Kettlebell training has increased in popularity during the last decade. One of the most frequently used kettlebell exercises is the kettlebell swing, which can be performed with 1 or 2 arms. In this study, the movement is primarily created by the hip extensors and transferred through the core and out to the arms and kettlebell. The core muscles have a key role in stabilizing the trunk allowing for effective transfer of forces from the hip to the arms holding the kettlebell (9).
This particular kettlebell exercise has shown beneficial effects. After conducting kettlebell swing exercise 3 times per week for 8 weeks, Jay et al. (6) found that the kettlebell-group increased their trunk extensor strength and lowered their pain intensity in neck, shoulders, and lower back significantly compared with the control group, which did not train. Another publication from the same authors with an identical intervention reported that the kettlebell swing–training group improved postural reactions to perturbation compared with the control group and that they were the only ones to improve their jump height during the 8-week period (7).
Different approaches have recently been proposed to increase core muscle activation during traditional resistance exercises. One of those is to perform exercises unilaterally instead of bilaterally (1,2,8,13,14). Theoretically, when performing an exercise unilaterally, the contralateral side has to increase muscle activation to avoid postural sway due to the increased torque in the trunk created by the external load. For example, Behm et al. (2) performed unilateral chest press and shoulder press exercises with both left and right arms, measuring the electromyographic (EMG) activity at the right side of the trunk. In general, they found the highest trunk muscle activation on the contralateral side when performing the resistance exercise unilaterally with the contralateral arm, second highest with bilateral execution, and the lowest activation was observed on the ipsilateral side with unilateral execution.
To our knowledge, only 2 studies have measured muscle activation by surface EMG in the kettlebell swing (12,16), and only 1 of these measured the trunk muscle activation. McGill and Marshall (12) compared muscle activation during bilateral kettlebell swing, kettlebell swing with kime (brief muscular pulsing at the top of the swing), kettlebell swing to kettlebell snatch, and kettlebell carrying. Similar trunk muscle activations were observed during these exercises for most trunk muscles, with the exception of greater activation in internal oblique when performing the swing with kime compared to the regular swing.
As we are not aware of studies comparing 1-armed and 2-armed kettlebell exercises with respect to trunk muscle activation, the aim of this study was to compare the activation of the core muscles during 1-armed and 2-armed kettlebell swing. Based on results from strength exercise studies, we hypothesized greater activation on the contralateral side and reduced activation on the ipsilateral side, in the 1-armed compared with the 2-armed swing. Furthermore, we expected higher activation on the contralateral compared with the ipsilateral side during 1-armed kettlebell swing.
Experimental Approach to the Problem
A within-subject crossover design was used to assess neuromuscular activity of the lower and upper erector spinae, rectus abdominis, and external oblique at both sides of the trunk in 1-armed and 2-armed kettlebell swing. Before the experimental test, 4 familiarization sessions were performed in a period of 2 weeks with at least 2 days between each session. In the experimental test, the participants performed 10 consecutive repetitions with a submaximal intensity (16-kg kettlebell). The order of the unilateral and bilateral swings was randomized and counterbalanced. After completing the exercises, the participants estimated the intensity using the Borg rating of perceived exertion scale (6–20).
Sixteen healthy men (age, 25 ± 6 years; body mass, 80 ± 8 kg; stature, 180 ± 7 cm) with 7 ± 7 years of resistance training experience volunteered for the study. None of them were familiar with the kettlebell swing exercise. Inclusion criteria were no injuries or pain that could reduce their performance in any of the tests. The participants agreed to refrain from any resistance training 72 hours before testing. All participants were informed verbally and in writing of the procedures and possible risks of the tests and provided written consent before they were included in the study. Ethics approval was obtained from the regional research ethics committee and conformed to the latest revision of the Declaration of Helsinki. The study had the approval by the Sogn og Fjordane University College Review Board and all appropriate consent pursuant to law was obtained before the start of the study.
The familiarization sessions lasted for approximately 30 minutes and consisted of 8 sets of each swing exercise with 10–15 repetitions per set. The weight of the kettlebell progressed from 8 to 16 kg from the first to last session. All sessions were led by a certified kettlebell instructor. The instructor had to approve the participants' technique before they were allowed to perform the experimental test. On the experimental test, all participants conducted the same warm-up routine. After a general warm-up on a treadmill or cycle, the participant performed a warm-up set, in both exercises, consisting of 10 repetitions with a 12-kg kettlebell.
