Impact of Myopia Control Spectacles on Retinal Image Contrast

Abstract

Purpose

Peripheral defocus and image contrast modulation are key strategies in myopia control, but both inherently reduce retinal image contrast. To date, no in vivo study has directly compared these effects. This study evaluated the contrast reduction profiles of different myopia control spectacles to better understand their underlying mechanisms.

Methods

Through-focus images (TFIs) were obtained with a double-pass instrument in a model eye and in human eyes wearing four spectacle designs: DIMS, Stellest, diffusion optics technology (DOT), and MyoCare. The naked eye served as a reference. Thirteen participants (mean age = 29.0 ± 3.5 years) were tested at the fovea and at 10° and 20° eccentricities in the temporal and superior fields. Lenses were decentered by 1.5 cm to assess off-axis effects. Model eye tests included only the central field. Retinal image contrast (RIC) was calculated as the coefficient of variation of pixel intensity in TFIs.

Results

In the model eye, RIC ranked from highest to lowest as follows: naked eye, DIMS, DOT, MyoCare, and Stellest. In vivo, DOT results were excluded due to a weak signal. A nonparametric test revealed significant differences in RIC among lenses at the fovea (P = 0.009) and at 10° in the temporal/superior fields (P = 0.05). Post hoc analysis showed that Stellest produced the greatest reduction in RIC, whereas MyoCare and DIMS demonstrated similar levels of contrast loss.

Conclusions

All myopia control lenses reduced retinal image contrast, with Stellest inducing the strongest reduction. MyoCare and DIMS showed comparable effects. Lens-induced optics primarily influenced contrast in the near periphery, while ocular optics dominated at larger eccentricities.

The rapid growth of myopia prevalence has raised significant concerns in the world¹ as highly developed myopia is associated with various sight-threatening pathologies.² A systematic review estimated³ that by the end of 2050, about 32% of adolescents in China will become high myopes. The prevention of myopia development is therefore a major public health priority. Among all interventions for myopia control, peripheral refraction manipulation has been popular in clinical activities and has been well developed in the past decade. In general, the technique for peripheral refraction manipulation can be categorized into two distinct technologies: peripheral defocus technology (PDT)^4–8 and diffusion optics technology (DOT; SightGlass Vision, Dallas, Texas, USA).^9–11

In the past four decades, PDT has been studied extensively, including animal studies^12–17 and clinical studies,^{4–6,18–24} and their optical characteristics have been extensively investigated.^25–29 The hypothesis for PDT assumed that the human eye always tries to compensate for the defocus signals in the periphery by modulating ocular shape, which means relative peripheral myopic defocus is the cue for slowing central myopia progression.^17,30 The representative products in clinics include DIMS (Defocus Incorporated Multiple Segments; Hoya Corporation, Tokyo, Japan),³¹ Stellest (Highly Aspherical Lenslet Target Technology; Essilor Luxottica, Charenton-le-Pont, France),⁶ and MyoCare (Zeiss, Oberkochen, Germany).^7,32 Those lenses introduce myopic defocus in the periphery by utilizing micro-lenslets or concentric annular rings. A recent study using the asymmetric multipoint defocus technique (higher defocus and more densely packed micro-lenslets) has also shown promising results for myopia management.^10,11

In contrast to PDT technology, the DOT technique does not change optical defocus but reduces the contrast of visual signals by adding numerous “micro-scattering” elements in the periphery. The development of DOT was based on the findings of cellular defects in cone photoreceptors in the males of high myopes.¹¹ It was found that the LVAVA haplotype caused higher activity of M and S cones and lower activity of L cones. It was hypothesized that the defect ultimately led to contrast enhancement after visual signal processing, as the photopigment optical density is different in those two groups of cells and thereby promotes myopia progression.

Although PDT and DOT are based on different mechanisms and optical designs, in principle, both techniques lead to a reduction of retinal image contrast. However, few studies have quantitatively compared the contrast reduction induced by different myopia control lenses (PDT or DOT), especially in peripheral visual fields under realistic viewing conditions. It is also possible that PDT and DOT may share the same mechanism for myopia control. Furthermore, it has not been investigated whether contrast degradation is induced solely by the spectacles or if there is a contribution from ocular optics as well.

The purpose of the study was to compare the retinal image contrast profiles produced by four commercial myopia control lenses (three PDT and one DOT technology) using a double-pass through-focus imaging system. By evaluating both model and real eyes at multiple eccentricities, we aim to isolate lens-induced optical effects and clarify their potential roles in myopia control.

