Thyroid Disease in Pregnancy

Healthcare · §663

Pregnancy fatigue, brain fog, weight you can't move, a creeping low mood — every one of those reads as normal first-trimester and sometimes isn't. For the first twelve weeks, the baby's developing brain runs entirely on hormones from the mother's thyroid, and a small but real fraction of women quietly run short — the symptoms blur with ordinary pregnancy until the damage is permanent. One blood draw catches it. The pill that fixes it is generic and cheap. Most obstetric workflows don't test it by default — you ask.

Test · As-needed Evidence Moderate Chapter Healthcare

The reason this entry matters is the silent first-trimester window. Catch it early and the pregnancy is uneventful, the child develops normally, the mother feels human again. Miss it and what gets lost — your child's IQ, weeks of postpartum sleep, a miscarriage you'll never explain — can't be retrieved. The test is cheap. The treatment is one of the oldest, safest pills in medicine. The work is asking for it.

Your thyroid sits low in your neck and makes two hormones — T4 and T3 — that set the pace of every cell. Pregnancy quietly puts it under one of the heaviest workloads of your life. The placenta pumps out hCG, the hormone your home pregnancy test is detecting, and hCG happens to look enough like the signal that normally tells the thyroid to work harder that it tugs on the same lever directly. Your thyroid revs up. Iodine demand — the raw material the gland needs — climbs by about half. The amount of thyroid hormone tied up in carrier protein in your blood roughly doubles. A healthy gland adapts; a gland that is already running on partial reserve, often without you knowing, tips over.

The piece that makes this matter so much more than the same problem outside pregnancy: for roughly the first twelve weeks, your baby has no working thyroid of its own. Every molecule of thyroid hormone in the developing brain came across the placenta from you. The fetal thyroid switches on around week 10–12 and isn't independent until around week 18–20 Glinoer 1997. Whatever you make in that window is what the baby gets to build with.

One more reason this is easy to miss: your blood test numbers shift in pregnancy. The pregnancy-driven rise in hCG pushes TSH — the brain's signal that normally goes up when the thyroid is underperforming — down. A reading of 0.2 that would look like overactive thyroid outside pregnancy is normal at twelve weeks. A reading of 3.5 that looks fine on a routine lab report is borderline-high for a pregnant woman Stricker et al. 2007. The reference ranges most labs print are not pregnancy ranges. Whoever is reading the result has to know that.

What goes wrong when it's missed

The single most important study in this whole field followed the children of mothers whose hypothyroidism had been silently present during pregnancy and then compared them to children of mothers with normal thyroid function. At ages seven to nine, the kids of the undiagnosed mothers averaged seven IQ points lower than the comparison group, and the share with an IQ at or below 85 was four times higher — 19% versus 5% Haddow et al. 1999. Seven points is the difference between a child who slides through a regular classroom and one who needs scaffolding. Nineteen percent below 85 is the difference between a typical childhood and a rough one.

Beyond the brain, untreated maternal hypothyroidism roughly doubles the risk of miscarriage, raises preterm birth, raises gestational hypertension and pre-eclampsia, and increases low birth weight Maraka et al. 2016. Untreated overactive thyroid — usually Graves disease — is in some ways even harder on a pregnancy: pre-eclampsia in roughly fifteen to twenty percent, preterm delivery in a quarter to a half, congestive heart failure in five to seven percent, and a small but real risk of thyroid storm, a medical emergency Cooper & Laurberg 2013.

Thyroid antibodies — the immune system's own attack on the gland, picked up on a blood test called TPO — are their own story. About one woman in ten carries them, often without any thyroid problem at all. Even when the thyroid hormone numbers look normal, women with TPO antibodies have roughly three times the miscarriage rate and about twice the preterm birth rate of women without them Thangaratinam et al. 2011. The largest analysis to date — combining individual data from forty-seven thousand pregnancies — confirmed the preterm-birth link is independent of TSH Korevaar et al. 2019.

A note on what the evidence does not show. Two large, careful trials starting treatment at thirteen to seventeen weeks — past the brain-development window described above — failed to improve children's IQ at ages three, five, and nine Lazarus et al. 2012 Casey et al. 2017 Hales et al. 2018. The most-debated implication is that for borderline cases, the leverage is in the first twelve weeks or not at all. The settled point is unchanged: overt disease, treated early, is a fixable problem.

