The Effect of Sleep Deprivation on the Vestibulo‐Ocular Reflex and Vestibular Adaptation in Healthcare Night‐Shift Workers

ABSTRACT

Objective

This study investigates the impact of sleep deprivation on the vestibulo‐ocular reflex (VOR) in healthcare night‐shift workers, focusing on basic VOR function and its adaptation during head movements with varying target distances.

Methods

A prospective study was conducted with 14 night‐shift healthcare workers, with a final analysis of 11 participants. Testing occurred under two conditions: After normal sleep and after a night shift. VOR gain was measured using the video head impulse test (vHIT) system for both active and passive head movements and for near and far targets. Participants also completed the computerized Dynamic Visual Acuity (c‐DVA) test to assess gaze stability and a self‐reported vigilance questionnaire.

Results

Participants experienced significantly shorter sleep durations and reported lower vigilance levels on the post‐night shift. VOR gain was notably reduced (7.5%) only during passive head movements toward a far target after a night shift. Other vestibular functions, such as distance modulation and other measurements including active head impulses, near‐target tests, and c‐DVA scores, showed no significant differences between normal sleep and sleep‐deprived states.

Conclusion

Sleep deprivation in healthcare night‐shift workers demonstrated a complex effect on VOR, with significant impairment in only one specific condition (passive‐far). Most other VOR functions, including adaptation to target distance, were maintained. This resilience may indicate an adaptive mechanism within the vestibular system under chronic sleep‐deprived conditions, warranting further research into the underlying protective processes.

Level of Evidence

3

This study examined the effects of sleep deprivation on the vestibulo‐ocular reflex (VOR) in healthcare night‐shift workers. Results showed a significant reduction in VOR gain (7.5%) during passive head movements toward far targets after a night shift, whereas other VOR functions and gaze stability remained unaffected. These findings suggest that the vestibular system may exhibit resilience under sleep‐deprived conditions, potentially due to adaptive protective mechanisms.

1. Introduction

The Vestibulo‐Ocular Reflex (VOR) holds great importance as one of the key reflexes in the human body, ensuring clear and steady vision whereas the head is in motion [1]. The VOR has an essential role in adaptation to different distances with physiological mechanisms to maintain stable foveal images during head motion. For example, adjusting the VOR gain (ratio of eye and head velocity amplitudes) as the viewing distance changes [2, 3]. The VOR's gain modulation based on target distance during head movements emphasizes higher gains for nearer targets, as was demonstrated by Judge et al. [4]. Chang and Schubert further expanded on the vergence‐mediated VOR gain and found it to be impaired in patients with unilateral vestibular hypofunction [5]. Any disruption to the VOR's proper functioning can result in an unstable gaze and blurred vision, posing a risk during activities demanding precise accuracy, such as operating upon a patient or driving [6].

The ubiquity of sleep deprivation has been on the rise in recent years. The prevalence of short sleep duration (less than 7 h per night) has increased by 25% from 2010 to 2018 in working American adults [7]. Healthcare providers showed a higher prevalence of sleep deprivation (36.3%–45.4% vs. 35.6% in the general adult working population) and often experience chronic sleep deprivation [7, 8]. Sleep deprivation causes deterioration in human cognitive functions such as vigilant attention with slower reaction times, processing speed, memory, and constructive thinking [9, 10, 11]. It is linked to car accidents, work‐related accidents, and lower work performance [12, 13, 14]. A comprehensive review of meta‐analyses and systematic reviews found a link between night‐shift work and car accidents, type 2 diabetes mellitus, and cancer [15].

Various studies demonstrate the adverse effect of sleep deprivation on the vestibular contribution to postural control [16, 17, 18, 19], which is maintained by stabilizing gaze and adjusting muscle tone and body position in response to head movements through the vestibulo‐spinal reflex [20]. However, only a few studies investigate the direct influence of sleep deprivation on the VOR. These studies were mainly executed using the rotatory chair test system [10, 21, 22]. They found a mixed effect of sleep deprivation on the VOR gain. Some demonstrated a significant increase in VOR gain [10, 21] and some demonstrated no effect [22].

