An a priori power analysis was conducted using G*Power 3.1.9.7 software to estimate the sample size for a repeated-measures analysis of variance (RM-ANOVA, within-subject design) (Faul et al., 2007). With an alpha level of 0.05, a power (1-β) of 0.80, and an expected medium effect size (partial ηp² = 0.06, corresponding to Cohen’s f ≥ 0.25), the minimum required sample size was determined to be 12 participants. Consequently, 12 healthy male university students (age 21.3 ± 1.6 years) were recruited for this study.

All participants provided written informed consent before the commencement of the study. The study protocol was approved by the Institutional Review Board (IRB) of Kangwon National University (IRB No.: KWNUIRB-2025-07-002-001). Inclusion criteria required participants to be healthy males aged 18–30 years with no known cardiovascular, metabolic, or neuromuscular diseases, and the ability to perform low-load resistance training. Exclusion criteria included a history of diagnosed thrombosis, abnormal electrocardiogram results after exercise, resting blood pressure exceeding 160/100 mmHg, and a difference in bilateral thigh circumference greater than 0.5 cm.

The study employed a randomized crossover, repeated-measures design. To eliminate order effects, participants were randomly assigned to one of four pressure sequences based on a balanced Latin square (Williams design). The procedure consisted of a familiarization session (Visit 0) and four formal experimental visits (Visits 1–4), with at least 7 days between sessions. All visits were conducted at the same time of day for each participant.

Data are presented as mean ± standard deviation (SD). Normality was assessed using the Shapiro-Wilk test, and sphericity was tested using Mauchly’s test (with Greenhouse-Geisser correction applied where necessary). A one-way RM-ANOVA was used for %EMGmax and RPE, while a two-way (Time × Pressure) RM-ANOVA was used for BLa and blood markers. Significant differences were further analyzed using Bonferroni post hoc tests. The significance level was set at p < 0.05, and effect sizes were reported as partial eta squared ηp².

The repeated-measures ANOVA revealed a significant main effect of occlusion pressure on the activation levels of all analyzed muscles (all p < 0.001, ηp² = 0.58–0.76). Overall, %EMGmax increased with rising pressure, reaching peak levels at 70% AOP (52.57%–54.57% EMGmax). Neuromuscular activation outcomes across conditions are summarized in Table 2.

Summary results for serum biomarkers are presented in Table 5 (descriptive statistics and RM-ANOVA p-values for Time, Pressure, and Time × Pressure).

Interleukin-6 (IL-6). Two-way repeated-measures ANOVA revealed a significant main effect of time for IL-6 (F = 348.31, p < 0.001, ηp² = 0.97), no significant main effect of pressure (F = 1.37, p = 0.270, ηp² = 0.11), and a significant time × pressure interaction (F = 11.70, p < 0.001, ηp² = 0.52). Post-hoc comparisons indicated that IL-6 after exercise was significantly higher in the 80% AOP condition than in the 0% AOP condition (p < 0.05), suggesting an occlusion-pressure–dependent amplification of the acute inflammatory response.

Testosterone (TST). For TST, there was a significant main effect of time (F = 54.11, p < 0.001, ηp² = 0.83) and a significant time × pressure interaction (F = 4.59, p = 0.009, ηp² = 0.29), whereas the main effect of pressure was not significant (F = 0.10, p = 0.957, ηp² = 0.01). Post-hoc tests showed that TST increased from Pre to Post in the 60%, 70%, and 80% AOP conditions (all p < 0.05), while no significant change was observed in the 0% AOP condition (p = 0.403).

The present findings demonstrate that increasing arterial occlusion pressure (AOP) during low-load isokinetic exercise augments neuromuscular activation and metabolic stress up to a certain threshold. Specifically, compared with 60% AOP, both 70% and 80% AOP elicited greater %EMGmax and blood lactate responses. However, increasing pressure from 70% to 80% AOP did not confer additional neuromuscular or metabolic benefits within the present task configuration, while perceptual strain increased markedly at 80% AOP. This pattern suggests diminishing returns beyond ∼70% AOP rather than a loss of efficacy at higher pressures. Thus, 70% AOP appears to provide a similar physiological stimulus (i.e., no additional gains relative to 80% AOP) with a lower perceptual cost under the specific velocity, load, and time frame examined in this study.

The results of this study elucidate how acute physiological responses during low-load resistance exercise are modulated by the magnitude of arterial occlusion. Overall, compared to 60% AOP, both 70% and 80% AOP conditions induced significantly higher muscle activation (%EMGmax) and blood lactate (BLa) levels. However, differences between 70% and 80% were small for %EMGmax and early lactate accumulation, although 70% remained higher than 80% at Post-10, suggesting a diminishing marginal return in neuromuscular and metabolic responses once pressure reaches a specific threshold. Conversely, the Rating of Perceived Exertion (RPE) at 80% AOP was significantly higher than at 70% AOP, indicating that extreme occlusion pressures substantially elevate the perceptual burden. This aligns with the current understanding that higher pressures in BFRT may disproportionately reduce exercise tolerability without providing additional physiological benefits (Patterson et al., 2019). In summary, while AOP≥70% provides a robust physiological stimulus, increasing pressure from 70% to 80% yields no further metabolic or electromyographic gains but significantly elevates subjective fatigue, suggesting that 70% AOP serves as a superior “stimulus-tolerance” equilibrium point for healthy adults. Notably, this pattern is consistent with recent squat-based findings using 0%–80% AOP (Zhu et al., 2025), while extending the evidence to an open-chain isokinetic dynamometry model.

