Effects of dual-task conditions on cervical spine movement variability

Abstract

BACKGROUND:

The potential to accurately perform cervical movements during more challenging tasks might be of importance to prevent dysfunctional motion characteristics. Although sensorimotor function during dual-task conditions are of increasing interest in biomedical and rehabilitation research, effects of such conditions on movement consistency of the neck have not yet been investigated.

OBJECTIVE:

In this crossover MiSpEx(Medicine in Spine Exercise)-diagnostic study, we aimed to explore differences between single and dual-task conditions on cervical movement variability.

METHODS:

Nineteen healthy participants (9 male; 24.5 $\pm$ 3.3 y) performed 10 repetitive maximal cervical movements in (1) flexion/extension and (2) lateral flexion, during one single- and during two dual-task test conditions (cognitive, motor) in a randomised and cross-over sequence. Latter consisted of a working memory n-back task ( $n=$ 2) and a repetitive ankle movement task. Range of motion (RoM) was assessed using an external three-dimensional ultrasonic movement analysis system. Coefficient of variation (CV) for repetitive RoM was analysed for differences between conditions and controlled for variances in intra-individual movement characteristics.

RESULTS:

Friedman and post-hoc Bonferroni-adjusted confidence intervals for differences from single- to dual-task values revealed changes in CV in flexion/extension from single-task to motor dual-task ( $+$ 0.02 $\pm$ 0.02 (97.5%CI: 0.01; 0.03); $p<$ 0.05) but not to cognitive dual-task condition ( $+$ 0.01 $\pm$ 0.02 (97.5%CI: 0.003; 0.02)) nor for lateral flexion ( $p>$ 0.05). Pearson regression analyses revealed a linear negative ( $p<$ 0.01) influence of CV in flexion/extension on differences from single to both cognitive (R ${}^{2}=$ 0.47) and motor dual-task (R ${}^{2}=$ 0.55). Results for lateral flexion are comparable, baseline CV negatively impacts differences to cognitive (R ${}^{2}=$ 0.2) and motor dual-task performance (R ${}^{2}=$ 0.76; $p<$ 0.01).

CONCLUSIONS:

Participants with comparable low cervical CV at single-task display a profound increase during dual-task conditions while participants with a higher variability remained almost stable or showed a decrease. The results point toward a complex interrelationship of motion patterns and adaptation processes during challenging tasks in respect of cervical CV.

Keywords

Kinematics MiSpEx motion analysis coefficient of variation dual-task

1. Introduction

Movement variability is commonly defined as the normal variation occurring during natural motor performance [9]. Its physiological meaning in human kinematics is under ambiguous discussion. On one hand, to maintain a high target precision during repeated movements is suggested to display economically performed motion patterns [10]. On the other hand, the amount of variability during repetitive movements reflects inherent functional features of the neuromuscular system [2], recently delineated to be of relevance for new motion patterns’ learning processes [11, 12]. Normal motor skills are accompanied by an “optimal” level of variability [13], higher and even lower levels may consequently be seen as symptomatic. Higher than “optimal” variability is seen as noisy or unstable whereas lower variability values are characterized as biologically rigid or unchanging systems [13]. Such non-standard variability consequently is rated as non-physiologic.

Although of particular interest in gait performance research and shown to impact gait variability [14], the impact of divided attention tasks (e.g. dual-tasks) on movement variability of simple cervical spine movement performance have not yet been investigated. The variability of repeated movements differs between healthy participants and neck pain patients [3]. As unphysiologic motion characteristics are seen as potential risk factors for the development of nonspecific neck pain and increased variability may be non-physiologic, the potential to accurately perform cervical movements during more challenging (e.g. dual-) tasks might be of importance to prevent dysfunctional motion characteristics and thus musculoskeletal disorders. Although speculative, maintaining “optimal” CV values may consequently be of importance as well in neck kinematics.

The aim of our study was to further explore the effects of simple motor tasks’ adaptations (dual-task, cognitive and motor) on movement variability in unimpaired participants during repetitive cervical movements. We hypothesize that these dual-task conditions increase the motor variability in participants without movement constrictions during simple cervical motion patterns.

2. Materials and methods

2.1 Ethical aspects

Our randomised cross-over study was approved by the local ethical review committee for research and conducted in accordance to the ethical standards set by the declaration of Helsinki (Helsinki, 1964; Fortaleza, 2013).