Methods

Participants

This study was conducted at Laboratorio de Optica of the University of Murcia (LOUM) in Spain from November 2024 to June 2025. The inclusion criteria include the following: (1) age between 18 years and 35 years, (2) spherical equivalent refraction (SER) ranging from −6.00 to +1.00 diopters (D) with astigmatism less than 1.50 D, (3) corrected distance visual acuity of 0.0 logMAR or better, (4) general normal ocular health, and (5) good compliance with the experimental protocols. The exclusion criteria include the following: (1) any ocular pathology, (2) known accommodative dysfunctions, (3) systemic diseases affecting vision, (4) current use of medications influencing pupil or accommodation function, and (5) inability to maintain stable fixation during the testing procedure. Only the right eye of the participants was examined during the study.

The experimental protocol was fully explained to the participants prior to the initiation of the study. All tests complied with the Declaration of Helsinki. The study was approved by the Institutional Review Board of the University of Murcia (M10/2023/080).

Spectacles

For real eyes, three peripheral defocus-based spectacles were included in the study: DIMS (Hoya Corporation), Stellest (Essilor Luxottica), and MyoCare (Zeiss). DIMS spectacles have a central clear zone for distant vision and micro-lenslets in the periphery that provide 3.50 D myopic defocus for myopia control. Stellest spectacles utilize H.A.L.T. technology that features 11-star rings made of micro-lenslets. Unlike the previous designs, MyoCare spectacles have concentric cylindrical refractive patterns around the central clear zone, which, in addition, to a 4 D myopic defocus in the periphery, introduces also approximately 8 D of astigmatism.

DOT was used for evaluating the contrast value compared with PDT techniques in a model eye. DOT is a novel spectacle lens designed for myopia control based on the theory that lower image contrast can slow axial elongation. The lens maintains central vision clearly while subtly diffusing peripheral light, creating a scattering effect that does not rely on imposed defocus.

Instruments

A double-pass prototype instrument³³ was used to evaluate the contrast of the retinal image. The instrument creates a spot on the retina by using a 780-nm laser, and then the point spread function (PSF) can be recorded by a CMOS camera. A high-speed tunable lens was used to introduce the through-focus images (TFIs) of PSF from −10.00 to +10.00 D in 0.25 D steps. The measurement time is less than 1 second. Since the instrument uses an open-view configuration, the participant can rotate the eye at a specific angle. For investigating the contrast of retinal images after fitting myopia control spectacles, a custom frame was mounted atop the device. The lens moved toward the participant's nasal side by approximately 1.5 cm, ensuring the laser beam always passed through the periphery of the lens. Figure 1 shows the prototype instrument used in this study. The objective refraction of the eye was determined by a Visual Adaptive Optics Simulator (VAO; Voptica S.L., Murcia, Spain).³⁴

Figure 1.

(A) The participant positioned for measurement. (B) Off-axis positioning of the lens.

Experimental Protocol

To ensure all the participants meet the inclusion criteria, objective refraction and best-corrected distance visual acuity were assessed using the VAO instrument. Then, the double-pass instrument was used to evaluate the contrast of through-focus retinal images. The measurement sequence for myopia control lenses (DIMS, Stellest, MyoCare, and DOT) was randomized to reduce the impact of confounding factors such as visual fatigue, inaccurate fixation, and unwanted accommodation response. The measurement always started with the naked eye, which served as a reference for comparing image contrast among the glasses. Five printed Maltese crosses with the side corresponding to 1° visual field were mounted on the wall at a distance of 3 m from the center, temporal 10°, 20°, and nasal 10°, 20°. Peripheral refraction (PR) and relative peripheral refraction (RPR; RPR = PR – central refraction) were determined using the double-pass instrument.

In addition to the real eyes, the myopia control lenses were tested on a model eye that actually reproduced the optical properties of the human eye at the foveal position (Diestia Systems, Athens, Greece). In this setup, a 4.00-mm pupil was used. The model eye was mounted on the right side of the chin rest using an M4 screw, ensuring stable positioning and proper alignment of the optical path. The same horizontal nasal displacement used for the human eye (1.50 cm) was applied to the model eye when holding the myopia control lenses. Only the through-focus images from the central field, together with the experimental conditions, were evaluated (naked eye and decentered myopia control lenses: DIMS, MyoCare, Stellest, and DOT). Off-axis measurements were not performed for the model eye, as it was not optimized for peripheral imaging, and off-axis illumination would substantially increase the amount of oblique astigmatism. To better understand the intersection between the laser beam and the lens surface in the study, a schematic illustration is shown in Figure 2.

Figure 2.

The alignment strategy for the laser beam and lens region in the human eye (A) and artificial eye (B). The laser beam passed through essentially the same region of the lenses during the experiment.

