Virtual Reality

, Volume 19, Issue 3, pp 267–275

Simulator sickness incidence and susceptibility during neck motion-controlled virtual reality tasks

  • Julia Treleaven
  • Jenna Battershill
  • Deborah Cole
  • Carissa Fadelli
  • Simon Freestone
  • Katie Lang
  • Hilla Sarig-Bahat
Original Article

DOI: 10.1007/s10055-015-0266-4

Cite this article as:
Treleaven, J., Battershill, J., Cole, D. et al. Virtual Reality (2015) 19: 267. doi:10.1007/s10055-015-0266-4
  • 750 Downloads

Abstract

To determine the incidence, severity, and predisposing factors to simulator sickness (SS) when using the neck virtual reality (VR) device in asymptomatic individuals to understand the risk of provoking SS in the development of neck VR as a rehabilitation tool. Thirty-two participants used the VR system. Postural stability was measured before and after each VR module [range of motion (ROM), velocity, and accuracy]. The duration of each module was recorded, and participants reported their SS using a visual analogue scale (SS–VAS)/100 mm. Following the VR assessment, participants completed the Motion Sickness Susceptibility Questionnaire (MSSQ) (child and adult subsections) and Simulator Sickness Questionnaire (SSQ). The incidence of motion sickness during the VR immersion was 28 %, and the mean severity was 17.2 mm on VAS. There was a significant difference in ROM time, total time, MSSQ score, and SSQ score (p < 0.05) between those who reported any level of SS–VAS and those with no SS–VAS. The SS–VAS score displayed significant positive correlations with SSQ score, change in postural stability time pre to post, ROM time, and total time. Results indicate a relatively high incidence but low severity of SS which was associated with the MSSQ child subsection score and exposure time.

Keywords

Virtual reality Motion sickness Velocity Neck Rehabilitation 

1 Introduction

Patients with neck pain tend to demonstrate several impairments related to physical function including altered neck movement kinematics, such as reduced neck movement velocity and smoothness (Feipel et al. 1999; Öhberg et al. 2003; Röijezon et al. 2010; Sarig-Bahat et al. 2010; Sjölander et al. 2008). Considering the importance of quick and precise head movement for daily function (Corneil et al. 2002; Liu et al. 2003; Selbie et al. 1993), rehabilitation addressing these qualities may be important for the management of chronic neck pain.

A recent and novel method to assess and manage neck kinematics involves using virtual reality (VR) including a head-mounted display (HMD) where the interaction is controlled by neck movement (Sarig-Bahat et al. 2009, 2010). VR has been recognised as a potentially effective modality for neck rehabilitation due to its potential to reduce pain and anxiety levels (Sharar et al. 2008), be engaging and motivating, and improve exercise compliance and motor learning (Bryanton et al. 2006; Mirelman et al. 2009; Rizzo and Kim 2005; Wulf and Su 2007). Despite these benefits, an adverse impact of being immersed in a VR environment is motion sickness (MS) (Golding and Gresty 2005). In fact, it has been reported that about 80–95 % of individuals interacting with a HMD experience some level of MS, with 5–50 % experiencing symptoms severe enough to end participation (Ling et al. 2013).

The four most documented signs and symptoms of MS include pallor, cold sweating, nausea, and vomiting (Bos et al. 2008; Eisenman 2009; Money 1970; Warwick-Evans et al. 1998) and, when induced in a VR environment, are collectively recognised as simulator sickness (SS) (Keshavarz and Hecht 2011). Other side effects include disorientation, disturbances to balance, eyestrain, blurred vision, fatigue, drowsiness, and lack of coordination (Keshavarz and Hecht 2011). Investigation into the relationship between VR and SS using a HMD device has previously shown that these effects could compromise the health and safety of users for several hours afterwards (Stanney et al. 1999a, b).

Previous literature has attempted to predict those who may have a predisposition to SS. Exposure time, age, female gender, pre-existing postural instability, MS susceptibility (Bos et al. 2008; Golding 2006b; Money 1970; Stanney et al. 2003; Stoffregen and Smart 1998; Villard et al. 2008), and the menstrual cycle (Golding 2006b) have been considered to be possible factors; however, evidence regarding the predictability of postural sway and the menstrual cycle is limited. As a result, it is suggested that individuals should be screened for SS susceptibility, and that those susceptible be warned of VR side effects (Stanney et al. 2003). Further, evidence suggests that exposure should be kept to approximately 15 min of initial immersion in VR using a HMD in order to limit SS (Stanney et al. 2003).

