Impact and visualization of scotomatic glare in central visual field perception

Strong monochromatic point light sources such as Light Emitting Diodes (LED) or Lasers have been increasingly used in recent decades. This also raises the risk of misuse resulting in glare phenomena and associated visual impairment. The objective of this prospective and partially blinded study was the visualization and characterization of glare-induced scotomas in visual field by dazzling with monochromatic point light sources in terms of disability and discomfort glare. Automated threshold perimetry under dazzling by LED exposure at three different wavelengths (470, 530 and 625 nm) and four different intensities (25, 50, 75, and 100%) was performed in 31 healthy subjects resulting in 434 visual field examinations. Visual disability was measured by sensitivity loss in the central 30 ◦ as compared to unexposed controls and visualized by reconstruction of mean visual fields for each group via backward-calculation. Psychological glare was assessed by subsequent questionnaire and evaluated based on the de Boer rating scale of discomfort. Increasing glare intensities resulted in a significant decrease in mean sensitivity for all wavelengths tested, paralleled by an increase of discomfort glare. The loss of sensitivity was scattered over all quadrants with accentuation of the corresponding mean exposure area. Reconstructed visual fields confirmed visual impairment in all quadrants at an extent of at least 30 ◦ . We conclude that even off-axis light exposure may affect central visual field perception. Our results extend previous research on directed light interaction and contribute in explaining its incapacitating impact on human performance.


Introduction
The misuse of laser pointers and similar handheld devices, once considered a trivial prank, has grown to a public health problem of epidemic relevance, with more than 100,000 documented events of laser strikes against aircraft over the past 20 years (Wawrzyński et al., 2022;Carroll & Richards, 2018;Murphy, 2024).Considering the multitude of respective press releases in road and rail transportation, sporting events and political riots, the estimated rate of unknown cases is probably much higher.
The development of ever more powerful light sources requires intense research in the field of glare.Monochromatic point light sources with a small angle such as visible diode lasers and Light Emitting Diodes (LEDs) are of special interest in this context.The way how vision is affected by directed monochromatic light emitted in the visible spectrum depends on a variety of related details including distance, wavelength, emissive power, beam divergence, accommodative state of the eye, environmental and ambient light conditions.Although the direct view into a strong coherent light source is always hazardous, the remote risk of retinal injury from distant illumination is considered generally low (Bhavsar, Michel, Greenwald, Cunningham, & Freund, 2021;Marshall, O'Hagan, & Tyrer, 2016), implicating that glareinduced visual impairment has to be regarded the main hazard for mobility-related tasks i.e., road and air traffic.(DeMik et al., 2013;Palakkamanil & Fielden, 2015).
According to the Commission Internationale de l' Éclairages (CIE), glare refers to a condition of vision in which there is discomfort and/or a reduction in the ability to see details or objects, caused by an unsuitable distribution of luminance or an unsuitable range of luminance values (Vos et al., 2002).From this definition, two types of glare can be derived, discomfort glare and disability glare.While discomfort glare is defined as a psychophysical sensation that causes annoyance and avoidance behavior, disability glare relates to a physiological light interaction that impairs vision up to the extent of visual incapacitation (Parsons, 1910).
The multiple aspects of discomfort and disability glare with its widespread pathophysiological effects on the human visual system have been extensively studied since the last century (Parsons, 1910;Holladay, 1926;Stiles, 1947;Mainster & Turner, 2012).This is particularly true for environmental glare induced by traffic-or work-related lighting conditions.Scatter is considered an important perpetrator in this context, leading to enhanced glare sensitivity and consecutive avoidance behavior (van den Berg, 1991;Ortiz-Peregrina et al., 2020).Accordingly, intraocular straylight reduction has been a major topic in the fields of automotive headlight concepts and occupational healthcare.Studies include polarized filters, anti-reflective coatings, as well as surgical reduction of ocular opacities by refractive laser surgery and artificial lens implants (van den Berg, Franssen, Kruijt, & Coppens, 2013;Hammond, 2012;Spadea, Maraone, Verboschi, Vingolo, & Tognetto, 2016;Xu et al., 2015;Łabuz, Reus, & van den Berg, 2015;Łabuz, Reus, & van den Berg, 2016).None of these studies, however, precisely defined the functional interference of visible monochromatic directed light such as laser or LED irradiation with human visual perception.As a result, the specific mechanism of visual degradation by such point light sources remains unexplained.
Given the impact of central retinal light processing, studies on corresponding visual field defects are surprisingly rare.As even studies on subthreshold central irradiation are missing in this context, we hypothesize that concerns of macular damage through experimental exposure to high-luminance light might be one of the main reasons for this scientific gap, particularly when different wavelengths are taken into account.Hence, while some aspects such as visual acuity, contrast sensitivity, color vision, and afterimages under visible laser exposure have been explored (Reidenbach et al., 2014;Reidenbach, Dollinger, Ott, Janßen, & Brose, 2008), the specific impact of directed monochromatic light on visual field perception is essentially unknown.As a result most studies available on this topic have been focusing on visible light interaction in the presence of scatter in front of the retina (Wood, Wild, & Crews, 1987;Bergin et al., 2011;Oleszczuk, Bergin, & Sharkawi, 2012;Anderson, Redmond, McDowell, Breslin, & Zlatkova, 2009).Such type of glare, however, might be completely different from the type of glare that is caused by local overload of retinal photoreceptors exposed to excess doses of collimated light.
This pilot study prospectively investigated the effect of eccentric dazzling glare on visual field perception in ophthalmologically healthy subjects.The experimental design was based on a technical upgrade of the perimetric device used, with light emitting diodes (LEDs) simulating the effects of monochromatic laser exposure.The objectives were (1) the measurement of the extent of visual field deficiency including perceived discomfort under different wavelengths and light intensities and (2) the visualization of resulting glare scotomas including backwardcalculation of sensitivity losses.

