Introduction

Strength training is a common practice undertaken by both the competitive athlete and fitness enthusiast. Athletes and practitioners utilize strength training in an effort to improve relevant physical skills, such as muscular power, strength or endurance. Strength training for sport includes a variety of unilateral (one limb) and bilateral (two limbs simultaneously) exercises that are employed in a sport specific manner [2]. However, recent research has uncovered that limb dominance can influence muscular involvement in bilateral resistance training exercises and sporting movements. More specifically, when participants perform a bench press exercise, their dominant side muscles achieve a greater level of muscle activation than their nondominant muscles [30]. Similar findings have been uncovered with the barbell squat exercise both with and without the imposition of muscle fatigue [24]. Over time, this asymmetry in muscle activation can lead to a muscular imbalance (due to specificity) and even injury due to overreliance on the dominant limb [12, 37].

Muscle imbalances between dominant and nondominant limbs are common in high-level athletes [37]. Moreover, repetitive tasks due to sport and position specific practice can exacerbate these asymmetries due to overreliance on the dominant limb [12]. These progressively increasing discrepancies between limbs are potentially predicated by the relatively permanent nature of unconsciously stored instructions for movement [29]. Therefore, the very practice that improves performance (training) also serves as a catalyst for greater injury susceptibility [37], and sport specific training could intensify this stimulus [28, 34, 35].

Due to these concerns and similar considerations, numerous strength training techniques and modalities have been brought to the forefront of strength and conditioning which have been proposed to mitigate a shortcoming of traditional strength or sport training. A novel technique, termed “offset loading” or “asymmetrical loading”, may be the next unique training method used by practitioners to potentially reduce asymmetries and improve bilateral performance in their clients or athletes. Offset, or asymmetrical, loading is the practice of offsetting a planned portion of a given training load towards one side of a resistance implement (typically a barbell) [27, 41], thereby creating a “loaded” or “heavier” side and a “de-loaded” or “lighter” side.

Jarosz et al. [27] were the first to demonstrate the acute neuromuscular adaptations to offset loading (OSL) in the bench press. These researchers discovered that OSL (“loaded”/“heavier” on the nondominant side) with as little as 2.5% of the total load led to increased muscle activation in the nondominant muscles. When compared to traditional loading (TDL), this led to a more symmetrical activation of nondominant and dominant muscle groups. Saeterbakken et al. [41] reported similar results by finding that OSL can influence acute muscle activation by increasing activation in the muscle exposed to the heavier offset load. Since muscle activation is often regarded as a measure of neural drive [8], it is clear that OSL in the bench press may be a viable option to modify asymmetrical neural drive patterns between limbs during strength training. Additionally, this unique training method may influence long-term adaptations (i.e., muscle strength and hypertrophy) due to these acute changes in muscle activity and the novelty of the loading pattern. Seminal research from Barnett et al. [10] demonstrated that performing incline bench press causes greater muscle activation of the clavicular head of the pectoralis major. Furthermore, recent research has reported that subjects performing incline bench press exhibited significantly greater changes in upper pectoral muscle thickness compared to subjects performing horizontal bench press [16]. Thus, it is plausible that increases in neural drive elicited by different bench press modalities may mediate hypertrophic adaptations; however, no studies to date have examined this outcome with offset loading.

Indeed, the current literature investigating OSL is minimal; therefore, the proverbial door is left open for further investigation. In fact, the two aforementioned studies [27, 41] are the only ones to date that have investigated the acute effects of OSL, and both studies are consistent with reporting higher muscle activation in muscles exposed to the heavier side of the offset load. This leads to the following question: does higher muscle activation previously reported to be induced by OSL enhance hypertrophy and strength outcomes relative to TDL. Hence, the purpose of this study was to compare the effects in pectoralis major muscle thickness and muscle strength from OSL to that of TDL in the bench press exercise. Additionally, we also observed the impact of OSL on perceptual measures of recovery, exertion, and mental effort relative to TDL.