What to actually do

The core action is a blood draw. You want it ordered ideally before you stop birth control, or at the absolute latest at the very first visit after a positive pregnancy test — not at the routine 10–12 week appointment. The single most important test is TSH. Add free T4 if TSH is abnormal, and add TPO antibodies if you have ever had a miscarriage, a preterm birth, an autoimmune condition, infertility, or a family member with thyroid disease.

If you already have known hypothyroidism and you are on levothyroxine, do not wait for an appointment to adjust the dose. The standing recommendation is to increase your dose by about thirty percent — the practical version is two extra tablets per week, e.g. a double dose on Mondays and Thursdays — starting the day your home pregnancy test turns positive, then have TSH retested within four weeks Alexander et al. 2017. Waiting for the first prenatal visit means three to six weeks of relative underdosing during exactly the window that matters most for the baby's brain.

If you're newly diagnosed during pregnancy and need to start levothyroxine, the usual starting dose is around 1.6 micrograms per kilogram of body weight, once daily, with TSH rechecked every four weeks until it's in range. Take it on an empty stomach, thirty to sixty minutes before food, and separate it by at least four hours from your prenatal vitamin, iron, or calcium — any of those drops absorption by twenty to forty percent.

For overactive thyroid in pregnancy, treatment is two specific anti-thyroid drugs — and which one matters by trimester. The drug used the rest of the time, methimazole, carries a small but real risk of birth defects when used in the first trimester. The standing recommendation is to switch to a related drug called propylthiouracil for the first trimester and switch back to methimazole at the start of the second Cooper & Laurberg 2013 Andersen et al. 2013. The goal is to dose just enough — not aggressively normal numbers, because too much anti-thyroid drug crosses the placenta and suppresses the baby's thyroid.

What most pregnancy advice gets wrong

"I'm tired because I'm pregnant — every pregnant woman is." Often true. Sometimes a treatable thyroid problem. The two are indistinguishable by feel, and the test that separates them is one tube of blood.

"My TSH was normal — the lab said so." The lab's printed reference range is for non-pregnant adults. A TSH of 3.8 with the comment "within normal limits" is actually borderline-high in pregnancy, and a TSH of 0.2 — which would mean overactive thyroid outside pregnancy — is normal at twelve weeks Stricker et al. 2007. Ask whoever ordered the test to read it against pregnancy ranges. If they don't know what those are, ask for a referral.

"I'm not high-risk — I don't need it tested." The official high-risk list — age over 30, BMI over 40, family history, autoimmune disease, prior miscarriage or preterm birth, prior thyroid problem, infertility — covers most pregnant women in the United States. Multiple analyses have shown a screen-by-risk-factor approach misses roughly a third of cases Negro et al. 2010. The American Thyroid Association formally calls universal screening "acceptable," and many practicing endocrinologists do it; ACOG hasn't moved that far ACOG 2020. The blood draw is cheap; the asymmetry of harms — preventable IQ damage on one side, an unnecessary lab test on the other — is the argument for asking even if you don't formally qualify.

"Postpartum I'm just exhausted and emotional — it's depression." Maybe. Or maybe it's postpartum thyroiditis — an autoimmune attack on the thyroid in the months after delivery that affects roughly five to ten percent of all postpartum women, classically with an overactive phase from one to six months and an underactive phase from four to eight months Stagnaro-Green 2012. The underactive phase looks like postpartum depression: low energy, low mood, weight gain, mental slowing. The standard postpartum depression screen does not include a thyroid test. Ask for TSH before, or alongside, antidepressants. About half of the women whose thyroid recovers do — the other half need lifelong treatment.

"Treating borderline cases helps the baby." The honest answer is: probably only if treatment starts in the first twelve weeks. Two large trials starting at thirteen-to-seventeen weeks didn't find an IQ benefit at age three, five, or nine Lazarus et al. 2012 Casey et al. 2017. The leverage, if it exists, is at the very beginning. That is the reason to test pre-conception or at six weeks, not at twelve.

Who should be paying the most attention

Some women are at much higher baseline risk than the population average, and these are the situations where the case for testing — and for testing early — is hardest to argue against:

Already on levothyroxine for hypothyroidism. Day one of pregnancy, take two extra tablets a week. Don't wait for the appointment.
Prior miscarriage or preterm birth. Add TPO antibodies to the workup. If they come back positive, you and your doctor have a much clearer picture of the postpartum risk too.
Type 1 diabetes or any other autoimmune condition. Roughly three times the baseline risk of thyroid autoimmunity. Test before pregnancy, not during.
Family history of thyroid disease or autoimmune disease. Especially mother or sister.
Hyperemesis gravidarum. Severe morning sickness sometimes co-presents with biochemical overactive thyroid driven by very high hCG levels. Usually self-resolves by twenty weeks without treatment, but needs to be sorted from true Graves disease, which doesn't resolve Cooper & Laurberg 2013.
Postpartum mood symptoms at 1–8 months. Even with no risk factors. TSH first, then talk about antidepressants.
Low-dairy, low-seafood, no-iodized-salt diet. The US is generally iodine-replete, but a diet that avoids the three main sources can leave you marginal during the period of highest demand Pearce et al. 2016.