Materials and methods
The experimental design was set up in collaboration with the Fraunhofer Institute of Optronics, System Technologies and Image Exploitation (IOSB), Ettlingen, Germany, and the Department of Ophthalmology, German Air Force Centre of Aerospace Medicine, Cologne/Fuerstenfeldbruck.

Participants
Subject recruitment was performed at the German Air Force Centre of Aerospace Medicine in Fuerstenfeldbruck, Germany.Informed consent was obtained from all participants in accordance to the tenets of the Declaration of Helsinki.Ethical approval was granted by the Bavarian State Medical Association in Munich, Germany (Ref. 19032, Nov. 05, 2019).
All study participants were ophthalmologically healthy subjects with normal anterior and posterior eye segments, BCVA of 20/20 or better, and normal intraocular pressure (<22 mmHg).In addition, clear optical media and intact visual field were required.Applicants with ophthalmic diseases, in particular glaucoma, optic disc drusen, corneal or lens opacities as well as high refractive error (myopia <-6D, hyperopia >+5D, astigmatism >2D) were excluded.Of 48 volunteers recruited, 31 fully completed the test series and were subsequently included into statistical analysis.Of these, 16 were female and 15 male.The average age was 32.2 (median 31 yrs) ranging from 18 to 58 yrs at baseline examination.Age stratification included three subgroups: 18-29 yrs (n = 13), 30-40 yrs (n = 13), and > 50 yrs old participants (n = 5).

