Research Question

Does initiating levothyroxine treatment based on Thyroid Stimulating Hormone (TSH) > 2.5 mIU/L or thyroid autoimmunity improve pregnancy continuation rates in recurrent pregnancy loss (RPL) patients?


We conducted a retrospective cohort study of 1064 RPL patients. Subjects were classified as either euthyroid (TSH 0.1 - ≤ 2.5 mIU/L), borderline-subclinical hypothyroid (borderline-SCH, TSH 2.5 – ≤ 4 mIU/L), or subclinical hypothyroid (SCH, TSH 4 – ≤ 10 mIU/L). For subjects with ≥2 pregnancy losses and a subsequent pregnancy with known outcome, we compared the pregnancy continuation rate past 10 weeks of treated and untreated borderline-subclinical hypothyroid (n = 98) and untreated euthyroid (n = 279) subjects, and between subjects with positive and negative thyroid peroxidase antibody (TPOAb) tests.


72% were euthyroid (721/992), 19.4% (192/992) were borderline-SCH, and 5.4% (54/992) were subclinically hypothyroid (SCH). 21% of subjects received treatment with levothyroxine. 58.7% of subjects had a TPOAb test, which was positive in 9.25% (37/400) in euthyroid, 16.5% (22/133) in borderline-SCH subjects, and 35.3% (12/34) in SCH subjects. Treatment did not improve pregnancy continuation rates in borderline-SCH subjects (p = 0.392). There was no difference in pregnancy outcomes based on TPOAb status and treatment for borderline-SCH subjects (p = 0.868), or based on TPOAb status for euthyroid subjects (p = 0.267).


Treatment of hypothyroidism in pregnancy should be initiated based on a TSH > 4 mIU/L. Treatment initiation based on thyroid autoimmunity or a TSH > 2.5 mIU/L may result in overtreatment.


Recurrent pregnancy loss
Subclinical hypothyroidism
Pregnancy loss
thyroid peroxidase antibodies
Thyroid Stimulating Hormone, autoimmune hypothyroidism