Introduction

Eye strain is a common ocular symptom in eye clinic patients with normal best-corrected visual acuity [1, 2]. It is a growing health problem that has been exacerbated during the COVID-19 pandemic due to increased screen time from telework and e-learning [3–8]. Digital eye strain due to virtual learning and other new technologies is an emerging social burden [4–8] and the prevalence of eye strain has been reported to be as high as 68.5% [5] and 81% [6] among visual terminal display users. Pavel et al. [6] reviewed and discussed numerous contributing factors for modern eye strain, including environmental and work factors, device-related factors, uncorrected refractive errors, ocular surface disorders, blink, and musculoskeletal modifications. Lotfy et al. [7] conducted a cross-sectional online survey of 412 participants and multiple logistic regression analysis revealed that studenthood and increased nighttime screen use were independent risk factors for digital eye strain by an odds ratio (OR) of 10.60 and 3.99, respectively. Presbyopia also seems to have developed earlier in life during the COVID-19 pandemic probably due to stress and increased digitalization leading to eye strain even in the younger population [9]. However, most reports describe review and survey results only and real examination data have not been sufficiently documented. Eye strain is a symptom with multiple presentations, including ocular pain, headache, heaviness, blurred vision, irritation, itching, lacrimation, and photophobia [5, 10]. It encompasses a broad spectrum of vision-related and nonvision-related pathologies, in addition to problems related to visual load, air conditioning, lighting, and ergonomics [4–8].

We have previously proposed a classification of eye strain (eye fatigue) and hypothesized that eye strain originates from both visual and nonvisual etiologies, with corneal hyperesthesia and photosensitivity playing major roles in nonvisual eye strain [11]. Neuropathic pain and intrinsically photosensitive retinal ganglion cells (ipRGCs) underlie this classification model, whereby corneal pain might be partly mediated by ipRGCs [12, 13]. IpRGCs are a subtype of RGCs that signal to the central biological clock to control circadian rhythm [14]. They are partly associated with photophobia and ocular pain in response to blue-light in the presence of melanopsin [14] and absence of rod and cone photoreceptor input [15]. IpRGCs project into the ciliary marginal zone [16] and iris [17], mediating innate light aversion to tissues in the anterior segments that receive trigeminal innervation [13]. Clinical investigations of glaucoma suggested ipRGCs are distributed in the retinal ganglion cell complex (GCC) [18–20] and thinning of the GCC may correspond to decreased ipRGC numbers or function. Although ophthalmological characteristics of ipRGCs are not determined clinically due to their low cell numbers, blurred vision may develop with a thinning of the GCC, as reported in glaucoma [21, 22]. Further, ipRGC-driven light sensitivity may be relieved with short wavelength light filtering eyewear to block blue-light and reduce light scattering [23–25]. Our previous investigation indicated GCC thickness was positively correlated with eye strain [11] and our hypothesis of ipRGC-associated eye strain is supported by numerous other investigations [26–31].

Dry eye (DE) is a major etiology of eye strain [10]. Unstable tear film and uneven corneal epithelial surface can disturb optical quality and induce higher order aberration and impaired functional visual acuity [32–35]. DE-associated visual symptoms can be treated with topical medications [36, 37] and they are effective in improving short tear break-up time (TBUT), superficial punctate keratopathy (SPK), and subjective symptoms including eye fatigue, blurred vision, photophobia and pain.

Eye strain has an unmet need in eye practice, mainly due to the difficulty in identifying causality, especially in cases without any detectable vision-related disorders. Few studies have evaluated eye strain in relation to objective examination data. This study aimed to investigate any association between abnormal ocular examination results and the common visual symptoms of eye strain, blurred vision and photophobia, to complement our previous association of eye strain with retinal thickness in a larger sample size [11]. In addition, we attempted to identify risk factors for each symptom and explore relevant management for patients with normal vision complaining of visual symptoms.

Methods

Results

The demographics and examination results in relation to each symptom are shown in Table 1. The prevalence of each symptom was 31.8% for eye strain, 22.5% for blurred vision, and 16.0% for photophobia. The prevalence of each abnormal examination value was 53.0% for short TBUT, 23.5% for SPK and 6.8% for central involvement of SPK, and 28.8% for a low Schirmer test value. The mean age was significantly older in cases with blurred vision and photophobia. All symptoms were more prevalent in women than men. Short TBUT and SPK were more prevalent in individuals with all three symptoms compared with those without. Regarding retinal thickness, mean GCC was thicker in those with eye strain than in those without eye strain, and thinner in those with blurred vision than in those without blurred vision.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Comparison of parameters between groups with and without visual symptoms.

https://doi.org/10.1371/journal.pone.0293320.t001

Fig 1 and S2 Table show the univariable and age and sex adjusted logistic regression analyses of examination results in relation to the presence of the three visual symptoms. Individuals with short TBUT and SPK were more likely to present with eye strain, blurred vision and photophobia, while patients with thick GCC were more likely to present with eye strain, and patients with thin GCC were more likely to present with blurred vision. Schirmer test value, peripapillary NFL thickness and full macular thickness were not associated with the three symptoms tested.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Univariable logistic regression analyses of dry eye-related signs and retinal thickness in relation to the presence of eye strain, blurred vison, and photophobia.

