Abstract
BACKGROUND:
There are no studies on the scoliotic population in which lateral abdominal muscle (LAM) was measured.
OBJECTIVE:
The aim of the study was to deliver the first results on LAM elasticity assessed by shear wave elastography (SWE) in right-side and left-side thoracolumbar scoliosis patients.
METHOD:
Twelve patients (mean age 12.25) with thoracolumbar scoliosis were included in the study. Muscle thickness and SWE of the obliquus external (OE), obliquus internal, and transversus abdominis (TrA) were measured with an ultrasound scanner. Measurements were taken at rest and during isometric contraction.
RESULTS:
An analysis showed that in right-side scoliosis, the OE muscle on the convex side was stiffer by 7.25 kPa compared to the concave side. The OE muscle on the convex side in right-side scoliosis was also stiffer by 11.6 kPa compared to the convex side in left-side scoliosis. In left-side scoliosis, the TrA muscle on the concave side was stiffer by 7.84 kPa compared to the convex side.
CONCLUSION:
Changes in LAM elasticity of thoracolumbar scoliosis were observed. A different pattern of LAM stiffness in different curve directions may be expected. OE and TrA shear moduli are the most sensitive to change in adolescent spinal deformation.
Introduction
Scoliosis is defined as a deviation of the spine, consisting of a lateral curvature with vertebrae rotation within the curve. In scoliosis definition, a Cobb angle measured in an anterior-posterior radiograph should be at least 10 degrees and always associated with axial rotation of the vertebrae. The cause and type of scoliosis may be neuromuscular, congenital, related with some syndromes, or idiopathic. In medical treatment, the most common type is idiopathic (without an obvious cause) occurring and/or diagnosed in adolescence (adolescent idiopathic scoliosis – AIS).
Much research has focused on AIS aetiology, but the cause has not been discovered [1]. Theoretically, lateral abdominal muscles (LAM) may have a significant role in scoliosis development and treatment [2], as they (mainly the transversus abdominis muscle) are considered crucial in spinal and pelvic stability [3]. Recently, some studies on LAM have shown differences in LAM thickness at rest and during some contractions in AIS compared to a control group [2,4–6].
To date, LAM in AIS patients were examined by assessing thickness and thickness change during some movements and by assessing side-to-side asymmetry using bright-mode ultrasound imaging [2,4–6]. LAM were also analysed in patients with severe thoracic scoliosis by intramuscular electromyography [7]. Both methods (B-Mode ultrasound imaging and electromyography) have some pros and cons, but in assessing LAM, they may not be fully informative [8]. Thus, knowledge about LAM in scoliotic populations is still very limited.
Recently, a non-invasive and real-time ultrasound shear wave elastography method has been used to quantify muscle elasticity and is called shear wave elastography (SWE). This method allows assessment of muscle elasticity by estimating an apparent shear modulus. A shear modulus measured by SWE is linearly related to passive and active muscle force [9,10]. Hence, SWE seems to be useful for inferring muscle tension or stiffness [9,11]. To our knowledge, there are no studies on the scoliotic population in which LAM elasticity was measured. Previous studies have confirmed that in adults the SWE method is potentially useful in assessing LAM elasticity with acceptable reliability [12–14]. Additionally, our unpublished data confirmed that SWE can be used to assess LAM elasticity in AIS patients with high reliability. Thus, the aim of this preliminary study was to deliver the first results on LAM elasticity assessed by SWE in right-side and left-side thoracolumbar AIS patients.
Methods
Setting and study design
This was an observational study conducted at the Stokrotka Health Resort for paediatric patients (under 18) in the Silesia region of Poland. The study was approved by the local medical ethics committee. All participants and their parents provided written informed consent to participate in the study.
Participants
A sample of 12 participants (mean age, 12.25 ± 2.22 years; sex, 83.3% girls; mean body height, 149.7 ± 11.5 cm; mean body mass, 44.5 ± 15.8 kg; mean BMI, 19.3 ± 4.82 kg/
Out of 12 included participants, only seven were able to perform proper isometric contraction by obtaining the dictated electromyographic (EMG) activity level. Thus, data from seven participants were presented for contraction and SWE ratio measurements.
Examiner
All ultrasound procedures were performed by one investigator, a physiotherapist with six years of experience in assessing abdominal muscle thickness in adolescent and adult populations. The examiner had not previously used SWE. Before testing, the examiner underwent four hours of theoretical training and 20 hours of practical training with an instructor who had extensive experience with SWE in the medical field. Before the study, the examiner examined 40 adolescents under the supervision of the instructor. Additionally, the next 24 participants were examined by the examiner for a reliability study (article under review).