Nevertheless, the vast majority of the current body of evidence focussed on the incidence and features of SS following VR immersion with a HMD has been done in exclusive subject populations such as the military or aviation (Braithwaite and Braithwaite 1990; Kennedy et al. 1990, 1997; Stanney and Kennedy 1998). In addition, most of the existing research into SS and predictive features has been done using aircraft and automobile simulators where motion is controlled by the hands and arms and not by head control (Frank et al. 1988; Kennedy et al. 1990; Stanney and Hash 1998). Thus, there is a need to consider SS and potential factors that might affect this when using the neck VR (head motion-controlled VR interaction) for the purpose of, or in the context of neck rehabilitation in the general population. In order to understand the characteristics of SS and potential risk factors for SS when using the neck VR device, investigation should occur amongst a healthy population prior to a patient population.

This study aims to determine the incidence, severity, and predisposing factors to SS when using the VR amongst asymptomatic individuals. It is hypothesised that SS will occur when using the neck VR device, and that older age, female gender, postural instability, previous susceptibility to MS, and increased exposure time will be associated with SS.

2 Method

2.1 Ethics

Ethical clearance was gained from the University of Queensland and adhered to throughout the study.

2.2 Participants

Healthy male and female individuals between the ages 18 and 80 were sought for this study. Individuals were included if they met inclusion criteria: no current neck pain, no visual pathology, no neurological or systemic conditions that affect balance or postural control, and no vestibular conditions.

2.3 Apparatus and virtual stimuli

The VR system used in this study consisted of a HMD with a three-dimensional motion tracker built in, with a sampling rate of 30 Hz (Wrap™ 1200VR by Vuzix, Rochester, New York). During the VR session, the virtual pilot flying the red airplane was controlled by the patient’s head motion and interacts with targets appearing and moving in four directions (flexion, extension, right rotation, left rotation). Neck motion controlled the interaction with the virtual targets, and body motion was restricted as the participants’ trunk was strapped to the back of the seat. The participants could not change the given virtual environment, but the software adapted according to their neck motion performance, i.e. feedback was provided with success hit, and failure adapted the following target to an easier mode. The interactions challenged range, velocity, and accuracy of neck motion. The 3D virtual environment was developed using the 3D graphics software Unity Pro (Unity Technologies, San Francisco). The custom-made software includes three modules: range of motion (ROM), velocity, and accuracy. After completion of each module, the system generates a full kinematic report for each patient. A full description of the VR neck assessment and its outcome measures can be found at Sarig-Bahat et al. (2010).

2.4 Subjective measures

Participants completed three questionnaires regarding motion sickness:

Simulator sickness visual analogue scale (SSVAS) was used to document SS intensity on a scale of 0–10, 0 being no symptoms and 10 being the worst possible symptoms. Participants marked their level of symptoms on a 100-mm line.

The Simulator Sickness Questionnaire (SSQ) (Kennedy et al. 1993) identifies 15 symptoms of SS, and participants are asked to rate them as 0 (none), 1 (slight), 2 (moderate), or 3 (severe). Symptoms on the questionnaire include general discomfort, fatigue, headache, eyestrain, difficulty focusing, increased salivation, sweating, nausea, difficulty concentrating, blurred vision, dizzy eyes open, dizzy eyes closed, vertigo, stomach awareness, and burping. Scores are divided into three subsections of nausea, disorientation, and oculomotor, with higher scores indicating greater levels of SS. Total scores are also calculated with possible scores ranging from 0 to 168. It has been shown that a total score greater than 20 represents significant SS (Webb et al. 2009).

