- Email: [email protected]
Perception of subjective visual vertical and horizontal in patients with chronic neck pain: A cross-sectional observational study
Manual Therapy 17 (2012) 133e138
Contents lists available at SciVerse ScienceDirect
Manual Therapy journal homepage: www.elsevier.com/math
Original article
Perception of subjective visual vertical and horizontal in patients with chronic neck pain: A cross-sectional observational study Sharon Docherty a, *, Rebekka Schärer b, Jeff Bagust a, B. Kim Humphreys b a b
Anglo-European College of Chiropractic, 13-15 Parkwood Road, Bournemouth BH5 2DF, UK University Orthopaedic Hospital of Balgrist, Forchstrasse 340, 8008 Zurich, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 April 2011 Received in revised form 31 October 2011 Accepted 8 November 2011
Previous studies have shown that chronic neck pain (CNP) patients have a larger spread of perceptual errors for subjective visual vertical (SVV) than those exhibited by asymptomatic controls. The current study investigated whether this was also the case for perception of subjective visual horizontal (SVH) and whether there was a correlation between the two measurements. Fifty patients with CNP were compared with a group of 50 age- and gender-matched controls. All subjects were required to complete a test to measure SVH as well as SVV using the computerised rod and frame (CRAF) test. These tests were conducted under various frame conditions. No difference was found between the errors of the CNP and control groups in the absence of a surrounding frame. When a tilted frame was added to the CRAF test, the range of errors observed in the CNP group increased for both SVV and SVH. In particular, significantly more CNP patients fell outside the reference range of errors and a subgroup of patients, characterised by higher neck pain disability indices, was identified who demonstrated higher than expected errors for both SVV and SVH. However no conclusion could be drawn with regards to the direction of error asymmetry and laterality of pain as those patients with unilateral pain exhibited errors both towards and away from the affected area. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Subjective visual vertical Subjective visual horizontal Chronic neck pain
1. Introduction Very little is known about the causes of chronic neck pain (CNP) (Benyamin et al., 2009) this, in part, could be due to the wide variety of signs and symptoms that are present in this group of patients (Passatore and Roatta, 2006). Altered movement and muscle activation patterns in the neck and shoulder region are among these signs (Falla et al., 2004) as well as reduced postural control (Karlberg et al., 1995). Subjective visual vertical (SVV) and horizontal (SVH), are measures of an individual’s internal representation of their position in space relative to gravity, and therefore important in the maintenance of posture and equilibrium (Mazibrada et al., 2008). Disturbances to this perception can occur as a result of errors in sensory input from one of the three major systems involved; the vestibular; visual; and proprioceptive systems, or at the level of integration of these signals within the brain (Anastasopoulos et al., 1999). As a result, measurements of SVV, and to a lesser extent SVH, are increasingly being employed in the assessment of conditions * Corresponding author. Tel.: þ44 1202 436 242; fax: þ44 1202 436 312. E-mail address: [email protected] (S. Docherty). 1356-689X/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2011.11.002
such as Menière’s Disease, vestibular dysfunction and dizziness (Guerraz et al., 2001; Hafström et al., 2004a,b; Mazibrada et al., 2008; Pagarkar et al., 2008; Kumagami et al., 2009) as well as brain injury due to, for example, stroke (Saj et al., 2005a,b). Based on work using healthy subjects, Betts and Curthoys (1998) proposed that it is incorrect to conclude that perception of SVV and SVH are “invariably orthogonal to one another”. However, Hafström et al. (2004a,b) have demonstrated a high correlation between the errors produced for SVV and SVH in patients with unilateral vestibular deafferentation and suggested that the two measures are of equal clinical relevance for this group. Previous studies conducted on CNP patients have shown that they have significantly larger errors in SVV, than asymptomatic controls (Grod and Diakow, 2002; Bagust et al., 2005). In particular, Bagust et al. (2005) found a wider range of errors in the neck pain group compared to the controls, suggestive that there may be a subgroup of neck pain patients who have associated disturbances of their perception of verticality. The aim of the current study was to investigate whether CNP patients exhibit increased errors in subjective visual horizontal (SVH), not previously investigated in this group of patients, as well as SVV, compared to asymptomatic controls. In addition, we
134
S. Docherty et al. / Manual Therapy 17 (2012) 133e138
examined whether there was a relationship between laterality of pain and asymmetry of perception of SVV and SVH. 2. Methods 2.1. Participants Fifty participants with CNP were recruited from four Chiropractic Clinics in the Canton of Zürich and from the Physiotherapy Department of the University Orthopaedic Hospital of Balgrist, Zürich. To be included, participants must have been between 30 and 65 years of age and experienced neck pain with or without arm pain for at least four weeks duration. A control group of 50 age- and gender-matched subjects was recruited at the University Orthopaedic Hospital of Balgrist and by advertising in a local paper. Control subjects had not experienced significant neck or upper thoracic pain in the previous month i.e. pain that had limited their daily activities or for which they had sought care. Subjects were excluded from the study if they had a recent history of cervical spine surgery, bone pathology such as neoplasm, infection or serious congenital anomalies, vestibulo-cochlear disease affecting balance or medication use which may affect balance, vision and co-ordination. All subjects gave their written consent to participate after being informed about the aims and methods of the study. The study was approved by the Zürich Cantonal Ethics Committee. For the CNP group, pain intensity was measured using the 11point (0e10) Numerical Rating Scale (NRS). The NRS is the most frequently used scale for measuring pain intensity and has been shown to be psychometrically robust and practicable (Bolton and Wilkinson, 1998). In addition, patients’ self-reported disability was measured using the Neck Disability Index (NDI). The NDI is the most commonly used questionnaire for neck pain research (Vernon, 2008). It consists of 10 categories in which a score of 0e5 is possible for each category. Total NDI scores of 30 are considered the cut-off for moderate levels of pain and disability (Johnston et al., 2008). The duration, type of onset and laterality of symptoms were also recorded for each patient. 2.2. Procedures The CNP group completed the NRS for pain intensity and the NDI for neck pain disability. Control participants were asked if they were currently experiencing any neck or upper back pain to ensure that they were asymptomatic at the time of testing. All participants were required to complete both of the tests for the perception of vertical and horizontal. The order in which the two tests were performed was randomly assigned for each subject. These were performed with the subject sitting in a comfortable position, their head was not restrained. Between tests subjects were given a short break. The tests consisted of a modified version of the computerised rod and frame (CRAF) test (see Docherty and Bagust, 2010) viewed by the participant through a pair of Olympus Eye-trek FMD 200 video glasses. Wearing the video eyeglasses resulted in an image that spanned a viewing angle of 30 23 degrees of the visual field (the equivalent of viewing a 1.42 m screen from a distance of 2 m). Both the horizontal and vertical tests comprised 18 presentations, the first two of which were for instruction purposes and were not included in the analysis. The first presentation in each test always contained the upright, untilted frame. This was used to confirm that the glasses were positioned correctly. The second presentation comprised of a tilted frame and was used to establish whether the
