Secondary Logo

Journal Logo

Original Research

Electromyographic Response of Global Abdominal Stabilizers in Response to Stable- and Unstable-Base Isometric Exercise

Atkins, Stephen J.; Bentley, Ian; Brooks, Darrell; Burrows, Mark P.; Hurst, Howard T.; Sinclair, Jonathan K.

Author Information
Journal of Strength and Conditioning Research: June 2015 - Volume 29 - Issue 6 - p 1609-1615
doi: 10.1519/JSC.0000000000000795
  • Free



The importance of a strong core, and allied postural stability, is of great interest to trainers and conditioners. Static and dynamic core stability are key components of functional posture, with neuromuscular control a key corollary of postural alignment (13). Neuromuscular control of the core is based on reactive (feedback) control mechanisms (4,25). Sports such as swimming require the maintenance of a streamlined position, either in the prone or supine orientation, in an unstable medium. To date, there is little research available to support the role of a strong core in enhancing the performance of elite level swimmers, particularly when focussed on land-based drills. Given the innate “instability” of the swimmer when in water, a greater emphasis on offsetting this perturbation is of great interest to coaches and conditioners.

Somatosensory feedback, relating to the orientation of key body segments, enables continuing adjustment of overall position. This adjustment is especially important in situations involving unstable bases and perturbation. Traditionally, exercises designed to activate and train core muscles have been undertaken using stable bases. More recently, unstable-base exercises using Swiss balls, Bosu balls, or suspension straps have allowed a greater level of instability to be introduced in core training. Such instability may reflect the relatively unstable nature of the body when performing in water. To date, there is very limited information on the effectiveness of these newer unstable-base training techniques, such as suspension training, when compared with more traditional forms of exercise using a stable base. This is particularly evident when reviewing existing literature relating to youth athletes.

In younger athletes, the importance of postural control has been associated with integrative neuromuscular training as part of a broader strength and conditioning program (17). Enhancing core endurance and strength would seem to be a vital component of performance-related fitness and has been proposed to form a key element in balance and postural stability (4). This association between core stability and enhanced mobility of the limbs has long been promoted (18). Core muscles can be defined as global and local units (3,18). These units are dependent on the position of the muscle in the core region and attachment sites. Local units support vulnerable spinal structures and allow more forceful actions, such as performance training, to be undertaken by the global elements of the core (3). Measurement of muscle activation in global muscles is often easier due to their surface proximity. Enhanced function and activation of global musculature, including rectus abdominus (RA), external obliques (EO), and erector spinae (ES) have been shown to improve stability and muscle coordination and reduce injury risk (5,22). Sadly, little research has sought to determine the effect of different land-based core strengthening exercises to enhance the neuromuscular response of global abdominal stabilizers.

Training of the core musculature is traditionally undertaken using stable isometric bracing techniques, such as the “plank” exercise. Recently, unstable-base techniques, such as Swiss ball and suspension straps, become very popular in training and conditioning contexts. Unstable-base training can induce changes in the resistance vector, a changing base of support throughout the movement and the effect of a fixed point pendulum (9). To date, there is a very limited evidence base relating to the use of unstable-base techniques, particularly regarding peer reviewed articles. Schoffenstall et al. (19) reported no significant difference in activation of global core muscles when comparing suspension training with other abdominal training exercises. This study used an isometric contraction similar to those undertaken in classic bracing exercises, such as plank work. Information on the use of unstable-base activities in youth athletes is absent from the literature.

The purpose of this study was to determine neuromuscular activation of global core musculature in 3 different conditions, in a population of elite youth swimmers. Core bracing is an important corollary of postural stability and has value in optimizing body alignment in swimming. Land training enhances swim performance (6) with core activation techniques often used in athletic settings. Good posture can decrease drag, improve catch and fixing phase of the stroke, encourage good body roll for more distance per stroke, and increase production and transferring of force from the arms to the upper core area to increase power output (14). It is hypothesized that there will be no differences in activation between anterior, posterior, and lateral core musculature when using stable- and unstable-base techniques. Results from this study will allow coaches and conditioners to assess the efficacy of 3 different forms of land-based core stability exercise, related to activation of superficial abdominal musculature.


