Long-term risk of hypothyroidism after thyroid abnormalities identified in pregnancy

When thyroid analytes are assessed during pregnancy, subclinical hypothyroidism and thyroid peroxidase antibodies may be identified in otherwise asymptomatic individuals. However, the long-term risk of progression to overt hypothyroidism in this population remains poorly defined. Understanding this trajectory is essential for informing long-term thyroid surveillance strategies.

To determine the long-term incidence and timing of overt hypothyroidism among individuals with subclinical hypothyroidism, thyroid peroxidase antibodies, or both identified during pregnancy and to compare that risk against matched normothyroid controls.

This retrospective cohort study included 718 individuals with subclinical hypothyroidism, thyroid peroxidase antibodies, or both identified during early pregnancy at a single tertiary-care public health system between 2000 and 2003. Participants were followed for up to 25 years, with a median follow-up time of approximately 21 years. Overt hypothyroidism was defined by an elevated thyroid-stimulating hormone with low free thyroxine, a documented diagnosis, or initiation of levothyroxine. Cumulative incidence was compared across antenatal thyroid phenotype groups using Kaplan-Meier methods and Cox proportional hazards models adjusted for age. A nested matched cohort analysis compared outcomes between individuals with antenatal thyroid abnormalities and normothyroid controls.

During the follow-up period, 238 of 718 participants (33.1%) developed overt hypothyroidism. Progression occurred in 82 of 155 (52.9%) individuals with both subclinical hypothyroidism and thyroid peroxidase antibodies, compared with 27 of 108 (25.0%) with subclinical hypothyroidism alone and 129 of 455 (28.4%) with thyroid peroxidase antibodies alone (P<.001). Individuals with both abnormalities had the highest cumulative incidence and the shortest time to progression. In the nested matched cohort analysis, overt hypothyroidism developed in 34 of 118 (28.8%) individuals with antenatal thyroid abnormalities compared with 12 of 118 (10.2%) normothyroid controls (P<.001).

Approximately one in 3 individuals with subclinical hypothyroidism, thyroid peroxidase antibodies, or both during pregnancy developed overt hypothyroidism over 25 years-nearly 3 times the rate of matched normothyroid controls. The risk was greatest among those with both abnormalities. These findings support consideration of phenotype-guided long-term thyroid surveillance in this population.