Introduction

Strongman training (ST) is an unconventional style of resistance training that incorporates compound movements including lifting and pulling oddly shaped objects, such as sleds, stones, anvils, keg barrels, logs, tractor tires, and trucks (^{8,9,16,22,23,31}). Strength and power athletes such as football, wrestling, hockey, rugby, and basketball players may benefit from ST during their general physical training (GPT). Typically, the GPT stage of the periodization cycle focuses on building a solid physiological foundation of strength, speed, and flexibility consisting of high volume regimens. It is suggested that strongman exercises are ideal for this phase because of the large amount of time under tension and activated muscle mass for some of the lifts, such as the keg carry and chain drag (⁴⁴). In addition, recent research has reported that strongman competitors integrate all phases of resistance training (i.e., hypertrophy, strength, power) into their conditioning program. Winwood et al. (⁴³) surveyed 167 strongman competitors and found that while ST programs incorporate a variety of training modalities, 97% of the competitors are focused on a universal training goal, which is increasing maximal strength.

Critics of ST believe that this training increases the risk of injury because of the lack of specific movement patterns (⁸). The need to purchase and store the specific equipment required for ST has also contributed to the low incidence of its use (¹⁶). Despite these negative opinions, ST has been shown to complement traditional resistance training (⁸). Proponents of ST feel that the added sense of competition and variety incorporated into the workouts is beneficial for athletes (^8,9). In addition, previous studies have reported that ST has the potential to generate physiological adaptations that lead to increased strength and power. Keogh et al. (²²) reported that during heavy sled pulls, subjects with faster trials demonstrated greater step length, step rate, and knee extension at toe-off, implying greater propulsive forces in the anterior-posterior plane. Increased propulsive forces may assist in contact sports where forceful collisions generated from high levels of horizontal body momentum may occur. Keogh et al. (²³) also reported large significant increases in heart rate and blood lactate responses after 2 sets of the tire flip, suggesting that ST imposes a high degree of physiological demand on the body. In addition, Berning et al. (⁹) reported that heavy truck pulls require extreme physiological arousal, resulting in an increased lactic acid accumulation, which is a response associated with increased testosterone secretion (²⁰).

The importance of the anabolic response from exercise-induced endogenous testosterone levels is a critical modulator of muscle hypertrophy (⁴⁰). A systematic review from Vingen et al. (⁴⁰) reported that heavy resistance exercise (RE) initially downregulates muscle androgen receptor (AR) content followed by an upregulation during recovery. This increases the binding of testosterone to the intracellular AR-inducing gene transcription, resulting in a net protein balance. Although testosterone can be elevated from a complex interaction of acute program variables (i.e., volume, choice of exercise, rest intervals), it is well documented that testosterone is elevated from short-term high-intensity anaerobic exercise (²⁶), specifically hypertrophy RE training routines. Generally, this consists of performing 3–4 sets of 10–12 repetitions using loads of approximately 75–80% of 1 repetition maximum (1RM) with short rest periods (1–2 minutes) (^{1,26,29,30,35}). Inducing an anabolic endocrine response (elevating testosterone levels) results from RE programs that incorporate heavy resistance using large muscle groups (^12,26,34,35). The structure and nature of ST is similar to this RE regimen, and for that reason, it is possible that performing ST may acutely elevate postexercise (PST) testosterone levels and result in skeletal muscle changes that are critical for muscle hypertrophy.

Although previous studies have examined the metabolic effects, biomechanics, and muscle activation of ST (^9,22,31), research has not examined the hormonal response to ST, specifically free testosterone levels. Therefore, the purpose of this study was to examine the testosterone response from 2 novel ST protocols compared with an established hypertrophic (H) protocol that acutely elevates testosterone levels. The results from this study will add to the limited body of knowledge pertaining to the training response from ST. Strength practitioners may be able to use this information by implementing an alternative method of resistance training that results in testosterone elevation. Incorporating these strategies into individual's training may potentially result in an optimal anabolic milieu critical for muscle hypertrophy. Based on the nature of the ST exercises, which incorporate multi-joint compound movements at a high intensity, our primary hypothesis was that salivary testosterone would be acutely increased immediately PST using either of the ST routines. We also hypothesized that the ST protocols would elicit a larger pre- to post-testosterone response (spike) compared with the H protocol because of the greater number of lower-body exercises included in our strongman protocols.