Experimental Approach to the Problem

This experiment was designed to test a hypothesis that 3 different static stability exercises, using both stable and unstable bases, would activate core muscles equally. The key dependent variables were the root mean square amplitudes measured using surface electromyography (sEMG), taken from the RA, EO, and ES sites. Independent variables were the 3 exercises used; static “plank,” Swiss ball “plank,” and suspension “plank.” Surface EMG in RA, EO, and ES was investigated while performing these static exercises. Maximal voluntary contraction (MVC) amplitude was determined using isometric activation manipulation exercises, designed to stimulate the RA, EO, and ES musculature. From these MVC measurements, the percentage of peak amplitude achieved within each exercise condition will be determined.


Eighteen amateur, elite-level, age-group male swimmers gave written informed consent to take part in this study. A written informed consent document was presented to all subjects. For all subjects under the age of 18 years, signatures were obtained from both athletes and parent(s)/guardian(s) before enrollment in the study. The research was approved by the School of Psychology Ethics Committee, at the University of Central Lancashire, and operated in accordance with the principles of the Declaration of Helsinki. Testing was undertaken between January and March. All swimmers competed at county or national level. Descriptive characteristics of the group were age, 15.9 ± 2 years; stature, 166.7 ± 11.1 cm; body mass, 64.7 ± 13.1 kg. The age range of subjects was 13–19 years. Subjects competed at county to national level, in England, and trained on a minimum of 4 days per week. All participants had had no recent history musculoskeletal injury and were free from infections that would preclude engagement in stress testing. On the day of testing, subjects reported having abstained from caffeine and exercise in the previous 12 hours. An a priori power analysis calculation was made using the Hopkins method. A moderate effect size, and 80% power measure, revealed that 19 participants would be required for this study.


Surface electromyography assessment was undertaken using an 8 channel data logger allied to standard differential EMG surface electrodes (Biometrics Ltd.; DLK900, Cwmfelinfach, Gwent, United Kingdom). Electrical activity was obtained at 1,000 Hz, from the right side RA, EO, and ES sites, using bipolar electrodes with an interelectrode distance of 20 mm. All electrodes were placed in alignment with the muscle pennation on the bellies of the muscles in accordance with the SENIAM guidelines (11). The ground electrode was placed on the posterior portion of the left hand. The skin was prepared with abrasive paper and cleaned with ethanol wipes to reduce skin impedance.


Maximal voluntary contraction measurements from RA, EO, and ES muscles were determined using existing activation manipulation techniques (23). These techniques used a forma of static bracing to generate peak neuromuscular activation values in target musculature. Each participant underwent a normalization procedure on a separate day to obtain MVCs. The techniques for MVC determination were as follows. Upper trunk flexion was used to activate the RA. This took the form of an adapted sit up, with knees flexed at 45°. With the feet held in position and resistance placed across the trunk at a line superior to the xiphisternal joint, the participant was asked to forcefully flex the trunk. A lying side bend was used to determine upper trunk EO MVC. With the feet secured, participants were asked to bend upward while lying on their side. Resistance was placed at a level superior to the xiphisternal joint. An adapted lower trunk extension was performed for ES. Participants lay in the prone position, with feet secured by a second tester. The participant raised shoulders and trunk from the floor, with resistance placed at a level equivalent to the inferior border of the scapula. All MVC exercises were held for 10 seconds with participants encouraged to achieve maximal voluntary levels of contraction. To achieve this, verbal feedback was provided to ensure participants were able to isolate the required musculature. On a separate testing occasion, participants returned to the laboratory to undertake the 3 exercise conditions. With sEMG electrodes attached, participants completed the prone “plank” (Figure 1), Swiss ball “plank” (Figure 2), and suspension training (ST) “plank” (Figure 3) in a randomized order. Before testing, all participants were provided with a familiarization session outlining key coaching points relating to each exercise.

Figure 1:
Participant position when holding the static prone plank exercise.
Figure 2:
Participant position when holding the static prone plank exercise using a Swiss ball.
Figure 3:
Participant position when holding the static prone plank exercise using suspension straps.

The prone plank was undertaken using previous guidelines (15). To ensure internal consistency of technique, for each test, the angle of participants shoulder was kept consent at 45° flexion. Simple goniometry was used to ensure this consistency. The spine was observed to ensure that it remained in the neutral position. On a countdown command “3, 2, 1, Go,” the plank position was held by participants for 10 seconds (15), with the test repeated 3 times, with 3 minutes rest in between. The Swiss ball plank and ST plank exercises were undertaken using similar shoulder angle and neutral spine technique to the prone version. Similar familiarization, countdown, and duration were used for these exercises.