subject understood the task. The remaining 16 presentations consisted of four replicates where the frame (a white square presented on a homogenous black background) was either e (i) absent (giving no visual references); (ii) untilted (frame0, providing visual reference cues for vertical and horizontal); tilted 18 degrees (providing confusing visual reference cues) in either a (iii) clockwise (frameþ18,); or (iv) counter clockwise direction (frame18). The two white dots used to represent the ends of the ‘rod’ were displayed in the centre of the screen (within the frame if present) and had two starting positions e tilted 20 degrees in either a clockwise or counter clockwise direction from gravitational vertical/horizontal. The order of presentation of these permutations of frame and dots was assigned by the computer from a bank of 4 randomised sequences. The subject’s task was to rotate the dots using the right and left mouse buttons to a position perceived to be vertical or horizontal depending on the test. The dots rotated around their virtual midpoint in 0.5 degree increments. When the participant was satisfied with the alignment of the dots, the programme was advanced to the next presentation by pressing the space bar of the computer keyboard. This task took approximately 11 s per presentation. Positioning error was recorded by the computer as degrees from gravitational vertical/horizontal. Neither the subject nor the operator had access to the recorded errors until the end of the recording session. 2.3. Statistical analysis Recorded errors from the CRAF test were used to calculate the signed and unsigned (absolute) means for the four frame condition (n ¼ 4 in each case) for each participant. Unless otherwise stated, reported errors are the absolute (unsigned) values. All statistical analyses were performed using SPSS 17.0 and Instat (Graphpad, Inc). Data were tested for normality using the KolmogoroveSmirnov method. In general, the data obtained from patients did not follow a normal distribution and so non parametric statistics were used. To test for differences between the errors produced under different frame conditions, a Repeated Measures Analysis of Variance (ANOVA) was used with post hoc TukeyeKramer Multiple Comparison Tests. All control group data were normally distributed and so the reference ranges for error distributions were calculated using mean þ 2SD of these data. Unless otherwise stated this was based on the unsigned data. Differences between neck pain patients and controls were investigated using the ManneWhitney U test for 2 independent samples. The 95% confidence intervals (CIs) for the proportion of participants with errors greater than 2SD of the control means were calculated using the Wilson method and to illustrate differences between the proportion of patients and controls in this category, the Newcombe method was used to calculate the CIs (Newcombe and Altman, 2000). 3. Results 3.1. Participants Both the CNP and control group consisted of 50 individuals of whom 40 were female. There was no significant difference between the groups with regards to age (mean 48 years in each, Table 1). In the CNP group, the median pain rating was 4.5 on the NRS with a mean (SD) NDI of 33 16. 3.2. Comparison of CNP and control groups 3.2.1. No visual reference frame As can be seen from Table 1 and Fig. 1, in the absence of the cues given by a surrounding frame, both CNP and control mean errors
S. Docherty et al. / Manual Therapy 17 (2012) 133e138
135
Table 1 Comparison of CNP patients (n ¼ 50) and asymptomatic controls (n ¼ 50). CNP (n ¼ 50)
Controls (n ¼ 50)
Age (years)*
48.1 8.7
47.9 8.7
SVV Signed ( ) Unsigned ( )
0.03 0.66 (0.00) 0.90 0.51 (0.75)
0.37 0.75 (0.25) 0.87 0.47 (0.81)
SVH Signed ( ) Unsigned ( )
0.07 1.16 (0.25) 1.04 1.71 (0.75)
0.01 0.48 (0.00) 0.56 0.30 (0.62)
Frame effect Untilted vertical ( ) Tilted vertical ( ) Untilted horizontal ( ) Tilted horizontal ( )
0.87 0.47 3.73 4.50 0.91 1.95 4.04 3.07
0.75 0.35 1.83 1.09 0.48 0.20 2.23 1.54
(0.75) (2.00) (0.63) (2.78)
U/t* 0.115
(0.63) (1.66) (0.50) (1.81)
p .610
862.0 1249.5
.007 .997
995.5 809.5
.078 .002
1249.50 867.0 809.50 685.0
.997 .008 .002 <.001
Figures given are mean standard deviation (median). U and t are the test statistics for the ManneWhitney U and two sample t-test respectively (*indicates where t-test was used). SVV and SVH indicate positional errors (in degrees) for vertical and horizontal tests where no frame is present. “Tilted” refers to the mean of the errors (in degrees) generated when the tilted frame was present (data for both tilted frame conditions have been combined). Results in bold indicate significance of <.05.
fell within a range considered normal for SVV (90% within 1.74 degrees) and SVH (90% within 1.37 degrees). Although statistically significant differences were found between the groups with regards to the signed vertical and unsigned horizontal errors, the difference between the median values was less than 0.5 degrees in each case (i.e. one mouse click). The unsigned data from the control group were used to establish an upper limit for the reference range of errors. In the absence of a frame, the upper limit (mean þ 2SD) for SVV was 1.8 degrees compared to 1.2 degrees for SVH. Two control subjects (4%) were found to be outside of this range for both SVV and SVH. In the CNP
30
a
25 Frequency
patients 20
controls
15 10 5 0 0
30
0.5
1 1.5 2 Error (degrees)
b
>2.5
patients
25 Frequency
2.5
controls
20 15 10 5 0 0
0.5
1 1.5 2 Error (degrees)
2.5
>2.5
Fig. 1. Distribution of positioning errors in the absence of a frame for the (a) vertical and (b) horizontal alignment test in CNP patients (n ¼ 50) and asymptomatic controls (n ¼ 50). Errors are unsigned degrees from gravitational vertical and horizontal.
group, 4 (8%) patients were outside of this range for SVV and this increased to 10 (20%) in the case of the SVH test. 3.2.2. Visual reference frame present When the groups were given the visual frame of reference of an untilted frame, their errors were not significantly different from those observed when no reference frame was given for either vertical or horizontal (p > .05, Table 1). Tilting the frame by 18 degrees in either a clockwise (frameþ18) or counter clockwise direction (frame18) resulted in significant increases in positioning error for both patients and controls for the vertical and horizontal tests (p < .001 in all cases). There were no significant differences between the errors produced for frameþ18 and frame18 (p > .05) and so the resultant errors were combined to produce a tilted frame score for each person for each test. The distributions of these errors are shown in Fig. 2a and b respectively. There were significant differences between the CNP and control groups for both the vertical and horizontal test (Table 1). Although the difference between the medians for these tests were still small (less than 1 degree), there were differences in the numbers of subjects falling above the reference range (denoted as dotted lines in Fig. 2). A comparison of the proportion of participants above the reference range (vertical: 4.0 degrees, horizontal: 5.3 degrees) revealed that there were more individuals from the CNP group than controls (vertical: 20% of CNP patients compared to 8% of controls; horizontal: 26% of CNP patients compared to 6% of controls). These are shown in Fig. 3 with the corresponding 95% CI in each case. The median errors for the CNP patients falling into this category were 9.94 degrees and 6.38 degrees for vertical and horizontal, respectively. Of the 50 CNP patients, a subgroup of 8 subjects exhibited higher than normal errors in both the horizontal and vertical tests. Investigating these patients further revealed that they scored higher on the NDI than patients whose errors fell within the reference range (U ¼ 74.0, p ¼ .016) but there was no difference between the subgroups with regards the NRS score (U ¼ 116.0, p ¼ .176). There were also no significant differences between these groups with regards to the type of onset (c2 ¼ 1.086, p ¼ .419) and duration (c2 ¼ 0.501, p ¼ .666) of the neck pain. Comparison of the results for the horizontal and vertical tests showed that both CNP patients and controls had significantly larger errors in the horizontal test (patients: Z ¼ 2.930, p ¼ .003; controls: Z ¼ 2.433, p ¼ .015). There was however a significant positive correlation between the results for the horizontal and vertical tests in both the CNP (Spearman’s rho ¼ 0.733, p < .001) and control (Spearman’s rho ¼ 0.656, p < .001) groups.
