目的：构建学龄前儿童近视发生风险的评估模型，并评估其在临床应用中的价值。

方法：回顾性分析2020年1月至2024年12月在我院建立屈光档案的学龄前儿童资料(选取屈光度较低眼，若双眼屈光度相同，选取右眼)，按照7∶3比例分为训练集与验证集。详细收集患儿各项资料，以儿童等效球镜度数(SE)为因变量，采用多元线性回归模型，筛选对学龄前儿童近视相关生物学指标有影响的指标。随后根据是否发生近视(根据屈光度评估，SE≤-0.50 D可诊断为近视)将训练集儿童分为近视组与未近视组。随后将多元线性回归模型筛选的影响因素纳入作为评估变量，采用Logistic回归构建风险评估方程。通过受试者工作特征(ROC)曲线、Hosmer-Lemeshow检验及决策曲线，验证评估模型的区分度、校准度及临床应用价值。

结果：本研究纳入学龄前儿童550例550眼，训练集385例385眼(男190例，女195例)，验证集165例165眼(男82例，女83例)。训练集儿童中近视组34例34眼(男16例，女18例)，未近视组351例351眼(男174例，女177例)。多元线性回归分析显示，眼轴长度(AL)/角膜曲率半径(CR)比值、脉络膜厚度(SFCT)、前房深度(ACD)、年龄、父母近视史、户外活动时间为学龄前儿童近视发生的影响因素(均P<0.05)。根据上述影响因素，构建Logistic回归风险评估模型，获得评估模型Logit(P)=0.894×AL/CR-1.312×SFCT+1.182×ACD+2.063×年龄+0.763×父母近视史-5.868。评估基于训练集所构建的学龄前儿童近视风险评估模型的效能，结果显示，ROC曲线下面积(AUC)为0.939(95%CI：0.912-0.965)，H-L检验为χ2=1.487，P=0.993，决策曲线提示在阈值0.03-0.81、0.89-0.97范围内，净收益率>0，提示模型评估学龄前儿童近视风险与实际风险具有较好区分度、校准度及临床应用价值。在独立验证集中进行内部验证，结果显示，ROC曲线的AUC为0.967(95%CI：0.919-1.000)，H-L检验为χ2=3.818，P=0.873，决策曲线提示在阈值0.01-0.99范围内，净收益率>0，进一步提示模型评估与实际风险具有较好一致性。

结论：年龄、父母近视史、AL/CR、SFCT、ACD为学龄前儿童近视发生的影响因素，据此构建的Logistic回归风险评估模型评估价值较高，可有效评估个体近视发生概率，为个体化预防学龄前儿童近视发生风险提供参考。

METHODS: A retrospective analysis was conducted on the data of preschool children who established refractive profiles in our hospital from January 2020 to December 2024. The eye with lower refractive error was selected. The right eye was chosen in cases where the refractive error of both eyes was identical. The data were divided into a training set(n=385)and a validation set(n=165)in a 7∶3 ratio. The data of the children were collected in detail, and the children's spherical equivalent(SE)was used as the dependent variable. The multiple linear regression model was used to screen the indicators that affected the biological indicators related to myopia in preschool children.Subsequently, the children in the training set were divided into a myopia group and a normal vision group based on whether myopia occurred(SE≤-0.50 D can be diagnosed as myopia according to refractive power assessment). Subsequently, the influencing factors selected by the multiple linear regression model were included as predictive variables, and a risk assessment equation was constructed using Logistic regression. The discrimination, calibration and clinical application value of the assessment model were verified through the receiver operating characteristic(ROC)curve, Hosmer-Lemeshow test and decision curve.

RESULTS: This study included 550 preschool children(550 eyes). The training set comprised 385 children(385 eyes; 190 boys and 195 girls), and the validation set comprised 165 children(165 eyes; 82 boys and 83 girls). Among the children in the training set, the myopia group included 34 children(34 eyes; 16 boys and 18 girls), and the non-myopic group included 351 children(351 eyes; 174 boys and 177 girls). Multiple linear regression analysis showed that the ratio of axial length(AL)to corneal curvature radius(CR), subfoveal choroidal thickness(SFCT), anterior chamber depth(ACD), age, parental myopia history, and outdoor activity time were risk factors for myopia in preschool children(all P<0.05). Based on the above risk factors, a Logistic regression risk assessment model was constructed, and the evaluation model Logit(P)was obtained as 0.894×AL/CR-1.312×SFCT+1.182×ACD+2.063×age+0.763×parental myopia history-5.868. The effectiveness of the myopia risk assessment model for preschool children based on the training set was evaluated. The results showed that the area under the ROC curve(AUC)was 0.939(95%CI: 0.912-0.965), the H-L test was χ2=1.487, P=0.993, and the decision curve indicated that within the threshold range of 0.03-0.81 and 0.89-0.97, the net return rate was greater than 0. This suggested that the model had good discrimination, calibration, and clinical application value in evaluating the myopia risk and actual risk of preschool children. Internal validation was conducted in an independent validation set, and the results showed that the AUC of the ROC curve was 0.967(95%CI: 0.919-1.000), the H-L test was χ2=3.818, P=0.873, and the decision curve indicated that within the threshold range of 0.01-0.99, the net return rate was greater than 0, further suggesting that the model assessment had good consistency with the actual risk.

CONCLUSION: Age, parental history of myopia, AL/CR, SFCT, and ACD are risk factors for myopia in preschool children. The Logistic regression risk assessment model constructed based on these factors has high evaluation value and can effectively evaluate the probability of myopia occurrence in individuals, providing reference for individualized prevention of myopia risk in preschool children.