Surface Electromyography Signal Processing

Electromyography processing was performed using Biometrics v7.5 Analysis (Biometrics Ltd.; DLK900). The EMG analysis equipment was set to sample at 1,000·s−1 and trace sensitivity of 1 V. Raw sEMG analysis was performed using different methods for peak average amplitude over 100 m·s−1 to eliminate frequent potential anomalies (21). For peak amplitude, the raw sEMG data were rectified using the biometrics software. Averages were compiled, and the highest value for each muscle and group could be identified. The EMG signals from each muscle/condition were full-wave rectified and filtered using a 20-Hz zero-lag Butterworth low-pass fourth filter, thereby creating a linear envelope. Surface electromyography data from each muscle were normalized to the MVC for each muscle. Electromyography measures extracted were the mean normalized amplitude during the pedal cycle.

Statistical Analyses

After the determination of a descriptive profile, differences in type of training method and muscle activation were examined using a mixed factorial (3 × 3) analysis of variance with the level of significance set at p ≤ 0.05. Bonferroni post hoc analysis was used to assess differences between sEMG activation of the RA, EO, and ES when performing a prone plank using ST, Swiss ball, and the floor. This technique was also used to compare sEMG activation between MVC and the 3 exercise conditions. The criterion for significance was set at p ≤ 0.05. Effect size was determined using a simple Eta square calculation (η2). The Shapiro-Wilk statistic revealed a normal distribution for all variables. Statistical analysis was completed using the software packages SPSS version 20.0 (SPSS, Inc., Chicago, IL, USA).


In each test condition, MVC levels were significantly higher than those recorded during the focussed exercises (p = 0.001). The percentage of maximal voluntary contraction achieved for each exercise is outlined in Table 1.

Table 1:
Average percentage of maximal muscle activation achieved with the 3 different training methods when compared with a maximal voluntary contraction, for anterior, lateral and posterior musculature.

Suspension training exercises yielded the highest percentage of MVC for RA and ES, when compared with a prone plank or Swiss ball exercise. Rectus abdominus activation was significantly higher when using suspension than either the prone or Swiss ball plank (p = 0.04). For EO, the highest level of MVC was achieved using the prone plank technique vs. either suspension or Swiss ball (p = 0.002). There were no significant differences in ES activation between any of the exercises (p > 0.05). For all muscles and conditions, the Swiss ball plank elicited the lowest percentage MVC.

Figure 4 outlines the profile of differences between MVC and exercise test condition for the RA.

Figure 4:
Peak muscle activation of while performing prone plank using 3 different exercises (ST, static, Swiss ball) vs. MVC of the rectus abdominis obtained through exercise manipulation (all values are given in mean ± SD). *Significant difference between suspension training and static/Swiss ball exercise (p = 0.04). **Significant difference between MVC amplitude and all forms of exercise (p = 0.001). ST, Suspension training; MVC, maximal voluntary contraction.

Figures 5 and 6 reveal the amplitude differences between EO and ES, related to MVC, for each exercise.

Figure 5:
Peak muscle activation of while performing prone plank using 3 different exercises (ST, static, Swiss ball) vs. MVC of the external obliques obtained through exercise manipulation (all values are given in mean ± SD). *Significant difference between static plank and suspension/Swiss ball exercise (p = 0.002). **Significant difference between MVC amplitude and all forms of exercise (p = 0.001). ST, Suspension training; MVC, maximal voluntary contraction.
Figure 6:
Peak muscle activation of while performing prone plank using 3 different exercises (ST, static, Swiss ball) vs. MVC of the erector spinae obtained through exercise manipulation (all values are given in mean ± SD). *Significant difference between MVC amplitude and all forms of exercise (p = 0.001). ST, Suspension training; MVC, maximal voluntary contraction.


The aim of this study was to investigate, using surface electromyography (sEMG), the neuromuscular activation of the anterior, posterior, and lateral global core muscles while performing a static straight arm prone plank exercise on stable and unstable bases. Our hypothesis proposed that there would be no differences in muscle activation between global core musculature, irrespective of stable vs. unstable bases being used. All testing was undertaken using elite age-group swimmers for whom strong core stability is an important corollary of good position in water.

