Core training is an essential aspect of conditioning and rehabilitation programs in an attempt to promote lower back health, improve general stabilization, and maximize force generation during athletic performances (1,3,10,21). Although multiple definitions exist of what the core entails; for the purposes of this investigation, the core will be referred to as the most outer layer of superficial musculature of the trunk (i.e., rectus abdominis [RA], external obliques [EO], lumbo-sacral erector spinae group [LSES], and rectus femoris [RF]). Superficial musculature of the trunk is primarily responsible for spinal movements, such as flexion, extension, postural stability, and resisting disruptions of a neutral spinal column (e.g., rotation, torsion) (10,11,19). Chronic training of the core musculature has been linked to improved lower-back health, increased athletic performance, and decreased incidence of fall-related injuries (1,11,19).
An emerging trend in core training is the utilization of instability devices (e.g., Swiss ball, suspension device, BOSU balls, etc.). These devices are primarily used with the intent of increasing the intensity of traditional exercises by providing further muscular demands of stability and balance. This is often accomplished through a reduction in stable surface area and contact points to increase core musculature effort to maintain balance and spinal stability (17,21). Although decreasing surface contact area and stabilization does elicit a greater need for balance (13), manufacturers claim that greater muscular recruitment, particularly in the core, will be needed to match these demands. This may result in increasing electromyographic (EMG) activity, muscle fiber recruitment, and improved core endurance and strength that may benefit individuals during rehabilitation, sports performance, and maintenance of general health and wellness (10,13,17,21).
Previous research supports the claims that unstable surface training elicits a greater response of the core musculature during traditional movements (2,6,12,13,16–18). For instance, isometric planks performed upon suspension devices and Swiss balls have been shown to increase core activity when compared to the traditional stable method (2,11,17). However, core exercises that involve dynamic movements upon instability devices have conflicting results. Swiss ball curl-ups have been reported to produce significantly less activation in the RA (6) and EO (14) when compared to the stable movement. Additionally, Saeterbakken et al. (14) showed no differences in RA activation when curl-ups were performed on a BOSU ball. Another popular exercise, the pike (PK), has become an important exercise because of its unique combination of dynamic and isometric contractions needed to properly perform the movement. While executing the PK, the spinal column is held in a neutral position (i.e., isometric contraction) while the individual also flexes at the hips (i.e., dynamic contraction). This exercise has been previously shown to elicit increased levels of EMG activity in the core musculature as compared to common abdominal exercises (i.e., crunch and bent-knee sit-up) (7).
With an increasing growth and anecdotal claims of increased core stability, limited data exist when cross-comparing multiple commercial instability devices. Additionally, the PK has only been examined by Escamilla et al. (7) with the use of one instability device (i.e., Swiss ball). Thus, the purpose of this investigation was to compare muscular activation, via EMG, between the RA, EO, RF, and LSES during pike variations performed with and without instability devices. Based upon previous findings, the authors hypothesized that the instability devices would elicit greater activation of the examined muscles as compared to the stable movement.
Experimental Approach to the Problem
A repeated measures design was used to determine the EMG differences during an abdominal exercise performed on stable and unstable surfaces. Twenty subjects completed 5 pike variations during 1 visit to the laboratory. A familiarization period was held a day or 2 before the testing in which subjects were instructed on proper technique and given adequate time to accustom themselves with each instability device. Electromyographic activity was measured via surface electrodes, whereas subjects performed 5 repetitions of each pike variation (i.e., traditional pike [PK], BOSU pike [BOSU], suspension training device pike [ST], Swiss ball pike [SB], and Core Coaster pike [CC]). The independent variables within the current study were the pike variations; whereas the dependent variables were the EMG values, presented in %MVC, obtained for each of the muscles examined (i.e., RA, EO, RF, and LSES). As a supplement to the reader, Table 1 provides a summary of the abbreviations used within the current study.
