The current project is a parallel control experimental study comparing men and women following a 6-week resistance training intervention in conjunction with partial blood-flow restriction. (Clinical Trial #: NCT05615831). Participants underwent baseline testing separated into two visits within a one-week span and presented at the Cardiometabolic Exercise and Lifestyle Laboratory at the University of New Brunswick for each visit. Participants began six weeks of BFRT within one week of the last baseline testing visit. Following the intervention, participants underwent follow-up testing at least two days [23], but no more than one week [10], following the last exercise session. An overview of the study design and timeline can be seen in Fig 1. All exercise sessions were supervised by research staff at the Cardiometabolic Exercise and Lifestyle Laboratory at the University of New Brunswick. Participants were told to wear comfortable clothes. Not every session was necessarily completed at the same time of day for all participants, as such, time of day was recorded for each exercise session and testing visit. All participants provided written and informed consent prior to participation. The project was reviewed and approved by the University of New Brunswick Research Ethics Board (REB 2021-124).

A power calculation was performed using G-power software (version 3.1.9.4, Germany) to determine the appropriate sample size for statistical significance. Based on an alpha of 0.05, a power of 0.8, and an effect size of 0.4, we determined the required total sample size to be eight participants for a repeated measures analysis of variance (ANOVA). However, we anticipate a dropout percentage of 13% as reported by Høgsholt et al. (2022) [24]. Furthermore, it has been suggested that in order to detect interaction effects between sexes, as well as main effects, the sample size needs to be four-times the size [25]. To account for dropout rate, and to ensure adequate ability to detect interaction and main effects, we attempted to recruit approximately 20 per group for a total of 40 participants.

Participants undertook 6-weeks of whole-body resistance training in conjunction with blood flow restriction to their exercising limbs. The intervention consisted of three supervised exercise sessions per week consisting of five different exercises: knee flexion (hamstring curl), knee extension, leg press, chest press, and seated row. As the exercise load was quite low, no warmup procedure occurred prior to beginning the exercise session. The exercise load was individualized to 30% of each participant’s 1-RM for each exercise. Participants were required to complete 75 total repetitions broken into four sets for each exercise. The sets were broken up in the following manner: set 1: 30 repetitions; set 2: 15 repetitions; set 3: 15 repetitions; set 4: 15 repetitions, as this protocol has previously been used in blood flow restriction research to induce muscle hypertrophy in a variety of populations [29,30] and has been suggested by multiple reviews [31,32]. At week 4, participants had their 1-RM reassessed to adjust the 30% 1-RM exercising loads. This occurred during the first exercise session of week 4. As such, following the 1-RM reassessment, participants performed two sets per exercise (30 reps and 15 reps) using the newly adjusted 30% 1-RM weight, before returning to the original rep scheme for their next session.

Blood flow restriction cuffs were placed at the most proximal portion of the exercising limb (just above biceps brachii on the arm and near the inguinal crease on the thigh), which is what has previously been used in BFRT research [8,29–31,33–36]. Blood flow restriction was achieved using the KAATSU C3 device (KAATSU Global, Inc., Huntington Beach, CA, USA). The KAATSU arm and leg cuffs are 5 cm wide, respectively, and are single-bladder cuffs. Cuffs were inflated to 60% of each individual’s total limb occlusion pressure as this has been shown to be a safe and effective pressure to induce muscular adaptations [15], and is within the recommended pressure range for BFRT [8,31]. Each participant’s total limb occlusion pressure (LOP) was estimated using equations developed by Loenneke et al. (2015) listed here [33]:

Although these equations have modest R-squared values of 0.61 and 0.49, respectively [33], they have been successfully utilized in a BFRT intervention of trained athletes leading to significant improvements in muscle hypertrophy and functional outcomes with no reported side effects [37]. Furthermore, although LOP was estimated, a study comparing the impact of various cuff pressures showed similar effects on muscle size, torque, strength, and endurance between 40% and 90% pressures [38]. These findings suggest that greater pressures are not necessarily more effective at inducing adaptations and that small pressure fluctuations won’t necessarily serve as greater or lesser stimuli during training. Cuffs remained inflated during the rest in between sets of each exercise but were deflated for the rest period between exercises [29,39–42]. The set rest was 60 seconds, and the rest between exercises was four minutes.

