Therapists, trainers, and coaches have emphasized the importance of abdominal exercises for years. Reasons have included sport performance, injury prevention and rehabilitation (especially low back pain), and aesthetics. Much of the interest has centered on the perceived need to stabilize the “core,” which has generated a variety of abdominal exercises designed to target specific muscles. Several studies have examined the interplay between the hip flexors and the abdominal muscles during a variety of exercises (2,3,10,20). The general consensus has been that high levels of hip flexor activity during abdominal strengthening exercises are undesirable. Ways in which the abdominal muscles can be optimally activated while minimizing activation of the hip flexors would seem to have practical importance.
Abdominal muscle activity has been found to be very dependent on the position of the pelvis during the execution of the exercise. In particular, a posterior pelvic tilt has been found to have a marked influence on the activation of abdominal musculature (9,20,23). Shirado and colleagues (21) reported that pelvic alignment could influence the electromyographic (EMG) activity of the trunk flexors and extensors during isometric trunk exercises. Full flexion of the lumbar spine has been reported to be unnecessary for maximum electrical activity of the abdominal muscles, suggesting that it is the position of the pelvis that influences the activation of the trunk muscles (19). Although there are some studies that have examined the effect of a posterior pelvic tilt on activation of the trunk musculature, the effect of an anterior tilt or neutral position of the pelvis has not been clearly elucidated. Many therapists and exercise specialists advocate the maintenance of a neutral spine and pelvis (17,18) during abdominal exercises in order to facilitate carryover into functional activities. In addition, observation of people performing a variety of abdominal exercises reveals that most do not prevent moving from a neutral to an anterior tilt, which potentially changes the purpose of the exercise and may predispose them to a risk of low back problems (13). It would seem important to define more clearly the effect of pelvic position, especially neutral and anterior tilt positions, on the activation of the trunk flexors.
The Janda sit-up (Figure 1), devised by Czech physician Vladimir Janda and also referred to as a heels-press sit-up, has received popularity in part because it is purported to decrease hip flexor activity during the sit-up movement through reciprocal inhibition (10). By actively contracting the hamstrings muscles, an individual will theoretically deactivate the hip flexors (10). However, there is little published evidence to support or refute this theory.
The purpose of this study was to determine the influence of pelvis position on the relative activity of selected abdominal and hip musculature. A second purpose of the study was to compare the relative muscle activity during an isometric hold at approximately 45° of the Janda sit-up to the relative activity of the muscle in the 3 pelvic positions. It was hypothesized that the neutral and posterior pelvic tilt would increase the activation of the anterior trunk muscles and that anterior pelvic tilt would increase the activation of the rectus femoris (reflecting the iliopsoas). In addition, it was hypothesized that the Janda sit-up would produce high levels of activation in the anterior trunk muscles while decreasing the activation of rectus femoris.
Experimental Approach to the Problem
Participants assumed a supine position and were fitted unilaterally with surface EMG electrodes on the upper rectus abdominis (URA), lower rectus abdominis (LRA), lower abdominal stabilizers (LAS), external obliques, rectus femoris, and biceps femoris muscles. Participants were asked to perform a Janda sit-up (exercise 1) and hold a position during the sit-up with the trunk at approximately 45° to the bench while contracting the hamstrings. The second exercise involved an isometric double straight leg lift (DSLL) (exercise 2) in each of 3 pelvis positions: anterior tilt, neutral, and posterior tilt. Each contraction was randomly allocated and held for 5 seconds. Two trials of each exercise were performed with a 30-second rest between trials and a 3-minute rest between each different exercise test position. The EMG activity of each muscle was monitored across each condition.
A convenience sample of 15 subjects was selected to participate in this study. All participants were male, with a mean age of 25.9 ± 8.4 years, mean height of 177.4 ± 9.5 cm, and mean weight of 78.9 ± 11.9 kg. The participants were instructed on the nature of the study and the equipment and apparatus involved and were provided with the opportunity to clarify this information. All subjects were either competitive rugby players (n = 11) or recreational athletes (racquet sports, running) (n = 4) who presently had no apparent or known musculoskeletal injuries. All participants had extensive experience with resistance training and performing a variety of abdominal exercises throughout their training career. Furthermore, they all scored in the excellent category in the partial curl-up test of the Canadian Physical Activity, Fitness, and Lifestyle appraisal. Fourteen participants were able to complete the exercise movements correctly. One participant was unable to maintain the pelvic tilt position during the isometric portion of the leg raise activity. His data were not included in the statistical analysis. Each subject was required to read and sign a consent form before participation. The Human Investigation Committee, Memorial University of Newfoundland, approved this study.
