TSH and Fertility — What Your Thyroid Result Actually Means (India 2026)

Whatever your laboratory's reference range says is normal for a woman who is not pregnant. The American Society for Reproductive Medicine's 2024 guideline says that range should decide, that 4.12 mIU/L is a reasonable upper limit where a laboratory gives none, and that the lower thresholds used in pregnancy should not be applied before pregnancy.

TSH, thyroid-stimulating hormone, is the pituitary's signal telling the thyroid how hard to work. It rises when thyroid output falls, so a high TSH points to an underactive thyroid, not an overactive one, and it is easy to read the number the wrong way round. On its own it says less than it seems to: it is read with free T4 (whether the thyroid is keeping up) and sometimes with thyroid peroxidase (TPO) antibodies (whether the immune system is involved).

ASRM also advises telling women that a TSH between 2.5 and 4 mIU/L is not associated with a higher risk of miscarriage. A result in that band is not, on that guidance, a reason to delay trying or to postpone treatment.

It is a pregnancy number. Government of India guidance for pregnant women, the Ministry of Health's 2014 national guideline, uses trimester-specific ranges: up to 2.5 mIU/L in the first trimester and up to 3 mIU/L after that. Under that guideline, a pregnant woman at higher risk is screened and, if her TSH is above the range for her trimester, started on levothyroxine. A history of infertility is on its list of reasons to screen.

Before pregnancy the picture is different. About 95 in every 100 people without thyroid disease have a TSH below 2.5, but the normal range runs higher than that, so 2.5 is not the upper limit of normal (ASRM 2024). ASRM also notes that insisting on a TSH below 2.5 has caused significant delays in women starting fertility treatment or an IVF cycle.

The cut-off chosen changes how many women are given a label. In a study of pregnant women across 11 Indian cities, 13.13% were classed as hypothyroid using a TSH cut-off of 4.5, and 36.07% using trimester cut-offs of 2.5 and 3.0 (Dhanwal 2016). Nothing about the women changed between those two numbers; only the line did.

So it is reasonable to ask a clinic which range it is using and why. Using the Indian pregnancy ranges once you are pregnant is a legitimate national standard; applying them to a woman who is not yet pregnant is a choice the current fertility guidance does not support.

Four different things get called a thyroid problem, and they are not degrees of the same thing:

• Overt hypothyroidism: TSH raised and free T4 low. The thyroid is not keeping up. This is a diagnosis in its own right and is treated; the trials below did not study it.
• Subclinical hypothyroidism: TSH above the laboratory's range with a normal free T4. ASRM says the label only applies once thyroid function has been stable for weeks and there has been no recent severe illness, which is a reason to repeat a borderline result before acting on it.
• Antibodies with normal thyroid function: TPO antibodies present, TSH and free T4 normal. This is the group studied in the largest trials.
• Hyperthyroidism: an overactive thyroid, a different condition with its own management, covered by the American Thyroid Association's 2026 guidelines.

These are common in India. In an eight-city survey of 5,360 adults, 15.86% of women were hypothyroid, 8.02% of all adults had subclinical hypothyroidism and 21.85% had TPO antibodies (Unnikrishnan 2013, a study funded by a manufacturer of levothyroxine). ASRM's pooled figure for subclinical hypothyroidism, from studies using the 97.5th percentile as the upper limit, is 3.47%.

The evidence does not show that it does, and current US fertility guidance has changed its advice because of that. ASRM's 2024 guideline recommends not treating women who want to conceive, and who have subclinical hypothyroidism, with levothyroxine, because treatment has not been shown to reduce pregnancy loss or improve pregnancy or live birth rates. It also advises telling women that subclinical hypothyroidism is not associated with miscarriage or with higher obstetric risk.

Two other bodies stop short of a firm no. NICE's fertility guideline reviewed the evidence and asked for a trial rather than recommending treatment. The Cochrane review found a signal in favour of treatment (live birth relative risk 2.13), but from a single trial of 64 women, graded low quality, and drew no clear conclusions.