Methods

Study Participants

Participants were recruited via social media advertisements as well as electronic communications with previous study participants at the laboratory. Thirty participants volunteered but only 21 qualified for the study. The inclusion criteria were males aged 18 to 45 years; at least 5 years of resistance training experience with the barbell bench press, and a bench press one repetition maximum (1RM) of at least their body mass; English literate, and the ability and willingness to sign an informed consent form. Exclusion criteria were musculoskeletal injury within 6 months of study enrollment, prior use of OSL techniques; history of drug, alcohol, or anabolic steroid use; daily usage of non-steroidal anti-inflammatory drug; or any serious health ailments that would impair study participation (thrombosis, high blood pressure, arrhythmias). For study enrollment, all participants signed an informed consent form detailing the procedures and inherent risk of the study, which was approved by an external Institutional Review Board and in agreement with the Declaration of Helsinki [48]. A total of 20 of 21 enrolled participants completed the study, as one participant of the TDL group did not attend post-testing. Participants were instructed to maintain their habitual diet and refrain from ergogenic aid consumption during the training period. Baseline descriptive characteristics for the 20 participants that completed the study can be found in Table 1. Height and body mass were confirmed using a stadiometer (Seca, Model 217; Chino, CA, USA) and body composition variables were collected via multi-frequency bioelectrical impedance (InBody 770 Body Composition Analyzer; Cerritos, CA, USA) following an overnight fast (≥ 10 h). Prior to the body composition assessment, participants were given a manufacturer supplied antibacterial tissue (NaCl 0.9%, antimicrobial) to wipe the hands and feet to reduce surface bacteria and control surface electrical conductivity.

Table 1 Participant baseline descriptive characteristics
Full size table

Study Design and Protocol

Following enrollment and approximately 1 week prior to pre-testing, participants underwent one familiarization session and one baseline session separated by at least 48 h to understand particulars of testing procedures and generate baseline data. Day one of familiarization included numerical scales used to assess perceptual measures, body and hand positioning for the bench press, bench press warm-up protocol, and bench press exercise session protocol. To familiarize themselves with a bench press exercise session, all participants performed the bench press exercise for 3 sets of 10 repetitions at 60% of their self-estimated 1RM. The 3 sets were randomized in loading styles as one set was performed with OSL on the dominant side, one with OSL on the nondominant side, and one with TDL. The perceived recovery status (PRS) scale was administered before initiating the control session and rating of perceived exertion (RPE) and mental effort (ME) scales were assessed 5 min after the control session. On the baseline testing day, muscle thickness (MT) of the dominant and nondominant pectoralis major (D-PM and ND-PM, respectively) was measured, and participants completed an assessment for bench press 1RM. Pectoral dominance corresponded to hand dominance, which was determined by which hand participants use during throwing.

After completing pre-testing, participants were quartile ranked based on the sum of MT (D-PM plus ND-PM thickness) and participants within each quartile were randomly assigned to TDL and OSL groups in a counterbalanced fashion using an online software [44] used in prior research [7, 47]. Assessment evaluators were blinded to the participant’s condition. Following randomization, participants underwent OSL or TDL on the bench press twice a week, for 4 weeks (8 sessions). All participants completed training sessions on the same two days each week (Mon/Thurs or Tue/Fri). Additionally, participants performed their respective training groups at the same time of day on every visit to the lab (± 1 h). During this training program, participants were instructed to refrain from upper body pressing exercises utilizing the chest musculature as the primary mover (e.g., incline and decline bench press, chest flies, push-ups, dips, and variations thereof). For each training session, participants completed 5 sets on the bench press with 2 min rest between sets. This rest period length was chosen as previous reviews have suggested shorter rest periods may optimize hypertrophy whereas long rest periods may improve strength to a greater degree [18]. Thus, the selection of an intermediate rest period may allow for optimal adaptation in both traits. The repetition scheme progressed from 12 repetitions per set in Week 1 to 6 repetitions per set in Week 4, with an increase in intensity each week to promote strength adaptations. In the OSL group, the offset load distribution was randomized between the right and left sides of the barbell for sets 1–4 with the load being reset to balanced distribution on the 5th set (mimicking TDL). Participants were blinded to the randomization order of offset loads, which was chosen to reduce the impact of prior knowledge of load placement that may have led some participants to immediately compensate for the offset load. Prior to each training session, participants completed the PRS scale and, 5 min following each session, they provided RPE and ME using standardized numerical scales. After completing 8 training sessions in a 4-week period, participants returned to the laboratory on Week 5 for post-testing. Post-testing was completed 48–72 h following the last training session and consisted of MT measurements and bench press 1RM.