The Short-Form Motion Sickness Susceptibility Questionnaire (MSSQ) (Golding 2006a) predicts individual differences in MS susceptibility caused by various stimuli (cars, buses, trains, aircraft, small boats, ships, swings in playgrounds, roundabouts in playgrounds, funfair rides). Participants rate how often they felt sick or nauseated from these stimuli (0—never, 1—rarely, 2—sometimes, 3—frequently) in their childhood experience before age 12 (MSSQ–Child) and their experience over the last 10 years (MSSQ–Adult). Total scores (MSSQ–Total) range from 0 to 54 (a score of 0–27 is possible for each of the child and adult subsections) with higher scores indicating greater levels of MS susceptibility.

2.5 Postural Stability Measures

Postural stability was measured as the time in seconds (up to 30 s) each participant could maintain tandem stance and single leg stance (right and left) eyes closed (Bohannon et al. 1984; Brotherton et al. 2005). The average time of these three tests was calculated. For participants aged 45 and over, balance testing also included an additional narrow stance eyes closed test as it is known that decline in balance can occur in tandem stance after age 45 (Vereeck et al. 2008). This testing was performed immediately pre- and post-NVR session. The difference in average balance times pre and post was also calculated (pre- and post-balance).

2.6 Procedure

All participants completed the consent form and a general form regarding age, gender, medical history, headaches, dizziness, and menstruation at the time of assessment. Postural stability was then assessed, followed by the NVR assessment. In order to retrieve isolated neck motion during the NVR, the participant’s torso was strapped to the back of the seat. A short introductory practice of each of the three modules (ROM, accuracy, and velocity) was then performed to familiarise the subject with the virtual interaction. The subjects then completed each module randomly ordered by drawing the options from a hat. Each module was timed, and SS–VAS was recorded after each module along with the time of symptoms’ occurrence and subsides. In order to avoid suggestive SS, participants were asked whether they were ‘feeling well’ after each assessment, rather than whether they were ‘experiencing motion sickness’. There were 2-min breaks between modules if SS–VAS was 0. In the event of any reported symptoms (SS–VAS > 0), participants waited until it subsided (up to 5 min) and continued after this time if they felt well enough to do so. The postural stability tests were then reassessed after all three modules were completed. Lastly, participants completed the MSSQ and SSQ, to avoid suggestion of motion/SS (Young et al. 2007). The SS–VAS scores from each module were totalled to give a total SS–VAS score during the assessment. At 24 h post-assessment, all participants were contacted via phone or email to ask about any delayed or ongoing symptoms. Participants rated any delayed or ongoing symptoms using the SS–VAS and SSQ.

2.7 Statistics

Data were explored for outliers and normality, revealing the majority of the measures was not normally distributed. Participants were divided into two groups based on their SSQ scores: nil or negligible SSQ scores (NIL–SS total score below 20) and significant SSQ score (SIG–SS total score greater than 20), as this cut-off has previously been shown to be of sufficient discomfort to warrant consideration of SS (Webb et al. 2009). Nonparametric statistics (Mann–Whitney U) were used to compare the two groups with respect to age; postural stability time pre, post, and pre-post differences; total SS–VAS; time for each module; total time; MSSQ–Child; MSSQ–Adult; and MSSQ–Total scores. Fisher’s exact test was used to compare groups with respect to gender. Further, in order to investigate the SS ratings during the assessment (rather than with the SSQ after all modules were completed), the subjects were grouped according to their SS–VAS total score where a score of 0 (NSS–VAS) was compared to subjects with any total SS–VAS of sickness (SSS–VAS). Mann–Whitney U and Fisher’s exact tests were used to compare between these groups with the same variables used previously.

Nonparametric correlations using Spearman’s Rho were also used to identify relationships between age; gender; postural stability time pre, post, and pre-post differences; total SS–VAS; time for each module; total time; MSSQ–Child; MSSQ–Adult; MSSQ–Total; and SSQ scores. A regression analysis was performed using any variables where statistically significant correlations were found to SS in order to determine the factors that most influenced SS.

