Abstract
Purpose: To identify retinal morphologic changes associated with pregnancy in patients with diabetes and evaluate potential risk factors for disease progression. Methods: A retrospective chart and image review was performed for 35 patients (70 eyes) with type 1 or type 2 diabetes mellitus (DM) who received prenatal care and had a delivery between January 2010 and December 2022 at Duke University Health System. Multimodal retinal imaging (spectral-domain optical coherence tomography, color fundus photography, and fluorescein angiography [FA]) was assessed pre-pregnancy and post-delivery for morphologic changes, including disorganization of the retinal inner layers (DRIL), diabetic retinopathy (DR) severity (based on Early Treatment Diabetic Retinopathy Study letter level), and central subfield thickness. Clinical data were extracted via chart review. Results: DRIL prevalence increased from 75.5% of eyes (37/49) pre-pregnancy to 88.1% (52/59) post-delivery (P < .001). Average DRIL extent nearly doubled (38.0% pre-pregnancy to 76.4% post-delivery; P < .001), and 38.6% of eyes (27/70) showed new or progressive DRIL after pregnancy. DR prevalence increased from 72.7% of eyes (40/55) pre-pregnancy to 92.1% (58/63) post-delivery, with a significant increase in proliferative DR prevalence after pregnancy (P < .001). Overall, the severity of both proliferative and nonproliferative DR significantly increased following pregnancy (both P < .001). DRIL progression from pre-pregnancy to post-delivery correlated with DR progression (P = .039). No significant associations were found between DR progression and baseline age, DM type, smoking, glycosylated hemoglobin level, body mass index, preeclampsia history, or hypertension treatment. New or worsening retinal ischemia (characterized as foveal avascular zone enlargement and/or peripheral nonperfusion) was identified on the longitudinal FA images available from 4 patients. Conclusions: Pregnancy is associated with the progression of DRIL and DR in diabetic women. Early identification of changes is important for timely intervention and optimizing visual outcomes.