Effect of Bang® Pre-Workout Master Blaster® combined with four weeks of resistance training on lean body mass, maximal strength, mircoRNA expression, and serum IGF-1 in men: a randomized, double-blind, placebo-controlled trial

The study employed a block-randomized, double-blind, placebo-controlled, parallel design (Fig. 1). Participants completed an entry session during which the requirements of the study were explained, informed consent was obtained, and testing exercises were familiarized. Participants completed two testing sessions (Pre and Post) in the morning separated by 4 weeks of resistance exercise combined with supplementation of BMB or placebo (PLA). Participants were instructed to complete a 3-day diet recall, fast for at least 10 h, and refrain from exercise for at least 48 h prior to each testing session. On the morning of the testing session, participants reported to the human performance laboratory where height and body mass measurements were obtained. Participants then rested for 5 min while seated in a chair after which hemodynamics were measured. After hemodynamic measurements, the participants completed a body composition assessment using dual-energy x-ray absorptiometry (DXA). After the DXA scan, a venous blood sample and skeletal muscle biopsy sample were obtained. Lastly, the participants completed a maximal strength assessment of the lower- and upper-body via measurement of squat and bench press one-repetition maximum (1-RM), respectively. Participants were block-randomized to BMB or PLA based on resistance training status and maximal squat strength. Post-testing sessions were identical to the pre-testing sessions and were performed at approximately the same time of day as the pre-testing session for each participant.

Sixteen recreationally-active men completed the study (BMB group: n = 8, age = 22.5 ± 2.9 years; height = 181.7 ± 9.2 cm; PLA group: n = 8, age = 22.5 ± 3.1 years; height = 175.3 ± 8.1 cm). Each group had a large, but similar, variance of resistance training experience. The average self-reported resistance training experience was 3.19 ± 2.96 years with a range of less than 1 year of experience (n = 3) to 8 years of experience for the PLA group and 2.94 ± 2.44 years with a range of less than 1 year of experience (also n = 3) to 7 years of experience. Participants did not consume dietary supplements (except multivitamins/multiminerals, caffeine, and/or protein powder) for at least 1 month prior to entering the study. Participants completed a health history questionnaire and a physical activity questionnaire prior to completing the study to assess health status and exercise training experience. Exclusion criteria included a history of or current health condition including diabetes, cardiovascular disease, arrhythmias, thyroid disease, hypogonadism, pulmonary disease, liver or kidney disease, musculoskeletal disorders, neuromuscular or neurological diseases, autoimmune disease, cancer, peptic ulcers, or anemia. Participants were familiarized to the study protocol via a verbal and written explanation outlining the study design and signed an informed consent document approved by the University of South Alabama Institutional Review Board (IRBNet #: 966357; Approval Date: 10/11/2016). All experimental procedures involved in the study conformed to the ethical consideration of the Declaration of Helsinki.

The resistance training program was initiated 2 to 3 days after the pre-testing session. Participants completed a four-week periodized resistance training program consisting of two lower-body and two upper-body sessions per week for a total of 16 sessions. Each resistance exercise session was supervised by study personnel and consisted of seven exercises with 60 to 120 s rest between sets. The resistance training protocol is outlined in Table 1.

Heart rate and blood pressure were determined in the seated position after resting for 10 min. Heart rate was measured by palpation of the radial artery for 30 s. Blood pressure was assessed with a mercurial sphygmomanometer and stethoscope (Welch Allyn, Skaneateles Falls, NY) using standard procedures.

Total body mass (kg) and height (cm) were determined using a calibrated scale and stadiometer (Seca model 700, Seca Corporation, Chino, CA). Body composition was measured by DXA (Horizon Wi, Hologic, Bedford, MA, USA).

Assessment of maximal strength was determined using a 1-RM test for the squat exercise followed by the bench press exercise at both the pre- and post-testing sessions. The procedures for obtaining the 1-RM measurement were the same for both exercises. Participants warmed-up by cycling on an Airdyne bicycle (Schwinn, Vancouver, WA) for 5 min at a self-determined pace followed by completion of 8 to 10 repetitions at approximately 50% of estimated 1-RM. The participant rested for approximately 2 minutes and then completed 3–5 repetitions at approximately 70% of estimated 1-RM. The weight was then increased conservatively and the participant attempted to lift the weight for one repetition. If the lift was successful, the participant rested for 2 minutes before testing the next weight increment. This procedure continued until the participant failed to complete the lift successfully. The 1-RM was recorded as the maximum weight that the participant was able to lift for one repetition.