3 Results

Thirty-two participants (14 males, 18 females), with a mean age of 36 ± 2.93 and with no current neck pain, enrolled in the study, and all completed the experiment. Table 1 shows the mean values with their ranges and standard deviation of age, balance scores, and the questionnaire results. Only one participant was menstruating during the assessment. Narrow stance was not included in the table, because everyone required to complete this balance assessment reached the maximum time of 30 s. Twenty-one subjects (65 %) had a SSQ score less than 20 and were included in the nil simulator sickness (Nil–SS) group whilst 11 (35 %) had a SSQ score greater than 20 and were included in the significant simulator sickness (Sig–SS) group. During the assessment, nine participants (28 %) reported feeling sick and rated this on the SS–VAS scale and thus were included in the significant SS–VAS group, whilst 23 (72 %) reported no SS–VAS so were included in the Nil SS–VAS group. No participants reported any symptoms of SS at the 24-h follow-up.
Table 1

Ranges, mean, and standard deviation (SD) of variables (n = 32) measured

 

Min

Max

Mean

SD

Age

18.00

64.00

36.16

16.5

Postural stability (sec)

 Tandem pre

6.00

30.00

26.63

7.5

 Tandem post

3.00

30.00

27.56

6.8

 Single leg pre

4.50

30.00

19.64

8.0

 Single leg post

3.50

30.00

21.30

9.1

 Average pre

7.00

30.00

23.13

6.9

 Average post

3.25

30.00

24.43

6.8

 Average pre-post difference

−15.25

12.00

−1.30

5.9

Simulator sickness VAS (/100 mm)

 ROM

0.00

70.00

8.10

17.0

 Velocity

0.00

80.00

6.30

19.0

 Accuracy

0.00

40.00

2.80

9.0

 Total (/300)

0.00

170.00

17.20

40.0

VR time (sec)

 ROM module

65.00

417.00

212.47

95.2

 Velocity module

120.00

416.00

220.34

76.8

 Accuracy module

50.00

70.00

54.66

3.4

 Total

273.00

861.00

487.47

113.5

Simulator Sickness Questionnaire (/168)

0.00

112.20

19.28

23.1

Simulator Sickness Questionnaire—24 h later (/168)

0.00

0.00

0.00

0.0

Motion Sickness Susceptibility Questionnaire

 Child (/27)

0.00

24.00

7.11

7.1

 Adult (/27)

0.00

18.00

6.05

5.8

 Total (/54)

0.00

40.88

13.16

12.1

VAS visual analogue scale

Table 2 shows the comparison between those with negligible SS Questionnaire scores <20 (Nil–SS group) and those with significant SSQ scores >20 (Sig–SS group). There was a significant difference in the ROM module SS–VAS (p = 0.006) and total SS–VAS (p = 0.006) between the nil and significant SS groups. Mann–Whitney U value was 110.00. No other significant differences were found.
Table 2

Comparison between negligible Simulator Sickness Questionnaire scores (NSSQ (under 20, n = 21) and significant Simulator Sickness Questionnaire scores (SSSQ above 20, n = 11)

Variable

NSSQ

SSSQ

Mean

SD

Mean

SD

Age

38.90

17.6

30.91

13.4

Postural stability (sec)

 Tandem pre

25.62

8.8

28.55

3.3

 Single leg pre

18.93

8.4

21.00

7.4

 Tandem post

28.10

5.8

26.55

8.5

 Single leg post

21.14

9.0

21.59

9.8

 Average pre

22.27

7.7

24.77

5.1

 Average post

24.62

5.9

24.07

8.6

 Pre-post difference

−2.34

5.4

0.70

6.5

Simulator sickness VAS (/100 mm)

 In ROM

1.40*

7.0

20.90*

24.0

 In velocity

0.50

2.0

17.30

30.0

 In accuracy

0.00

0.0

8.20

15.0

 Total (/300)

1.90*

7.0

46.40*

58.0

VR time (sec)

 ROM time

205.1

101.4

226.55

84.9

 Velocity time

201.62

59.2

256.09

95.6

 Accuracy time

54.14

2.2

55.64

5.1

 Total time

460.86

97.5

538.27

21.3

Simulator Sickness Questionnaire (/168)

7.48*

6.9

42.08*

27.3

Motion Sickness Susceptibility Questionnaire

 Child (/27)

5.69

6.6

1.40

9.8

 Adult (/27)

5.88

6.3

1.40

6.4

 Total (/54)

11.57

12.4

2.70

16.2

VAS visual analogue scale

* Significant difference between groups (p < 0.05)

