Fulton Jane M, Leung Tsz Wing, McCullough Sara J, Saunders Kathryn J, Logan Nicola S, Lam Carly S Y, Doyle Lesley
Centre for Optometry and Vision Science, Ulster University, Coleraine, UK.
Research Centre for SHARP Vision (RCSV), The Hong Kong Polytechnic University, Kowloon, Hong Kong.
Ophthalmic Physiol Opt. 2025 Jan;45(1):89-99. doi: 10.1111/opo.13416. Epub 2024 Nov 18.
The Predicting Myopia Onset and progression (PreMO) risk indicator, developed using data generated from white children in the UK, incorporates age, spherical equivalent refraction (SER), axial length (AL) and parental myopia to stratify the likelihood of developing myopia. This study evaluated the PreMO's predictive accuracy using prospective datasets from independent samples of children in Hong Kong (HK) and an ethnically diverse cohort of children in the United Kingdom.
Non-myopic children (SER > -0.50 D) aged 6-8 and 9-10 years were scored using the PreMO risk indicator framework, integrating baseline cycloplegic SER, AL and parental myopia data. Scores were assigned risk categories as follows: 0 = no risk, 1-3 = low risk, 4-6 = moderate risk and 7-9 = high risk. SER at ≥15 years of age was used to define refractive outcomes as 'myopic' or 'not myopic'. PreMO's predictive accuracy was analysed via Receiver Operator Characteristic curves, with Youden's J-Index identifying the optimal risk score threshold. Sensitivity, specificity and area under the curve were determined and compared with those of singular predictors, that is, SER < +0.75 D and AL ≥ 23.07 mm at 6-8 years.
In the cohort of children aged 6-8 years, a PreMO risk score ≥ 4 exhibited high sensitivity in predicting myopia onset in UK (0.97) and HK (0.94) children, with high specificity in UK (0.96) and moderate specificity in HK (0.64) children. In UK children aged 6-8 years, the PreMO outperformed singular predictors such as SER and AL. Among HK children aged 9-10 years, the PreMO score maintained high sensitivity (0.90) and moderate specificity (0.72).
A PreMO risk score ≥ 4 is a strong predictive indicator for future myopia onset, particularly in UK children. Despite high sensitivity in both UK and HK cohorts, specificity varied, indicating the need for contextual application of the tool, particularly in pre-myopic Asian children.
预测近视发生与进展(PreMO)风险指标是利用英国白人儿童的数据开发的,纳入了年龄、等效球镜度(SER)、眼轴长度(AL)和父母近视情况,以分层评估患近视的可能性。本研究使用来自中国香港(HK)儿童独立样本以及英国不同种族儿童队列的前瞻性数据集,评估了PreMO的预测准确性。
对年龄在6 - 8岁和9 - 10岁的非近视儿童(SER > -0.50 D),使用PreMO风险指标框架进行评分,该框架整合了基线散瞳SER、AL和父母近视数据。分数被分配到以下风险类别:0 = 无风险,1 - 3 = 低风险,4 - 6 = 中度风险,7 - 9 = 高风险。将15岁及以上的SER用于定义屈光结果为“近视”或“非近视”。通过受试者工作特征曲线分析PreMO的预测准确性,用尤登指数确定最佳风险评分阈值。确定敏感性、特异性和曲线下面积,并与单一预测指标(即6 - 8岁时SER < +0.75 D和AL ≥ 23.07 mm)进行比较。
在6 - 8岁儿童队列中,PreMO风险评分≥4在预测英国儿童(0.97)和香港儿童(0.94)近视发生方面表现出高敏感性,在英国儿童中特异性高(0.96),在香港儿童中特异性中等(0.64)。在6 - 8岁的英国儿童中,PreMO优于SER和AL等单一预测指标。在9 - 10岁的香港儿童中,PreMO评分保持高敏感性(0.90)和中等特异性(0.72)。
PreMO风险评分≥4是未来近视发生的有力预测指标,尤其是在英国儿童中。尽管在英国和香港队列中敏感性都很高,但特异性有所不同,这表明该工具需要根据具体情况应用,特别是在近视前期的亚洲儿童中。