The squat exercise was performed using a Smith machine (Maxicam, Muscle Dynamics, Paramount, CA) to help standardize form. In addition, squats were performed down to a squat box (Elitefts™, London, OH) to standardize squat depth to 90 degrees of knee flexion for all participants. For the squat to be considered successful, participants were required to squat down until lightly touching the box before beginning the concentric portion of the lift. The bench press exercise was performed in a power rack using an adjustable bench (Hammer Strength, Life Fitness, Rosemont, IL). Participants were required to touch the chest with the barbell before performing the concentric portion of the lift in order to be considered successful.

Venous blood from the antecubital vein was collected at rest using a Vacutainer apparatus and needle (Becton, Dickinson and Company, Franklin lakes, NJ). Blood samples used for complete blood count (CBC) analysis were collected in EDTA tubes and inverted to prevent clotting. Blood samples used for comprehensive metabolic panel (CMP) and IGF-1analysis were collected using serum separator tubes, allowed to stand at room temperature for 10 min, and then centrifuged. CBC and CMP analyses were outsourced to LabCorp Inc., Birmingham, AL. Serum used for the IGF-1 assay was removed and aliquoted into 1.5 mL tubes and immediately frozen at − 80 °C for later analysis.

Percutaneous muscle biopsies (~ 30 mg) were obtained at rest from the middle portion of the vastus lateralis muscle at the midpoint between the patella and the greater trochanter of the femur at a depth between 1 and 2 cm based on previously-used procedures [11]. The same leg and general location (determined by pre-biopsy markings) was biopsied at each testing session. The biopsy area was shaved clean of leg hair and cleaned with rubbing alcohol. A small area of the cleaned skin ~ 2 cm in diameter was anesthetized with a 1.5 mL subcutaneous injection of 1% lidocaine hydrochloride (Hospira, Lake Forest, IL). After, the biopsy site was further cleansed by swabbing the area with povidine-iodine. Once anesthetized, a pilot hole was created using a sterile 12-gauge needle followed by insertion of a 14-gauge fine needle aspiration biopsy instrument (Pro-Mag Ultra Automatic Biopsy Instrument, Argon Medical, Gainesville, FL) was inserted into the skin at an approximate depth of 1 cm to extract the muscle sample using three passes. After removal, adipose tissue was trimmed from the muscle specimens. Specimens were immediately immersed in 500 μL of RNAlater stabilization solution (Life Technologies, Carlsbad, CA) and stored at − 80 °C for later analysis.

Serum samples were analyzed in duplicate for IGF-1 (ALPCO, Salem, NH) using enzyme-linked immunosorbent assay (ELISA) following the manufacturer’s supplied protocol and absorbances were measured at a wavelength of 450 nm using a microplate reader (SpectraMax Plus 384, Molecular Devices, Sunnyvale, CA). Concentrations of the unknown samples were calculated using data reduction software (SoftMax Pro, Molecular Devices, Sunnyvale, CA). Serum IGF-1 assays were performed using a 1:21 sample dilution with an intra-assay coefficient of variance of 7.6%.

Total RNA was isolated from muscle samples using the mirVana PARIS kit according to manufacturer’s specifications (Life Technologies, Carlsbad, CA) as previously described [12]. cDNA synthesis and real-time polymerase chain reaction (RT-PCR) were performed using the qScript® microRNA cDNA Synthesis Kit (QuantaBio, Beverly, MA) and PerfeCTa® SYBR® Green SuperMix (QuantaBio, Beverly, MA). Primers for miRs (miR-15a-5p, miR-23a-5p, miR-23b-5p, miR-126-3p, miR-16-5p, miR-361-5p, miR-320a, miR-186-5p; Additional file 1: Table S1) were commercially synthesized (Integrated DNA Technologies, Coralville, IA). Reactions totaling 25 μL consisting of 5 μL of miRNA cDNA template, 12.5 μL of PerfeCta SYBR Green SuperMix (Quantabio, Beverly, MA), 0.5 μL of the PerfeCTa Universal PCR Primer, 0.5 μL of the target miRNA primer, and 6.5 μL of nuclease-free water were added to each well. Each reaction was amplified using RT-PCR on a qTower 2.2 (Analytik Jena US LLC, Beverly, MA). The amplification profile was run for an initial pre-incubation/activation phase at 95 °C for 2 min and then for 40 cycles of 95 °C for 5 s and 60 °C for 30 s according to manufacturer specifications (QuantaBio, Beverly, MA). Fluorescence was measured after each cycle. Relative miR expression was determined by the 2^-ΔΔCt method using the geometric mean of three miRNAs (miR-361-5p, miR-320a, miR-186-5p) as a reference [5, 13, 14]. Data were expressed with post-testing levels normalized to pre-testing levels for each group. Intra-assay coefficients of variance for miR-186, − 320, − 361, − 15, − 16, −23a, −23b, and − 126 were 0.51, 0.82, 0.94, 0.79, 0.67, 0.95, 0.56, and 0.86%, respectively.