Experimental setup
Visual field examination was performed using an automated Octopus TM 900 perimeter (V2.3.1) by Haag-Streit, Köniz, Switzerland, that was modified by installation of a dazzling light source on an adjustable carrier slide along the inner horopter area.For safety reasons, incoherent Light Emitting Diodes were used instead of coherent laser light.LEDs were interchangeable to choose between three different wavelengths of 625 nm (red), 530 nm (green), and 470 nm (blue).The maximum optical radiation reaching the study subjects' eyes varied dependent on wavelength due to technical and safety reasons (BGBl, 2010;BAuA, 2013).Continuous down regulation of luminous intensity was provided by a 4-channel DC4100 LED driver from Thorlabs, Bergkirchen, Germany.For the purpose of this study, bounded intervals were defined at 100%, 75%, 50%, and 25% of nominal power at the output of the LED driver.
We measured (1) the radiant flux (μW) and (2) the irradiance (μW/ cm 2 ) at eye level at all degrees of illumination.In addition to these radiometric quantities, we specified photometric quantities such as (3) the illuminance (lux) at eye level and (4) the luminance (cd/m 2 ) of the light source.Because of the multitude of variables to be considered in the measurements of light emitted by LED, we calculated the illuminance by radiometric measurements based on the CIE "Principles Governing Photometry" (BIPM, 2019).(5) Retinal illuminance (Td) was calculated from the luminances and the mean values of pupil areas (mm 2 ).A summary can be found in Table 1.For further details about radiometric and photometric quantities, pupil sizes and the estimated retinal illuminances please refer to supplementary material Table S1.
Eccentricity of the dazzling LEDs was chosen to be 15 • inferior and 10 • left off-center thus avoiding direct irradiation exposure of the fovea.White-on-white-perimetry was performed using a standard 30-2 automated setting with a 6 • paraxial grid resulting in 76 data points each.Contrary to conventional, photopic white-on-white perimetry, which is supposed to run at a background luminance of 10 cd/m 2 , background lighting in the perimeter hemisphere was switched off for the purpose of this study.The resulting background luminance was 0.255 cd/m 2 , comparable with mesopic light conditions.The maximum light sensitivity to be measured was device-dependently limited to 40 dB.Sensitivity threshold was defined as the stimulus luminance perceived with a 50% probability.

Subjective glare
For measurement of discomfort glare, de Boer's dazzling scale (De Boer, 1967;Wittlich, 2010) was applied.The scale, which is still the most common visual perception scale in the field of automotive and public lighting, provides five rating options: unbearable (1), disturbing (3), just admissible (5), satisfactory (7), and unnoticeable (9).Intercepts between these ratings are considered arbitrary and left undefined for reasons of discriminatory power in statistical evaluation.For graphical representation please see Fig. 1.

Test procedure
Routine ophthalmological examination including OCT of the macular region was performed in all subjects followed by initial visual field testing without dazzling.Only right eyes were investigated, while nontested eyes were occluded.Experimental visual field testing was performed by successive exposure to one wavelength at four light intensities with 100%, 75%, 50%, and 25% each.Sequences were randomized and blinded with regard to exposure levels.In order to ensure reliable test results, the test procedure was aborted in case of unacceptable numbers of fixation losses and/or false positive or negative errors of more than 15%, and repeated another day.Blink and pupil position checks were implemented for compliance control.In case of successful testing, participants were asked to assess subjective discomfort according to de Boer's rating scale.A maximum of two examinations per day were conducted, with a minimum recovery interval of five hours in order to allow possible afterimages to subside and to protect the macula from excessive light exposure.After completion of experimental testings, ophthalmic examinations including unexposed visual field and OCT were repeated in order to exclude any light-induced injury in study participants.

Statistical analysis
For statistical evaluation, measured sensitivities of each data point were extracted from the perimeter software (Eyesuite V2.04, Haag Streit) and transcribed to an external relational database.Analysis, statistical tests and graphical visualization were performed using Excel TM and Python TM -based software (Virtanen et al., 2020).For testing of sample mean differences, Mann-Whitney-U and Wilcoxon signed rank tests were used.For testing of group variances, univariate and multivariate linear regression analysis based on ordinary least square model fit was performed.Statistical significance levels were uniformly set to 5%.Reconstruction of visual fields was performed by backward calculation of average point sensitivity losses and visualized using grayscale encoding based on threshold definitions as provided by Haag Streit (Racette et al., 2019).Grayscale intercepts were empirically set to 5 dB steps.

Results
Statistical analysis included 434 completed visual fields with glare and unexposed controls, resulting in a total of 32,984 data points, excluding individual discomfort ratings.Statistical results are presented according to the dichotomous study concept by differentiation into objective and subjective test results where objective results refer to the physiological disability glare, while subjective results refer to the psychophysical discomfort glare.