2.2 Participants

A sample of healthy and adult volunteers was recruited by personal request of one of the authors. Sample number was based on a sample size calculation using the algorithm given in G*POWER 3.0.2 for the primary outcome ( $\alpha=$ 5%, $\beta=$ 0.2, f $\geqslant$ 0.3 as a middle effect size). Following application of inclusion and exclusion criteria, each participant signed informed consent prior to study enrolment. Participants were considered eligible if they fulfilled the following inclusion criteria: adults with full-time employment without neck pain history. Exclusion criteria encompassed previous surgery of the spine, any persistent or passed cervical spine complaints based on space-occupying, inflammatory, traumatic or systemic processes and intake of painkilling drugs, analgesics or muscle relaxants within the previous 48 hours. Participants with impaired vision, dizziness, inflammatory rheumatic diseases, osteoporosis, fracture and/or surgery within the last 6 months, severe cardiovascular, pulmonary, neurological, cancerous, endocrinologic and psychiatric diseases, fibromyalgia, haemophilia and pregnancy were excluded, alike. Inclusion and exclusion criteria were assessed anamnestically by interviewing the participants.

2.3 Outcome assessment and analyses

Demographic, anthropometric and physical activity (amount of exercise and leisure physical activity per week in hours, one question) data were initially assessed. Twenty-two persons were contacted to include 19 participants (10 female, 9 male; mean 24.5 range 19–32) years) between 48 kg and 92 kg body weight (median $=$ 67 kg).

All participants performed afterwards 10 repetitive maximal cervical flexion/extension movements, followed by 10 lateral flexion (bending) manoeuvres, each at three different conditions and in an upright sitting position. The tasks were performed in a randomised and counterbalanced cross-over sequence. Participants were instructed to perform over the maximal range of motion in a self-determined and comfortable velocity with eyes open. Neither additional instruction nor feedback was given. For device familiarisation, all participants initially performed three familiarisation trials prior to the first trial.

A non-invasive external three-dimensional ultrasonic movement analysis system (Zebris CMS 70, Zebris Medical GmbH, Isny, Germany) was used to collect cervical kinematic data. Sufficient test–retest reliability of this measurement system (ICC $=$ 0.87; SEM $=$ 5.47 ${}^{\circ}$ ; $p<$ 0.001) has been demonstrated [15]. Three ultrasound microphones, determining a local coordinate system, were used to track low-mass ultrasonic markers (diameter 10 mm, weight 3 g each). Three markers, mounted on a small, lightweight T-plate, were attached to the lateral aspect of the shoulder girdle, and a second similar triangular base plate was mounted on a rig and positioned to the head. Head movements were defined as movements of a single, rigid body system with respect to the shoulder embedded coordinate system.

Data were sampled at 30 Hz, real time monitored and stored on a data processor for the following offline analysis. Coefficient of variation (CV $=$ SD/mean) to determine the intra-subject variability in per cent of maximal range of motion was considered as primary outcome and computed as follows: oscillation amplitude was calculated for each of the 10 repetitive movements; they were afterwards normalized to the maximal amplitude what from the CV was computed. Movements were performed during the three conditions: (1) standard single task – (2) cognitive dual task (standard movements and cognitive task) – (3) motor dual task (standard movements and motor task), each performed by each subject in a randomly sequenced order. The cognitive task thereby consisted of a working memory n-back task ( $n=$ 2) [16], a task often used in dual task conditions [17]. Letters from A-Z were randomly used as stimuli. Participants were told to left-button click on a mouse if current stimulus matches the one from two steps earlier during the auditory two-back cognitive task. The mouse was placed on the anterior thigh. To avoid a potential influence on the movements in comparison to single- or motor dual-task, the mouse was hold during all tasks in this position. The motor task included both-sided repetitive dorsal flexion of the upper ankle joints in a self-determined velocity. Compliance during both tasks was optically controlled by an investigator. All tests were performed during the same session using a 20 minute washout-period in-between tasks.

Sagittal plane CV was afterwards categorized using validated cut-off values [8] to distinguish potentially symptomatic from asymptomatic CV values. This dichotomisation is further used for as a grouping variable (group one below cut-off and group two above cut-off).

