Endocrine Abstracts

Objectives: An increased risk of birth defects with the use of antithyroid drugs (ATDs) in early pregnancy has been found in large epidemiological studies. It remains debatable whether this risk associates with the medical treatment, maternal thyroid function or thyroid autoimmunity. We aimed to evaluate the risk of birth defects in relation to maternal thyroid function and thyroid autoantibodies in early pregnancy.

Methods: Retrospective study including pregnant women and their live-born children from the Danish National Birth Cohort (DNBC), 1997-2003, and the North Denmark Region Pregnancy Cohort (NDRPC), 2011-2015. Pregnant women with known thyroid disease were not included to exclude any exposure to ATDs. Blood samples drawn in early pregnancy (median week 10) were used for retrospective assessment of maternal TSH and free T4 in both cohorts (Dimension Vista and ADVIA Centaur, Siemens Healthineers). Thyroid peroxidase antibodies (TPO-Ab) and thyroglobulin antibodies (Tg-Ab) were assessed in the full NDRPC as well as TSH-receptor antibodies (TRAb) in a sub-cohort of the NDRPC. The applied cut-offs were TPO-Ab > 60 U/ml, Tg-Ab > 33 U/ml, and TRAb > 0.27 U/ml. Information on birth defects was obtained via linkage to Danish registers that also held information on potential confounders. Logistic regression was used to calculate adjusted odds ratio (aOR) with 95% confidence interval (95% CI) for birth defects in the pooled cohort investigating the continuous explanatory variables of TSH and free T4.

Results: Altogether 20,399 live-born pregnancies were studied including 7,433 pregnancies from the DNBC and 12,966 from the NDRPC. A total of 702 children (3.4%) had birth defects in the pooled cohort, and maternal median TSH (no birth defects; 1.13 mIU/l, birth defects; 1.19 mIU/l, P = 0.1) and free T4 (no birth defects; 15.5 pmol/l, birth defects; 15.7 pmol/l, P = 0.3) did not differ by outcome of birth defects, which was substantiated in adjusted analyses (TSH; aOR 1.02 (95% CI: 0.99-1.05), free T4; 1.01 (0.98-1.05)). In the NDRPC, the prevalence of birth defects was overall 4.1% and did not differ by maternal autoantibody status (TPO-Ab positive; 4.1%, P = 0.9; Tg-Ab positive; 4.4%, P = 0.4, TRAb positive; 3.9%, P = 0.7).

Conclusions: In a large cohort of Danish pregnant women, no evidence was found that maternal thyroid function or thyroid autoantibodies in early pregnancy associate with the risk of birth defects. Results inform the debate regarding the underlying mechanisms for birth defects in children born to mothers with thyroid disease.