RÉSUMÉ
<p><b>OBJECTIVES</b>To explore the association between elevated light levels in classrooms and change in vision acuity among elementary and secondary students.</p><p><b>METHODS</b>A total of 4 elementary (grade 1-5) and secondary (grade 7-8) schools in urban and rural areas in Sujiatun, Shenyang, China were selected by cluster sampling as experimental schools, and lighting systems have been rebuilt to improve the ambient light levels in 56 classrooms in November 2012. The control schools were chosen for the comparable academic burden and adjacent location to experimental schools, 4 schools in all. Cluster sampling of all students in the selected schools as the subjects was carried out. A total of 2 092 students were chosen as experimental group and 1 595 students were in the control group. The luxmeter was used to measure illuminance of classrooms in two groups at baseline, and intervention for 1 month, respectively.Students in both groups were underwent 3 times for vision acuity examination by standard logarithmic visual acuity chart at baseline, intervention for 6 month and intervention for 1 year, respectively. The light levels of desk and blackboard in two groups were compared by Wilcoxon test. Multivariate analysis of covariance with repeated measures was performed to assess three vision acuity results between groups.</p><p><b>RESULTS</b>After intervention, the average illuminance of desk (117.5 vs 532.5 lx, Z = -5.38, P < 0.001) and blackboard (75.6 vs 423.5 lx, Z = -5.38, P < 0.001) and uniformity of desk (Z = -4.28, P < 0.001) with new lighting were improved significantly than that with old lighting, however the uniformity of blackboard was lower than baseline significantly (0.64 vs 0.70, Z = -2.34, P = 0.019). The average scores of vision acuity in students at baseline, intervention for 6 month and intervention for 1 year were 4.87 ± 0.23, 4.84 ± 0.25 and 4.85 ± 0.23 in experimental group, and 4.88 ± 0.22, 4.84 ± 0.25 and 4.81 ± 0.27 in control group, respectively. The significant differences between groups were found and F values were 1.41, 0.13, 19.99, P values were 0.235,0.724, <0.001. At last the average vision acuity in experimental group were significantly better than that in control group either among elementary (4.90 ± 0.20) vs (4.87 ± 0.21) score, F = 13.61, P < 0.001 or secondary students (4.73 ± 0.28) vs (4.68 ± 0.32) score, F = 14.25, P < 0.001.</p><p><b>CONCLUSIONS</b>Visual acuity loss could be decreased in students with elevated light levels which may slow the response to myopiagenic stimuli for eyes, therefore the ambient light levels of blackboard and desk in classroom should be improved.</p>