be commended for their methodological considerations, particularly for noting the use of standard cuff pressures for all BFR participants. Prescription of a uniform pressure most likely causes variable restriction between participants, and thus the BFR exercise field is now adopting methods to personalise restriction based on variable subjective parameters. However, the authors rightly showed that there was little variation in both blood pressure and limb circumference between participants, and so any effect of individual differences was likely minor.

Overall, the authors used thorough analytical techniques to explore the role of inflammatory markers, macrophage infiltration, cellular stress and muscle damage in BFR training. Together, the 3‐ and 1‐week studies show that BFR training exerts small but significant changes in inflammation, macrophage infiltration and cellular stress in skeletal muscle, despite minor indications of muscle damage. The authors highlighted an interesting finding: BFR exercise increased the number of centrally located nuclei in the skeletal muscle sections, but non‐BFR training did not. The presence of centrally located nuclei indicates recent fusion of satellite cells to the muscle fibre. This phenomenon can occur following the regeneration of damaged muscle fibres. Alternatively, this phenomenon can occur following the addition of myonuclei for muscle hypertrophy, independent of muscle damage. Given that minor muscle damage occurred in this study, authors suggested that the centralised nuclei were a result of myonuclei addition for muscle hypertrophy. This theory was supported by previous findings that 3 weeks of BFR training causes muscle hypertrophy and proliferation of myogenic satellite cells (Nielsen et al. 2012).

In conclusion, these findings suggest that low‐load BFR training can elicit muscle hypertrophy without inducing significant muscle damage and oxidative stress. This hypertrophy is often comparable to heavy‐load resistance training. These results add further support for the use of low‐load BFR training in populations who are at risk of low muscle mass, including bedrest patients, ageing populations and musculoskeletal rehabilitation patients.

Additional information

Competing interests

None declared.