Lean body mass was estimated using dual-energy x-ray absorptiometry (DXA) prior to the 6-week BFRT intervention, and again following the intervention. Body composition was estimated using a Hologic Horizon^® DXA System (Hologic Canada ULC, Mississauga, ON, Canada). Lean body mass constitutes that which is not fat mass nor bone mineral mass [43]. Participants presented to the laboratory following a 12-hr fast and were asked to refrain from exercise for a 24-hr period prior to testing for consistency, and as dietary intake and exercise could alter lean mass [44]. Participants were instructed to wear loose-fitting clothing with no metal (buckles, zippers, buttons, etc.) and then instructed to lie supine on the scanner’s table and remain still for the duration of the scan. Arms were placed at the participants’ sides with palms facing medially and thumbs pointed upwards. For individuals too large for the width of the table, they were positioned with one arm outside of the scan area and results of the scanned arm were duplicated. The coefficient of variation in our lab for lean mass is 0.6% and for body fat percentage is 0.7%. This was performed on 33 people (males, n = 10) with a mean age of 23.4 years and a mean body mass index (BMI) of 25.6.

Anthropometric measurements, muscular power and endurance, and strength were measured for exploratory purposes and sample description. Participants’ height and weight were measured to the nearest 0.5 cm and 0.1 kg, respectively, according to the CSEP protocol [45]. Weight was measured using a calibrated column scale (SECA^® model #213). Height was measured using a standardized stadiometer. With no shoes, feet together, and arms at their side, height was taken following an inhalation. Hip and waist circumference were measured using an anthropometric tape measure and recorded to the nearest 0.5 cm. For hip and waist measurements, participants stood with their feet shoulder-width apart and their arms folded across their chest. Waist circumference was measured at the upper lateral border of the iliac crest following a normal exhalation, hip circumference was measured around the widest portion of the buttocks after a normal exhalation [46].

Strength was assessed by 1-RM for the five exercises used during the intervention. 1-RM was measured during the second baseline testing visit, at the midpoint of the study during the first exercise session of week four, and again during the second testing visit in the follow-up testing. Each participant’s 1-RM was determined using the following protocol: one set of 6–10 repetitions, followed by one set of 3–5 repetitions, followed by small incremental increases for one repetition until a failure is achieved within seven attempts. If no failure was achieved within seven attempts, the 1-RM for that exercise was redone prior to their first exercise session.

Muscular power was estimated using the squat jump equation derived by Sayers et al. (1999) and is as follows [47]:

This equation was chosen as it has been shown to be more accurate than previously used power estimation equations and was developed from a large and diverse population which enhances our external validity [47]. Jump height was recorded using the Perform Better^® Just Jump System. The Just Jump System has been validated against a 3-camera motion analysis system for estimating vertical jump height in a sample of males and females between the ages of 18–25 [48]. Participants were instructed to stand on the mat with their feet shoulder-width apart, place their hands on their hips, lower into the jump position (knees at approximately a 90° angle), hold for 2 seconds, explode upwards as high as possible, and land back on the mat. Participants performed three squat jumps separated by a 60 second recovery period. The highest jump was used to estimate muscle power.

Dynamic balance was measured using the Y-Balance Test. Briefly, after no more than four practice attempts, participants started by balancing on their left leg and then reached forward as far as they could and touched down. The distance was recorded, and the process was repeated two more times. The same process was then followed when balancing on the right leg. This was performed three times in each direction, alternating between balancing on the left and right feet. All six reaches per direction (left then right) were performed before moving to another direction.

Muscular endurance of the dominant knee extensors and flexors was assessed using a Humac^® NORM isokinetic dynamometer system (Computer Sports Medicine, Inc., Stoughton, MA, USA). Prior to testing, participants performed a 5-minute walking warmup. The participants were seated and secured to the device using straps across the trunk and thighs. The positioning of the seat was adjusted to the comfort level of the participant, so long as the approximate axis of the knee (through the lateral femoral epicondyle) was aligned with the dynamometer’s mechanical axis, and recorded so the same settings were used following the intervention. Range of motion was then prescribed on an individual basis (0° corresponds to full knee extension). Prior to testing, participants performed five repetitions at 120°/s as a familiarization. Upon completion of the familiarization, participants were given a two-minute recovery period before testing commenced. The testing protocol consisted of 30 reciprocal maximal contractions of the knee extensors and flexors performed at 180°/s, as previously described [49]. Total work, mean power per repetition, and peak torque were recorded.

To test for normality within the sample, Shapiro-Wilk test was performed and confirmed with a visual examination of the data. General characteristics of the sample are presented as mean ±  standard deviation (SD) for continuous variables and n (%) for categorical variables, along with their effect sizes. Effect sizes were calculated using Hedges g formula, which is calculated by dividing the difference between the means by the pooled weighted standard deviation. Hedges g was selected as a measure of effect size as it incorporates a correction for small sample sizes to reduce risk of bias and provide a more accurate estimate of effect [50]. Effect sizes were interpreted as follows: Small effect: 0.00 ≤  g ≤  0.50, Medium effect: 0.50 ≤  g ≤  0.80, Large effect: g ≥  0.80. Differences in baseline and post-intervention values, grouped by sex, were analyzed using paired sample t-tests. For baseline and general characteristics, independent t-tests were performed to assess differences between groups. A repeated measures ANOVA was performed to determine whether there were significant main effects for an interaction between time and sex with changes in primary and exploratory outcomes. Data management and statistical analyses were performed using SPSS (IBM Corp. Released 2023. IBM SPSS Statistics for Windows, Version 29.0, Armonk, NY: IBM Corp). A p ≤  0.05 was considered significant.

