When to Treat Subclinical Hypothyroidism
Treat subclinical hypothyroidism when TSH is >10 mU/L, or when TSH is between the upper limit of normal and 10 mU/L in specific high-risk populations including pregnant women, women planning pregnancy, those with infertility, and younger adults (<65-70 years) with cardiovascular disease or risk factors. 1
General Population Approach
TSH >10 mU/L
Always treat with levothyroxine regardless of symptoms or antibody status. This threshold represents profound subclinical hypothyroidism with clear progression risk and potential cardiovascular harm. 1, 2
TSH 4.5-10 mU/L (Mild Subclinical Hypothyroidism)
First, confirm the diagnosis - repeat TSH and free T4 in 4-6 weeks, as 62% of elevated TSH levels normalize spontaneously. 1, 3
If persistent, treatment decisions depend on:
DO NOT treat routinely in the general adult population, as randomized controlled trials show no improvement in symptoms or cognitive function when TSH <10 mU/L. 3, 4
Consider treatment in these specific scenarios:
- Positive TPO antibodies (higher progression risk to overt hypothyroidism)
- Symptomatic patients with clear hypothyroid symptoms (though evidence for benefit is weak)
- Cardiovascular disease or risk factors in patients <65 years old
- Dyslipidemia requiring optimization 5, 2, 6
Age-Specific Considerations
Elderly patients (≥70 years): Generally do not treat subclinical hypothyroidism regardless of TSH level. TSH naturally rises with age (upper limit of normal reaches 7.5 mU/L by age 80), and treatment may cause harm including increased cardiovascular events and atrial fibrillation. 3, 5
**Younger adults (<65 years):** Lower threshold for treatment, particularly with TSH >7 mU/L and cardiovascular risk factors, as observational data suggests potential cardiovascular benefit. 4, 7
Pregnancy and Preconception
This represents a critical exception where treatment thresholds are substantially lower due to fetal neurodevelopment concerns.
First Trimester
Treat if TSH >4.0 mU/L (not the traditional 2.5 mU/L cutoff from older guidelines). Recent evidence shows adverse outcomes increased with treatment when TSH was 2.5-4.0 mU/L, while benefits emerged at TSH >4.0 mU/L. 8
- TSH >10 mU/L: Always treat (considered overt hypothyroidism in pregnancy)
- TSH 4.0-10 mU/L with normal free T4: This is subclinical hypothyroidism in pregnancy
- Consider treatment - evidence shows 55% reduction in miscarriage risk when TSH 4.1-10 mU/L
- Stronger indication if TPO antibody positive - 70% reduction in preterm birth and 83% reduction in neonatal admission 8
Second and Third Trimester
Do not initiate treatment for mild subclinical hypothyroidism diagnosed after the first trimester. Strong evidence shows no benefit and potential harm. Instead, recheck TSH in 4-6 weeks. 1
The critical window for fetal brain development is early first trimester; late treatment provides no neurodevelopmental benefit.
Women Planning Pregnancy or With Infertility
TSH >10 mU/L: Treat with levothyroxine before attempting conception. 1
TSH 4.0-10 mU/L:
- Confirm with repeat testing in 4-6 weeks
- If persistent, may start low-dose levothyroxine (25-75 mcg/day) targeting TSH 0.5-2.5 mU/L
- Delay fertility treatment until euthyroid if newly diagnosed 1
Important: Do NOT treat euthyroid TPO antibody-positive women. Three high-quality randomized trials since 2017 show no benefit. Instead, monitor TSH every 3-6 months preconception. 1
Treatment Targets
When treatment is initiated:
- Pregnancy/preconception: TSH 0.5-2.5 mU/L 1
- General population: TSH within normal reference range (typically 0.5-4.5 mU/L) 6
- Avoid overtreatment: Suppressed TSH increases atrial fibrillation and osteoporosis risk 9
Common Pitfalls
- Treating single elevated TSH without confirmation - Always repeat testing as most normalize spontaneously
- Overtreating elderly patients - Age-adjusted TSH reference ranges are critical
- Using outdated 2.5 mU/L pregnancy cutoff - Current evidence supports 4.0 mU/L threshold
- Treating asymptomatic mild subclinical hypothyroidism expecting symptom improvement - Placebo-controlled trials show no benefit
- Initiating treatment in second/third trimester - This causes harm without benefit