Advancements in measurement technology, like the video head impulse test (vHIT), have opened new possibilities for assessing the functioning of the VOR under more realistic and natural conditions [23]. Moreover, the vHIT system allows us to examine how different situations affect the VOR, such as active and passive head movements and the reflex adaptation to an object at different distances [4]. These, in turn, enhance our ability to gain deeper insights into the mechanisms of the VOR.

In this study, we aimed to elucidate how sleep deprivation among healthcare night‐shift workers influences two levels of VOR function during high‐frequency daily functional head rotation. The first level is the effect of sleep deprivation on basic VOR function. The second level is the effect of sleep deprivation on vestibular adaptation abilities generated by vergence modulation. We hypothesized that sleep deprivation would have a negative effect on the performance of the VOR and VOR adaptation.

2. Materials and Methods

We conducted a prospective interventional measurement study. The participants are healthy healthcare night‐shift worker individuals, that is, medical doctors from different residency programs—pediatrics, otolaryngology, and general surgery. All participants work night shifts four to six times a month. Participants were recruited through an announcement by the author (N.L.) among colleagues, inviting those interested to take part in the study. Those who responded were subsequently approached by the author (N.L.), who provided further details about the study. Participation was entirely voluntary. This study received approval from the Helsinki Committee of the hospital to ensure compliance with ethical guidelines and standards (approval number 8879‐21‐SMC).

Participants were tested in two conditions: (1) Normal night sleep (control night) and (2) after a night shift (intervention night). The two examinations were held at least 2 weeks apart. Examinations took place at the vestibular laboratory of the medical center.

Control night conditions: Participants were admitted prior to their regular workday at 7:00–8:00 a.m. and were tested by the vHIT system (EyeSeeCam, Interacoustic, A/S Denmark). Test protocol included four test modes of physiologic function of the VOR using a combination of two parameters: head movement method and target distance. Head movement could be passive (tester rotates the participants' head at high velocity), or active (participant self‐induced head rotation). The target distance was either far (1.5 m) or near (15 cm). The combination was conducted in a random order. Randomization was done using the RANDBETWEEN formula in Microsoft Excel, producing random numbers. The vHIT test was followed by a behavioral computerized dynamic visual acuity (cDVA) test. Finally, each participant completed a questionnaire assessing their previous night's experience and their level of vigilance using a five‐point scale: 1—very tired, 2—tired, 3—neutral, 4—alert, 5—very alert. None of the participants were on a “night float” rotation, as this practice does not exist in our country. Additionally, none of the participants had worked a night shift in the evening prior to the control night, ensuring they had at least two consecutive nights of normal sleep.

Intervention night conditions: Participants were again admitted after a night shift at 7:00–8:00 a.m. The participants were tested with the same routine described above.

The proposed study was conducted in a “real world” setting. Participants were not required to avoid sleep during the night shift, and they were not monitored on either night. These conditions are more representative of the participants' real state in daily life at the expense of results purity.

The main measurement tool was the vHIT system, a device that is used to assess the physiological function of the VOR [23]. The vHIT software (EyeSeeCam software/OtoAccess) calculates the VOR gain as the ratio between the eye velocity and the head velocity per head impulse and the average of all impulses. In our study, the vHIT was used in the horizontal plane. The vHIT system technically measures the movement of one eye (the right eye in our study) and the head movements in both directions (right and left when examining the lateral semicircular canal). The VOR gain is automatically calculated separately for right and left head thrust. A comparison between VOR gains for rightward and leftward head rotations showed no significant differences in all situations tested. Thus, the gains used in this study were calculated as an average of the right and left head thrust gains. In addition to VOR gain, we also measured the adaptation of the VOR to different target distances by calculating the “adaptation ratio”‐ dividing the Near VOR gain/Far VOR gain*100. The accepted head velocity is at least 150°/s, and a minimum of 12 head thrusts were performed in each direction.

