A split decision based on vision: an empirical approach to classify short-distance sprinters with a vision impairment.

Athletes with vision impairment (VI) competing in track athletics typically compete in one of three 'classes' based on the severity of their VI. The creation of these classes, however, was arbitrary and not based on evidence. The aim of this study was to investigate the relationship between monocular visual acuity (VA) and sprinting performance in athletes with a VI to establish the optimal cut-off point(s) between competition classes. Historical data including VA and sprinting performance were collected through the International Paralympic Committee's Sport Data Management System. Pearson correlations were conducted to investigate the relationship between VA and sprinting performance. VA was significantly associated with 100, 200 and 400 m race times (r = 0.26; r = 0.26; r = 0.24 [all p < 0.001], respectively). Decision tree analysis suggested splitting data into two classes with a VA cut-off at 2.1 logMAR. Stability assessment confirmed a split into two classes but showed considerable variability in the cut-off points between 2.1 and 3.2 logMAR. A two-class system would provide legitimate competition for athletes with VI in short-distance track athletics. However, other visual functions (e.g. contrast sensitivity) should be considered in future classification research.