Surface Electromyography Preparation and Placement
It has been suggested that a valid EMG signal is compromised when the muscles of interest are performing a dynamic contraction (4,8). As the joint moves through a range of motion, the distance between the muscle and the detection surface changes, which results in a change in the EMG amplitude. It is recommended that isometric contractions be used to control for movement during surface EMG testing. Although this detracts from the ecological validity, it does increase the validity and reliability of the EMG signal (8). An effective start for analyzing the effect of pelvis position on trunk muscle activity would be to use quantified and controlled EMG procedures.
The electrode placement sites were prepared by shaving, exfoliating with sandpaper, and wiped with isopropyl alcohol. Participants were placed in a supine position on a plinth, providing support to the entire length and width of the body. Electromyographic surface electrodes (Kendall Medi-trace 100 series; Kendall, Chikopee, MA) were placed in parallel with the muscle fibers with an interelectrode distance of 2 cm. A ground electrode was placed at the nearest bony prominence for each pair of active electrodes. The 6 muscle sites were the URA, LRA, external obliques, LAS (reported to represent activation of the internal obliques and transversus abdominis [1,5,6]), biceps femoris, and rectus femoris. The rectus femoris was used to approximate the activity of the deep hip flexors, namely, the iliopsoas muscle group (14). Landmarking for the URA was achieved by measuring 3 cm lateral to the midline and midway between the xiphoid process and the umbilicus. The LRA was positioned 3 cm lateral to the midline and 2 cm inferior to the umbilicus. Additional electrodes were placed superior to the inguinal ligament and 1 cm medial to the anterior superior iliac spine for the lower abdominals. McGill et al. (14) reported that surface electrodes adequately represent the EMG amplitude of the deep abdominal muscle within a 15% root mean square difference. However, Ng et al. (15) indicated that electrodes placed medial to the anterosuperior iliac spine would receive competing signals from the external obliques and transverse abdominis with the internal obliques. Based on these findings, the EMG signals obtained from this abdominal location are described in the present study as the LAS, which would be assumed to include EMG information from both the transverse abdominis and internal obliques. The external obliques were positioned superior to the anterior superior iliac spine at an oblique angle, at the level of the umbilicus. Biceps femoris electrodes were positioned at the midpoint of the muscle belly of the biceps femoris. Rectus femoris electrodes were positioned at the most proximal aspect of the muscle belly. All muscle sites were measured on the right side of the body only.
The participants were instructed on proper technique to complete a maximal anterior pelvic tilt and maximal posterior pelvic tilt. The anterior pelvic tilt was achieved by asking the participants to tilt the pelvis forward in order to create as much space as possible between the plinth and the lower back area. The posterior pelvic tilt was achieved by asking the participants to flatten their lower back into the plinth. Manual guidance was also provided during the instruction and familiarization period to ensure proper technique and understanding. The neutral position was described as the participants' normal, comfortable resting supine position. One investigator was positioned by the side of each participant to ensure proper pelvic positioning during data collection as well as palpating the anterosuperior iliac spine as a way of monitoring pelvic position. A second investigator was positioned at each participant's feet to ensure proper leg lifting during the exercise. Participants were instructed to keep their head resting on the plinth and to rest their hands by their sides. Participants began in a supine position on the plinth with their legs straight and their feet placed on a stable bench 15 cm in height. For the anterior pelvic tilt position, participants were asked to assume the proper position. A reference mark was placed on the lateral malleolus and the bench supporting the feet. This mark would be used to ensure the same starting position for the second trial of the exercise. The participants were asked to raise their feet off the support 5 cm, hold the position for 5 seconds, and return to the support. A 30-second rest period was provided before a second trial was performed. The same procedure was followed for the neutral and posterior pelvic tilt positions. A 3-minute rest period was provided between the anterior, neutral, and posterior tilt trials. The order of exercises was randomized. If the position was not held properly, then the position and the data acquisition was terminated and attempted again after an appropriate rest period.