The 1-armed and 2-armed kettlebell swing exercises were performed in a randomized and counterbalanced order. In the 1-armed swing, the dominant arm was used. Participants placed their feet with a preferred distance between the feet, and the same placement was used in the subsequent test. The exercise started with the kettlebell in the upper position, with the help of a test-leader holding the kettlebell. On the participants' verbal signal, the test-leader dropped the kettlebell and the participant performed 11 consecutive repetitions in both swing exercises (Figures 1 and 2). The first repetition was used to adjust to ones preferred tempo and discarded from the analysis. The repetitions were performed in a self-selected, but controlled tempo. It was emphasized that the hip extensors should generate the motion in their hips, with knees slightly bent during the whole movement. Furthermore, exercises were performed with straight arms and a neutral position in their lower back throughout (Figure 1A). To ensure the same range of motion of every repetition, the kettlebell had to touch the test-leader's hand gently at the highest part of every repetition (Figures 1 and 2). This point corresponded to the height, in which the participant's arm(s) were parallel to the floor (Figure 1B). A 16-kg kettlebell was used in all tests (Sportsmaster Pro kettlebell; Sportsmaster, Nesbru, Norway). Two-to-3 minutes rest was given between each test.
A linear encoder (Ergotest Technology AS, Langesund, Norway; sampling frequency of 100 Hz) was used to control the swing time and range of motion of the swings and identify the beginning and end of the hip extension (the kettlebell moving upward) and hip flexion (kettlebell moving downward) in addition to the lower and upper phases of these movements. The lower and upper phases were defined from the kettlebells' range of motion. Independent of kettlebell moving upward or downward (hip extension or flexion), the lower phase was defined from the lowest kettlebell position during a swing to the midpoint of the trajectory of the kettlebell. The upper phase was defined from the midpoint to the highest position of the kettlebell during the repetition. The linear encoder was synchronized with the EMG recording system (MuscleLab 4020e; Ergotest Technology AS).
Before placing the gel-coated self-adhesive electrodes (Dri-Stick Silver circular sEMG Electrodes AE-131; NeuroDyne Medical Corp., Cambridge, MA, USA), the skin was shaved, abraded, and washed with alcohol. The electrodes (11-mm contact diameter and a 2-cm center-to-center distance) were placed along the presumed direction of the underlying muscle fiber on the lower and upper erector spinae, rectus abdominis, and external oblique according to the recommendations by SENIAM (5). The electrodes were placed on the ipsilateral and contralateral sides of the core. The ipsilateral and contralateral sides were defined from the 1-arm swing, but the terms are also used for the same side in the 2-armed swing. The raw EMG signal was captured analogously. Furthermore, to minimize interfering signals from the surrounding muscles, the raw EMG signal was bandpass amplified and filtered (fourth-order Butterworth filter) using a preamplifier located close to the sampling point. The preamplifier had a common mode rejection ratio of 100 dB, high cut frequency of 600 Hz, and low cut frequency of 8 Hz. The EMG signals were converted to root mean square (RMS) signals using a hardware circuit network (frequency response, 0–600 kHz; averaging constant, 100 milliseconds; total error, ±0.5%). Finally, the RMS-converted signal was sampled at 100 Hz continuously moving time window using a 16-bit A/D converter. Commercial software (MuscleLab V8.13; Ergotest Technology AS) was used to analyze the stored EMG data. The mean EMG amplitude obtained during the entire swing of the 10 repetitions (i.e. both hip extension and flexion) was used to calculate the RMS EMG of the whole set. In addition, 6 repetitions (repetition number 1, 2, 5, 6, 9, and 10) were divided into 4 phases, lower and upper phase of the hip extension and flexion. The mean EMG amplitude of each phase obtained during the 6 repetitions was used to calculate the RMS. Approximately 3 to 5 minutes after the experimental test, 2 maximal voluntary isometric contractions (MVCs) for both abdominal and lower back muscles were measured. For the abdominal muscles, an isometric sit-up was used. The participants sat with a 90° angle in both hip and knees holding on to a fastened rope. For the lower back muscles, a resisted isometric back extension in the Biering-Sorenson position was used (15). The participants were instructed to obtain their maximal force as quickly as possible. Each MVC lasted for a minimum of 3 seconds (10,11). The MVCs with the greatest EMG amplitudes were chosen for subsequent normalization of the kettlebell swing EMG data. Each muscle was normalized to the EMG amplitude obtained from the same muscle during the MVC.