Disability glare
Native visual fields displayed threshold mean sensitivities ranging from 34.1 dB to 39.7 dB (37.8 dB ±1.5 dB).In experimental visual fields, increasing glare intensities were accompanied by a general decrease in MS with accentuation of the inferonasal quadrant, that displayed significantly lower sensitivities under glare than the remaining quadrants (p < 0.001).The effect was seen at all wavelengths.In contrast, inter-quadrant calculation in unexposed subjects did not display any significant disparities (0.977 > p > 0.272).The recorded loss of sensitivity was highest at 530 nm (green), followed by 470 nm (blue), while the least impairment was observed at 625 nm (red) for all intensities (p ≤ 0.001) (Fig. 2).Regression analysis was based on the assumption that the mean sensitivity is related inversely proportional to the square of LED power.The adjusted coefficient of determination (R 2 adj ) for the overall model ranged from 0.125 for 625 nm (F(1, 122) = 18.63, p < 0.001) to 0.224 for 470 nm (F(1,122) = 36.58,p< 0.001) and 0.252 for 530 nm (F(1, 122) = 42.45,p < 0.001) indicating a good model fit.
Corresponding standard deviations ranged from 2.0 dB at 470 nm to 2.2 dB at 530 nm and 3.3 dB at 625 nm reflecting group homogeneity in the distribution of disability glare across different wavelengths.
Visual field reconstruction (Fig. 3) indicated that sensitivity losses were strongly dependent on local glare at all wavelengths and intensities, with its highest scotomatic impairments around 15 • in the lower left quadrant and 10-15 • off-centered in the area of maximum dazzle.While this pattern was nearly identical in all visual fields investigated, the highest rates of relative scotomas were seen in 530 nm green, followed by 470 nm blue and 625 nm red LEDs.Increasing intensities of light caused increasing rates of scotomas in all cases.
In order to gain additional insight into the topographical and relative distributions of dazzle scotomas, we calculated the amounts of decibel (dB) change from baseline to experimental visual fields in LED exposed study subjects.Supplementary material Figure S1 shows the differences of disability glare caused by different wavelengths and intensities elaborated in detail by choosing an intercept of 5 dB.The most  Stratification by gender did not show any differences regarding the MS of male and female study participants at all light intensities and wavelengths (0.999 > p > 0.236).
Age stratification, in contrast, revealed a statistically significant difference in MS for persons ≥ 50 yrs as compared to the remainder of the test group (p < 0.001), while no statistical difference was seen in the 18-29 and 30-49 yrs groups (0.798 > p > 0.100).If age is included into linear regression analysis, statistics confirm that correlations of decreasing MS with increasing light intensity were considerably higher in elderly study participants across all wavelengths.The R 2 adj for the agerelated overall model ranged between 0.385 for 625 nm (F(2, 121) = 39.50, p < 0.001), 0.433 for 470 nm (F(2, 121) = 47.92,p < 0.001) and 0.561 for 530 nm (F(2, 121) = 79.67,p < 0.001).

Discomfort glare
Statistical analysis of individual ratings documented a stepwise increase of subjective discomfort with increasing light intensities which was observed at all wavelengths.At 25% glare intensity, the mean value  corresponded to the assessment "still acceptable" in general.Perceived discomfort continued to increase above 50% and 75% to almost "disturbing" at 100% light intensity.(Fig. 4) Exposure levels of 100% were associated with a general increase of discomfort glare which was statistically significant in nearly all cases as compared to lower intensities (470 nm: p < 0.002; 530 nm: p < 0.034; 625 nm: p < 0.001).The only exception were red wavelengths at 100% that caused a marginal insignificant increase of discomfort as compared to 75% glare intensities (p = 0.075).
The increase of reported discomfort was not significantly different by gender for all wavelengths and intensities (0.967 > p > 0.057) except for 530 nm at 100%.At this wavelength male study subjects reported higher discomfort for green exposures (p = 0.014).Stratification by age was inconsistent with regard to the groups of 18-29 and 30-40 yrs old participants and showed no statistically significant wavelengthdependent differences between < 50 and ≥ 50 yrs old participants (0.978 > p > 0.067).