The Janda sit-up was performed in a supine, crook-lying position (Figure 1). A padded bar was placed at the back of the lower leg and held in place manually by one of the investigators. This bar provided an object against which each participant was able to contract the hamstring muscles, by attempting to perform bilateral knee flexion. This bar was held manually by an investigator in order to ensure that consistent hamstring contraction occurred throughout the entire exercise trial. The participants were instructed to contract the hamstring muscles, perform the sit-up, and hold for 5 seconds at approximately 45° before returning to the start position. A 30-second rest period was provided before the second trial. The order of pelvic position and the Janda sit-up was randomly assigned.
The isometric BSLL was used in this study to allow us to maintain a relatively constant torso and leg position, while changing only the pelvis position. We do acknowledge that with any change in pelvis rotation there will be changes in the rest of the kinetic chain, both above and below the pelvis. However, this exercise would provide the most consistency in the upper and lower body segments, allowing us to examine the influence of the pelvis on abdominal muscle activity.
Electromyographic Data Collection
Electromyographic data were collected during the concentric and isometric contractions of each exercise. The EMG signals were amplified (MEC 100 amplifier; Biopac Systems Inc., Santa Barbara, CA), monitored, and directed through an analog-digital converter (Biopac MP100) to be stored on the computer (Sona, St. John's, Newfoundland, Canada). The EMG signals were collected over 15 seconds at 2000 Hz and amplified (×500). The EMG activity was sampled at 2000 Hz with a Blackman 61-dB band-pass filter between 10 and 500 Hz, amplified (Biopac Systems MEC bipolar differential 100 amplifier, Biopac Systems, Inc.; input impedance = 2 MΩ common mode rejection ratio >110 dB minimum (50/60 Hz), noise >5 UV) and analog-to-digitally converted (12 bit) and stored on personal computer for further analysis. The EMG signal was rectified and integrated over the 5-second static (isometric) contraction period of the movement. An average of the 2 trials was obtained, and the mean integrated value used for statistical analysis. Similar to previous published research from this laboratory (6), absolute rather than normalized EMG data were analyzed because it was a repeated-measures design that was completed in a single experimental session (no change in electrode position). Since the focus was on changes in activation of individual muscles and not between muscles or individuals, normalization of the electromyogram was not considered necessary.
A 1-way, repeated measures analysis of variance (GBStat; Dynamic Microsystems, Silver Spring, MD) was performed to detect differences in muscle activation for each muscle, relative to pelvic position and Janda sit-up exercise. When statistical significance was found, the Dunn's (Bonferroni) post hoc test was used to reveal the differences. Descriptive statistics include mean ± SD.
Upper Rectus Abdominis
For the URA site, the Janda sit-up demonstrated the highest EMG activity. Relatively, the anterior pelvic tilt position showed 70.9% less activity in the URA (p < 0.01). The EMG activity in the neutral position was 52.1% less than that seen in the Janda sit-up (p < 0.01). There was no significant difference between the Janda sit-up and the posterior pelvic tilt position. The anterior position demonstrated 57% less activity than the posterior pelvic tilt position (p < 0.05) (Figure 2).
Lower Rectus Abdominis
For the LRA site, the Janda sit-up elicited the highest EMG activity. This was significantly different than the anterior pelvic tilt position (p < 0.01), which showed 68.4% less activity, and the neutral position (p < 0.05), which showed 46.3% less activity. There was no significant difference between the Janda sit-up and the posterior pelvic tilt position for LRA activity. The anterior pelvic tilt position showed significantly less (56.6%) activity in the LRA than in the posterior pelvic tilt position (p < 0.05) (Figure 3).
The rectus femoris site demonstrated the highest activity in the anterior pelvic tilt position. This was significantly different from the Janda sit-up (p < 0.05), which showed 38.1% less activity. There were no other significant rectus femoris differences when compared to the other test positions. The Janda sit-up was not significantly different from the posterior pelvic tilt or neutral positions (Figure 4).