A 2-way analysis of variance (ANOVA) with repeated measures was used to assess differences in the EMG activity for each of the muscles measured. The independent variables were exercise (1-armed and 2-armed kettlebell swing) and side (contralateral and ipsilateral). When a difference was detected by ANOVA, paired t-tests with a Bonferroni post hoc correction were applied to determine where the differences lay. Paired t-tests were used to check for systematic differences in swing time and range of motion between both conditions. Statistical analyses were performed with SPSS version 17.0 (SPSS, Inc., Chicago, IL, USA). All results are presented as mean ± SDs and Cohen's d effect size (ES). An effect size was considered small, medium, and large at 0.2, 0.5, and 0.8, respectively. Statistical significance was accepted at p ≤ 0.05.
Electromyographic data for the whole movement are presented in Figures 3–6, whereas the values of the different phases are presented in Tables 1 and 2. An example of the EMG signals from rectus abdominis and erector spinae during the different swings is shown in Figure 3.
When analyzing the average EMG activity of the whole set, there was an interaction between exercise and side in the EMG data of the upper erector spinae (F = 26.513; p < 0.001). Post hoc analysis for the ipsilateral side showed a higher activation when performing the 2-armed compared with the 1-armed swing (40.2 ± 12.2% of MVC vs. 36.0 ± 12.7%; p = 0.006; ES = 0.34). For the contralateral side, a 7% nonsignificant greater activation was observed for the 1-armed compared with the 2-armed swing (p = 0.060). When performing the 1-armed swing, the activation of the contralateral side was higher than for the ipsilateral side of the 1-armed and for the 2-armed swing (44.7 ± 12.3% vs. 36.0 ± 12.7%; p < 0.001; ES = 0.70 and 44.7 ± 12.3% vs. 40.2 ± 12.2%; p = 0.026; ES = 0.37, respectively). Furthermore, during the 1-armed swing, the ipsilateral side showed a lower EMG activity compared with the opposite side during 2-armed swing (36.0 ± 12.7% vs. 41.6 ± 10.5%; p < 0.001; ES = 0.49). There was no difference in activation between the sides during the 2-armed kettlebell swing (p = 0.494, Figure 4). When analyzing the different phases of the repetitions, no interaction between exercise and side was observed (F = 0.330–2.250; p = 0.156–0.575). However, there were main effects for side in both the lower phase of the hip extension (F = 7.210; p = 0.018) and the lower phase of the hip flexion (F = 10.504; p = 0.006), and also for exercise in the lower (F = 7.312; p = 0.017) and upper phase (F = 6.184; p = 0.026) of the hip extension. In the lower phase of the hip extension, post hoc analysis showed a 30% higher activation of the contralateral side when performing the 1-armed compared with the 2-armed swing (p = 0.028; ES = 0.53). Furthermore, there was a 41% higher activation of the contralateral compared with the ipsilateral side when performing the 1-armed swing (p = 0.009; ES = 0.70). In the upper phase of the hip extension, the 2-armed swing led to a 21% higher activation of the ipsilateral side compared with the 1-armed swing (p = 0.050; ES = 0.52). In the lower phase of the hip flexion, the contralateral side had a 21% higher activation compared with the ipsilateral side performing the 1-armed swing (p = 0.018; ES = 0.52). Finally, there was no main effect for side in the 2 upper phases (F = 0.077–1.010; p = 0.332–0.786) nor exercise (F = 0.218–3.923; p = 0.068–0.648, Tables 1 and 2) in both phases of the hip flexion.