The total number of participants that completed the study was 38: 19 males and 19 females. An overview of the baseline descriptive characteristics is outlined in Table 1. Briefly, the mean age of males was 24.2 ±  2.76 years, and was 22.7 ±  3.37 years (p = 0.135) for females. Approximately 66% of the total sample were White (males =  68.4%, females =  60%). There were no statistically significant differences at baseline in weight, BMI, waist circumference, or steps per day (all p-values > 0.05). High-density lipoprotein levels were significantly higher in females compared to males (p = 0.002).

Table 2 describes the impact of six-week BFRT on body composition in males and females. Baseline fat mass was not significantly different between males and females (p = 0.303), but baseline lean mass was significantly greater in the male group (p < 0.001). Males significantly increased weight, BMI, and waist circumference (p = 0.006, 0.046, 0.025), whereas females saw no significant changes in these anthropometric measures. A significant interaction was observed for waist circumference (p = 0.022) and lower limb lean mass (p = 0.028). No significant changes were observed in males’ or females’ body fat percentage or fat mass.

Males (p = 0.009, g = 0.11) and females (p = 0.023, g = 0.08) significantly increased lean mass following six-week BFRT (Figs 3A and 3B). Similar results were observed for the impact of BFRT on relative lean mass in males (p = 0.011, g = 0.09) and females (p = 0.020, g = 0.08; Figs 4A and 4B). Repeated measures ANOVA revealed a time effect for lean body mass (p < 0.001) but no interaction between time and sex (p = 0.279, Fig 3C). Similar results were observed for relative lean mass, with a time effect (p < 0.001) and no interaction between time and sex (p = 0.472, Fig 4C).

Table 3 shows changes in performance outcomes, while Figs 5A–C and 6A–C show changes in 1-RM from baseline to post-testing for chest press and knee extension, respectively, for males and females. Exception for females’ leg press (p = 0.110), males and females significantly improved knee extension (male: p < 0.001.; female: p < 0.001), knee flexion (p < 0.001; p < 0.001), chest press (p < 0.001; p = 0.004), and seated row (p = 0.003; p = 0.004) following 6-week BFRT. Repeated measures ANOVA revealed a time effect for leg press (time: p = 0.002), knee extension (p < 0.001), knee flexion (p < 0.001), chest press (p < 0.001), and seated row (p < 0.001), while an interaction effect was only observed for knee extension (p = 0.035), knee flexion (p = 0.043), chest press (p = 0.003), and seated row (p = 0.038).

Both males and females increased peak power produced during the vertical jump (male: p < 0.001; female: p = 0.002), and males increased jump peak power significantly more than females as an interaction between time and sex was observed (p = 0.039). Both males and females significantly improved their mean power, peak torque, and total work for their dominant knee flexors (all p-values < 0.05). Females improved mean power, peak torque, and total work for their dominant knee extensors (all p-values < 0.05). Males improved mean power (p = 0.031) and total work (p = 0.016), but not peak torque (p = 0.055) for their dominant knee extensors. There were no significant differences in the improvements between males and females as indicated by the interactions between time and sex (all p-values > 0.05).

Our work has some limitations that need to be highlighted. First, our intervention had a short duration of only six weeks, and it is possible that some sex-specific differences in outcomes could have occurred over longer durations of BFRT. Second, we did not include a control condition in our study which limited the conclusion drawn from this study. Finally, although we asked participants to maintain their current lifestyle, including diet, we did not control diet every week for the duration of the study. While it has been recommended to include females in exercise science research with 1) pre-defined and standardized inclusion criteria and 2) to adapt the experimental design with consideration of menstrual cycle or hormonal contraceptive factors [65], recent evidence suggests that the impact of the menstrual cycle or hormonal contraceptives is nonexistent [66,67]. Although we did not control for menstrual cycle, current evidence suggests that it is unnecessary, and the emphasis should be to include females in exercise physiology research that is adequately powered to address their underrepresentation and to answer the research question [67]. However, our work is strengthened by the compliance of participants to the study – every participant completed 100% of the exercise sessions. Also, exercise sessions were supervised in very small groups by research staff, allowing for a tightly controlled environment. Furthermore, our work is strengthened by our sufficiently powered sample to directly compare males and females, allowing us to draw insights on potential sex differences in lean mass, muscle strength, and performance outcomes following BFRT.