Over the past two decades, lumbar spine stability has become an integral part of the low back pain assessment and treatment strategies, especially given its potential link to injury mechanisms and the ongoing clinical efforts directed toward enhancing stability in patients [1]. Furthermore, an increasing number of researchers and clinicians consider the strategy used by patients to activate their abdominal muscles to be central to the stability theme. It has been demonstrated that bracing, defined as the increase of torso stiffness by the activation of all abdominal muscles and back extensors muscles, produces greater stability than hollowing, which consist of the activation of the transversus abdominis and internal oblique muscles in healthy subjects [1, 2]. As a corollary, a variety of trunk coactivation exercise protocols are frequently used in daily clinical practice for low back pain prevention in healthy patients, rehabilitation in low back pain patients, or in order to evaluate trunk muscle function. The quadratus lumborum, external oblique, internal oblique, iliocostalis, longissimus and intertransversalis are believed to act as spine stabilizers when contracting bilaterally and as lateral trunk flexors by pulling the rib cage toward the hip when contracting unilaterally [3].

McGill et al. [4] suggested that a difference between left and right endurance time of the trunk flexors, extensors and lateral flexors muscles would predict who is at greater risk of back problem. As no exercise can evaluate all muscles involved in lumbar spine stability, evaluation in the 3 planes has to be done separately. Side bridge exercise protocols have been suggested to evaluate torso muscles in frontal plane. Such exercises are usually executed in a position where the participant lies down sideways with support of one arm and are named after the side of the arm support (e.g. left side bridge = left arm support) [5] Variants to the protocol initially described by McGill have also been described [5–7].

A wide range of use of these exercises has been presented in the literature. Some authors proposed comparing muscle balance by evaluating holding times, whereas others assessed the use of maximal voluntary contraction or isometric contraction of short duration to evaluate muscle recruitment.

McGill et al. [8] evaluated the holding times of 75 healthy subjects (mean age of 23 years) during side bridge protocols and obtained mean times of 81 ± 34 sec and 85 ± 36 sec for right and left side bridge respectively. Other authors [9] reported similar mean times, i.e. 87.5 ± 36.4 sec and 92.0 ± 45.8 sec for right and left side bridges respectively, in a group of 24 healthy subjects (mean age of 35.3 years). McGill [10] has proposed that endurance scores during side bridges could be interpreted by using a right side to left side holding time ratio. A discrepancy of over 0.05 in the ratio would suggest unbalanced endurance. McGill et al. [8] also examined the intra-rater reliability of this test in 5 subjects on 5 consecutive days and at 8 weeks (follow-up) and got an excellent reliability coefficient of over .96. Evans et al. [9] reported high intra-rater and inter-rater reliability with the lowest coefficient being .81 and .82 respectively.A number of studies have also provided information about muscle recruitment during side bridge tests. McGill et al. [11] reported ipsilateral (left side during side bridges with left supporting arm) trunk muscle activation of the quadratus lumborum, external oblique, rectus abdominis and lumbar erector spinae during isometric side bridges. These muscles presented 54%, 40%, 22% and 24% of their maximal voluntary contraction activity (MVCA) respectively. However, the small sample (4 participants) does not allow for much generalization. Ekstrom et al. [12] evaluated muscle activation during 5-second hold side bridges and reported an activation of up to 72% of MVCA of the gluteus medius, 69% of the external oblique, between 34% and 42% of the longissimus thoracis, lumbar multifidus and rectus abdominis and less of 21% of the gluteus maximus and harmstring muscles. In 2008, Ekstrom et al. [13] used the same protocol to evaluate the longissimus thoracis and lumbar multifidus activation on both sides during side bridges. The authors reported greater activation of the ipsilateral longissimus thoracis than of the lumbar multifidus (48-49% vs. 32-33% of MVCA respectively) and greater activation on the ipsilateral side than on the contralateral side (48-49% vs 7-8% of MVCA for the longissimus thoracis and 32-33% vs 14% of MVCA for the lumbar multifidus). They also evaluated muscle activation during maximal resistance in side bridge without trunk support (legs fixed and trunk unsupported) and reported greater activation of the ipsilateral (left muscle when left side up) longissimus thoracis at L1 than of the lumbar multifidus (54-58% and 38-39% of MVCA respectively). They demonstrated a greater activation on the ipsilateral side than on the contralateral one (54-58% vs. 7-9% of MVCA for longissimus thoracis and 38-39% vs. 12-13% of MVCA for lumbar multifidus). In light of these studies, the external oblique, gluteus medius and back extensor muscles on the ipsilateral side of the side bridge seem to be the muscles with the greatest levels of activation during this exercise.

Although muscle activation during side bridge endurance and trunk lateral flexion has been described in a few studies, muscle recruitment and fatigability protocols have not been extensively studied. Therefore the purpose of our study was to quantify fatigue parameters in various trunk muscles during a modified side bridge endurance test (i.e. a lateral isometric hold test on a 45° roman chair apparatus). A second objective was to determine which primary trunk muscles are fatigued during this functional task. We hypothesized that the ipsilateral external oblique and erector spinae muscles would exhibit the highest fatigue indices during the modified side bridge protocol.