Overall, our results showed that using the unstable suspension training method, when performing a prone plank, elicited the highest peak muscle activation of anterior musculature (RA). This is compared with the use of a Swiss ball or more traditional plank exercise on the floor. However, the traditional plank exercise did stimulate greater activation of the lateral musculature (EO). There was limited activation of the posterior musculature (ES) within any exercise condition. As expected, maximal muscle activation using static bracing techniques was significantly higher than in any of the exercise conditions. This confirms previous findings (23,2). Peak amplitude measurements obtained in the investigation must be used with caution, however, as the manipulation exercises undertaken to obtain these have been shown to only give approximately 80% true representation of MVC (23).

The use of unstable-base exercises, such as Swiss ball and suspension training, gives rise to an inherently greater level of instability when performing traditional exercises such as a prone plank. Such perturbation was emphasized through the higher activation of anterior musculature when using suspension training. Of the 3 methods chosen for this study, suspension elicits a far greater level of instability and can be considered effective in enhancing bracing of anterior core musculature. However, perturbation when using a Swiss ball did provide similar responses when compared with the more traditional plank exercise.

Of great, interest to this research team was the seemingly poor activation of both lateral and posterior musculature, most notably when compared with anterior musculature. Increased, and balanced, activation of the posterior musculature has been proposed to ensure good posture and muscle performance (7). Our results show that for all conditions, a noticeable imbalance between posterior and anterior core muscle activation was evident. This was surprising, given our prediction that unstable-base exercises would facilitate similar levels of posterior activation. Posterior activation is vital in developing overall postural stability (24), with the complete core proposed to work together to gain overall postural stability (1,10). The seeming inability of both stable- and unstable-base static exercises to engage posterior musculature presents conditioners with a challenge. In swimmers, power output of the limbs is a key performance indicator (16). These levers need to be attached to a strong core, enabling force to be expressed on the water. Despite a very limited research base, the potential for core strength training affecting a better “in-water” body position, thereby reducing drag (20), should warrant further investigation.

We report the greatest activation of superficial musculature to occur anteriorly. Cross talk between the superficial anterior musculature and underlying transverse abdominis has been proposed to occur during the prone plank exercise (12). The magnitude of transverse abdominis activation is eminently difficult to assess, as with all deep lying musculature. Given the importance of activating these deep lying muscles in the performance of plank exercises (8), further studies may wish to focus on invasive EMG, enabling the deep/superficial cross talk relationship to be fully explored.

From the practical application perspective, the use of a prone plank, in swimming-related training, requires further assessment. This is irrespective of the stable- or unstable-base debate. The prone plank is performed in a “swim-like” position, but it does not simulate water-based activities, particularly instability or neuromuscular characteristics. In contrast, exercises using suspension training, which can be adapted to include the limbs on an unstable base, may be more specific to actual swimming posture; more so than Swiss ball training where the hand opposition is a relatively well controlled. This also reflects the unstable base associated with swim performance. Regarding training intervention, open kinetic chain exercise, such as suspension training, may exaggerate the recruitment of the primary core stabilizers, notably the anterior muscles. With underlying cross talk in operation, this would act as a useful adjunct to activating both superficial and deep lying musculature.

In conclusion, swimmers require excellent postural alignment, in water, to limit drag. Improvements in core strength are proposed as one mechanism to achieve this. It is vital that appropriate core activation techniques are practiced and trained to ensure optimal strength development. Our results suggest that unstable-base training, using suspension techniques, could be a more effective strategy for meeting these aims. This is particularly related to the anterior abdominal musculature. Further work is required to ascertain how best to engage lateral and posterior core musculature and to determine engagement of the limbs using suspension techniques.

Practical Applications

To best replicate the open kinetic chain demands of swimming, coaches and athletes may wish to use unstable-base training methods such as suspension straps. These are readily available and make for a very useful alternative to more traditional land-based core strengthening exercises, such as plank work. Traditional land-based exercise, such as the static plank, do not engage the limbs effectively, thereby having limited replication of the demands of swimming. Similarly, Swiss ball plank work does not activate core musculature as readily as when using suspension training. A stable and strong core, trained using unstable methods, would better reflect the demands of swimming. Suspension training does provide a more intensive activation of the anterior musculature, but the lateral and posterior musculature requires a greater level of stimulus. Coaches and conditioners would be advised to investigate more appropriate mechanisms of activating these muscles, during conditions of isometric bracing. This would include a focus on recruitment techniques and maintenance of tension during isometric activation, incorporating the limbs.