Methods

Experimental Approach to the Problem

All subjects acted as their own control and were randomized in counterbalanced order to complete 1 of 3 RE protocols consisting of an H, ST, and a mixed (XST) hypertrophy/strongman. Saliva samples were collected pre-exercise (PRE), immediately postexercise exercise (PST), and 30 minutes postexercise (30PST) to determine free testosterone concentrations. To identify which RE protocol elicited the greatest testosterone response, a 2-way analysis of variance (ANOVA) group by time (3 × 3), with training program as the independent variable and salivary testosterone as the dependent variable, was used to analyze the data set. All subjects completed a medical health history questionnaire and demonstrated that he did not have any health-related contraindications to physical activity. All procedures were approved by the University Institutional Review Board Committee.

Subjects

Sixteen men (Table 1) volunteered to participate in this study. Of the 16 subjects, 15 completed all 3 protocols. One subject was dropped from the final analysis due to the inability to complete 2 of the RE sessions. Ten subjects were recreational lifters (training at least 4 days per week (>2 years), 2 were collegiate athletes (wrester and football player), 2 were competitive bodybuilders (lightweight and light-heavyweight), 1 was a competitive powerlifter (heavyweight), and 1 was a competitive Olympic lifter (lightweight).

All subjects were recruited from Gridiron, Inc., which is a high-intensity strength training facility that caters to high school, collegiate, and recreational athletes. Each subject was familiar with the ST regimen, although none of the subjects were competitive strongman athletes and thus were not familiar with the proper technique for each exercise. None of the subjects were taking any dietary or performance-enhancing supplements. The risks and benefits of the investigation were explained to each subject, and all subjects signed an informed consent before participating in any facet of the study. Each subject was screened before participation to ensure that they were medically and physically able to complete the exercises. All subjects completed a medical health history questionnaire and demonstrated that he did not have any health-related contraindications to physical activity.

Preliminary Testing Procedures and Dietary Controls

Subjects recorded all dietary intakes during the 2 days before each resistance training session and were instructed to match their caloric intake before each training sessions. In the 2 days before each resistance training session, the subjects refrained from exercise and ingestion of alcohol and stimulants. Analysis of the dietary records indicated that subjects consumed an average of 2500 calories per day, of which ∼44% was carbohydrate, ∼29% was fat, and ∼27% was protein. All dietary logs were analyzed by a certified sports nutritionist (Certified International Society of Sports Nutritionist) using Diet Power software (Danbury, CT, USA). In addition, subjects were instructed to drink 400–500 ml of water the night before and 400–500 ml the morning of their RE session to ensure adequate hydration. Adequate hydration was operationally defined as urine specific gravity ≤1.020 (²), and this was measured before the RE session via urine refractometry (ATAGO, Bellevue, WA, USA). Urine refractometry is a measure of hydration by measuring the density (concentration) of a urine sample (²). This value is reported as urine specific gravity, which is compared with the specific gravity of water (1.00). This measure was included because hydration status can influence testosterone responses to RE (²¹).

