High myopia has emerged as a growing public-health concern because of its rising prevalence and the disproportionate burden of vision-threatening complications that follow excessive axial elongation. Population-level estimates and trend analyses demonstrate a steep recent increase in myopia and a substantive rise in high myopia among children and young adults, with projections indicating continued growth in affected cohorts worldwide. The epidemiological expansion of high myopia translates to larger absolute numbers of eyes at risk for chorioretinal degeneration, with attendant increases in disability, healthcare utilization, and socioeconomic impact (1, 2).

The clinical spectrum of chorioretinal damage in highly myopic eyes encompasses a range of atrophic, tractional, and neovascular phenotypes that often coexist and drive progressive visual loss (3). Atrophic changes include tessellated fundus, diffuse and patchy chorioretinal atrophy, and macular atrophy; tractional entities include myopic traction maculopathy with foveoschisis and macular detachment; neovascular complications include myopic choroidal neovascularization that may produce fibrovascular scarring. Contemporary classification frameworks, such as the atrophy-traction-neovascularization (ATN) system and related staging schemes, permit standardized description of disease subtypes and severity and facilitate longitudinal assessment of progression and outcomes. Comprehensive clinical description must integrate multimodal imaging findings from fundus photography, optical coherence tomography, and angiography to capture the full phenotypic complexity (4, 5).

Risk predictors for chorioretinal damage in high myopia include biometric, demographic, and structural imaging markers that modulate susceptibility and temporal course. Longer axial length and greater myopic refractive error are consistently associated with higher risk and more severe maculopathy, and axial elongation rate predicts earlier onset and faster progression in younger cohorts. Age and baseline severity of myopic maculopathy also stratify progression risk. Emerging quantitative markers from OCT and vascular imaging, such as choroidal thinning, reduced choriocapillaris density, and alterations in deep capillary plexus metrics, show promise as mechanistic biomarkers that correlate with atrophy and tractional changes. Despite these insights, many studies remain cross-sectional, focused on selected clinical samples, or limited to single imaging modalities, constraining integrated risk modeling across populations (6, 7).

Recent work in systemic biomarker discovery illustrates how routine inflammatory indicators can stratify disease susceptibility, supporting the rationale for identifying objective predictors in high-risk ocular conditions (8, 9). Advances in ocular immunology and imaging, including detailed characterization of immune-mediated uveitis and the development of high-resolution microvascular and wavefront assessment tools, highlight how modern analytical approaches can refine detection of early structural changes in eyes vulnerable to chorioretinal injury (10–13).

The main objective of the research was to describe the clinical spectrum and rates of ATN-detected chorioretinal losses in a high myopic cohort of a tertiary eye care unit, and to estimate its connections with clinical and imaging variables routinely assessed. Instead of attempting to identify completely new biomarkers, this study itself was meant to reaffirm and generalize existing findings about high myopia, axial elongation, and chorioretinal pathology in an unrepresentatively representative group, and to translate it into a risk-stratification model that can be understood clinically in the form of standardized ATN grading and multimodal imaging. This work contributes novelty by combining: (1) a large retrospective sample spanning pediatric and adult age ranges to capture age-dependent patterns; (2) systematic application of an established classification to enable phenotype-specific risk modeling; and (3) integration of axial length, refractive error, OCT-derived structural metrics, and clinical variables into multivariable predictive models. The manuscript addresses a gap in the literature where prior reports have been limited by small cohorts, restricted age bands, or single-modality assessments, and where population-level retrospective analyses that relate standardized phenotypes to measurable predictors are scarce. The primary objective is to quantify the association strength of candidate predictors with discrete chorioretinal outcomes and to produce clinically applicable predictors that inform screening and follow-up strategies for eyes at highest risk of irreversible visual loss.

The substantial size of this cohort (n = 1,420) strengthens the statistical confidence in observed demographic and clinical trends. The mean age (47.8 years, range 18–78) and the predominance of women (56.1%) are in line with global high-myopia populations, where middle-aged adults form a large proportion of those needing subspecialty care. Both the mean spherical equivalent (−11.9 D) and mean axial length (28.4 mm) highlight a particularly severe degree of myopia, exceeding average levels found in other large-scale studies. Comorbid systemic diseases, especially hypertension and diabetes, were prevalent and may interact with ocular pathology, amplifying the risk of chorioretinal complications via vascular and metabolic mechanisms, underscoring the need for comprehensive interdisciplinary screening in high-risk patients (17, 18).