What changes when it's caught

If you have overt hypothyroidism that's been picked up early and treated, the felt change in the first four to eight weeks of levothyroxine is usually obvious: the bone-deep tiredness lifts, the brain fog clears, body temperature normalizes, the bowels start working again, the low mood gets noticeably lighter. Many women describe the experience as realizing they had been operating on three-quarter battery for years without knowing what full power felt like.

What you don't see is the harder part to communicate, because it's the version that doesn't happen. The miscarriage that doesn't occur. The preterm delivery that doesn't occur. The seven-IQ-point gap that doesn't open up in your child's reading scores at age seven Haddow et al. 1999. Treated overt hypothyroidism in pregnancy returns these risks to near-population baseline; that is the closest thing in medicine to a clean save.

Postpartum, if you and your obstetrician have agreed to recheck TSH at three and six months — especially if your TPO antibody test was positive — the catch is sharper still. A woman in the underactive phase of postpartum thyroiditis at five months postpartum can be sleepwalking through the day, weeping in the car, and a hundred days deep into the wrong diagnosis. A single TSH redirects the next year of her life. About half of women with postpartum thyroiditis recover on their own; the other half become long-term hypothyroid Stagnaro-Green 2012. Either way, treatment is the same one cheap pill that has been on the formulary since the 1950s.

Where this goes wrong in practice

Waiting for the 10-week appointment to get tested. The leverage is at six weeks, not ten. If your obstetrician's first slot is at ten weeks, request a TSH order over the phone or through the patient portal as soon as the home test is positive — most offices will accommodate it.

Already-hypothyroid woman who doesn't adjust her own dose. The two-extra-tablets-per-week move on the day of the positive pregnancy test is the protective default for women on stable levothyroxine. Waiting for the doctor to adjust at the first appointment costs three to six weeks of relative underdosing during the most sensitive window Alexander et al. 2017.

Reading the result against the wrong reference range. A non-endocrinologist reading the lab without flagging pregnancy ranges will call 3.8 normal. It isn't, in pregnancy. If your number is borderline and the response is "looks fine," ask whether it's being read against trimester-specific ranges.

Methimazole in the first trimester. A woman with known Graves disease on long-standing methimazole who becomes pregnant should be moved to propylthiouracil immediately. Continuing methimazole through weeks 6–10 is the window where the specific birth-defect signal lives Andersen et al. 2013.

Treating postpartum depression without checking the thyroid. The default postpartum depression screen does not include TSH. Asking is on you, or on a clinician who has the experience to think of it.

Over-treating subclinical disease past the window. If the TSH was borderline-high and treatment didn't start until fifteen weeks, the level of effort spent monitoring may not be matched by a benefit to the baby — the trials are clear on that Lazarus et al. 2012 Casey et al. 2017. Worth knowing so you don't beat yourself up about a borderline number that was found at twenty weeks.

Cost in the United States: a TSH runs twenty to sixty dollars cash through direct-to-consumer labs and is almost always insurance-covered when ordered with a pregnancy diagnosis code. Adding free T4 and TPO antibodies brings the total to under a hundred dollars cash. Levothyroxine is one of the oldest, cheapest generics in pharmacy — roughly four to ten dollars a month without insurance. Propylthiouracil and methimazole are similar. The monitoring schedule during pregnancy is a TSH draw every four weeks until mid-pregnancy, then every six to eight weeks; postpartum monitoring for TPO-positive women is TSH at three and six months after delivery Alexander et al. 2017. Most major insurance plans cover the full schedule. If you're paying cash, the worst-case total for the full pregnancy-and-postpartum year is in the low hundreds.

A few related topics this entry deliberately doesn't cover: thyroid nodules and thyroid cancer found during pregnancy (a different workup); congenital hypothyroidism in the newborn (a fetal thyroid problem; universal newborn heel-prick screening catches it); thyroid disease outside pregnancy; iodine deficiency in non-pregnant adults; and the intersection of thyroid antibodies with infertility and IVF, which has its own evidence base and trial literature.