136
S. Docherty et al. / Manual Therapy 17 (2012) 133e138 14
tilted clockwiseeframe tilted counter clockwise) for either test. Patients whose errors were recorded as higher than normal were distributed between both groups for vertical (4 unilateral:5 bilateral) and horizontal (8 unilateral:5 bilateral). Individuals making up the group with high errors in both tests consisted of 4 unilateral and 3 bilateral patients.
a
12
Frequency
10 8 6 4
3.4. Direction of errors in unilateral patients
2 0 <0.5 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 >20
Error (degrees) patients
14
controls
b
12
Frequency
10 8 6 4 2 0
<0.5 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 >20
Error (degrees) patients
4. Discussion
controls
Fig. 2. Distribution of positioning errors in the presence of a tilted frame for the (a) vertical and (b) horizontal alignment test in CNP patients (n ¼ 50) and asymptomatic controls (n ¼ 50). Errors are unsigned degrees from gravitational vertical and horizontal. Reference line represents mean þ 2SD limit of controls.
3.3. Unilateral versus bilateral pain To investigate whether the differences observed between the CNP and control groups were influenced by the laterality of the patient’s pain, the CNP group was divided into those with bilateral (18 patients) versus unilateral (29 patients) pain. This information was missing for 3 of the patients and so they were excluded from further analysis. These results are presented in Table 2. Although there are significant differences between the groups with regards to the duration of their condition and the type of onset, there was no difference in either their VAS score or NDI. There were no significant differences between the unilateral and bilateral group with regards to their errors when the frame was tilted in either the horizontal or vertical test (Table 2). Furthermore there was no difference in the amount of asymmetry of error (frame
40.0 Patient
Control
Percentage
30.0
20.0
10.0
0.0 No Frame Vertical
No Frame Horizontal
To investigate if the location of the CNP influenced the direction of the perceptual errors in those patients exhibiting unilateral pain their asymmetry measures were regrouped according to whether the errors were towards or away from the painful side. It was found that there was no significant difference between the errors when the frame was ipsilateral to the pain compared to when the frame was tilted contralateral to the side of the pain (vertical: Z ¼ 0.054, p ¼ .957; horizontal: t ¼ 0.099, p ¼ .922). Of the 29 patients with unilateral CNP, 9 exhibited vertical error asymmetry (difference between the two tilted frame directions greater than 1 degree) in the direction of their painful side and 9 had contralateral asymmetry. The remaining 11 patients exhibited a difference of less than 1 degree between the sides. For horizontal, these figures were 9 ipsilateral, 6 contralateral and 14 with no defined asymmetry.
Frame Vertical
Frame Horizontal
Fig. 3. Proportion of individuals with mean positioning errors greater than 2SD of the corresponding control subject group in the vertical and horizontal alignment tests. The error bars show 95% CI of the proportions (Wilson method).
The current study was the first to investigate the perception of SVH as well as SVV in patients with CNP. In the absence of a surrounding frame, the range of observed errors was similar to those reported in the literature for control groups using both this system (Bagust et al., 2005; Docherty and Bagust, 2010) and other variations (Guerraz et al., 2001; Grod and Diakow, 2002; Pagarkar et al., 2008). There was no significant difference between the errors of the CNP patients and the asymptomatic controls as was found by Grod and Diakow (2002). The results for SVH covered a narrower range and this fell well within the ranges of errors reported by previous studies (Dai et al., 1989; Hafström et al., 2004a,b; Pagarkar et al., 2008). There was a significant difference between the CNP group and the controls for SVH, however this must be viewed with caution as the difference between the median errors was very small. Contrary to the findings of Lopez et al. (2006), there was no significant difference between the errors produced in the absence of a surrounding frame and when the additional visual cues of a surrounding untilted frame were presented. It could be argued that the edges of the video eyeglasses used in the current study provided additional horizontal and vertical cues however, Bagust (2005) reported greater and more significant errors associated with the use of the glasses compared to viewing the test on a computer monitor in a blacked out room. Furthermore, Grod and Diakow (2002) found no significant difference between these frame conditions in either their neck pain patients or controls. Presenting both groups with the conflicting stimulus of a tilted frame surrounding the “rod” resulted in an increase in the magnitude of observed errors. Although the errors for SVH under these conditions were higher than SVV in both CNP and control groups, there was a significant, positive correlation between these scores. For SVV, in spite of there being a statistically significant difference between the CNP and control groups this difference was again less than 0.5 degrees while the difference for SVH was only slighter greater than 0.5 degrees. In studies comparing asymptomatic controls to a particular group of patients, it is important to consider the resolution of the test system being used. In the current study, there were statistically significant differences between the CNP and control groups for SVV
S. Docherty et al. / Manual Therapy 17 (2012) 133e138
137
Table 2 Comparison of patients with unilateral versus bilateral pain.
Pain (VAS) Disability (NDI, %) Duration (% > 1 year)* Onset (% insidious onset)* Tilted vertical ( ) Tilted horizontal ( ) Asymmetry vertical ( ) Asymmetry horizontal ( )
Unilateral (n ¼ 29)
Bilateral (n ¼ 18)
U/c2
p
(4) 30.55 14.56 (26.00) 82.8 72.4 3.12 3.21 (2.00) 3.80 3.11 (2.56) 0.38 1.68 (0.30) 0.49 2.19 (0.40)
(5) 38.75 17.63 (36.00) 77.8 61.1 4.86 6.24 (2.06) 4.53 3.21 (3.63) 0.21 1.64 (0.55) 0.98 4.31 (0.30)
222.5 176.0 17.894 6.149 245.0 204.0 255.0 258.0
.393 .183 <.001 .013 .728 .212 .895 .956
Figures given are mean standard deviation (median). Only the median is given for Pain due to the nature of the data. U and c2 are the test statistics for the ManneWhitney U and chi-square test respectively (*indicates where chi-square test was used). “Tilted” refers to the mean of the errors (in degrees) generated when the tilted frame was present (data for both tilted frame conditions have been combined). “Asymmetry” is the difference between the error (in degrees) generated when the frame is tilted clockwise and when it is tilted counter clockwise. Results in bold indicate significance of <.05.
