This cross-sectional study was conducted between March and May 2024 across three distinct geographical regions of Shaanxi Province, China: Hanzhong (Southern Shaanxi), Xi'an (Guanzhong Plain), and Yulin (Northern Shaanxi). Using a multistage stratified random sampling approach, we first compiled a comprehensive list of all elementary schools within each region—433 in Southern Shaanxi, 1,169 in Guanzhong, and 312 in Northern Shaanxi. From these, 4 schools per region (totaling 12 schools) were randomly selected as primary sampling units. In each school, the three classes per level (Grades 1–6) were selected using a simple random sampling method from all available parallel classes at each grade level, ensuring representativeness while maintaining logistical feasibility. All enrolled students underwent standardized visual acuity testing and non-cycloplegic autorefraction assessments. Exclusion criteria included: (1) diagnosed ocular pathologies (amblyopia, strabismus, congenital glaucoma, or congenital cataracts); (2) history of ocular surgery; (3) incomplete examination data. After applying these exclusion criteria, the final analytic sample comprised 8,207 participants, distributed as follows: 2,724 from Southern Shaanxi, 2,761 from Guanzhong, and 2,722 from Northern Shaanxi. A flowchart showing the students affected by the various exclusion criteria is presented in Figure 1. A map illustrating the Shaanxi region, along with the approximate location of the schools involved can be found in Figure 2.

The study protocol was approved by the Medical Ethics Committee of Xi'an central hospital (Approval No. LW-2024-027) and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants' legal guardians prior to data collection.

Visual acuity assessment was performed by certified ophthalmologists using the standardized E-Chart logarithmic visual acuity system (GB 11533-2011). Participants were positioned at 5 meters from the chart under ambient illumination of 300–500 lux. Monocular measurements followed standardized protocols, with right eyes examined prior to left eyes, recording uncorrected visual acuity (UCVA) values.

Refractive measurements were conducted without cycloplegia using the Topcon KP-800 automated refractometer (Topcon Corporation, Tokyo, Japan), operated by licensed optometrists. Three consecutive measurements were obtained for each eye (right eye first) to determine spherical power, cylindrical power, axial orientation, and spherical equivalent. The final refractive values represented the arithmetic mean of triplicate measurements. All data were automatically transmitted via secure cloud-based interface to the Shaanxi Provincial Student Vision Surveillance Database.

To ensure data integrity, a multi-level quality control system was implemented: (1) Daily calibration of instruments using NIST-traceable reference standards; (2) Random duplicate testing of 5% participants with <5% inter-test discrepancy threshold for spherical equivalent (95% limits of agreement within ±0.25 D); (3) Automated outlier detection algorithms flagging measurements beyond ±3SD from cohort means for manual verification.

In addition to student vision assessments, a structured parental questionnaire was distributed to the legal guardians of all participants to assess knowledge, attitudes, and practices (KAP) related to myopia prevention. The questionnaire was developed based on validated KAP frameworks and reviewed by ophthalmology and public health experts for content validity. It included 15 items covering awareness of myopia risk factors, understanding of preventive strategies (e.g., outdoor activity, screen time management), and engagement with school-based vision programs. Questionnaires were distributed via school communication channels and collected within one week. A total of 7,894 completed responses (96.2% response rate) were included in the final analysis.

Myopia classification was established according to internationally recognized criteria (3–5). Screening myopia was characterized by a spherical equivalent (SE) ≤ −0.50 diopters (D) in at least one eye accompanied by uncorrected visual acuity <5.0 on standard logarithmic charts (3). The severity of refractive error was stratified as follows: low myopia (−3.00 D < SE ≤ −0.50 D), moderate myopia (−6.00 D < SE ≤ −3.00 D), and high myopia (SE ≤ −6.00 D) (4). Pre-myopia status was identified as eyes demonstrating borderline refractive values within the range of −0.50 D < SE ≤  + 0.75 D, in accordance with the WHO definition for individuals at risk of myopia development (5).

All statistical analyses were performed using SPSS (Version 27.0; IBM Corp.) and Microsoft Excel (Office 2019) software. Continuous variables with approximately normal distribution were expressed as mean ± standard deviation (SD), as confirmed by Shapiro–Wilk normality testing. Refractive error measurements demonstrated symmetrical interocular agreement across all age groups (Mann–Whitney U test, P > 0.05 for right vs. left eye spherical equivalent comparisons). Consequently, right-eye data were selected as representative values for subsequent analyses to maintain methodological consistency. Categorical variables were reported as frequency distributions with percentages. Intergroup comparisons of categorical outcomes were evaluated using Pearson's chi-square test. A two-tailed P value <0.05 was established as the threshold for statistical significance.

