The literature concerning SCH and pregnancy is vast, and often contradictory, leading to confusion among clinicians. Contributing to the confusion is that some observational studies report a modest association between SCH and adverse pregnancy outcomes. To date, however, randomized clinical trials show no benefit of thyroxine treatment in these cases. I explore these contradictory pieces of evidence below.
Is SCH associated with adverse pregnancy outcomes due to low thyroxine levels?
There is conflicting literature about the association of SCH and adverse reproductive outcomes. A meta-analysis of 47,045 pregnant women reported that the preterm birth rate for women with SCH and euthyroid women (normal TSH and normal free thyroxine levels) was 6.1% and 5.0%, respectively (odds ratio [OR], 1.29; 95% CI, 1.01–1.64).4 Interestingly, pregnant women with normal TSH levels but a low free thyroxine level also had an increased rate of preterm birth (7.1% vs 5.0%; OR, 1.46; 95% CI, 1.12–1.90).
Although observational studies report an association between SCH and adverse reproductive outcomes, multiple randomized clinical trials conducted in women with SCH or hypothyroxinemia have failed to demonstrate that thyroxine replacement improves reproductive outcomes. For example, in a study of 794 pregnant women with elevated TSH and/or low free thyroxine levels randomly assigned to thyroxine treatment (0.15 mg daily) or no treatment, there was no difference in preterm birth rate (5.6% vs 7.9%, P = .2), mean birth weight (3.5 kg vs 3.3 kg, P = .15), gestational age at delivery (40.1 vs 40.2 weeks, P = .10), or the intelligence quotient of children at 3 years (99 vs 100, P = .40).5
In another study, 674 pregnant women with mild SCH (mean TSH, 4.4 mU/L) were randomly assigned to receive thyroxine (0.1 mg daily and dose adjusted to achieve a normal TSH level) or placebo. In this study there was no difference between the thyroxine treatment or placebo groups in preterm birth rate (9% vs 11%, P = .44), gestational age at delivery (39.1 vs 38.9 weeks, P = .57) or intelligence quotient of children at 5 years (97 and 94, P = .71).6
The same investigators also randomized 524 pregnant women with isolated hypothyroxinema (mean free thyroxine level, 0.83 ng/dL) and normal TSH level (mean, 1.5 mU/L) to thyroxine (0.05 mg daily and dose adjusted to achieve a normal free thyroxine level) or placebo.6 In this study there was no difference in preterm birth rate (12% vs 8%, P = .11), gestational age at delivery (39.0 vs 38.8 weeks, P = .46) or intelligence quotient of children at 5 years (94 and 91, P = .31).6
When large randomized clinical trials and observational studies report discrepant results, many authorities prioritize the findings from the randomized clinical trials because those results are less prone to being confounded by unrecognized factors. Randomized trials do not demonstrate that mild SCH or isolated hypothyroxinemia have a major impact on pregnancy outcomes.
Thyroid antibodies, fertility, miscarriage, and preterm birth
Some observational studies report that the presence of thyroid antibodies in a euthyroid woman reduces fecundity and increases the risk for miscarriage and preterm birth. For example, a meta-analysis of 47,045 pregnant women reported that the preterm birth rate for women with and without antithyroid antibodies was 6.9% and 4.9%, respectively (OR, 1.33; 95% CI, 1.15–1.56). However, in euthyroid women with antithyroid antibodies, low-dose thyroxine therapy has not been shown to improve fertility, or reduce miscarriages or preterm birth rate.
Continue to: In a large randomized clinical trial, 952 euthyroid women...