and both SVV and SVH in the presence of a tilted frame however the differences between the median errors was less than the resolution of the equipment and therefore such results cannot be considered of clinical relevance. A more useful approach would be to compare the distribution of errors compared to a reference range based on the errors exhibited by the control group. As well as resulting in higher errors than the no frame condition, presenting the CNP group with a tilted frame also resulted in an increase in the number of individuals outside of the reference range (4e10 for SVV). These results are not unique; Bagust et al. (2005) found a similar group of neck pain patients in their study. Additionally, studies investigating patients with peripheral nervous system dysfunction found that the addition of a stimulus providing the subject with conflicting sensory input (e.g. tilted body position, vibration of proprioceptors) resulted in increases in the difference between errors observed for patients and controls in the perception of both vertical (Bronstein, 1999; Lopez et al., 2008; Mazibrada et al., 2008) and horizontal (Karlberg et al., 2002). Even though the change was not as marked for SVH (increase from 10 to 13 patients outside range), the majority (8 out of 10) of the patients showing larger than normal errors for SVV with a tilted frame were also present in this subgroup for SVH. Closer inspection of these patients revealed that they had a higher level of reported disability than the others but were not in more pain. Furthermore, in contrast to the findings of Bagust et al. (2005) insidious onset does not seem to be a characteristic of this group of patients. This subgroup of CNP patients appeared to show an increased reliance on visual signals for the determination of SVV and SVH. This is in line with previous descriptions of patients suffering from Parkinson’s disease (Azulay et al., 2002.), or unilateral or bilateral vestibular loss (Bronstein et al., 1996; Hafström et al., 2004a,b; Lopez et al., 2006). Neck proprioceptors have been shown to contribute to the perception of SVV and SVH (Karlberg et al., 1995, 2002). The current observations might be explained if the neck pain is associated with a disturbance of the input from neck proprioceptors resulting in a change in the balance between the sensory systems that the brain uses to determine SVV and SVH, forcing increased emphasis to be placed on the visual input. The patients presenting with unilateral neck pain provided an opportunity to test this suggestion. It was reasoned that if unilateral neck pain were associated with a unilateral proprioceptive disturbance which was compensated by increased visual dependence, an asymmetry in the perceived SVV or SVH would be expected when the visual frame of reference was tilted ipsilateral or contralateral to the neck pain. The results obtained did not provide support for this hypothesis as there was no significant difference between the errors when the frame was ipsilateral to the pain compared to when the frame was tilted contralateral to the side of the pain. The numbers of patients with unilateral pain in the current study was
relatively small, and this aspect of the relationship between SVV and SVH and neck pain would merit further investigation in a larger group of subjects. Alternatively, as has been reported for patients with long term unilateral vestibular dysfunction, compensation mechanisms may be highly individualised (Hafström et al., 2004a; Lopez et al., 2006) and relate to individual differences such as level of field dependence prior to the condition and the degree of ocular torsion induced by the frame tilt (Pansell et al., 2006).
References Anastasopoulos D, Bronstein A, Haslwanter T, Fetter M, Dichgans J. The role of somatosensory input for the perception of verticality. Annals of the New York Academy of Sciences 1999;871(1):379e83. Azulay JP, Mesure S, Amblard B, Pouget J. Increased visual dependence in Parkinson’s disease. Perceptual and Motor Skills 2002;95:1106e14. Bagust J. Assessment of verticality perception by a rod-and-frame test: preliminary observations on the use of a computer monitor and video eye glasses. Archives of Physical Medicine and Rehabilitation 2005;86:1062e4. Bagust J, Rix GD, Hurst HC. Use of a computer rod and frame (CRAF) test to assess errors in the perception of visual vertical in a clinical setting e a pilot study. Clinical Chiropractic 2005;8(3):134e9. Benyamin R, Singh V, Parr AT, Conn A, Diwan S, Abdi S. Systematic review of the effectiveness of cervical epidurals in the management of chronic neck pain. Pain Physician 2009;12(1):137e57. Betts GA, Curthoys IS. Visually perceived vertical and visually perceived horizontal are not orthogonal. Vision Research 1998;38:1989e99. Bolton JE, Wilkinson RC. Responsiveness of pain scales: a comparison of three pain intensity measures in chiropractic patients. Journal of Manipulative and Physiological Therapeutics 1998;21:1e7. Bronstein AM. The interaction of otolith and proprioceptive information in the perception of verticality: the effects of labyrinthine and CNS disease. Annals of the New York Academy of Sciences 1999;871(1):324e33. Bronstein AM, Yardley L, Moore AP, Cleeves L. Visually and posturally mediated tilt illusion in Parkinson’s disease and in labyrinthine defective subjects. Neurology 1996;47:651e6. Dai MJ, Curthoys IS, Halmagyi GM. Linear acceleration perception in the roll plane before and after unilateral vestibular neurectomy. Experimental Brain Research 1989;77(2):315e28. Docherty S, Bagust J. From line to dots: an improved computerised rod and frame system for testing subjective visual vertical and horizontal. BMC Research Notes 2010;3(1):9. Falla D, Bilenkij G, Jull G. Patients with chronic neck pain demonstrate altered patterns of muscle activation during performance of a functional upper limb task. Spine 2004;29(13):1436e40. Grod JP, Diakow PR. Effect of neck pain on verticality perception: a cohort study. Archives of Physical Medicine and Rehabilitation 2002;83(3):412e5. Guerraz M, Yardley L, Bertholon P, Pollak L, Rudge P, Gresty MA, et al. Visual vertigo: symptom assessment, spatial orientation and postural control. Brain 2001; 124(8):1646e56. Hafström A, Fransson P-A, Karlberg M, Magnusson M. Idiosyncratic compensation of the subjective visual horizontal and vertical in 60 patients after unilateral vestibular deafferentation. Acta Oto-Laryngologica 2004a;124(2): 165e71. Hafström A, Fransson P-A, Karlberg M, Magnusson M. Ipsilesional visual field dependency for patients with vestibular schwannoma. Neuroreport 2004; 15(14):2201e4. Johnston V, Jimmieson NL, Jull G, Souvlis T. Quantitative sensory measures distinguish office workers with varying levels of neck pain and disability. Pain 2008; 137(2):257e65.
138
S. Docherty et al. / Manual Therapy 17 (2012) 133e138
Karlberg M, Aw ST, Halmagyi GM, Black RA. Vibration-induced shift of the subjective visual horizontal: a sign of unilateral vestibular deficit. Archives of Otolaryngology e Head & Neck Surgery 2002;128(1):21e7. Karlberg M, Persson L, Magnusson M. Reduced postural control in patients with chronic cervicobrachial pain syndrome. Gait & Posture 1995;3:241e9. Kumagami H, Saino Y, Baba A, Fujiyama D, Takasaki K, Takahashi H. Subjective visual vertical test in patients with chronic dizziness without abnormal findings in routine vestibular function tests. Acta Oto-Laryngologica 2009;129: 46e9. Lopez C, Lacour M, Léonard J, Magnan J, Borel L. How body position changes visual vertical perception after unilateral vestibular loss. Neuropsychologia 2008; 46(9):2435e40. Lopez C, Lacour M, Magnan J, Borel L. Visual field dependenceeindependence before and after unilateral vestibular loss. Neuroreport 2006;17(8):797e803. Mazibrada G, Tariq S, Pérennou D, Gresty M, Greenwood R, Bronstein AM. The peripheral nervous system and the perception of verticality. Gait & Posture 2008;27(2):202e8.
Newcombe RG, Altman DG. Proportions and their differences. In: Altman DG, Machin D, Bryant TN, Gardner MJ, editors. Statistics with confidence. London: British Medical Journal; 2000. p. 45e56. Pagarkar W, Bamiou D-E, Ridout D, Luxon LM. Subjective visual vertical and horizontal: effect of the preset angle. Archives of Otolaryngology e Head & Neck Surgery 2008;134(4):394e401. Pansell T, Sverersten U, Ygge J. Visual spatial clues enhance ocular torsion response during visual tilt. Experimental Brain Research 2006;175:567e74. Passatore M, Roatta S. Influence of sympathetic nervous system on sensorimotor function: whiplash associated disorders (WAD) as a model. European Journal of Applied Physiology 2006;98(5):423e49. Saj A, Honore J, Bernati T, Coello Y, Rousseaux M. Subjective visual vertical in pitch and roll in right hemispheric stroke. Stroke 2005a;36(3):588e91. Saj A, Honore J, Davroux J, Coello Y, Rousseaux M. Effect of posture on the perception of verticality in neglect patients. Stroke 2005b;36(10):2203e5. Vernon H. The neck disability index: state-of-the-art, 1991e2008. Journal of Manipulative and Physiological Therapeutics 2008;31(7):491e502.