The cross-sectional study enrolled 8,207 primary school students stratified into three geographical regions: Southern Shaanxi (n = 2,724; 33.2%), Guanzhong (n = 2,761; 33.6%), and Northern Shaanxi (n = 2,722; 33.2%). The cohort had a mean age of 8.78 ± 1.52 years (range: 6–12 years), with balanced gender distribution across regions (χ² = 1.85, P = 0.39).

As detailed in Table 1, spherical equivalent (SE) progression demonstrated significant grade-level dependence in all regions. Southern Shaanxi exhibited the most gradual myopic shift, with mean SE values progressing from −0.43 ± 0.78 D in Grade 1 to −1.32 ± 1.28 D in Grade 6. The Guanzhong region showed accelerated refractive changes, advancing from −0.44 ± 0.90 D to −2.25 ± 1.95 D across the same grade span. Northern Shaanxi displayed intermediate progression rates, ranging from −0.38 ± 1.39 D (Grade 1) to −2.18 ± 2.22 D (Grade 6). Regional comparison of progression slopes revealed statistically significant differences (ANCOVA adjusted for age, F = 18.34, P < 0.05), with Guanzhong demonstrating the steepest annual myopic progression rate (−0.30 D/year, 95% CI: −0.34 to −0.26).

The age-standardized myopia prevalence demonstrated significant geographical variation across Shaanxi Province (χ² = 185.3, P < 0.05). Southern Shaanxi exhibited the lowest prevalence (30.43%; 95% CI: 28.7–32.2), followed by Guanzhong (42.96%; 95% CI: 41.1–44.8) and Northern Shaanxi (48.02%; 95% CI: 46.2–49.8). Grade-stratified analysis revealed a consistent positive correlation between educational progression and myopia risk (Southern χ² = 340.7, Guanzhong χ² = 346.2, Northern χ² = 271.4; all P < 0.05, Figure 3).

Low myopia (−3.00 D < SE ≤ −0.50 D) predominated across all regions, accounting for 70%–86% of cases. However, Northern Shaanxi demonstrated significantly higher age-adjusted prevalence of high myopia (SE ≤ −6.00 D) compared to other regions (4.06% vs. 2.53% in Guanzhong vs. 0.12% in Southern; χ² = 28.4, P < 0.05).

Consistent female predominance was observed across all regions:

Southern Shaanxi: 34.00% (F) vs. 26.91% (M); χ² = 16.1, P < 0.05

Guanzhong: 48.02% (F) vs. 37.99% (M); χ² = 28.3, P < 0.05

Northern Shaanxi: 52.54% (F) vs. 44.13% (M); χ² = 19.2, P < 0.05

The regional hierarchy (Northern Shaanxi > Guanzhong > Southern Shaanxi) persisted within gender subgroups (Male: χ² = 92.4; Female: χ² = 100.1; both P < 0.001), with Northern Shaanxi females exhibiting the highest overall prevalence (Tables 2–5).

The age-adjusted pre-myopia prevalence exhibited inverse geographical distribution compared to established myopia patterns (χ² = 185.3, P < 0.05). Southern Shaanxi demonstrated the highest prevalence (40.60%), followed by Guanzhong (34.19%) and Northern Shaanxi (33.73%). A significant negative correlation with grade progression was observed across all regions (Southern Shaanxi χ² = 242.3, Guanzhong χ² = 311.8, Northern Shaanxi χ² = 198.9; all P < 0.05) (Figure 4).

Consistent male predominance was identified in pre-myopia distribution:

Southern Shaanxi: 38.73% (F) vs. 42.45% (M); χ² = 3.9, P < 0.05

Guanzhong: 30.01% (F) vs. 38.28% (M); χ² = 21.0, P < 0.05

Northern Shaanxi: 29.17% (F) vs. 37.64% (M); χ² = 21.7, P < 0.05

The regional hierarchy (Southern Shaanxi > Guanzhong > Northern Shaanxi) remained significant within gender subgroups (Males: χ² = 8.0; Females: χ² = 33.9; all P < 0.05) (Tables 6–8).

Our study reveals significant regional disparities in myopia prevalence among primary school students across three distinct geographical divisions of Shaanxi Province: Southern Shaanxi, Guanzhong, and Northern Shaanxi. The hierarchical pattern of myopia prevalence—Northern Shaanxi > Guanzhong > Southern Shaanxi—persists even when controlling for grade-level variations, with Northern regions additionally demonstrating greater severity of refractive errors.

