Age-related kinematic behavior on fast head-neck rotation target task in non-painful individuals aged from 8 to 85 years old

Kinematic behavior during fast cervical rotations is a useful parameter for assessing sensorimotor control strategies in neck-pain patients. However, in non-painful adults and children, the influence still needs to be explored. We assessed the impact of age on sensorimotor control strategies with computed kinematic variables (total time, mean speed, acceleration, deceleration, overshoots, maximum, minimum speed) during fast rotation task in 80 volunteers stratified in four different age-groups: Children, Young and Old adults and Seniors. Significant effects in Children and Seniors were observed for total time, mean speed, acceleration and deceleration. The test allows to discriminate sensorimotor control strategies between different age groups. Age must therefore be considered as a key parameter when analyzing results


Introduction
The lifetime prevalence of neck pain is almost 70%, including pre-and adolescent patients, and this increases with age up to the age of 60 (Blanpied et al., 2017;Stahl et al., 2004).
Nevertheless, underlying causes of neck pain are poorly understood. Because head rotation is a regular movement performed during everyday activities, this movement represents one of the issues that are of interest in patients with neck pain (Dugailly, Coucke, Salem, & Feipel, 2018).
Interestingly, previous studies supported that fast-rotational movements of the head-neck complex may be applied in clinical assessment and/or therapeutic management in patients with idiopathic neck pain (de Zoete, Osmotherly, Rivett, Farrell, & Snodgrass, 2016;Descarreaux, Passmore, & Cantin, 2010;Roijezon et al., 2010;Sarig Bahat, Takasaki, Chen, Bet-Or, & Treleaven, 2015). In view of these reports, fast axial rotation of the neck represents a critical motion to assess.
A target test using the DidRen Laser can be applied in response to real visual targets to standardize a head rotation motor task about 30° (Hage & Ancenay, 2009;Sarig Bahat, Sprecher, Sela, & Treleaven, 2016). Firstly, this test is believed to rely on the integrity of neuromusculoskeletal structures and of the sensorimotor control system (Treleaven, 2017) i.e. input from the visual, vestibular and proprioceptive systems, particularly the richly innervated cervical spine, which controls head and eye movement and postural stability (Treleaven, 2017). Secondly, the axial head rotation about 30° will preserve the motion from the strain of the passive system (joint capsules, facets, intervertebral disks and ligament) (Panjabi, 1992).
Target-directed head-neck complex movements are fundamental components of individuals' daily activities and usually require high levels of motion speed and accuracy. Speed-accuracy tradeoff could be considered as a "signature" of the decision process (Heitz, 2014). This varies according to which motion behavior is emphasized: accuracy or speed (Zhang & Rowe, 2014)] during, for author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint instance, an aiming task that encompasses amplitude of the movement, the size and position of the target (Descarreaux et al., 2010;Paul M Fitts, 1954;Heitz, 2014;Zhang & Rowe, 2014).
Lack of sensorimotor control of head-neck complex can be explained (Passmore, Burke, & Lyons, 2007;Uthaikhup, Jull, Sungkarat, & Treleaven, 2012) by degeneration of vestibular, visual and neuromuscular functions due to aging or even neck pain in adults and due to immaturity of the central and peripheral nervous and musculoskeletal system in healthy children (Assaiante, 2012;Assaiante, Barlaam, Cignetti, & Vaugoyeau, 2014).
Sensorimotor performance of the head-neck complex can be assessed through various motor tasks requiring complex sensory input and motor output functions (de Zoete et al., 2016). These include the cephalic repositioning test, which estimates the proprioceptive and vestibular systems without interference of the visuomotor system. By using a laser beam placed on the head, the ability of blindfolded participants to actively relocate the head in a neutral position is evaluated after an active rotation movement of the head-neck complex (Revel, Andre-Deshays, & Minguet, 1991). In other settings, the cervical proprioceptive reflex is assessed without focusing on the vestibular system by not including a cervical motor output component of the head-neck complex rotational kinematics. During this test, rotation of the trunk without movement of the head is measured with a 3D Fastrak electronic goniometer, by analyzing the effect of the modified cephalic repositioning test with a "neck torsion test" (Chen & Treleaven, 2013).