Instrumentation
The eMotion EMG system (Mega Electronics LTD, Kuopio, Finland) was used to simultaneously assess rectus abdominis (RA) muscle activity on both body sides. Two pairs of self-adhesive Ag/AgCl electrodes (Bio-Lead-Lok B, Jozefow, Poland) with a diameter of 36 mm were used. Prior to the electrode placement, the position of RA tendinous intersections was checked by B-mode USI. This allowed to locate electrodes on one belly muscle close to the navel. Reference electrodes were always placed between and laterally to the active electrodes (Fig. 1A). EMG signals were notch filtered (50 Hz), bandpass filtered (20–500 Hz), and sampled at 1 kHz. In order to reduce impedance between the electrodes and skin, the skin surface was cleaned. The EMG signal from the RA was used to assess the participant’s maximal voluntary contraction (MVC) and to control the level of abdominal muscle activation during isometric contraction.
The EMG electrodes placement (A) and position to maximal voluntary contraction assessment (B). Black rectangles on figure A show the transducer position during shear wave elastography measurement.
A PBU (Stabilizer Pressure Biofeedback-Chattanooga Group, USA) was used to help obtain 10% of MVC of the RA. The PBU was routinely placed under the lumbar spine, where the inferior edge of the device was aligned with the posterior superior iliac spine marks.
An Aixplorer ultrasound scanner (Product Version 12.2.0., Software Version 12.2.0.808, Supersonic Imagine, Aix-en-Provence, France), coupled with a linear transducer array (2–10 MHz; SuperLinear 10–2, Vermon, Tours, France), was used in the SWE mode to measure shear modulus of the LAM (obliquus external – OE, obliquus internal – OI, transversus abdominis – TrA) on both body sides. The probe was placed laterally to the navel and transversely to the long axis direction of the body (along the line of muscle fibers of TrA) (black rectangles on Fig. 1A). The transducer was not positioned within the direction of the muscle fibres for each muscle, because probe orientation had no effect on shear modulus of the LAM [15]. Before the measurements, the probe was adjusted to ensure parallel orientation of the OE, OI, and TrA fascial borders. A necessary quantity of hypoallergenic transduction gel was used to diminish the applied load to the skin, as pressure has a great impact on SWE measurements.
Shear modulus measurements were taken in the semi-supine position at rest and during isometric contraction. In the resting stage, the knees were in 90° flexion and the upper limbs placed along the sides of the trunk. The participants were asked to breathe comfortably, and the SWE image was taken at the end of normal expiration. Two separate SWE images were collected from both body sides. Each image was inspected for artefacts and/or missing values (unfilled regions within the elastic map). If any were present, the measurement was repeated.
In the isometric contraction stage, the MVCs were first calculated in a semi-supine posture with arms crossed and hands on the shoulders, knees in 90° flexion, and feet flat on a therapeutic couch. An examiner put his hands on the patient’s elbow joints and resisted the flexion force applied by the patient (Fig. 1B). Each patient was encouraged to push as much as possible for five seconds. This procedure was repeated four times with 30-second intervals. The maximum root-mean-square (RMS) EMG of RA over the four MVCs was considered the maximal activation of RA. Based on it, 10% of maximal activation of the RA for each participant was calculated. Next, the PBU was placed under the lumbar spine and inflated to 40 mmHg pressure, and the patients were asked to slowly press a PBU-inflated bag without motions of the lower legs and head. Two examiners simultaneously controlled the PBU value and EMG activity level to ensure the participant achieved 10% maximal activation. The consistency between PBU and EMG values was checked five times.
During the SWE measurements in the isometric contraction stage, participants were asked to (1) breathe comfortably; (2) slowly press the PBU-inflated bag using the lumbar spine until reaching the previously established PBU value (patients continuously control the pressure value on the PBU); (3) try to hold the proper PBU value; and (4) hold their breath and keep the PBU value. During point four, the SWE image was stored. In this way, three images were collected from the right and left sides of the body, separately. Each image was inspected for artefacts and/or missing values (unfilled regions within the elastic map). If any were present, the measurement was repeated. Examiners were blinded regarding patients’ personal data and characteristics of the scoliosis during the whole procedures.
Data analysis
Muscle elasticity was calculated from the images stored in the ultrasound scanner after collecting data from all participants. In each image, the Q-Box TM quantitative tool was used to quantitate muscle shear modulus. Three separate circles inside the fascial edge of each muscle were positioned, and the muscle elasticity within the circle was automatically calculated (Fig. 2). Mean values of elasticity from nine separate circles from three images in the isometric contraction state and six separate circles from two images in the resting state were taken for further analysis.
Shear wave elastography of lateral abdominal muscles.
