Design and participants
To analyze the effect of direct vibration on trunk muscles, we assessed the maximal and rested contraction of trunk muscles after applying direct vibration during stabilization exercise. We measured the patients’ anatomical and electrophysiological features using a randomized comparison study design. We measured contracted and rested muscles using ultrasound and surface electromyography. The participants performed preprogrammed spinal stabilization exercises and then scheduled three visits to measure the contraction of trunk muscles. The protocol was approved by the institutional review board of Korea University Anam Hospital in Korea. Each participant signed a written informed consent before participating in the study. All procedures were performed in accordance with relevant guidelines including a ‘declaration of Helsinki’.
Participants were enrolled in this study based on the following inclusion criteria: non-specific CLBP for at least three months (11), and aged between 20 and 60.
Participants were excluded if they displayed symptoms of recognizable pathologies such as infection, malignancy, inflammatory disease, structural deformity, overt neurologic signs, history of trauma, abdominal or lumbar surgery, and pregnancy (suspected or confirmed). The exclusion criteria from the clinical trial were: deterioration of back pain, severe physical or physiological damage, refusal of patient, drastic reduction of compliance, absence on three or more sessions, and any intervention of other treatments.
We used random allocation because the subjects’ task or living environment could not be controlled. We computerized randomized assignments before enrollment. The subjects were enrolled in the order they were diagnosed with non-specific CLBP and approved for participation in this study.
Participants (n = 62) were randomly divided into two groups. Each group received a protocol including twelve sessions of treatment over four weeks and a follow-up visit 8 weeks after the end of the treatment. A flow chart describing the number of subjects considered for this clinical investigation is shown in Fig. 1. The study was conducted per the guidelines of the consolidated standards of reporting trials (CONSORT).
Intervention and device
The spinal stabilization exercise protocol for the rehabilitation of CLBP patients is based on commonly advocated spinal stabilization exercises (12–14). The spinal stabilization exercise program consists of five different exercise types repeated in each session: upper-body extensions, alternate arm and leg lifts, alternate arm and leg extensions, diagonal curl-ups, and curl-ups. The subjects exercised three times a week in our gym under the supervision of experienced physical therapists. They helped the subjects to maintain accurate postures for 10–15 seconds and encouraged the completion of the 30 minutes exercise protocol, including warm-up, cool-down, and proper rests.
We designed and manufactured a vibration device (330x180x120mm) for this study only. The target functions of the device were as follows: adjustable frequency from 0 to 100Hz, adjustable intensity (0-4mm, peak acceleration of 0-6g) in three steps, gradual start and end of vibration for compliance, exterior and materials designed to maintain contact on the body surface above the lumbar paraspinal muscles and increase the vibration transmission efficiency, interior structured to prevent loss of generated vibration, and immediate shutdown and removal during application.
Outcomes
We obtained ultrasonographic images of the rectus abdominis (RA), the external oblique (EO), the internal oblique (IO), the transversus abdominis (TrA) muscles, the lumbar erector spinae (LES), and the lumbar multifidus (LM) muscles at L4/5 level during both resting and activation states. We determined the patients’ position and transducer location based on previous studies (15, 16). We measured the thickness of the contracted EO, IO, and TrA during an abdominal drawing-in maneuver (ADIM), of the RA during curl-ups, and of the LES and LM during maximal resistive lumbar extension on prone (17). We calculated the change in muscle thickness due to contraction for all muscles using the following equation: [thickness ratio = thickness contracted / thickness rest]. The transducer was vertically placed for the measures. We placed the inferior border immediately above the umbilicus and moved laterally from the midline. For the LES and LM muscles, we placed a curvilinear transducer longitudinally along the spine just lateral to L4 spinous process to identify the L4/5 zygapophyseal joint (12).
We recorded the muscle activity of the bilateral EO, IO, TrA, RA, LES, and LM muscles using surface EMG. We determined the patients’ position and the testing protocol based on previous studies (18). We positioned electrodes on the abdominal muscles (EO, IO, TrA, and RA muscles) according to McGill (19). We placed the electrodes on the LES muscles by positioning them 4cm laterally from the L4 spinous process at the level of the iliac crest (20). For the LM muscles, we placed the electrodes above the posterior superior iliac spine on the spinous processes of the lumbar and sacral vertebrae (21, 22). We verified the quality of the EMG signal to ensure correct electrode placement (23), and used reference voluntary contractions (%RVCs) to normalize muscle activation. We collected the raw EMG data using the wireless EMG system TeleMyo 2400R which was connected to a Noraxon Myosystem 1200 unit (Noraxon, Scottsdale, AZ, USA) with a sampling rate of 1000 Hz. The mean activities of the muscles were expressed as the percentage of RVC. We calculated the ratios of muscle activation as the % RVCs of IO relative to RA (IO/RA ratio), TrA relative to RA (TrA/RA ratio), and LM relative to LES (LM/LES ratio).
We measured the pain intensity using a visual analog scale (VAS) and evaluated the LBP-correlated functional disability using the Oswestry Disability Index (ODI). The pain and disability levels were assessed at T0 (initial), T1 (after four weeks), and T2 (after 12 weeks).
Statistical analysis
We performed all the statistical analyses with SPSS version 22.0 (SPSS Korea Data Solution Inc., Seoul, Korea). We used Wilcoxon’s signed-ranks tests to compare the muscle thickness ratio, the sEMG data, the disability level, and the pain intensity of the subjects before and after training; and Mann-Whitney U tests to identify any significant difference in the baseline data between the two groups. We also used a repeated-measures analysis of variance (rm-ANOVA) to determine the presence of a temporary interaction of the three different measures within each group. We presented the three epidemiologic and measured variables and two explanatory variables of symptom in a regression analysis of each group. Statistical significance was considered at p < 0.05.