Interventions to reduce harm to women and their children from untreated low levels of thyroid hormone in pregnancy

The thyroid is a butterfly-shaped gland at the front of the oesophagus/throat that produces thyroid hormone. Thyroid hormone helps the body to make energy, keeps body temperature regulated and assists other organs in their functions. Hypothyroidism (a deficiency of thyroid hormone) is a relatively common illness that can cause fatigue, constipation, muscle cramps and weakness, hair loss, dry skin, intolerance to cold, depression and weight gain. Medication is with levothyroxine. Selenium is a trace element that changes the expression of selenoproteins. These act as antioxidants and appear to decrease thyroid inflammation in autoimmune thyroiditis. Pregnant women with subclinical hypothyroidism have abnormal thyroid hormone levels but no symptoms. They are at a increased risk of miscarriage, pre-eclampsia and preterm birth with impaired neuropsychological development in the child.

We identified four randomised studies involving only 362 women with hypothyroidism. In one trial of 115 women with thyroid autoantibodies but normal thyroid hormone levels, levothyroxine clearly reduced the risk of preterm birth by 72% compared with no treatment. The risk of women developing pre-eclampsia was not reduced, but there was a trend toward a reduction in miscarriage. In a study of 169 women with autoimmune hypothyroidism, supplementation with selenium did not decrease preterm birth rates or pre-eclampsia, but appeared to reduce moderate to severe inflammation of the thyroid gland and thyroid dysfunction after the birth. The third and fourth studies looked at different doses of levothyroxine on thyroid hormone levels.

Levothyroxine is an established treatment for women with symptomatic hypothyroidism, but it may also benefit women with low thyroid levels who do not have symptoms. Selenium also shows promise for women with hypothyroidism but needs further testing.

Authors' conclusions: 

This review found no difference between levothyroxine therapy and a control for treating pregnant euthyroid women with thyroid peroxidase antibodies for the outcome of pre-eclampsia, however a reduction in preterm birth and a trend towards reduced miscarriage with levothyroxine was shown. This review also showed no difference for pre-eclampsia or preterm birth when selenium was compared with placebo, however a promising reduction in postpartum thyroiditis was shown. Childhood neurodevelopmental delay was not assessed by any trial included in the review.

Given that this review is based on four trials of moderate risk of bias, with only two trials contributing data (n = 284), there is insufficient evidence to recommend the use of one intervention for clinical or subclinical hypothyroidism pre-pregnancy or during pregnancy over another, for improving maternal, fetal, neonatal and childhood outcomes.

Read the full abstract...

Over the last decade there has been enhanced awareness of the appreciable morbidity of thyroid dysfunction, particularly thyroid deficiency. Since treating clinical and subclinical hypothyroidism may reduce adverse obstetric outcomes, it is crucial to identify which interventions are safe and effective.


To identify interventions used in the management of hypothyroidism and subclinical hypothyroidism pre-pregnancy or during pregnancy and to ascertain the impact of these interventions on important maternal, fetal, neonatal and childhood outcomes.

Search strategy: 

We searched the Cochrane Pregnancy and Childbirth Group’s Trials Register (31 March 2013).

Selection criteria: 

Randomised controlled trials (RCTs) and quasi-randomised controlled trials that compared a pharmacological intervention for hypothyroidism and subclinical hypothyroidism pre-pregnancy or during pregnancy with another intervention or placebo.

Data collection and analysis: 

Two review authors assessed trial eligibility and quality and extracted the data.

Main results: 

We included four RCTs of moderate risk of bias involving 362 women. In one trial of 115 women, levothyroxine therapy to treat pregnant euthyroid (normal thyroid function) women with thyroid peroxidase antibodies was not shown to reduce pre-eclampsia significantly (risk ratio (RR) 0.61; 95% confidence interval (CI) 0.11 to 3.48) but did significantly reduce preterm birth by 72% (RR 0.28; 95% CI 0.10 to 0.80). Two trials of 30 and 48 hypothyroid women respectively compared levothyroxine doses, but both trials reported only biochemical outcomes. A trial of 169 women compared the trace element selenomethionine (selenium) with placebo and no significant differences were seen for either pre-eclampsia (RR 1.44; 95% CI 0.25 to 8.38) or preterm birth (RR 0.96; 95% CI 0.20 to 4.61). None of the four trials reported on childhood neurodevelopmental delay.

There was a non-significant trend towards fewer miscarriages with levothyroxine, and selenium showed some favourable impact on postpartum thyroid function and a decreased incidence of moderate to advanced postpartum thyroiditis.