Association Between Refractive Status and Ocular Biometric Parameters Among Children and Adolescents — 10 PLADs, China, 2020–2024

In 2018, China elevated myopia prevention and control to a national strategic priority; however, myopia prevalence among Chinese children and adolescents has remained persistently high, with overall rates reaching 52.7% by 2020 (1) and demonstrating a concerning trend toward younger age of onset. Axial myopia represents the predominant form of myopia in Chinese youth (2), and accumulating evidence indicates that the axial length-to-corneal radius ratio (AL/CR) correlates more strongly with refractive status than axial length (AL) alone (3–5). Investigating the relationships between ocular biometric parameters and myopia development provides fundamental insights into myopia pathogenesis. Compared to automated refractometers that measure only refractive error, biometric parameter assessment offers the distinct advantage of identifying high-risk populations earlier, enabling early intervention before visual acuity deteriorates. Although previous studies have examined correlations between ocular biometric parameters and myopia in Chinese students, most have focused on specific regions, with national-scale investigations remaining limited. Therefore, leveraging the comprehensive monitoring system for hyperopic reserves among children and adolescents established across 10 provincial-leveladministrative divisions (PLADs) in China beginning in 2020, this study examined AL, corneal radius (CR), and AL/CR in students from these ten regions. We analyzed the distribution of refractive parameters and their associations with refractive status to further elucidate ocular growth patterns and refractive development in students, while evaluating the monitoring effectiveness of ocular biometric parameters for refractive development to inform myopia prevention and control strategies.

The National Disease Control and Prevention Administration of China coordinated a cross-sectional survey of hyperopic reserves across 10 PLADs from November 2020 to July 2024. These included Liaoning Province, Beijing Municipality, Zhejiang Province, Guangdong Province, Henan Province, Hunan Province, Shanxi Province, Shaanxi Province, Shandong Province, and Chongqing Municipality. Students from senior kindergarten through high school were selected using multistage cluster sampling at each site. Initially, 67,260 participants were surveyed. Exclusions included 22 individuals with diseases affecting visual development or interfering with ophthalmic examinations, 6,030 with missing spherical equivalent data or abnormal post-mydriasis values, 171 outside the 5–18 year age range, and 767 with missing or abnormal right eye AL or corneal radius (CR) values. The final sample comprised 60,270 participants: 14,475 kindergarten students, 28,482 lower-grade primary school students (grades 1–3), 11,317 upper-grade primary school students (grades 4–6), 3,496 junior high school students, and 2,500 senior high school students. All participants provided informed consent, and the study received institutional ethics committee approval (batch number: 2022 [24]). Myopia was defined as spherical equivalent refraction (SER) ≤−0.50 D in either eye under cycloplegia (6). Classifications included non-myopia (SER>−0.50 D), mild myopia (−3.00 D≤SER≤−0.50 D), and moderate myopia (−6.00 D≤SER<−3.00 D). Cycloplegic refraction employed 0.5% compound tropicamide eye drops administered four times at 5-minute intervals, with refraction performed 30 minutes after the final instillation. Intraocular pressure measurement was conducted when clinically indicated. All regions selected ocular biometric measurement instruments based on local resources, with standardized equipment brands and models used whenever possible.

Statistical analyses were conducted using SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Since SER, AL, and AL/CR did not follow normal distributions, we reported data as median values with interquartile ranges and used non-parametric tests for comparisons. The Mann–Whitney U test was employed for two-group comparisons, while the Kruskal–Wallis test was used for multiple-group analyses. Correlations between variables were assessed using Spearman’s rank correlation coefficient. To determine optimal cut-off points for AL and AL/CR in predicting myopia onset, we calculated the Youden index and constructed receiver operating characteristic (ROC) curves with corresponding areas under the curve and 95% confidence intervals. When Youden indices were equivalent, we prioritized sensitivity over specificity. Statistical significance was established at P<0.05.