Synopsis
Fig. 5 summarizes the statistical interrelations of subjective and objective glare impairments according to different wavelengths and light intensities.As compared by relative change from baseline values, the differences between rated discomfort and measured visual field degradation were most pronounced at 625 nm (red), and least distinct at 530 nm (green).Perceived glare sensitivity displays higher relative changes than measured disability in all cases.While changes in visual field defects ranged in the 1 to 25% area, subjective ratings changed up to 60% as compared to unexposed baseline measurements.Perceived discomfort was highest at 470 nm, whereas most visual field deficiencies were seen at 530 nm indicating that blue wavelengths were perceived more annoying or distracting as green or red colored lights.However, the effect was not consistently observed in all study subjects at equal ranges.

Discussion and conclusions
This study investigates the extent to which retinal exposure to glare derived from LED lights and limited to discrete wavelengths causes a reduction in perimetric visual field sensitivity.We hypothesize that beyond scattered light causing disability and loss of contrast sensitivity, long-range neuroretinal mechanisms may contribute in the reduction of peripheral sensitivity which is likely to be in addition to the pure optical effects of adding a veil of light over the central or paracentral retina.Given the increasing frequencies of laser-and flashlight-induced dazzle events in civilian as well as in police and military-related environments, the resulting transient reduction in effective visual field might be a stronger predictor of human incapacitation than conventional dazzling models as known so far.
The test setup addressed quantification and qualification of visual field degradation using conventional visual field testing by static perimetry.Complementary to the work of Reidenbach et al. (Reidenbach et al., 2008;Reidenbach et al., 2014) who researched the impact of laser dazzling on visual acuity, pupillary reflex, and retinal afterimages, the present study was designed with regard to the assessment of scotomatic glare and its accompanying subjective discomfort under mesopic conditions.To our knowledge, this is the first scientific report describing visualization and characterization of glare-related visual field scotomas induced by monochromatic directed light.
Our results indicate that eccentric overexposure with monochromatic light affects the entire visual field up to 30 • of measurement in an intensity-and wavelength-dependent manner, including absolute and relative scotomas scattered around the source of glare.In addition to the known mechanisms such as forward light scatter in the eye (Stiles & Parsons, 1929), which tends to flood the inner eye causing a more generalized image of visual field impairment (veiling glare) (Vos, 2003), the loss of glare-related sensitivity seen in this study was directly linked to retinal correspondence with the external source of light (scotomatic glare) (Mainster & Turner, 2012).
The specific perimetric loss of sensitivity induced by this type of glare has to be differentiated from the decrease in contrast sensitivity known from scatter patterns that are induced by haze or opacification of the eye's refractive media (Wood et al., 1987).In intraocular straylight studies, the effect has been explained by inter-reflections of light within the eye (Bergin et al., 2011;Oleszczuk et al., 2012), which is probably not the predominant mechanism regarding a point-shaped source of light (Patterson, Bargary, & Barbur, 2015).Consistent with other studies on this topic (Bargary, Jia, & Barbur, 2014), the findings in this study rather suggest that photoreceptor saturation may play a key role in the induction of disability and discomfort glare.
Paradoxically, a central glare depression without peripheral depression was seen in reconstructed visual fields.A possible explanation of this observation might relate to the high density of cone receptors in the center of the retina, maintaining the most effective sensitivity of luminance difference discrimination even in case of superimposition by glare.Hence, the lowest loss of sensitivity was observed in this area.In addition, we hypothesize a static reinforcement of this mechanism by a Stiles-Crawford-Effect (SCE)-like phenomenon, induced by the eccentric position of the LED which hindered the emitted light to pass through the center of the pupil (Westheimer, 2008).In this case, a disproportionate reduction of glare will occur, with the result of only a slight reduction of mean sensitivity in the central visual field.The presence of adaptive pupillary reflexes under glare (Reidenbach et al., 2008), the interreflection of light within the eye, and the angular dependence of straylight might have further modulated the specific appearance of reconstructed visual fields which would rather support a multifactorial mechanism of action than a single causative event (van den Berg et al.,

