In this study, taking into account 25% mean difference in FRR for bilateral erector spinae between the two groups as well as α-level of 0.05 (type I error) and β-value of 0.10 (type II error), 25 patients with NSCNP and 25 healthy volunteers were recruited. Taking into consideration a type I error of 0.05, type II error of 0.1 (power=90%), and σ of 11, 25 adults with non-specific CNP participated in this quasi-experimental study. They were recruited from among those admitted to physical therapy and orthopedic clinics in Tehran province. The patients were examined by an experienced physiotherapist and orthopedic surgeon and diagnosed with NSCNP. Based on the inclusion criteria, the study was conducted on:
1) Patients aged 20 to 45 years.
2) Cases who had experienced persistent pain for at least three months without a specific pathology (e.g. no history of degenerative disease, discopathy, radiculopathy).
3) Cases who without a history of any type of head, shoulder, and lumbar pain during the past two years.
Based on the exclusion criteria, the following cases were excluded from the study:
1) Patients with a pain score more than 5 cm based on a visual analog scale (VAS).
2) Cases who had participated in a neck rehabilitation program during the past three months.
3) Patients with a history of systemic disease and any type of cervical or shoulder trauma.
The method was approved by the Research Ethics Committee of School of rehabilitation- Shahid Beheshti University of Medical Sciences research committee (approval ID= IR.SBMU.RETECH. REC.1398.405). We explained clearly all steps of study to subjects and wrote informed consent before beginning the study.
All cases were tested in a 45- min session in a laboratory. Before the test, pain severity was determined using VAS. In addition, the subjects completed NDI questionnaire. Then, each subject was asked to sit on an adjustable stool with hips and knees at an angle of 90°, with feet on the floor positioned shoulder width-apart, with arms relaxed by their side, looking at an eye-level point. Because of the effect of different seated posture on cervical spine alignment, during the test each subject was asked to maintain a neutral lumbar lordosis i.e. the midpoint between full flexion and extension that was determined by a physiotherapist [24, 25]. Therefore, seated posture was standardized in a neutral lumbar lordosis. Furthermore, each subject was asked to perform neck full flexion and extension and the mid-position was set as the neutral head position [25, 26]. The upper thorax was stabilized tightly by a belt at the level of T1-T7 spine to immobilize this region. Before the test, all the cases had become informed about the protocol of the test and the starting neutral position was determined. Then, each subject was asked to maintain a neutral starting position for 4s (phase 1), complete cervical full forward flexion for 4s (phase 2), sustain full flexion for 4s (phase 3), and perform re-extension to starting position for 4s (phase 4) . The protocol of the test was performed in three trials. A digital metronome was used to control the effects of speed and cumulative daily loading on FRR, and all the steps of the test were carried out in the morning. All parts of the protocol was performed by an experienced physiotherapist. SEMG signals in CES muscles were recorded bilaterally and simultaneously using a electrogoniometer during the test, before and after the intervention.
Electromyography (EMG) signals were recorded using a SEMG device (Datalog, UK). Bipolar disposal EMG surface electrodes (Ag–AgCl) were applied on the skin, in line with muscle fibers. Inter-electrode distance was set to be 2cm and electrode leads were taped on the skin . SEMG data were recorded in right and left CES muscles, 2- cm lateral to the spinous process of C-4 [6, 28]. In order to decrease the impedance, the subjects’ skin was shaved, abraded, and washed by water before placing the electrodes. A ground electrode was applied to the left wrist . The sampling rate of EMG device was 1,000 Hz and the band-pass filter frequency was between 20-480 Hz. Neck flexion and extension angles were recorded by an electrogoniometer sensor (sampling rate of 1,000 Hz, Biometrics) synchronized with EMG data.
NDI questionnaire is used to show the functional changes following the treatment of CNP patients . All the patients completed the Persian version of this questionnaire to assess self-reported disability, both before and after the eight weeks of intervention. It consists of 10 items, with seven items associated with activities of daily living (ADL), two items associated with pain, and one item associated with concentration. All the items of the questionnaire are scored from 0-50, with a higher score representing a higher level of disability [29, 30]. The Persian version of NDI is a reliable and valid instrument to assess disability and functioning in CNP patients (Cronbach's alpha: 0.88, Intraclass correlation: 0.90) .
VAS (which is scored from 0-10 cm, with a higher score indicating a more pain) was used to measure the intensity of neck pain. Patients were asked to determine a point along this line based on the severity of their pain during the last month. According to Bijur et al., VAS is a valid and reliable instrument to measure the severity of pain .
Low-load craniocervical flexion exercises (CCFE) were performed by the subjects in crook lying position to carry out deep flexor muscles training (longus capitis and longus colli, the main cervical stabilizer muscles) presented by Jull et al. . The patient's head was positioned horizontally, without any pillow under the head. The uninflated pressure sensor cuff was placed under the patient's neck and was inflated to reach a pressure of 20 mm Hg. Both the patients and the physiotherapist received visual feedback. In the first step, each patient was taught to activate the deep muscles rather than superficial muscles and perform the exercise slowly and precisely, not strongly. In the second phase, the patients were asked to perform CCFE (as, if saying yes) and try to achieve a sequence of five targets, with five increments of 2 mmHg, to finally reach 30 mmHg. Moreover, at the beginning of the exercise, the patients’ tolerance level was determined; it is the level at which a subject could maintain the contraction for five seconds. The treatment sessions began with a preset level and increased progressively in a range of motion (ROM), in amount of pressure and in time to maintain contraction from 5 to 10 seconds. Each level consisted of three repetitions, each lasting 10 seconds, in a predetermined level of pressure [33, 34]. Throughout the treatment sessions, through touching the patient's neck, it was tried to ensure that the patient was not using rapid movements as well as movements by the superficial neck flexor muscles, such as sternocleidomastoids and scalenes. The exercise was performed for eight weeks at home, 30 minutes daily, and three sessions per week (24 sessions) under the supervision of a physiotherapist. In addition, the patients were advised to stop the practice and report it to the physiotherapist when the exercise exacerbated their pain.
Data and Statistical Analysis
Raw EMG data were collected, rectified, and filtered. Root mean square (RMS) values (time window of 50 ms) were obtained to determine the parameters of FRP. FRR values were calculated through dividing the maximum EMG in the re-extension phase by the average EMG in the relaxation phase . Inverse FRR ≤ 40% was used to determine the occurrence of FRP . These measurements were carried out both before and after the intervention.
Furthermore, the distributions of VAS, NDI, FRR, and inverse FRR before and after the intervention were calculated using paired t-test. Student's t-test was used to investigate the association between the average change of VAS, NDI, and FRR and gender and duration of pain. The comparison of inverse FRR frequency after the intervention was performed using McNemar's test. The standard error of measurement (SEM) and smallest detectable change for individual patients (SDCind) was calculated using ( ) and , respectively. Data were analyzed in the Stata software (version 14) and p<0.05 was considered statistically significant in all the statistical tests.