This geographical pattern likely reflects the complex interplay between educational stressors and preventive interventions. Students in Guanzhong reported significantly longer near-work duration and higher participation in after-school tutoring, yet demonstrated better myopia control outcomes. Parental questionnaires revealed that 71.2% of Guanzhong caregivers could accurately describe ≥3 myopia prevention strategies, compared to 53.1% in Northern regions (χ² = 18.43, p < 0.001). These findings corroborate previous evidence that caregiver health literacy significantly mediates myopia progression through optimized behavioral interventions (6, 7).

The relatively favorable ocular outcomes in Southern Shaanxi may be multifactorial: (1) traditional agricultural lifestyles potentially provide more cycloplegic outdoor exposure; (2) lower population density reduces visual crowding effects. However, our study design precludes definitive attribution of these environmental factors to observed prevalence differences.

Our longitudinal observations demonstrate a consistent positive correlation between grade progression and myopia prevalence across all three geographical regions in Shaanxi Province, aligning with established epidemiological patterns in East Asian populations (8, 9). This progression likely reflects the cumulative impact of multiple risk factors: 1.

Academic Intensity

Our study revealed a distinct geographic gradient in pre-myopia prevalence across Shaanxi Province, with Southern Shaanxi demonstrating the highest detection rate, followed by Guanzhong and Northern Shaanxi. Notably, we observed a significant inverse correlation between school grade levels and pre-myopia prevalence, with male students exhibiting higher rates compared to females. These findings contrast with previous reports by Sun (17) that found no significant demographic associations, potentially reflecting regional differences in myopia progression dynamics.

These epidemiological findings reveal that primary school students in northern Shaanxi exhibit significantly higher prevalence rates of myopia and high myopia compared to other geographical regions in China (17). This distinct epidemiological pattern highlights the critical need for implementing effective strategies to decelerate myopia progression during childhood, particularly given the substantial ocular morbidity associated with high refractive errors. Notably, high myopia substantially elevates the risk of developing potentially sight-threatening ocular complications including cataract formation, retinal detachment, macular degeneration, and progressive retinal degeneration (18–20).

This investigation has several methodological limitations that warrant consideration. First, the utilization of non-cycloplegic autofocus autorefractometry for large-scale vision screening—while operationally practical for epidemiological studies—may yield systematically lower refractive error measurements compared to gold-standard cycloplegic refraction (21, 22). This methodological constraint necessitates cautious interpretation of refractive data, particularly when defining precise myopia severity thresholds. However, the protocol demonstrates superior feasibility for population-level preliminary screening given the prohibitive time requirements (average 45–60 min per subject) and logistical challenges associated with cycloplegic examinations in school-based settings. To reconcile diagnostic accuracy with operational practicality, future iterations should adopt a two-phase approach: initial mass screening through non-cycloplegic autorefraction followed by stratified random sampling for confirmatory cycloplegic refraction in representative subgroups (n ≥ 10% of total cohort).

Our population-based study reveals three critical epidemiological patterns of pediatric myopia in Shaanxi Province: 1.

Significant regional disparities—Northern Shaanxi demonstrates a substantially higher prevalence of myopia and high myopia compared to Guanzhong and Southern Shaanxi potentially associated with differential sunlight exposure patterns and educational pressure metrics.

Primary prevention should commence at preschool age (3–6 years), leveraging digital vision monitoring systems to establish longitudinal refractive profiles.

Secondary prevention requires standardized dynamic surveillance protocols:

Quarterly cycloplegic refractions for pre-myopes (SE: + 0.75 D to −0.50 D). Personalized interventions combining outdoor time optimization, and optical/pharmacological control for fast progressors (>0.5 D/year progression).

Tertiary prevention mandates multidisciplinary coordination between ophthalmologists, educators, and public health authorities to implement region-specific mitigation policies addressing northern Shaanxi's elevated high myopia risk.

It is important to acknowledge that refractive error assessments in this study were conducted without cycloplegia, which might lead to an overestimation of myopia prevalence, particularly among younger children with strong accommodative responses. Non-cycloplegic measurements tend to produce more negative spherical equivalent values due to residual accommodation, thereby increasing the likelihood of false-positive myopia diagnoses. While the use of automated refractometry and quality control procedures improved measurement consistency, the absence of cycloplegia remains a methodological limitation. Future large-scale screenings incorporating cycloplegic protocols, particularly in younger cohorts, are warranted to provide more accurate estimates of true refractive status.