Neck proprioception assessment with a laser beam device: reliability in participants without neck pain and differences between participants with and without neck pain

Proprioception deficits have previously been reported in patients with non-specific chronic neck pain (NSCNP), with a comprehensive and valid battery of tests still required. This study aimed to investigate the test-retest and inter-rater reliability of cervical proprioception in participants without NSCNP and to examine differences in proprioception between participants with and without NSCNP. Twenty participants without NSCNP and 20 age- and sex-matched participants with NSCNP were recruited. Proprioception tests were sequentially performed in random order, in four head-to-neutral movement directions (starting positions at mid-flexion, mid-extension and mid-right/mid-left rotation head-neck positions and end position at neutral head-neck posture) and two head-to-target movement directions (starting position from neutral head-neck posture and end positions at right and left 45° rotation), with a laser beam device secured onto their forehead. Participants performed all tests in sitting at a 1-m distance from a whiteboard. The average deviations of the laser beam mark from set targets marked on the whiteboard represented proprioception deficits. The two-way random, absolute agreement model of the intraclass correlation coefficient (ICC), the standard error of the measurement (SEM) and the smallest detectable difference (SDD) were used as measures of reliability. Between-group differences were examined with the independent samples t test. The reliability of the laser beam device in participants without neck pain varied from poor to good. The following tests demonstrated good reliability: test-retest ‘Head-to-neutral from flexion’ (ICC: 0.77–0.78; SDD: 5.73–6.84 cm), inter-rater ‘Head-to-neutral from flexion’ (ICC: 0.80–0.82; SDD: 6.20–6.45 cm) and inter-rater ‘Head-to-neutral from right/left rotation’ (ICC: 0.80–0.84; SDD: 5.92–6.81 cm). Differences between participants with and without NSCNP were found only in head-to-neutral from flexion (4.10–4.70 cm); however, those were within the limits of the SDD values of the HtN from flexion test. The laser beam device can be reliably used in clinical practice only in the aforementioned head-neck movement directions, based on the findings of the present study. The between-group differences noted involved only the head mid-flexion to neutral test, possibly denoting proprioception deficits only in this movement direction, for reasons that require further evaluation.

Page 2 of 9 Ntenezakos et al. Bull Fac Phys Ther (2021) 26:36 Background Chronic neck pain is one of the most common musculoskeletal disorders, defined as the presence of symptoms in the cervical spine longer than or equal to three months, without any specific aetiology linked to symptoms. The usual causes of cervical pain may occur as a result of an injury, prolonged incorrect posture and/or movement of the cervical spine, nerve root pressure, muscle sprains and whiplash-associated disorders [1]. The symptoms associated with NSCNP include headache, migraine, vision disorders, severe pain, range of movement limitations, numbness along the shoulder and upper extremities, and a feeling of fatigue [2]. Proprioception deficits in the cervical spine have also been identified [3]. Proprioception is defined as the sense of movement (kinesthesia), sense of orientation in space (joint position sense) or sense of effort [3,4]. Proprioception sensory input is provided primarily in the periphery and is integrated in the central nervous system, contributing to static control, stability, and other conscious senses [5].
It has been indicated in a recent systematic review that a number of different measurement tools and tests have been used to assess neck proprioception, commonly by assessing deviation from set targets, a method using joint position error (JPE) as an outcome [6]. A recent study that conducted a factor analysis of seven different sensorimotor control (SMC) tests concluded that JPE tests might be assessing a unique proprioception aspect, separate from other aspects that the other six tests examine [7]. Regarding the measuring instrument utilized to assess neck proprioception JPE, the laser beam device has been used in many studies [8][9][10][11][12]. However, further exploration of the methods of administration of cervical proprioception JPE tests is required, considering that previous reliability studies with the laser beam device presented non-uniform findings, with intra-rater reliability values ranging in participants without NSCNP from moderate to good [8,9], or moderate to excellent [10]. Similarly, in participants with NSCNP reliability JPE values ranged from moderate to excellent [10], to moderate [11] and in mixed group of participants with and without NSCNP from moderate-poor [12]. In addition, a systematic review with meta-analysis that examined the known-groups validity of these tests between participants with and without neck pain has not been invariably demonstrated, with overlapping JPE test scores revealed in several test directions that were examined, resulting in non-significant differences between populations [6]. Therefore, the primary aim of this study was to investigate the reliability of a laser beam device to assess cervical proprioception in healthy participants. Furthermore, a secondary aim was to perform a comparison of cervical proprioception between two groups of age-and sexmatched participants with or without NSCNP, to assess the known-groups validity of the cervical proprioception assessment methods employed.

