1. Introduction

Along the line of myopia control research, it was found that spending more time outdoors can significantly delay myopia onset with the underlying protective mechanisms of outdoor light yet to be determined [1,2]. Potential factors include the difference in light intensity, spectral composition or spatial frequency content of outdoor environments [3–5].

In the context of spectral composition, the longitudinal chromatic aberration (LCA) was suggested as a sign-decoder of defocus and, therefore, as a directional guidance of eye growth [6–8]. However, contradictory results were found between animal species when raised in narrow-band light. On the one hand, monkeys and tree shrews reared in long-wavelength light showed relatively shorter eye lengths and hyperopia [9–11]. On the other hand, for chicks, mice and guinea pigs short-wavelength light presented an inhibitory signal for eye growth and myopia development [12–15]. Adding to the debate, two recent studies on humans showed a similar tendency as in previous non-primate animal models [16,17].

However, the previous studies used commercial lightning devices and filters and could not fully control for LCA-induced defocus and luminance. Therefore, it needs to be clarified whether the growth responses of the eye are triggered by the light chromaticity per se or rather by the parameters defocus and luminance that were not controlled for. In the current study, it was possible to control for both factors using a novel interference-based optical setup. Consequently, the influence of narrow-band, isoluminant and aberration-free light on axial length was investigated.

2. Materials and methods

2.1 Participants’ data

Written informed consent was obtained from each participant and all experimental procedures adhered to the tenets of the Declaration of Helsinki and were in accordance with the guidelines of the independent ethics committee of the medical faculty at the University of Tübingen. Participants were excluded in case of structural or functional ocular abnormalities. Normal color vision was ensured using Ishihara plates (Kanehara Trading Inc., Japan). The mean age of the included 18 participants was 24.6 $\pm$ 3.4 years and ranged from 20 to 31 years. Furthermore, the right and left eyes had mean spherical equivalents of −0.74 $\pm$ 1.22 D and −0.59 $\pm$ 1.31 D (as measured by objective refraction, ZEISS iProfiler+, Carl Zeiss Vision GmbH, Germany). Axial length was measured via swept-source biometry (ZEISS IOL Master 700, Carl Zeiss Meditec AG, Germany) and average axial lengths of the subjects were 23.47 $\pm$ 0.79 mm and 23.49 $\pm$ 0.89 mm for the right and left eyes, respectively.

2.2 Interference-based optical setup

This study used a custom optical setup to generate interference fringe stimuli on the retina. The details of this setup has been described in detail elsewhere [18]. In brief, the system is based on a spatial light modulator that generates two laterally separated wavefronts. These two wavefronts are focused into the nodal plane of the eye, thereby generating interference fringes on the retina [19]. Due to the Maxwellian view configuration, the system provides aberration-free stimulus presentation as it bypasses the eye’s optics. The initial optical design of this setup provided a field of view (FoV) of 1.5$^\circ$. To increase the stimulus FoV to the desired 12$^\circ$, an extra telescope, relaying the pupil and retina conjugate planes, was added to the system and the final focusing lens was changed from 400 mm to 50 mm focal length (AC254-050-A-ML, Thorlabs, Germany). Light source was a supercontinuum laser (SuperK Compact, NKT Photonics, Denmark) providing spatially coherent light of a wide spectrum. The light was spectrally filtered to achieve the different experimental conditions needed. For the narrow-band light conditions, narrow bandpass filters with a full width at half maximum (FWHM) of 10 nm and center wavelengths at 450 nm (FB450-10, Thorlabs, Germany), 550 nm (FB550-10), and 650 nm (FB650-10) were used. For each participant, a holder fixed to a custom dental impression (bite bar) was used to immobilize and control the position of the head during the stimulation period. Accurate centering of the eye in front of the setup aperture was ensured by a camera (DMK 27AUP031, The Imaging Source Europe GmbH, Germany) monitoring the pupil position.

