Refractive status and prevalence of myopia among Chinese primary school students.

<bold>Background: </bold>The aim of this study was to investigate the prevalence of myopia in key (university-oriented) and non-key elementary schools in China using a traditional and a new criterion for myopia diagnosis in an epidemiological study.<bold>Methods: </bold>This school-based, cross-sectional study examined students from four key schools and seven non-key schools. Non-cycloplegic autorefraction and visual acuity (VA) were performed on each student. Myopia was defined as a spherical equivalent (SE) refractive error not better than -1.00 D. A questionnaire was also administered.<bold>Results: </bold>Of the 13,220 students examined, 6,546 (49.5 per cent) had myopia using the criterion of SE not better than -1.00 D. However, 2,246 (34.3 per cent) of these myopes had VA ≥ 0 logMAR in both eyes, indicating they were not functioning as myopes. Thus, a second myopia criterion was adopted: SE refractive error not better than -1.00 D + uncorrected VA ≥ 0 logMAR in at least one eye. By this definition, only 32.5 per cent of the overall sample had myopia. Students in key schools had a higher prevalence of myopia than those in non-key schools (53.8 per cent versus 44.7 per cent) by the initial criterion. By the new criterion, the prevalence of myopia was 41.2 per cent versus 22.7 per cent. Myopia was equal in grade 1 of both school types, but accelerated faster in key schools, where there was a much higher prevalence of myopia by fourth grade, and continued up to 79.2 per cent prevalence by sixth grade based on SE refractive error not better than -1.00 D.<bold>Conclusion: </bold>Students in more competitive university-oriented elementary schools developed myopia much faster than those in regular schools, although they started with the same level of myopia. Since one-third of the 'myopes' had VA ≥ 0 logMAR in both eyes, they would not be prescribed a correction, or be clinically treated as myopes. A new criterion of SE refractive error not better than -1.00 D + uncorrected VA ≥ 0 logMAR in at least one eye was tested. This criterion is more clinically appropriate and could be used in future epidemiological studies. [ABSTRACT FROM AUTHOR]