All subjects were recruited via flyers and word of mouth as a nonprobability sampling. Twenty physically active men (n = 10; mean age = 25.9 ± 5.61 years; mean height = 175.2 ± 8.5 cm; mean weight = 81.3 ± 6.9 kg) and women (n = 10; mean age = 22.8 ± 1.8 years; mean height = 166.1 ± 8.5 cm; mean weight = 63.0 ± 10.4 kg) participated in this investigation. All subjects were currently active in resistance-training programs of at least 3 sessions per week with a minimum duration of 30 minutes. Participants completed health history and medical questionnaires before testing to ensure that they were free from cardiorespiratory, musculoskeletal, metabolic, and neurological disorders. Subjects with any prior injuries that would otherwise affect muscular activation were excluded from data collection. Subjects were asked to report to the Human Performance Laboratory for 1 visit. Participants were asked to come to the laboratory having refrained from heavy intensity exercise 24 hours before testing. Upon arrival, participants reviewed and signed the informed consent document and completed a 24-hour history questionnaire to verify adherence to pretest instructions. This investigation was approved by the University's Institutional Review Board.
All EMG values were collected using a BIOPAC MP150 BioNomadix Wireless Physiology Monitoring System with a sampling rate of 1.0 kHz. Data were analyzed using Acqknowledge 4.2 software (BIOPAC System, Inc., Goleta, CA). The EMG signals were bandpass-filtered at a 20–400-Hz frequency while using a fourth order Butterworth filter. The root mean square EMG signals were recorded throughout each exercise; however, only repetitions 3 through 5 were used for data analysis. The first 2 repetitions were not used because of the low reliability of EMG activity seen in previous studies between multiple repetitions of an exercise (5). Mean RMS data were then normalized to the maximal voluntary contraction and reported as %MVC.
Before electrode placement (Biopac EL504 disposable Ag-AgCl), participant skin sites were prepped for application through shaving, exfoliation, and alcohol cleansing to reduce impedance from dead surface tissue and oils. The electrode placement chosen for this investigation were consistent with Cram and Kasman (5). Electrodes for the RA were placed 2 cm to the right of the umbilicus and 3 cm apart (vertically) directly over the muscle fibers. The EO electrodes were placed 15 cm lateral to the umbilicus, halfway between the iliac crest and the bottom of the ribs at a slightly oblique angle (i.e., 25 degrees). The LSES electrodes were vertically placed 2 cm parallel from the L-3 vertebrae at a distance of 2 cm apart. The electrode placement for the RF was positioned vertically near the midline of the thigh, halfway between the anterior superior iliac spine and proximal border of the patella. A ground surface electrode was placed directly over the right and left anterior superior iliac spine.
Maximal Voluntary Contractions
Maximal voluntary contractions (MVC) were collected postelectrode placement to normalize all EMG data. Each MVC was performed 3 times per each muscle group for 6 seconds a piece. The middle 2 seconds of each contraction were then averaged to obtain the reference value (i.e., MVC). To obtain the MVC for the RA, subjects performed an isometric crunch against a matched resistance (i.e., resistance forceful enough to elicit an isometric contraction from the participant). To obtain MVC of the RF, subjects were to lie down on their back and attempted to flex the hip against a matched resistance. The EO MVC was captured through a matched resistance side crunch. Lastly, the MVC for the LSES was performed with the participant in a prone lying position. An isometric (i.e., matched resistance) back extension was then performed. Once all MVC's were collected, the exercise trials were performed.
After completion of the MVCs, the 5 exercises performed in a randomized order are as follows: pike (PK), pike with a BOSU ball (BOSU), pike with a suspension training device (ST), pike with a Swiss ball (SB), and pike with Core Coaster (CC). All subjects were taught proper technique and familiarized with each exercise before data collection. If a subject was not able to maintain proper form as instructed, then all data were omitted from the analysis process. Each variation of the pike was performed at a rate of 4 seconds per repetition using a metronome set at 60 beats per minute (i.e., 2 seconds eccentric, 2 seconds concentric) and repeated for a total of 5 repetitions. During data collection, each subject was allowed a 3-minute rest between each exercise to prevent fatigue of the trunk musculature. The exercises were randomized for each participant, thereby eliminating fatigue error. The proper technique of each exercise used in this study is as follows:
- Pike (PK) (Figure 1): Subjects were instructed to assume a standard push-up position on an exercise mat with their arms fully extended and hands on the ground directly beneath the shoulders. The feet were placed together with only the toes in contact with the ground. Subjects were then instructed to “pike” by flexing at the hips slowly and under control until a 90o angle had been formed between the shoulders, hips, and legs. The subjects were told to maintain a rigid torso, neutral head and spine, and extended legs position throughout each exercise.