Muscle Thickness

Two-dimensional, B-mode ultrasound (GE LOGIQ P6, General Electric Company; Boston, MA, USA) imaging with a 12 MHz broad-spectrum linear-matrix array transducer (Model ML6-15, General Electric Company; Boston, MA, USA) was used to assess MT of the D-PM and ND-PM with participants lying supine. Water soluble hypoallergenic ultrasound gel (Aquasonic® CLEAN® Ultrasound Gel, Parker Laboratories, INC; Fairfield, NJ, USA) was applied to the transducer to optimized spatial resolution and then placed at the marked measurement site with minimal force to avoid compression. Clavicle length was measured with a Gulick tape measure (Blue Jay™; Middetown, NY, USA) with participants laying supine with the elbow fully extended and the shoulder fully adducted to mimic the anatomical position. In this position a distinguishable mark was made at the clavicle midpoint. The ultrasound transducer was perpendicularly placed beneath the clavicle midpoint between the third and fourth costas of the rib cage [36, 49] and onscreen electronic calipers with a single perpendicular line from bottom to the top of the muscle sheath (between the superior and inferior layer of the facia) were used to measure MT of the D-PM and ND-PM. An example of a resulting image is provided in Fig. 1. Two images for MT were captured and measured to the nearest 0.01 cm. The resulting measurements of the two images were averaged and used for data analysis. Additionally, the sum of both pectoral MT sites (ΣMT) was used for analysis. The MT assessments were performed at the same time of the day at pre-testing and post-testing. In an attempt to minimize training-induced muscle swelling, images were obtained 48–72 h after the latest training session before the commencement of the study and 48–72 h after the last training session at the end of the study. This time window was adapted from other research assessing muscle hypertrophy via ultrasonography [4, 19, 40], and this is consistent with research showing that an acute increase in MT returns to baseline levels within 48 h following a resistance training session in trained individuals [11]. Additionally, to improve reliability, images were saved prior to making the measurement at pre-testing and images were notated for the distinct point along the fascia where the measurement was taken on the images using the crosshair tool from the manufacturer’s software to ensure the same measurement location at post-testing. The intraclass coefficient of variation and typical error for PM-MT was computed to be 1.5% and 0.04 cm, respectively, which was determined prior to the start of the study by assessing ten different participants with similar characteristics to the current participants on three different occasions separated by 24 h. Using the variance components from a one-way analysis of variance (ANOVA), the ICC and 95% confidence interval was calculated at 0.992 (0.987 to 0.998).

ICC =\({\sigma }^{2}\left(b\right)/({\sigma }^{2}\left(b\right) + {\sigma }^{2}\left(w\right))\)

$${\sigma }^{2}\left(w\right)=pooled variance within subjects$$
$${\sigma }^{2}\left(b\right)=pooled variance between subjects$$
Fig. 1
figure 1

Example ultrasonographic image of pectoralis major muscle thickness. Yellow dashed line represents onscreen digital calipers used for pectoralis major muscle thickness measurement (2.62 cm in this example). AT subcutaneous adipose tissue, PM pectoralis major muscle, IIC internal intercostal muscle

Full size image

One-Repetition Maximum Muscular Strength

Symmetrically loaded bench press 1RM was determined at pre-testing. All participants performed a standardized warm-up that involved 5-min on a cycle ergometer at approximately 140 W followed by a general dynamic warm-up that included movements for both the lower and upper body. Following the general warm-up, participants began a specific warm-up on the bench press by performing 10 repetitions at 40% self-estimated 1RM, 8 repetitions at 60% self-estimated 1RM, 5 repetitions at 70% self-estimated 1RM, and 3 repetitions at 80% self-estimated 1RM with 2-min of rest given between each warm-up set. The self-estimated 1RM was the load that the participant himself estimated to be his one repetition maximum on the symmetrically loaded bench press [21]. After the specific warm-up, participants began performing single rep attempts at 90% of their estimated 1RM. Five minutes rest were given between each attempt and each successful attempt resulted in an additional 2.5–10 kg added to the next attempt. This procedure continued until participants reached either actual or voluntary failure on a selected load (i.e., if a participant declined the next attempt, the highest previous successful attempt was considered their 1RM). All participants reached their bench press 1RM within 4 attempts. No supportive equipment of any kind was allowed during bench press 1RM testing. Hand spacing on the barbell was controlled at 150% biacromial width [22] for all participants and thumbs were required to be wrapped around the bar. A successful attempt required the participants to keep their back, head, shoulders, and buttocks in contact with the bench at all times and feet flat on the floor during the lift and the barbell had to touch the xiphoid process and return to the starting position with full elbow extension without any bouncing of the barbell or assistance from the spotters [23]. Two experienced spotters (National Strength and Conditioning Association-Certified Strength and Conditioning Specialists) were present for all attempts to ensure safety and judge successful attempts. Verbal encouragement was given to all participants during 1RM attempts. All 1RM testing was performed using a 20.4 kg power barbell and standard power rack (EliteFTS; London, OH, USA). Using the variance components from a one-way analysis of variance (ANOVA), the ICC and 95% confidence interval was calculated at 0.997 (0.994 to 0.999).