Familiarization and Repetition Maximum (3RM) Testing

Before participating in the training protocol, predicted maximal strength for the subjects was assessed. Each subject was familiarized with each of the REs and strongman exercises. After familiarization and instruction, subjects performed a repetition to failure protocol (⁴¹) for the squat, leg press, bench press, and seated row. For the squat exercise, subjects positioned the bar across the middle portion of the trapezius muscle, with the hands grasping the bar at a position near the weight plates. After removing the bar from the support rack, the subjects lowered the bar slowly by flexing at the knees until the bottom of thigh was parallel with the floor. Bench press repetitions followed the standard “touch and go” protocol. The bar was required to touch the chest before pressing to “full arms” extension. Subjects placed hands slightly wider than shoulder width apart on the barbell, and feet were placed on the ground during all sets. Subjects performed a warm-up lift of 5–10 repetitions at 40–60% perceived maximum exertion. Following a 1-minute rest with light stretching, subjects completed 3–5 repetitions at 60–80% of perceived maximum exertion. Following a 3–5 minute rest, a weight that was approximately 85% of their probable 1RM was loaded on to barbell. Using the selected weight, subjects performed as many repetitions to failure as possible. The final number of valid repetitions was recorded. A prediction equation was used to determine 1RM based on the number and weight of the repetitions (1RM = (0.033 rep wt) × repetitions + rep wt) (¹³).

Saliva Collection and Analysis

Salivary testing is a common method to assess testosterone levels in sports science research (¹⁸). Saliva testing is noninvasive and stress free and offers real-time sampling when blood collection is either undesirable or difficult to implement (²⁸). Salivary sampling is typically used to monitor training intensity (^6,12,18) and allows samples to be collected in a convenient manner on a repeated basis. Salivary testosterone reflects the circulating level of free testosterone, which is reported to be more physiologically relevant than total blood levels (³³).

Saliva samples were collected using a Salimetrics Oral Swab (SOS) nontoxic inert polymer. Subjects placed the SOS under the tongue for 2 minutes. After 2 minutes, the subjects placed the swab into a 30 × 10-mm cylinder, which was then placed into a capped conical centrifuge tube. Samples were immediately stored at −20° C (CMF151L—EdgeStar) and later shipped to Salimetrics, Inc. (State College, PA, USA) for singlet measure (10% duplicate) of free testosterone (picograms per milliliter) assay concentration. Subjects rested for 10 minutes before providing resting saliva samples (PRE). A second saliva sample was collected PST. The subjects relaxed for 30 minutes before supplying a third saliva sample (30PST). Subjects were instructed to avoid hard foods, drinking hot fluids, and brushing their teeth for approximately 30–60 minutes before testing, to minimize any risk of blood contamination in the saliva. Subjects replicated this sampling procedure for all 3 protocols. The average intra-assay and interassay coefficient of variation for testosterone was 3.7 and 14%, respectively, which are within acceptable range (Salimetrics, Inc.). Salivary analysis using this method has been shown to be reliable and acceptable for research purposes (³).

Description of Strongman Training Exercises and Equipment

The description, technique, and execution for the ST exercises were repeated from a previous published study (³¹) (Figure 1).

Tire Flip

Subjects were instructed to flip a large tractor tire (Figure 2) with a mass of 225–340 kg. The tire was a General Tire, model 23.5-25, with a standing height of 1.60 m and a width of 0.61 m. The tire was initially positioned on its side. Each subject took a deep squat position and hooked their hands under the bottom of the tire. Keeping the hips low, the subject drove their hips, knees, and ankles into full extension. After triple extension, the subjects dropped into a catch position and positioned their hands to an overhand driving position. The subjects then drove the base of their hands forcefully into the tire by extending their arms and pushing the tire over end to the ground once again. Each subject flipped the tire a total distance of 18.3 m.

Chain Drag

Each subject dragged a 170- to 225-kg chain (Figure 3) for 22.86 m. Subjects grasped 2 handles connected to the chain and began the pull by leaning back. The handles were maintained against the abdomen at all times by keeping the elbows flexed. Once the chain was pulled the required distance, it was placed down to rest.

Farmers Walk

In proper deadlift position (feet flat and shoulder width apart with back straight, shoulders back, head up, and chest up), the subject held one loaded bar (Figure 4) with a mass ranging from 29.5 to 46 kg in each hand. The load was carried forward 22.86 m and then set down. The distance carried (20 m) for this exercise was similar to other reported ST practices (⁴³); however, the load used was lighter than that encountered in competition. According to Winwood et al. (⁴³), 46.6% of strongman competitors use a heavier load than that in competition to increase grip and carrying strength. The lack of competition experience in our subjects resulted in using a lesser load than that typically observed with more experienced competitors.