Data Analysis

A custom MATLAB application (MathWorks, Natick, MA, USA) was used to identify the best-focus image. Typically, the best-focus image corresponds to the highest image contrast relative to defocused images, provided there is no substantial astigmatism. However, this relationship can be disrupted by factors such as scattering, diffraction, higher-order aberrations, and severe astigmatism, in which case the best-focus image may not exhibit the highest contrast. To address this, the best-focus image was visually selected in MATLAB with the objective of targeting the circle of least confusion in the conoid of Sturm for the through-focus image series. Representative examples are shown in Figure 3.

Figure 3.

Examples of a through-focus retinal PSF image. Through-focus retinal PSF images from participant 1 at the central and temporal 20° positions are shown as examples. Each row corresponds to one experimental condition (naked eye, DIMS, MyoCare, Stellest, and DOT), and each column represents the PSF image at a given defocus level (from −3.0 to +3.0 D in 1.5 D steps). The composite PSF arrays were generated using a colormap, with the maximum intensity normalized to the highest pixel value among all retinal images for the participant and eccentricity (A, B), or using an adaptive color scale across all PSF images to better visualize the distribution of features (C, D).

The coefficient of variation (CV) measures relative variability, and it is defined as the ratio of the standard deviation to the mean.

$$\begin{array}{c}\hfill C_{\mathrm{cv}}=\frac{\mathrm{SD}\left(\mathrm{pixel}\mathrm{values}\right)}{\mathrm{mean}\left(\mathrm{pixel}\mathrm{values}\right)}\end{array}$$

This image intensity statistic was used to evaluate the contrast in the retinal images. The calculation was performed within a circular region of interest on the sensor corresponding to a 0.5° visual field, centered on the weighted centroid of the PSF image (radius = 128 pixels; PSF image dimensions = 512 × 512 pixels).

In addition to CV, image contrast was also evaluated using the root mean square (RMS) contrast formula:

$$\begin{array}{c}\hfill C_{\mathrm{RMS}}=\mathrm{SD}\left(\mathrm{pixel}\mathrm{values}\right)\end{array}$$

with selected examples shown in Figure 4B.

Figure 4.

Retinal image characteristics (CV and RMS) as a function of relative defocus for two of the participants. Image contrast was evaluated using two metrics: the CV (A1–A4) and the RMS (B1–B4) formula. Data from the DOT lens were excluded because the recorded retinal image brightness was near the detection limit of the instrument. Results from participants 1 (A1, A2, B1 and B2) and 3 (A3, A4, B3 and B4) at the central (A1, B1, A3 and B3) and temporal 20° (A2, A4, B2, and B4) positions are shown as examples. “Ref,” “DIM,” “MYC,” and “STL” denote the naked eye, DIMS, MyoCare, and Stellest conditions, respectively. RMS contrast plots are presented exclusively in this figure and do not appear elsewhere in the article.

The contrast comparison among the different experimental conditions (naked eye, DIMS, MyoCare, and Stellest) was performed under both best-focus and defocus conditions. The defocus image for each participant was selected based on their natural relative peripheral refraction, and thus only peripheral results were evaluated in this analysis. These two analysis types were chosen to explore retinal contrast distribution when participants wore myopia control spectacles under distance-viewing conditions. For example, myopic individuals wearing spectacles often exhibit relative peripheral hyperopia in outdoor environments, where light from the central field is focused on the retina (best-focus condition), while light from the peripheral field is focused beyond the retina (defocus condition). Representative through-focus images from two participants at two eccentricities, along with corresponding contrast plots, are shown in Figure 4.

Data are presented as the median and interquartile range (Q25, Q75) for contrast values (non-Gaussian distribution) and as mean ± 1 standard deviation for other metrics. Statistical analyses were conducted using GraphPad Prism (GraphPad Software, San Diego, CA, USA). Differences among experimental conditions were assessed using the Kruskal–Wallis test, followed by Dunn's post hoc test for multiple comparisons. A two-tailed P value <0.05 was considered statistically significant.

To assess whether an individual's intrinsic peripheral optics influences the contrast reduction induced by myopia control spectacles, the lens producing the lowest contrast at each eccentricity for each participant was identified and analyzed.

Results

Participants

Fifteen participants were initially recruited. Data from two participants were excluded due to extremely low retinal reflex intensity, which could compromise the reliability of the results. Consequently, data from 13 participants were included in the final analysis (mean age = 29.00 ± 3.51 years; SER = −1.77 ± 1.77 D). Detailed demographic information for the participants from VAO is provided in Table 1.

Table 1.

Demographics of the Volunteers (N = 13; 8 Males, 5 Females)

Characteristic	Mean	SD	Minimum	Maximum
Age (y)	29.00	3.51	23	34
S (D)	−1.47	1.68	−6.5	−0.25
C (D)	−0.60	0.31	−1	0
M (D)	−1.77	1.77	−7	−0.25
J0 (D)	0.15	0.25	−0.25	0.49
J45 (D)	−0.05	0.18	−0.38	0.27

Two participants were excluded from the analysis because the brightness of the retinal reflex was close to the sensor's detection limit. C, cylindric power; J0, astigmatism component J0; J45, astigmatism component J45; M, spherical equivalent refraction; S, spherical power.

