Astigmatic changes in orthokeratology

Purpose. To investigate and compare changes in refractive astigmatism and corneal toricity in orthokeratology (ortho-k) using the Thibos vector analysis. Methods. Consecutive records of 74 patients (aged 7 to 16 years) who were fitted with spherical ortho-k lenses at the Optometry Clinic of The Hong Kong Polytechnic University between April 2000 and November 2003 and who had been on the treatment for at least 6 months were reviewed. Manifest refractive error and topographic data were retrieved from the right eye for data analysis. Changes in refractive astigmatism and corneal toricity were evaluated using the Thibos vectors (refractive vectors, RJ0and RJ45; corneal vectors, CJ0and CJ45among the astigmats (refractive astigmatism 0.75 diopters cylinder (DC) or above) and the non-astigmats (refractive astigmatism 0.50 DC or less). Results. The mean changes in refractive astigmatism, RJ0and CJ45were 0.51 ± 0.51 diopter (D), -0.31 ± 0.35 D, and 0.09 ± 0.16 D, respectively, among the astigmats; and 0.01 ± 0.36 D, 0.00 ± 0.17 D, and 0.03 ± 0.13 D, respectively, among the non-astigmats. RJ0 was significantly reduced, and RJ45 was significantly increased only in the astigmats (one sample t tests, astigmats: p < 0.01; non-astigmats: p > 0.10). Although there were significant correlations between the initial RJ0and CJ45with their changes in all subjects and in the astigmats, only 21 to 55% of the variability could be explained by the regression equations (Pearson r: all subjects -0.63 < r < -0.62; astigmats -0.74 < r < -0.46, p < 0.001). No significant changes in corneal toricity and no correlations between the changes in refractive astigmatism and changes in corneal toricity were found (p > 0.13). Conclusions. Spherical ortho-k leads to significant changes in R0and CJ45, but only 38% of the variance of the changes in all subjects could be explained by relationship between changes and the initial RJ0or RJ45. No correlation between changes in refractive astigmatism and changes in corneal toricity was found, and ortho-k treatment did not induce astigmatism in our non-astigmatic subjects.