The observed increases in %EMGmax with higher AOP should be interpreted cautiously. Surface EMG amplitude reflects the summed electrical activity of active motor units and does not directly quantify neural drive or motor unit firing rates. Therefore, elevated %EMGmax under BFRT conditions likely represents a composite effect of altered recruitment strategies, fatigue-related compensation, and peripheral signal amplification. Importantly, the apparent plateau between 70% and 80% AOP does not indicate a physiological ceiling of neuromuscular activation, but rather reflects the responses within the present isokinetic task constraints.

Consistent with previous evidence, we observed significantly higher %EMGmax under elevated AOP conditions (especially≥70%), reinforcing the efficacy of BFRT in enhancing muscle activation during low-load tasks (Scott and Loenneke, 2022). Mechanistically, BFRT induces local hypoxia and the accumulation of metabolic by-products (e.g., H+, lactate, inorganic phosphate), which collectively augment metabolic stress (Lixandrao and Ugrinowitsch, 2018). Under such conditions, low-threshold Type I muscle fibers fatigue prematurely due to inadequate oxygen supply, necessitating the compensatory recruitment of high-threshold Type II muscle fibers to maintain the target torque output (Loenneke et al., 2014). Furthermore, the accumulation of metabolites stimulates Group III/IV afferent fibers, which enhances central drive and further facilitates high-threshold motor unit recruitment (Scott and Loenneke, 2022). Our findings suggests that 70% AOP may reach the necessary threshold to trigger near-maximal recruitment compensation within the present task constraints in an isokinetic model. These mechanistic pathways are broadly consistent with prior syntheses and experimental work in BFRT (Slysz et al., 2016; Centner et al., 2019; Patterson et al., 2019). However, because within-session set-by-set EMG trajectories were not quantified in the present dataset, we cannot determine whether activation plateaued early versus increased progressively across sets; future studies should report set-level activation patterns to strengthen dose–response interpretations.

Metabolic stress is a hallmark of BFRT, characterized by restricted venous return that shifts energy metabolism toward anaerobic glycolysis (Patterson et al., 2019). In this study, lactate responses at 70% and 80% AOP were significantly higher than those at lower pressures, confirming that these levels provide a potent metabolic stimulus. Interestingly, the lack of a significant difference between 70% and 80% AOP suggests a “plateau trend” in metabolic response: once a sufficient degree of blood flow restriction is achieved, further increases in pressure do not linearly increase anaerobic glycolysis contribution (Patterson et al., 2019). Notably, at Post-10 min, 70% AOP maintained significantly higher BLa than 80% AOP, which may be attributed to complex microvascular dynamics and altered metabolite washout rates under varying high-pressure gradients. The relatively elevated baseline blood lactate levels may reflect dietary or recovery-related factors prior to testing; however, all conditions were compared within subjects under identical procedures, thereby preserving internal validity.

RPE increased significantly with AOP, with 80% AOP being significantly higher than both 60% and 70% AOP. RPE may also have been influenced by cuff-related discomfort/pain, particularly at higher AOP, because participants were not asked to differentiate exertion from pain-related sensations. This sharp increase in perceived exertion likely stems from the intense mechanical compression of the cuff and the amplified feedback from Group III/IV afferents sensitive to severe ischemia and metabolic acidification (Mouser and Mattocks, 2017). Given that long-term adherence to BFRT is highly dependent on subjective tolerability, the high perceptual burden associated with 80% AOP (RPE >16) may hinder protocol feasibility and sustainability (Patterson et al., 2019). Thus, 70% AOP appears to be a more viable prescription for maximizing stimuli while maintaining acceptable tolerance.

The significant Time × Pressure interactions observed for IL-6 and TST suggest that higher occlusion pressures amplify acute systemic responses. The pronounced increase in IL-6 at 80% AOP indicates a greater acute inflammatory or myokine response, potentially linked to a larger metabolic disturbance (Moriggi et al., 2021). Regarding TST, the significant elevation across all BFR conditions (60%–80% AOP) relative to the control condition suggests that the restricted blood flow environment is a primary driver of acute anabolic hormonal responses, likely mediated by lactate-induced signaling (Geng et al., 2021). Given the sensitivity of testosterone to circadian variation, standardized testing times across visits are important when interpreting these acute endocrine responses.

Several limitations should be acknowledged. First, muscle oxygenation was not directly assessed (e.g., via near-infrared spectroscopy), which limits mechanistic inference regarding local ischemia-hypoxia dynamics across pressure levels. Second, participants were not asked to separately rate cuff-related pain/discomfort; therefore, the extent to which occlusion discomfort contributed to RPE cannot be quantified. Third, EMG amplitude was normalized to EMGmax derived from MVC trials; we did not quantify between-visit reliability (e.g., ICC) of MVCMAX/EMGmax, and variability in maximal effort could have influenced normalized EMG outcomes. Fourth, venous blood was sampled at a fixed post-exercise time point (5 min); however, the full time course of IL-6 and testosterone responses was not characterized, and peak responses may occur at different time windows. Fifth, the absence of formal equivalence testing means that non-significant differences between 70% and 80% AOP should not be interpreted as statistical equivalence. Finally, arterial occlusion pressure is posture-dependent; because AOP was determined under standardized measurement conditions, the effective occlusion during the seated isokinetic task may have differed slightly due to hydrostatic effects. These limitations should be considered when generalizing the findings.