Are calves trainable? Low-intensity calf muscle training with or without blood flow restriction: a randomized controlled trial

Study design

A parallel, two-group, randomized study design was used to assess the effects of an 8-week, twice-weekly low-intensity BFR or low-intensity conventional calf muscle RT (noBFR) on functional and structural adaptations. Participants were assigned to groups using blocked randomization via studyrandomizer.com to ensure equal distribution. The independent variables were the two training modalities (BFR and noBFR). The dependent variables were calf circumference (CC), leg stiffness (LS), maximum plantar flexion strength tests, and isokinetic ankle strength tests. Prior to the investigation, ethical clearance was obtained from the IST University of Applied Sciences ethics committee (Ethic Decision No. 022023IST233) according to the Declaration of Helsinki [31].

Participants

Before the recruitment of volunteers, an a priori power analysis for repeated measures analyses of variance (2 groups, 4 measurements) was conducted using G*Power software (Version 3.1.9.7, Universität Düsseldorf, Düsseldorf, Germany) to determine the number of participants needed. The analysis was based on detecting a medium effect size (f=0.25) as seen in a previous study [24], with a significance level of α=0.05 and a desired power of 1-β=0.95 [32]. According to this power analysis, a total of 36 subjects would have been sufficient. However, to account for possible dropouts, a total of 43 healthy men and women between the ages of 18 and 45 were initially verbally recruited in a rehabilitation center. Participants had to be healthy with no absolute or relative contraindications to BFR training [33] and no lower extremity injury in the past 6 months prior to this study. Additionally, volunteers were required to have at least two years of RT experience and not to be taking any legal (i.e., nutritional supplements) or illegal performance-enhancing substances (i.e., anabolic steroids). Informed consent was obtained from all individuals included in this study. One subject was eliminated because it was technically not possible to determine arterial occlusion pressure (AOP). Six participants did not reach the minimum of 12 out of 16 training sessions (75 % adherence) and were therefore also eliminated from the study. Data from 36 subjects were included for final analysis (BFR: 7 females, 32.9 ± 8.8 years, 171 ± 6 cm, 62.6 ± 8.5 kg; 11 males, 28.4 ± 3.6 years, 181 ± 8 cm, 82.2 ± 13.0 kg; noBFR; 8 females, 29.6 ± 3.4 years, 166 ± 5 cm, 65.6 ± 6.6 kg; 10 males, 28.6 ± 4.9 years, 181 ± 7 cm, 84.5 ± 12.9 kg).

Testing

All tests were conducted in a rehabilitation center by the same researchers, using the same equipment, and at a similar time of the day. All participants were familiar with the tests employed in this study, having undergone these assessments previously. Participants were instructed to refrain from any rigorous exercise 48 h prior to the start of the study.

Anthropometry

First, participants’ body weight (BW) was measured using a digital scale (ES-26M-W, Renpho, Eastvale, California, USA) and height was asked for descriptive purposes. Subsequently, CC of both shanks was measured by a well-trained, experienced member of the research team 10 cm below the patella using a tape measure. Such measurements have been previously used and reported to demonstrate excellent reliability [34]. The mean of both sides of a participant was used for further analysis.

Leg stiffness

Following a low intensity, 10-min warm-up on a stationary bike (Cybex 770C Upright Fitness Bike, Cybex International, Inc., Owatonna, Minnesota, USA), LS was assessed using the Optojump photocell system (Microgate Srl, Bolzano, Italy) [24]. Participants first completed 20, 10 and then 7 hops with their hands on hips, with increasing intensity and 30, 60, and 90 s of rest between sets, respectively. They then completed two trials of a seven-maximal hopping test with 90 s of rest between trials. Volunteers were instructed to jump as high as possible with minimal ground contact time. Mean flight and ground contact times from the seven jumps of a single trial, along with participants’ BW, were used to calculate LS using the formula proposed by Ruggiero et al. [35]. The best LS value from the two trials was used for the analysis.

One-repetition maximum

Thereafter, 1RM testing was performed for the bilateral calf raise with extended knees using a seated leg press (Quadrus Funktionsstemme, Kanzlsperger GmbH, Berngau, Germany). Participants first completed three warm-up sets (10, 5, and 2 repetitions) with increasing intensity (80, 100, 120 % of their BW), and 60, 90, and 120 s of rest between sets. The 1RM was then assessed by increasing the resistance for a single repetition until a valid attempt could no longer be performed [36, 37]. A repetition was considered valid if the weight could be lifted through the full range of motion (ROM), verified using a digital goniometer (EasyAngle, Meloq AB, Stockholm, Sweden). The 1RM was identified in all participants within a maximum of three trials.