2.4 Statistics

All statistical calculations were conducted using SPSS 20 (SPSS Inc., Chicago, IL, USA) or BiAS for Windows (version 10, 2012, Goethe-University Frankfurt). Statistical procedures were executed after the examination of underlying assumptions for group comparisons and regression analyses. Data were analysed for (1) differences between conditions and for (2) variance explanation of baseline values (independent variable) to dual-task-induced differences (dependent variable) Following the dichotomisation based on the cut-offs for CV, data were (3) statistically analysed for differences between groups and conditions. The a priori level of significance was set at 5% for all statistical analyses.

3. Results

No participant had to be excluded after study enrolment. Participants were physically active for (range) 2 to 21 hours per week (median $=$ 9 h/week).

Friedman and post-hoc Bonferroni-Holm-adjusted Wilcoxon tests [18] revealed significant differences in movement variability in flexion/extension between the motor dual- and the standard task but not between the cognitive dual-task and the standard task nor in lateral flexion (Table 1).

Table 1
Statistical characteristics for the repeated measure trial. In case of significant Friedman tests, Bonferroni-adjusted post-hoc-Wilcoxon tests were performed

	Flex/Ext			Latflex
	Cognitive	Standard	Motor	Cognitive	Standard	Motor
	(dual)	(single)	(dual)	(dual)	(single)	(dual)
CV
Median	0.03	0.02	0.05	0.04	0.04	0.05
Range	0.01–0.07	0.01–0.06	0.01–0.09	0.01–0.07	0.01–0.07	0.03–0.07
Friedman characteristics	Chi ${}^{2}=$ 9.6; $p<$ 0.01			Chi ${}^{2}=$ 4.1; $p>$ 0.05
Post-hoc Wilcoxon	$p>$ 0.05		$p<$ 0.05
Effect size dz			0.34

CV $=$ coefficient of variation, Flex $=$ flexion in sagittal plane, Ext $=$ extension in sagittal plane, Latflex $=$ lateral flexion in frontal plane, $p=$ (adjusted) p-value.

The dichotomization based on the cut-offs (sagittal plane only) revealed a numerical equality between groups (below cut-off: 10 participants; above cut-off: 9 participants). Thus, 10 out of the 19 participants display a CV described as asymptomatic. All CV values for these allocated groups are displayed in Fig. 1.

Figure 1.

Boxplots for the allocated (below cut-off vs. above cut-off) group data for sagittal plane (Flexion / Extension) kinematic data. Each plot displays median, percentile 50 and range of data (whisker bars) incl. outliers. CV $=$ coefficient of variation.

Friedman testing, in case of significance followed by Bonferroni-Holm-adjusted Wilcoxon testing, revealed that participants below the cut-off-value showed an increase of the CV from the standard baseline assessment to both dual task conditions ( $p<$ 0.01); participants above the cut-off remained stable ( $p>$ 0.05). Groups differed at standardized baseline ( $p<$ 0.001), but not at dual task conditions (each $p>$ 0.05).

Pearson regression analyses revealed a significant negative influence of baseline coefficient of variation in both movement planes and for all outcomes ( $p<$ 0.01, Fig. 2).

Figure 2.

Individual data and regression curves for motion analysis outcomes ( $n=$ 19 healthy participants) including individual predictive 95% confidence intervals ; independent variable: baseline CV-value, dependent variable: dual-task-induced differences in CV [%]. CV $=$ coefficient of variation, Flex $=$ flexion in sagittal plane, Ext $=$ extension in sagittal plane, LatFlex $=$ lateral flexion in frontal plane.

4. Discussion

The present study explored effects of simple motor tasks’ adaptations on cervical movement variability. Unimpaired participants with a comparable low variability during single-task assessment display a profound increase of this variability during dual-task conditions while participants with a higher initial variability remained almost stable or showed a decrease. Thus, the results prove our hypothesised association.

These findings are comparable to those from other trials who assessed kinematics during dual-task-condi-tions: postural sway during static posturography increases under such conditions [4] and gait characteristics might be impaired by dual-tasks, likewise [5]. The baseline CV-values are not different to those found in other studies under single task conditions [6]. In this previous study, we found a sagittal plane CV of 0.03 (95%-CI $=$ 0.023, 0.038) in asymptomatic participants.

