Skip to main content
ENDO 2026 Main banner

Thyroid Biology and Disease

ORF41-04 - Levothyroxine Treatment in Women with High-Normal Thyrotropin, Thyroperoxidase Antibodies or Both, Prior to or during Early Pregnancy: An Individual Participant Data Meta-Analysis of 8 Randomised Trials

Monday, June 15, 2026
9:45 AM - 10:00 AM CT
Location: Room W183A
Disclosure(s):
Robin Peeters, MD,PhD: No financial relationships to disclose
Kristien Boelaert, MD, PhD: No relevant disclosure to display

Background
Subclinical hypothyroidism and thyroperoxidase antibody (TPOAb) positivity are associated with subfertility, miscarriage, and preterm birth. Despite widespread use of levothyroxine (LT4) in women of reproductive age with high-normal thyrotropin (TSH) concentrations or TPOAb positivity, evidence for benefit remains uncertain. We evaluated whether LT4 therapy initiated preconception or in early pregnancy improves miscarriage and live birth outcomes in women with high-normal TSH, TPOAb positivity or both.

Methods
We conducted an individual-participant data meta-analysis of randomised controlled trials (RCTs in which LT4 was started preconception or during the first trimester in women with TSH concentrations < 5.0 mU/L, with or without TPOAb positivity. Primary outcomes were miscarriage ( < 24 weeks) and live birth (≥34 weeks). We used one-stage generalised linear mixed models with random intercepts for each RCT or two-stage random-effects meta-analyses where appropriate.

Findings
A total of 2,360 women from eight randomized trials were included (1,166 LT4; 1,194 placebo or no treatment). In those with a TSH between 2.5-5.0 mU/L, the rate of miscarriage was 14.6% (70/480) in the LT4 group versus 21.7% in the control group (odds ratio [OR] 0.57, 95% CI 0.40–0.81), and the live birth rate was 72.1% (405/563) in the LT4 group versus 68.9% (399/579) in the control group (OR 1.32, 0.98--1.78).

However, the favourable effect of LT4 was only present amongst women with a history of recurrent miscarriages and not in those with one or no previous miscarriage. In women with ≥2 previous miscarriages and high-normal TSH, the rate of miscarriage was .25.4% (36/142) in the LT4 group versus .46.1% (70/152) in the control group (OR 0.39, 0.24–0.64 vs OR 1.09, 0.72-1.65 for ≤1 miscarriages), and the live birth rate was .58.9% (103/175) in the LT4 group versus .46.5% (80/172) in the control group (OR 1.65, 1.07–2.53 vs OR 0.94, 0.73-1.22 for ≤1 miscarriages).

Interestingly, we identified a biological gradient for TSH whereby a higher TSH at diagnosis was associated with a larger reduction in miscarriage risk, without a clear risk-threshold.

Interpretation
LT4 therapy does not confer universal benefit in women with a high-normal TSH or thyroid autoimmunity, but it offers clinically meaningful benefit in women with recurrent miscarriages. In women with recurrent miscarriages, LT4 might also have beneficial effects for those with a TSH < 2.5 mu/L. Future studies are needed to define the optimal LT4 treatment decision threshold and further explore the underlying biological mechanisms of these effects.