For including both sensory and motor component, other tests such as the "virtual reality test" (Sarig Bahat et al., 2016) and "The Fly" (Kristjansson, Hardardottir, Asmundardottir, & Gudmundsson, 2004) use a tracking system placed above the head to respectively measure the accuracy of the ability to follow a virtual target (the fly). Finally, the "DidRen Laser" is a test based on a system developed by our team in the late 2000s. It includes also both sensory and motor component by using a laser beam placed on the head of the participant with the aim to induce fast, low amplitude, and accurate rotational movements of the head-neck complex in response to real author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint visual targets placed in front of the participant (Hage & Ancenay, 2009). Sensorimotor performance is assessed using the time difference between two successful hits of the targets. This means that, in view of the participant's performance of speed-accuracy trade-off, the shorter the time, the faster and the more accurately the task is accomplished (Heitz, 2014;Liu & Watanabe, 2012;Zhang & Rowe, 2014). In 2009, we showed that the DidRen Laser was a simple and reliable device with a good reproducibility in asymptomatic and symptomatic adult individuals (Hage & Ancenay, 2009).
Compared to the other tests, the DidRen Laser is very easy to implement in a clinical setting, but it assesses only temporal variables. This does not allow us to gain insight into sensorimotor strategies adopted by individuals.
Our previous results now need to be extended with a detailed kinematic analysis carried out by a motion capture system. In a recent study by Sarig Bahat et al. (2016), modified kinematics of the head-neck complex were reported in asymptomatic people over 60 compared to young and middle-aged adults (Bahat, Igbariya, Quek, & Treleaven, 2016). This study did not include measurements in children, which we consider essential to provide a better understanding of headneck complex kinematics and normative data across all stages of the lifespan and to ensure that appropriate age-matched comparisons to participants with neck pain can be made (Hoy, Protani, De, & Buchbinder, 2010).
The aim of this study was to analyze the effect of age from 8 to 85 years old on sensorimotor control strategies adopted by asymptomatic individuals using kinematic variables derived from rotational motion of head-neck complex during the execution of the DidRen Laser test. author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the University hospital (Brussels, Belgium) and among the researchers' acquaintances. All participants volunteering for the study were informed about the nature of the study. Written consent was obtained from the participants, or from the legal representatives of participating children, before the start of the study. The study was approved by the local ethics committee and conducted in accordance with the declaration of Helsinki.
Participants suffering from any neuromusculoskeletal, neurologic disorder, impaired comprehension, non-correctable visual impairment, deafness, dizziness or any vestibular disorders diagnosed that could influence the performance of head-neck complex rotation, were excluded.
Further inclusion criteria were absence of neck pain episodes in the 6 months prior to the study and to ensure that we analyzed asymptomatic participants, a Neck Disability Index (NDI) score of less than or equal to 4% (Kato et al., 2012;Vernon & Mior, 1991). Similarly, all participants were asked to complete a Visual Analogue Scale (VAS) to confirm the absence of pain on the testing day (Ogon, Krismer, Söllner, Kantner-Rumplmair, & Lampe, 1996).

Instrumentation
The DidRen Laser was used (see (Hage & Ancenay, 2009) for a detailed description of the system and its components). In brief, it is made up of three photosensitive sensors with photocells, a fine laser beam, and a computer equipped with customized software that calculates the time between two consecutive "hits" of the sensors and displays the results. A chair without armrests is author/funder. All rights reserved. No reuse allowed without permission.