Dietary intake data for (24-h recalls) were collected and analyzed using the Automated Self-Administered 24-h (ASA24) Dietary Assessment Tool, version 2016, developed by the National Cancer Institute, Bethesda, MD [15]. The participants’ diets were not standardized, but participants were instructed not to change their dietary habits during the course of the study. A 3-day diet recall was completed by the participants before each testing session.

Data for each group at each time point were checked for normality of distribution using the Shapiro-Wilk test. Of the 46 variables analyzed statistically, 11 had at least one dataset of each group at either time point not normally distributed according to the Shapiro-Wilk test (mean cell hemoglobin, monocyte count, eosinophil count, basophil count, glucose, potassium, bilirubin, aspartate aminotransferase, alanine aminotransferase, miR-15, and miR-23a). Data for these variables were first analyzed non-parametrically and resulted in similar outcomes to the parametric tests employed; thus, results of the parametric tests are presented. A separate general linear model was utilized for analysis of each variable to determine the effect of each supplement (between-factor) over time (within-factor) on hemodynamics, body composition, maximal strength, serum IGF-1, skeletal muscle miRNA expression, blood safety markers, and dietary intake. Effect sizes for interaction effects were calculated as partial eta-squared (ƞ²). If no significant interaction was observed, main effects were analyzed using paired samples t test for time comparisons and independent samples t test for group comparisons. If a significant interaction was observed, simple main effects were analyzed using paired samples t test for time comparisons for each group and independent samples t test for group comparisons at each time point. Effect sizes for main effects and simple main effects were calculated as Cohen’s d using Excel (Microsoft Corp., Redmond, WA). Statistical analyses were performed using SPSS Statistics 22.0 (IBM Corp.; Armonk, NY) and an a priori probability level of ≤0.05 was adopted.

No significant interaction effects were observed for kilocalorie (p = 0.98; partial n² <  0.01), protein (p = 0.57; partial n² = 0.02), fat (p = 0.60; partial n² = 0.02), or carbohydrate (p = 0.47; partial n² = 0.04) intake (Table 2). No significant differences for the main effect of time were observed for kilocalorie (p = 0.87; Cohen’s d = 0.05), protein (p = 0.82; Cohen’s d = 0.07), fat (p = 0.38; Cohen’s d = 0.25), or carbohydrate (p = 0.58; Cohen’s d = 0.16) intake. No significant differences for the main effect of group were observed for kilocalorie (p = 0.61; Cohen’s d = 0.18), protein (p = 0.29; Cohen’s d = 0.37), fat (p = 0.96; Cohen’s d = 0.03), or carbohydrate (p = 0.99; Cohen’s d < 0.01) intake.

No significant interaction effects were observed for heart rate (p = 0.77; partial n² = 0.03), systolic blood pressure (p = 0.59; partial n² = 0.02), or diastolic blood pressure (p = 0.17; partial n² = 0.13; Fig. 3a-c). No significant differences for the main effect of time were observed for heart rate (p = 0.54; Cohen’s d = 0.11) or diastolic blood pressure (p = 0.34; Cohen’s d = 0.25). A significant decrease in systolic blood pressure was observed for the main effect of time (p = 0.05; Cohen’s d = 0.37). No significant differences for the main effect of group were observed for systolic blood pressure (p = 0.23; Cohen’s d = 0.43). A significant difference for the main effect of group was observed for heart rate (p = 0.01; Cohen’s d = 0.95) and diastolic blood pressure (p = 0.02; Cohen’s d = 0.90) with both significantly higher for the BMB group.

A significant interaction between group and time was observed for total body mass (TBM; p < 0.01; partial n² = 0.56). A significant increase in TBM was observed over time for the BMB group (+ 3.19 kg, 95% CI, 1.98 kg, 4.40 kg, p < 0.001; Cohen’s d = 0.24), but not the PLA group (+ 0.44 kg, 95% CI, − 0.50 kg, 1.39 kg, p = 0.30; Cohen’s d = 0.02). No difference between groups was observed for TBM at the pre-testing (p = 0.39; Cohen’s d = 0.44) or post-testing (p = 0.56; Cohen’s d = 0.30) time points (Fig. 4a).