Substance + claimed effects

Thyroid disease in pregnancy covers the spectrum of maternal thyroid dysfunction encountered before, during, and after gestation: overt and subclinical hypothyroidism, overt and subclinical hyperthyroidism (most commonly Graves disease), isolated hypothyroxinemia, thyroid autoimmunity in euthyroid women (positive TPOAb and/or TgAb), and postpartum thyroiditis. The pregnancy state itself imposes a transient physiologic load — hCG-driven TSH-receptor cross-stimulation, ~50% rise in iodine demand, ~50% rise in thyroxine-binding globulin, and placental deiodinase-mediated T4 turnover — that shifts laboratory reference ranges and unmasks marginal thyroid reserve Glinoer 1997. The claimed effects this entry covers holistically: (1) miscarriage and preterm birth risk, including in TPO-antibody-positive euthyroid women; (2) fetal neurocognitive development, especially the pre-12-week window before fetal thyroid onset when the fetus depends entirely on maternal T4; (3) maternal day-to-day energy, mood, and cognition during and after pregnancy; (4) postpartum thyroid dysfunction (transient or permanent); and (5) screening, dose, and timing of intervention. Scope is maternal thyroid axis from preconception through one year postpartum. Out of scope: thyroid nodules and cancer in pregnancy (a distinct workup), neonatal congenital hypothyroidism (a fetal axis problem, screened separately), and non-pregnant thyroid disease.

Evidence by addressing question

mechanism

The hCG–TSH cross-reactivity. Pregnancy hCG and TSH share a common alpha subunit and partial beta-subunit homology; hCG weakly stimulates the TSH receptor. Serum hCG peaks around 10–12 weeks at ~50,000–100,000 IU/L, and TSH falls in mirror image — often below the non-pregnant lower limit of 0.4 mIU/L, sometimes to undetectable, with FT4 transiently elevated Glinoer 1997. This is normal physiology and not Graves disease, though it can be misread as such; it also drives gestational transient thyrotoxicosis when hCG is supraphysiologic (multiples, molar pregnancy, hyperemesis gravidarum) Cooper & Laurberg 2013.

Maternal-to-fetal T4 transfer and the pre-12-week window. The fetal thyroid does not begin meaningful hormone production until ~10–12 weeks of gestation and is not fully autonomous until ~18–20 weeks. Before that, all fetal triiodothyronine in brain tissue derives from maternal T4 transported across the placenta and locally deiodinated to T3 by fetal brain deiodinases Glinoer 1997. Maternal hypothyroxinemia in the first trimester is therefore the highest-leverage exposure for fetal neurocognition; T4 transfer at term contributes a smaller fraction of fetal hormone but is detectable in congenital hypothyroidism cohorts who reach term with measurable cord T4 Alexander et al. 2017.

Iodine demand. Renal iodide clearance roughly doubles and fetal-placental uptake adds a sink; daily requirement rises from 150 to 220–250 µg/day. In an iodine-replete mother the thyroid expands output; in mild deficiency, the gland enlarges (the physiologic "pregnancy goiter") and a marginal axis tips into hypothyroidism Pearce et al. 2016.

Trimester-shifted reference ranges. Because hCG suppresses TSH and TBG-bound T4 rises, applying non-pregnant ranges to pregnant serum produces false positives (apparent hyperthyroidism in 1st trimester) and false negatives (a TSH of 3.5 looks "normal" by non-pregnant standards but is elevated for pregnancy) Stricker et al. 2007. The 2017 ATA guidelines recommend laboratory- or population-specific trimester ranges, defaulting to an upper TSH of ~4.0 mIU/L (revised up from the earlier 2.5/3.0/3.0 trimester cuts) when local data are unavailable Alexander et al. 2017.

Autoimmunity without dysfunction. TPO and Tg antibodies signal an axis under autoimmune attack with reserve already partly consumed. Affected women are at higher baseline risk of progression to subclinical or overt hypothyroidism during pregnancy as demand rises, and meta-analyses find independent associations between TPO positivity and miscarriage / preterm birth even with normal TSH and FT4 — mechanism is debated (direct placental autoimmunity vs marginal hormonal reserve) Thangaratinam et al. 2011.

evidence

Overt hypothyroidism — settled. Maternal overt hypothyroidism (elevated TSH with low FT4, or TSH >10 mIU/L regardless of FT4) is consistently associated with miscarriage, gestational hypertension, preterm birth, low birth weight, and offspring IQ deficits; levothyroxine treatment normalizes these risks. Haddow et al. 1999 is the landmark cohort: 62 children of mothers with untreated thyroid deficiency during pregnancy averaged 7 IQ points below matched controls at age 7–9; 19% scored ≤85 vs 5% of controls. Universal levothyroxine in overt disease is the standing recommendation of every major guideline body — ATA, ACOG, Endocrine Society Alexander et al. 2017 ACOG 2020.

