Crystalline Lens Tilts Less With Higher Myopia in Children With Limited Age-Related Effect

Abstract

Purpose

To evaluate tilted crystalline lens morphology in children across different orientations and their correlations with age and spherical equivalent (SE).

Methods

This cross-sectional study encompassed 966 eyes of 493 children (5–15 years; SE, −6.00 to +3.00 diopters [D]). The binocular crystalline lens biometrics in 16 orientations were measured using swept-source anterior segment optical coherence tomography. Parameters included the anterior radius of the lens (LAR), posterior radius of the lens (LPR), thickness, and diameter, as well as lens tilt and decentration. Group comparisons were performed among non-, low-, and moderate-myopic eyes.

Results

Lens tilt and decentration demonstrated semi-periodic variations (R² = 0.984 and R² = 0.789, respectively). Lenses predominantly tilted temporal–inferiorly, with corresponding temporal decentration. The non-myopia group had significantly higher LAR and LPR (more curved) than the low- and moderate-myopia groups (all P < 0.001). As myopia increased, lens tilt consistently decreased across all refractive subgroups (all P < 0.001), whereas other parameters either reached a turning point at approximately −4.00 D or showed no linearity. Age-stratified polynomial analysis revealed that the positive correlation between lens tilt and SE remained strong within the range of −4.18 D to +3.00 D.

Conclusions

Although the lens maintains a greater tilt toward the temporal and inferior regions, it exhibits less tilt and becomes more orthotopic in eyes with higher myopia across different age groups in children.

Translational Relevance

This study established crystalline lens tilt as a robust, less age-dependent biomarker that could enhance the prediction of myopia progression and facilitate individualized risk assessment in pediatric refractive development.

Introduction

The prevalence of myopia has risen dramatically worldwide, with particularly high rates (80%–90%) observed among children in East and Southeast Asia, resulting in a greater psychological burden, reduced societal well-being, and increased risks and economic costs.^1–3 Although the myopia epidemic is driven by complex genetic and environmental factors, its core mechanism is a disruption of refractive equilibrium. In this imbalanced developmental process, the crystalline lens plays a pivotal role.^4,5

As the most dynamically active optical component within the eye, the crystalline lens assumes a crucial role in modulating refractive development to maintain emmetropization, particularly during childhood.^5–8 To compensate for continuous axial elongation, the lens undergoes a series of precise and dynamic remodeling processes, such as thinning and flattening, to actively reduce its refractive diopter.^8,9 This enables the maintenance of a sharp focus on the retina. The developmental patterns of these dimensional parameters, including lens thickness and radius of curvature, are relatively well understood.^4,10–12

However, the crystalline lens is not simply a lens with adjustable diopter; rather, it is a three-dimensional structure that occupies a specific location within the eye. Its spatial orientation, especially lens tilt, has generally been disregarded as a crucial spatial parameter, particularly in pediatric populations.⁹ The tilted refractive medium can modify its effective refractive diopter and induce higher-order aberrations such as coma, thus affecting the optical quality.^13–15 More significantly, lens tilt may serve as a potential indicator of axial elongation, which could further mirror the structural remodeling and biomechanical stresses that the eye experiences during rapid growth phases.¹⁶ Although it is recognized that the morphological parameters of the lens undergo comprehensive changes during myopic development, it remains uncertain whether its spatial parameters, such as tilt, follow the same developmental trajectories or display a distinct association with myopia progression.⁹

With the advancement of high-precision imaging modalities, such as swept-source optical coherence tomography (SS-OCT), quantitative analysis of the three-dimensional position of the pediatric crystalline lens has become achievable.¹⁷ Nevertheless, the existing literature mainly concentrates on dimensional parameters, such as the radius of curvature and thickness of the lens, whereas its spatial parameters have attracted relatively limited attention.⁹ As a result, few studies have systematically quantified the symmetry and orientational preferences of lens tilt in substantial cohorts of myopic children, or their correlation with refractive error and progression patterns.

