PURPOSE: Most astigmats have a similar level of astigmatism in each eye. However, there is controversy over whether the astigmatic axes in fellow eyes typically show direct or mirror symmetry. We carried out a statistical analysis designed to address this issue. METHODS: The median absolute difference in the astigmatic axes of fellow eyes was calculated for a sample of 50 995 astigmats (subjects with at least 0.25 D of astigmatism in each eye). This was done, firstly, for a 'direct symmetry model' in which the difference in axis was calculated as |AxisR - AxisL| and secondly, for a 'mirror symmetry model' in which the difference in axis was calculated as |AxisR - (180 - AxisL)|. RESULTS: Under the direct symmetry model, the median absolute difference in the axis of astigmatism between fellow eyes was 20 degrees. Under the mirror symmetry model, the median absolute difference in the axis of astigmatism between fellow eyes was significantly lower, at 10 degrees (p < 10e-100). Comparable results were found when the analysis was restricted to subjects with: lower levels of astigmatism (< or =1.00 D), higher levels of astigmatism (>1.00 D), against-the-rule astigmatism, with-the-rule astigmatism or oblique astigmatism (all p < 10e-100). CONCLUSION: Our results show that mirror, rather than direct, symmetry is the norm.
PURPOSE: Most astigmats have a similar level of astigmatism in each eye. However, there is controversy over whether the astigmatic axes in fellow eyes typically show direct or mirror symmetry. We carried out a statistical analysis designed to address this issue. METHODS: The median absolute difference in the astigmatic axes of fellow eyes was calculated for a sample of 50 995 astigmats (subjects with at least 0.25 D of astigmatism in each eye). This was done, firstly, for a 'direct symmetry model' in which the difference in axis was calculated as |AxisR - AxisL| and secondly, for a 'mirror symmetry model' in which the difference in axis was calculated as |AxisR - (180 - AxisL)|. RESULTS: Under the direct symmetry model, the median absolute difference in the axis of astigmatism between fellow eyes was 20 degrees. Under the mirror symmetry model, the median absolute difference in the axis of astigmatism between fellow eyes was significantly lower, at 10 degrees (p < 10e-100). Comparable results were found when the analysis was restricted to subjects with: lower levels of astigmatism (< or =1.00 D), higher levels of astigmatism (>1.00 D), against-the-rule astigmatism, with-the-rule astigmatism or oblique astigmatism (all p < 10e-100). CONCLUSION: Our results show that mirror, rather than direct, symmetry is the norm.