In the biceps femoris site, the Janda sit-up provided the highest EMG activity. This was significantly higher than all other test positions (p < 0.01). The neutral, anterior, and posterior pelvic tilt positions demonstrated 91.1%, 88.6%, and 87.8% less biceps femoris activity, respectively. There were no other significant differences in biceps femoris activity among the pelvic tilt positions (Figure 5).
There were no statistically significant differences in external obliques EMG activity when comparing the 4 test positions (p = 0.09). The Janda sit-up and the neutral pelvis positions showed the greatest difference in EMG activity.
Lower Abdominal Stabilizers
There were no statistically significant differences in LAS EMG activity when comparing the 4 test positions.
The major findings of this study show that changing the position of the pelvis significantly changes the pattern of activation of the URA, LRA, and rectus femoris. This is in agreement with a study by Shields and Heiss (20) who found that the double straight leg lowering exercise, while maintaining posterior pelvic tilt, achieved greater abdominal muscle activation compared to a typical crunch exercise. Posterior pelvic tilting has also been found to activate the rectus abdominis to a greater degree than in the abdominal hollowing exercise (9). Other studies have identified high levels of rectus abdominis activity during the posterior pelvic tilt maneuver (24) and leg lifting exercise (2). This differs from the results of Urquhart et al. (23) who found the internal oblique muscle more active than the rectus abdominis during a posterior pelvic tilt. In the Urquhart et al. study, participants were asked to gently and slowly rock their pelvis backward. Urquhart et al. (23) describe this as a gentle effort, corresponding to a 2 on the Borg scale. The present methodology differed in that the posterior pelvic tilt was accompanied by the isometric DSLL, a much more demanding task. Our study was in agreement, however, with the authors' conclusion that abdominal muscle activity was dependent on body position, including lumbopelvic motion or position.
There is general agreement that an individual cannot preferentially activate the URA versus LRA (7,12) unless highly trained (19). The results of the present study also found similar activation patterns for the URA and LRA throughout the exercises. Moreover, it has been found that no single exercise is able to optimally recruit all the abdominal musculature simultaneously (3). Therefore, a comprehensive, individualized program is required to sufficiently challenge each of the abdominal muscles (3) in different planes of movement.
The anterior pelvic tilt position provided the highest EMG activity in the rectus femoris and the lowest EMG recordings in both the URA and LRA. The anterior tilt may place the rectus femoris and underlying iliopsoas muscle group in a more optimal length position. This will change the muscle length-tension relationship and produce higher contractile forces. As the rectus femoris is in an optimal position, the LRA and URA will be placed in a relatively lengthened position. For the LRA and URA, the change in length-tension relationship may place the muscles in a disadvantageous position and cause a reduction in contractile forces. Furthermore, several authors have cautioned against the use of the BSLL because of the risk of low back injury caused by increased shear and compressive forces (3,9). Invariably, individuals may adopt an anterior pelvic tilt position when performing sit-ups or leg raises that can be considered contraindicated considering the increased shear and compressive forces (3,9) placed on the lower back by stronger hip flexors.
A secondary finding showed that the Janda sit-up produced relatively high levels of URA and LRA activity and low levels of rectus femoris activity; however, this was not significantly different from the posterior pelvic tilt. Our results regarding the inability of the Janda sit-up to significantly reduce hip flexor activity in comparison to the posterior pelvic tilt are in agreement with Juker et al. (10) who found no decrease in psoas activity using the “press heels” sit-up. The Janda sit-up is identical to a traditional bent-knee sit-up when considering the trunk flexion component. The difference is in the contraction of the hamstring muscles during the exercise. As this is a sit-up movement, it is typically performed in a posterior pelvic tilt start position (16). Participants were not instructed regarding pelvis position before the Janda sit-up trials. Therefore, pelvis position was not controlled during this exercise. This may account for some of the similarities between the Janda sit-up and the results from the posterior pelvic tilt position. The contraction of the hamstrings during the Janda sit-up purportedly reduces hip flexor activation through reciprocal inhibition (10). Our data cannot conclude whether the low rectus femoris activity can be attributed to reciprocal inhibition through contraction of the hamstring musculature. The Janda sit-up did demonstrate the highest biceps femoris activity as anticipated. However, the rectus femoris activity was not significantly different from the posterior pelvic tilt position.