This study’s main objectives were to quantify fatigue parameters in various trunk muscles during a lateral isometric hold test in healthy participants and to determine which primary trunk muscles are fatigued during this functional task. The ipsilateral external oblique, during both lateral isometric hold tests, and the contralateral L5 erector spinae, during the left lateral isometric hold test, showed significantly steeper negative NMF_slope than the other side’s respective muscles. The results also showed that all recorded trunk muscles presented fatigue parameters during lateral isometric hold tests on both sides. Thereby, these results provide a preliminary understanding of the trunk muscles tested during a lateral isometric hold task.

Even if the results showed a NMF_slope significantly different from 0 for the 8 muscles recorded, the percentage of decay did not exceed 0.07%s^-1 which may raise questions with regard to the level of fatigue generated in individual muscles during the lateral hold. Despite the lack of normative data for fatigue parameters, other studies have reported steeper negative NMF_slope during trunk endurance protocols. Plamondon et al. [18] reported L3 erector spinae NMF_slope (%s^-1 ± SD) of 0.10 ± 0.07 and 0.13 ± 0.05 for women and men respectively during modified Sorenson intermittent contraction tasks. Mannion et al. [16, 17] reported NMF_slope values up to 0.46 ± 0.19 during modified Biering-Sorenson test. The results of the present study could suggest that muscles not recorded in our experiment could act as primary contributors during the lateral isometric hold task. In fact, the quadratus lumborum, which activity can be predicted with surface EMG placed on L3 erector spinae [15], has been reported to act as a trunk lateral flexor when contracted unilaterally [3, 19], and some studies reported high levels of activity in this muscle during side bridges [11, 20]. Lower limb muscle activity was not recorded in the present study, but gluteus medius, gluteus maximus and hamstring activation during different side bridges exercises has also been reported [6, 12, 13].

As often seen in motor tasks, there are certainly multiple muscle recruitment strategies (redundancy) that can be selected by the central nervous system to optimize task performance during a lateral isometric hold test [21]. During this task, more than one muscle contribute to the torque generated, and a changing combination of muscle forces to maintain the isometric contraction may be an efficient strategy to improve performance. Motor variability has been shown to be an efficient strategy to reduce the development of muscle fatigue [22]. Our results suggest that a variable trunk co-contraction strategy where numerous muscles contribute to the generation of isometric force is selected during a lateral isometric hold test. A few authors also reported global activation of trunk muscles instead of specific muscle recruitment during trunk endurance task. Page et al. [23] reported fatigue of various trunk muscles (i.e. abdominal muscles, lumbar erector spinae, biceps femoris) during sustained isometric contractions of abdominal muscles whereas Plamondon et al. [18] reported fatigue of lumbar erector spinae and hip extensor muscles (hamstring and gluteus maximus) during an intermittent modified Sorenson contraction task. Furthermore, other studies have reported the recruitment of several trunk muscles (e.g. quadratus lumborum, external oblique, rectus abdominis, lumbar and thoracis erector spinae, gluteus medius, lumbar multifidus, gluteus maximus and hamstrings) during side bridge tasks [11–13].

Our participants showed mean holding times similar to the ones reported in previous studies where lateral flexors endurance was evaluated using the side bridge position [8, 9]. However, although only healthy participants were included in the study, a mean difference of more than 5% between both side holding times was observed. According to McGill, such differences in holding times during the side bridge position would suggest trunk muscle imbalance [10]. The relatively high variability of the right on left holding time ratio (i.e. a standard deviation of 11.8%) suggests a wide variation of this ratio in healthy populations. Differences in testing protocols (isometric lateral hold versus side bridge) may explain these differences. However, one might question the clinical value of the proposed criteria suggesting that differences of more than 5% between right and left holding times characterize individuals with a history of disabling back troubles or an increased risk of back trouble [10]. Others studies are necessary to validate the use of a holding time ratio derived from the lateral isometric hold test as a normative data to evaluate muscle balance and function.

A few limitations need to be considered when interpreting the present results. For this study, only healthy young adults were recruited, and consequently generalization to other healthy or clinical populations may be limited. Only eight trunk muscles (surface EMG) were recorded, and other studies are necessary to evaluate the possible contribution and fatigue of several other trunk or lower limb muscles that could be recruited during lateral isometric hold task.

This study was conducted in order to identify muscles evaluated during a lateral isometric hold task. Although the fatigue indices suggest that ipsilateral external oblique and contralateral L5 erector spinae are significant contributors, all recorded muscles were active during the lateral isometric hold task. The lateral isometric hold, while different from the side bridge assessment for endurance because it does not involve support of the floor, seems to be adequate for the evaluation of lateral trunk flexors. Studies involving individuals with low back pain are needed to evaluate the clinical relevance of this procedure.