Contents lists available at SciVerse ScienceDirect
Manual Therapy journal homepage: www.elsevier.com/math
Original article
Perception of subjective visual vertical and horizontal in patients with chronic neck pain: A cross-sectional observational study Sharon Docherty a, *, Rebekka Schärer b, Jeff Bagust a, B. Kim Humphreys b a b
Anglo-European College of Chiropractic, 13-15 Parkwood Road, Bournemouth BH5 2DF, UK University Orthopaedic Hospital of Balgrist, Forchstrasse 340, 8008 Zurich, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 April 2011 Received in revised form 31 October 2011 Accepted 8 November 2011
Previous studies have shown that chronic neck pain (CNP) patients have a larger spread of perceptual errors for subjective visual vertical (SVV) than those exhibited by asymptomatic controls. The current study investigated whether this was also the case for perception of subjective visual horizontal (SVH) and whether there was a correlation between the two measurements. Fifty patients with CNP were compared with a group of 50 age- and gender-matched controls. All subjects were required to complete a test to measure SVH as well as SVV using the computerised rod and frame (CRAF) test. These tests were conducted under various frame conditions. No difference was found between the errors of the CNP and control groups in the absence of a surrounding frame. When a tilted frame was added to the CRAF test, the range of errors observed in the CNP group increased for both SVV and SVH. In particular, significantly more CNP patients fell outside the reference range of errors and a subgroup of patients, characterised by higher neck pain disability indices, was identified who demonstrated higher than expected errors for both SVV and SVH. However no conclusion could be drawn with regards to the direction of error asymmetry and laterality of pain as those patients with unilateral pain exhibited errors both towards and away from the affected area. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Subjective visual vertical Subjective visual horizontal Chronic neck pain
1. Introduction Very little is known about the causes of chronic neck pain (CNP) (Benyamin et al., 2009) this, in part, could be due to the wide variety of signs and symptoms that are present in this group of patients (Passatore and Roatta, 2006). Altered movement and muscle activation patterns in the neck and shoulder region are among these signs (Falla et al., 2004) as well as reduced postural control (Karlberg et al., 1995). Subjective visual vertical (SVV) and horizontal (SVH), are measures of an individual’s internal representation of their position in space relative to gravity, and therefore important in the maintenance of posture and equilibrium (Mazibrada et al., 2008). Disturbances to this perception can occur as a result of errors in sensory input from one of the three major systems involved; the vestibular; visual; and proprioceptive systems, or at the level of integration of these signals within the brain (Anastasopoulos et al., 1999). As a result, measurements of SVV, and to a lesser extent SVH, are increasingly being employed in the assessment of conditions * Corresponding author. Tel.: þ44 1202 436 242; fax: þ44 1202 436 312. E-mail address: [email protected] (S. Docherty). 1356-689X/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.math.2011.11.002
such as Menière’s Disease, vestibular dysfunction and dizziness (Guerraz et al., 2001; Hafström et al., 2004a,b; Mazibrada et al., 2008; Pagarkar et al., 2008; Kumagami et al., 2009) as well as brain injury due to, for example, stroke (Saj et al., 2005a,b). Based on work using healthy subjects, Betts and Curthoys (1998) proposed that it is incorrect to conclude that perception of SVV and SVH are “invariably orthogonal to one another”. However, Hafström et al. (2004a,b) have demonstrated a high correlation between the errors produced for SVV and SVH in patients with unilateral vestibular deafferentation and suggested that the two measures are of equal clinical relevance for this group. Previous studies conducted on CNP patients have shown that they have significantly larger errors in SVV, than asymptomatic controls (Grod and Diakow, 2002; Bagust et al., 2005). In particular, Bagust et al. (2005) found a wider range of errors in the neck pain group compared to the controls, suggestive that there may be a subgroup of neck pain patients who have associated disturbances of their perception of verticality. The aim of the current study was to investigate whether CNP patients exhibit increased errors in subjective visual horizontal (SVH), not previously investigated in this group of patients, as well as SVV, compared to asymptomatic controls. In addition, we
134
S. Docherty et al. / Manual Therapy 17 (2012) 133e138
examined whether there was a relationship between laterality of pain and asymmetry of perception of SVV and SVH. 2. Methods 2.1. Participants Fifty participants with CNP were recruited from four Chiropractic Clinics in the Canton of Zürich and from the Physiotherapy Department of the University Orthopaedic Hospital of Balgrist, Zürich. To be included, participants must have been between 30 and 65 years of age and experienced neck pain with or without arm pain for at least four weeks duration. A control group of 50 age- and gender-matched subjects was recruited at the University Orthopaedic Hospital of Balgrist and by advertising in a local paper. Control subjects had not experienced significant neck or upper thoracic pain in the previous month i.e. pain that had limited their daily activities or for which they had sought care. Subjects were excluded from the study if they had a recent history of cervical spine surgery, bone pathology such as neoplasm, infection or serious congenital anomalies, vestibulo-cochlear disease affecting balance or medication use which may affect balance, vision and co-ordination. All subjects gave their written consent to participate after being informed about the aims and methods of the study. The study was approved by the Zürich Cantonal Ethics Committee. For the CNP group, pain intensity was measured using the 11point (0e10) Numerical Rating Scale (NRS). The NRS is the most frequently used scale for measuring pain intensity and has been shown to be psychometrically robust and practicable (Bolton and Wilkinson, 1998). In addition, patients’ self-reported disability was measured using the Neck Disability Index (NDI). The NDI is the most commonly used questionnaire for neck pain research (Vernon, 2008). It consists of 10 categories in which a score of 0e5 is possible for each category. Total NDI scores of 30 are considered the cut-off for moderate levels of pain and disability (Johnston et al., 2008). The duration, type of onset and laterality of symptoms were also recorded for each patient. 2.2. Procedures The CNP group completed the NRS for pain intensity and the NDI for neck pain disability. Control participants were asked if they were currently experiencing any neck or upper back pain to ensure that they were asymptomatic at the time of testing. All participants were required to complete both of the tests for the perception of vertical and horizontal. The order in which the two tests were performed was randomly assigned for each subject. These were performed with the subject sitting in a comfortable position, their head was not restrained. Between tests subjects were given a short break. The tests consisted of a modified version of the computerised rod and frame (CRAF) test (see Docherty and Bagust, 2010) viewed by the participant through a pair of Olympus Eye-trek FMD 200 video glasses. Wearing the video eyeglasses resulted in an image that spanned a viewing angle of 30 23 degrees of the visual field (the equivalent of viewing a 1.42 m screen from a distance of 2 m). Both the horizontal and vertical tests comprised 18 presentations, the first two of which were for instruction purposes and were not included in the analysis. The first presentation in each test always contained the upright, untilted frame. This was used to confirm that the glasses were positioned correctly. The second presentation comprised of a tilted frame and was used to establish whether the