Participants
A total of 20 healthy participants were recruited, from the staff and students of Metropolitan College. The sampling method used was convenience sampling. Additionally, data from 20 participants with NSCNP in a previous study [13] were utilized. The participants were age-and sex-matched between the two groups, with the 'individual matching' metho d [14], to avoid confounding related to either of those anthropometric variables, previously shown to affect proprioception measures [15] or sexrelated cervical muscles' orientatio n [16].
The inclusion criteria for participants with NSCNP were age between 18 and 65 years and neck pain history of more than or equal to three months. All participants with non-specific chronic neck pain had previously consulted a medical practitioner (orthopaedic doctor) and had a diagnosis following a thorough clinical examination. They all had a history of either repeated or continuous episodes of pain limited to the neck area, with no radicular pain symptoms in the upper extremities, collectively exceeding a time-period of 3 months [1]. Participants with NSCNP with upper quadrant pathology or spinal dysfunctions other than neck pain were excluded. Admission criteria for healthy participants without NSCNP were as follows: age between 18-65 years and no history of neck pain. Exclusion criteria for all participants were clinical disorders that could affect proprioception, such as obesity, osteoporosis, spinal arthropathy, spinal disorders, spinal surgery, rheumatic diseases, cancer and pregnancy.

Ethics approval and consent to participate
The study was approved by the Ethics Committee of Metropolitan College (approval number 208/2017). All procedures were performed in compliance with relevant laws and institutional guidelines and the rights of participants were protected at all times, according to the Declaration Keywords: Cervical spine, Sensorimotor, Position sense, Proprioception, Neck pain, Disability of Helsinki. All participants signed an informed consent form prior to participation in this study.

Equipment and material
A laser beam device was constructed, by fixing a standard laser pointer to a commercially available head-mounted elastic band for action cameras. The laser pointer was secured with duct tape in the space available for mounting the action camera onto the strap. The laser pointer was positioned exactly in the centre of the participants' forehead and was parallel to the ground when the device was placed on the participants' head. In addition to the laser beam device, a blindfold was also used, to occlude the participants' vision, therefore eliminating all visual feedback during the proprioception experiment. The participants with NSCNP were asked to report their usual daily life average pain intensity and their current pain intensity before the experimental proprioception testing procedure on a pain visual analogue scale (P-VAS). Participants with NSCNP were asked to mark their pain intensity level with a point on a 10-cm straight line extending between 0 (no pain) and 10 (maximum pain). The validity and reliability of the VAS for the evaluation of chronic musculoskeletal pain has been previously established [17]. The participants with NSCNP were also asked to fill in the Neck Disability Index (NDI), a functional impairment self-report subjective rating scale in relation to neck pain. Ten categories of questions are assessed, with six possible answers for each category and scores ranging from 0 (no disability) to 50 (maximum score). The score is then converted into a percentage score. The validity and reliability of this scale in evaluating cervical dysfunction in all categories of patients with cervical pain has been previously established [18]. The cross-culturally adapted version of the NDI scale in Greek was utilized [19].

