Evaluation of retinal and choroidal microvasculature parameters by OCTA in patients with premature ovarian...

Postmenopausal cohorts show reduced vessel density on optical coherence tomography angiography (OCTA), yet OCTA data specific to premature ovarian insufficiency (POI) are lacking. We aimed to characterize retinal and choroidal microvasculature in women with POI using OCTA and to compare these findings with age-matched women with regular menses.

In this cross-sectional study, 36 women with POI and 52 healthy controls underwent right-eye best-corrected visual acuity testing, Schirmer I, tear breakup time (TBUT), and OCTA (RTVue XR Avanti/AngioVue). Quantitative outputs included macular superficial capillary plexus/deep capillary plexus (SCP/DCP) vessel and flow indices, choriocapillaris flow density, subfoveal choroidal thickness (SFCT), choroidal vascularity index (CVI), and optic nerve head (ONH)/radial peripapillary capillary (RPC) flow area. Imaging was standardized and analyses were masked.

POI participants had significantly lower TBUT (4.41 ± 0.69 vs 12.32 ± 2.21 s; p = 0.001) and Schirmer scores (6.86 ± 2.20 vs 13.76 ± 3.70 mm; p < 0.001). SCP/DCP metrics (flow density, non-perfusion, mean and parafoveal vessel density) and retinal nerve fiber layer (RNFL) thickness did not differ between groups (all p > 0.05). In contrast, POI showed reduced choriocapillaris flow density (1.75 ± 0.11 vs 1.92 ± 0.04 mm²; p < 0.001), thinner SFCT (282.61 ± 43.50 vs 305.36 ± 32.61 µm; p = 0.006), and lower macular CVI (0.624 ± 0.04 vs 0.699 ± 0.04; p = 0.001). At the disc/peripapillary region, ONH vessel density and RPC vessel density were similar (all p > 0.05), but flow areas were smaller in POI: ONH flow area (1.59 ± 0.06 vs 1.67 ± 0.04 mm²; p < 0.001), RPC flow area (1.39 ± 0.12 vs 1.47 ± 0.18 mm²; p = 0.011), and choroidal flow area (1.50 ± 0.09 vs 1.57 ± 0.09 mm²; p < 0.001). Temporal peripapillary choroidal thickness was also lower in POI (177.05 ± 28.33 vs 211.88 ±  32.89 µm; p < 0.001), whereas nasal thickness and nasal/temporal disc CVI were not different (both p > 0.05).

To our knowledge, this is the first OCTA study in POI, demonstrating a consistent microvascular pattern-lower choriocapillaris flow, reduced SFCT, and diminished ONH/RPC flow areas-despite preserved SCP/DCP metrics and RNFL thickness. These findings are compatible with estrogen-related microvascular dysregulation and suggest that OCT/OCTA can sensitively detect subclinical ocular perfusion changes in hypoestrogenic states. Confirmation in larger, longitudinal cohorts is needed to establish prognostic value and to test modifiability (e.g., with hormone therapy).