Perceptual Measures

The perceptual measures collected in this trial were perceived recovery status (PRS) scale, rating of perceived exertion (RPE) and mental effort (ME). For assessing perceptual measures, standard instructions and anchoring procedures were explained during the familiarization session. The PRS is a numerical scale from 1 to 10 with visual descriptors of “very poorly recovered” “adequately recovered” and “very well recovered” for perceived recovery are presented at numbers 0, 5, and 10, respectively. The PRS scale was administered to participants following the bench press warmup, prior to the training session [32]. The purpose of the scale was to assess the participant’s recovery from the previous exercise session.

RPE was collected approximately 5 min after the completion of every training session using the CR-10 Borg RPE scale [14] so that the fatigue from the final bench press set did not influence the RPE rating. Participants were asked to select a number on the scale to rate their overall effort. A rating of 0 was said to be no effort (rest) and a rating of 10 was considered to be maximal effort.

To assess ME, an adapted version of the Paas scale, a 9-point numerical scale ranging from 1 to 9, was used. This scale was shown to be reliable in a study testing problem solving and cognitive load [38] and has since been used to explore the effectiveness of graphical representation in learning [26]. For this study, the scale was used to assess ME during the training sessions. Five minutes following bench press cessation, the participants were instructed to “Please rate the mental effort required to complete the workout.” The rating descriptors are: 1 = very, very low mental effort; 2 = very low mental effort; 3 = low mental effort; 4 = rather low mental effort; 5 = neither low nor high mental effort; 6 = rather high mental effort; 7 = high mental effort; 8 = very high mental effort; 9 = very, very high mental effort. All perceptual measures were collected in isolation from other participants to promote accuracy and recorded as arbitrary units (a.u.).

Training Intervention

Prior to each training session, participants completed a standardized warm-up protocol that began with a general warm-up (cycle ergometer and dynamic warm-up) then a specific warm-up (20% 1RM × 15, 40% 1RM × 10, 60% 1RM × 5). Then, participants began their prescribed load and repetition scheme as described in Table 2. Offset load distributions were set such that there was approximately 52.5% of the adjustable load on the loaded side and approximately 47.5% on the de-loaded side (Table 3) for a 5% offset. For example, a prescribed load of 60% of a 118 kg 1RM would be set at 70.8 kg. The adjustable load is the prescribed load (70.8 kg) minus the barbell mass of 20.4 kg, equaling 50.4 kg of adjustable load. Adjusting to the balanced load used in TDL would amass 25.2 kg on each side. In OSL, a 50.4 kg adjustable load would create a loaded side of 26.5 kg and a de-loaded side of 23.9 kg, accounting for 52.5% and 47.5% of the adjustable load, respectively. All loads were rounded to the nearest 0.2 kg. The TDL group maintained a balanced load distribution for all sessions. In the OSL group, loaded and de-loaded sides were randomly applied to the dominant and non-dominant sides for sets 1–4 such that each side (dominant and non-dominant) was equally exposed to each loaded outcome (loaded and de-loaded) within each training session. The 5th set of every OSL session was reverted back to balance load (mimicking TDL). In both OSL and TDL groups, whenever participants failed to complete their prescribed repetitions in a given set, the absolute load was reduced by 10% for the following set. Bench press volume was calculated for each session as (volume = load × reps × sets).

