BMC Pregnancy Childbirth. 2026 Feb 23. doi: 10.1186/s12884-026-08818-7. Online ahead of print.
ABSTRACT
BACKGROUND: Gestational diabetes mellitus (GDM) may be associated with occult abnormalities in choroidal microcirculation.This study utilizes ultra-widefield swept-source OCTA (SS-OCTA) and introduces the three-dimensional choroidal vascularity index (3D-CVI) as a volumetric quantitative parameter to systematically characterize the choroidal microcirculatory characteristics in GDM.
METHODS: This cross-sectional study utilized SS-OCTA with a 24 × 20 mm scan range. A 3 × 3 grid (9 regions) was applied to assess an effective area of interest measuring 17 × 17 mm. Built-in automated algorithms quantified choriocapillaris plexus vascular density (CCP VD), 3D-CVI, and choroidal thickness (CT). Statistical analyses included one-way ANOVA and general linear models, with Pearson/Spearman correlation for association analysis.
RESULTS: 138 women were enrolled (n = 46/group), comprising the GDM, pregnant controls (PC), and non-pregnant controls (NC), with mean ages of 31.98 ± 4.43, 30.74 ± 3.66, and 31.00 ± 4.26 years.Significant differences were observed in macular grid CCP VD among the three groups (46.26 ± 1.41%, 46.28 ± 1.43%, and 45.41 ± 1.85%; P = 0.012), after adjusting for age and axial length(AL), at Age = 35.38 years, GDM was higher than both PC and NC (P = 0.005,P < 0.001), and PC was higher than NC (P = 0.033). In pregnancy, GDM showed significantly higher macular grid CCP VD than PC at longer AL (AL = 23.98 mm, P = 0.015). For 3D-CVI, significant differences were found in the macular grid (40.37 ± 3.30%,41.50 ± 3.23%,41.96 ± 2.70%,P = 0.043)/nasal-superior(39.59 ± 2.90%,40.65 ± 2.69%,41.43 ± 2.71%;P = 0.007);After adjustment, GDM remained lower than NC at the mean age level(macular grid: P = 0.047; nasal-superior: P = 0.012), while GDM and PC showed no significant difference.No significant differences were found in CT across regions (all P > 0.05).Correlation analysis showed inverse associations of age with mean CCP VD, 3D-CVI, and CT (r = – 0.56/-0.49/-0.47,all P < 0.05). Gestational age was negatively correlated with CCP VD and CT (r = – 0.46/-0.23,all P < 0.05). AL, mean ocular perfusion pressure, fasting plasma glucose, and glycated hemoglobin showed no significant correlations ( all P > 0.05).
CONCLUSIONS: Macular grid CCP-VD increased overall in pregnancy versus NC.In pregnancy, GDM exceeded PC at higher age or AL.3D-CVI differed only in macular grid and nasal-superior, lower in GDM than NC yet not from PC.Given the structural differences between mild-to-moderate GDM and PC may be subtle.3D-CVI may serve as an auxiliary indicator for regional choroidal remodeling, and combined with CCP, it can comprehensively assess the choroidal pathophysiology in GDM.Longitudinal gestation-to-postpartum follow-up is needed.
PMID:41731441 | DOI:10.1186/s12884-026-08818-7
