分析浙江省金华市儿童眼生物学参数与近视眼进展风险的关联，探讨多参数联合评估在近视眼进展风险预测中的应用价值。

收集2014年1月至2023年12月于浙江省金华地区进行近视眼防控筛查的6~12岁儿童9567人(9567只眼)的临床资料。其中，男性4872人(4872只眼)，女性4695人(4695只眼)；年龄6~12岁，平均年龄(8.0±2.0)岁。随访5年，收集研究对象一般人口学信息、检测眼球生物学参数并进行非睫状肌麻痹验光。根据5年近视眼等效球镜屈光度(SE)进展情况，分为无进展组、缓慢进展组及快速进展组。年龄、身体质量指数(BMI)及SE呈偏态分布，采用中位数(四分位间距)表示，组间比较采用Kruskal-Wallis秩和检验；眼轴长度(AL)、角膜曲率半径(CR)、AL/CR及眼压(IOP)分布近似正态，采用±s表示，组间比较采用单因素方差分析。性别以例数和百分比表示，组间比较采用Pearson χ²检验。近视眼进展具有有序等级特征，进行Jonckheere-Terpstra趋势检验，作为补充趋势分析。对总体差异有统计学意义的变量进一步进行两两比较，并采用Bonferroni法校正。以近视眼进展等级为因变量，分别进行单因素和多因素有序Logistic回归分析，计算β系数、Z值、Waldχ²值、OR值、95%置信区间及P值。将基线年龄、BMI、基线SE、基线AL/CR及基线IOP纳入多因素Logistic回归分析。采用方差膨胀因子评估自变量间多重共线性，方差膨胀因子<5认为不存在明显多重共线性。

经相关性检验，所有受试者左右眼SE呈高度正相关，有统计学意义(r＝0.89，P<0.05)，故本研究以右眼数据作为主要分析对象。本研究儿童年龄中位数为8.0(7.0，9.0)岁，身高中位数为130.0(124.0，140.0)cm，体质量中位数为26.5(23.0，32.0)kg，BMI中位数为15.63(14.42，17.36)kg/m²。所有受试者平均AL、CR、AL/CR及IOP分别为(23.30±1.07)mm、(7.81±0.25)mm、(2.99±0.13)及(15.23±2.83)mmHg(1 mmHg＝0.133 kPa)，SE中位数0.00(-1.00，0.38)D。正视、低度近视眼及远视者分别为5458只眼、2250只眼及1160只眼，分别占57.0%、24.0%及12.0%。本研究无进展组767人(767只眼)、缓慢进展组3359人(3359只眼)、快速进展组5441人(5441只眼)。Kruskal-Wallis秩和检验显示，三组基线年龄、BMI及SE的总体差异均有统计学意义(H＝245.07，20.30，622.58；P<0.05)。单因素方差分析显示，三组基线AL、CR、AL/CR及IOP的总体差异均有统计学意义(F＝277.14，10.28，392.62，9.70；P<0.05)。Jonckheere-Terpstra趋势检验显示，随着近视眼进展等级升高，年龄、BMI、AL、AL/CR及IOP呈递增趋势，而CR和SE呈递减趋势(Z＝15.09，2.73，24.11，-4.01，31.34，3.92，-20.99；P<0.05)。经Bonferroni法校正，两两比较结果显示，年龄、BMI、AL、AL/CR及SE在无进展组、缓慢进展组与快速进展组之间差异均有统计学意义(P<0.05)。性别分布在三组间总体差异无统计学意义(χ²＝2.190，P>0.05)。屈光状态分布在三组间总体差异有统计学意义(χ²＝811.670，P<0.05)，且无进展组与缓慢进展组、无进展组与快速进展组、缓慢进展组与快速进展组之间的两两比较差异均有统计学意义(χ²＝91.05，312.08，571.32；P<0.05)。多因素有序Logistic回归分析显示，年龄和BMI为基线学龄期人群近视眼进展的独立保护因素(OR＝0.947，0.983；95%CI：0.919~0.976，0.967~0.999；P<0.05)，SE、AL/CR及IOP为近视眼进展至更高等级的独立危险因素(OR＝1.177，712.007，1.019；95%CI：1.124~1.232，399.544~1275.193，1.004~1.033；P<0.05)。其中，AL/CR的关联强度最大，其每升高0.1，儿童进展至更高近视眼等级的风险约增加93%。共线性诊断显示，所有自变量的方差膨胀因子均<5，无明显多重共线性。

多项眼生物学参数与儿童近视眼进展密切相关。其中，AL/CR是预测近视眼进展的核心指标，IOP可作为附加参考指标。以AL/CR为核心的多指标联合评估策略，有助于近视眼快速进展的高危个体早期识别，为儿童近视眼的早期筛查、分层管理及公共卫生防控策略制定提供循证依据。

The aim of this study is to analyze the association between ocular biological parameters and the risk of myopia progression in children in Jinhua City of Zhejiang Province, and to explore the application value of combined multi-parameter assessment in predicting the risk of myopia progression.