Table 2 Bench press exercise prescription
Full size table
Table 3 Offset loading examples
Full size table

Statistical Analysis

Raw data for muscle strength, muscle thickness, and perceptual measures are presented as mean ± standard deviation (SD). Normality and variance were assured via the Shapiro Wilk and Levene’s test, respectively. Baseline characteristics, cumulative volume load, and overall study averages of perceptual measures were compared between groups using parametric, two-tailed, unpaired t-test. A two-way mixed ANOVA was performed on dependent variables of strength and muscle thickness assuming group (i.e., TDL and OSL) and time (i.e., Pre and Post) as fixed factors and participants as a random factor. In the event of a significant F value, a post hoc test with Tukey’s adjustment was applied for multiple comparison purposes. The alpha level was set at 0.05.

Magnitudes of mean changes were assessed by standardization (i.e., an effect size; the mean difference divided by the appropriate SD). For between-group comparisons, Cohen’s d effect sizes were calculated as (M1 minus M2) / SDpooled where M1 and M2 are the mean differences (Post minus Pre) for each group and SDpooled is the pooled standard deviation of the changes from each group. For within-group mean changes, Cohen’s dz effect sizes were calculated by dividing the mean difference (Post minus Pre) by the SD of the mean differences. Perceptual measures and bench press volume data are reported as the overall study average and Cohen’s d effect sizes were determined by (M1 minus M2) / SDpooled where M1 and M2 are the overall study averages for OSL and TDL, respectively, and SDpooled is the pooled standard deviation from each group. Uncertainty in all outcome measures was expressed with 90% confidence intervals (CI). Non-clinical magnitude-based inferences were used to provide an interpretation for all differences with inferences based on the standardized thresholds for small, moderate, and large differences set at 0.2, 0.6, and 1.2 of the standardized mean differences. The chance of the difference being substantial or trivial was interpreted using the following scale: possibly, 25%–75%; likely, 75%–95%; very likely, 95%–99.5%; and most likely, > 99.5%. If the qualitative chances of the effect being greater or lower were both greater than 5%, the effect was assessed as unclear. Otherwise, the effect was interpreted as the observed chance [25].

Results

No differences were found between groups for characterization parameters (P > 0.05, Table 1) or baseline 1RM (P = 0.427), 1RM:BM (P = 0.615), ΣMT (P = 0.860), MT of the D-PM (P = 0.863) or N-PM (P = 0.862). No differences were detected for cumulative volume load (OSL: 27,057.1 ± 5642.3 kg, TDL: 25,077.3 ± 4673.7 kg; P = 0.411). The percentage of prescribed repetitions completed was 96.0% ± 5.9% and 97.6% ± 3.5% for the OSL and TDL groups, respectively (failed normality testing; nonparametric Mann Whitney P = 0.858). Additionally, the attendance for the training sessions was 100% for both groups.

There were no group-by-time interactions for MT of the D-PM (P = 0.136) or ND-PM (P = 0.085), ΣMT (P = 0.071), bench press 1RM (P = 0.179), or 1RM:BM (P = 0.142). Time effects were detected for each of the 5 aforementioned hypertrophy and strength variables (each variable P < 0.001). No differences were found between groups for overall study averages in RPE (OSL: 6.61 ± 1.41, TDL: 5.81 ± 1.88 a.u.; P = 0.292), ME (OSL: 5.84 ± 1.54, TDL: 4.91 ± 2.06 a.u.; P = 0.261), or PRS (OSL: 7.52 ± 0.81, TDL: 7.68 ± 0.76 a.u.; P = 0.652).

Using non-clinical magnitude-based inferences, no clear substantial differences were found between groups for characterization parameters (Table 3) or for cumulative volume load [OSL: 27,057.1 ± 5642.3 kg, TDL: 25,677.3 ± 4673.7 kg; ES (90%CI) = 0.26 (−0.48, 1.01); unclear, 55/29/16].

Both study groups were independently evaluated (Post vs. Pre) for their likelihood of obtaining small (0.2), moderate (0.6), and large (1.2) changes in dependent variables for muscle thickness and strength (Table 4). There were most likely (100/0/0) small magnitude increases for all muscle thickness and strength variables in the OSL group and the inference for a small increase in the TDL group ranged from very likely to mostly likely. For muscle thickness outcomes, very likely to most likely moderate increases were produced in the OSL group and likely to very likely moderate increases were produced in the TDL group. There were possibly to likely large increases in muscle thickness outcomes in the OSL group and the TDL group produced a possible large increase. The change in relative strength (1RM:BM) and absolute strength (1RM) was likely to most likely a moderate increase in both groups. Furthermore, the OSL group very likely produced a large magnitude increase in both of these strength measures whereas this evaluation was likely to possibly trivial in the TDL group.