The secondary measurement tool was the c‐DVA test (VORTEQTM, Micromedical Technologies, IL USA), which provides an assessment of the implication of head movements on gaze stability [24]. The test relies on the response of the participant to head movement, making it a behavioral function test. The participant is first instructed to identify optotypes appearing on a screen to set a baseline static visual acuity score, measured in units of Logarithm of the Minimum Angle of Resolution (logMAR). Then the participant needs to identify these optotypes while moving their head. An optotype briefly flashes only when participants move their head at a velocity of 100°–150°/s and follow a 2 Hz frequency of a metronome. Head movement is done in four directions—up, down, right, and left. The difference between the two conditions (static vs. dynamic) is compared, giving the change in gaze stabilization affected by the head rotation (using the VOR). A difference of 0.2 logMAR score or higher is considered to be pathological.

Our sample size estimation was based on data from a previous study by Chang and Schubert [5], which examined vestibular adaptation in individuals with unilateral vestibular hypofunction and healthy controls. Specifically, we utilized their reported convergence VOR data in healthy adults. According to their findings, the mean VOR gain for far targets was 1.07 ± 0.09, whereas for near targets, it was 1.24 ± 0.15. Using PS Power and Sample Size Calculations, version 3.1.2 we calculated that a minimum of six participants would be necessary to achieve a statistical power of 0.8, with a significance level of 0.05.

Descriptive statistics are presented in absolute numbers, percentages, mean, and standard deviation. We used a paired student t‐test to assess differences in performance with the same participants on different conditions (normal sleep and sleep deprivation). All tests were two‐tailed except for the comparison of VOR gain between normal sleep and sleep deprivation due to the hypothesis that VOR gain would be lower in sleep‐deprived participants. We used a mixed‐effects model with sleep condition as a within‐subject fixed effect, sleeping duration and vigilance score as covariates, and a random intercept for participants to account for individual variability. The model allowed us to evaluate the effects of sleep condition and covariates while controlling for repeated measures. The Fisher exact test was used to compare categorical variables. Vergence modulation was calculated as the ratio between VOR gain toward the near target and far target. The difference in active versus passive impulses was calculated as the ratio between VOR gain in these two conditions. Significant results were defined as a p value < 0.05. Statistical analysis was performed using IBM SPSS software version 26.

3. Results

The study population consisted of 14 medical residents (Table 1). Three participants dropped out of the study: One participant due to a history of neurosurgical intervention, one due to severe myopia with high‐index lenses, and one did not complete the test. Thus, 11 participants were eventually included in the final analysis. Seven participants (64%) were male, the average age was 34.1 (SD 3.8) years, and none had any prior neurologic, vestibular, or sleep disorders.

TABLE 1.

Questionnaires results.

	Normal sleep	Sleep deprivation	p
Sleep length, hours, average ± SD	6.2 ± 0.6	1.5 ± 1.4	< 0.001
Continuous sleep, ratio	75%	8%	< 0.05
Coffee 24 h, cups, average ± SD	2.4 ± 2	3.3 ± 2.5	0.08
Alcohol 24 h, servings, average ± SD	0.1 ± 0.3	0 ± 0	0.3
Vigilance scale, average ± SD	4 ± 0.7	2.3 ± 0.6	< 0.001

During the night shift, sleep duration was 24% of normal night (p < 0.001). Sleep during regular nights was largely unbroken, with 75% of participants reporting continuous sleep compared with only 8% during the night shift (p < 0.05). Participants were asked before the test how vigilant they felt, with one being extremely tired and five being fully alert. Significant differences were found, as participants had a score 75% lower after night shifts when compared to normal night sleep (p < 0.001) (Table 1).

VOR gain was tested in four different test modes comprised of far (1.5 m) versus near (15 cm) targets and passive versus active head impulse. First, we compared the difference in VOR gain in all situations between normal sleep and sleep deprivation (Table 2). Passive impulse—far target VOR gain was 7.5% significantly lower after the night shift (Figure 1a and Table 2). VOR gains comparison in other modes, that is, passive head thrust toward near target and active head thrust toward far and near targets, did not demonstrate significant differences due to sleep deprivation (Table 2). Figure 1b shows an example of a single participant vHIT output tested by passive head movement toward a distant target (1.5) and illustrates the decline in VOR gain while being sleep deprived.