No significant interaction effect was observed for fat mass (p = 0.39; partial n² = 0.05) or body fat % (p = 0.99; partial n² < 0.01). The main effect of time was not significant for fat mass (p = 0.64; Cohen’s d = 0.02) or body fat % (p = 0.11 Cohen’s d = 0.11). Likewise, the main of effect of group was not significant for fat mass (p = 0.39; Cohen’s d = 0.46) or body fat % (p = 0.36; Cohen’s d = 0.49; Fig. 4b and c).

A significant interaction between group and time was observed for LBM (p < 0.01; partial n² = 0.41). A significant increase in LBM was observed over time for the BMB group (+ 3.15 kg, 95% CI, 1.80 kg, 4.49 kg, p < 0.01; Cohen’s d = 0.54), but not PLA (+ 0.89 kg, 95% CI, − 0.14 kg, 1.93 kg, p = 0.08; Cohen’s d = 0.08). No difference between groups was observed for LBM at the pre-testing (p = 0.50; Cohen’s d = 0.35) or post-testing (p = 0.86; Cohen’s d = 0.09) time points (Fig. 4d).

A significant interaction between group and time was observed (p = 0.02; partial n² = 0.32) for combined strength (squat + bench 1-RM). A significant increase in combined strength was observed over time for the BMB group (+ 34.38 kg, 95% CI, 21.75 kg, 47.00 kg, p < 0.01; Cohen’s d = 0.68) and the PLA group (+ 18.75 kg, 95% CI, 11.88 kg, 25.62 kg, p < 0.01; Cohen’s d = 0.33). No difference between groups was observed for combined strength at the pre-testing (p = 0.51; Cohen’s d = 0.34) or post-testing (p = 0.22; Cohen’s d = 0.64) time points (Fig. 5a).

Individually, a significant interaction between group and time was observed for squat 1-RM (p = 0.04; partial n² = 0.27). A significant increase in squat 1-RM was observed over time for the BMB group (+ 23.86 kg, 95% CI, 16.75 kg, 30.97 kg, p < 0.01; Cohen’s d = 0.78) and the PLA group (+ 14.20 kg, 95% CI, 7.04 kg, 21.37 kg, p < 0.01; Cohen’s d = 0.44). No difference between groups was observed for squat 1-RM at the pre-testing (p = 0.37; Cohen’s d = 0.46) or post-testing (p = 0.13; Cohen’s d = 0.80) time points (Fig. 5b). No significant interaction between group and time was observed for bench press 1-RM (p = 0.08; partial n² = 0.20). A significant increase was observed for the main effect of time (p < 0.01; Cohen’s d = 0.31), with no significant difference observed for the main effect of group (p = 0.45; Cohen’s d = 0.27; Fig. 5c).

A significant interaction between group and time was observed for white blood cell count (p = 0.04; partial n² = 0.28), platelet count (p < 0.01; partial n² = 0.42), lymphocyte count (p < 0.01; partial n² = 0.47), creatinine (p < 0.01; partial n² = 0.48), and calcium (p = 0.03; partial n² = 0.31). White blood cell count (p = 0.04; Cohen’s d = 0.63), platelet count (p = 0.05; Cohen’s d = 0.25), and lymphocyte count (p = 0.01; Cohen’s d = 0.40) decreased in the PLA group over time. No significant effect of time was observed for PLA for creatinine (p = 0.96; Cohen’s d = 0.01) or calcium (p = 0.23; Cohen’s d = 0.64). Lymphocyte count (p = 0.05; Cohen’s d = 0.70) and creatinine (p < 0.01; Cohen’s d = 0.96) increased over time in the BMB group. No significance for time was observed in the BMB group for white blood cell count (p = 0.27; Cohen’s d = 0.60), platelet count (p = 0.06; Cohen’s d = 0.32), or calcium (p = 0.07; Cohen’s d = 0.54). At the pre-testing time point, lymphocyte count (p = 0.05; Cohen’s d = 1.07) was significantly higher for the PLA group, with no significant difference between groups for white blood cell count (p = 0.38; Cohen’s d = 0.44), platelet count (p = 0.74; Cohen’s d = 0.17), creatinine (p = 0.07; Cohen’s d = 0.98), or calcium (p = 0.82; Cohen’s d = 0.09). At the post-testing time point, serum creatinine was significantly higher in the BMB group (p < 0.01; Cohen’s d = 1.64); whereas, calcium was significantly higher in the PLA group (p = 0.02; Cohen’s d = 1.35). No significant difference between groups was observed for white blood cell count (p = 0.13; Cohen’s d = 0.81), platelet count (p = 0.16; Cohen’s d = 0.74), or lymphocyte count (p = 0.83; Cohen’s d = 0.11) at the post-testing time point,