This study aimed to characterize the spatial features of the crystalline lens in a large pediatric sample and to investigate the quantitative relationship between lens tilt and refractive parameters, including spherical equivalent (SE) and axial length (AL). This study also assessed the influence of age on this relationship, thereby elucidating the distinctive role and potential clinical predictive value of lens tilt in refractive development and myopia progression in children.

Methods

Participants

This cross-sectional study adhered to the tenets of the Declaration of Helsinki, and ethical approval was approved by the Institutional Review Board/Ethics Committee of Sun Yat-sen University (Guangzhou, China). This study recruited children who visited the Zhongshan Ophthalmic center between June and August 2021. These participants were acquired for initial screening or subsequent follow-up. Participants were recruited based on the following inclusion criteria: (1) children 5 to 15 years of age with a cycloplegic SE refraction ranging from +3.00 to −6.00 diopters (D) in both eyes; (2) astigmatism of 2.00 D or less; (3) best-corrected visual acuity (BCVA) of 0.0 logMAR or better in both eyes, and (4) a history of being prescribed single-vision corrective spectacles for refractive errors.

The exclusion criteria included (1) anisometropia, defined as an interocular difference in SE of 2.00 D or more; (2) prior or current use of myopia control interventions, including atropine, orthokeratology, or defocus-incorporating spectacles; (3) any history of ocular trauma or intraocular surgery; (4) clinical abnormalities on slit-lamp examination and intraocular pressure exceeding 21 mm Hg in either eye; (5) any medical history of ocular conditions other than refractive errors, such as amblyopia or strabismus, as well as diagnosed anisometropia or preexisting systemic diseases; and (6) inability to cooperate with study procedures or unsuitability for participation at the investigators’ discretion. Based on the SE after cycloplegia, we categorized the participants’ eyes into non-myopia (−0.50 D ≤ SE ≤ +3.00 D, including both emmetropes and hyperopes), low-myopia (−3.00 D ≤ SE ≤ −0.75 D), and moderate-myopia (−6.00 D ≤ SE ≤ −3.25 D) groups.

Ocular Biometric Evaluation

Before cycloplegia induction, anterior segment parameters were obtained using swept-source anterior segment OCT (CASIA 2; Tomey, Nagoya, Japan) in a dimly lit room to enable enough pupillary dilation for at least 5 minutes to minimize the potential influence of accommodation. The CASIA 2 system performed automated scans across 16 orientations (e.g., initial scan at 0°–180° [horizontal orientation], followed by 11.25°–191.25°, with an 11.25° step interval [180°/16], as shown in Supplementary Fig. S1). The sign for lens tilt and decentration is determined by a CASIA 2 built-in algorithm. The corneal reflection line serves as the zero reference. In each cross-sectional B-scan, the angle between the lens tilt line and the reflection line is defined as negative when the tilt is directed toward the left side. For example, along the 11.25° to 191.25° scanning meridian, the temporal side appears on the left; because the lens tilts temporally, the reported tilt value is negative. Conversely, along the 168.75° to 348.75° meridian, the temporal side is on the right, so a temporal tilt yields a positive value. Thus, the observed sign reversals between adjacent meridians are a computational artifact of the measurement system rather than a true biological reversal of lens orientation. The same sign convention applies to lens decentration. The measured parameters included anterior radius of the lens (LAR), posterior radius of the lens (LPR), lens thickness (LT), lens diameter (LD), lens tilt, and decentration.

Refractive error (sphere and cylinder) was assessed via autorefraction under cycloplegia, which was achieved by administering 0.5% tropicamide eye drops five times at 5-minute intervals followed by a 30-minute wait for full pupil dilation as indicated by no significant light reflection. AL was measured using the ZEISS IOLMaster 500 (Carl Zeiss Microscopy, Jena, Germany).