subject understood the task. The remaining 16 presentations consisted of four replicates where the frame (a white square presented on a homogenous black background) was either e (i) absent (giving no visual references); (ii) untilted (frame0, providing visual reference cues for vertical and horizontal); tilted 18 degrees (providing confusing visual reference cues) in either a (iii) clockwise (frameþ18,); or (iv) counter clockwise direction (frame18). The two white dots used to represent the ends of the ‘rod’ were displayed in the centre of the screen (within the frame if present) and had two starting positions e tilted 20 degrees in either a clockwise or counter clockwise direction from gravitational vertical/horizontal. The order of presentation of these permutations of frame and dots was assigned by the computer from a bank of 4 randomised sequences. The subject’s task was to rotate the dots using the right and left mouse buttons to a position perceived to be vertical or horizontal depending on the test. The dots rotated around their virtual midpoint in 0.5 degree increments. When the participant was satisfied with the alignment of the dots, the programme was advanced to the next presentation by pressing the space bar of the computer keyboard. This task took approximately 11 s per presentation. Positioning error was recorded by the computer as degrees from gravitational vertical/horizontal. Neither the subject nor the operator had access to the recorded errors until the end of the recording session. 2.3. Statistical analysis Recorded errors from the CRAF test were used to calculate the signed and unsigned (absolute) means for the four frame condition (n ¼ 4 in each case) for each participant. Unless otherwise stated, reported errors are the absolute (unsigned) values. All statistical analyses were performed using SPSS 17.0 and Instat (Graphpad, Inc). Data were tested for normality using the KolmogoroveSmirnov method. In general, the data obtained from patients did not follow a normal distribution and so non parametric statistics were used. To test for differences between the errors produced under different frame conditions, a Repeated Measures Analysis of Variance (ANOVA) was used with post hoc TukeyeKramer Multiple Comparison Tests. All control group data were normally distributed and so the reference ranges for error distributions were calculated using mean þ 2SD of these data. Unless otherwise stated this was based on the unsigned data. Differences between neck pain patients and controls were investigated using the ManneWhitney U test for 2 independent samples. The 95% confidence intervals (CIs) for the proportion of participants with errors greater than 2SD of the control means were calculated using the Wilson method and to illustrate differences between the proportion of patients and controls in this category, the Newcombe method was used to calculate the CIs (Newcombe and Altman, 2000). 3. Results 3.1. Participants Both the CNP and control group consisted of 50 individuals of whom 40 were female. There was no significant difference between the groups with regards to age (mean 48 years in each, Table 1). In the CNP group, the median pain rating was 4.5 on the NRS with a mean (SD) NDI of 33 16. 3.2. Comparison of CNP and control groups 3.2.1. No visual reference frame As can be seen from Table 1 and Fig. 1, in the absence of the cues given by a surrounding frame, both CNP and control mean errors
S. Docherty et al. / Manual Therapy 17 (2012) 133e138
135
Table 1 Comparison of CNP patients (n ¼ 50) and asymptomatic controls (n ¼ 50). CNP (n ¼ 50)
Controls (n ¼ 50)
Age (years)*
48.1 8.7
47.9 8.7
SVV Signed ( ) Unsigned ( )
0.03 0.66 (0.00) 0.90 0.51 (0.75)
0.37 0.75 (0.25) 0.87 0.47 (0.81)
SVH Signed ( ) Unsigned ( )
0.07 1.16 (0.25) 1.04 1.71 (0.75)
0.01 0.48 (0.00) 0.56 0.30 (0.62)
Frame effect Untilted vertical ( ) Tilted vertical ( ) Untilted horizontal ( ) Tilted horizontal ( )
0.87 0.47 3.73 4.50 0.91 1.95 4.04 3.07
0.75 0.35 1.83 1.09 0.48 0.20 2.23 1.54
(0.75) (2.00) (0.63) (2.78)
U/t* 0.115
(0.63) (1.66) (0.50) (1.81)
p .610
862.0 1249.5
.007 .997
995.5 809.5
.078 .002
1249.50 867.0 809.50 685.0
.997 .008 .002 <.001
Figures given are mean standard deviation (median). U and t are the test statistics for the ManneWhitney U and two sample t-test respectively (*indicates where t-test was used). SVV and SVH indicate positional errors (in degrees) for vertical and horizontal tests where no frame is present. “Tilted” refers to the mean of the errors (in degrees) generated when the tilted frame was present (data for both tilted frame conditions have been combined). Results in bold indicate significance of <.05.
fell within a range considered normal for SVV (90% within 1.74 degrees) and SVH (90% within 1.37 degrees). Although statistically significant differences were found between the groups with regards to the signed vertical and unsigned horizontal errors, the difference between the median values was less than 0.5 degrees in each case (i.e. one mouse click). The unsigned data from the control group were used to establish an upper limit for the reference range of errors. In the absence of a frame, the upper limit (mean þ 2SD) for SVV was 1.8 degrees compared to 1.2 degrees for SVH. Two control subjects (4%) were found to be outside of this range for both SVV and SVH. In the CNP
30
a
25 Frequency
patients 20
controls
15 10 5 0 0
30
0.5
1 1.5 2 Error (degrees)
b
>2.5
patients
25 Frequency
2.5
controls
20 15 10 5 0 0
0.5
1 1.5 2 Error (degrees)
2.5
>2.5
Fig. 1. Distribution of positioning errors in the absence of a frame for the (a) vertical and (b) horizontal alignment test in CNP patients (n ¼ 50) and asymptomatic controls (n ¼ 50). Errors are unsigned degrees from gravitational vertical and horizontal.
group, 4 (8%) patients were outside of this range for SVV and this increased to 10 (20%) in the case of the SVH test. 3.2.2. Visual reference frame present When the groups were given the visual frame of reference of an untilted frame, their errors were not significantly different from those observed when no reference frame was given for either vertical or horizontal (p > .05, Table 1). Tilting the frame by 18 degrees in either a clockwise (frameþ18) or counter clockwise direction (frame18) resulted in significant increases in positioning error for both patients and controls for the vertical and horizontal tests (p < .001 in all cases). There were no significant differences between the errors produced for frameþ18 and frame18 (p > .05) and so the resultant errors were combined to produce a tilted frame score for each person for each test. The distributions of these errors are shown in Fig. 2a and b respectively. There were significant differences between the CNP and control groups for both the vertical and horizontal test (Table 1). Although the difference between the medians for these tests were still small (less than 1 degree), there were differences in the numbers of subjects falling above the reference range (denoted as dotted lines in Fig. 2). A comparison of the proportion of participants above the reference range (vertical: 4.0 degrees, horizontal: 5.3 degrees) revealed that there were more individuals from the CNP group than controls (vertical: 20% of CNP patients compared to 8% of controls; horizontal: 26% of CNP patients compared to 6% of controls). These are shown in Fig. 3 with the corresponding 95% CI in each case. The median errors for the CNP patients falling into this category were 9.94 degrees and 6.38 degrees for vertical and horizontal, respectively. Of the 50 CNP patients, a subgroup of 8 subjects exhibited higher than normal errors in both the horizontal and vertical tests. Investigating these patients further revealed that they scored higher on the NDI than patients whose errors fell within the reference range (U ¼ 74.0, p ¼ .016) but there was no difference between the subgroups with regards the NRS score (U ¼ 116.0, p ¼ .176). There were also no significant differences between these groups with regards to the type of onset (c2 ¼ 1.086, p ¼ .419) and duration (c2 ¼ 0.501, p ¼ .666) of the neck pain. Comparison of the results for the horizontal and vertical tests showed that both CNP patients and controls had significantly larger errors in the horizontal test (patients: Z ¼ 2.930, p ¼ .003; controls: Z ¼ 2.433, p ¼ .015). There was however a significant positive correlation between the results for the horizontal and vertical tests in both the CNP (Spearman’s rho ¼ 0.733, p < .001) and control (Spearman’s rho ¼ 0.656, p < .001) groups.