Keg Carry

Each subject lifted a bar keg (Figure 5) loaded with sand with a mass of 63–75 kg from the ground, walked forward 13.7 m, and set it down.

Atlas Stone Lift

Each subject lifted a large stone (Figure 6) comprised of poured concrete with a diameter of 0.43 m and a mass of 110 kg from the ground to chest height. The subject started in a deadlift position and the arms and hands were wrapped around the stone while the spine was curved over the stone with the chest facing the ground. The subject proceeded to wrap his forearms around the stone and rolled it up to the chest so the stone sat on his lap in a deep-seated position. Subjects were allowed to adjust their hands and arms if necessary and stood straight up, thrusting hips forward.

Resistance Exercise Sessions

The primary objective of the RE sessions (Table 2) was to induce an acute endogenous testosterone response immediately PST. Salivary testosterone production from the novel ST protocols was compared with a previously established H protocol that has been documented to increase testosterone PST. The duration, rest periods, and intensities of all protocols were equal, and all exercise sets were performed to failure. Exercise selection (large multi-joint movements) and loading (75% 1RM) were chosen because they reflected commonly prescribed RE protocols that produce an acute hormonal response (²⁶).

The training intervention required 1 RE session per week across a 3-week period for a total of 3 sessions. A recovery period of at least 1 week between each session was required. To avoid the potentially confounding effects of circadian rhythm on hormonal concentrations, each session was scheduled at the same time and on the same day of each week. Before the first session, all subjects were instructed to replicate their pre-session behavior when returning to each of the subsequent sessions (e.g., sleep, exercise, and diet). In the ST protocol, subjects were required to perform 3 sets to muscle failure of 5 different ST exercises with a 2-minute rest interval between sets and exercises. Exercises consisted of the tire flip, chain drag, farmers walk, keg carry, and stone lift, in that order. For this type of training, it was difficult to quantify optimal resistance because of variability in equipment. If failure did not occur during the first set, then additional weight was added for activities such as the farmers walk and the chain drag exercises. For the stone lift, keg carry, and tire flip, heavier pieces of equipment were used.

In the H protocol, subjects performed 3 sets of 10RM to failure with an initial load of 75% of predicted 1RM maximum in each of the back squat, leg press, bench press, and seated row exercises, in that order. This protocol is similar in exercise selection to previously published protocols (^6,32); however, it differs in exercise order. During each set, when muscle failure was reached, assistance was provided until the subject completed the remaining repetitions. Resistance was then reduced for subsequent sets.

The XST session consisted of performing exercises from both the ST and H protocols. Subjects alternated from a ST to H exercise for a total of 5 exercises. All subjects performed 3 sets to failure of the following: tire flip, back squat, chain drag, bench press, and stone lift, in that order. The prescribed load (75% of predicted 1RM) for the traditional H exercises in the XST session was the same load as in the H protocol.

In all cases, a 2-minute rest interval was provided between sets and exercises. This regimen has been shown to acutely elevate testosterone levels (^29,34). All RE sessions were supervised by the same certified exercise physiologist, and all sessions were conducted at Gridiron, Inc. This ensured that all the required equipment for the ST sessions was available, all subjects completed the training in same environment, and all subjects received the same verbal instruction and encouragement for each session.

Statistical Analyses

Descriptive statistics (mean ± SD) were calculated for all dependent variables. Differences between the groups for PRE, PST, and 30PST salivary measures (testosterone level = dependent variable) were analyzed using 2-way factorial ANOVA with repeated measures. If significant main effects were attained, a Tukey post hoc test was used to assess differences between the groups. Fifteen subjects per group were recruited for the present study to detect a potential large (1.0–1.2) effect size for between-treatment differences for a power of 0.80 at significance level (α) of 0.05. This sample size was estimated using calculations from previous work (¹⁴) and is similar to sample sizes of previous studies (^6,7,12) comparing the effects of different resistance training protocols on free testosterone levels.