Model Fitting

Given the sinusoidal variation of crystalline lens parameters across the 16 meridians, a sine function was fitted using R 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria):

$$\begin{array}{c}\hfill Y=A\times sin\left(2X+\phi \right)+C\end{array}$$

or

$$\begin{array}{c}\hfill Y=A\times sin\left(X+\phi \right)+C\end{array}$$

where Y is a biometric parameter, and X is the orientation angle. The coefficient of X determines periodicity: 2 for anterior/posterior curvature radii, thickness, and diameter (one peak per full 180° rotation scanning), and 1 for tilt and decentration (one peak per full 360° rotation). A is amplitude (peak deviation from mean), φ represents phase shift (orientation of maximum/minimum values), and C is mean value of the parameter. The goodness of fit was assessed using the coefficient of determination (R²).

Statistical Analysis

Binocular data were included to describe the morphological symmetry and orientational periodicity. To avoid any binocular interaction effect and high inner correlation, only right-eye metrics were chosen for intragroup comparison and correlation analysis. All statistical analyses and graph creation were performed using R 4.3.2. The results are represented as mean ± SD (range). The normality of the data was evaluated using the Shapiro–Wilk test and confirmed through a homogeneity of variance test. Both independent and paired-sample t-tests were conducted to compare the lens parameters in contiguous orientations to show continuous alterations. Interclass comparisons among the three refractive groups were conducted by one-way analysis of variance (ANOVA), with pairwise post hoc t-test comparisons and Bonferroni or Tamhane’s correction for equal or unequal sample sizes. The power calculation of ANOVA for the right-eye comparison was 0.999 (with an effect size of 0.25, α of 0.05, and a sample size of 483 divided into three subgroups) using G*Power 3.1.9.7. Categorical variables were compared using the χ² test. Pearson's correlation coefficient and multiple linear regression analysis were applied to examine the correlations between lenticular and ocular parameters. A polynomial fitting model was further used to analyze the turning point of the correlation.⁹ P < 0.05 was considered statistically significant.

Results

Demographic and Binocular Metrics

A total of 966 eyes were recruited in this study (946 binocular metrics from 473 children, 13 right and seven left monocular metrics from 20 children). Binocular metrics were employed in the symmetrical and sinewave fitting model, and monocular data were utilized for group comparisons and relationship analysis. Baseline metrics, including age, gender, SE, and AL, are shown in Table 1. Based on refractive error subgroup classification, no discrepancy in gender was observed (all P > 0.05), but significant differences in age were observed (all P < 0.001).

Table 1.

Baseline Metrics of Participants

					P
Metric	Non-Myopia (N = 324)	Low Myopia (N = 409)	Moderate Myopia (N = 232)	P (ANOVA)	1 vs. 2	1 vs. 3	2 vs. 3
OD:OS, n	159:165	203:207	124:108	—	0.161a	0.663a	0.416a
Male:female, n (% male)	182:148 (56.2)	209:201 (51.0)	126:106 (54.3)	—	0.906a	0.309a	0.338a
Age (y), mean ± SD (range)	8.12 ± 2.27 (5 to 15)	10.03 ± 2.31 (5 to 15)	10.78 ± 2.37 (6 to 15)	<0.001	<0.001 b	<0.001 b	<0.001 b
SE (D), mean ± SD (range)	0.77 ± 0.88 (−0.5 to 3)	−1.7 ± 0.63 (−3 to −0.75)	−4.23 ± 0.83 (−6 to 3.25)	<0.001	<0.001 b	<0.001 b	<0.001 b
AL (mm), mean ± SD (range)	22.91 ± 0.76 (21.32 to 25.74)	24.28 ± 0.8 (21.94 to 26.54)	25.24 ± 0.79 (23.46 to 27.59)	<0.001	<0.001 b	<0.001 b	<0.001 b

Bold indicates statistical significance.

^aχ² test (1, non-myopia; 2, low-myopia; 3, moderate-myopia).

^bOne-way ANOVA with post hoc Bonferroni correction.

