Retinal vessel diameters confound the relationship of pregnancy to retinopathy and infant outcomes in T1D.

OBJECTIVES: To determine whether: 1) retinal vessel diameters in pregnant and non-pregnant women with type 1 diabetes (T1D) alter the relationship of pregnancy to severity of diabetic retinopathy (DR); and 2) retinal vessel diameters in early pregnancy alter the relationship of severity of DR in the mother to severe adverse outcome in the infant. METHODS: Two cohorts of women with T1D, one composed of pregnant women and the other of non-pregnant women of child-bearing age, were recruited in Wisconsin. Baseline examinations (including retinal photography and collection of diabetes-related characteristics) were conducted, with follow-up approximately one year later. Retinal images were graded according to the modified Airlie House classification protocol, and retinal vessel diameters were measured from digitized images. The latter were included in analyses as central retinal arteriolar equivalent (CRAE) and central retinal venular equivalent (CRVE). RESULTS: In multivariate models: 1) Pregnancy was significantly associated with both incidence (OR = 4.43, CI = 1.42-13.79) and progression (OR = 2.62, CI = 1.52-4.51) of DR. Neither CRAE nor CRVE were significant, but their addition modestly altered the associations of pregnancy to worsening DR; 2) Baseline retinopathy (OR = 1.28, CI = 1.05-1.57) was associated with severe adverse outcome in the infant. This association was unchanged by adjustment for retinal vessel diameters. CONCLUSIONS: Pregnancy is associated with worsening DR in women with T1D. DR severity in early pregnancy is associated with severe adverse outcome in the infant. The retinal vessel diameters CRAE and CRVE were not associated with these outcomes but were modest confounders of the association of pregnancy to worsening DR.