TABLE 2.

Video head impulse test results—basic VOR gain, VOR adaptation, and VOR gain ratio of active and passive impulses.

Test type	Setting		Normal sleep	Sleep deprivation	p
Basic VOR gain	Passive impulse	Far target	0.95 ± 0.08	0.89 ± 0.1	0.03
		Near target	1.14 ± 0.09	1.13 ± 0.11	0.38
	Active impulse	Far target	1.02 ± 0.05	1 ± 0.1	0.24
		Near target	1.22 ± 0.12	1.07 ± 0.12	0.19
VOR adaptation to different target distances	Passive impulse	Near/far target ratio	1.21 ± 0.1	1.28 ± 0.1	0.14
	Active impulse	Near/far target ratio	1.2 ± 0.11	1.23 ± 0.11	0.91
VOR gain ratio of active and passive impulses	Far target	Active/passive impulse ratio	1.08 ± 0.09	1.14 ± 0.19	0.32
	Close target	Active/passive impulse ratio	1.07 ± 0.12	1.13 ± 0.08	0.67

FIGURE 1.

(A) Change in vestibulo‐ocular reflex gain with passive head movement toward far target between normal sleep and after night shift. Each line represents a different participant. (B) Participant number 3 vHIT output was tested by passive head movement toward a distant target as an example of the decline in VOR gain after night shift. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

Adaptation measurements (near VOR Gain/far VOR Gain) revealed higher VOR gain to near targets compared with far targets, with an average raise of 20%–30%. However, no differences were found in the adaptation ratio in both active and passive head impulses between normal sleep and sleep deprivation (Table 2).

Active head impulses were characterized by higher VOR gain compared to passive head impulses after normal sleep (7%–8% higher) and even more so after night shift (13%–14% higher), but the differences between them were not significant (Table 2).

Results of the mixed‐effect model for the differences in the VOR gain and VOR adaptation showed that none of the fixed effects reached statistical significance. Specifically, there was no significant difference in VOR gain nor adaptation between the normal sleep and night shift conditions (p > 0.05). Additionally, neither sleeping duration (p > 0.05) nor vigilance score (p > 0.05) significantly predicted VOR gain or adaptation.

There were no significant differences in the c‐DVA test between regular night and night shift (Table 3). The c‐DVA test done after a regular night yielded an expected increase of 1.4, 1.3, 1.6, and 1.6 logMAR scores between static and dynamic head movement to the right, left, upward, and downward, respectively. The same test done after the night shift also showed a normal increase of 2, 1.8, 1.6, and 1.6 logMAR scores, respectively.

TABLE 3.

Computerized dynamic visual acuity (cDVA) results.

Movement direction	Normal sleep	Sleep deprivation	p
Right movement	1.4 ± 1	2 ± 1.4	0.31
Left movement	1.3 ± 1.1	1.8 ± 1.3	0.34
Up movement	1.6 ± 1.2	1.6 ± 1.1	0.79
Down movement	1.6 ± 1.2	1.6 ± 1.1	0.87

4. Discussion

This study aims to evaluate the effect of sleep deprivation on the VOR and VOR adaptation in healthcare night shift workers using the vHIT system. Our results were mixed—on the one hand, basic and clinical evaluation of VOR gain had some unfavorable effects with significantly lower VOR gain while sleep‐deprived at the passive‐far target condition. On the other hand, other tests of active impulses, dynamic visual acuity, and surprisingly even vestibular adaptation did not show a significant deterioration after the night shift. Importantly, even though the study was not done under controlled conditions but in “real world” conditions, the questionnaire results showed significant differences in sleep duration and vigilance levels between regular night and night shift. These equivocal results might demonstrate a possible complex effect of sleep deprivation on VOR in healthcare night‐shift workers.