Next, the 1RM in the unilateral seated calf raise with 90° flexed knees (NA9, Kaphingst GmbH, Lahntal, Germany) was estimated using the Epley formula [38]. The reason we estimated rather than actually testing the 1RM was because the seated calf raise machine did not offer enough resistance for our trained participants. In a previous study, the Epley formula was found to have a low average error [38] and was deemed accurate for machine exercise testing of plantar flexion [39]. The warm-up consisted of 10 repetitions at 70 % BW for each leg. Following a 60 s rest period, participants completed as many valid repetitions as possible at 130 % of their BW per leg. The criteria for a valid repetition correspond to the requirements described in the previous section. Subsequently, the 1RM was calculated from the repetitions completed and the resistance used. The mean of both sides of a participant was used for further analysis.

Isokinetic strength

Afterwards, isokinetic strength testing of plantar- and dorsiflexion was performed using the CON-TREX^® MJ dynamometer. Isokinetic testing has been reported to be reliable and valid for assessing ankle strength parameters [40, 41]. For this, an ankle adapter adjusted at 4 cm height (PHYSIOMED ELEKTROMEDIZIN AG, Schnaittach, Germany), the trunk, the pelvis at 20° hip flexion, and the tested limb was fixed with constraining and Velcro straps. In addition, a hard roll of foam with a diameter of 10 cm was placed under the knees for support. The arms were crossed over the chest and the leg not being tested was placed in a rest position. The participants then performed five submaximal exercise cycles (repetitions) at 60°/s as a specific warm-up. Each cycle included a movement from full plantar flexion to full dorsiflexion (with −3° in each position for safety reasons) and back. Subsequently, to determine mean peak torque in Newton-meter two tests with five repetitions at 60°/s and one test with 25 repetitions at 210°/s and maximal contraction per leg were carried out with 2 min of rest in-between.

Post-testing occurred three to five days after the last training session [24]. All tests were carried out identically to the pre-tests.

Cuff pressure

In the participants allocated to the BFR group, the individual AOP was determined prior to the first training session following the protocol proposed by Vehrs et al. [42]. For this, a 7 cm wide pneumatic BFR cuff (Occlusion Cuff Elite^®, The Occlusion Cuff^®, Belfast, United Kingdom) was applied to participants’ right leg as proximally as possible. Volunteers were then placed half-sitting on a therapy bench to prevent the cuff from being compressed by the benches surface. Subsequently, pulse sounds were monitored dorsomedially at the level of the ankle joint on the anterior tibial artery using a hand-held Doppler device (The Occlusion Cuff Vascular Doppler, The Occlusion Cuff^®, Belfast, United Kingdom). The cuff was then gradually inflated every 10 s by 10 mmHg starting from 50 mmHg until pulse sounds were no longer detectable. This pressure was defined as 100 % AOP. During the intervention, 60 % AOP was used in the BFR group, as described earlier [43]. The mean AOP was 209 ± 54 mmHg, with a range from 130 to 330 mmHg.

Intervention

The intervention lasted eight weeks and consisted of two RT sessions per week. All participants agreed to refrain from any additional lower-body RT for the duration of the study. Additionally, they were instructed to maintain their normal dietary habits to minimize dietary bias. Training sessions were carried out on non-consecutive days with a minimum of 48 and maximum of 96 h between. All training sessions were supervised by a member of the research team or a qualified therapist with a trainer to participant ratio of 1:1–4. Prior to each training session, a standardized warm-up was completed, consisting of 10 minutes of low intensity cycling on a stationary bike (Cybex 770C Upright Fitness Bike, Cybex International, Inc., Owatonna, Minnesota, USA). Thereafter, two calf exercises were performed in randomized order. The bilateral calf raises with knees extended and unilateral seated calf raises, as described in the testing section. For both exercises, one set of 30 and three sets of 15 repetitions using 30 % of their pre-intervention 1RM with 30 s of inter-set rest were completed as described by Patterson and colleagues [43]. Cadence of every repetition was set as 2 s eccentric and 1 s concentric, with no isometric hold at the top or the bottom. Each repetition was performed with full ROM, either with or without BFR, according to group assignment. In the BFR group, the cuffs were placed and inflated to 60 % AOP before the first set of the first exercise, as described previously, and not loosened or taken off during the inter-set rest periods until the last set of the second exercise was completed.