Our results point towards a complex interrelationship of simple motion patterns and adaptation processes during more challenging tasks in respect of cervical motor variability. Decreases in the accuracy and thus increases in the variability may picture sensorimotor malfunction, e.g. in chronic musculoskeletal conditions like low back [1] or neck pain [3, 19]. A comparable increase in movement variability was described in unimpaired participants simulating impairments during motion patterns, likewise [20]. Nevertheless, the relevance of CV on new motion patterns’ learning processes was recently shown [11, 12]. In this sense, one may speculate that the value of the movements’ variability may be used to predict participants who will learn a specific motor task faster. Likewise, one may construe CV as – during simple and well-known single-tasks – unwanted motor noise and – during new and challenging dual-tasks – as both unwanted motor noise and supporting learning processes. During simple and well-known single-tasks, low CV may be physiologic and, during more challenging new tasks, higher CV may consequently be physiologic or unphysiologic. As unphysiologic motion characteristics are seen as potential risk factors for the development of nonspecific neck pain, the potential to accurately perform cervical movements during more challenging (e.g. dual-) tasks might be of importance to prevent dysfunctional motion characteristics and thus musculoskeletal disorders. However, the definitive allocation of the CV and its dual-task triggered changes in either noise in sensory and motor processing [1] or sensorimotor integration functional features of the neuromuscular system [2] nevertheless still be of particular interest. Until this point of time, maybe assessments of CV to distinguish asymptomatic from symptomatic or even physiologic from nonphysiologic movement behaviour ought to be assessed in well-known situations and during single task conditions.

Potential mechanisms may be found in the intersection of these two meanings of movement variability. The motor system provides a neural circuit mechanism for the active control of motor output variability to enhance learning processes [12]. It is able to both, measure the value of motor errors and the knowledge of how adjustments in motor output reduce them [21]. Large repositioning errors thus may be followed by significant improvements in sensorimotor accuracy. Our findings therefore might have impact on the development of motor learning processes and the detection of cervical spines’ kinematic malfunctions. In healthy participants, having an increased CV may indicate a slight sensorimotor restriction. Cervical movement variability changes during challenging tasks may in addition be of relevance in monitoring learning processes.

The dual-tasks’ unknown validity in implementing distraction may slightly limit our results validity The fact that nearly the half of our (healthy) sample shows rather high measures of variability under dual tasks conditions might also be reflective of either a natural range of variability or a questionable validity of cut-off-criteria. Future study is warranted to 1) recheck the herein found association in a sample of neck pain patients and to 2) unravel the ambiguity of the movement variability by separating motor malfunction and learning processes. As cervical movement behavior is affected by the amount of physical activity [7], the same analyses for low compared to high active persons may be of importance for future study. Such study could have an impact on sports medical training programs and foster multi-tasking rehabilitation settings as well in neck pain patients To reveal if variability remains locally restricted during more challenging tasks or not, e.g. to reveal if cervical dual task remotely affects postural control, further remains an open question.

5. Conclusions

Participants with lower cervical movement variability at single-task display a profound increase during dual-task conditions while participants with a higher variability remained almost stable or showed a decrease. These results point toward a complex interrelationship of motion patterns and adaptation processes during challenging tasks in respect of cervical CV. Maintaining an optimal CV during more challenging tasks might be of importance to prevent dysfunctional motion characteristics.

Conflict of interest

None declared. No financial or personal relationships with other people or organizations have inappropriately influenced our work.

Footnotes

Acknowledgments

The present study was initiated and funded by the German Federal Institute of Sport Science and realized within MiSpEx – the National Research Network for Medicine in Spine Exercise.

References

Laird

Gilbert

Kent

Keating

. Comparing lumbo-pelvic kinematics in people with and without back pain: a systematic review and meta-analysis. BMC Musculoskelet Disord 2014; 15: 229.

Preatoni

Hamill

Harrison

Hayes

van Emmerik

REA

Wilson

et al. Movement variability and skills monitoring in sports. Sports Biomech 2013; 12(2): 69–92.

Vogt

Segieth

Banzer

Himmelreich

. Movement behaviour in patients with chronic neck pain. Physiother Res Int (Physiotherapy Research International) 2007; 12(4): 206–12.

van Daele

Hagman

Truijen

Vorlat

van Gheluwe

Vaes

. Decrease in postural sway and trunk stiffness during cognitive dual-task in nonspecific chronic low back pain patients, performance compared to healthy control subjects. Spine 2010; 35(5): 583–9.