Procedure
All participants completed a questionnaire with demographic characteristics. Participants with NSCNP additionally completed the NDI and P-VAS scales. Participants with NSCNP had also completed the test-retest and inter-rater reliability of the proprioception methods involved [13]. All questionnaires and consent forms were collected by one of the principal investigators (ZD), with assessors (ND and MM) remaining blinded to participants' grouping.
Each participant was then asked to sit on a stable chair at a set distance of 1 m away from a whiteboard to perform the proprioception tests. The measurement position was standardized for all participants by placing their knees and hips at 90°, their pelvis and low back resting firmly against the back of a chair and their hands resting on their knees. All measurements were performed in a Physiotherapy Lab, where external auditory distractions were maintained at a minimum and temperature was kept constant, at 23 °C. The participants' vision was occluded throughout the experiment with a blindfold and the laser beam device was secured on their forehead ( Fig. 1), following standard methodology [6,7].
Two different types of JPE tests were performed in relation to their starting and end positions: four head to neutral (HtN) and two head to target (HtT) tests. Prior to the initiation of each test, the participants were asked to assume their relaxed, neutral head position (NHP) and memorize it. The examiner marked the exact point (point 0) where the laser beam was projected on the whiteboard with a black marker. When participants felt confident about memorizing this position, each test was initiated. Each test was administered in random order, to eliminate any order effects.
For the HtN tests, the examiner passively moved each participant's head in the following directions: right rotation, left rotation, flexion and extension, at approximately 45°. From each of these positions, participants were instructed to actively return their head to the NHP and the new point where the laser beam projected on the whiteboard was marked (point A). After the completion of this test, the examiner used a tape measure to record the distance between points 0 and A. This distance/deviation represented the proprioception deficit (JPE) of each participant for the examined movement. Five repetitions were performed for each movement direction and five scores of cervical proprioception JPE were obtained for each movement direction. For the two HtT tests, the chair was placed at a 45° angle to the whiteboard (rotated either at 45° to the right or to the left). Maintaining the same conditions, the participants' heads were passively rotated at 45°. This was the target position and was marked on the whiteboard with a black marker (point 0). Participants were asked to memorize this position and their heads were returned to their neutral positions. Immediately after, they were asked to rotate their head to the target position and the new laser beam position was marked again on the whiteboard (point A). Again, five repetitions were performed for each movement direction and five scores of JPE were recorded for each HtT test movement direction. For all tests performed the average deviation from the set target of the first three repetitions and all five repetitions represented the overall proprioception deficit (JPE) per movement direction [20].
The whole procedure was performed by examiner A and after a 5-min break, it was repeated by examiner B. Then, participants were asked to leave and return after 45-60 min, for the same procedure to be repeated by examiner A. During this period, patients were asked not to smoke, eat or participate in any activities that could affect their neck condition. The examiners were blinded to each other's measurements.

Statistical analysis
The distribution of all variables was checked using the Kolmogorov-Smirnov test. Test-retest and inter-rater reliability was examined using the second model of intraclass correlation coefficient (ICC) (two-way random, absolute agreement). The averages of the first three and all five trials per test were used for the reliability and between-group differences analyses. ICC values of <0.50, between 0.50 and 0.75, between 0.75 and 0.90 and >0.90 are considered to indicate poor, moderate, good and excellent reliability, respectively [21]. The associated standard error of measurement (SEM), (SEM=SD√1-ICC) and smallest detectable difference (SDD), (SDD=1.96√2SEM) were calculated for each ICC [22]. The smallest detectable difference (SDD) represents the minimum amount of change that can be considered a real change, beyond the error of the measurement [22].
Independent samples t-tests were performed to test for differences between the two groups. An a priori power calculation was performed in the G*Power 3.1.9.7 for Windows program [23] to establish the required sample size for the known-groups validity part of this study. It was calculated that to detect a between-group mean difference of 4 cm with a common standard deviation of 4.5 cm, a-level (2-sided) of 0.05 and 0.80 power level, the required sample would be n=20 per group. Additionally, an a priori power calculation was performed in an online web-based sample size calculator for reliability studies and for an expected ICC=0.85, measurements taken at two occasions, a lowest acceptable ICC=0.55, a-level (2-sided) of 0.05 and 0.80 power level, the required sample would be n=21 (https:// wnari fin. github. io/ ssc_ web. html). The significance level was set at p=0.05 for all comparisons. The Statistical Package for Social Sciences (SPSS, version 26.0) was used for all data analyses.