The high prevalence of chorioretinal lesions in this cohort mirrors the global burden in highly myopic populations. Notably, myopic macular degeneration affected 35.1%, lacquer cracks 18.2%, tessellation 42.5%, and posterior staphyloma 29.3%. The elevated rates of these pathologies emphasize the structural vulnerability associated with axial elongation, as also documented in cross-sectional, population-based studies. The deployment of multimodal imaging for lesion verification ensures diagnostic accuracy and aligns with expert recommendations for managing high myopia complications. These findings are consistent with those from East Asia, yet the precise rates can be influenced by ethnicity, recruitment source, and age distribution. Together, they stress the critical need for routine retinal assessment in this at-risk demographic (18–20).

The imaging-based prevalence rates highlight the heterogeneous spectrum of chorioretinal pathology in eyes with high myopia. The substantial proportion of eyes exhibiting diffuse (26.1%) and patchy (16.1%) atrophy underscores the frequent occurrence of structural retinal changes. The identification of foveoschisis (11.9%) and tractional maculopathy (13.8%) aligns with recognized complications tied to progressive axial elongation and biomechanical stress in myopic degeneration. Myopic CNV, though less common (6.0%), remains clinically significant due to its vision-threatening potential. Notably, 8.4% of eyes demonstrated overlapping phenotypes, reflecting the complex interplay of degenerative and tractional processes. These findings reinforce the necessity of multimodal imaging for precise phenotypic classification and inform targeted surveillance strategies for high-risk individuals.

Eyes with visible structural pathology displayed a profile of greater age, longer axial length, and more severe refractive errors vs. non-pathological eyes a relationship that has been repeatedly established. Furthermore, systemic vascular comorbidities such as hypertension and diabetes occurred more frequently among those with ocular pathology, suggesting an additive or possibly synergistic role in disease progression. OCT-based evaluation revealed marked choroidal thinning, a recognized harbinger of advanced retinal disease in myopia. The statistical significance of these differences (e.g., axial length, spherical equivalent, and age with all p < 0.01) further supports active monitoring of older, more myopic patients particularly those with other systemic health challenges (17, 21, 22).

Multivariable logistic regression within the cohort identified axial length, spherical equivalent, hypertension, and choroidal thinning as robust independent predictors of chorioretinal pathology. Each increment in axial length and refractive error substantially increased risk, corroborating literature showing the linear and region-specific impact of these variables on choroidal structure and function. The observed AUC of 0.86 for the risk model demonstrates strong discriminatory power. Notably, systemic hypertension was an independent variable, hinting at mechanisms involving compromised choroidal perfusion. These findings provide a strong basis for integrating biometric and systemic variables into routine risk-stratification algorithms to refine surveillance and preventive strategies in high myopia (17, 21).

Stratification by axial length and age further refined the risk profile, highlighting a steep increase in pathologic lesion rates for eyes with axial length exceeding 30 mm and among patients over 50 years. This trend is echoed in robust cross-sectional analyses, confirming that both advancing age and progressive axial elongation independently predict increased chorioretinal morbidity. Sensitivity analyses, such as the exclusion of treated choroidal neovascularization cases, revealed minimal change in the regression output, underscoring the generalizability of these associations. As such, both biometric and demographic stratification procedures are invaluable for tailoring clinical vigilance and intervention thresholds (18, 21).

OCT imaging in the cohort documented progressive subfoveal choroidal thinning and increased outer retinal layer anomalies corresponding to increasing refractive severity. Quantitative correlation matrices confirmed strong inverse relationships between axial length and choroidal thickness and robust positive associations linking myopic severity to cumulative lesion burden. These findings are in direct agreement with meta-analyses and multi-ethnic population studies, highlighting that both structural and functional retinal metrics degrade as myopic progression worsens. Incorporating such imaging biomarkers into clinical risk models enhances the predictive value for future vision-threatening complications (17, 21).

Eyes harboring structural lesions experienced significantly poorer best-corrected visual acuity (BCVA) and incurred higher rates of clinical interventions and follow-up visits most commonly for anti-VEGF therapy in choroidal neovascularization cases. This elevated medical need is corroborated by long-term outcome studies, where the presence and severity of myopic macular degeneration portends a marked decline in BCVA, particularly with age progression (23, 24). These adverse clinical outcomes translate to increased demand for healthcare resources and underscore the importance of early identification and aggressive management of pathology to mitigate vision loss (19).

Population projections suggest that as many as one in three individuals with severe myopia could require intervention for chorioretinal disease over the next decade. Such forecasts are grounded in the increasing prevalence of high myopia and associated complications worldwide. The regional adaptation of age and axial length distribution data enhances local applicability and strengthens the rationale for targeted, high-frequency surveillance in at-risk groups. Integrating these insights into public health policies could substantially improve early detection, clinical outcomes, and resource allocation in regions facing a surge in myopia-related vision loss (19).