Orientational Characteristics of the Crystalline Lens Tilt Metrics

Figure 1 shows 16 orientation crystalline lens tilt metrics derived by CASIA 2, including lens tilt, and decentration. The box plot shows the binocularly symmetric and half-periodic fluctuation of crystalline lens tilt and decentration (Figs. 1A, 1C). To further present the continuous alteration, a standard sinewave function was used to describe this correlation and showed a good fitting index (R² = 0.984 and R² = 0.789, respectively). This symmetry was generally due to the definition of intraocular metrics given by the CASIA 2 techniques. A planar scatterplot of crystalline lens tilt metrics and axial distributions are shown in Figures 1B and 1D. The crystalline lens tilt showed significant temporal and inferior preference (192.54° ± 15.92° for right eyes), and decentration showed temporal and superior preference.

Figure 1.

Binocular crystalline lens tilt metrics in 16 orientations. (A, C) Cross-section images and metrics of the crystalline lens were derived by CASIA 2 using a two-dimensional lens analysis model that included lens tilt and decentration. Each orientation represents 11.25°. Statistical significance is indicated by asterisks (*) for independent-sample t-tests or pound signs (^#) for paired-sample t-tests: */^#P < 0.05, **/^##P < 0.01, ***/^###P < 0.001; ns, no statistical significance. (B, D) Orientation‑dependent metrics are shown as polar plots. The radial distance from the center represents the magnitude of the metric (tilt in degrees or decentration in millimeters), and the angular coordinate indicates the meridian (0° = temporal side). For the right eye (OD), the radial scale is inverted; the center corresponds to −10 (maximum absolute value) and decreases to 0 at the periphery. For the left eye (OS), the center is 0 and increases radially to 10 at the periphery. This difference in scaling was determined by a CASIA 2 built-in algorithm.

The fluctuation of other morphological metrics was also observed when comparing the contiguous relatively oblique orientation (i.e., for LAR, 33.75°–67.50° and 112.50°–146.25° in right eyes; all P < 0.05, one-sample Student's t-test) (indicated by an asterisk in Supplementary Fig. S2A). This was also shown in another statistical analysis (for contiguous orientation, except for the comparison between 168.75° and 180.00° in right eyes, P < 0.001, paired-sample t-test) (indicated by a pound sign in Supplementary Fig. S2A). For LT, LD, and LPR, no fluctuation was observed (Supplementary Figs. S2B–S2D).

Factors on Crystalline Lens Morphology and Tilt Metrics

No significant binocular discrepancy was observed in the lenticular metrics, except for lens tilt and steep value of LPR (P = 0.014 and P = 0.024, respectively) (Supplementary Table S1). The non-myopia group showed lower values in LAR (Fig. 2A), LPR (Fig. 2B), and LD (Fig. 2D) but higher values in LT (Fig. 2C, within panels), compared with the low- and moderate-myopia groups (all P < 0.001). To avoid any binocular interaction effect, only right-eye metrics were chosen for further analysis. The higher myopia group showed contiguous reduction in crystalline lens tilt (all P < 0.05) (Fig. 2E) but no differences in lens decentration (all P > 0.05) (Fig. 2F). This continuous alteration was not observed in lenticular morphological metrics; however, no significant discrepancy in these metrics was observed when comparing the moderate-myopia and low-myopia groups (all P > 0.05). Detailed comparisons of right eyes are shown in Supplementary Table S2.

Figure 2.

Association between crystalline lens morphology and tilt metrics and refractive errors. Shown is the linear regression and polynomial fitting analysis between crystalline lens morphology and tilt metrics and refractive errors. The linear analysis with 95% confidence intervals is shown with standard error bars, and the colors of the dashed regions indicate the group classification. The light purple line and the dark dashed curve indicate the linear and polynomial associations, respectively, for all participants. The perpendicular dashed line indicates the inflection point; the x-axis location is light purple. The panels within the violin boxes represent the distribution and intragroup comparisons of the metrics. Statistical significance is indicated for one-way ANOVA with Bonferroni correction: *P < 0.05, **P < 0.01, ***P < 0.001; ns, no statistical significance.