Table 4 Within-group data for muscle strength and thickness with inferences evaluated at small, moderate, and large magnitudes
Full size table

There was likely a small difference in the changes from Pre to Post between the two groups for all muscle thickness and strength variables (Table 5). When evaluating Pre to Post changes between groups at a moderate magnitude, D-PM MT, ND-PM MT, ΣMT, and 1RM:BM were possibly greater/possibly the same (trivial) in the OSL group compared to the TDL group (Table 5). Between-group differences evaluated at large magnitudes were unclear and did not yield substantial differences.

Table 5 Between-group data for muscle strength and thickness
Full size table

When evaluated at small magnitudes, overall study RPE [OSL vs. TDL: 6.61 ± 1.41 vs. 5.81 ± 1.88 a.u; ES (90%CI) = 0.45 (− 0.3, 1.21); 72/23/5] and overall study ME [5.84 ± 1.54 vs. 4.91 ± 2.06 a.u.; ES (90%CI) = 0.48 (− 0.27, 1.24); 74/21/5] were possibly greater in the OSL group compared to the TDL group. Overall study PRS was unclear [7.52 ± 0.81 vs. 7.68 ± 0.76; ES (90%CI) = − 0.19 (− 0.90, 0.52); 19/50/51]. No clear substantial differences were found when evaluating perceptual measures at moderate and large magnitudes.

Discussion

The primary purpose of this experiment was to investigate the hypertrophy and strength adaptations to OSL compared to TDL in the bench press exercise. Our secondary purpose was to observe the impact of OSL on perceptual measures of PRS, RPE, and ME relative to TDL. The findings did not yield an interaction effect for any variables assessing muscle hypertrophy or strength, and overall study averages of perceptual measures were not different between groups. However, magnitude-based inference analyses indicated likely beneficial effects for OSL compared to TDL for all strength and hypertrophy variables and possibly greater effects for RPE and ME. Thus, according to our findings, it is plausible to assume that a four-week mesocycle of OSL can enhance adaptations in hypertrophy and strength and our findings may be used to justify its inclusion in strength and conditioning programs.

Muscle Thickness

We report favorable likelihoods for each group to small and moderate increases in muscle thickness outcomes. The groups began to differ when evaluated at a large magnitude as the OSL was likely to produce this magnitude of change for ND-PM MT and ΣMT whereas the TDL group possibly produced this magnitude. When comparing groups, we report a likely greater effect (evaluated at a small magnitude) in OSL over TDL for MT in the ND-PM (9.5% vs. 5.0%) and D-PM (12.3% vs. 7.7%) muscles, and ΣMT (11.0% vs. 6.4%). The greater MT seen in the OSL group may be attributed to the novel training stimulus of OSL. Practitioners often examine training volume in respect to hypertrophy adaptations [42], however, in our study, no differences in training volume were found between groups. Fonseca et al. [20] previously demonstrated that utilizing a variety of novel exercises was more effective at inducing hypertrophy in the quadriceps muscles as opposed to employing a single volume-matched exercise (barbell squats).

The novelty of a loading pattern and its effects on muscular hypertrophy is not completely understood. Mechanical tension is likely the primary driving stimulus of muscular hypertrophy [45], thus, training variables designed to maximize mechanical tension are presumed to induce greater hypertrophy. Importantly, muscles experience this mechanical tension at the individual muscle fiber level through somatosensory receptors called mechanoreceptors. The primary mechanism through which mechanoreceptors detect mechanical tension is deformation of their plasma membrane due to muscular force production [15]. Therefore, for a muscle fiber to detect mechanical tension and, eventually, undergo hypertrophic remodeling, a muscle fiber likely must be activated during a given exercise. This can be inferred through previous research showing that the most activated portion of a muscle exhibits the greatest hypertrophy response following a long-term strength training program [46].

The necessity of mechanical tension as a driver of muscle hypertrophy is underlined when comparing long-term training adaptations to endurance training and strength training. While both exercise modes can induce high levels of metabolic stress and muscle damage, the discrepancy in mechanical loading is a major difference between each type of exercise [45]. Since the OSL group randomized the order and placement of the offset load, a plausible theory is that the increased load lifted by the PM muscles during OSL, as well as higher muscle activation in the PM during dynamic OSL repetitions, as evidenced by Jarosz et al. [27], led to likely greater increases in PM thickness compared to TDL.