Previous studies on the effect of sleep deprivation on VOR gain primarily utilizing the rotatory chair system have also yielded mixed results. Collins et al. found no difference in VOR gain between sleep‐deprived participants and controls [22]. In contrast, Quarck et al. observed no change in low‐frequency VOR gain but noted an increase in fast acceleration VOR after 26–29 h of sleep deprivation [10]. Wolfe & Brown reported a decrease in VOR under slower rotatory conditions, attributing it to attention deficits from sleep deprivation [21]. Our study differs from these studies mainly by the measurement tool. The rotatory chair tests the response of the vestibular system to mid‐high frequencies, whereas the vHIT system utilized in our study stimulates response to higher frequencies exclusively. Thus, VOR gain measured with vHIT is not affected by other somato‐sensory systems [25]. Some studies show disagreement in measurements between rotatory chair and the vHIT system. Judge et al. demonstrate that vHIT tends to identify patients with mild bilateral vestibular loss as unilateral loss, unlike the rotatory chair [4]. Lany and Peterson found the same trend in 25% of the participants in their study of young adults and children with cochlear implants, among other trends of disagreements, suggesting inherent differences in the measurement techniques [26]. Our results showed a negative effect of sleep deprivation on VOR gain in only one mode (passive thrust, far target), raising the question of whether the lack of widespread significant differences was due to the technical measurement differences or due to a true physiological mechanism.

The long‐term effects of sleep deprivation in healthcare night‐shift workers might have caused the relatively mild effect of sleep deprivation on VOR gain and no effect on vergence modulation adaptation seen in our study. Medical residents regularly need to perform during and after 24‐h shifts and often suffer from the cumulative effect of sleep deprivation. A few studies elucidated and reviewed the long‐term effects of sleep deprivation on healthcare night‐shift workers on different cognitive fields. Hudson et al. reviewed how it affects attention and described that chronic sleep deprivation may affect baseline state and vulnerability to vigilant attention deficit [27]. Another recent literature review showed long‐term decline in cognitive performance and memory performance, mainly in participants who worked for many years (10 years) [28]. However, no studies examined the long‐term effect of sleep deprivation in healthcare night‐shift workers on the vestibular system in general and specifically on the VOR. Our results suggest that the VOR gain and the ability of the vestibular system to adapt to different distances might be relatively protected in individuals who suffer from chronic sleep deprivation. Thus, it is possible that the vestibular system is less affected by cumulative sleep deprivation in comparison to other cognitive domains. These differences may point to the fundamental function of the vestibular system and its role in survival in comparison to other cognitive functions that were more affected by sleep deprivation.

In their review of the neurocognitive effect of sleep deprivation, Goel et al. discuss studies that found faster accumulating cognitive deficits (mainly psychomotor and working memory) with acute total sleep deprivation in comparison to partial sleep deprivation over days [29]. The authors suggest that neuroplastic adaptation to gradual chronic sleep deprivation might explain these findings. The mild effect of chronic sleep deprivation on VOR may also result from adaptive neuroplastic mechanisms in the central vestibular pathways or the intrinsic resilience of the vestibular system under extreme conditions. Further research is required to better understand this phenomenon and to investigate the protective mechanisms that may help maintain VOR function during prolonged sleep deprivation.

Our findings align with previous research on the vHIT system regarding the differences in VOR gain based on target distance, head thrust direction, and the distinction between active and passive thrusts. VOR gain was consistently higher when participants faced a near target (15 cm) compared to a distant one (1.5 m), supporting the concept of vestibular adaptation to distance modulation, as observed in earlier studies [4]. Concerning the direction of the head thrust, our results corroborate those of Park et al. who found no significant effect of thrust direction (toward vs. away from the recorded eye) on VOR gain [30]. This contrasts with the findings of Faranesh et al. who reported statistically higher gains when the thrust was toward the recorded eye in healthy individuals [31]. Additionally, although the increase in VOR gain observed with active head impulses versus passive impulses has not been extensively studied using the vHIT system, Black et al. suggested that active head impulses are less sensitive for detecting peripheral vestibulopathy [32]. Studies on the rehabilitative effects of active versus passive head thrusts, as conducted by Schubert et al. demonstrated that increases in VOR gain were independent of whether the training involved active or passive head impulses. However, the outcomes for passive head movements were notably more variable [33, 34]. The consistency of our results with previous studies underscores the reliability of the vHIT system as a tool for measuring the VOR. Still, we need to be concerned regarding the validity of the current technology of the vHIT system to recognize small differences in vestibular function.