Concerning the complete set, there was an interaction between exercise and side in the EMG data of rectus abdominis (F = 6.582; p = 0.022). Post hoc analysis showed a lower activation on the contralateral side during 1-armed swing compared with the ipsilateral side during both 1-armed and 2-armed swing (18.6 ± 12.6% vs. 31.9 ± 23.9%; p = 0.044; ES = 0.69 and 18.6 ± 12.6% vs. 35.7 ± 26.7%; p = 0.038; ES = 0.81). No significant differences between the other conditions were observed (p = 0.070–0.977, Figure 5), despite a nonsignificant 42% lower activation on the contralateral side during 1-armed swing compared with the contralateral side during 2-armed swing (p = 0.070). When analyzing the different phases, an interaction between side and exercise in the upper phase of the hip extension was detected (F = 7.376; p = 0.017), but not in the other phases (F = 0.006–2.045; p = 0.176–0.938). There was, however, a main effect for side in the lower phase of the hip extension (F = 4.897; p = 0.045). In the lower phase of the extension phase, post hoc analysis showed a 139% higher activation on the ipsilateral compared with the contralateral side when performing the 1-armed swing (p = 0.020; ES = 0.99). In the upper phase of the hip extension, there was a 38% higher activation on the contralateral side performing the swing with 2 arms compared to 1 arm (p = 0.021; ES = 0.51). On the ipsilateral side, a nonsignificant 29% greater activation during the 1-armed compared with the 2-armed swing was observed (p = 0.071). Concerning side, there was a 66% higher activation on the ipsilateral compared with the contralateral side during 1-armed swing (p = 0.023; ES = 0.72). There was no main effect for side in any of the phases during hip flexion (F = 0.330–0.963; p = 0.344–0.576) nor for exercise in the lower phase during hip extension or both phases during the flexion (F = 0.026–0.881; p = 0.319–0.874, Tables 1 and 2).
For the lower erector spinae and external oblique, there were no significant interactions (F = 0.002–2.860; p = 0.111–0.966) nor main effect for exercise or side (F = 0.002–4.346; p = 0.055–0.969, Figures 6 and 7) when analyzing for either the whole set or the different phases.
In the 1-armed swing compared to the 2-armed swing, the range of motion was 6.1 cm longer from the upper to the lower position (p < 0.007), and the participants used 0.8 seconds (p < 0.001) more to complete the 10 repetitions. The mean intensity (Borg's scale) of both exercises was 13 (somewhat hard) (SD ± 1).
The study showed for the first time that 1-armed kettlebell swing induced greater activation of the contralateral side of the upper erector spinae than that of the ipsilateral side and greater than during 2-armed swing. Interestingly, however, the activation of the contralateral side of rectus abdominis was substantially lower than that of the ipsilateral side, and also during 2-armed swing. The lower erector spinae or external oblique was similarly activated on both sides during both swing exercises.
The highest activation of the upper erector spinae was found on the contralateral side during 1-armed swing as hypothesized. A possible explanation could be the lower part of the movement where performing the kettlebell swing with 1 arm would allow the kettlebell to travel further through the legs before starting the hip extension, leading to a rotation of the spinae. In this study, the data from the linear encoder showed a difference between the conditions with the kettlebell traveling 6 centimeters further in the 1-arm swing. During the hip extension, the contralateral side of the upper erector spinae then would have to be more activated to rotate the spinae back and transfer the force from the hip extensors to the arm holding the kettlebell. This explanation is supported by the analysis of the different phases. The only difference between the exercises was found on the contralateral side in the lower phase of the hip extension where the rotation is expected to occur. Furthermore, in the 2 lower phases, there was a difference between both sides during the 1-armed swing with the contralateral side being superior. These findings seem plausible during a rotation because the torque on 1 side of the spine would have to exceed the other to create the rotation during acceleration or deceleration of the kettlebell. Another explanation proposed in strength exercise studies with similar results is increased activation of the contralateral side during unilateral execution to avoid lateral flexion (2,13). However, it is likely to be of less importance here than in the earlier studies, as in our experiment, the kettlebell was moving in the centerline of the body, whereas in the earlier, the external load was held lateral to the trunk. Nor, is this argument supported by the findings in the different phases because it would be expected to find differences on the contralateral side in all phases and not just the lower phase of the hip extension.
The results for the upper erector spinae extend findings from previous strength exercise studies, which found higher activation of the contralateral side when performing unilateral upper body exercises (2,13). Behm et al. (2) found higher activation of the contralateral side during unilateral execution of chest presses and shoulder presses, compared with the contralateral side during both unilateral and bilateral presses. Similar findings were recently reported by Saeterbakken and Fimland (13) who compared unilateral and bilateral executions of shoulder presses among well-trained men.