Differences in the posterior pelvic tilt and Janda sit-up are seen when we examine their relationship to the neutral pelvis position. For both the URA and LRA sites, the Janda sit-up demonstrated significant differences from the neutral position; however, the posterior pelvic tilt position did not. This may be explained by the investigators' definition of neutral pelvis. The participants were asked to maintain their normal, comfortable supine position. The discrepancy of neutral for each participant may have influenced the results. In addition, anatomically, the neutral position may be closer in range of available motion to the posterior tilt than the anterior direction. This may account for the lack of significant difference in muscle activity when comparing the neutral position to the posterior pelvic tilt position.
When we examine the overall trend of muscle site activity, a pattern emerges. As the participant moves from a posterior pelvic tilt position through neutral to the anterior pelvic tilt position, the relative activity of the URA, LRA, and rectus femoris becomes reversed. During the posterior pelvic tilt and Janda sit-up, there are high levels of activity in both the URA and LRA and low activity in the rectus femoris. In the neutral position, the level of activity of the URA and LRA decreases, although this was shown to be only significantly different from the Janda sit-up. The activity of the rectus femoris increased slightly when mean EMG activity was examined; however, the change was not significant. In the anterior pelvic tilt position, the URA and LRA exhibited their lowest activity levels, while the rectus femoris shows the highest level of activity.
There was no significant difference between the exercises in the amount of EMG activity in the LAS or external obliques muscle sites. This differs from Shields and Heiss (20), who found varying levels of oblique muscle activity during their isometric double straight leg lowering exercise. The finding in the present study would suggest that the stabilizing role of the LAS (24) was similar for all pelvic positions as well as the Janda sit-up. As the DSLL is not a trunk flexion exercise, a significant difference in the activity of a trunk flexor such as the external obliques might not be expected. The Janda sit-up, however, is a trunk flexion exercise, but it did not show significant differences in external obliques activity compared to the 3 different pelvis positions. During these exercises, the external obliques probably also act as a stabilizer (2).
The biceps femoris muscle site was also unaffected by a change in pelvis position. This hip extensor muscle may be expected to have little activation during a hip flexion type of activity. During the Janda sit-up, there was significantly greater biceps femoris activity compared to the other test positions. This is to be expected as the participant is instructed to actively contract the hamstrings while performing the Janda sit-up.
The results of this study will be of value when instructing persons in correct posture during supine abdominal strengthening activities. There is evidence showing that specific exercise instruction is important for a client to learn and retain the proper technique and form of an exercise (11). Particular attention should be given to individuals with increased lumbar lordosis or very weak abdominal muscles. Several authors have stressed the potential increase in lumbar compression and shear force with some abdominal exercises. The BSLL is not recommended for individuals who have known lumbar pathologies or very weak abdominal musculature (3,10). These individuals may be at risk of moving into an anterior pelvic tilt position due to postural habit or fatigue while exercising (9,22). By changing the rotation of the pelvis, the focus of the strengthening exercise may shift from the abdominals to the hip flexors. These results will add to the existing and emerging scientific literature regarding the relationship between the pelvis, hip, and lumbar spine and the interplay of the supporting musculature.
From these data, we can conclude that a change in pelvis position demonstrates significant differences in URA, LRA and rectus femoris muscle activity, as measured by surface electromyography. When considering pelvis position independently, the highest abdominal muscle activity occurs in the posterior pelvic tilt position. The Janda sit-up also seems to be effective in producing significant activation of the rectus abdominis.
The National Science and Engineering Research Council of Canada supported this research.
1. Anderson, K and Behm, DG. Trunk muscle activity increases with unstable squat movements. Can J Appl Physiol
30: 33-45, 2005.
2. Andersson, EA, Nilsson, J, Ma, Z, and Thorstensson, A. Abdominal and hip flexor muscle activation during various training exercises. Eur J Appl Physiol
75: 115-123, 1997.
