Longitudinal Data on Chinese Children Suggest That Cutoff Values for Defining Premyopia Should Be Age- and Location-Specific

Purpose: The International Myopia Institute definition for premyopia includes a spherical equivalent refraction (SER) range of greater than −0.50 to +0.75 diopters (D), based largely on data from North American children. Whether this range is appropriate for children in other regions is uncertain. Methods: We pooled individual-level longitudinal data from three prospective studies of children in southern China: the Guangzhou Twin Eye Study, the Guangzhou Longitudinal Outdoor Activity Study, and the Zengcheng School Myopia Study. Children aged 6 to 12 years who were nonmyopic at baseline and had at least two follow-up visits were included. Results: Of the 6550 children included (45.5% girls), mean (SD) baseline age was 7.98 (2.03) years, and median follow-up was 3.0 (interquartile range, 2.0–5.0) years. During follow-up, 2587 children (39.5%) developed myopia. The optimal baseline SER for predicting future myopia varied systematically by baseline age and follow-up duration: thresholds became more hyperopic with longer follow-up and less hyperopic with older baseline age. For predicting myopia over approximately 3 to 5 years, the optimal SER cutoff was around +1.00 D for children aged 6 to 7 years, +0.75 D for those aged 8 to 9 years, and from +0.37 D to +0.75 D for those aged 10 to 12 years. The model area under the curve ranged from 0.791 to 0.958. Conclusions: The optimal SER cutoff for predicting future myopia onset depends on age and population. In Chinese children, +1.00 D is a more appropriate cutoff value for defining premyopia among children in the early years of primary school.