Examples of Through-Focus PSF Images

Through-focus PSF images (TFI) from participant 1 at two eccentricities (central and temporal 20°) are shown in Figure 3 to illustrate the combined effects of intraocular optics and lens optical properties on the retinal image. Participant 1 exhibited mild central myopia (SER = −1.21 D).

Overall, PSF images from the naked eye closely resembled those obtained with DIMS lenses, although DIMS exhibited a slightly more dispersed edge pattern, accompanied by peripheral brightness attenuation. MyoCare lenses produced a pronounced horizontal focusing effect and, under defocus, showed stronger scattering compared to both the naked eye and DIMS. Rotated astigmatic focal lines can be identified for defocus at +3.0 D. Stellest lenses demonstrated peripheral scattering similar to MyoCare but with substantially less central focusing and a more dispersed pattern. For the DOT lens, the retinal reflex was barely discernible when applying a unified color scale across experimental conditions. The retinal reflex observed in DOT is very likely the reflection of the lens surface.

CV contrast (Fig. 4A) of the TFI from the two participants is presented in Figure 4. In general, contrast curves exhibit an approximately symmetric distribution around the best-focus images in most cases. Nevertheless, in a few cases, the best-focus image does not show the highest contrast. Instead, the plots shift by a few diopters (see Stellest curve for participant 1 at the central position), and the curve does not show a symmetric feature. The contrast profiles obtained from the naked eye and DIMS lenses show comparable patterns, whereas MyoCare shows moderate performance for reducing image contrast, and Stellest shows the strongest performance for reducing image contrast. RMS contrast (Fig. 4B) indicates that the absolute fluctuation of image contrast from the naked eye is much higher than the rest of the experimental conditions. All measured PSF images (except those beyond 3.0 D defocus) have been provided as Supplementary Material for readers interested in further image details.

Comparison of Retinal Contrast Across Different Lenses

The descriptive data of contrast value are summarized in Table 2 and presented in Figure 5 with medians and quartiles. The general trend from the boxplot shows that the naked eye has the highest image contrast, followed by the values reduced in the order DIMS, MyoCare, and Stellest.

Table 2.

Descriptive Statistics of Peripheral Refraction and Image Contrast

			Contrast Value, Median (Q25, Q75)
	Diopter (Mean±SD)		Contrast-BF				Contrast-DF
	PR	RPR	Ref	DIM	MYC	STL	Ref	DIM	MYC	STL
0	−1.54 ± 1.87	—	2.8 (2.08, 3.18)	2.05 (1.67, 3.07)	1.76 (1.39, 2.70)	1.48 (1.29, 2.12)	—	—	—	—
T10	−1.69 ± 1.52	−0.15 ± 0.50	2.26 (1.69, 2.92)	1.96 (1.54, 2.76)	1.68 (1.40, 2.27)	1.36 (1.27, 2.04)	1.99 (1.61, 2.84)	1.79 (1.50, 2.52)	1.64 (1.38, 2.21)	1.31 (1.14, 2.00)
T20	−1.83 ± 1.15	−0.29 ± 0.94	2.11 (1.25, 2.46)	1.94 (1.41, 2.54)	1.68 (1.24, 2.18)	1.30 (1.13, 1.81)	1.70 (1.26, 2.40)	1.49 (1.17, 2.39)	1.38 (1.06, 2.12)	1.13 (1.02, 1.79)
S10	−1.42 ± 1.60	0.12 ± 0.69	1.88 (1.74, 2.86)	1.67 (1.36, 2.31)	1.54 (1.38, 2.19)	1.33 (1.19, 1.82)	1.90 (1.43, 2.86)	1.66 (1.22, 2.30)	1.43 (1.07, 2.12)	1.29 (1.10, 1.83)
S20	−1.62 ± 1.62	−0.08 ± 1.02	1.56 (1.21, 2.29)	1.48 (1.10, 2.08)	1.19 (0.97, 1.95)	1.16 (1.05, 1.71)	1.49 (1.15, 2.32)	1.48 (1.12, 2.08)	1.06 (0.91, 1.87)	1.13 (1.04, 1.70)

Peripheral refraction and image contrast at various eccentricities. Contrast-BF and Contrast-DF refer to image contrast calculated from the best-focus image or RPR-based defocus image. Ref, DIM, MYC, and STL mean the results from reference (naked eye), DIMS, MyoCare, or Stellest. Data are presented as mean ± 1 standard deviation for refraction, and median (interquartile range [Q25, Q75]) for contrast.

Figure 5.