Results

Pre-exercise Testosterone

There was no significant difference in PRE testosterone levels between the groups (p ≥ 0.05). Pre-exercise testosterone scores for each group were 167.89 ± 77.56 pg·ml⁻¹, 188.40 ± 130.50 pg·ml⁻¹, and 222.12 ± 135.88 pg·ml⁻¹ for the H, ST, and XST groups, respectively. In addition, there was no significant difference in PRE urine specific gravity scores between the groups (p ≥ 0.05). The PRE urine specific gravity scores were 1.0178 ± 0.0096, 1.0212 ± 0.0083, 1.017 ± 0.0082, for the H, ST, and XST, respectively. Within-subject intraclass correlation (r) for PRE urine specific gravity across each protocol was 0.95.

Testosterone Response

There was a significant main effect (p ≤ 0.05) for time between the PRE and PST testosterone level within each group with a nonsignificant trend (p = 0.061) between the groups (Figure 7). Absolute PST to PRE delta scores (testosterone spikes) for the H, ST, and XST protocols were 225.23 ± 148.01 pg·ml⁻¹, 132.04 ± 98.09 pg·ml⁻¹, and 122.10 ± 140.67 pg·ml⁻¹, respectively (Figure 8). A significant simple main effect for time occurred between PST and 30PST for the H group (p ≤ 0.05), with a nonsignificant trend (p = 0.07) between the groups. Absolute PST to 30PST delta scores for H, ST, and XST groups were 109.87 ± 38.36 pg·ml⁻¹, 8.79 ± 22.04 pg·ml⁻¹, and 40.79 ± 29.45 pg·ml⁻¹, respectively (Figure 9). Spaghetti graphs for individual acute testosterone response for all 3 exercise protocols are depicted in Figures 10–12.

Discussion

The aim of the present study was to compare the effects of 2 different ST protocols and an established H protocol, that were all similar in duration and number of sets, on salivary testosterone. These data are congruent with previous testosterone work, suggesting that testosterone is significantly elevated when combining large compound movements using training intensities of 75% 1RM. The current study identified distinct PRE to PST testosterone spikes within each protocol, with a nonsignificant trend of greater testosterone release after the H protocol.

The most important finding is that our strongman protocols resulted in a significant increase from baseline in endogenous salivary testosterone levels of 74% and 54% for the ST and XST levels, respectively. This is comparable with previous recognized hypertrophy protocols (^4,12,15) but not with the current H protocol, which produced a robust increase of 137%. Several studies support RE as the most effective way to acutely increase testosterone, which in turn stimulates strength and muscle hypertrophy (^15,29,34) and is causally linked to strength and weight gain, thus maximizing functional gain (⁵). Recently, the relationship between elevated endogenous testosterone levels and hypertrophy function has been challenged (⁴²), and it is suggested that this premature assumption is based on individuals with deficient levels of anabolic hormones (i.e., growth hormone and testosterone). However, a much larger body of evidence supports the integral role that the acute hormonal response to RE has on muscle hypertrophy (^36,40) and its role in strength training adaptation (^19,27). Those in support of an endogenous testosterone response stand by the belief that RE causes an initial downregulation on AR content in the target tissue (i.e., skeletal muscle) followed by a subsequent upregulation during the recovery period, thus increasing free testosterone uptake facilitating protein synthesis (^35,37).

As expected, salivary testosterone spikes were observed after all 3 protocols (Figure 7). The testosterone spikes in our sample were substantially higher than results from previous studies that have also evaluated large multi-joint compound movements early in the session using intensities ranging from 70 to 80% of 1RM to failure with short rest periods. Previous literature examining different resistance loading schemes have reported a broad range of PST testosterone spikes, ranging from 4 to 13% (^6,7) to larger spikes of 45–90% (^4,12,15). Because of this large range, it is difficult to compare our results (54–137%) with previously established normative values. The abnormal response in our H group was attributed to 6 of our subjects reporting extreme spikes ranging from 165 to 493%. This introduced a large amount of variability for the hormonal responses, indicated by the large standard errors. The sampling procedure or variability within the population studied may have contributed to this variation, which has also been reported in other studies using similar sampling methods (^6,34) examining 3 time points (PRE, PST, and 30PST).