Results
The demographic characteristics and all proprioception variables for each of the groups of participants, and the symptoms variables of participants with NSCNP conformed to a normal distribution (p>0.05), according to the Kolmogorov-Smirnov test.
Demographics of the two groups examined and the symptoms of participants with NSCNP are presented in Table 1. No significant differences were identified in the between-group comparisons of the demographic characteristics of participants, using the independent samples t-test.
Reliability in healthy participants was calculated from the average of the first three, as well as of the five repetitions for each movement direction, to examine whether higher reliability levels could be achieved with either of those averaging methods. The test-retest reliability level in participants without NSCNP varied from low to good (Table 2). Specifically, the test-retest reliability of HtN from flexion (ICC: 0.77-0.78; SEM: 2.07-2.47 cm; SDD: 5.73-6.84 cm) was good, with either of the two methods of averaging repetitions. In addition, the test-retest reliability of HtT at 45o right . For the remaining test directions, the test-retest reliability was low. The inter-rater reliability level in participants without NSCNP also varied from low to good (Table 3). The inter-rater reliability in three out of six tests (HtN from right rotation, left rotation and flexion) was good (ICC: 0.80-0.84; SEM: 2.14-2.46; SDD: 5.92-6.81 cm), with either of the two methods of averaging repetitions. Also, inter-rater reliability of HtN from extension (ICC: 0.66; SEM: 3.35 cm; SDD: 9.27 cm) and HtT at 45° left rotation (ICC: 0.59; SEM: 2.27 cm; SDD: 6.28 cm) were moderate, when the average score of all 5 repetitions was considered. However, the inter-rater reliability in the HtT at 45° right rotation test was low with either method of averaging repetitions.
Differences in proprioception JPE values between participants with and without NSCNP were not significantly different (Table 4), except for the HtN from the flexion test (p<0.05). However, the between-group differences for this test were between 4.10 and 4.70 cm, within the limits of the SDD values of the HtN from the flexion test.