Brach

McGurl

Wert

Vanswearingen

Perera

Cham

et al. Validation of a measure of smoothness of walking. J. Gerontol. A Biol. Sci. Med. Sci. 2011; 66(1): 136–41.

Niederer

Vogt

Pippig

Wall

Banzer

. Local Muscle Fatigue and 3D Kinematics of the Cervical Spine in Healthy Subjects. Journal of Motor Behavior 2016; 48(2): 155–63.

Chen

Solinger

Poncet

Lantz

. Meta-analysis of normative cervical motion. Spine 1999; 24(15): 1571–8.

Niederer

Vogt

Rickert

Wilke

Banzer

. Age-related Cut-Offs for Cervical Movement Behaviour to Distinguish Chronic Idiopathic Neck Pain Patients from Unimpaired Subjects. European Spine Journal (Eur Spine J) 2015; 24(3): 493–502.

Stergiou

Harbourne

Cavanaugh

. Optimal movement variability: a new theoretical perspective for neurologic physical therapy. J Neurol Phys Ther 2006; 30(3): 120–9.

10.

Vogt

Banzer

. Dynamic testing of the motor stereotype in prone hip extension from neutral position. Clin Biomech (Bristol, Avon) 1997; 12(2): 122–7.

11.

Herzfeld

Shadmehr

. Motor variability is not noise, but grist for the learning mill. Nat Neurosci 2014; 17(2): 149–50.

12.

Miyamoto

Castro

LNG

Ölveczky

Smith

. Temporal structure of motor variability is dynamically regulated and predicts motor learning ability. Nat Neurosci 2014; 17(2): 312–21.

13.

Stergiou

Decker

. Human movement variability, nonlinear dynamics, and pathology: is there a connection? Hum Mov Sci 2011; 30(5): 869–88.

14.

Lamoth

van Deudekom

van Campen

Appels

Vries

de Pijnappels

. Gait stability and variability measures show effects of impaired cognition and dual tasking in frail people. J NeuroEngineering Rehabil 2011; 8(1): 2.

15.

Cagnie

Cools

Deloose

Cambier

Danneels

. Reliability and Normative Database of the Zebris Cervical Range-of-Motion System in Healthy Controls with Preliminary Validation in a Group of Patients with Neck Pain. J Manipulative Physiol Ther 2007; 30(6): 450–5.

16.

Kirchner

. Age differences in short-term retention of rapidly changing information. J Exp Psychol 1958; 55(4): 352–8.

17.

Verhaeghen

Zhang

. What is still working in working memory in old age: dual tasking and resistance to interference do not explain age-related item loss after a focus switch. J Gerontol B Psychol Sci Soc Sci 2013; 68(5): 762–70.

18.

Holm

. A Simple Sequentially Rejective Multiple Test Procedure. Scand J Stat 1979; 6(2): 65–70.

19.

Sjölander

Michaelson

Jaric

Djupsjöbacka

. Sensorimotor disturbances in chronic neck pain–range of motion, peak velocity, smoothness of movement, and repositioning acuity. Man Ther 2008; 13(2): 122–31.

20.

Dvir

Prushansky

Peretz

. Maximal versus feigned active cervical motion in healthy patients: the coefficient of variation as an indicator for sincerity of effort. Spine 2001; 26(15): 1680–8.

21.

Smith

Ghazizadeh

Shadmehr

. Interacting adaptive processes with different timescales underlie short-term motor learning. PLoS Biol. 2006; 4(6): e179.

Effects of dual-task conditions on cervical spine movement variability

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Materials and methods

2.1 Ethical aspects

2.2 Participants

2.3 Outcome assessment and analyses

2.4 Statistics

3. Results

Table 1 Statistical characteristics for the repeated measure trial. In case of significant Friedman tests, Bonferroni-adjusted post-hoc-Wilcoxon tests were performed

5. Conclusions

Conflict of interest

Footnotes

Acknowledgments

References

Table 1
Statistical characteristics for the repeated measure trial. In case of significant Friedman tests, Bonferroni-adjusted post-hoc-Wilcoxon tests were performed