Abstract
Purpose :
The prevalence of stereoanomaly, an inability to correctly detect either crossed or uncrosed binocular disparity, is contested. We performed a cross-sectional study to determine the prevalence of stereoanomaly and to establish normative clinical values for crossed and uncrossed stereoacuity.
Methods :
Crossed and uncrossed stereoacuity was measured across four visits on a sample of 46 visually normal adults using three stereotests (Randot, TNO, Frisby). Interleaved staircases, random presentation and an extended range of disparity levels were used to mitigate inherent limitations of clinical stereotests.
Results :
Three-way repeated measures ANOVA revealed significant main effects of visit (p<0.01), stereotest (p<0.01) and disparity direction (p<0.01). Significant interactions between visit and stereotest (p=0.02) and stereotest and disparity direction (p<0.01) were also found. Seven participants (15.2%) exhibited a difference of at least two-doublings between crossed and uncrossed stereoacuity, but only with the Randot test. We propose normative crossed and uncrossed stereoacuity ranges, based on 95% confidence intervals, and a suggested upper limit for a normal result, for both single and repeated measurements (Table 1).
Conclusions :
Differences between crossed and uncrossed stereoacuity, although statistically significant, were small and inconsistent across the three stereotests. No stereoanomalous individuals were identified. Test-related factors on the Randot test, rather than a physiological cause, are likely responsible for the more marked differences between crossed and uncrossed stereoacuity in some individuals with this test. Small differences are unlikely to be clinically relevant, thus either disparity direction can be used to clinically assess stereopsis in adults. Our results provide expected levels of stereoacuity in visually normal adults. Further investigation into the existence of stereoanomaly, and whether it can be reliably detected clinically, is warranted.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.