136
S. Docherty et al. / Manual Therapy 17 (2012) 133e138 14
tilted clockwiseeframe tilted counter clockwise) for either test. Patients whose errors were recorded as higher than normal were distributed between both groups for vertical (4 unilateral:5 bilateral) and horizontal (8 unilateral:5 bilateral). Individuals making up the group with high errors in both tests consisted of 4 unilateral and 3 bilateral patients.
a
12
Frequency
10 8 6 4
3.4. Direction of errors in unilateral patients
2 0 <0.5 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 >20
Error (degrees) patients
14
controls
b
12
Frequency
10 8 6 4 2 0
<0.5 1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 >20
Error (degrees) patients
4. Discussion
controls
Fig. 2. Distribution of positioning errors in the presence of a tilted frame for the (a) vertical and (b) horizontal alignment test in CNP patients (n ¼ 50) and asymptomatic controls (n ¼ 50). Errors are unsigned degrees from gravitational vertical and horizontal. Reference line represents mean þ 2SD limit of controls.
3.3. Unilateral versus bilateral pain To investigate whether the differences observed between the CNP and control groups were influenced by the laterality of the patient’s pain, the CNP group was divided into those with bilateral (18 patients) versus unilateral (29 patients) pain. This information was missing for 3 of the patients and so they were excluded from further analysis. These results are presented in Table 2. Although there are significant differences between the groups with regards to the duration of their condition and the type of onset, there was no difference in either their VAS score or NDI. There were no significant differences between the unilateral and bilateral group with regards to their errors when the frame was tilted in either the horizontal or vertical test (Table 2). Furthermore there was no difference in the amount of asymmetry of error (frame
40.0 Patient
Control
Percentage
30.0
20.0
10.0
0.0 No Frame Vertical
No Frame Horizontal
To investigate if the location of the CNP influenced the direction of the perceptual errors in those patients exhibiting unilateral pain their asymmetry measures were regrouped according to whether the errors were towards or away from the painful side. It was found that there was no significant difference between the errors when the frame was ipsilateral to the pain compared to when the frame was tilted contralateral to the side of the pain (vertical: Z ¼ 0.054, p ¼ .957; horizontal: t ¼ 0.099, p ¼ .922). Of the 29 patients with unilateral CNP, 9 exhibited vertical error asymmetry (difference between the two tilted frame directions greater than 1 degree) in the direction of their painful side and 9 had contralateral asymmetry. The remaining 11 patients exhibited a difference of less than 1 degree between the sides. For horizontal, these figures were 9 ipsilateral, 6 contralateral and 14 with no defined asymmetry.
Frame Vertical
Frame Horizontal
Fig. 3. Proportion of individuals with mean positioning errors greater than 2SD of the corresponding control subject group in the vertical and horizontal alignment tests. The error bars show 95% CI of the proportions (Wilson method).
The current study was the first to investigate the perception of SVH as well as SVV in patients with CNP. In the absence of a surrounding frame, the range of observed errors was similar to those reported in the literature for control groups using both this system (Bagust et al., 2005; Docherty and Bagust, 2010) and other variations (Guerraz et al., 2001; Grod and Diakow, 2002; Pagarkar et al., 2008). There was no significant difference between the errors of the CNP patients and the asymptomatic controls as was found by Grod and Diakow (2002). The results for SVH covered a narrower range and this fell well within the ranges of errors reported by previous studies (Dai et al., 1989; Hafström et al., 2004a,b; Pagarkar et al., 2008). There was a significant difference between the CNP group and the controls for SVH, however this must be viewed with caution as the difference between the median errors was very small. Contrary to the findings of Lopez et al. (2006), there was no significant difference between the errors produced in the absence of a surrounding frame and when the additional visual cues of a surrounding untilted frame were presented. It could be argued that the edges of the video eyeglasses used in the current study provided additional horizontal and vertical cues however, Bagust (2005) reported greater and more significant errors associated with the use of the glasses compared to viewing the test on a computer monitor in a blacked out room. Furthermore, Grod and Diakow (2002) found no significant difference between these frame conditions in either their neck pain patients or controls. Presenting both groups with the conflicting stimulus of a tilted frame surrounding the “rod” resulted in an increase in the magnitude of observed errors. Although the errors for SVH under these conditions were higher than SVV in both CNP and control groups, there was a significant, positive correlation between these scores. For SVV, in spite of there being a statistically significant difference between the CNP and control groups this difference was again less than 0.5 degrees while the difference for SVH was only slighter greater than 0.5 degrees. In studies comparing asymptomatic controls to a particular group of patients, it is important to consider the resolution of the test system being used. In the current study, there were statistically significant differences between the CNP and control groups for SVV
S. Docherty et al. / Manual Therapy 17 (2012) 133e138
137
Table 2 Comparison of patients with unilateral versus bilateral pain.
Pain (VAS) Disability (NDI, %) Duration (% > 1 year)* Onset (% insidious onset)* Tilted vertical ( ) Tilted horizontal ( ) Asymmetry vertical ( ) Asymmetry horizontal ( )
Unilateral (n ¼ 29)
Bilateral (n ¼ 18)
U/c2
p
(4) 30.55 14.56 (26.00) 82.8 72.4 3.12 3.21 (2.00) 3.80 3.11 (2.56) 0.38 1.68 (0.30) 0.49 2.19 (0.40)
(5) 38.75 17.63 (36.00) 77.8 61.1 4.86 6.24 (2.06) 4.53 3.21 (3.63) 0.21 1.64 (0.55) 0.98 4.31 (0.30)
222.5 176.0 17.894 6.149 245.0 204.0 255.0 258.0
.393 .183 <.001 .013 .728 .212 .895 .956
Figures given are mean standard deviation (median). Only the median is given for Pain due to the nature of the data. U and c2 are the test statistics for the ManneWhitney U and chi-square test respectively (*indicates where chi-square test was used). “Tilted” refers to the mean of the errors (in degrees) generated when the tilted frame was present (data for both tilted frame conditions have been combined). “Asymmetry” is the difference between the error (in degrees) generated when the frame is tilted clockwise and when it is tilted counter clockwise. Results in bold indicate significance of <.05.
