Women with type 1 diabetes are at a greater risk for adverse pregnancy outcomes (APO) compared with the general population, though factors relating to the occurrence of APO have not been well studied. We thus assessed predictors of APO among women with type 1 diabetes. Female participants (n=325; mean baseline age 28, duration 19 years) of a prospective cohort of childhood onset type 1 diabetes were selected for study. Among 320 with pregnancy data available, all pregnancies occurred after study entry (1986-88) for 61 (mean baseline age 21 and duration 14 years). APO were defined as tubal/ectopic pregnancies, spontaneous or induced abortions and stillbirths. Follow-up time was calculated as time in years from study entry to the date of the first adverse pregnancy outcome or date of last live birth for those with normal pregnancies. During 25 years of follow-up, 28 (45.9%) had at least one APO. With the exception of a greater number of total pregnancies in those with an APO (p=0.03), no other univariate differences were observed. In multivariable Cox models, a greater number of pregnancies (HR=1.49, p=0.03), log albumin excretion rate (HR=1.27, p=0.03) and proliferative retinopathy (HR=5.04, p=0.003) predicted APO. In multiple linear regression analyses, the number of pregnancies (β=0.58, p<0.0001) and the presence of proliferative retinopathy (β=0.75, p=0.01) directly predicted number of APO. In analyses excluding induced abortions from the outcome definition (19 adverse events (31.1%)), glomerular filtration rate was the only predictor of any APO (HR=0.98, p=0.08). A greater number of pregnancies (β=0.48, p<0.0001), non-HDL cholesterol (β=0.006, p=0.04), white blood cell count (β=-0.10, p=0.05) and proliferative retinopathy (β=0.74, p=0.02) predicted number of APO. Among women with type 1 diabetes, the presence of microvascular complications relates to both the occurrence and the number of subsequent adverse pregnancy outcomes.


T. Costacou: None. D. Rubinstein: None. T.J. Orchard: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.