Both linear and polynomial fitting were utilized to show the association among crystalline lens tilt, morphological metrics, and SE (Fig. 2). Lens tilt showed a significant linear association with SE (all P < 0.001) but decentration did not (all P > 0.05). LAR (Fig. 2A). LPR (Fig. 2B), LT (Fig. 2C), and LD (Fig. 2D) showed general linear associations with SE for all participants, but these were partially diminished when solely analyzing the moderate-myopia group (all P > 0.05, except for LT, P = 0.032, and LAR steep, P = 0.020). Considering this inflected phenomenon, a polynomial fitting model was utilized to represent the association.⁹ The inflection point was located closely to about −4.00 D but varied in different metrics, as shown in each panel in Figure 2 with straight dashed lines.

Table 2 shows both the univariable and multiple linear regression of lens tilt and morphological metrics. Lens tilt showed significant positive correlation with age but negative correlation with SE (Pearson's r = 0.284 and r = 0.377, respectively; both P < 0.001). Specific metrics showed gender-based relationship (i.e., males showed lower LAR flat and LPR mean and LPR steep values: P = 0.044, P = 0.035, and P = 0.004, respectively). The crystalline lens decentration showed no association with age, SE, or AL (all P > 0.05).

Table 2.

Crystalline Lens Morphological Metrics Derived by CASIA 2

	Univariable Linear Regression Analysis								Multiple Linear Regression Analysis
	Age		Sex		SE		AL		Age		Sex		SE
Metric	r a	P	r	P	r	P	r	P	B b	P	B	P	B	P
LAR mean (mm)	0.371	0.000	−0.078	0.091	−0.546	0.000	0.649	0.000	0.156	0.000	—	—	−0.474	0.000
LAR steep (mm)	0.411	0.000	−0.054	0.238	−0.528	0.000	0.627	0.000	0.217	0.000	—	—	−0.430	0.000
LAR flat (mm)	0.318	0.000	−0.092	0.044	−0.531	0.000	0.630	0.000	0.107	0.014	−0.094	0.016	−0.481	0.016
LPR mean (mm)	0.331	0.000	−0.096	0.035	−0.393	0.000	0.520	0.000	0.203	0.000	−0.107	0.010	−0.300	0.000
LPR steep (mm)	0.290	0.000	−0.129	0.004	−0.350	0.000	0.501	0.000	0.178	0.000	−0.137	0.001	−0.267	0.000
LPR flat (mm)	0.325	0.000	−0.056	0.222	−0.381	0.000	0.472	0.000	0.192	0.000	—	—	−0.295	0.000
LT (mm)	−0.292	0.000	0.077	0.090	0.439	0.000	−0.488	0.000	−0.118	0.010	—	—	0.384	0.000
LD (mm)	0.373	0.000	−0.090	0.047	−0.434	0.000	0.594	0.000	0.233	0.000	−0.106	0.009	0.329	0.000
Tilt (°)	−0.284	0.000	0.005	0.909	0.417	0.000	−0.377	0.000	−0.120	0.010	—	—	0.361	0.000
Decentration (mm)	0.037	0.413	0.095	0.037	−0.034	0.455	0.039	0.392	—	—	0.102	0.025	—	—

Bold indicates statistical significance.

^aPearson's correlation coefficient.

^bEstimated by multiple linear regression analysis.