and both SVV and SVH in the presence of a tilted frame however the differences between the median errors was less than the resolution of the equipment and therefore such results cannot be considered of clinical relevance. A more useful approach would be to compare the distribution of errors compared to a reference range based on the errors exhibited by the control group. As well as resulting in higher errors than the no frame condition, presenting the CNP group with a tilted frame also resulted in an increase in the number of individuals outside of the reference range (4e10 for SVV). These results are not unique; Bagust et al. (2005) found a similar group of neck pain patients in their study. Additionally, studies investigating patients with peripheral nervous system dysfunction found that the addition of a stimulus providing the subject with conflicting sensory input (e.g. tilted body position, vibration of proprioceptors) resulted in increases in the difference between errors observed for patients and controls in the perception of both vertical (Bronstein, 1999; Lopez et al., 2008; Mazibrada et al., 2008) and horizontal (Karlberg et al., 2002). Even though the change was not as marked for SVH (increase from 10 to 13 patients outside range), the majority (8 out of 10) of the patients showing larger than normal errors for SVV with a tilted frame were also present in this subgroup for SVH. Closer inspection of these patients revealed that they had a higher level of reported disability than the others but were not in more pain. Furthermore, in contrast to the findings of Bagust et al. (2005) insidious onset does not seem to be a characteristic of this group of patients. This subgroup of CNP patients appeared to show an increased reliance on visual signals for the determination of SVV and SVH. This is in line with previous descriptions of patients suffering from Parkinson’s disease (Azulay et al., 2002.), or unilateral or bilateral vestibular loss (Bronstein et al., 1996; Hafström et al., 2004a,b; Lopez et al., 2006). Neck proprioceptors have been shown to contribute to the perception of SVV and SVH (Karlberg et al., 1995, 2002). The current observations might be explained if the neck pain is associated with a disturbance of the input from neck proprioceptors resulting in a change in the balance between the sensory systems that the brain uses to determine SVV and SVH, forcing increased emphasis to be placed on the visual input. The patients presenting with unilateral neck pain provided an opportunity to test this suggestion. It was reasoned that if unilateral neck pain were associated with a unilateral proprioceptive disturbance which was compensated by increased visual dependence, an asymmetry in the perceived SVV or SVH would be expected when the visual frame of reference was tilted ipsilateral or contralateral to the neck pain. The results obtained did not provide support for this hypothesis as there was no significant difference between the errors when the frame was ipsilateral to the pain compared to when the frame was tilted contralateral to the side of the pain. The numbers of patients with unilateral pain in the current study was
relatively small, and this aspect of the relationship between SVV and SVH and neck pain would merit further investigation in a larger group of subjects. Alternatively, as has been reported for patients with long term unilateral vestibular dysfunction, compensation mechanisms may be highly individualised (Hafström et al., 2004a; Lopez et al., 2006) and relate to individual differences such as level of field dependence prior to the condition and the degree of ocular torsion induced by the frame tilt (Pansell et al., 2006).
References Anastasopoulos D, Bronstein A, Haslwanter T, Fetter M, Dichgans J. The role of somatosensory input for the perception of verticality. Annals of the New York Academy of Sciences 1999;871(1):379e83. Azulay JP, Mesure S, Amblard B, Pouget J. Increased visual dependence in Parkinson’s disease. Perceptual and Motor Skills 2002;95:1106e14. Bagust J. Assessment of verticality perception by a rod-and-frame test: preliminary observations on the use of a computer monitor and video eye glasses. Archives of Physical Medicine and Rehabilitation 2005;86:1062e4. Bagust J, Rix GD, Hurst HC. Use of a computer rod and frame (CRAF) test to assess errors in the perception of visual vertical in a clinical setting e a pilot study. Clinical Chiropractic 2005;8(3):134e9. Benyamin R, Singh V, Parr AT, Conn A, Diwan S, Abdi S. Systematic review of the effectiveness of cervical epidurals in the management of chronic neck pain. Pain Physician 2009;12(1):137e57. Betts GA, Curthoys IS. Visually perceived vertical and visually perceived horizontal are not orthogonal. Vision Research 1998;38:1989e99. Bolton JE, Wilkinson RC. Responsiveness of pain scales: a comparison of three pain intensity measures in chiropractic patients. Journal of Manipulative and Physiological Therapeutics 1998;21:1e7. Bronstein AM. The interaction of otolith and proprioceptive information in the perception of verticality: the effects of labyrinthine and CNS disease. Annals of the New York Academy of Sciences 1999;871(1):324e33. Bronstein AM, Yardley L, Moore AP, Cleeves L. Visually and posturally mediated tilt illusion in Parkinson’s disease and in labyrinthine defective subjects. Neurology 1996;47:651e6. Dai MJ, Curthoys IS, Halmagyi GM. Linear acceleration perception in the roll plane before and after unilateral vestibular neurectomy. Experimental Brain Research 1989;77(2):315e28. Docherty S, Bagust J. From line to dots: an improved computerised rod and frame system for testing subjective visual vertical and horizontal. BMC Research Notes 2010;3(1):9. Falla D, Bilenkij G, Jull G. Patients with chronic neck pain demonstrate altered patterns of muscle activation during performance of a functional upper limb task. Spine 2004;29(13):1436e40. Grod JP, Diakow PR. Effect of neck pain on verticality perception: a cohort study. Archives of Physical Medicine and Rehabilitation 2002;83(3):412e5. Guerraz M, Yardley L, Bertholon P, Pollak L, Rudge P, Gresty MA, et al. Visual vertigo: symptom assessment, spatial orientation and postural control. Brain 2001; 124(8):1646e56. Hafström A, Fransson P-A, Karlberg M, Magnusson M. Idiosyncratic compensation of the subjective visual horizontal and vertical in 60 patients after unilateral vestibular deafferentation. Acta Oto-Laryngologica 2004a;124(2): 165e71. Hafström A, Fransson P-A, Karlberg M, Magnusson M. Ipsilesional visual field dependency for patients with vestibular schwannoma. Neuroreport 2004; 15(14):2201e4. Johnston V, Jimmieson NL, Jull G, Souvlis T. Quantitative sensory measures distinguish office workers with varying levels of neck pain and disability. Pain 2008; 137(2):257e65.
138
S. Docherty et al. / Manual Therapy 17 (2012) 133e138
Karlberg M, Aw ST, Halmagyi GM, Black RA. Vibration-induced shift of the subjective visual horizontal: a sign of unilateral vestibular deficit. Archives of Otolaryngology e Head & Neck Surgery 2002;128(1):21e7. Karlberg M, Persson L, Magnusson M. Reduced postural control in patients with chronic cervicobrachial pain syndrome. Gait & Posture 1995;3:241e9. Kumagami H, Saino Y, Baba A, Fujiyama D, Takasaki K, Takahashi H. Subjective visual vertical test in patients with chronic dizziness without abnormal findings in routine vestibular function tests. Acta Oto-Laryngologica 2009;129: 46e9. Lopez C, Lacour M, Léonard J, Magnan J, Borel L. How body position changes visual vertical perception after unilateral vestibular loss. Neuropsychologia 2008; 46(9):2435e40. Lopez C, Lacour M, Magnan J, Borel L. Visual field dependenceeindependence before and after unilateral vestibular loss. Neuroreport 2006;17(8):797e803. Mazibrada G, Tariq S, Pérennou D, Gresty M, Greenwood R, Bronstein AM. The peripheral nervous system and the perception of verticality. Gait & Posture 2008;27(2):202e8.
Newcombe RG, Altman DG. Proportions and their differences. In: Altman DG, Machin D, Bryant TN, Gardner MJ, editors. Statistics with confidence. London: British Medical Journal; 2000. p. 45e56. Pagarkar W, Bamiou D-E, Ridout D, Luxon LM. Subjective visual vertical and horizontal: effect of the preset angle. Archives of Otolaryngology e Head & Neck Surgery 2008;134(4):394e401. Pansell T, Sverersten U, Ygge J. Visual spatial clues enhance ocular torsion response during visual tilt. Experimental Brain Research 2006;175:567e74. Passatore M, Roatta S. Influence of sympathetic nervous system on sensorimotor function: whiplash associated disorders (WAD) as a model. European Journal of Applied Physiology 2006;98(5):423e49. Saj A, Honore J, Bernati T, Coello Y, Rousseaux M. Subjective visual vertical in pitch and roll in right hemispheric stroke. Stroke 2005a;36(3):588e91. Saj A, Honore J, Davroux J, Coello Y, Rousseaux M. Effect of posture on the perception of verticality in neglect patients. Stroke 2005b;36(10):2203e5. Vernon H. The neck disability index: state-of-the-art, 1991e2008. Journal of Manipulative and Physiological Therapeutics 2008;31(7):491e502.