Although using saliva to examine free testosterone responses is simple and noninvasive, confounding factors, such as sampling procedure, interference of blood plasma in the saliva, gender, and age contribute to the variability of salivary testosterone level (¹⁷). Although RE will typically increase testosterone levels, some studies have reported no significant change or even decreased testosterone levels post-RE (^10,32). In our study, one subject's testosterone level decreased 46.7% immediately after the XST exercise protocol. This may be a result of the intensity of the protocol or the tight regulatory control of free testosterone in saliva (²⁵). Interindividual variability in pre-exercise to postexercise testosterone secretion can arise from differences in age (⁴), gender (²⁴), training experience (³⁸), and baseline values (²⁶), suggesting that individuals respond differently to different RE protocols during one's training cycle (⁶). These differences may have contributed to the large variability of the data and reduced the possibility of finding differences between our protocols.

Exercise selection and familiarity of the protocols may also have affected the results. Our subjects were more familiar with the H protocol exercises compared with the exercises in the ST protocol. It is plausible that the anxiety level due to the unfamiliarity of strongman lifts may have reduced the testosterone spike. This possibility is supported by previous literature examining the hormonal responses to different RE protocols in seasoned trainers (⁶). Beaven et al. (⁶) suggested that the novelty and stress of the situation are likely to be perceived based on experience. Thus, the stressors of the ST and XST sessions and the lack of familiarity of the exercises can suppress the actual physical nature of the stimulus. This psychological nature of the hormonal response in our subject pool may have caused a different response to protocols with which they were unfamiliar with or disliked.

A final factor to consider when interpreting the results from this study is the inability to accurately control and measure the volume of load over a given period, which has been shown to be a critical role in regulating testosterone secretion after a given RE program (^12,39). Because of the duration of some of the exercises, it becomes impractical to equate for load between ST and traditional training. For example, one does not perform the farmers walk, keg carry, and chain drag for 1 repetition. Therefore, it is not possible to assess 1RM strength, which is a limitation when properly loading the athlete using 1RM percentage, charts (i.e., 70–80% 1RM) and makes it difficult to equate loading across different protocols.

In conclusion, the findings of our study can be useful for strength practitioners aiming to develop effective alternate ways to induce endogenous testosterone, thus creating an optimal hormonal milieu for strength and hypertrophy adaptation. Although the generalizability of our results may be limited to highly trained subjects, the authors' purpose was to design a protocol that would elicit the highest endocrine response. Therefore, only this type of population was expected to complete such an intense protocol. Future research should focus on examining the endocrine response from various ST protocols with larger, stronger, and more experienced strongman athletes. This may provide a better understanding of the hormonal response from ST to subjects who are more familiar with the exercises.

Practical Applications

Strongman training has experienced steady growth and use in major collegiate programs, professional settings, and research interests (^{8,9,11,22,31,43,44}). This study supports that ST can be effective in eliciting a significant neuroendocrine response comparable with traditional hypertrophy training. Our findings have practical implications regarding exercise prescription for designing resistance training programs stimulating testosterone release. Strongman training is unique and exciting and provides an effective alternative to traditional resistance training; however, it does incorporate unorthodox movement patterns, which may blunt the testosterone response because of the perceived risk of injury. Once technique and familiarity of the movement patterns are achieved, coaches can manipulate the specific combinations of rest intervals, loading, and volume toward their desired training goals.

Acknowledgments

The authors would like to thank Gridiron, Inc., for providing the resources and equipment necessary to make this study possible. The authors also express their appreciation to all the subjects for their time and effort. This research was supported by the Hofstra University School of Education Dean's Research Grant. The authors disclose professional relationships with companies or manufacturers who will benefit from the results of the present study. Results of the present study do not constitute endorsement of the product by the authors or the National Strength and Conditioning Association.