A recent study by Keren et al. [31] investigated the effects of sleep deprivation on medical residents, finding a significant reduction in VOR gain following night shifts and a correlation between sleep duration and changes in VOR gain. The authors suggest that vHIT could serve as a useful screening tool for sleep deprivation, noting that participants were unable to accurately assess their fatigue levels via questionnaires. Their study specifically measured VOR gain in response to a distant target (1.5 m).

In contrast, the present study expands upon this by evaluating a broader spectrum of vestibular functions, including VOR gain in response to both near and far targets, as well as assessing active and passive movements and VOR adaptation. Despite the effects of sleep deprivation on VOR gain that both Keren et al. and our study found, our results indicate that most vestibular functions remained intact [35]. We hypothesize that the discrepancy in which our study found most functions to be unharmed may arise from the more comprehensive nature of our assessment, which captures a wider array of vestibular responses that may have adapted to sleep deprivation, in contrast to the more focused examination of a single VOR function in their study.

The suggestion by Keren et al. to use vHIT as a screening tool for sleep deprivation is promising, but our study indicates that relying solely on vHIT could miss important aspects of vestibular dysfunction [35]. Therefore, integrating vHIT with other diagnostic tools or questionnaires that assess fatigue and cognitive function could improve the accuracy of sleep deprivation screening.

In 2013, Scherer et al. investigated the effects of 26 h of sleep deprivation on dynamic visual acuity (DVA) in healthy military personnel. Their study found no change in DVA along the yaw axis and only a slight decline along the pitch axis [24]. They concluded that functional gaze stability remains preserved following short‐term sleep deprivation. In contrast, our study found no significant differences in DVA across all axes between normal conditions and sleep deprivation. This suggests that chronic adaptation to sleep deprivation may mitigate the decline in VOR gain measured using this tool too.

Few limitations are demonstrated in this research. The comparative prospective study was performed in “real world” conditions. Participants were not instructed on how many hours to sleep in both conditions, and they were not supervised by anyone during the night. This kind of research might harm the purity of the results but is nonetheless important to study, and potentially makes every significant result even more robust. To ensure that conditions represented true sleep deprivation, every participant completed a questionnaire that gives insight into the amount of time spent sleeping and other objective and subjective conditions. Another potential limitation is the small cohort size, with only 14 participants, three of whom were excluded for various reasons. While most results did not show significant differences between sleep deprivation and normal sleep, a larger cohort might reveal significant findings. However, the sample size calculated before the study was fewer than 10, and most results did not demonstrate a clear trend, suggesting that the true effect of sleep deprivation was demonstrated in our study.

5. Conclusions

In conclusion, sleep deprivation has a complex effect on the VOR in healthcare night‐shift workers. The significant influence on the ability to perform tasks that require utilization of the vestibular system might be somewhat subdued by long‐term working while sleep‐deprived. However, even in a healthy population, the combined effects of decreased vigilance, alertness, motor skills, and decision‐making following sleep deprivation warrant caution when performing high‐risk tasks requiring dexterity, such as surgical procedures. Further research is needed to determine whether the mild impact of chronic sleep deprivation on the vestibular system is due to adaptive neuroplastic mechanisms or the intrinsic resilience of the vestibular pathways under extreme conditions.

Conflicts of Interest

The authors declare no conflicts of interest.

Livneh N., Shimron H. B.‐r., Tessler I., et al., “The Effect of Sleep Deprivation on the Vestibulo‐Ocular Reflex and Vestibular Adaptation in Healthcare Night‐Shift Workers,” The Laryngoscope 135, no. 9 (2025): 3364–3370, 10.1002/lary.32160.