Additionally, each muscle thickness was measured on the SWE images (Fig. 3). As in previous studies, the images were saved on an external drive in JPEG format and transferred to a computer where they were further processed using Photoshop software (Adobe Systems, Inc., San Jose, CA, USA). A detailed protocol for editing the images was presented in a previous study [16]. For further analysis, the mean thickness value of each muscle was derived as a mean value from three separate images in the isometric contraction state and two images in the resting state.
Lateral abdominal muscles and fat tissue measurements.
Due to the fact that analysis of LAM thickness in the adolescent population without body mass normalisation can lead to incorrect interpretations of study results [17–19], it was recommended that allometric-scaled OE, OI, and TrA thicknesses should be analysed. The allometric parameters necessary for the scaling procedure were from a recently published paper. For the OE, OI, and TrA, they were 0.88, 0.72, and 0.61, respectively [16]. To normalise the muscle size measurements to body mass, the following equation was used:
For the collected data, two indices were also calculated using the following equations: (1) thickness ratio = muscle thickness during contraction/muscle thickness at rest; (2) elasticity ratio = elasticity value during contraction/elasticity value at rest. Calculations using each equation were performed for OE, OI, and TrA on both body sides, separately.
Differences in basic data between right- and left-side scoliosis patients were examined using an independent sample t test. SWE and muscle thickness were analysed using one-way analysis of variance (ANOVA) for repeated measurements, where the convex and concave sides of scoliosis were a repeated measurement and the between-subjects factor was the group (right-side scoliosis vs. left-side scoliosis). The results are presented as a mean difference and 95% confidence interval (CI). For all analyses, the threshold of the p value considered as significant was set at ≤0.05. All statistical analyses were performed with the use of STATISTICA software.
Results
Participants
Participants with left-side scoliosis were significantly heavier by 18.3 kg (95% CI 1.4–35.3) and had a higher BMI value by 5.9 (95% CI 0.9–10.9) compared to participants with right-side scoliosis. With regard to age, body height, and Cobb angle, both groups were similar (Table1).
Characteristics of the participants
Characteristics of the participants
∗significant differences.
With regard to actual and allometric-scaled values of the LAM, the only significant main effect was found for the OE muscle (interaction), but detailed post-hoc Tukey's analysis revealed no significant differences in the examined population (Table2, Fig. 4). However, for the allometric-scaled OE thickness in right-side scoliosis, the difference between the convex and concave body side was at borderline significance (p = 0.06), suggesting thicker OE muscle on the concave side.
Results of repeated measures one-way ANOVA
Results of repeated measures one-way ANOVA
∗significant differences; 1concave side vs convex side; 2scoliosis direction (concave vs convex) x scoliosis type (right-side thoracolumbar scoliosis vs left-side thoracolumbar scoliosis).
Actual (left) and allometric-scaled (right) values of lateral abdominal muscles.
For SWE measurements in the supine resting position, the significant main effect was found for OE and TrA muscles (Table2). A detailed post-hoc analysis showed that in right-side scoliosis the OE muscle on the convex side was stiffer by 7.25 kPa (95% CI 0.14–14.3, p = 0.04) compared to the concave side. The OE muscle on the convex side in right-side scoliosis was also stiffer by 11.6 kPa (95% CI 3.66–19.6, p < 0.01) compared to the convex side in left-side scoliosis. Additionally, in left-side scoliosis, the TrA muscle on the concave side was stiffer by 7.84 kPa (95% CI 0.54–16.2, p = 0.04) compared to the convex side (Fig. 5).
With regard to the muscle elasticity ratio, the only significant main effect was found for the TrA muscle (main effect scoliosis direction). A subsequent post-hoc analysis showed that on the convex body side (in both types of scoliosis – right and left) during isometric contraction, the increase in stiffness of the TrA muscle was higher (p = 0.07) than on the concave side (Table2, Fig. 6).
Shear wave elastography values of lateral abdominal muscles.
Thickness and shear wave ratio of lateral abdominal muscles.
This was the first study in which LAM stiffness was measured in the supine resting position and during isometric contraction (as explained in the method section) in adolescent thoracolumbar scoliosis patients using SWE. The preliminary results have shown that during the resting stage in right-side scoliosis, the OE muscle on the convex side was stiffer than on the concave side, and this muscle was also stiffer compared to the convex side in left-side scoliosis. In turn, left-side scoliosis patients had a stiffer TrA muscle located on the concave side compared to the convex side. For the SWE ratio (contraction state divided by rest state), there were no significant differences, but a possible tendency was observed that during contraction the TrA muscle increased its stiffness on the convex side to a greater degree than on the concave side in both scoliosis directions.