Of the 11 participants who scored above 20 on the SSQ, four subjects did not record feeling sick on the VAS during the neck VR assessment. When comparing those who scored feeling sick on the SS–VAS (SSS–VAS) to those who did not (NSS–VAS), there was a significant difference in ROM time (p = 0.024), total time (p = 0.009), MSSQ–Child (p = 0.043), and MSSQ–Total (p = 0.034) (Table 3).
Table 3

Comparison between NSS–VAS (nil simulator sickness visual analogue scale score of 0) (n = 23) and SSS–VAS (simulator sickness visual analogue scale score of greater than 0) (n = 9)

Variable

NSS–VAS

SSS–VAS

Mean

SD

Mean

SD

Age

36.61

17.3

35.00

15.1

Postural stability (sec)

 Tandem pre

25.61

8.5

29.22

2.3

 Single leg pre

19.57

8.6

19.83

6.7

 Tandem post

28.91

4.8

24.11

9.8

 Single leg post

22.37

8.8

18.56

9.9

 Average pre

22.59

7.8

24.53

4.2

 Average post

25.64

5.8

21.33

8.3

 Pre-post difference

−3.05

5.1

3.19

5.6

Simulator sickness VAS (/100 mm)

 ROM SS–VAS

0.00

0.0

28.90

23.0

 Velocity SS–VAS

0.00

0.0

22.20

31.0

 Accuracy SS–VAS

0.00

0.0

10.00

16.0

 Total SS–VAS (/300)

0.00

0.0

61.10

56.0

VR time (sec)

 ROM time

192.35*

93.1

263.89*

84.6

 Velocity time

207.17

57.2

254.00

109.9

 Accuracy time

54.74

4.0

54.44

1.7

 Total time

454.26*

92.7

572.33*

122.5

Simulator Sickness Questionnaire (/168)

6.84*

6.7

30.94*

12.0

Motion Sickness Susceptibility Questionnaire

 Child (/27)

5.35*

5.6

11.61*

8.6

 Adult (/27)

5.02

5.7

8.68

5.5

 Total (/54)

10.37*

10.7

20.29*

13.1

VAS visual analogue scale

* Significant difference between groups (p < 0.05)

Table 4

Correlations between variables

 

SSQ

SS–VAS

Age

Pre-post balance

ROM time

Velocity time

Accuracy time

Total time

MSSQ–Child

MSSQ–Adult

MSSQ–Total

SSQ

1.00

0.618**

−0.15

0.29

0.14

0.17

−0.04

0.20

0.25

0.17

0.23

SS–VASs

0.618**

1.00

0.04

0.374*

0.433*

0.17

0.01

0.478**

0.32

0.31

0.34

Age

−0.15

0.04

1.00

−0.06

−0.06

0.380*

0.21

0.17

−0.09

−0.10

−0.08

Pre-post balance

0.29

0.374*

−0.06

1.00

0.20

0.07

−0.10

0.19

0.17

0.25

0.18

ROM time

0.14

0.433*

−0.06

0.20

1.00

−0.29

−0.20

0.721**

0.28

0.23

0.27

Velocity time

0.17

0.17

0.380*

0.07

−0.29

1.00

0.18

0.34

−0.08

−0.04

−0.03

Accuracy time

−0.04

0.01

0.21

−0.10

−0.20

0.18

1.00

−0.05

−0.20

−0.17

−0.15

Total time

0.20

0.478**

0.17

0.19

0.721**

0.34

−0.05

1.00

0.27

0.25

0.30

MSSQ–Child

0.25

0.32

−0.09

0.17

0.28

−0.08

−0.20

0.27

1.00

0.834**

0.958**

MSSQ–Adult

0.17

0.31

−0.10

0.25

0.23

−0.04

−0.17

0.25

0.834**

1.00

0.941**

MSSQ–Total

0.23

0.34

−0.08

0.18

0.27

−0.03

−0.15

0.30

0.958**

0.941**

1.00

Nausea

0.734**

0.530**

−0.01

0.366*

0.02

0.03

−0.15

−0.04

0.13

0.11

0.12

Oculomotor

0.871**

0.513**

−0.29

0.24

0.18

0.15

0.03

0.25

0.11

0.03

0.10

Disorientation

0.781**

0.466**

−0.02

0.06

0.16

0.19

−0.16

0.25

0.418*

0.355*

0.401*

SSSVAS Significant Simulator Sickness Visual Analogue Scale, SSQ Simulator Sickness Questionnaire, MSSQ Motion Sickness Susceptibility Questionnaire, ROM range of motion