3. Axler, CT and McGill, SM. Low back loads over a variety of abdominal exercises: searching for the safest abdominal challenge. Med Sci Sports Exerc
29: 804-811, 1997.
4. Basmaijan, JV and De Luca, CJ. Muscles Alive: Their Functions Revealed by Electromyography
(5th ed.). Baltimore: Williams & Wilkins, 1985.
5. Behm, DG, Burry, SM, Greeley, GED, Poole, AC, and Mackinnon, SN. An unstable base alters limb and abdominal activation strategies during the flexion-relaxation response. J Sports Sci Med
5: 323-332, 2006.
6. Behm, DG, Leonard, A, Young, W, Bonsey, A, and MacKinnon, S. Trunk muscle EMG activity with unstable and unilateral exercises. J Strength Cond Res
19: 193-201, 2005.
7. Clark, KM, Laurence, EH, and Sinyard, J. Electromyographic comparison of the upper and lower rectus abdominis during abdominal exercises. J Strength Cond Res
17: 475-483, 2003.
8. De Luca, CJ. The use of surface electromyography in biomechanics. J Appl Biomech
13: 135-163, 1997.
9. Drysdale, CL, Earl, JE, and Hertel, J. Surface electromyographic activity of the abdominal muscles during pelvic-tilt and abdominal-hollowing exercises. J Athletic Train
39: 32-36, 2004.
10. Juker, D, McGill, S, Kropf, P, and Steffen, T. Quantitative intramuscular myoelectric activity of lumbar portions of psoas and the abdominal wall during a variety of tasks. Med Sci Sports Exerc
30: 301-310, 1998.
11. Karst, GM and Willett, GM. Effects of specific exercise instructions on abdominal muscle activity during trunk curl exercises. J Orthop Sports Phys Ther
34: 4-12, 2004.
12. Lehman, GJ and 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
81: 1096-1101, 2001.
13. McGill, SM. The biomechanics of low back injury: implications on current practice in industry and in the clinic. J Biomech
30: 465-475, 1997.
14. McGill, SM, Juker, D, and Kropf, P. Appropriately placed surface EMG electrodes reflect deep muscle activity (psoas, quadratus lumborum, abdominal wall) in the lumbar spine. J Biomech
29: 1503-1507, 1997.
15. Ng, JK, Kippers, V, and Richardson, CA. Muscle fiber orientation of abdominal muscles and suggested surface EMG electrode positions. Electromyogr Clin Neurophysiol
38: 51-58, 1998.
16. Norris, CM. Abdominal muscle training in sport. Br J Sports Med
27: 19-26, 1993.
17. Panjabi, MN. The stabilizing system of the spine. Part 1. Function, dysfunction, adaptation, and enhancement. J Spine Disord
5: 383-389, 1992.
18. Panjabi, MN. The stabilizing system of the spine. Part 2. Neutral zone and stability hypothesis. J Spine Disord
5: 390-397, 1992.
19. Sarti, MA, Monfort, M, Fuster, MA, and Villaplana, LA. Muscle activity in upper and lower rectus abdominis during abdominal exercises. Arch Phys Med Rehabil
77: 1293-1297, 1996.
20. Shields, RK and Heiss, DG. An electromyographic comparison of abdominal muscle synergies during curl and double straight leg lowering exercises with control of the pelvic position. Spine
22: 1873-1879, 1997.
21. Shirado, O, Toshikazu, I, Kaneda, K, and Strax, TE. electromyographic analysis of four techniques of isometric trunk muscle exercises. Arch Phys Med Rehabil
76: 225-229, 1995.
22. Shirazi-Adl, A, Sadouk, S, Parnianpour, M, Pop, D, and El-Rich, M. Muscle force evaluation and the role of posture in human lumbar spine under compression. Eur Spine J
11: 519-526, 2002.
23. Urquhart, DM, Hodges, PW, Allen, TA, and Story, IH. Abdominal muscle recruitment during a range of voluntary exercises. Manual Ther
10: 144-153, 2005.
24. Vezina, MJ and Hubley-Kozey, CL. Muscle activation in therapeutic exercises to improve trunk stability. Arch Phys Med Rehabil
81: 1370-1379, 2000.