Statistics

The summary statistics are presented as mean ± standard deviation. Relevant data were initially tested and verified for normal distribution via Shapiro-Wilk tests and variance homogeneity. To examine baseline group differences (BFR vs. noBFR) for age, height, and BW, several 2 (group: BFR vs. noBFR) × 2 (sex: male vs. female) repeated measures analyses of variance (rANOVA) were conducted. Furthermore, separate 2 (group: BFR vs. noBFR) × 2 (time: pre vs. post) × 2 (sex: male vs. female) repeated measures analyses of variance with covariate (rANCOVA) were conducted for each outcome measure: CC, LS, leg press calf raise 1RM, seated calf raise 1RM, plantar flexion 5RM, dorsal extension 5RM, plantar flexion 25RM, and dorsal extension 25RM. Pre-intervention individual values were used as covariates. Effect sizes for rANOVA are provided as partial eta squared (η ² _p), with ≥0.01, ≥0.06, ≥0.14 indicating small, moderate, and large effects, respectively [33]. In case of significant interaction, group, time, or sex effects, relevant Bonferroni post-hoc tests were subsequently computed. Additionally, to examine group differences in total training sessions, an independent t-test was calculated. For pairwise effect size comparison, standardized mean differences (SMD) were calculated as the differences between means divided by the pooled standard deviations (trivial: |SMD|<0.2, small: 0.2≤|SMD|<0.5, moderate: 0.5≤|SMD|<0.8, large: |SMD|≥0.8) [33]. Statistical analyses were performed using R in its integrated development environment RStudio (version 4.1.1; The R Foundation for Statistical Computing). For all calculations, an α-level of 0.05 was used as threshold for statistical significance.

Output parameters

The rANCOVA revealed no statistically significant interaction (0.056≤p≤0.941; 0.01≤η ² _p≤0.06) or group (0.162≤p≤0.996; 0.01≤η ² _p≤0.02) effects for all output parameters (Table 1). While seated calf raise 1RM revealed statistically significant time effects (p=0.046; η ² _p=0.17;), all other output parameters showed no relevant time effects (0.275≤p≤0.900; 0.01≤η ² _p≤0.06) regarding the rANCOVA results. However, leg press calf raise 1RM (Figure 3), seated calf raise 1RM (Figure 3), plantar flexion 5RM (Figure 4), dorsal extension 5RM (Figure 4) and plantar flexion 25RM (Figure 5) revealed moderate to large pairwise pre to post comparison effect sizes (0.35≤SMD≤1.11) for both BFR and noBFR. Except for CC (Figure 2), dorsal extension 25RM (Figure 5) and LS (0.077≤p≤0.120; 0.19≤η ² _p≤0.57; Figure 2), all other output parameters revealed statistically significant sex effects (0.005≤p≤0.024; 0.20≤η ² _p≤0.68). Sub sequent performed post hoc tests revealed statical significant higher CC (Figure 2), LS (Figure 2), leg press calf raise 1RM (Figure 3), seated calf raise 1RM (Figure 3), plantar flexion 5RM (Figure 4), dorsal extensions 5RM (Figure 4), plantar flexion 25RM (Figure 5), and dorsal extensions 25RM (Figure 5) values (p≤0.001; SMD≥0.77) for men compared to female participants.

Figure 2:

Pre (circles) to post (triangles) data of both intervention groups: blood flow restriction (BFR) and conventional resistance training without BFR (noBFR). Calf circumference (left) and leg stiffness (right) are displayed. Individual male (blue) and female (orange) data are presented. Mean with standard derivation is given in black.

Figure 3:

Pre (circles) to post (triangles) data of both intervention groups: blood flow restriction (BFR) and conventional resistance training without BFR (noBFR). Leg press calf raise one-repetition maximum (1RM) (left) and seated calf raise 1RM (right) are displayed. Individual male (blue) and female (orange) data are presented. Mean with standard derivation is given in black.

Figure 4:

Pre (circles) to post (triangles) data of both intervention groups: blood flow restriction (BFR) and conventional resistance training without BFR (noBFR). Plantar flexion five-repetition maximum (5RM) (left) and dorsal extension 5RM (right) are displayed. Individual male (blue) and female (orange) data are presented. Mean with standard derivation is given in black.