2013).
Despite similar radiant flux and irradiance levels, the highest loss of mean sensitivity was found at 530 nm (green), followed by 470 nm (blue), and the least sensitivity loss was seen at 625 nm (red).This implicates a higher rate of perimetric scotomas under dazzling with 470 nm (blue) rather than with 625 nm (red), which was unexpected since measurement and calculation of photometric quantities at eye level revealed illuminances at 625 nm (red) being about 8 times brighter than the ones at 470 nm (blue), and 530 nm (green) being up to 15 times brighter than 470 nm (blue).In search of explanations for this observation, we finally addressed the spectral sensitivity of retinal photoreceptors.Nevertheless, while photometric quantities at 530 nm clearly corresponded with the peak sensitivity of human L and M cone receptors at wavelengths of approximately 550 nm (Bowmaker & Dartnall, 1980;Müller, Frings, & Möhrlen, 2019), the extent of sensitivity loss at 470 nm and 625 nm still did not match the referring illuminance levels.
However, these presumptions are restricted to photopic environments.There is common consens that these interrelations will change dependent on ambient luminance (BIPM, 2019).Since it is well known that depending on the adaptation state of the eye, brightness perception for different wavelengths varies according to ambient light conditions, it can be assumed that the mesopic environment in the present study induced a shift of the spectral luminous efficiency function towards blue wavelengths (Stockman & Sharpe, 2006).The representation of the luminous efficiency functions in Fig. 6 illustrates it graphically.This view is supported by the results of scotopic luminous flux calculations as listed in the supplementary material (Table S1).Notwithstanding of these results, it has to be noted that especially for smaller glare point sizes as used in our test setup, the effective retinal illuminance and the calculated retinal illuminance will not be identical.Apart from this, wavelength-dependent factors beyond inherent photoreceptor sensitivity, such as red-dominated fundus reflectance and blue-dominated scatter from cornea and lens (Coppens, Franssen, & van den Berg, 2006), might have contributed (Ginis, Perez, Bueno, Pennos, & Artal, 2013;Williamson et al., 2017).
Evaluation of de Boer ratings showed an increase in discomfort with increasing glare intensity for all wavelengths.In contrast to visual impairment, no significant difference could be elaborated in glare ratings with regard to different wavelengths.This is contrary to the findings of Flannagan et al. (Flannagan, Sivak, Ensing, & Simmons, 1989) who reported strong wavelengths-dependencies using monochromatic sources of light.However, it has to be questioned in how far the light emitted by projector with filters is comparable with the monochromatic light derived by LED.Nevertheless, as compared to the much less pronounced disability glare in form of scotomas at 625 nm (red), discomfort ratings at this wavelength were fairly substantial as shown in the synopsis part of the results.There are some studies indicating that discomfort glare might be related to retinal illuminance rather than the illuminance of the pupil plane or the luminous flux entering the eye (Bargary et al., 2014).Taken together, as compared to psychophysical sensitivity loss, the differentiation of data points regarding discomfort was much less evident, which was most probably due to the high variation of individual/subjective estimates.This might support the hypothesis that the sensation of discomfort is more related to the spatial distribution of light on the retina than to the illuminance or total amount of light (Bargary et al., 2014).
Subgroup analysis did not reveal any significant gender-related differences regarding the statistical impact of disability glare.Age-related differences were found in > 50 years old study participants, which was expected as straylight is known to increase with the extent of agerelated opacities of refractive media (Aslam, Haider, & Murray, 2007;Bergin et al., 2011) related to higher scatter coefficients and equivalent veiling luminance in older individuals.With regard to discomfort glare no age or gender preferences could be found which appears to be in agreement with the working groups of Saur (Saur, 1969) and Pierson (Pierson, Wienold, & Bodart, 2018) who reported similar findings at a large interindividual variance (Marié, Montés-Micó, Martínez-Albert, García-Marqués, & Cerviño, 2021).
Our study has limitations.As is true for most laser-related investigations in humans, safety concerns have limited the experimental possibilities and optional designs (BAuA, 2013;BGBl, 2010).In the setup described here, laser emission was simulated by replacement through monochromatic LEDs of different luminance.It is possible that the specific effects of laser radiation would have been different from the  LED-induced effects in this study.Depending on distance and retinal spot size, it can be assumed that a visible laser will introduce a higher irradiance at eye level which in turn increases retinal illuminance.Accordingly, the effect on perception is probably underestimated in this study as compared to the glare derived from commercially available laser pointers.Furthermore, the source of light was moved out of the visual axis to an eccentric position in order to protect the central retina from excessive visual light exposure.There is no doubt that such precautional measures have influenced the final results by introducing additional variables and biometric confounding (Foutch, Stringham, Smith, Novar, & Garcia, 2009).Moreover, the luminance of wavelengths used for glare induction was based on different initial values in order to avoid excess doses of light at 100% exposure.This was particularly true for 470 nm (blue), which had to be technically lowered on behalf of the ethics comittee.We emphasize the use of isoluminant glare sources in future studies for reasons of comparability.Nevertheless, mesopic environments involve a variety of influencing factors such as rod-cone interactions, receptor saturation, mixed spectral sensitivities and different topical distributions of rods and cones, and the changes during the transition from rods to cones, resulting in inconsistent estimates of the specific retinal mechanism involved (Stockman & Sharpe, 2006).Finally, we tend to suggest that de Boer's scale, although representing the most widely used rating scale for glare environments, might be less useful in small-study populations and real-life applications (Theeuwes, Alferdinck, & Perel, 2002).In order to control for this bias, larger study collectives and more detailed rating scales (Fotios, 2015) have to be taken into consideration.
Although our results do not provide insight into critical safety parameters beyond scotomatic visual field defects, they certainly complement previous research on the extent of visual degradation caused by laser dazzling in pedestrians, drivers, and pilots (STO, 2018).Hence, glare-induced visual field reduction might be a stronger predictor of driver or pilot incapacitation than visual impairment only.Since the test setup reflected realistic conditions as seen in mesopic civilian and military environments, our results might contribute to improve and develop anti-dazzle training devices such as situational awareness simulators for professional drivers, flight simulators for pilots and training simulators for police and military (Coelho, Freitas, & Williamson, 2016;Földes et al., 2024).