Concerning rectus abdominis activation, the contralateral side during 1-armed swing was lower compared with the ipsilateral side during both 1-armed and 2-armed swings, which was in contrast to our hypothesis. Still, higher activation at the ipsilateral side during 1-armed swing could be related to the findings in the upper erector spinae. The increased activation on the contralateral side of the upper erector spinae could cause a demand for increased coactivation of the ipsilateral side of the rectus abdominis, as this could help stabilize and protect the trunk during explosive rotation in the 1-armed swing exercise. The findings in the lower phase of the hip extension support this argument with the ipsilateral activation being greater than the contralateral during the 1-armed swing and in contrast to the activation of the upper erector spinae in the same phase. This pattern of antagonist activation would most likely lead to an increased transfer of the forces from the hip extensors to the arms and kettlebell. In the upper phase of the hip extension, the rectus abdominis most likely acts as an antagonist and decelerates the movement. Therefore, in this part of the movement, it would not be each side alone, but the interplay of both sides that is of importance. This could explain the activation of the ipsilateral side during 1-armed swing being greater than the contralateral side during the same exercise and the statistical tendency to being greater compared with the ipsilateral side during 2-armed swing. Importantly, if we combine both sides in each exercise, the mean activation would be equal (15.3% of MVC).
When comparing our results for rectus abdominis with strength exercise studies, several have found increased activation on the contralateral side during unilateral execution (2,13,14). However, those studies used upper body exercises with the external load lateral to the truncus (shoulder presses or chest presses). This should lead to an increased stress on the contralateral side of the trunk to avoid a rotation and or lateral flexion. In contrast, the external load in our experiment is kept close to the centerline of the truncus.
In contrast to our hypothesis, there were no differences between exercise and side in the lower erector spinae when analyzing the whole set or different phases of the movement. Earlier studies have found differences in the lower back muscles when comparing unilateral and bilateral exercises (1,2,13). They have, however, used different location of the electrodes. We placed these electrodes in the lower parts of the lumbar region where the primary movements possible are flexion and extension. In general, the activation seemed higher during hip extension compared with the hip flexion. This may indicate that the lower erector spinae contributes as an extensor during the swing. Furthermore, because the lower back should be in a neutral position throughout parts of the movement, the erector spinae would also contribute to stabilize the spine. Because the external load is identical in both conditions, the demand for extension and stabilization would be similar, which could explain the lack of difference in muscle activation between both exercises.
Furthermore, the lack of significant difference in neuromuscular activity of the external oblique is also in contrast with our hypotheses and to earlier studies examining core muscle activation during unilateral and bilateral executions of strength exercises of the leg, shoulder, and chest muscles (1,13,14). These previous studies demonstrated an increased activation of the contralateral side when performing the exercise unilaterally instead of bilaterally. Common for these studies is the use of exercises, which challenges the balance in the frontal or horizontal plane (Bulgarian squat, unilateral shoulder press, and cable press), increasing the demand of the contralateral core muscles to avoid a lateral flexion or rotation of the truncus. In other words, the lack of differences between contralateral and ipsilateral activations could be explained by the fact that the kettlebell is moving in the centerline of the truncus, in the sagittal plane. Another possible explanation could be the load of the weights. Unlike our study, the studies mentioned above used maximal or close to maximal intensity, which could increase the demand of contralateral activation of the external oblique during unilateral execution of the exercises.
Several of the previous studies have only measured the contralateral side when performing unilateral resistance exercises and compared this with the activation obtained during bilateral execution (1,2,8,13). However, as is clearly illustrated in Figures 3 and 4, ipsilateral activation of erector spinae is substantially lower than that obtained during bilateral execution. Therefore, reporting muscle activation on both sides of the truncus gives a more in depth understanding of the core muscle activation during unilateral and bilateral training.
There are some limitations to our study. Only healthy men with no earlier kettlebell experience were recruited, and the results can therefore not necessarily be generalized to other populations. In addition, because the load was submaximal, the same conclusions cannot be drawn to execution of the kettlebell swing with maximal or near maximal loading. However, as we recruited inexperienced participants who performed the exercises with submaximal loading, the findings are more relevant for the general population. Also, surface EMG provides only an estimate of the muscle activation, and there is a possible risk of crosstalk from neighboring muscles (4). Furthermore, there are additional methodological limitations with dynamic EMG recordings (3).
In conclusion, performing the kettlebell swing with 1 arm provides higher muscle activation on the contralateral side of the upper erector spinae and on the ipsilateral side of the rectus abdominis, compared with the opposite sides, and these activation levels were also greater than those obtained during 2-armed swing. The differences observed for these muscles in the lower and upper phases suggest that the muscles have different tasks in the different regions. No differences for the external oblique and lower erector spinae were observed.
