Ocular Magnification Correction Influences Glaucoma-Related Microvascular Measurements in OCT Angiography

Purpose: To evaluate the impact of axial length (AL) handling strategies on optical coherence tomography angiography (OCTA)-derived vessel density and its interpretation in glaucoma. Methods: This cross-sectional study included 1091 participants (1823 eyes). Vessel density was quantified using fixed region of interest (ROI) and AL-rescaled ROI approaches. Comparisons of glaucoma versus normal eyes were evaluated using generalized estimating equation (GEE) models adjusted for age, sex, systemic hypertension, OCTA signal strength index, and AL. Associations between vessel density and visual field mean deviation (VF MD) were evaluated using multivariable GEE models. Results: Glaucoma eyes had longer AL than normal eyes (P < 0.001). Under the fixed-ROI approach, deep vessel density was positively associated with AL, whereas superficial vessel density showed no significant association. After rescaling, both were inversely associated with AL (P < 0.001). Although rescaling increased apparent glaucoma-normal differences, estimates were similar after AL adjustment (superficial: β = -0.35 vs. -0.36 mm⁻1; deep: β = -1.03 vs. -1.09 mm⁻1). In the age- and sex-matched cohort (544 participants; 890 eyes), differences remained similar between approaches (superficial: β = -0.293 vs. -0.310 mm⁻1; deep: β = -0.963 vs. -0.992 mm⁻1). Both vessel density measures were associated with VF MD after AL adjustment. AL rescaling improved deep vessel density model fit and yielded slightly higher discrimination in sensitivity analyses allowing nonlinear AL effects (superficial: area under the curve [AUC] = 0.884 vs. 0.868; deep: AUC = 0.895 vs. 0.869). Conclusions: Ocular magnification correction influenced OCTA measurements, but its incremental impact on glaucoma-related associations after AL adjustment was small. The clinical relevance remains uncertain.