Age-Stratified Factors on Lens Tilt

Interestingly, the association between AL or SE and lens tilt demonstrated consistency. Considering the potential effect of age on crystalline lens morphology and refractive development, age-stratified analysis of lens tilt was further conducted (Fig. 3). In the recruited participants (5–15 years of age), both tilt and SE showed negative correlation with age, indicating a higher myopic eye and less tilted lens in older children (Figs. 3A, 3B). Less tilted lenses were observed in elongated and higher myopic eyes (Figs. 3C1, 3D1). The age of the recruited children was stratified into subgroups with 2- or 3-year intervals to ensure relatively comparable sample sizes across groups (Figs. 3C2, 3D2). Inflection points of the polynomial fitting model were still observed in association with SE (−4.18 D and R² = 0.157 for 5–7 years; −5.69 D, R² = 0.166 for 13–15 years) and AL (22.08 mm and R² = 0.050 for 5–7 years; 21.27 mm and R² = 0.129 for 13–15 years). However, for the age groups of 8 to 9 years and 10 to 12 years, abnormal inflection points and fitting coefficients were observed. Nevertheless, a linear model can still represent this association with SE (slope = 0.162°/D, P = 0.002 for 8–9 years; slope = 0.252°/D, P < 0.001 for 10–12 years) or AL (slope = −0.339°/mm, P < 0.001 for 8–9 years; slope = −0.379°/mm, P < 0.001 for 10–12 years; not shown).

Figure 3.

Effect of age-stratified axial length and refractive errors on crystalline lens tilt. (A, C1, D1) Linear regression and polynomial fitting analysis between factors and crystalline lens tilt. The red line and 95% confidence interval region show the linear association, and the dark blue dashed curve shows the polynomial fitting association. (B) Linear regression and polynomial fitting analysis between age and SE. (C2, D2) Age-stratified polynomial fitting analysis of effect of AL or SE on lens tilt. The perpendicular line indicates the inflection point with x-axis location; R² and number of eyes are indicated by color.

Discussion

In the present study, we observed that lens tilt and decentration showed binocular symmetry and distinct directional preferences, in agreement with previous studies. Lens tilt exhibited a linear correlation with SE, which decreased proportionally with increasing myopia degree, whereas lens decentration did not exhibit significant correlations with age, SE, or AL. Age-stratified analysis of lens-tilt factors further emphasized the relatively independent effect of SE on lens tilt.

In the horizontal direction (from nasal to temporal), the lens was predominantly tilted toward the temporal side relative to the reference corneal reflection point captured by CASIA 2, with both lens tilt and decentration showing bilateral symmetry, consistent with previous research findings.^16,18,19 Lenses in this cohort were tilted temporally; however, the literature shows considerable variation in the precise orientation (temporal/inferior).^16,20 Furthermore, data specifically from pediatric cohorts remain limited. Chen et al.²⁰ reported that the average cataractous lens tilt, with respect to the corneal topography axis, was 5.16° inferiorly and 0.22 mm temporally. Lu et al.,¹⁶ however, observed a predominant tilt of the lens toward the superior temporal quadrant in relation to the visual axis, which was partially due to the selection of the reference axis. The optical axis (which coincides with the corneal topographic axis) is located temporally to the visual axis, a configuration resulting from the temporal and slightly inferior displacement of the fovea relative to the point where the optical axis intersects the retina.^21,22 When the visual axis is used as a reference, it yields measurements suggesting an inferiorly and temporally tilted cornea and a relatively greater lens tilt, given that the lens naturally tilts in similar directions, as observed in this study and that of Chen et al.¹⁶

Numerous studies have identified a critical developmental stage during childhood when the lens undergoes an inflected transition and starts to thicken, and lens diopter reduction is slower.^11,23,24 This compensatory relationship between lenticular morphology/diopter and refractive status follows a non-linear, biphasic developmental pattern, marked by critical inflection points.⁴ Polynomial regression analysis further revealed these inflection points in correlations among certain lens parameters (e.g., LAR, LPR, LT, LD) and SE, occurring within the range between −5.05 D and −3.76 D. This observation suggests a potential shift in lens morphological adaptation or response patterns during moderate myopia, consistent with findings from prior studies.^9,25 Cheng et al.⁹ observed a decoupling of lens power (LP) and AL in eyes > 27 mm among young individuals with high myopia. Similarly, a longitudinal study by Rozema et al.²⁵ involving children in Singapore demonstrated that individuals developing myopia exhibited accelerated axial growth and loss of LP before myopia onset, followed by a rapid deceleration in LP loss after onset. The LP was significantly influenced by lens morphological (thickness and curvature) and tilt parameters, indicating an inflection point in LP changes during the early stages of myopia.⁹ Moreover, variations in curvature radius can impede, halt, or even reverse the steepening process.^17,24,26 Overall, despite the non-linear change in LT, all morphological metrics demonstrated inflection points, with the most flattened lenses observed in the moderate myopia range, indicating the presence of multiple critical transitions in lens compensatory mechanisms during myopia progression.