While these results certainly require further investigation in future research, it is plausible that using OSL provides a novel stimulus which can produce greater muscle hypertrophy over TDL. Importantly, these effects were observed in as little as a 4-week mesocycle; interventions of greater duration may discover more significant findings. Moreover, it is also conceivable that the training status (1RM:BM ~ 1.3) and experience (~ 10 years) of our participants underlined the effect of the novel stimulus. It is unclear if untrained or recreationally trained individuals would garner the same hypertrophic benefits of OSL compared to TDL.

One-Repetition Maximum Muscular Strength

Both groups exhibited to mostly likely moderate increases in bench press 1RM (7.1% vs. 4.9%) and 1RM:BM (6.9% vs. 4.8%). When evaluated at a large magnitude, the strength improvements in the OSL group remained very likely substantial and the TDL improvements were possibly trivial. In the evaluation of a small magnitude of the between-group differences, there was a likely greater increase for OSL compared to TDL in bench press 1RM (82/15/3) and 1RM:BM (85/12/3). These results are certainly interesting given the chasm in bench press specificity between groups. By study design, the OSL group only performed 20% of their training volume with a symmetrical (TDL) bench press load. Thus, it would have been reasonable to expect a greater increase in bench press 1RM in the TDL group. However, it could be interpreted from previous research that multiple variations of bench press training can improve 1RM in the TDL bench press. More specifically, Langford et al. [31] found that groups performing machine, barbell, isokinetic, and water-filled log bench press training for 10-weeks were all able to increase TDL bench press 1RM to a similar degree.

Thielen et al. [43] recently reported similar improvements in squat 1RM in a group performing traditional squat training and a group undergoing suspended load squat training. Intriguingly, the study participants utilized in this study were collegiate baseball players with at least 2-years of resistance training experience on average and a relative squat strength of approximately 1.75 times body mass. Therefore, while the specificity of strength improvement in resistance training is still a valid principle, it may not be regarded as a rule in scenarios such as ours, Thielen et al. [43], or Langford et al. [31]. Regardless, the degree of specificity differences between OSL and TDL may not be great enough to produce unique strength adaptations to each mode of bench press, but this outcome certainly warrants further investigation.

The likely greater PM MT seen in the OSL group also potentially influenced strength outcomes as Akagi et al. [1] reported a significant relationship (r = 0.866; P < 0.001) between PM cross-sectional area and bench press 1RM performance. While it is generally believed that muscle size more so influences 1RM performance of single joint exercises that are not as reliant on skill [9], it is possible from the results of Akagi [1] and our own that prime mover muscle size still likely plays a large role in 1RM performance of the bench press exercise.

Finally, progressive overload is a largely accepted principle of strength and conditioning and widely acknowledged as a primary prerequisite to improvements in muscular strength. Bompa and Buzzichelli [13] (pp. 41–52) note that both changing exercises and changing loading patterns can constitute a form of progressive overload. OSL likely falls into one or both categories and, thus, may be instituted as a form of progressive overload beyond TDL bench press training. Due to the training experience of our participants, the inherent progressive overload from OSL may have influenced the likely greater strength gains seen in this group.

Perceptual Measures

An intriguing, but not unexpected, finding was reported in regard to the perceptual measures of RPE and ME as the OSL group showed a possibly greater effect in both measures compared to the TDL group. Due to the novelty of the technique and the uncomfortable nature of the loading pattern, it is conceivable that OSL provoked greater perceived exertion despite a similar volume load to TDL. Greater RPE has been reported when training on unstable versus stable surfaces [3], suggesting that exercise with greater stability requirements elevates exertion and this could be applied to OSL in theory.