1. Arokoski JP, Vaito T, Airaksinen O, Kankaanpaam M. Back and abdominal muscle function during stabilization exercises. Arch Phys Med Rehabil 82: 1089–1098, 2001.
2. Behm DG, Leonard AM, Young WB, Bonsey WAC, Mackinnon SN. Trunk muscle electromyographic activity with unstable and unilateral exercises. J Strength Cond Res 19: 193–201, 2005.
3. Bergmark A. Stability of the lumbar spine: A study in mechanical engineering. Acta Orthop Scand 230: 20–24, 1989.
4. Bird SP, Stuart W. Integrating balance and postural stability exercises into the functional warm-up for youth athletes. Strength Cond J 34(3): 73–79, 2012.
5. Cissik JM. Programming abdominal training, Part 1. Strength Cond J 24(1): 9–15, 2002.
6. Cronin J, Jones J, Frost D. The relationship between dry-land power measures and tumble turn velocity in elite swimmers. J Swim Res 17: 17–23, 2007.
7. Ershad N, Kahrizi S, Abadi MF, Zadeh SF. Evaluation of trunk muscle activity in chronic low back pain patients and healthy individuals during holding loads. J Back Musculoskelet Rehabil 22: 165, 2009.
8. Faries MD, Greenwood M. Core Training: Stabilizing the confusion. Strength Cond J 29(2): 10, 2007.
9. Fitness Anywhere. TRX Suspension Training Group Course. User's Guide. San Francisco, CA: Fitness Anywhere, Inc, 2010.
10. Gamble P. An integrated approach to training core stability. Strength Cond J 29(1): 58–68, 2007.
11. Hermens HJ, Freriks B, Merletti R, Hägg GG, Stegeman D, Blok J, Rau C. European Recommendations for Surface Electromyography (SENIAM): CD-ROM. Enschede, the Netherlands: Roessingh Research and Development, 1999.
12. Hildenbrand K, Noble L. Abdominal muscle activity while performing trunk-flexion exercises using the Ab Roller, ABslide, FitBall, and conventionally performed trunk curls. J Athl Train 39: 37–43, 2004.
13. Hori N, Newton RU, Nosaka K, Stone MH. Weightlifting exercise enhance athletic performance that requires high load speed strength. Strength Cond J 27(4): 50–55, 2005.
14. Kluemper M, Uhl TL, Hazelrigg H. Effect of stretching and strengthening shoulder muscles on forward shoulder posture in competitive swimmers. J Sport Rehabil 15: 58, 2006.
15. Lehman GJ, McGill SM. Quantification of the differences in electromyographic activity magnitude between the upper and lower portions of the rectus abdominis muscle during selected trunk exercises. Phys Ther 8: 1096–1101, 2001.
16. Maglischo EW. Swimming Fastest. Champaign, IL: Human Kinetics, 2003.
17. Myer GD, Faigenbaum AD, Chu D, Falkel J, Ford KR, Best TM, Hewett TE. Integrative training for children and adolescents: Techniques and practices for reducing sports-related injuries and enhancing athletic performance. Phys Sportsmed 39: 74–84, 2011.
18. Nelson N. Diaphragmatic breathing: The foundation of core stability. Strength Cond J 34(5): 34–40, 2012.
19. Schoffenstall J, Titcomb DA, Kilbourne BF. Electromyographic response of the abdominal musculature to varying abdominal exercise. J Strength Cond Res 24: 3422–3426, 2010.
20. Scibek JS. The Effect of Core Stabilization Training on Functional Performance in Swimming. Master's Thesis. Chapel Hill, NC: University of North Carolina, 1990.
21. Stanton R, Reaburn PR, Humphries B. The effect of short-term Swiss ball training on core stability and running economy. J Strength Cond Res 18: 522–528, 2004.
22. Tyson A. Lumbar stabilization. Strength Cond J 21(5): 17–18, 1999.
23. Vera-Garcia F, Grenier SG, McGill SM. Abdominal muscle response during curl-ups on both stable and labile surfaces. Phys Ther 80: 564–569, 2000.
24. Willardson JM. Core stability training: Applications to sports conditioning program. J Strength Cond Res 21: 979–985, 2007.
25. Zazulak BT, Hewett TE, Reeves NP, Goldberg B, Cholewicki J. Deficits in neuromuscular control of the trunk predict knee injury risk. Am J Sports Med 35: 1123–1130, 2007.

core; sEMG; MVC; abdominals; suspension

Copyright © 2015 by the National Strength & Conditioning Association.