PURPOSE: To study retinal blood flow (RBF) during pregnancy in subjects with and without diabetes and to relate the changes to progression in diabetic retinopathy. METHODS: RBF in a major temporal retinal vein was measured, where possible, during all three trimesters (T1 to T3) and the postpartum period (PP) using laser Doppler velocimetry (measuring velocity [V]) and monochromatic fundus photographs (measuring diameter [D]). RESULTS: In the subjects without diabetes (n = 19), no significant change in RBF was demonstrated. V was significantly greater in T3 than in PP (P = 0.01). D was significantly smaller in T1 to T3 than in PP (P < or = 0.01). RBF in the subjects with diabetes was significantly higher in T2 and T3 than in PP (P < 0.05). V increased from T1 to T2 (P = 0.04) and decreased from T2 to PP (P = 0.001) and from T3 to PP (P = 0.002). The only significant change in D was a smaller value in T1 than in PP (P = 0.003). However, it was only those subjects whose DR progressed (n = 11) who had a significant increase in RBF during pregnancy (P = 0.0001). CONCLUSIONS: These data suggest that the hyperdynamic circulation present in pregnancy led to compensatory retinal vessel constriction (autoregulation) in the group without diabetes; this was largely absent in the patients with diabetes, with resultant increase in retinal blood flow associated with worsening retinopathy. This finding lends further support to the hypothesis that increased retinal blood flow is a pathogenic mechanism for diabetic retinopathy.
PURPOSE: To study retinal blood flow (RBF) during pregnancy in subjects with and without diabetes and to relate the changes to progression in diabetic retinopathy. METHODS: RBF in a major temporal retinal vein was measured, where possible, during all three trimesters (T1 to T3) and the postpartum period (PP) using laser Doppler velocimetry (measuring velocity [V]) and monochromatic fundus photographs (measuring diameter [D]). RESULTS: In the subjects without diabetes (n = 19), no significant change in RBF was demonstrated. V was significantly greater in T3 than in PP (P = 0.01). D was significantly smaller in T1 to T3 than in PP (P < or = 0.01). RBF in the subjects with diabetes was significantly higher in T2 and T3 than in PP (P < 0.05). V increased from T1 to T2 (P = 0.04) and decreased from T2 to PP (P = 0.001) and from T3 to PP (P = 0.002). The only significant change in D was a smaller value in T1 than in PP (P = 0.003). However, it was only those subjects whose DR progressed (n = 11) who had a significant increase in RBF during pregnancy (P = 0.0001). CONCLUSIONS: These data suggest that the hyperdynamic circulation present in pregnancy led to compensatory retinal vessel constriction (autoregulation) in the group without diabetes; this was largely absent in the patients with diabetes, with resultant increase in retinal blood flow associated with worsening retinopathy. This finding lends further support to the hypothesis that increased retinal blood flow is a pathogenic mechanism for diabetic retinopathy.