- Pike on BOSU ball (BOSU) (Figure 2): This variation called for the subjects to assume a push-up position with the arms fully extended and hands positioned directly beneath the shoulders on the ground and feet placed together upon the flat side of a BOSU ball. Subjects then performed the pike until the 90o angle had been achieved and slowly returned to the starting position.
- Pike on Swiss ball (SB) (Figure 3): Subjects were to assume the above-stated position, but with the feet placed upon the Swiss ball. Next, subjects performed the pike by flexing at the hips and pulling the Swiss ball toward the upper trunk until a 90o angle was reached.
- Pike with suspension training device (ST) (Figure 4): Before performing this exercise, a suspension device (TRX Suspension Trainer) was secured overhead to a door frame. The handles were placed approximately 20–30 cm inches above the ground as to create a horizontal positioning for each subject. The subjects then assumed a push-up position with each foot placed within the suspension device cradles. Following this, the subject performed the pike as previously described.
- Pike on Core Coaster (CC) (Figure 5): Participants performed the pike with each foot placed upon a Core Coaster device using the same techniques as described prior.
All subjects successfully completed the trials and no data were omitted from analysis. Data analysis was performed using SPSS/PASW Statistics version 22.0 (Somers, NY). Means and standard deviations were calculated for the %MVC values of the RA, EO, LSES, and RF. Repeated measures analysis of variance (ANOVA) was used to determine if the normalized value for each muscle group was significantly different across the varying exercises. A priori statistical significance was set to a value of p ≤ 0.05. A Bonferroni post hoc was used for a follow-up procedure. A Cohen's d statistic (4) was calculated as the effect size of the differences in %MVC values and Hopkin's scale of magnitude (9) was used where an effect size of 0–0.2 was considered trivial, 0.2–0.6 was small, 0.6–1.2 was moderate, 1.2–2.0 was large, >2.0 was very large.
Means (±SD), p-values, and effect sizes for all %MVC values of examined musculature across the 5 pike variations are shown in Table 2. For the RA and LSES, %MVC values for the PK were significantly lower compared with all instability device pikes. Hopkins scale of magnitude determined large differences between the PK and each of the instability device pikes for the RA. No differences existed between any of the instability devices for RA or LSES activation. The effect size differences between the unstable pikes for the RA were small or trivial. The effect sizes for the ES for all exercises were found to be small or moderate.
In terms of the EO, the PK was significantly lower than the remaining exercises; however, differences existed between the unstable pikes. Electromyographic values of the EO during the BOSU pike were determined to be significantly lower than the ST, SB, and CC; whereas the SB and CC were also found to be significantly lower than the ST. Effect sizes between the PK and remaining exercises for the EO were determined to be moderate and large, whereas the magnitude of differences between the instability device pikes were small to moderate.
Lastly, for the RF, the PK produced significantly lower EMG values than the instability pikes. Furthermore, activation of the RF during the BOSU pike was significantly lower than the ST. Hopkins scale of magnitude found all differences between the PK and all instability pikes to be of moderate size although the effect sizes between the unstable pikes were either trivial or small.
Limited literature exists comparing multiple commercial instability devices with popular abdominal movements. Thus, the purpose of this investigation was to compare the EMG differences between the RA, EO, LSES, and RF during pike variations performed with several commercial fitness devices compared with a stable surface. Findings were consistent with previous literature, showing instability devices provide increased EMG activity of the abdominal wall and surrounding musculature when compared to their stable counterparts (2,6,11,12,18,20). One explanation is that instability devices may increase the amount of spinal column disturbances (e.g., rotation, shearing, etc.), thereby demanding further requirements of the core musculature. Previous research, by Snarr and Esco (16), stated that prime movers may take on a “multi-role” approach when introduced into an unstable environment. For instance, the RA's primary actions are lumbar flexion, posterior pelvic rotation, and lateral flexion (8); however, the RA may also serve as a dual role in stabilizing the trunk during movement. Thus, an increase in activation of the core musculature may occur when various forms of instability are introduced compared with a stable surface.