Discussion
This study assessed the within-day reliability (test-retest and inter-rater) of a laser beam device with a battery of six repositioning accuracy tests (HtN and HtT) in  participants without NSCNP. This device is low-cost and tests performed with this evaluation tool are not timeconsuming compared to other tools that evaluate cervical proprioception. The methodology used for assessing cervical proprioception in participants without NSCNP demonstrated adequate test-retest reliability in three out of six tests (HtN from mid-flexion, HtT at 45° right and at 45° left rotation) and adequate inter-rater reliability in five out of six tests (all, apart from the HtT at 45° right rotation). The average of five repetitions over three repetitions towards each test direction is recommended, based on our test-retest and inter-rater reliability data, as ICC values tended to increase and SEM and SDD values tended to decrease when more repetitions were used. For the participants with NSCNP included in the current study, following the same experimental procedure, moderate to good same-day test-retest (ICC: 0.56-0.81, SEM: 2.09-5.01 cm, SDD: 5.79-13.87 cm) and moderate to excellent inter-rater reliability (ICC: 0.59-0.92, SEM: 1.75-3.27 cm, SDD 4.85-9.05 cm) were also reported [13]. The test-retest reliability level in previous cervical proprioception studies utilizing the laser beam as a measuring instrument was variable, ranging from moderate to good [8,9] or moderate to excellent [10] in participants without NSCNP. Specifically, in a recent study examining test-retest reliability (n=24, same day-12h interval) of 4 HtN tests and calculating JPE by averaging 5 repetitions per test, the reliability values of HtN from flexion (ICC:0.75, SEM: 0.49 cm), extension (ICC: 0.81, SEM: 0.64 cm), and right and left rotation (ICC: 0.64, SEM: 0.63 cm) [8]. In another similar previous study that examined the effect of 10 different averaging methods on reliability (n=40, same day-1h interval), it was reported that when the HtN from rotation test was executed in its original form (10 trials for each test), the reliability level was moderate to good (horizontal plane ICC: 0.80, SEM: 0.9o; vertical plane ICC: 0.52, SEM: 0.7 o) [9]. In a more recent study examining test-retest reliability (n=33, same day), 4 different tests were administered bilaterally (HtN rotation tests, trunk to target rotation tests, HtT rotation tests and figure of eight tests), calculating JPE by averaging 6 repetitions per test [10]. Reliability values for the healthy participants were good to excellent for the within-session intra-rater reliability (ICC: 0.75-0.93, SEM: 0.5-1.2°); however, for the between-session intra-rater reliability, they were moderate to good (ICC: 0.55-0.83, SEM: 0.7-1.1°) [10]. Although the SEM values were in general of low magnitude, the authors of that study remained sceptical about the clinical utility of the tests, since error was masking the JPE differences identified between participants with and without NSCNP [10]. In another study, the intra-rater reliability values in 41 participants with NSCNP (same examiner repeating the procedure after 1 h, HtN from maximal cervical rotation test) was moderate (ICC=0.68) [11]; however, no SEM values were reported in parallel. Under a different methodology set-up, with the laser beam positioned 1 m behind the participant, and the laser beam projected to a cm ruler attached to a cap worn by the participants, 2 different studies reported from poor to good (ICC: 0.48-0.80) [24] to poor-moderate (ICC: 0.02-0.66) results for three different HtN tests (from rotation, flexion, and extension) [12].
In all the above studies the HtN tests were conducted by a return to neutral from an end of range movement in each test direction. The methodology we employed was slightly different, as participants were requested to match the neutral position upon return from a mid-range respective movement direction, to provide less end-ofrange feedback to participants; however, this variation in test performance did not provide substantially different reliability estimates. Possible reasons for increased reliability may have been the larger samples (n=33-41) recruited in some of the previous studies [9,10]. Additionally, it was observed that when proprioception JPE was calculated from tests that included at least 5 repetitions, ICC and error values improved [9], whereas moderate-low reliability was reported in studies that only three repetitions per test were performed [9,12]. Additionally, the known-groups validity of the proprioception assessment methods used requires rigorous assessment. To this end, the present study further examined the differences in JPE between two age-and sex-matched populations of participants with and without NSCNP. Statistically significant between-group differences were observed only in the HtN from the flexion test. However, these differences were within the SDD values for this particular test direction; therefore, they cannot be considered clinically significant. The results of this study coincide with those of three previous studies that included patients with NSCNP and healthy participants, where statistically significant between-group JPE differences were reported only in HtN from flexion [25][26][27]. In contrast, other studies [10,11,28,29] have reported that healthy participants have significantly less JPE compared to participants with NSCNP for HtN from rotation tests, also.
Considering the between-group differences in JPE identified in this study (in HtN from flexion) it seems that JPE was primarily affected in sagittal plane movements in participants with NSCNP. However, forward head posture or motor control between-group differences were not examined in this study. Therefore, as a hypothesis to be tested in future studies, it is proposed that alterations in neck posture, such as forward head posture [30], may sequentially lead to proprioception acuity diminution related to this particular plane of movement [31,32] in conjunction with pain presence. Motor control exercises that are effective in reducing symptoms in patients with NSCNP [33] may also be re-training the proprioception acuity of the cervical spine [34], possibly contributing in parallel to the therapeutic effect.
As a limitation to this study, no separate familiarization session was included, which could have improved the reliability level of the tests employed, and is proposed to be included in the future. Furthermore, the vertebral level of pain in participants with NSCNP recruited was not examined, as in NSCNP the distribution of pain is widespread. However, the region of cervical pain may affect JPE values, especially if located in the upper cervical spine [35]. Future studies may need to classify participants with NSCNP according to their primary area of symptoms, if possible. In addition, the participants recruited for this study constituted a sample of convenience. Future investigations could recruit a larger sample of participants with NSCNP, following certain neck pain categorization criteria [36], to provide more population-specific results. Finally, psychosocial characteristics of patients with NSCNP (anxiety and fear) reported to correlate with disability levels [37,38] and JPE values [10], require further study.

Conclusions
In a group of healthy participants without NSCNP, the laser beam device JPE tests demonstrated variable testretest and inter-rater reliability. Therefore, these can be used with confidence in clinical practice only in particular testing directions with the least measurement error. The between-group differences between participants with and without NSCNP involved only one of the six movement directions examined. Future diagnostic and therapeutic studies are required to confirm these findings in a larger sample of participants and subgroups of neck pain patients with a range of pain, disability, and psychological characteristics.