* Significance at p < 0.05; ** p < 0.01

Significant positive mild-to-moderate correlations were found between SS–VAS scores and SSQ total (r = 0.618), total time (r = 0.478), ROM time (r = 0.433), and pre-post postural stability difference (r = 0.374). A trend was found between SS–VAS and MSSQ–Child (r = 0.320) (Table 4). No other significant correlations were found for SSQ or SS–VAS and the other variables.
Table 5

Sensitivity and Specificity of Motion Sickness Susceptibility Questionnaire–Child cut-off of 10 score for each group—nil simulator sickness visual analogue scale score of 0, and significant simulator sickness visual analogue scale score >0 (n = 9)

 

Significant simulator sickness VAS

Nil simulator sickness VAS

Total

Motion Sickness Susceptibility Questionnaire–Child <10

3

19

22

Motion Sickness Susceptibility Questionnaire–Child >10

6

4

10

Total

9

23

32

VAS visual analogue scale

The regression analysis used the three significant measures that correlated to SS–VAS: total VR time, pre-post balance difference, and MS susceptibility in childhood. Results of the regression showed that the combination of total VR time and MS in childhood accounted for 41 % of the variance of SS–VAS.

Table 5 shows a post hoc analysis of sensitivity and specificity of MSSQ–Child for SSS–VAS. A cut-off MSSQ–Child score of 10 correctly identified 86 % in the NSS–VAS group, and MSSQ score greater than 10 correctly identified 60 % in the SSS–VAS.

4 Discussion

The results of this study demonstrated an approximate 30 % incidence with low severity of SS whilst using the neck motion-controlled VR device in asymptomatic individuals. It also showed how the individual factor of MS susceptibility and the VR exposure time may impact on the incidence and severity of SS as rated by the sickness VAS rather than the SSQ. Such knowledge could assist in identifying those who may develop sickness during the assessment and potentially reduce SS. This is vital considering the emerging development of VR being used for both assessment and management of neck pain, as well as for other non-physiotherapy-related domains.

4.1 Incidence and severity of simulator sickness

When using SS rated by VAS > 0, the incidence of sickness during the VR immersion was approximately 28 % (n = 9), and the severity of such sickness was on average only 17.2 out of 100 mm. A slightly higher incidence of 34 % (n = 11) was found when using SSQ > 20, and the mean overall SSQ score was 19.3. This reported incidence of symptoms is comparable with a wide range of results reported in earlier studies of approximately 10–60 % (Braithwaite and Braithwaite 1990; Merhi et al. 2007; Regan and Price 1994), confirming that NVR, like other VR application, provokes side effects of SS. However, the fact that such symptoms were not severe enough to warrant high sickness VAS scores or SSQ scores was an interesting finding. When reviewing the severity of SS reported in previous studies encompassing VR immersion, it is clear that severe adverse symptoms are sufficiently common (Braithwaite and Braithwaite 1990; Lampton et al. 2000; Regan and Price 1994), and at times have led to participants having to withdraw from the experiments altogether. This was the case in a study by Stanney et al. (1999a, b), where 19 % withdrew with an average SSQ score of 49.9, whilst those who completed the testing reported an average SSQ score of 30.35. Similarly, in a study by Stanney et al. (2002), 12 % of participants withdrew with an average SSQ score of 33.53, and those who were finishers scored an average of 23.22 on the SSQ. In the current study, none of the participants had to withdraw due to severe symptoms, and there were no reports of any prolonged or latent symptoms 24 h post-testing. Compared to other literature, the lower severity of SS reported in the current study may be due to the differing devices and equipment used. These findings are clinically important, suggesting that whilst using a neck VR device can still provoke symptoms of SS, the severity of these symptoms will likely be low compared to other HMD devices, although higher than other devices as such as desktop and reality theatre viewing. (Sharples et al. 2008) Furthermore, Sharples et al. found high inter- and intra-participant variabilities, supporting the current findings of individual susceptibility to VR side effects. (Sharples et al. 2008) It is therefore yet established how to identify individuals at high risk of SS, prior to exposing them to the virtual experience. Nevertheless, results of this study would suggest that those with previous high levels of MS as a child, may, but not necessarily, suffer from SS and should be closely monitored.

