The effect of filters and varying illumination on contrast sensitivity in eyes with moderate to severe visual impairment

To investigate the effect of filters and illumination on contrast sensitivity in persons with cataract, pseudophakia, maculopathy and glaucoma to provide a guide for eye care providers in low vision rehabilitation. A within-subjects experimental design with a counter-balanced presentation technique was employed in this study. The contrast sensitivity of eyes with cataract, pseudophakia, maculopathy and glaucoma was measured with filters (no filter, yellow, pink and orange) combined with increasing illumination levels (100 lx, 300 lx, 700 lx and 1000 lx) using the SpotChecks™ contrast sensitivity chart. The data were analyzed using descriptive statistics and two-way repeated measures ANOVA. The yellow filter at 100 lx significantly improved contrast sensitivity in the maculopathy group. There were no significant improvements with either intervention in the rest of the groups. There was, however, a significant interaction between filters and illumination in the cataract group. There were small improvements in contrast sensitivity at low illumination levels with the yellow filter in the maculopathy group, and this could be considered in clinical practice and low vision rehabilitation. Overall, filters at most illumination levels did not benefit most groups.


Introduction
Contrast sensitivity (CS) is the ability of the eye to detect the lowest difference in luminance between an object and its background [1]. Poor contrast sensitivity can occur in the presence of good visual acuity or before loss in visual acuity can be perceived. All the major causes of irreversible visual impairment such as glaucoma, age-related macular degeneration, retinitis pigmentosa and diabetic retinopathy severely affect contrast sensitivity even when structural changes are too subtle for a confident diagnosis to be made [2][3][4][5]. Unlike visual acuity testing which provides information on the ability to resolve fine detail, contrast sensitivity testing provides more detail on functioning in a real-world environment [6]. It has therefore gained a lot of importance in the visual functioning assessment of persons living with low vision, due to its impact on their vision-related quality of life.
To increase independence and the general quality of life of low-vision patients with reduced contrast sensitivity, interventions to increase both surround and focused illumination are often administered [7]. Another common method of improving contrast sensitivity is the use of spectral filters [8,9].
Spectral filters work by the selective inhibition and transmission of light of different wavelengths [10]. It is known that an inversely proportional relationship exists between light scattering intensity of the eye and the wavelength of incident light, and hence, blue light which is of short wavelength is scattered more and therefore is implicated in reduced contrast sensitivity [11]. There is evidence which suggests that spectral filters like the yellow filter limit the transmission of blue light, resulting in less light scatter, and improving the quality of the retinal image [10,12].
Currently, spectral filters are being prescribed to patients with reduced CS based on anecdotal reports of subjectively perceived improvements in CS with the filters. While most studies found positive results (objective outcome measure) with the spectral filters in persons with low vision [8,9,13,14], and in healthy eyes [15,16], others found neutral or negative results [17,18]. The controversy on the benefit of spectral filters may be because most studies did not factor in the effect of filters on each specific ocular disease, although it is known that the mechanism of CS function loss is different for each condition [19].
Additionally, although filters and illumination have been studied over the years as tools for improving CS, the potential of the coupling effect of these two interventions has rarely been studied especially for various ocular conditions. This study, therefore, sought to investigate the unique and synergistic effect of spectral filters and illumination on CS measures in patients with cataract, pseudophakia, maculopathy and glaucoma.

Materials
This study employed the SpotChecks™ contrast sensitivity chart (Precison-Vision.com) (Fig. 1) which is a near chart consisting of spherical targets of decreasing contrast levels. It employs minimum detectability and has an advantage over letter charts like the Pelli-Robson chart in terms of decreasing learning effects in within-subject designs. The test comprises 25 levels (arranged in ascending order of discrimination difficulty) of 5 spherical contrast targets of equal contrast level each. The SpotChecks™ CS chart has been shown to have good repeatability and strong agreement with the Pelli-Robson chart in both adults [20] and children [21].

