Associations between ocular parameters and corneal astigmatism in young Chinese adults: a cross-sectional study

Abstract

Purpose: This study aims to explore the relationships between key ocular optical parameters and corneal astigmatism, providing new insights into the factors influencing astigmatic changes in the human eye. Setting: The study was conducted at a tertiary eye care center with a population of adults. Design: A cross-sectional, observational analysis of ocular measurements. Methods: A total of 935 adults (507 males, 428 females; aged 21–44 years) participated. Inclusion criteria were adults with no history of ocular disease or surgery. Exclusion criteria included individuals with systemic diseases affecting ocular health. Ocular measurements including central corneal thickness (CCT), anterior chamber depth (ACD), axial length (AL), lens thickness (LT), vitreous thickness (VT), white-to-white diameter (WTW), spherical power (SPH), and cylindrical power (CYL) were obtained using the Suoer SW-9000 μm Plus biometer and Nidek ARK-1 Autorefractor. Corneal astigmatism (ΔK) was calculated as the difference between K2 and K1 within the 3 mm optical zone. Shapiro–Wilk, Pearson correlation, and stepwise multiple linear regression analyses were applied. Results: ΔK was significantly correlated with age (r = -0.07, p = 0.030), AL (r = 0.10, p = 0.002), VT (r = 0.10, p = 0.002), SPH (r = -0.13, p < 0.001), and CYL (r = -0.77, p < 0.001). After multicollinearity diagnostics, VT was excluded (VIF > 10). Multiple linear regression identified AL (β = -0.269, 95% CI: -0.415 to -0.123, p = 0.002, VIF = 3.75), SPH (β = 0.072, 95% CI: 0.012 to 0.132, p = 0.018, VIF = 3.39), and cylindrical power (β = -0.806, 95% CI: -0.909 to -0.703, p < 0.001, VIF = 2.31) as significant predictors, with an adjusted R² of 0.608. Conclusions: AL, SPH, and CYL significantly influence corneal astigmatism, suggesting corneal morphology is closely linked to ocular characteristics, which may aid in targeted approaches to diagnosis and management.