Further, the type of HMD hardware used in the current study may have contributed to the relatively high frequency (albeit low severity) of SS. The Vuzix HMD had a tracking sample rate of 30 Hz, but its consistency varied, and there were some cases of drift resulting in a skewed visual output, probably due to errors of the gyroscopic sensors. Latency and drift of the tracking device may create a discrepancy between the timing of the visual display and the actual motion experienced, possibly stimulating SS (Draper et al. 2001; Meehan et al. 2003). Thus, it is possible these hardware features may have had an impact on the presence of SS. Future studies should consider using newer and more advanced HMDs with higher sampling rates, and minimal latency and drift to potentially minimise SS further.

4.2 Simulator sickness and sickness VAS

Interestingly, the findings of the study showed that some individuals who reported symptoms on the SSQ did not always report a SS–VAS during immersion. This may be explained by ‘demand characteristics’ (Young et al. 2007), whereby the participants were exposed to a set of symptoms which may have prompted them to become more attentive and sensitive to them, and thus recorded the perceived ‘appropriate response’ rather than an actual response. Whilst sickness VAS did correlate moderately with SSQ (r = 0.618; p < 0.01), unlike the sickness VAS, the results of the study failed to demonstrate any associations between SSQ and potential predictors of SS. To our knowledge, no previous studies have used sickness VAS to monitor symptoms throughout a VR experience, and thus, no correlations between sickness VAS and motion or simulator sickness have previously been observed. In this study, to gain a sickness VAS, we deliberately asked whether participants were feeling ‘well’ rather than ‘feeling sick’ and also administered the SSQ after the neck VR assessment to avoid anticipatory SS (Young et al. 2007). Nevertheless, the SSQ lists a number of symptoms, which might still elicit a suggestive response. Thus, sickness VAS might be an alternate measure to SSQ, avoiding this phenomenon of ‘demand characteristics’, as well as enable easy monitoring during, rather than following, the VR assessment.

Further, SSQ has been shown to decrease with repeated exposure (Kennedy et al. 2000). The moderate relationship between SSQ and sickness VAS would suggest that sickness VAS should also reduce with exposure. Such a reduction in adverse symptoms with repeated exposure is important information for participants to make informed decisions regarding their future participation in VR environments. Future research will be required to confirm this.

4.3 Factors associated with simulator sickness

4.3.1 Duration of VR immersion

No relationship was found between VR exposure time and SSQ; however, a positive significant relationship (r = 0.478, p ≤ 0.01) was identified between the total time taken to complete the modules and sickness VAS. This is in line with other studies that have found that a longer VR exposure time is associated with greater symptoms of sickness; however, sickness in these studies was measured using the SSQ (Jerome et al. 2005; Moss and Muth 2011). In particular, the present study found that the longest module, ROM, produced the most significant responses in terms of sickness VAS. Such findings may be attributed to the device set-up. For the ROM module, the parameters were set so the individual initially had of the first targets positioned at 30° from the centre to each direction. With each successful trial, the ROM increased by a further 5° until the individual missed three targets and the module ended. Therefore, it can be inferred that individuals capable of greater neck rotation ROM would accomplish more increments and hence longer exposure time within the module. Thus, longer VR sxposure time will be predisposing them to a greater susceptibility to SS. Screening an individual for their available neck ROM, particularly rotation, and setting the device parameters accordingly (beginning at a larger required ROM) may be highly beneficial for reducing adverse symptoms associated with neck VR. Additionally, past studies have recommended limiting VR exposure time to 15 min in order to reduce the incidence of SS (Stanney et al. 2003). However, in the current study, no participants exceeded this threshold and still approximately one-third experienced SS. This further highlights the importance of limiting VR exposure time in order to reduce the incidence of SS. Future research that aims to determine the optimal exposure time for the neck VR and any benefit of modifying the ROM module as suggested above to reduce exposure time should be explored further. This is potentially important for the progression of this device for use in neck rehabilitation as recently this system was used in a trial in 32 patients with chronic neck pain (Sarig Bahat et al. 2015). Four out of 40 eligible patients withdrew from the study due to side effects (Sarig Bahat et al. 2015). The information gained from the current study especially regarding VR exposure time may help to limit the degree of SS in the future.

