Background/aims To evaluate the impact of dry eye on reading performance.

Methods Out-loud and silent reading in patients with clinically significant dry eye (n=41) and controls (n=50) was evaluated using standardised texts. Dry eye measures included tear film break-up time, Schirmer's test and corneal epithelial staining. Symptoms were assessed by the Ocular Surface Disease Index (OSDI).

Results The dry eye group had a greater proportion of women as compared with the control group but did not differ in age, race, education level or visual acuity (p≥0.05 for all). Out-loud reading speed averaged 148 words per minute (wpm) in dry eye subjects and 163 wpm in controls (p=0.006). Prolonged silent reading speed averaged 199 wpm in dry eye subjects versus 226 wpm in controls (p=0.03). In multivariable regression models, out-loud and sustained silent reading speeds were 10 wpm (95% CI −20 to −1 wpm, p=0.039) and 14% (95% CI −25% to −2%, p=0.032) slower, respectively, in dry eye subjects as compared with controls. Greater corneal staining was associated with slower out-loud (−2 wpm/1 unit increase in staining score, 95% CI =−3 to −0.3 wpm) and silent (−2%, 95% CI −4 to −0.6 wpm) reading speeds (p<0.02 for both). Significant interactions were found between OSDI score and word-specific features (longer and less commonly used words) on out-loud reading speed (p<0.05 for both).

Conclusions Dry eye is associated with slower out-loud and silent reading speeds, providing direct evidence regarding the functional impact of dry eye. Reading speed represents a measurable clinical finding that correlates directly with dry eye severity.