Supplementary Table S5 highlights distinct age-related trends in the prevalence of principal chorioretinal lesions among high-myopia eyes. Diffuse atrophy, patchy atrophy, foveoschisis, and myopic CNV all show a progressive increase in frequency with advancing age. Eyes from patients under 18 years exhibit the lowest rates of all lesion types, while those aged 60 years or older display the highest burdens, most notably, diffuse atrophy (27.5%), patchy atrophy (14.2%), foveoschisis (18.9%), and myopic CNV (7.1%). This age gradation suggests cumulative risk for structural retinal damage as myopia duration and severity intensify over time, reinforcing the importance of early monitoring and targeted intervention for older individuals at greatest risk of vision-threatening complications (24–26).

The supplementary analyses further strengthen the validity and clinical relevance of this study. Performance metrics (Supplementary Table S6) support the robustness of the multivariable logistic regression model, with an area under the ROC curve (AUC) of 0.86 indicating excellent capacity to differentiate between eyes with and without chorioretinal lesions (27). The acceptable Brier score and calibration statistics particularly the Hosmer–Lemeshow p-value and a near-unity slope affirm model reliability and minimal risk of overfitting. Supplementary Table S7 reveals that although combined lesion patterns in high myopia are relatively uncommon, notable proportions of eyes exhibit dual or triple pathologies. Patchy atrophy plus foveoschisis or diffuse and patchy atrophy affect 2.0%−2.7% of eyes, highlighting the complexity of disease presentation (27, 28).

The prediction model showed good internal discrimination and acceptable calibration, suggesting that these routinely measured variables can be used to stratify eyes by their likelihood of having chorioretinal damage; however, external validation in independent cohorts remains necessary before broader implementation. The calibration slope close to 1.0 reflects apparent calibration estimated on the full dataset and is likely somewhat optimistic; although 5-fold cross-validation partly addresses this, more rigorous optimism-corrected approaches (e.g., bootstrap-based shrinkage) and external validation will be important in future work. Future ML models, e.g., random forests or neural networks, may improve AUC beyond 0.86). However, traditional models are prioritized in the cross-sectional design and focus on interpretable logistic regression for clinical translation (29, 30).

The sensitivity analysis shown in Supplementary Table S8 confirms that exclusion of treated CNV does not substantially alter the regression findings; all key predictors retain statistical significance, and the direction of effect remains stable. This demonstrates the model's resilience to potential confounding effects from therapeutic intervention. Population projections in Supplementary Table S9 illustrate the substantial clinical burden that could arise with shifting prevalence rates. Even with conservative assumptions, tens of thousands of eyes could be impacted, while rapid growth scenarios in urban or high-risk populations may result in more than 49,000 affected individuals emphasizing the scale of the public health challenge (22, 25, 31).

Although age, axial length, refractive error, and reduced subfoveal choroidal thickness are established correlates of myopic chorioretinal pathology, this study contributes several clinically relevant additions. First, it applies the ATN classification to a large, real-world high-myopia cohort from a tertiary center in China, providing detailed prevalence estimates of specific ATN lesion patterns in an underrepresented population. Second, by integrating these routinely available parameters into a multivariable GEE logistic model with good discrimination (AUC 0.86), we translate known factors into a practical risk-stratification tool that reflects how they jointly relate to chorioretinal damage in everyday practice. Third, the quantified per-unit gradients (per-mm axial length and per-10 μm choroidal thickness) can help clinicians identify eyes at particularly high risk that may warrant closer surveillance and earlier imaging.

This study has a number of limitations in which the findings should be taken. This is a hospital, tertiary eye care cohort study, hence, is prone to selection biases including the bias of Berkson where more severe or symptomatic patients will present and be recruited. Therefore, the rate of chorioretinal lesions found in this cohort probably over represents the burden in high myopia population in general and the figures generated in Table 7 should be seen as estimates of the burden based in hospitals and not the actual numbers based in the population. We adhered to IMI guidelines by combining both SE and AL to define high myopia since both of them represent refractive status but AL represents structural axial elongation and thus both of them give a more complete picture of high-myopia phenotypes. Community sampling, which will involve a population-based sampling, will be necessary to have objective estimates of disease burden as well as to determine the generalizability of our findings to various definitions of high-myopia and care contexts.

High myopia is swiftly transforming from a personal refractive challenge into a global public-health emergency, with sight-threatening complications rising sharply in prevalence. This clinical study demonstrates a considerable population burden of chorioretinal damage among high-myopia eyes, most notably in diffuse atrophy, foveoschisis, and patchy atrophy. Multivariable analysis identified older age, greater axial length, more severe myopic refractive error, and reduced choroidal thickness as reliable predictors enabling early risk assessment and stratification. These results offer a concise epidemiological overview and practical groundwork for refining screening protocols. The findings underscore an urgent need for tailored surveillance and intervention programs within clinical and community settings. As myopia rates surge worldwide, these insights are instrumental for guiding public-health strategies aimed at preventing irreversible vision loss and improving long-term ocular outcomes.