Mental effort can be defined as the degree of strain experienced by the mobilization of controlled cognitive processing with limited energetic resources [39]. If a task requires executive function or self-regulation, ME is needed the instant the individual begins mobilizing the task. In exercises performed at uncomfortable intensities, ME is necessary to sustain neural drive and force output throughout the bout [5] and inhibit the bodily afferences that emerge with physical fatigue [6]. Thus, one potential influential factor in the ME data may be that OSL is inherently more intense than TDL. Indeed, Saeterbakken et al. [41] found that a 5% offset load resulted in a 12% reduction in 1RM load relative to a traditionally-loaded 1RM. While participants in both groups achieved a similar number of successful repetitions in each training session, participants in the OSL group may have been training at a higher relative intensity due to the disparity in 1RM load between 5% offset conditions and traditionally-loaded bench press. An additional explanation may lie within the inherent elaborate processing, which may occur when comparing OSL to TDL. Specifically, we adjusted loads on the final (5th) set in all OSL sessions to a balanced (traditional) load as we believed that OSL would force lifters to compare and contrast their previous OSL sets to the final TDL set, thereby forcing their TDL bench press back into a high cognitive effort phase in order to enhance strength transfer. According to the elaborate processing model, this comparison and contrast situation would induce a more robust cognitive experience [33].

Finally, PRS measures between groups resulted in an unclear effect. While the novelty of OSL may have impacted this measure early on in the training protocol, previous research has shown that the repeated bout effect can significantly ameliorate muscle damage and perceived soreness even following a single training session [17]. Therefore, over a 4-week training program, one would not expect OSL to result in higher levels of muscle soreness following the initial sessions.

Study Limitations

As is common in strength training literature, a primary limitation of this study is the small sample size. Indeed, this limitation may make it difficult to apply our results to the general population, however, research into OSL is certainly in its infancy and future investigations will help develop a deeper understanding of this novel training method. Furthermore, the current study utilized participants who were trained (1.31 and 1.26 1RM:BM for OSL and TDL, respectively) and well experienced (~ 10 years on average) on the barbell bench press exercise. Other populations (e.g., untrained, elderly, or adolescents) may experience force imbalances from OSL to an extent that increases injury risk. Therefore, the performance outcomes and the safety of OSL should not be generalized to all populations. Also, during participant recruitment, the authors did not account for bench press training frequency prior to (< 6 months) the commencement of the study. Even though both study groups had similar bench press strength at baseline, and similar training experience, it is possible that prior training frequency could have been a confounding variable influencing the response to the training protocol.

Additionally, the length of the training period and not conducting ultrasonography measures on secondary movers of the bench press (anterior deltoid and triceps brachii) may be considered limitations. However, our results suggest that adaptations took place in both groups, indicating that 4-weeks appears to be sufficient for detecting resistance training adaptations, but performing longer training programs may provide further clarification of future results. The PM was the only muscle group subjected to ultrasonography as it is the primary mover of the criterion movement in this study (barbell bench press).

Lastly, another limitation is the randomization of offset loads in the OSL group. This method was chosen in order to reduce participant anticipation of load position which may have resulted in pre-planned compensation patterns. Importantly, the participants were blinded to the randomization order of the offset loads. Future studies should endeavor to better understand the use of targeted OSL, in which OSL is employed in the case of a known asymmetry, which may persist following musculoskeletal injuries, surgery, or even sport-specific training. Since this is the first study investigating adaptations to a training program involving OSL, future studies should experiment with targeted OSL to uncover a greater understanding of this novel training method. For instance, a crossover study design implementing OSL and TDL conditions may be useful in more long-term studies to verify the theory that OSL is likely a form of progressive overload.

Conclusion

This study demonstrated, for the first time, the training adaptations to OSL versus TDL in the barbell bench press exercise. TDL has been well-established as a beneficial training modality for muscular adaptations such as strength and hypertrophy. The novelty in the current study is that magnitude-based analyses supported a likely beneficial effect from OSL in muscle thickness and strength adaptations compared to TDL; thus, suggesting a progressive exercise effect that can be used to promote continual adaptations. The outcomes of magnitude-based inferences reported in this investigation permits sports practitioners and coaches the opportunity to judge the practical relevance of OSL. While the concept and application of OSL certainly merits future inquiry, OSL is a unique training method that can be applied in strength and conditioning settings to assist in optimizing hypertrophy and strength outcomes or when progressive overload beyond the TDL bench press is required.

Future investigations should endeavor to confirm our findings as well as uncover the long-term impacts of OSL on injury risk and overall performance. Ultimately, an athlete cannot display maximal physical skills while sidelined with an injury, and any training modality that can alter injury risk is worthy of study. OSL may be a novel training method that reduces asymmetrical neural drive between dominant and nondominant limbs, and further research is warranted for investigating these long-term outcomes and effects on injury rates in sport.