The power of each chromatic condition was measured at the participant’s pupil position (sensor: 818-UV/DB; power meter: 843-R, Newport Spectra-Physics GmbH, Germany). During the study light input power was then adjusted by using neutral density filters, ensuring an equal retinal illuminance of 4.9 td (based on the equation provided by [20]) across the different chromatic conditions. The Maxwellian view condition of the setup renders the pupil size of the subject neglectable, thus the regarding luminance value is given by about 5 cd/m$^{2}$.

2.3 Stimuli

The optical setup projects interference fringes - a stripe pattern stimulus - directly on the retina, covering a central retinal area of about 12$^\circ$. An example of the stimulus for all three color conditions is shown in Fig. 1. The spatial frequency and rotation of the fringes were changed randomly every five seconds to provide a wide range of mid-spatial frequencies that are assumed to play an important role in emmetropization [21]. The fringe’s spatial frequency covered a range from 3 to 18 cpd in steps of 3 cpd, in orientations of 0$^\circ$, 45$^\circ$, 90$^\circ$ and 135$^\circ$.

 

Fig. 1. Simplified schematic of the interference based aberration-free chromatic stimulator. The output from a supercontinuum source was bandpass filtered to achieve the desired light condition. To generate two laterally separated wavefronts in the pupil of the eye a spatial light modulator (SLM) was used. The left subset shows the checkerboard patterned mask on the SLM. Only the shown first diffraction order path was used for the stimulus formation, while the zeroth and higher diffraction orders were blocked by the field stop. An additional telescope was added to the previous published design to achieve a field of view of 12$^\circ$. Red lines denotes the conjugated plane of the SLM and the retina (r), black lines of the pupil (p). The right subset shows the retinal aberration-free stimulus with the three different light conditions applied and exemplary rotations as well as spatial frequencies: long-wavelength (L, 650 $\pm$ 5 nm), mid-wavelength (M, 550 $\pm$ 5 nm), and short-wavelength (S, 450 $\pm$ 5 nm).

Download Full Size | PDF

2.4 Study protocol

The study protocol started at 9 a.m. for all participants. Prior to each measurement cycle, the participants underwent a 10 min washout period - looking at a plain wall in distance - to minimize the effects of confounding factors, such as accommodation, on axial length [22]. The participants’ left eyes were then stimulated for 20 min in the optical setup, while their right eyes were occluded and served as control eyes. Before and after the illumination period, biometric measurements of both eyes were acquired using the ZEISS IOL Master 700 (Carl Zeiss Meditec AG, Germany). The entire process was repeated for all three randomized chromaticity conditions.

2.5 Data analysis and statistics

Data and statistical analysis was performed with the Statistics toolbox of MATLAB (MATLAB 2020a, The MathWorks, Inc., USA). The Lilliefors test ensured normal distribution of pre-vs-post differences of axial length for further statistical analysis. Paired sample t-tests were applied to evaluate whether the axial length changes for the separate stimulus conditions differed significantly from zero and whether test and control eye were statistically significant from each other. A two-factor repeated-measures analysis of variances (ANOVA) was performed with "eye" (2 levels: OD and OS) and "chromaticity" (3 levels: S, M, L) as within-subject factors to examine the individual and interaction effects of these factors on the response variable axial length change.

3. Results

The mean changes in axial length in the left eye are shown in Table 1 and in Fig. 2. There were no significant pre-vs-post-illumination changes for any illumination condition neither in the test (S: p = 0.14, M: p = 0.65, L: p = 0.27) nor in the control eye (S: p = 0.84, M: p = 0.28, L: p = 0.96). Changes of the test and control eye did not differ significantly for any chromaticity condition (S: p = 0.26, M: p = 0.22, L: p = 0.55). The two-way repeated-measures ANOVA showed no significant effect neither in chromaticity (F = 0.54; p = 0.59), eye (F = 0.25; p = 0.63), nor in the interaction between both factors (F = 1.12; p = 0.34).

 

Fig. 2. Axial length changes from baseline in the test eye (OS, filled bars) and control eye (OD, patterned bars) for each chromaticity condition (S = 450 nm, M = 550 nm, L = 650 nm). Results are given as mean $\pm$ standard error.

