A meta-analysis found that pregnant women with normal thyroid function but circulating thyroid auto-antibodies have an increased risk of miscarriage and pre-term birth that may be marked; treatment with levothyroxine may reduce the risk.

Miscarriage and preterm birth are among the commonest complications of pregnancy. There is evidence that thyroid auto-immunity is a risk factor for miscarriage and premature birth: it is also relatively common, occurring in between 6% and 20% in an unselected population of women, and 17% to 33% in women with a history of recurrent pregnancy loss. The authors of this paper carried out a systematic review and meta-analysis to clarify the evidence on the association and determine whether treatment with levothyroxine was beneficial. They carried out a comprehensive literature search for studies that evaluated the association between thyroid auto-antibodies and pregnancy outcomes. They then pooled odds ratios (OR) from individual studies with the random effects model, separating those for cohort and case-control studies.

The literature search identified 30 articles with 31 studies (19 cohort and 12 case-control; n=12,126) that assessed the association between thyroid auto-antibodies and miscarriage. Five studies (n=12,566) evaluated the association with preterm birth.

Of the 31 studies evaluating miscarriage, 28 showed a positive association between thyroid auto-antibodies and miscarriage. Meta-analysis of the cohort studies showed that the odds of miscarriage were nearly tripled with the presence of thyroid auto-antibodies (OR 3.90; 95% CI 2.48 to 6.12; P<0.001). For case-control studies the OR for miscarriage was 1.80 (95% CI 1.25 to 2.60; P=0.002). Risk of preterm birth also increased by double with the presence of thyroid auto-antibodies (OR 2.07; 95% CI 1.17 to 3.68; P=0.01).

Two randomised studies (n=187) evaluated the effect of treatment with levothyroxine on miscarriage. Both showed a fall in miscarriage rates, and meta-analysis showed a significant 52% relative risk reduction in miscarriages with levothyroxine (relative risk [RR] 0.48; 95% CI 0.25 to 0.92; P=0.03). One of these (n=115) also reported on the effect of levothyroxine on the rate of preterm birth, and noted a 69% relative risk reduction (RR 0.31; 95% CI 0.11 to 0.90).

The authors conclude that presence of maternal thyroid auto-antibodies is strongly associated with miscarriage and preterm delivery, and discuss possible mechanisms for this effect. There is evidence that treatment with levothyroxine can attenuate the risks, but this is based on only two controlled trials with a small number of patients, both carried out in the same unit. They recommend that a large, placebo-controlled randomised trial is needed, with live birth as the primary outcome.

An accompanying Editorial comments on the study. The authors discuss possible mechanisms, and call for more research both on the mechanisms involved and on treatment.