Visual loss in high myopia without pathologic myopia: a 5-year longitudinal study on the role of Choroidal Vascularity Index

Background To determine whether a progressive reduction in Choroidal Vascularity Index (ΔCVI) is independently associated with unexplained visual loss (UVL) in highly myopic (HM) eyes without pathologic myopia (PM), and to evaluate the predictive utility of baseline CVI as a structural biomarker. Methods This longitudinal study included 126 HM eyes (axial length ≥26.0 mm or SE ≤–6.00 D) without PM and with ≥5 years of optical coherence tomography (OCT) follow-up, including 35 eyes with UVL—defined as ≥0.1 logarithm of the minimum angle of resolution best-corrected visual acuity loss in the absence of new anatomical, refractive or media-related changes— and 91 control eyes with stable vision. A 1:2 propensity score analysis was performed matching 32 eyes from UVL group and 62 control eyes. CVI was computed from subfoveal OCT B-scans acquired with Heidelberg Spectralis at baseline and at 60 months. ΔCVI was defined as the difference between follow-up and baseline. Multivariate logistic regression and ROC analyses were performed on the whole cohort to identify independent predictors of UVL. Results ΔCVI was significantly greater in UVL eyes (mean –1.87%) than in controls (–0.61%, p < 10⁻¹³). Baseline CVI did not differ significantly after matching (p = 0.12). In multivariate analysis, ΔCVI emerged as the strongest predictor of UVL (OR=10.4 per 1% CVI decrease; p < 0.001), with an area under the curve (AUC) of 0.90 and a Youden index of 0.66. In contrast, baseline CVI had poor predictive power (AUC=0.69) and did not improve model performance when added. Conclusions Longitudinal reduction in CVI, but not baseline CVI, is strongly associated with UVL in HM eyes without PM. ΔCVI may represent an early biomarker of functional decline before visible macular damage.