Keratoconus is associated with increased ocular surface symptoms and dry eye indicatorsPeople with keratoconus experience more severe dry eye symptoms

BackgroundOcular-surface abnormalities are increasingly recognized in keratoconus (KC), but the magnitude and consistency of these changes remain uncertain. We evaluated ocular-surface symptoms, tear-function parameters, and dry-eye-related categorical outcomes in patients with KC compared with non-KC controls.MethodsPubMed, Embase, Web of Science, and the Cochrane Library were searched from inception through September 1, 2026. Comparative studies reporting the Ocular Surface Disease Index (OSDI), fluorescein tear break-up time (TBUT), noninvasive TBUT (NITBUT), the Schirmer I test (SIT), or a defined dry eye disease (DED) classification were eligible. Meibomian gland dysfunction (MGD) and meibomian gland atrophy (MGA), when reported by otherwise eligible studies, were summarized as secondary ocular-surface findings. Two reviewers independently screened studies, extracted data, and assessed methodological quality using an adapted Newcastle–Ottawa Scale (NOS). Continuous outcomes were quantitatively pooled when the analysis unit and measurement methods were sufficiently defined and comparable; otherwise, study-level findings were summarized without calculating a pooled estimate. Random-effects models were estimated by restricted maximum likelihood with Hartung-Knapp adjustment.ResultsSeven studies involving 359 participants with KC and 333 controls were included. OSDI scores were significantly higher in KC (MD, 19.45 points; 95% CI, 12.27 to 26.63; I2 = 79.9%). Fluorescein TBUT was shorter, although the pooled difference was not statistically significant (MD, −2.74 s; 95% CI, −7.81 to 2.32; I2 = 93.8%). All three studies reporting NITBUT found lower values in KC, but eye selection or data aggregation was unclear in two studies; therefore, a pooled NITBUT estimate was not calculated. SIT findings varied across studies, and a pooled estimate was not calculated because statistical independence of eye-level data and sufficiently comparable testing procedures could not be established across an adequate number of studies. The OSDI finding remained robust across sensitivity and leave-one-out analyses.ConclusionsPatients with KC had higher OSDI scores. The pooled difference in fluorescein TBUT was not statistically significant. Lower NITBUT and SIT values were reported in several studies, but pooled estimates for these outcomes were not calculated because of limitations in eye-level data handling and measurement comparability. Larger prospective multicenter studies using clearly defined analysis units and standardized tear-testing protocols are needed.Systematic Review Registrationhttps://www.crd.york.ac.uk/PROSPERO/, identifier CRD420261282391.