Thyroid antibodies raise miscarriage risk, but levothyroxine did not improve births

Many women are told they carry thyroid peroxidase antibodies — a sign the immune system is reacting to the thyroid gland — after tests done for miscarriage or fertility problems. Those antibodies are linked to a higher risk of miscarriage and preterm birth even when thyroid hormone levels themselves test normal [s1]. It is a natural next step to ask whether a low dose of the thyroid hormone levothyroxine, a cheap and familiar pill, might lower that risk. Two randomised trials have now tested exactly that, and both found no benefit for live births [s1][s2].

What the larger trial did

The TABLET trial, run across 49 hospitals in the United Kingdom, screened 19,585 women who were trying to conceive and had either miscarried before or been diagnosed with infertility [s1]. From that group it identified women who had thyroid peroxidase antibodies but normal thyroid function, and randomly assigned 952 of them to take either 50 μg of levothyroxine once daily or a matching placebo, starting before conception and continuing to the end of pregnancy [s1].

The measure that mattered was live birth after at least 34 weeks of gestation. Follow-up for that outcome was near-complete, at 98.7% (940 of 952 women) [s1]. Similar proportions became pregnant — 266 of 470 women (56.6%) on levothyroxine and 274 of 470 (58.3%) on placebo — and the live-birth rate was almost identical: 37.4% (176 of 470) with levothyroxine versus 37.9% (178 of 470) with placebo [s1]. The relative risk was 0.97 (95% confidence interval 0.83 to 1.14, P=0.74), and the absolute difference was -0.4 percentage points (95% CI -6.6 to 5.8) [s1]. There were no significant differences in pregnancy loss, preterm birth or newborn outcomes [s1].

A second trial reached the same place

A separate trial, T4LIFE, asked a narrower version of the question: does levothyroxine help women with antibodies and a history of recurrent pregnancy loss specifically? Run in the Netherlands, Belgium and Denmark, it enrolled women aged 18 to 42 who were antibody-positive, had two or more pregnancy losses and had normal thyroid function, assigning them to levothyroxine or placebo before conception [s2].

The trial was stopped early because recruitment was slow, with 187 of a planned 240 women included [s2]. Among them, 47 of 94 women (50%) in the levothyroxine group and 45 of 93 (48%) in the placebo group had a live birth — a relative risk of 1.03 (95% CI 0.77 to 1.38) and an absolute difference of 1.6 percentage points (95% CI -12.7 to 15.9) [s2]. The investigators' conclusion was direct: they do not advise routine levothyroxine for antibody-positive women with recurrent pregnancy loss and normal thyroid function [s2].

What this does and does not mean

The two trials answer one specific question and answer it consistently: for women whose thyroid hormone levels are already normal, adding levothyroxine because of positive antibodies does not raise the chance of a baby [s1][s2]. That is worth knowing, because the pill is easy to prescribe and the antibodies are common, so it is the kind of low-effort intervention that spreads before the evidence catches up.

None of this changes the picture for women who actually have an underactive thyroid. Overt hypothyroidism in pregnancy is a different situation with clear reasons to treat, and these trials deliberately enrolled women with normal thyroid function to isolate the effect of the antibodies alone [s1][s2]. The results also do not say antibodies are harmless — the raised risk of miscarriage and preterm birth is real [s1]. What they say is that a levothyroxine tablet is not the lever that removes it.

Safety was reassuring but not perfectly clean. In TABLET, serious adverse events occurred in 5.9% of the levothyroxine group and 3.8% of the placebo group, a difference that was not statistically significant (P=0.14) [s1]; in T4LIFE, adverse events were reported in 7% and 8% of the two groups, none tied to the study procedure [s2].

The liftable summary: positive thyroid antibodies with normal thyroid function are a marker of higher miscarriage risk, but two randomised trials found that treating them with levothyroxine did not improve live-birth rates. A raised antibody result on its own is not a reason to start the drug.

This article is informational and is not medical advice. Thyroid testing and treatment in pregnancy should be decided with a clinician.