Conclusions
In summary, our results indicate that LED dazzling with subthreshold amounts of monochromatic light tends to affect the complete central visual field of exposed individuals with relative and absolute scotomas scattered around the source of glare, even when exposure is decentered and the light is not precisely focused by physiological accommodation.In addition, increasing glare intensities are accompanied by decreasing sensitivities in visual field perception and an increase in subjective discomfort.The wavelength of the applied light illumination appears to have more influence on the disability glare than on discomfort glare.Although the specific mechanisms of visual degradation by distant monochromatic light exposure remain unexplained, our results certainly complement previous research on the extent of visual degradation caused by laser and other point light sources.Further research is needed to explore the mechanism of glare induced by high-luminance light exposure and its dependency on discrete wavelengths of the visible spectrum.

Ethics/human subject approval
Ethical approval was granted by the Bavarian State Medical Association in Munich, Germany (Ref. 19032, Nov. 05, 2019).Informed consent was obtained from all participants in accordance to the tenets of the Declaration of Helsinki.No identifiable details about participants were included for publication.

Funding Source Declaration
The perimeter device was provided as a temporary item of loan by Fraunhofer IOSB, Ettlingen, Germany.Furthermore, this research has not been funded by any scientific, industrial, or governmental institution.The present work was performed in fulfillment of the requirements for obtaining the degree Doctor of Medicine (Dr.med.) of the first author (Diana Hering).

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
prominent provider of central and paracentral visual field deficiencies was 530 nm.The inferonasal quadrant was clearly affected with the highest frequency of absolute scotomas (n = 71), comparable with the depth of scotomas as found in the blind spot area (n = 70).Notably, increased frequencies of absolute scotomas occurred in the 30 • range of visual fields (n = 15).

Fig. 3 .
Fig. 3. Scotomatic glare in reconstructed visual fields due to paracentral LED illumination with 3 different wavelengths and 4 different light intensities.Each data point represents the arithmetic mean in dB of 31 individual visual fields.(Grayscale adapted to Haag-Streit Visual Field Digest (Racette et al., 2019)).

Fig. 5 .
Fig. 5. Relative change of MS and rating on the de Boer scale for 470nm, 530 nm and 625 nm.

Table 1
Measured radiometric and photometric quantities and calculated retinal illuminance.
Fig. 1. "de Boer scale" as used in our test setup.D. Hering et al.