Non-adjusted results are Fig 1A and age and sex adjusted results are Fig 1B. *P < 0.01. Odds ratios and 95% confidence intervals are indicated. Abnormal parameters were defined as follows: tear break-up time ≤ 5 s, Schirmer test value ≤ 5 mm, the presence of corneal epitheliopathy (positive corneal staining). Retinal thickness value was binarized by mean of the present cohort; thicker than mean = 1, thinner than mean = 0. Men = 1, women = 0. Abbreviations: GCC, ganglion cell complex thickness; NFL, retinal nerve fiber layer thickness; Whole macula, full thickness of whole macula.

https://doi.org/10.1371/journal.pone.0293320.g001

Discussion

The present study confirmed the results of recently published works [4–7], and linked underlying pathologies with patients’ symptoms by demonstrating ORs calculated with objective data. The current study found short TBUT and SPK were significant risk factors for three visual symptoms with considerable ORs. This supports previous reports of both SPK and short TBUT being known causes of impaired visual function [33–35]. In the present cohort, short TBUT was observed as early disruption of tear film in the whole corneal area, whereas SPK was mostly located inferiorly, suggesting visual symptoms could develop even with local SPK not involving the central area. The current results confirm previous clinical findings that visual symptoms of DE could be treated with eyedrops by ameliorating TBUT and SPK [36, 37].

DE patients often complain of photophobia and it is generally believed to derive from light scattering on the ocular surface. However, in addition to developing from the optic media and retina, photophobia could also develop from the brain, systemic nervous system or psychological factors [39, 40]. Burstein et al. [39] introduced four definitions of photophobia: 1) abnormal sensitivity to light; 2) ocular discomfort, also termed photo-oculodynia; 3) exacerbation of headache by light; and 4) general aversion to light. Of these, ocular discomfort is often also applicable for eye strain-related photophobia and pain in DE and non-DE individuals since light inevitably enters the eyes while the eyes are open, especially when they are watching a digital display. The authors hypothesized that light may indirectly activate intraocular trigeminal nociceptors, that, in turn, activate second-order nociceptive neurons in the spinal trigeminal nucleus to induce ocular discomfort or pain in the eye. Other investigators have reported that a cervical sympathetic ganglion block reduced spontaneous pain and light sensitivity, and increased production of tears [40]. Based on these investigations, the trigeminal and sympathetic nervous systems may be involved in a subtype of eye strain.

The present study has further confirmed GCC thickening is a risk factor for eye strain. IpRGCs in the GCC may be associated with eye strain that causes discomfort or pain. This could be evoked through the trigeminal circuit after receiving short wavelength light emitted from lighting and displays [26–31]. Blue-light filtering eyewear may be effective for ocular symptoms related to ipRGC activation [23–25], however, it should be noted that sensitivity to light may depend on age, race and genetic factors [41, 42] and blue-light filtering does not work in certain populations. Thinning of the GCC was a risk factor for blurred vision and is a typical finding in Parkinson’s disease [43] and vision-threatening diseases, such as glaucoma [21, 22], ischemic optic neuropathy [44] and optic neuritis [45].

Female gender was a significant risk factor for visual symptoms in the present study. Women complain of their symptoms more strongly than men, and genetic, hormonal, psychological and socio-cultural factors have been suggested to be associated with this observation [46]. One study also described girls may refrain from reporting their symptoms in adolescence [47]. As such, this gender difference should be recognized in the management of DE and other diseases. Age is also a considerable risk factor for eye strain. Younger individuals may suffer a more vigorous accommodative load and photo-oculodynia than the elderly due to better light transmission and retinal light sensitivity. Meanwhile, older individuals suffer DE-associated eye strain due to a susceptive ocular surface and decreased lacrimal function [48].

The current results support the contribution of symptom-oriented clinical reasoning combined with conventional statistics [49, 50], where practitioners first listen to the patient then correlate symptoms with examination results. In order to eliminate risk factors for eye strain, practitioners may prescribe DE eyedrops and consider recommending light blocking strategies through eyewear, shields and digital displays after excluding organic and functional disorders. Detailed examination may be required if a patient with a thin GCC suffers blurred vision; namely, a visual field test, macular threshold and focal electro-retinogram. An OR could be a convenient threshold indicator to speculate the probability of symptoms in relation to particular objective data.

The present study has several strengths. First, the sample size is large enough to have the statistical power to calculate the OR in first-visit patients in a multiple logistic regression analysis. Second, the present analysis was performed on patients visiting before the COVID-19 pandemic. During the pandemic, many people changed lifestyle habits through a shift to telework and quarantining. This has likely impacted ocular, systemic and mental status and as such, means any generalizations from data during this period are unreliable. The current results may have sufficient generalizability since this is a two-center study with minimized bias.

The current study has some limitations. First, a functional visual acuity, contrast sensitivity, higher order aberration and accommodation test should have been performed to evaluate the severity of visual function in detail. Visual acuity corrected with patients’ spectacles should also have been examined since inappropriate correction and excessive visual load are major causes of eye strain. Second, autonomic nerve activity could be further accessed to determine the neural contribution to visual symptoms since it is closely associated with eye strain-related symptoms including ocular discomfort, pain and photophobia. Eye pain could also be further evaluated with a validated questionnaire (such as the short-form McGill pain questionnaire) and esthesiometers. Third, the anatomy, physiology and function of human ipRGCs remain unclear. The measurement of ipRGC function with pupillary reflex or electroretinogram would be needed to determine the possibility of developing ocular symptoms. We must acknowledge the limitation of selection bias, as this study is clinic-based and people with any concerns may visit eye clinic.

In conclusion, the current study successfully identified female gender, short TBUT, and SPK as significant risk factors for eye strain, blurred vision, and photophobia with considerable ORs, and further confirmed GCC thickening is a risk factor for eye strain. Moreover, the present results contribute to understanding the association of visual symptoms with DE and short-wavelength light activating ipRGC-mediated ocular discomfort and pain and should help in the clinical reasoning and management of common visual symptoms in daily eye practice.

Supporting information