No significant interaction between group and time was observed for red blood cell count (p = 0.18; partial n² = 0.12), hemoglobin (p = 0.41; partial n² = 0.05), hematocrit (p = 0.65; partial n² = 0.02), mean corpuscular volume (p = 0.36; partial n² = 0.06), mean cell hemoglobin (p = 0.19; partial n² = 0.12), mean corpuscular hemoglobin concentration (p = 0.84; partial n² < 0.01), neutrophil count (p = 0.48; partial n² = 0.04), monocyte count (p = 0.14; partial n² = 0.15), eosinophil count (p = 0.12; partial n² = 0.16), basophil count (p = 0.33; partial n² = 0.07), glucose (p = 0.40; partial n² = 0.05), blood urea nitrogen (p = 0.15; partial n² = 0.14), sodium (p = 0.46; partial n² = 0.04), potassium (p = 0.24; partial n² = 0.10), chloride (p = 0.42; partial n² = 0.05), carbon dioxide (p = 0.75; partial n² = 0.01), protein (p = 0.80; partial n² = 0.01), albumin (p = 0.83; partial n² < 0.01), globulin (p = 0.61; partial n² = 0.02), albumin/globulin ratio (p = 0.56; partial n² = 0.03), bilirubin (p = 0.28; partial n² = 0.08), alkaline phosphatase (p = 0.25; partial n² = 0.09), aspartate aminotransferase (p = 0.41; partial n² = 0.05), or alanine aminotransferase (p = 0.46; partial n² = 0.04). No significance was observed for the main effect of time for red blood cell count (p = 0.63; Cohen’s d = 0.06), hemoglobin (p = 0.99; Cohen’s d < 0.01), hematocrit (p = 0.37; Cohen’s d = 0.15), mean corpuscular volume (p = 0.11; Cohen’s d = 0.26), mean cell hemoglobin (p = 0.85; Cohen’s d = 0.02), mean corpuscular hemoglobin concentration (p = 0.27; Cohen’s d = 0.30), neutrophil count (p = 0.38; Cohen’s d = 0.23), monocyte count (p = 0.38; Cohen’s d = 0.22), eosinophil count (p = 0.06; Cohen’s d = 0.44), basophil count (p = 0.33; Cohen’s d = 0.19), blood urea nitrogen (p = 0.73; Cohen’s d = 0.07), sodium (p = 0.09; Cohen’s d = 0.51), potassium (p = 0.29; Cohen’s d = 0.40), chloride (p = 0.41; Cohen’s d = 0.26), carbon dioxide (p = 0.11; Cohen’s d = 0.67), globulin (p = 0.13; Cohen’s d = 0.52), albumin/globulin ratio (p = 0.33; Cohen’s d = 0.23), bilirubin (p = 0.95; Cohen’s d = 0.02), alkaline phosphatase (p = 0.49; Cohen’s d = 0.05), aspartate aminotransferase (p = 0.44; Cohen’s d = 0.25), or alanine aminotransferase (p = 0.48; Cohen’s d = 0.20). Likewise, no significance was observed for the main effect of group for red blood cell count (p = 0.09; Cohen’s d = 0.63), hemoglobin (p = 0.18; Cohen’s d = 0.49), hematocrit (p = 0.13; Cohen’s d = 0.55), mean corpuscular volume (p = 0.75; Cohen’s d = 0.11), mean cell hemoglobin (p = 0.46; Cohen’s d = 0.27), mean corpuscular hemoglobin concentration (p = 0.67; Cohen’s d = 0.15), neutrophil count (p = 0.16; Cohen’s d = 0.51), monocyte count (p = 0.32; Cohen’s d = 0.36), eosinophil count (p = 0.07; Cohen’s d = 0.68), basophil count (p = 0.16; Cohen’s d = 0.51), glucose (p = 0.47; Cohen’s d = 0.26), blood urea nitrogen (p = 0.09; Cohen’s d = 0.63), sodium (p = 0.12; Cohen’s d = 0.57), potassium (p = 0.54; Cohen’s d = 0.22), chloride (p = 0.57; Cohen’s d = 0.20), carbon dioxide (p = 0.43; Cohen’s d = 0.28), protein (p = 0.85; Cohen’s d = 0.07), albumin (p = 0.61; Cohen’s d = 0.18), globulin (p = 0.64; Cohen’s d = 0.17), albumin/globulin ratio (p = 0.60; Cohen’s d = 0.19), alkaline phosphatase (p = 0.31; Cohen’s d = 0.36), aspartate aminotransferase (p = 0.49; Cohen’s d = 0.25), or alanine aminotransferase (p = 0.51; Cohen’s d = 0.24). A significant main effect for time was observed for glucose (p = 0.01; Cohen’s d = 0.72) protein (p = 0.02; Cohen’s d = 0.71), and albumin (p = 0.03; Cohen’s d = 0.41). Glucose and albumin were significantly increased at the post-testing time point compared with pre-testing; whereas, protein decreased from pre-to-post-testing. A significant main effect for group was observed for bilirubin (p = 0.04; Cohen’s d = 0.79) with the PLA group significantly higher compared with the BMB group. Although some statistical changes were observed, all mean values were still within the normal clinical reference range (Table 3).