Boxplots showing median, quartiles (Q25, Q75), and whiskers of min/max values of image contrast at various eccentricities and experimental conditions. (A, B) are the contrast results calculated based on the best-focus (BF) image. (C, D) are the contrast results calculated from the defocus image based on the individual's relative peripheral defocus (RPD) value. (A, B) show the same data grouped by eccentricity and lens type, respectively. The same grouping strategy is applied to (C, D). The details are summarized in Table 2.

For the contrast value at best-focus images (Figs. 5A, 5B), the nonparametric test (Table 3) indicated that only center (P = 0.009), T10 (P = 0.05), and S10 (P = 0.05) showed significant differences among different myopia control lenses, and the post hoc comparison further indicated that only the contrast value from Stellest was significantly lower than the naked eye. The intrinsic relative peripheral defocus-based contrast value (Figs. 5C, 5D) showed a similar trend to that observed in the best-focus images. However, for statistics, only T10 showed significant differences, and only Stellest was significantly lower than the naked eye in post hoc test.

Table 3.

Multiple Comparison for Image Contrast

	PR					RPR
	P	Ref	DIM	MYC	STL	P	Ref	DIM	MYC	STL
0	0.009*	A	AB	AB	B	—	—	—	—	—
T10	0.05*	A	AB	AB	B	0.05*	A	AB	AB	B
T20	0.30					0.31
S10	0.05*	A	AB	AB	B	0.21
S20	0.12					0.21

The Kruskal–Wallis test was performed to evaluate the differences among the conditions at different angles (*P < 0.05). The letter displayed in the cell is a compact way to show the results of the post hoc test: columns with the same letter indicate no significant differences between the two groups.

Another general trend is that the contrast value decreases as eccentricity increases. For example, the contrast value from the periphery is always lower than that from the center, and the contrast value from T20/S20 is always lower than that from T10/S10. More details about contrast reduction can be found in Table 2.

Figure 6 shows the through-focus contrast for both the model eye (eccentricity = 0°) and the human eye (only the average is presented, eccentricity = 0°, temporal 10°, and temporal 20°). In general, the model eye showed higher contrast than the human eye, and this trend was maintained in subsequent human eye measurements. However, the absolute contrast levels decreased with increasing eccentricity, and the distinctive patterns observed in the model eye became progressively less apparent in the far periphery. Notably, the contrast curves of the model eye and the human eye almost overlapped at temporal 20° for the reference condition and the DIMS lens. In summary, the reference condition (naked eye) showed the highest contrast, followed by DIMS, MyoCare/DOT, and Stellest.

Figure 6.

Relative defocus as a function of retinal PSF image contrast measured with the model eye and with the human eyes (mean values). “Ref,” “DIM,” “STL,” “MYC,” and “DOT” correspond to the naked eye condition or to measurements taken with a single DIMS, Stellest, MyoCare, or DOT lens positioned 1.5 cm in front of the eyes. The model eye used a planar surface to simulate the human retina. Through-focus contrast was analyzed in the central direction for the model eye over a range of −5.00 to +5.00 D relative to the best-focus image and repeated in human eyes in the central direction and at temporal 10° and temporal 20° over a range of −2.50 to +2.50 D. Human eye data beyond ±2.50 D were excluded because of increased background noise.

Table 4 summarizes the experimental conditions that produced the greatest contrast reduction effect for each individual at the central field and at temporal eccentricities of 10° and 20°. While the overall statistical analysis indicated that Stellest generally yielded the lowest image contrast among the tested lenses, in some cases, MyoCare or DIMS produced a greater contrast reduction than Stellest. Furthermore, the lens associated with the lowest image contrast for a given participant could vary across eccentricities (notably in participant 2). Specifically, in the central visual field, 0%, 7.7%, 7.7%, and 84.6% of participants exhibited the lowest image contrast with the naked eye, DIMS, MyoCare, and Stellest, respectively. In the periphery, the distribution differed. For instance, at temporal 10°, the proportions were 0% for the naked eye, 7.7% for DIMS, 38.5% for MyoCare, and 53.4% for Stellest.

Table 4.

The Lenses With the Lowest Retinal Image Contrast for Individuals at Various Eccentricities

