Quantitative analysis of retinal microvascular changes in gestational hypertension and preeclampsia using artificial...

Hypertensive disorders of pregnancy (HDP) are among the leading causes of maternal and perinatal morbidity and mortality worldwide. Current clinical diagnosis primarily relies on late-onset indicators including elevated blood pressure and proteinuria, which fail to meet clinical demands for early identification and risk stratification. The retinal microvasculature shares high structural and physiological homology with cerebral and placental vasculature, providing an ideal non-invasive window for systemic microcirculation assessment. In this exploratory cross - sectional study, artificial intelligence - based quantitative fundus analysis was employed to assess retinal microvascular parameters in 36 pregnant women (72 eyes). The retinal arterial diameter and the mean macular venous diameter were notably narrower in both the gestational hypertension (GH) and preeclampsia (PE) groups when compared with the healthy control group (all p < 0.001), and there was no significant difference between the two disease groups. In univariate analyses, serum uric acid exhibited a stronger negative correlation with the arterial diameter (r = -0.614) than diastolic blood pressure (r = -0.436). In multivariable generalized estimating equation (GEE) models that accounted for inter - eye correlation and maternal covariates, serum uric acid remained independently associated with a narrower retinal arterial caliber, while diastolic blood pressure was independently associated with a reduced macular venous caliber. No independent associations were detected for the arteriovenous ratio or fractal dimension. Vascular density and the arterial branching angle were significantly correlated with 24-h urinary protein excretion, and the arterial branching angle was the sole parameter showing a between - group difference between GH and PE. These findings imply that the narrowing of the retinal vascular caliber represents a consistent microvascular alteration in hypertensive disorders of pregnancy (HDP), with differential arterial - venous association patterns observed in multivariable analyses. AI - assisted quantitative retinal assessment offers an objective framework for investigating systemic microvascular alterations in HDP and requires validation in larger longitudinal studies.