Figure 5:

Pre (circles) to post (triangles) data of both intervention groups: blood flow restriction (BFR) and conventional resistance training without BFR (noBFR). Plantar flexion twenty-five-repetition maximum (25RM) (left) and dorsal extension 25RM (right) are displayed. Individual male (blue) and female (orange) data are presented. Mean with standard derivation is given in black.

Participants data

Regarding the participant data, the 2 × 2 rANOVA revealed no statistical significant interaction (0.256≤p≤0.935; 0.01≤η ² _p≤0.04) and group (0.633≤p≤0.919; η ² _p≤0.01) effects for age, height and body mass. In contrast, age, height and body mass showed statistical relevant sex effects (0.001≤p≤0.006; 0.22≤η ² _p≤0.29). Subsequent performed post hoc tests revealed statical significant higher height and body mass values (0.001≤p≤0.006; 0.22≤η ² _p≤0.29) for male compared to female participants. In addition, age revealed no statistical relevant sex effects (p=0.088; η ² _p=0.09). Furthermore, post hoc testing revealed no significant group differences (0.56≤p≤0.754; 0.11≤SMD≤0.28) for age, height and body mass between BFR and noBFR. Apart from this, total number of sessions were not statistical different (p=0.99; MD=0.1 ± 1.4; SMD=0.07) between BFR (14.7 ± 1.6) and noBFR (14.8 ± 1.3).

Strengths and limitations

We also have to acknowledge some methodological limitations of our approach. Firstly, there was no time matched non-exercise control group. Moreover, the number of women and men in this study was too small to adequately investigate any differences in adaptations between the sexes. This should be investigated in future studies with larger numbers of participants from both sexes. Secondly, we determined CC using measuring tape. At the area of measurement, there is an overlap of gastrocnemius and soleus muscles, making it impossible to separately determine muscle growth between the two muscles. Additionally, we have no information on whether type I or type II muscle fibers were affected by any adaptive changes in response to our RT intervention, as we were technically unable to collect muscle biopsies. It has been shown that the molecular signature of low-intensity BFR within skeletal muscle differs from high-intensity RT without BFR. Due to technical limitations concerning weight loading on the seated calf raise machine, the Epley formula used to estimate the 1RM might have introduced a deviation from the actual 1RM. However, because the distribution of such an error was equal in both groups, we assume that it has not influenced our results, specifically not for the between-groups comparison. Finally, it should be noted that our total training volume, considered one primary driver for hypertrophy [65], was likely adequate [30], although it was distributed across two exercises. However, given the specificity of strength adaptations [66], the question remains whether significant increases in strength would occur with the same volume concentrated in a single exercise. This is a topic that could be explored in future studies.

On the other hand, this study is strengthened by its individual determination of AOP following a gold standard protocol. Additionally, the adequate sample size enabled robust statistical analysis, ensuring strong statistical power. Besides, a variety of different strength tests (1RMs, isokinetic testing) were conducted, allowing for valid conclusions from the study. Therefore, the presented data contribute to identifying a highly effective training regimen for inducing calf muscle growth and strength.

Conclusion and practical application

In conclusion, our findings demonstrate that low-load BFR does not offer superior benefits in terms of muscle hypertrophy and strength adaptations in young healthy adults compared to low-load RT when both are conducted at 30 % of 1 RM without muscle failure. Both training modalities resulted in significant strength improvements in seated calf raise, but neither produced notable increases in muscle circumference. This suggests that for individuals seeking strength enhancements, traditional low-load RT is a viable option sufficient to increase strength with a moderate volume of 150 contractions per week, without the need for additional BFR techniques. From a practical standpoint, these findings suggest, that low-load RT provides a safe and effective method to improve strength without the risks associated with higher-intensity training or the complexities of BFR. This may be especially important for older adults who need to maintain functional capacity and independence, as regular strength training can counteract age-related muscle loss and improve mobility. However, this should be investigated in future studies. Additionally, fitness professionals can incorporate low-load RT into training programs to provide clients with a method that balances efficacy and safety, promoting long-term muscle health and overall well-being. Thus, incorporating low-load RT into regular exercise regimens can support muscle maintenance and strength improvement across diverse populations, ensuring broad applicability and accessibility.