4.3.2 Motion Sickness Susceptibility Questionnaire

Findings of the current study revealed a trend towards a positive correlation between sickness VAS, rather than SSQ, and results of the MSSQ–Child (r = 0.32) and MSSQ–Total (r = 0.31). Further, when the groups were divided into those with no sickness (SS–VAS = 0) and those with some sickness (SS–VAS > 0), the sickness group had significantly higher MSSQ–Child and MSSQ–Total scores (on average double the no-sickness group). This suggests that those with higher MSSQ scores may report some sickness when using the device. Further, 86 % of individuals with a childhood MSSQ score <10 did not report sickness whilst using the device. Conversely, 60 % of people with a MSSQ score >10 reported some sickness. This finding would support the use of the MSSQ–Child prior to neck VR testing to at least help identify those who are unlikely to experience sickness during the assessment.

4.3.3 Postural instability

Evidence has suggested that postural control may play a crucial role in the development of nausea and disorientation, and specifically, that postural instability is strongly associated with MS and SSQ scores (Cobb & Nichols 1999). In the current study, no associations were found between pre- or post-postural stability and SSQ or sickness VAS; however, there was a positive association between the difference in the pre- and post-balance scores and sickness VAS (r = 0.374; p ≤ 0.05). This is interesting and suggests that the subjective complaints of sickness are relating to objective changes in postural stability, and thus, sickness may be related to altered sensorimotor control (Kennedy and Stanney 1996). However, this factor can only be assessed after exposure to VR and therefore would not be useful in helping to predict those who may get MS prior to exposure to the neck VR.

4.3.4 Factors not associated with simulator sickness

Contrary to what was predicted, age and gender did not correlate with scores of MSSQ, SSQ, or SS–VAS. This could be due to the limited sample size, both in terms of absolute numbers and age ranges.

4.4 Limitations

There are some limitations to the study. As mentioned, the present study failed to identify relationships between pre-postural stability and SSQ or sickness VAS; however, this may be a reflection of the method of assessment. Previous studies that revealed relationships between postural stability and SSQ scores following VR immersion tended to use more sophisticated measures (Cobb 1998; Cobb and Nichols 1999; Stoffregen and Smart 1998). Nevertheless, differences in pre-post postural stability measures were found, indicating that our method was sensitive to those with SS, but did not have predictive power. Using highly sophisticated methods would not be clinically practical when trying to predict those who might develop SS.

Sample size and self-exclusion were other limitations. Whilst this study established some clinically valuable predictors for sickness VAS, the sample size was small (n = 32), and consequently, the conclusions drawn from this study should be done so with caution and future studies may benefit from a larger sample size. Contributing to the limited sample size, individuals who self-perceived susceptibility to SS may have avoided volunteering for the study. This may have lead to an inappropriate representation of the general population; however, it would be a realistic representation of those who would seek or consent to neck VR as a treatment option.

5 Conclusion

This study aimed to determine the incidence, severity, and predisposing factors of SS in asymptomatic individuals using VR for neck motion assessment. This study found that using the Vuzix HMD provoked SS in about one-third of the participants, but these seem to be of a lesser severity than previously reported for other devices. Technical features of the Vuzix may have increased the discrepancy between target display and experienced physiological movement (Draper et al. 2001; Meehan et al. 2003), and thus, future research should also consider using HMDs with faster and more accurate tracking units.

Exposure time appears to be related to higher SS–VAS and should be considered in future to limit SS when using the device. Subjects should also be screened prior to VR participation using the MSSQ–Child and symptoms monitored throughout using SS–VAS. The reliability and validity of using VAS to measure SS symptoms in VR contexts should be considered as a direction for further investigation. Investigation into symptomatic individuals and treatment contexts will also be important avenues for future research.

Copyright information

© Springer-Verlag London 2015

Authors and Affiliations

  1. 1.The Department of PhysiotherapyUniversity of QueenslandBrisbaneAustralia
  2. 2.The Department of Physical TherapyUniversity of HaifaHaifaIsrael

Personalised recommendations