Download Full Size | PDF

Table 1. Axial length changes in microns for all illumination paradigms (S, M, L). Results are given as mean $\pm$ standard error.

View Table

4. Discussion

The current study investigated the effects of chromaticity on axial length without LCA-induced defocus. To the authors’ best knowledge, this is the first study to separate both factors - chromaticity and the LCA induced defocus - using an interference-based optical setup for the purpose of human myopia research. The study found no statistically significant changes in axial length after exposure to short-, mid- and long-wavelength light. Moreover, no significant differences between test eye and control eye or between chromaticity conditions were detected. Under these test conditions, it can be suggested that eye growth seems rather triggered by the LCA-induced defocus than chromaticity of the incident light. However, it is noteworthy that a tendency of axial length changes can be observed (see Fig. 2), with short-wavelength light leading to a general decrease in axial length and long-wavelength light to an increase in axial length in the test eye.

Compared to recent literature, the current study drew differing conclusions [16,17]. Lou and Ostrin [16] revealed a relative decrease after short-wavelength light and a significant increase after long-wavelength and polychromatic light in the stimulated eye. Thakur and colleagues [17] also showed a significant reduction in axial length after exposure to short-wavelength light in combination with a substantial increase after illumination with long-wavelength light. This response remained unchanged by the presence of hyperopic defocus. Therefore, both studies concluded that chromaticity seems to be the primary signal for eye growth regulation. Despite overlapping tendencies of axial length changes, the substantial differences in outcomes between the previous and current study might be caused by dissimilarities in the study designs. Both previous studies did not eliminate the LCA-induced defocus, as ambient room lighting with LEDs [17] or additional filters [16] was provided as light stimulus. Moreover, the human spectral sensitivity curve was not considered, thus, the perceived brightness of the light varied among chromaticity conditions. The longer exposure time (60 min vs. 20 min) could have also led to the observed variations.

Moreover, it was again observed that the human visual system seems to respond similarly to non-mammal animal models in response to narrow-band light [12–15]. Before, it was assumed that the human reactions might be more comparable to these of non-human primates due to the evolutionary similarities in ocular physiology as well as the process of emmetropization and the nature of the vision-induced refractive errors [8,23–25]. However, the most prominent study on rhesus monkeys and tree shrews showed an opposite reaction to the observed effects in humans with long-wavelength light [9–11].

This study also comes along with certain limitations that might have accounted for the small effect size. These include the relatively short exposure time of only 20 min. The 20 min time frame presents a trade-off between exposure duration versus participant comfort and attention. Within the chosen duration the participants were still able to steadily fixate onto the stimulus target, although this could only be verified subjectively and not objectively. Longer exposure times could have led to more apparent and secured results, but could have caused other undesired effects on axial length, such as proximal accommodation [22]. Despite having tested within a common constant time frame of 3 hours [16,17] per participant, future measurements should be performed on separate days at the same time of the day. This would also allow longer exposure times per visit. Due to technical limitations in the optical setup, a trade-off existed between the size of the stimulated area and the retinal illuminance. A larger stimulus area, compared to the original design of the system [18], was chosen in order to ensure stimulation of the central retina in the case of potential eye movements during the fixation period. In return, this led to a decrease of intensity of the stimulation light across the stimulated retinal area. However, not only the central retina [26] but also the peripheral retina [27] play an important role in emmetropization. As a consequence, it is not known whether a higher light intensity or an extended stimulus area have a higher impact on the observed biometric changes.

5. Conclusions and outlook

In conclusion, the study showed that the eye does not seem to respond to chromaticity signals alone under the tested conditions. With regard to previous work, it is suggested that defocus presents a stronger signal than the light chromaticity for eye growth regulation. Future studies are planned to investigate axial and choroidal changes using the same optical aberration-free setup but with a larger stimulation area, higher illuminance and polychromatic light in order to obtain more distinct and controlled results.