This study examined the visual status of 60,270 students aged 5–18 years across 10 PLADs in China, comprising 31,070 boys and 29,200 girls. The overall myopia prevalence was 29.24%. Myopia prevalence among kindergarten, lower-grade primary school, upper-grade primary school, junior high school, and senior high school students was 5.20%, 19.76%, 56.49%, 79.12%, and 83.28%, respectively. Student myopia prevalence demonstrated a significant increasing trend with advancing grade levels (χ²=15,989.103, P<0.001), with statistically significant differences between each grade after Bonferroni adjustment. The overall myopia prevalence was lower in boys (28.04%) than in girls (30.51%) (Table 1).

Correlation analysis demonstrated that SER values between left and right eyes, as well as biometric parameters between eyes, exhibited strong correlations (R>0.9). Consequently, only right eye data were included in subsequent analyses. The median SER values [M (Q25, Q75)] for kindergarten, lower-grade primary school, upper-grade primary school, junior high school, and senior high school students were 1.00 (0.63, 1.38) D, 0.63 (0.00, 1.00) D, −0.50 (−1.88, 0.38) D, −1.88 (−3.38, −0.38) D, and −2.38 (−3.75, −0.75) D, respectively. SER values progressively decreased with advancing grade levels, with statistically significant differences between all grade comparisons (P<0.001). Median AL values [M (Q25, Q75)] across grade levels were 22.56 (22.06, 23.02) mm, 23.00 (22.46, 23.53) mm, 23.81 (23.16, 24.45) mm, 24.34 (23.67, 25.01) mm, and 24.59 (23.82, 25.36) mm, respectively. AL demonstrated a consistent increase with grade progression, with significant differences between all grade levels (P<0.001). Corneal radius (CR) values [M (Q25, Q75)] remained relatively stable across grades: 7.77 (7.60, 7.94) mm, 7.77 (7.60, 7.95) mm, 7.79 (7.62, 7.96) mm, 7.79 (7.61, 7.97) mm, and 7.80 (7.64, 7.98) mm. The AL/CR ratio [M (Q25, Q75)] increased progressively across grade levels: 2.90 (2.86, 2.95), 2.95 (2.90, 3.01), 3.05 (2.98, 3.13), 3.13 (3.04, 3.21), and 3.15 (3.06, 3.24), with significant differences between all grade comparisons (P<0.001) (Table 2).

Female students consistently demonstrated SER values equal to or lower than their male counterparts across all grade levels. Male students exhibited greater AL and CR measurements compared to females in all grades. The AL/CR ratio was higher in male students across all grades except senior high school, with all differences reaching statistical significance (P<0.05) (Table 2).

Spearman correlation analysis revealed strong associations between ocular biometric parameters and refractive status. AL demonstrated a high negative correlation with SER (R=−0.657, P<0.001), while AL/CR showed an even stronger negative correlation (R=−0.750, P<0.001). Regardless of the student’s refractive classification, the correlation between AL/CR and SER consistently exceeded that observed between AL, CR, and SER (Table 3).

ROC curve analysis established optimal cut-off values for myopia prediction across all school grades. For AL, the cut-off points were 23.10 mm (kindergarten), 23.24 mm (lower-grade primary school), 23.72 mm (upper-grade primary school), 24.00 mm (junior high school), and 24.12 mm (senior high school). The corresponding areas under the ROC curve were 0.724 (95% CI: 0.704, 0.744), 0.784 (95% CI: 0.777, 0.791), 0.801 (95% CI: 0.793, 0.809), 0.824 (95% CI: 0.809, 0.840), and 0.858 (95% CI: 0.840, 0.876), respectively. For AL/CR, the optimal cut-off values were 2.97 (kindergarten), 2.99 (lower-grade primary school), 3.04 (upper-grade primary school), 3.07 (junior high school), and 3.10 (senior high school). The AL/CR parameter demonstrated superior predictive accuracy with areas under the ROC curve of 0.825 (95% CI: 0.808, 0.843), 0.854 (95% CI: 0.848, 0.860), 0.881 (95% CI: 0.875, 0.887), 0.885 (95% CI: 0.872, 0.897), and 0.882 (95% CI: 0.865, 0.899) for the respective grade levels.