Although no differences in muscle activation between the BOSU and remaining instability pikes existed within the RA or LSES, significant values were seen within the EO and RF. A possible justification may reside within the differences in pike techniques and devices themselves. The PK and BOSU technique both involved stationary feet and hand placements, whereas the others involved dynamic movement of the feet closer to the trunk to complete the exercise. Because the PK and BOSU involved no changes in location of the feet, to create a 90° angle at the hips, the subject had to passively “push back” the hips while actively flexing at the shoulder joint. However, during the ST, SB, and CC, subjects passively flex at the shoulder joint while actively flexing at the hips to complete the repetitions. Therefore, the movements that used active hip flexion (i.e., ST, SB, and CC) demonstrated the significantly higher RF values compared with the BOSU and PK. In terms of the LSES, the PK was shown to be significantly lower in activation compared with the instability devices. During exercises, such as the pike, the LSES is primarily responsible for the stabilization of the spinal segments and does not directly contribute to the main action (i.e., hip flexion). The %MVC values of the LSES are consistent with previous clinical research by Cholewicki et al. (3). Researchers reported similar values of approximately 10%MVC in musculature responsible for segmental spinal stabilization during higher intensity activities.
According to the current results, the ST provided the highest muscle activation values for each of the muscles examined. An unexpected result was that EO activation during the ST was significantly greater when compared to the remaining exercises, both stable and unstable. This finding is inconsistent with previous findings that demonstrated significantly greater EO activation during isometric planks when the feet where placed upon the Swiss ball versus a suspension training device (17). This inconsistency may possibly be explained by the differences between the devices themselves. The Swiss ball is a single, movable unit, whereas the suspension training device has 2 independent, freely-moving foot cradles. As the subject attempts to create the “pike,” they must maintain simultaneous control of the independent limbs as to create a fluid movement while not swinging to one side or the other. As a result, EO activation may have increased during the ST serving a multiple role to sustain a rigid trunk and prevent lumbar lateral flexion.
Although the current results agree with previous findings, they are also inconsistent with previous research that showed no significant differences or decreases in EMG activity while using instability devices (11,14,15). For instance, Schoffstahl et al. (15) found no significant differences in the RA, EO, or internal obliques during isometric pikes performed with a suspension device, Swiss ball, and other fitness devices (i.e., slide board and power wheel). The differing results can be explained such that the previous study (15) examined only isometric contractions; whereas the current examined dynamic movements. During dynamic motion on instability devices, larger disturbances to the spinal column and limbs may exist that elicit greater demands in muscular activity to maintain body position. Although previous research (16) has demonstrated that isometric movements upon instability devices can induce greater activation of the abdominal wall, it can be speculated that isometric pikes on unstable surfaces reduced the moment arm of resistance enough that multiplanar stability requirements diminished in the isometric environment.
This study is not without limitations such that the population tested is limited to adults aged 19–35 years of age. Electromyographic patterns may differ in elderly populations that may require more stability and balance to perform stable traditional movements as compared to younger adults because of diminished timing of motor unit firing, decreased muscle mass, and other sarcopenic effects on the neuromuscular system. Another limitation to this study is the use of a select few instability devices. Although multiple commercial devices exist, the current study was limited to 4 devices that offer variations in stability and are commonly used in fitness and rehabilitation settings. The third limitation is that this study is only measuring an acute bout of instability training and its effects on the core musculature. Future research may be warranted to determine the chronic effects of instability devices on the EMG patterns of the core musculature and effects on balance.