	Center				Temporal 10°				Temporal 20°
Participant	Reference	DIM	MYC	STL	Reference	DIM	MYC	STL	Reference	DIM	MYC	STL
1 (−1.21 D)				1				1				1
2 (−1.40 D)				1			1				1
3 (−2.13 D)				1			1					1
4 (−2.25 D)				1			1					1
5 (−0.75 D)				1				1				1
6 (−0.25 D)				1			1			1
7 (−6.50 D)	—	—	—	—	—	—	—	—	—	—	—
8 (−7.00 D)			1					1	—	—	—	—
9 (−0.25 D)	—	—	—	—	—	—	—	—	—	—	—	—
10 (−1.00 D)				1			1					1
11 (−0.38 D)				1				1				1
12 (−3.00 D)				1				1				1
13 (−1.63 D)				1				1				1
14 (−1.75 D)		1						1		1
15 (−0.38 D)				1		1						1
Sum	Reference = 0/13 (0%)				Reference = 0/13 (0%)				Reference = 0/13 (0%)
	DIM = 1/13 (7.7%)				DIM = 1/13 (7.7%)				DIM = 2/13 (15.4%)
	MYC = 1/13 (7.7%)				MYC = 5/13 (38.5%)				MYC = 1/13 (7.7%)
	STL = 11/13 (84.6%)				STL = 7/13 (53.4%)				STL = 9/13 (69.2%)

This table identifies the experimental condition (naked eye, DIMS, MyoCare, or Stellest) that yielded the lowest image contrast for each participant at various eccentricities. The value in parentheses in the first column indicates the participant’s central SER. Participants 7 (SER = −6.50 D) and 9 (SER = −0.25 D) were excluded from the analysis due to extremely low retinal brightness. A value of “1” marks the condition that produced the lowest image contrast relative to the other three conditions. For peripheral locations, contrast values were calculated at the defocus image based on relative peripheral refraction, whereas for the central location, they were calculated from the best-focus image.

Discussion

In the present study, the peripheral optical profiles associated with contrast reduction were evaluated in four types of myopia control lenses: DIMS, MyoCare, Stellest, and DOT. The experiments were conducted using both human participants and an artificial eye, with the naked eye serving as a reference condition. In real eyes, all tested lenses demonstrated the ability to reduce retinal image contrast, with the effect increasing in the following order: DIMS, MyoCare, and Stellest. The DOT lens could not be evaluated in real eyes due to insufficient retinal signals detected by the sensor. In the artificial eye, a similar trend was observed, and the DOT lens exhibited a moderate contrast-reducing effect. Participants’ intrinsic ocular refraction influenced contrast reduction both centrally and peripherally, leading to deviations from the trend observed in the artificial eye in some cases. These findings offer new insights into the underlying mechanisms of myopia control, particularly the interplay between peripheral defocus and contrast reduction, under the hypothesis that high retinal contrast serves as a risk factor for progressive myopia.

The optical design of micro-lenslets plays a critical role in the degree of contrast reduction observed in myopia control lenses based on peripheral defocus theory. In the present study, Stellest lenses exhibited the greatest reduction in contrast, followed by MyoCare and then DIMS. This raises an intriguing question: why does Stellest induce a stronger reduction in contrast than DIMS, given that both employ micro-lenslet technology? Theoretically, Stellest utilizes an aspherical lenslet design to generate optical defocus. This configuration introduces multiple focal points along the optical axis and enhances diffraction effects in the peripheral field, which may disrupt the distribution of optical power and lead to a more pronounced reduction in image contrast. In contrast, DIMS lenses adopt a conventional spherical lenslet design, resulting in a shallower depth of focus in the periphery and a more concentrated distribution of optical energy. This design, while potentially preserving better image clarity, may consequently produce a weaker contrast reduction effect. Thus, Stellest lenses demonstrate a stronger impact on reducing image contrast compared to DIMS. The horizontally elongated PSF observed with MyoCare can be attributed to multiple overlapping astigmatic images. The toroidal zone of the concentric lens design has a much smaller diameter than the pupil, causing several astigmatic focal lines to overlap. Additionally, the circular aperture of the pupil introduces edge effects, further stretching the image along the horizontal axis. Retinal contrast for the DOT lens was not quantitatively assessed in human eyes in this study due to the extremely low retinal signal, which could lead to inaccuracies in contrast estimation. Nevertheless, the reduced retinal reflectance suggests that if RMS contrast is used instead of CV contrast, the DOT lens may demonstrate the most significant reduction in image contrast. The RMS metric also highlights the influence of retinal brightness, underscoring the importance of ambient illumination in shaping visual image quality.

Retinal optics with myopia control lenses arises from the combined effects of ocular and lens optics. In the present study, the measurements obtained from human eyes reflect the interaction between these two components, whereas results from the artificial eye represent solely the optical characteristics of the lenses projected onto the retinal image plane. The comparison between the human and artificial eyes indicates that, within the measured region, the final peripheral optical effect is primarily governed by the optical properties of the myopia control lenses. However, it should be noted that the experiments were conducted over a limited field, with a maximum eccentricity of 20°, whereas the temporal visual field can extend up to 100°. Theoretically, peripheral optical aberrations in the human eye increase with eccentricity³⁵; therefore, beyond a certain eccentricity, ocular optics may become the dominant factor. The specific angle at which this shift occurs is likely to vary among individuals, depending on their peripheral refraction profile. Overall, Stellest exhibited the strongest contrast reduction effect compared with other peripheral defocus–based lenses. Nevertheless, in some cases, the greatest reduction occurred with DIMS or MyoCare at specific eccentricities. For example, participants 2, 3, and 4 (Table 4) showed the lowest retinal image contrast with Stellest at the central and temporal 20° positions, whereas at temporal 10°, the strongest reduction was observed with MyoCare. These variations suggest that ocular refraction should also be considered in optical designs—potentially through customization—when the objective is to reduce retinal image contrast.