From an anatomical perspective, the OE fibres are attached to the external surface and inferior borders of the last eight ribs and pass to the iliac crest, aponeurosis at the mid-clavicular line, xiphoid process, pubic crest, pubic tubercle, and linea alba. In thoracic and thoracolumbar idiopathic scoliosis, a rib hump is usually a hallmark seen from the back. A rip hump helps to determine the direction of the AIS, because posterior ribs on the convex side are pushed posteriorly (posterior rib hump), and anterior ribs on the concave side are pushed anteriorly (anterior rib hump). Thus, the OE muscle, due to its attachment to ribs, may be connected with scoliosis. Our study has shown that in right-side thoracolumbar scoliosis, the OE is stiffer on the convex side. This stiffness may restrict further chest rotation to the right (protective tension), as the OE muscle on one side helps trunk rotation to the opposite side. The OE muscle on the convex side may also be under stretching, and higher resting tension limited it. However, this possible pattern was not seen in left-side scoliosis, in which no side-to-side differences in OE stiffness were observed. Moreover, the convex OE stiffness value was higher in right-side scoliosis (≈25.5 kPa) compared to left-side scoliosis (≈14 kPa), whereas in both types of scoliosis the OE stiffness value (≈18 kPa) on the concave side was the same on average.
With regard to the deepest LAM muscle, our results showed that the TrA muscle on the concave side was stiffer compared to the convex side only in left-side scoliosis. The TrA fibres are attached to the inside part of the lower six ribs, the thoracolumbar fascia, iliac crest, inguinal ligament, linea alba, xiphoid process, and symphysis pubis. Their primary and well-known role is to maintain proper tone of the abdominal organs, exhale, and raise the intra-abdominal pressure needed for urination and defecation. Some other studies have shown that the TrA can increase (1) the stability of the sacroiliac joints [3] and (2) the thoracolumbar fascia tension [20]. Such an increase in thoracolumbar fascia affects intervertebral stiffness by transversal forces and takes part in controlling vertebrae rotation [21,22]. Some researchers have examined vertebrae rotation in connection to the AIS aetiology [23] and found that vertebral rotation measurements are important in the prognosis and treatment of the scoliotic spine [23–26]. Thus, the TrA tension partially affects the thoracolumbar fascia and may potentially contribute to vertebrae rotation inhibition. Barker et al. claim that tension generated in the abdominal muscles is transferred to the lumbar spine mainly by the middle layer of the thoracolumbar fascia [27,28], and, via them, the TrA may influence intersegmental movements [27]. The middle layer of the thoracolumbar fascia is also seen as a medial continuation of the TrA aponeurosis [29]. In our left-side scoliosis patients, the vertebrae have a tendency towards left rotation; thus, the transverse processes are positioned more anteriorly. This may explain the higher TrA stiffness on the concave side (≈20 kPa) and smaller TrA stiffness on the convex side (≈12 kPa). In the opposite, right-side scoliosis patients have an equal TrA stiffness value (≈14 kPa) on both sides. It may be that vertebrae in the lumbar region of the right-side scoliosis patients were less rotated compared to left-side scoliosis patients. This aspect was not measured in the study.
In a recently published paper, a shear modulus value of a right OE muscle was taken from 30 healthy adult volunteers (mean age 20), and the mean value was on average 7 kPA [12]. There are no other data on it. The TrA shear modulus value depends on research and varies from 4 kPa [12] to 13–15 kPa [13,14]. In our study, regardless of the body side, the OE stiffness was at least twice as high as results from the literature, whereas the TrA value was similar to those from Hirayama et al.’s (9,10) studies (with the exception of TrA stiffness on the concave side in left-side scoliosis) and much higher than in the study by MacDonald et al. (24). At this stage of knowledge, it is difficult to draw conclusions on TrA stiffness value, as even in almost the same examined populations [12–14] shear modulus differs highly. Thus, more studies on LAM shear modulus in healthy and scoliotic populations are warranted.
It is worth noting that this study was conducted on only 12 adolescents, in which one type of spinal deviation was analysed. Thus, a large case-control study is needed to assess more probable patterns of LAM elasticity in different types of spine morphologies. It also seems necessary to provide reference data on LAM elasticity in healthy adolescents to better understand changes reported here. Lastly, the LAM elasticity measurements were taken on the anterolateral part of the abdomen at the navel level, but regional differences in LAM morphology are confirmed [30]. This suggests variation in LAM function between regions.
In conclusion, some changes in LAM elasticity of thoracolumbar scoliosis were observed. A different pattern of LAM stiffness in different curve directions may be expected. The OE and TrA shear moduli are the most sensitive to change in adolescent spinal deformation.
Footnotes
Conflict of interest
None to report.
Ethical approval
The Local Ethics Committee of The Jerzy Kukuczka Academy of Physical Education in Katowice approved this study. Participants provided written informed consent prior to data collection.
Funding
The study was financed by the Polish National Science Centre (decision no. 2016/23/D/NZ7/02003).