Design
The study was an experimental one, which utilized a within-subjects design by exposing all subjects to each intervention. A counter-balanced presentation technique was employed in order to reduce learning and fatigue effects. Participants and sampling procedure The purposive sampling method was used to select persons with moderate to severe visual impairment (MSVI) caused by selected diseases known to reduce contrast sensitivity. MSVI was defined as corrected visual acuity of worse than 0.50logMAR (6/18) but better than 1.30logMAR (3/60). All participants were of African descent and had contrast sensitivities of not more than 1.65logCS to avoid ceiling effect. A lower highest contrast sensitivity was assumed because the population sampled was persons with visual impairment.
The researchers' judgment in selecting the participants was guided by a preliminary examination, consisting of an eye health examination to ascertain the cause of the reduction in vision, visual acuity testing and refraction to ensure that selected eyes indeed had a visual impairment fit to be classified as moderate to severe visual impairment, and a baseline contrast sensitivity measurement.
A total of eighty-four persons and ninetythree eyes with MSVI, comprising 35 (37.6%) eyes with cataract, 24 (25.8%) with pseudophakia (pseudophakic eyes with MSVI due to posterior capsular opacities as a complication of Small Incision Cataract Surgery), 16 (17.2%) with maculopathy (macular degeneration and macular scarring) and 18 (19.4%) with glaucoma were included. Recruitment was conducted at the Department of Optometry and Vision science eye clinic, University of Cape Coast, and the Bishop Ackon Memorial Christian Eye Centre. Selection of participants was based on known diagnosis. A slitlamp examination and indirect ophthalmoscopy were carried out on the anterior and posterior segments of the eyes, respectively, to confirm the diagnosis and to exclude persons with ocular comorbidities. Their ages ranged from 18 to 87 years. In accordance with the principle of autonomy, informed consent was sought from sampled participants. Persons with psychological disorders or who were generally unwell were excluded as they were likely to give unreliable subjective responses. Patients with ocular comorbidities were also excluded.

Preliminary procedures
Using the LogMAR chart, the visual acuity of each participant was assessed monocularly at 4 m. The anterior and posterior segments of the eye were examined using a Keeler slitlamp biomicroscope with a 90D volk lens to identify and confirm the underlying cause of the visual impairment. Retinoscopy with a Keeler streak retinoscope and subjective refraction were carried out to make sure that eyes recruited were in their corrected state. Individuals with two eyes having different VA and CS which satisfied the inclusion criteria had both eyes included in the study. Persons with equal VA and CS in both eyes had only their right eye included in the study. Individuals with two eyes but who had only one eye which was eligible had that eye included in the study irrespective of whether it was the right eye or the left eye.

Main experiment
The YF-170 digital light meter (Tenmars Electronics Co., Ltd.6f, 586, Rui Guang road, Neihu, Taipei, Taiwan) was used to set the ambient room illumination to 100 lx. Subsequent increments in focused illuminance on the CS chart were also measured with the same device. The SpotChecks™ chart was illuminated to four different levels of illumination; 100 lx, 300 lx, 700 lx, and 1000 lx and CS measured each time. Each illumination level was then combined with each of the filters guided by the matrix below. The SpotChecks™ chart was placed 40 cm away from the participants. They were asked to tick each round contrast target they could discriminate and stop when they thought they could not see any more targets. With each change in intervention, the contrast chart was also changed to avoid learning effects.
To ensure internal validity and limit order and sequence effects, a counterbalanced presentation technique (using a balanced Latin square design) was employed where the order of presentation of the various interventions was systematically shuffled up (Table 1).
Four orders of presentation were used; Informed consent was sought from all participants. All collected data were stored in a password protected folder, only accessible to the supervisor and the researcher. No names of participants or other identity giveaways were recorded. To honor the assurance of confidentiality given to participants, all collected data were coded.

Data analysis
All collected data were inputted into IBM SPSS version 25.0 and analysis carried out using descriptive tools and two-way repeated measures ANOVA, with age and visual acuity as covariates.