Why did older advice favour treatment? Much of it leaned on meta-analyses that included one Egyptian trial from 2010 whose effect was four to eight times larger than any other trial's (ASRM 2024). ASRM and the Cochrane reviewers both excluded that trial over concerns about the integrity of its data. One 2018 meta-analysis of IVF patients reported fewer miscarriages with levothyroxine; ASRM notes the effect disappears without that trial, and the meta-analysis's own subgroup of women with subclinical hypothyroidism alone showed no significant difference.

India's 2014 pregnancy guideline does treat TSH above its trimester ranges once a woman is pregnant. That is a different moment from trying to conceive, and the guideline predates the trials discussed here; it is still the national standard many Indian obstetricians follow.

Not while thyroid function is normal, on the best available evidence. Three randomised trials asked whether levothyroxine helps women with TPO antibodies and normal thyroid function, in three different groups, and none found a benefit:

• TABLET, UK, 952 women with a history of miscarriage or infertility: live birth 37.4% with levothyroxine and 37.9% with placebo.
• POSTAL, China, 600 women having IVF: live birth 31.7% versus 32.3%, miscarriage 10.3% versus 10.6%.
• T4LIFE, Europe, 187 women with two or more pregnancy losses: live birth 50% versus 48%. It stopped early because recruitment was slow.

ESHRE's 2022 recurrent pregnancy loss guideline turned this into a strong recommendation: euthyroid women with thyroid antibodies and recurrent pregnancy loss should not be treated with levothyroxine. ASRM recommends against screening women with infertility for thyroid antibodies when they have no symptoms.

A positive antibody result is still worth having on record, because it raises the chance of thyroid function changing later, and is a reason to recheck TSH rather than to start a tablet. Antibodies sit alongside other immune findings that are often over-treated, which we cover in what is proven about autoimmune disease and fertility.

Being told you carry an antibody linked to miscarriage, and in the same breath that the obvious tablet does not help, is a hard place to be left. It is still better than taking a treatment that costs money and hope without changing the outcome, and it does not close the search for other causes of loss.

When the thyroid is genuinely underactive: a raised TSH with a low free T4. ASRM notes that overt hypothyroidism can affect fertility, miscarriage and a baby's development, and it is treated as a medical condition in its own right.

If you already take levothyroxine, nothing on this page is a reason to stop. Thyroid hormone needs rise in pregnancy (ASRM), so tell whoever prescribes it as soon as a pregnancy test is positive rather than adjusting the dose yourself. Under the Indian guideline, TSH is rechecked six weeks after treatment starts.

Levothyroxine is not entirely free of downsides when it is not needed. ASRM lists palpitations, heat intolerance, sweating, sleep problems, anxiety and bone loss among its possible effects, as well as the cost and the delay of waiting for a number to come down.

Only if there was a reason, according to both NICE and ASRM. NICE's fertility guideline says to check thyroid function in women with possible fertility problems only if they have symptoms of thyroid disease. ASRM says to test TSH and T4 in women with signs or symptoms, including irregular cycles, rather than in everyone. Many fertility workups, including the one on our infertility workup page, include TSH as standard, alongside the other tests in our guide to fertility tests for women.

That matters because a test you did not strictly need can still produce a borderline number, and a borderline number tends to get treated. Repeating it, and checking free T4 alongside, often settles the question.

Not on the basis of living in India. The India Iodine Survey 2018-19, covering 21,406 households in every state and union territory, found median urinary iodine of 173.4 micrograms per litre in pregnant women, within the WHO's adequate range. Coverage is uneven: 76.3% of households nationally used adequately iodised salt, and 11 states and 2 union territories fell below that national figure. No trial retrieved for this page tested iodine, selenium or diet changes as a fertility treatment in thyroid autoimmunity.

• What was my TSH, what was my free T4, and which reference range were they read against?
• Was the range for a non-pregnant woman or a pregnant one?
• Has a borderline result been repeated?
• Is the treatment for an underactive thyroid, or for antibodies with normal function?
• If I become pregnant, who will recheck my TSH and when?

Your doctor may have a reason particular to you, such as a previous thyroid condition or symptoms, and that belongs in the conversation. For how TSH fits alongside the other hormone results, see FSH, LH and estradiol, AMH and the role of hormones in female reproduction.