Contrast reduction at the best focus does not necessarily correlate with the clinical efficacy of myopia treatment. For example, 1-year studies on myopia control efficacy, measured as the change in SER, have reported efficacy of 50% for DIMS,³¹ 21% for MyoCare,⁷ 67% for Stellest,⁶ and 74% for DOT.¹¹ However, these efficacy values were derived from separate clinical studies that likely differed in age distribution, baseline refractive error, follow-up duration, study design, and visual environment. Therefore, direct comparisons among lens designs should be interpreted cautiously. In the present study, the contrast reduction effect had the following order:

$$\begin{array}{c}\hfill \mathrm{Stellest}>\left(\mathrm{MyoCare}\approx \mathrm{DOT}\right)>\mathrm{DIMS}\end{array}$$

The apparent discrepancy between optical measurements and clinical outcomes can be attributed to several factors: (1) Field limitation of the current study: The present measurements were taken at a limited number of eccentricities within a relatively narrow field and were therefore largely dominated by lens optics. Such measurements may not capture the image degradation across the complete visual field. In the far periphery, more complex interactions between lens and ocular optics are likely to occur. (2) Neglect of accommodation response and visual environment: The present study did not account for accommodation dynamics or the visual environment (retinocentric view of refraction^36,37), both of which have a substantial influence on retinal image contrast. For instance, in a participant with a flat peripheral refraction profile and accommodative lag, the visual scene will introduce relative hyperopic defocus. If the accommodative lag is 1.00 D, model eye through-focus contrast curves (Fig. 6) predict image contrast reductions of approximately 39.1% for the reference, 49.5% for DIMS, 22.5% for Stellest, 44.5% for MyoCare, and 51.1% for DOT. Furthermore, the visual environment—particularly the dioptric map encountered indoors—strongly affects the peripheral refraction profile. For participants with accurate accommodation and a flat peripheral refraction profile, near-peripheral objects (closer to the observer than the fixation point) will produce hyperopic defocus, thereby reducing retinal image contrast. Differences in visual behavior among study populations may therefore contribute to the observed mismatch between contrast reduction and clinical myopia control efficacy. (3) Additional optical factors: Contrast reduction is only one component of the ocular optical changes induced by myopia control lenses. Other factors, such as higher-order aberrations, magnitude, and sign of peripheral defocus, may also adjust treatment efficacy. For instance, Singh and De Gracia³⁸ provide a comprehensive review of peripheral modification techniques (including DIMS, DOT, MyoCare, and Stellest, the same designs evaluated in the present study), highlighting how myopic defocus, contrast modulation, and optical aberrations may interact as potentially synergistic mechanisms.

PDT and DOT, although based on different theoretical principles, may share a common pathway to modulate myopia development—reducing contrast and degrading image quality. For instance, a theoretical framework investigated the characteristics of the modulation transfer function in the DIMS lens.³⁹ The study indicates that lenslet arrays in the periphery act as a low-pass spatial frequency filter, restricting high-frequency transmission and supporting the hypothesis of contrast reduction in PDT technology for myopia control.³⁹ However, further reducing contrast by incorporating scattering patterns into PDT designs does not necessarily guarantee better myopia control efficacy, as lower contrast could make the recognition of the focus point or the two primary focal lines of astigmatism more difficult for the human eye (the human eye may recognize peripheral refraction through the direction of the focal lines of astigmatism⁴⁰). Therefore, it is very likely that there is a threshold for achieving the maximum myopia control efficacy for the reduced image contrast, optical quality, or relative peripheral defocus. One critical piece of evidence is that DOT 0.2 (mild contrast reduction) did not demonstrate better myopia control efficacy than DOT 0.4 (greater contrast reduction).¹⁰ Another supportive piece of evidence comes from the clinical success of spectacles incorporating negative-power lenslets.⁴¹ However, it does not explain why relative hyperopic defocus promotes myopia development in animal studies.⁴² Moreover, the extremely low retinal brightness recorded by the double-pass device suggests that ambient illumination may influence the efficacy of DOT in myopia control, although further studies are required to validate this hypothesis. As a brief summary, contrast reduction may be considered a potential causal mechanism contributing to myopia control; its effectiveness is likely constrained by the physiological thresholds of the human visual system, particularly peripheral refraction and the capacity of photoreceptors to detect contrast changes.