Clinical data of 9567 children aged 6 to 12 years at baseline who underwent myopia screening and prevention programs in the Jinhua area of Zhejiang Province from January 2014 to December 2023 were collected. Among them, 4872 were boys and 4695 were girls, with a mean age of (8.0±2.0) years (ranging from 6 to 12 years). All participants were followed up for 5 years. Demographic information, ocular biometric parameters, and non-cycloplegic refractive data were obtained. According to the 5-year progression of spherical equivalent (SE), participants were classified into the non-progression group, slow progression group, and rapid progression group. Continuous variables were first tested for normality and further evaluated using histograms and Q-Q plots. Skewed distributions were described as median (interquartile range) and were performed using the Kruskal-Wallis rank-sum test between-group comparisons. Normal distributions were described as ±s and were performed using one-way analysis of variance between-group comparisons. Categorical variables were expressed as number and percentage and compared using the Pearson chi-square test. Myopia progression was performed by Jonckheere-Terpstra trend test, and pairwise comparisons with Bonferroni correction. Univariate and multivariable ordinal logistic regression analyses were performed with myopia progression grade as the dependent variable, and β coefficients, Z values, Wald χ² values, odds ratio (OR), 95% confidence intervals (CI), and P values were calculated. The multivariable model included baseline age, BMI, baseline SE, baseline AL/CR, and baseline IOP. Variance inflation factor (VIF) was used to assess multicollinearity among independent variables.

After correlation analysis, a strong positive correlation was observed between the SE of the left and right eyes in all subjects, with statistical significance (r=0.89, P<0.05). Therefore, the right eye data were used as the primary analysis object in this study. Among of them, the median age was 8.0 (7.0, 9.0) years, the median height was 130.0 (124.0, 140.0) cm, the median body mass was 26.5 (23.0, 32.0) kg, and the median BMI was 15.63 (14.42, 17.36) kg/m². The mean AL, CR, AL/CR, and IOP of all subjects were (23.30±1.07) mm, (7.81±0.25) mm, (2.99±0.13), and (15.23±2.83) mmHg (1 mmHg=0.133 kPa), respectively. The median SE was 0.00 (-1.00, 0.38) D. There were 5458 eyes with emmetropia, 2250 eyes with low myopia, and 1160 eyes with hyperopia, accounting for 57.0%, 24.0%, and 12.0%, respectively. Among of the 9567 participants, there were 767 cases (767 eyes) in the non-progression group, 3359 cases (3359 eyes) in the slow progression group, and 5441 cases (5441 eyes) in the rapid progression group. Kruskal-Wallis tests showed significant overall differences in baseline age, BMI, and SE among the three groups (H=245.07, 20.30, 622.58; P<0.05). One-way analysis of variance showed significant overall differences in baseline AL, CR, AL/CR, and IOP (F=277.14, 10.28, 392.62, 9.70; P<0.05). Jonckheere-Terpstra trend analysis showed that with increasing myopia progression grade, age, BMI, AL, AL/CR, and IOP showed increasing trends, whereas CR and SE showed decreasing trends (Z=15.09, 2.73, 24.11, -4.01, 31.34, 3.92, -20.99; P<0.05). After correction using the Bonferroni method, pairwise comparisons showed that there were statistically significant differences in age, BMI, AL, AL/CR, and SE among the non-progression group, the slow progression group, and the rapid progression group(P<0.05). There was no significant overall difference in sex distribution among the three groups (χ²=2.190, P>0.05). In contrast, refractive status distribution differed significantly overall (χ²=811.670, P<0.05), and all three pairwise comparisons were also significant (χ²=91.05, 312.08, 571.32; P<0.05). Multivariable ordinal logistic regression analysis showed that age and BMI were independent protective factors for myopia progression in the baseline school-age population (OR=0.947, 0.983; 95%CI: 0.919 to 0.976, 0.967 to 0.999; P<0.05), whereas SE, AL/CR, and IOP were independent risk factors for progression to a higher myopia grade (OR=1.177, 712.007, 1.019; 95%CI: 1.124 to 1.232, 399.544 to 1275.193, 1.004 to 1.033; P<0.05). Among them, AL/CR showed the strongest association, and each 0.1 increase in AL/CR was associated with an approximately 93% increase in the risk of progression to a higher myopia grade. Collinearity diagnostics showed that all VIF were <5, indicating no evident multicollinearity.

Multiple ocular biological parameters were closely associated with myopia progression in children and adolescents. Among them, AL/CR was the core indicator for predicting myopia progression, while IOP may serve as an additional reference indicator. A multi-parameter assessment strategy centered on AL/CR may facilitate early identification of individuals at high risk of rapid myopia progression and provide an evidence-based basis for early screening, stratified management, and public health prevention and control of myopia in children and adolescents.