Data presented as mean ± standard deviation. ^aDenotes significant increase from PRE to POST for main effect of time. ^bDenotes significant decrease from PRE to POST for main effect of time. ^†Denotes significant group x time interaction. *Denotes significant difference (p ≤ 0.05) from PRE to POST for group. **Denotes significant difference (p < 0.01) from PRE to POST for group. ^#Denotes significant difference (p ≤ 0.05) between groups at time point. ^##Denotes significant difference (p ≤ 0.01) between groups at time point.

No significant interaction between group and time was observed for serum IGF-1 (p = 0.34; partial n² = 0.44; Fig. 6). No significant main effect of time (p = 0.95; Cohen’s d = 0.01) or group (p = 0.77; Cohen’s d = 0.15) was observed.

No significant interaction effects between group and time were observed for miR-15 (p = 0.72; partial n² = 0.01), miR-16 (p = 0.55; partial n² = 0.03), miR-23a (p = 0.98; partial n² < 0.01), miR-23b (p = 0.57; partial n² = 0.03), or miR-126 (p = 0.71; partial n² = 0.01) expression.. A significant main effect for time was observed for miR-23a (p = 0.01; Cohen’s d = 1.04) and miR-23b (p = 0.05; Cohen’s d = 0.70) expression with both significantly increased at Post compared with Pre. No significant main effect of time was observed for miR-15 (p = 0.24; Cohen’s d = 0.40), miR-16 (p = 0.21; Cohen’s d = 0.39), or miR-126 (p = 0.33; Cohen’s d = 0.36). The main of effect of group was not significant for miR-15 (p = 0.64; Cohen’s d = 0.17), miR-16 (p = 0.16; Cohen’s d = 0.51), miR-23a (p = 0.67; Cohen’s d = 0.16), miR-23b (p = 0.21; Cohen’s d = 0.47), or miR-126 (p = 0.39; Cohen’s d = 0.34; Fig. 7).

This study was limited by the short duration of resistance training; therefore, the current results cannot be extrapolated to longer periods of resistance training, i.e. 6 months to multiple years, after which adaptations may be more or less robust compared with placebo. The study is also limited by the inherent inaccuracies associated with dietary recalls [40]. The participants were asked to not change their dietary habits and to report all food intake for 3 days prior to each testing session. Although no differences were observed between groups or over time for macronutrient or kilocalorie intake, it is possible that dietary intakes were not reported accurately which could result in missed effects resulting from dietary intake. Furthermore, we could not mask the stimulant effects of caffeine in the BMB versus the PLA supplement.

Hemodynamic measurements were assessed at rest, which does not account for any potential alterations in heart rate or blood pressure experienced during exercise. Furthermore, similar to the hemodynamic measurements, blood and muscle samples were collected at rest. Consequently, only differences in basal levels of serum IGF-1 and miRs were studied. Changes in acute skeletal muscle miR expression in response to resistance exercise as a result of BMB supplementation may exist, as previously observed with serum IGF-1 [10], but they would be unable to be detected with the design of the current study. Lastly, the study is limited by a relatively small sample size. While the sample size of the current study was large enough to detect significant interaction effects regarding LBM and maximal strength, a larger sample size would give a better representation of the true change to be expected in the study population as individual responses to resistance training and supplementation present with wide variability [41].