Purpose Evaluate average cyclotorsional and non-cyclotorsional components (NCY) of eye rotation from sitting to supine, and associate average cyclotorsion to different variables.

Methods Medical records of patients who underwent bilateral sequential laser refractive surgery were retrospectively evaluated. Recorded variables included the patient's age, refraction, automated keratometry, pupil displacement and eye rotation from sitting to supine position. Measured iris rotation (total rotation, TR) was decomposed into two components: NCY, defined as the common rotation component of each eye of the same patient, and cyclotorsional component (CY), defined as the assumed independent eye rotation for each eye in relation to the face, so that TR=NCY+CY. Cyclotorsion ratio (CR) was calculated as CR=|CY|/|TR|, and used to correlate CY with TR for each eye.

Results Data from 310 eyes of 155 patients were evaluated. TR was +1.43° ±3.41° (−8.30° to +9.20°). Average CYs and NCYs per patient were +1.43°±2.04° (−3.15± to +7.40°) and −0.28°±2.72° (−6.85° to +7.15°), respectively. TR demonstrated that 40.6% and 8.4% of patients presented bilateral excyclotorsion and incyclotorsion, respectively. When excluding NCYs, average CYs demonstrated that 74.2% of patients presented excyclotorsion and 23.9% presented incyclotorsion. CR demonstrated that TR represented from 75% to 125% of average CY in 19.68% of the eyes. TR overestimated and underestimated average CYs above these limits in 52.26% and 28.06% of the eyes, respectively. There was no statistical association between average CYs and the different variables.

Discussion This study demonstrates that most of the rotations previously attributed to torsional components were probably due to NCYs, such as postural misalignments. Apparently, the amplitude of cyclotorsional movements is smaller than observed in previous reports, and could not be associated with any studied variable.