Results of this investigation demonstrated that significant differences exist when performing abdominal exercises on stable and unstable surfaces. Instability devices are shown to be effective at increasing the RA, EO, RF, and LSES musculature when compared to a stable prone pike. Although no differences existed between the instability devices within the RA, a suspension training device was capable of producing a greater muscular demand in the EO, RF, and LSES compared with the other commercial fitness devices. These results indicate that with more freely moving instability devices (e.g., suspension device, Swiss ball, etc.), surrounding musculature (i.e., EO, RF, and LSES) may require greater muscular demands, thereby increasing %MVC activity. Based upon the %MVC values alone, practitioners should take note that traditional stable pikes may not offer a core musculature challenge to resistance-trained individuals but may act as a beginning variation for sedentary individuals. This information may also provide practitioners a useful comparison of available commercial instability devices, which will allow for proper progressions through conditioning and rehabilitation programs based upon muscular activation.
1. Anderson K, Behm DG. The impact of instability resistance training on balance and stability. Sports Med 35: 43–53, 2005.
2. Byrne JM, Bishop NS, Caines AM, Crane KA, Feaver AM, Pearcey GEP. Effect of using a suspension training system on muscle activation during the performance of a front plank exercise. J Strength Cond Res 28: 3049–3055, 2014.
3. Cholewicki J, Juluru K, McGill S. Intra-abdominal pressure mechanism for stabilizing the lumbar spine. J Biomech 32: 13–17, 1999.
4. Cohen J. Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum Associates, 1988.
5. Cram JR, Kasman GS. Introduction to Surface Electromyography. Gaithersburg, MD: Aspen Publishers, Inc., 1998.
6. Duncan M. Muscle activity of the upper and lower rectus abdominis during exercises performed on and off a Swiss ball. J Bodyw Mov Ther 13: 364–367, 2009.
7. Escamilla RF, Lewis C, Bell D, Bramblet G, Daffron J, Lambert S, Pecson A, Imamura R, Paulos L, Andrews JR. Core muscle activation during Swiss ball and traditional abdominal exercises. J Ortho Sports Phys Ther 40: 265–276, 2010.
8. Floyd RT. Manual of Structural Kinesiology (17th ed.). New York, NY: McGraw-Hill, 2009.
9. Hopkins W, Marshall S, Batterham A, Hanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc 41: 3, 2009.
10. Kibler WB, Press J, Sciascia A. The role of core stability in athletic function. Sports Med 36: 189–198, 2006.
11. Lehman GJ, Hoda W, Oliver S. Trunk muscle activity during bridging exercises on and off a Swiss ball. Chiropr Osteopat 13: 14, 2005.
12. Marshall PW, Murphy BA. Core stability exercises on and off a Swiss ball. Arch Phys Med Rehabil 86: 242–249, 2005.
13. McGill SM, Cannon J, Anderson JT. Analysis of pushing exercises: Muscle activity and spine load while contrasting techniques on stable surfaces with a labile suspension strap training system. J Strength Cond Res 28: 105–116, 2014.
14. Saeterbakken AH, Andersen V, Jansson J, Kvellestad AC, Fimland MS. Effects of BOSU ball(s) during sit-ups with body weight and added resistance on core muscle activation. J Strength Cond Res 28: 3515–3522, 2014.
15. Schoffstahl JE, Titcomb DA, Kilbourne BF. Electromyographic response of the abdominal musculature to varying abdominal exercises. J Strength Cond Res 24: 3422–3426, 2010.
16. Snarr RL, Esco MR. Electromyographic comparison of traditional and suspension push-ups. J Hum Kinet 39: 75–83, 2013.
17. Snarr RL, Esco MR. Electromyographical comparison of plank variations performed with and without instability devices. J Strength Cond Res 28: 3298–3305, 2014.
18. Snarr RL, Esco MR, Witte EV, Jenkins CT, Brannan RM. Electromyographic activity of rectus abdominis during a suspension push-up compared to traditional exercises. J Exerc Physiol 16: 1–8, 2013.
19. Tan S, Cao L, Schoenfisch W, Wang J. Investigation of core muscle function through electromyography activities in healthy young men. J Exerc Physiol 16: 45–52, 2013.
20. Vera-Garcia FJ, Grenier SG, McGill SM. Abdominal muscle response during curl-ups on both stable and labile surfaces. Phys Ther 80: 564–569, 2000.
21. Wahl MJ, Behm DG. Not all instability training devices enhance muscle activation in highly resistance-trained individuals. J Strength Cond Res 22: 1360–1370, 2008.