Experimental procedure
All participants received the same instructions about how the experimental procedure was to be conducted and more specifically that of the DidRen Laser test, by watching an explanatory video. The participants wore a helmet to which a laser beam is mounted and sat on a chair. They were instructed to keep their back against the backrest, the palm of hands on the thighs, feet flat on the floor with heels against a stop block placed at the feet of the chair and to refrain from talking during the test.
The procedure of one generated cycle was explained as followed: when the laser beam is pointed correctly at the target (during at least 0.5 s to lower chance), the LED's sensor lights up and author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint 8 the system emits a sound signal. As soon as the central sensor has been ''hit'', the participant must, as rapidly as possible, turn his/her head to maximum 30° to the right to hit the right-hand sensor.
He/she returns to the central sensor and then rotates the head to maximum 30° to the left to hit the left-hand sensor. The participant completes the cycle by hitting the central sensor for a third time.
Then, one cycle is defined systematically with 4 head rotations facing sensors in the same order: 1) from center to the right, 2) from right back to the center, 3) from center to the left and 4) back to the center. A complete trial is composed of 5 cycles. We used non-random generated cycles in order to facilitate intra and inter-subject comparisons.
As in 2009 (Hage & Ancenay, 2009), two DidRen Laser tests were conducted. The first trial was considered as a short warm-up to "familiarize" the participant with the experiment by emphasizing on the adequate sitting position and speed-accuracy execution. So only the second trial was used for data collection and analysis. We didn't conduct more than two trials to avoid possible fatigue which could lead to a lost precision (Longo & Meulenbroek, 2018).

Kinematic variables
The DidRen Laser software calculates each time taken by the participant to go from one target's ''hit'' (that is when participant stops during at least 0.5 s on the sensor) to another target.
Only the total time (TT, in s) to complete the 5 cycles (from the first to the last target) of a trial was included in the data collection.
Head rotational and shoulders-displacement were computed using ELICLINIC software (BTS, Italy) from X, Y, and Z coordinates at each frame. By successive numeric finite difference (n=±5 points), we calculated speed and acceleration of head. The following specific parameters were computed for each kinematics variable: head-rotation range of motion (in °); shoulders displacement (in °), overshoot (OS, in °) was computed as the difference between peak rotation author/funder. All rights reserved. No reuse allowed without permission.

Statistical analysis
To assess the effect of age on the average variables, a one-way ANOVA with post hoc Holm-Sidak method for pairwise multiple comparisons was carried out when normally distributed and with post hoc Dunn's method for pairwise multiple comparisons if normality test failed. All statistical procedures were performed with SPSS version 25 and SigmaPlot 13 with a significant level at 5%.
author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint

Results
Total sample size consisted of 87 participants from which 7 participants were excluded due to an NDI score >4%. The anthropometric characteristics of the participants are listed in Table 1.
All results are showed in Figure 2. A significant effect (p<0. author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint

Discussion
The results of the study reveal significant effects in asymptomatic Children and Seniors groups for four rotational kinematic variables: total time, mean speed, acceleration and deceleration.
These kinematic differences were observed during the DidRen Laser test, showing its capacity to discriminate age-related differences.
Velocity and the derivate forms were chosen to analyze the performance of age-group participants in accordance with previous studies (Roijezon et al., 2010;Sarig Bahat et al., 2016).
Sarig Bahat et al. (2016) showed that the most powerful age-groups differences were Velocity Peaks and Number of Velocity Peaks . With a different task and method of calculation, our kinematic variables such as Total Time, Mean Speed, Acceleration and Deceleration appeared to be very significant especially with the groups of Children and Seniors.
To reach the speed-accuracy required by our protocol: "turn your head faster as you can", participants needed to trade-off with their sensorimotor ability to perform the rotational movement.
In 1954, Fitt described the speed-accuracy trade-off that exists when people, including seniors and children (Passmore et al., 2007;Smits-Engelsman, Sugden, & Duysens, 2006), are involved in accuracy constrained target-aiming task. The theory was that movement time would increase linearly as the size of the targets decreases or the distance between the targets increases (Paul M Passmore et al., 2007).
In accordance with Fitts' law, we showed that Seniors and Children needed to be slower (Total Time, Mean Speed, Acceleration, Deceleration) (Descarreaux et al., 2010;Paul M Fitts, 1954;P. M. Fitts, 1992;Smits-Engelsman et al., 2006) to become as precise as Young and Old people (no significant differences for Overshoot in all age-groups). Our results have demonstrated the ability for Children and Seniors to adapt their speed to be at least as accurate as Young and Old Adults. Similarly, we have also demonstrated that both Adults groups have adapted their speed.
author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint Indeed, both Adults groups showed an average of speed 10 times less than the average of speed (Vmean ) obtained by Sarig Bahat's participants of the same ages . Actualy, even if Sarig Bahat's participants were asked to move their head fast, they could take longer time (seven seconds) to point the target after having produced the motion. So, they didn't need to tradeoff with their speed and accuracy as much as needed to accomplish the DidRen Laser test.
Even if the kinematic variables assessed in this study do not allow us to differentiate the subsystem(s) of the sensorimotor control, it is plausible that senior-related modifications can be due to dysfunction in neck proprioception, vestibular and visual systems (Passmore et al., 2007;Uthaikhup et al., 2012). Slower Children-related performances could be due to slow maturation of the sensorimotor system (Assaiante, 2012;Assaiante, Mallau, Viel, Jover, & Schmitz, 2005) and immaturity in the development of the central nervous system (Assaiante, 2012;Assaiante et al., 2005;Corporaal et al., 2017;Tau & Peterson, 2010).
The results of the present study should be seen in the light of several methodological clarifications. Firstly, standardization of instructions and better understanding of participants, especially for children and some older participants, were increased by using a tablet computer with an explanatory video of the DidRen Laser test. Secondly, to avoid measurement errors that could increase variability of our results, we standardized the posture of the head, trunk, limbs and the height of visual reference and the influence of the scapular girdle displacement while performing head-neck complex rotations (Gueguen, Vuillerme, & Isableu, 2012;Scotto Di Cesare et al., 2013;Scotto Di Cesare, Macaluso, Mestre, & Bringoux, 2015). Thirdly, to avoid influencing stability of the neck and for being more functional, we didn't secure the shoulder with a seat-belt (Cromwell, Aadland-Monahan, Nelson, Stern-Sylvestre, & Seder, 2001). Fortunately, our results concerning the scapular girdle rotation displacement showed that it was marginal for all participants and confirmed author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint that they respected the instruction not to move their shoulders during the test. Thanks to this option, the DidRen Laser test could be used more easily for future clinical study.
This study presents some limitations. First, neck ROM was limited to 30°, while this allowed us to avoid the strain of the neck passive system and to improve input from the neck, it corresponds to about 30 % of the total cervical amplitude and may not be perceived as clinically relevant.
Second, only head-neck rotational movement was assessed. As rotation seems to be a regular movement during daily activities, the assessment of other motion directions (e.g. flexion/extension) appeared to have limited interest. Third, we used for the test regular generating head rotation cycles.
Indeed, use randomly generated cycles might reduce an induced anticipatory motions and predictions of participants but that has not yet been investigated for this kind of test.
In conclusion, a difference was observed in sensorimotor control behavior in both children and senior individuals compared to adults. The DidRen test proves to be discriminant to sensorimotor control strategies linked to age. When using this test, age must be considered as a key parameter when analyzing results of patients versus non-painful individuals.

Declaration of interest
None.

Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint Fig. 2. Box-plots (interquartile range) for all age effect measurements. First four box-plots are for age-groups significant kinematic variables with post hoc comparisons: ★ P<0.05. The three horizontal box-plots are for nonsignificant kinematic variables. author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/519488 doi: bioRxiv preprint