However, this positive correlation between tilt and SE did not exhibit an inflection point (Figs. 2E, 3). In children, this observed correlation was less affected by age stratification. Age-stratified analysis revealed a decline in tilt and SE with increasing age among children ages 5 to 15 years. Hu et al.²⁷ utilized Scheimpflug imaging in a cohort with a median age of 13 years (range, 4–53 years) and observed a decreasing trend in lens tilt with advancing age. Conversely, Shen et al.²⁸ evaluated 252 patients with an average age of 43.63 ± 17.02 years, and they reported an elevation in lens tilt with age. The association between lens tilt and age lacks consensus, suggesting that diverse age brackets and ocular health conditions may exhibit distinct patterns.

In this study, we primarily concentrated on the lens tilt of children within a relatively narrow age range as compared to previous reports involving all age ranges, and we found a general positive correlation between lens tilt and SE.^16,26,28 The modest positive changes in horizontal (temporal preference) lens tilt can influence higher-order aberrations, particularly horizontal coma and spherical aberration, compensated by oblique light incidence on the eye.^29–31 Moreover, previous research has demonstrated a positive association between lens power and lens tilt.²⁶ In our study, more myopic eyes exhibited less lens tilt and likely lower lens power, which might possibly help maintain focused retinal images despite axial elongation. Thus, one potential explanation is that reduced lens tilt in more myopic eyes may reflect a compensatory mechanism, possibly serving to diminish aberrations and adjust lens power in a way that helps optimize visual quality during myopia progression. The observed decrease in lens tilt may also be explained by a reduced biomechanical shift and a less constrained anterior segment, created by axial growth and zonular weakness in myopic eyes.^28,32–34 Clarifying this mechanism warrants future longitudinal aberration- and biomechanics-based research. Nonetheless, it can be posited that tilt may serve as a more predictive factor than other lens parameters, considering the significant effect of age and refractive status on lens biometrics, such as lens thickness. With accessible optical coherence elastography and Brillouin biomechanics, future studies could focus on investigating zonular-related lens elasticity and biomechanics in myopic children, as well as their potential association with tilted lens formation.^35,36

This study has several limitations. First, it is a cross-sectional analysis involving myopic children. To fully understand the predictive capacity of lens tilt, longitudinal investigations in children and adolescents are imperative to delineate the developmental trajectory of lens tilt (both in terms of magnitude and orientation) and its association with emmetropia and refractive error progression. Second, the lens is a non-uniform optical entity, with the refractive index and biomechanics varying from the center to the peripheral region.^37,38 In addition to external morphological factors, future research should explore the impact of this heterogeneity on inflection points and periodic fluctuations. Third, this study primarily used SE as the refractive measure for lens tilt analyses, which may not fully capture the meridional interactions between tilt and astigmatism. Future studies should further explore vector‑based astigmatic components. Finally, the precise biological or clinical relevance of the inflection point warrants validation in longitudinal or observational cohorts.

The anterior surface radius of the lens fluctuates, and refractive error–related lens tilt persists throughout refractive progression, despite the complex morphological changes in the crystalline lens. Significantly, lens tilt can be regarded as a metric that is less dependent on age and could effectively depict refractive status among lenticular parameters.