Patients with USOP generally demonstrate hypertropia and excyclotorsion in the paretic eye. In this study, the excyclotorsion of USOP was more frequently observed in the paretic eye (71.4%) than in the non-paretic eye (28.6%). This is consistent with previous studies showing that 25% of cases of excyclotorsion could also be present in the non-paretic eye.[12, 13, 20].
In this study, the incidence of paradoxical ocular excyclotorsion in the non-paretic eye could be increased when patients had a fixation preference in the paretic eye or were aged <2 years. Initially, paralysis of the superior oblique may lead to ocular excyclotorsion in the paretic eye, followed by ocular dominance , cyclofusion , or a neural adaptation mechanism  in the non-paretic eye, which can induce ocular excyclotorsion in either eye.
According to Kim et al., the dominant eye could influence the alignment of the non-paretic eye through conjugate cycloversion eye movements. It has been demonstrated that neural adaptability for improving the efficiency of motor control and visual function may contribute to the torsional state in USOP. Prolonged fixation with the paretic dominant eye might act in a way that conversely decreases the amount of extortion in the eye, which induces excyclotorsion in the non-paretic eye through conjugate movements according to Hering’s law. Additionally, Kusher and Hairharan reported that long-term fixation with the affected eye in USOP could induce fundus torsional change in the affected eye.
In this study, 42.9% of patients with excyclotorsion in the non-paretic eye were under 2 years of age. This result could be explained by the immature binocular visual system and sensory adaptation of children. A lack of immediate motor correction for torsional misalignment and defective binocular fusion in USOP may disrupt cyclofusion, which, in turn, induces the eye to be more excyclorotated, especially in the case of a non-paretic eye. Graf et al. demonstrated that the resting position of human eyes became more excyclorotated after disrupting binocular fusion with 8 h of prolonged monocular occlusion. Shin et al. also demonstrated that a certain level of defective binocular fusion might disrupt cyclofusion and subsequently make the eyes more excyclorotated.
In an effort to overcome vertical diplopia, patients with poor sensory adaptation, especially children, would have a greater need for vertical fusion. The immaturity of fusion would contribute to excycloduction in the non-paretic eye. Repetitive sensorial and motor adaptations to torsional misalignment aggravate fundus extorsion in patients with USOP under 2 years of age.
This study demonstrates that the reduction of excyclotorsion after surgery was larger in the accordance group than in the discordance group, consistent with the results of a previous study. A possible explanation for this result might be a significant improvement in excyclotorsion in the paretic eye, which might be due to postoperative mechanical changes in the extraocular muscles, which has a direct effect on the torsional forces in the accordance group.
This study has a few limitations. First, it was conducted retrospectively, and most patients were pediatric patients. Second, although 98 patients were enrolled in this study, additional USOP patients might be needed to confirm the findings of this study. Third, the possibility of observer bias might exist for measurements of pre- and postoperative degrees of excyclotorsion using fundus examination in patients who were under 2 years of age.
In conclusion, the paretic eye may not coincide with excyclotorsion in USOP. Excyclotorsion in the non-paretic eye was observed in 28% of patients with USOP. Patients aged <2 years and with a fixation preference for the paretic eye might have a significantly greater propensity to undergo excyclotorsion in the non-paretic eye. Surgery for USOP can have a more obvious effect in improving excyclotorsion in the paretic eye.