Although the double-pass instrument provides objective quantification of retinal images with different myopia control lenses, it is important to note that optical quality and psychophysical perception are not equivalent. Human visual perception relies on two fundamental components: (1) the optical component, which includes ocular refraction and the physical properties of the world, and (2) neural adaptation, which is shaped by prior visual experience and cortical processing.⁴³ The measurements in the present study reflect only the former, optical component. Recent work by Pusti et al.⁴⁴ demonstrated that choroidal thickness exhibits a bidirectional response to local optical defocus, highlighting its sensitivity to the sign of defocus. However, this response disappeared after correction of the eye's native aberrations, emphasizing the role of empirical neural processing in defocus recognition. The associated increase in retinal image contrast after correction further suggests a physiological contrast threshold may exist in the human eye for detecting optical defocus. A novel hypothesis considering the human eye infers the sign and magnitude of optical defocus by decoding the orientation of peripheral blur images.^40,45,46 An important basis for the theory is the wavelength-dependent response characteristics of photosensitive cells.^47–50 Moreover, emmetropes and myopes exhibit different sensitivities to myopic and hyperopic defocus, with myopic individuals generally showing reduced sensitivity to negative power to positive power.^51,52 Such asymmetries suggest that neural encoding of optical defocus and contrast (or image blur) is not uniform across refractive error groups and may interact with the physical contrast alterations induced by myopia control lenses. Future work integrating optical measurements, contrast sensitivity testing, and neural modeling may help bridge this gap and provide a more comprehensive understanding of contrast-based mechanisms in myopia control.

The limitations of the study include the following: (1) Small sample size. Only 15 participants were initially recruited, which limited the application of stratified analysis to various refraction groups, especially as peripheral refraction can significantly vary with central refractive error. (2) The contrast analysis for the DOT lens in the human eye was unable to be performed due to the limited dynamic range of the sensor. (3) As a cross-sectional study, we were unable to perform the correlation analysis between contrast reduction and myopia control efficacy. (4) The contrast evaluation was restricted in a relatively small field, making the comparison between the results from the best-focus image and RPR image indistinguishable, weakening the role of intraocular optics in contrast reduction. (5) In real-world viewing, peripheral rays enter the pupil tangentially after passing through the peripheral zones of the spectacle lens, whereas in our setup, the incident beam entered the lens orthogonally. This angular discrepancy may attenuate the contrast reduction effect, particularly at large eccentricities, as tangential incidence typically induces stronger oblique astigmatism, coma, and field curvature while simultaneously diminishing the effective optical defocus. In this study, we rotated the eye rather than the instrument because the device was not originally designed for peripheral measurements. Enforcing tangential entry would have substantially prolonged the experiment and increased the risk of mechanical damage. (6) Additionally, the present study does not report contrast versus spatial frequency characteristics across lens types, as our measurements reflect only broadband contrast. Readers interested in frequency-dependent contrast predictions may refer to the recent theoretical analysis by Charman and Atchison,³⁹ which provides complementary insights.

Future directions for contrast-related research include the following. First and most importantly, there is the development of a new instrument capable of simultaneously measuring a two-dimensional peripheral refraction map and analyzing image contrast. Such an instrument should not rely on a Hartmann–Shack wavefront sensor, as the latter acquires only wavefront phase information rather than the image intensity profile. A detailed comparison of contrast-related information obtained from a double-pass instrument and a wavefront sensor has been reported by Papadogiannis et al.⁵³ Second, there is a need to expand the sample size to comprehensively characterize contrast reduction patterns in different refractive groups across both PDT and DOT lens designs. Third, longitudinal studies are needed to quantify the relationship between contrast profiles and myopia control efficacy. Fourth, an improved sensor with an extended dynamic range should be incorporated to capture the extremely low-intensity retinal profiles produced by DOT lenses. Finally, although the primary aim of the present study is to compare contrast reduction across different lens designs, future work may also consider the influence of lens surface treatments, such as antireflective or blue light–filter coatings, which can affect contrast measurements, particularly in model eye setups.

In conclusion, this study presents a novel analysis of image contrast across four myopia control spectacles, including DIMS, MyoCare, Stellest, and DOT, measured in both human and artificial eyes. The findings demonstrate that lens optics plays a dominant role in determining peripheral image contrast within the central 40° (20° on each side), with Stellest producing the strongest reduction among peripheral defocus–based designs. Variations between human and artificial eye results highlight the influence of intrinsic ocular optics, suggesting that personalized lens designs may be beneficial when targeting retinal image contrast reduction. These findings contribute to a deeper understanding of the role of retinal image contrast in myopia progression and shed light on potential shared mechanisms between PDT and DOT lenses when contrast reduction is considered a primary factor.