Discussion
The study showed a significant effect of filters only in the maculopathy group where at 100 lx, the yellow filter improved CS. There was no significant effect of either intervention in the other three groups.
In eyes with cataract, increasing illumination performed poorly in eyes with cataract confirming a study by Smedowski et al. [22] who concluded that the mainstay intervention of increasing illumination may rather be detrimental to persons with cataract. The mechanism of contrast sensitivity reduction in cataract is thought to be the result of increased light scatter, especially light of short wavelength, through the thickened opaque lens which reduces the quality of the retinal image and adversely affects the discrimination ability of the visual system especially in bright daylight [23]. Hence, higher levels of lighting increase light scatter and glare perception. None of the filters with or without illumination improved CS contrasting studies which have found the yellow filter to improve CS [9,15,16,18]. Filters interacted significantly with illumination showing that the effect of an increase in illumination may be impacted by the type of filter it is combined with. For example, while higher lighting improved CS when combined with the orange filter, it reduced CS when combined with the yellow filter (Fig. 2).  This study showed an insignificant effect of filters on CS in eyes with pseudophakia which agree with systematic reviews [24][25][26] that have generally concluded on inadequate evidence to support the use of filters in pseudophakic eyes (Fig. 3). This study found no effect of illumination on CS. To the best of the authors' knowledge, the effect of increasing illumination on CS in pseudophakic eyes has not been investigated. Contrast sensitivity loss in pseudophakia is mainly due to the absence of the blue light-blocking ability of the natural crystalline lens [27] and also to degradation of the retinal image as a result of chromatic aberration of intraocular lenses [28,29]. This has prompted researchers to investigate the effect of a blue-light filtering intraocular lens designed to resemble the natural crystalline lens. So far, the results have been conflicting and inconclusive.
Analysis in eyes with maculopathy showed that at 100 lx, the yellow filter significantly improved contrast sensitivity (Fig. 4) confirming studies by Ahmad et al. (2017) [8], Caballe-Fontanet et al. [30] and Wolffsohn et al. [13] Macular pigments, lutein, zeaxanthin and meso-Zeaxanthin have been shown to possess blue-light filtering ability, enhancing the quality of central vision by the reduction of the effects of blue light-chromatic aberration and disability glare [31][32][33][34]. Renzi and Hammond [35] showed that macular pigment density (MPOD) was positively correlated with contrast sensitivity due to its selective absorption of blue light. Since cone photopigment regeneration is adversely affected in maculopathy [36], contrast sensitivity is also reduced. Since macular pigment is yellow in color, researchers have investigated the effect of an external yellow filter in eyes with Age-related Macular Degeneration (AMD) [37,38] and macular scarring [8] and found positive results which are in agreement with the results of this study. Increasing illumination with no filter, however, had no significant main effect. In the absence of the glare-reducing ability of macular pigments, it is postulated that high intensities of light may induce glare which override the positive impact of illumination on visual discrimination. Further research on this is recommended.
In glaucoma, none of the filters significantly improved CS disagreeing with Ding et al. [39] and  Ahmad et al. [8] who found significant improvements with the yellow filter. Increasing illumination did not significantly affect the performances of the filters. Since patients with glaucoma experience high levels of glare and haloes due to pressure build-up in the eye, it was posited that filters, in reducing forward transmission of certain wavelengths, could improve the quality of the retinal image. The results of this study, however, showed a disagreement to that assertion (Fig. 5). Illumination also did not have a significant effect disagreeing with earlier studies [40,41]. There is still a lot of controversy concerning these interventions, especially the effect of filters, in persons with low vision. Much of this could be attributed to differing population characteristics, experiment set-up and choice of analysis. This study did not match age, visual acuity and gender in the selection of participants and significant differences existed in the mean age and visual acuity of the different groups. This was accounted for by employing these two variables as covariates in order to ensure that the results are not confounded. This study was delimited to only eyes with cataract, pseudophakia, maculopathy and glaucoma. Other contrast sensitivity-reducing ocular conditions such as retinitis pigmentosa, corneal opacities, and optic atrophy were excluded due to difficulty in obtaining adequate numbers to make sound analysis. The study was also delimited to the purposive sampling method as only a limited number of persons in the society could serve as primary data sources.

Conclusion
There were small improvements in contrast sensitivity at low illumination levels with the yellow filter in the maculopathy group, and this could be considered in clinical practice and low vision rehabilitation. Overall, filters at most illumination levels did not benefit most groups.