The kettlebell swing is a popular exercise, which can be performed with 1 or 2 arms. However, few studies have investigated the kettlebell swing, giving us limited research-based knowledge of any advantageous way to perform the swing. This study suggests that it could be useful for fitness practitioners, when executing the kettlebell swing with submaximal load, to perform 1-armed swing instead of 2-armed swing as this provides higher muscle activation, but as the muscle activation gained on 1 side of these core muscles seems to be lost on the other side, an equal number of sets with each arm should be performed to provide the same stimuli for the core muscles on both sides of the trunk. Furthermore, our results may benefit and improve performance of athletes competing in a sport with a demand of transferring forces from the legs to the arms, such as javelin, shot put, and handball. However, because the conclusions of this study are based on submaximal loads, they can also apply to the recreational trained or the patient with a desire or need to increase their core muscle strength.
The authors thank the participants for their positivity and participation in the study. This study was conducted without any funding from companies or manufacturers or outside organizations.
1. Andersen V, Fimland MS, Brennset O, Haslestad LR, Lundteigen MS, Skalleberg K, Saeterbakken AH. Muscle activation and strength in squat and bulgarian squat on stable and unstable surface. Int J Sports Med 35: 1196–1202, 2014.
2. Behm DG, Leonard AM, Young WB, Bonsey WA, MacKinnon SN. Trunk
muscle electromyographic activity with unstable and unilateral exercises. J Strength Cond Res 19: 193–201, 2005.
3. Farina D. Interpretation of the surface electromyogram in dynamic contractions. Exerc Sport Sci Rev 34: 121–127, 2006.
4. Farina D, Merletti R, Enoka RM. The extraction of neural strategies from the surface EMG. J Appl Physiol (1985) 96: 1486–1495, 2004.
5. Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 10: 361–374, 2000.
6. Jay K, Frisch D, Hansen K, Zebis MK, Andersen CH, Mortensen OS, Andersen LL. Kettlebell training for musculoskeletal and cardiovascular health: A randomized controlled trial. Scand J Work Environ Health 37: 196–203, 2011.
7. Jay K, Jakobsen MD, Sundstrup E, Skotte JH, Jorgensen MB, Andersen CH, Pedersen MT, Andersen LL. Effects of kettlebell training on postural coordination and jump performance: A randomized controlled trial. J Strength Cond Res 27: 1202–1209, 2013.
8. Jones MT, Ambegaonkar JP, Nindl BC, Smith JA, Headley SA. Effects of unilateral and bilateral lower-body heavy resistance exercise on muscle activity and testosterone responses. J Strength Cond Res 26: 1094–1100, 2012.
9. Kibler WB, Press J, Sciascia A. The role of core stability in athletic function. Sports Med 36: 189–198, 2006.
10. Marshall PW, McEwen M, Robbins DW. Strength and neuromuscular adaptation following one, four, and eight sets of high intensity resistance exercise in trained males. Eur J Appl Physiol 111: 3007–3016, 2011.
11. McBride JM, Cormie P, Deane R. Isometric squat force output and muscle activity in stable and unstable conditions. J Strength Cond Res 20: 915–918, 2006.
12. McGill SM, Marshall LW. Kettlebell swing, snatch, and bottoms-up carry: Back and hip muscle activation, motion, and low back loads. J Strength Cond Res 26: 16–27, 2012.
13. Saeterbakken AH, Fimland MS. Muscle activity of the core during bilateral, unilateral, seated and standing resistance exercise. Eur J Appl Physiol 112: 1671–1678, 2012.
14. Santana JC, Vera-Garcia FJ, McGill SM. A kinetic and electromyographic comparison of the standing cable press and bench press. J Strength Cond Res 21: 1271–1277, 2007.
15. Tse MA, McManus AM, Masters RS. Development and validation of a core endurance intervention program: Implications for performance in college-age rowers. J Strength Cond Res 19: 547–552, 2005.
16. Zebis MK, Skotte J, Andersen CH, Mortensen P, Petersen HH, Viskaer TC, Jensen TL, Bencke J, Andersen LL. Kettlebell swing targets semitendinosus and supine leg curl targets biceps femoris: An EMG study with rehabilitation implications. Br J Sports Med 47: 1192–1198, 2013.