Subclinical hypothyroidism — contested, with two large negative RCTs. Observational data link SCH (elevated TSH, normal FT4) to pregnancy loss, preterm birth, and modestly lower offspring IQ Maraka et al. 2016. But two large RCTs of antenatal treatment failed to improve child cognition: Lazarus et al. 2012 (CATS, n=21,846 screened, 794 treated, no IQ difference at age 3) and Casey et al. 2017 (n=1,203 with SCH or isolated hypothyroxinemia, no IQ difference at age 5). Follow-up of CATS at age 9 was also null Hales et al. 2018. Both trials began treatment at a median gestational age of 13–17 weeks — past the pre-12-week window of maximum dependence on maternal T4 — which is the trials' most-debated limitation. Negro et al. 2010 (n=4,562) found that universal screen-and-treat reduced adverse pregnancy outcomes vs case-finding only in the high-risk subgroup, but not overall.

TPO antibodies in euthyroid women — signal real, treatment null. Thangaratinam et al. 2011 meta-analysis: TPO+ women had ~3× miscarriage risk and ~2× preterm birth risk. Korevaar et al. 2019 IPD meta-analysis (n=47,045): isolated TPO positivity associated with ~33% higher preterm-birth odds even with normal TSH. But Dhillon-Smith et al. 2019 TABLET trial (n=940 euthyroid TPO+ women with prior miscarriage or subfertility, randomized to levothyroxine 50 µg or placebo from preconception) found no improvement in live birth rate (37.4% vs 37.9%). Treatment of euthyroid TPO+ women is therefore not supported.

Hyperthyroidism — settled. Untreated overt hyperthyroidism (most commonly Graves) raises pre-eclampsia, preterm delivery, low birth weight, congestive heart failure, and thyroid storm risk Cooper & Laurberg 2013 Nguyen et al. 2018. Antithyroid drug choice matters: methimazole carries a 2–4% absolute risk of distinctive embryopathy (aplasia cutis, choanal/esophageal atresia, omphalocele) when used in the first trimester; PTU carries a smaller and less specific malformation signal plus a rare hepatotoxicity risk. Andersen et al. 2013 Danish nationwide cohort (n=817,093) confirmed the methimazole–malformation association (adjusted OR ~1.7 for any birth defect). Standing recommendation: PTU in the first trimester, switch to methimazole at the start of the second.

Postpartum thyroiditis. Affects 5–10% of women in the year after delivery; classically biphasic (transient hyperthyroid phase at 1–6 months, hypothyroid phase at 4–8 months), but ~50% present with only one phase. ~50% with the hypothyroid phase progress to permanent hypothyroidism within 7 years; the rest recover. TPO antibody positivity in early pregnancy is the strongest predictor — ~33–50% of TPO+ women develop PPT vs ~5% of TPO− Stagnaro-Green 2012 Stagnaro-Green 2004.

Iodine. Mild-to-moderate iodine deficiency in pregnancy is associated with lower offspring IQ and language/motor development in observational cohorts; supplementation in deficient populations improves neonatal thyroid function and reduces goiter, but RCTs of cognitive outcomes in marginally deficient populations are mixed Pearce et al. 2016.

protocol

Pre-conception or first-prenatal visit screen. ATA 2017 recommends measuring serum TSH in any woman seeking pregnancy or newly pregnant who has any risk factor: prior thyroid disease, family history of autoimmune thyroid disease, prior preterm birth or miscarriage, infertility, age >30, BMI >40, head/neck radiation, prior thyroid surgery, type 1 diabetes or other autoimmune disease, presence of goiter, known TPO/Tg antibody positivity, residence in an iodine-deficient region, or use of amiodarone / lithium / recent iodinated contrast Alexander et al. 2017. Universal screening is not formally recommended by ATA or ACOG but is considered acceptable; many endocrinologists test universally in practice given the broad risk-factor list ACOG 2020.

Levothyroxine dose for overt hypothyroidism. Goal: TSH <2.5 mIU/L pre-conception; trimester-specific (or population-specific) range during gestation (default upper limit ~4.0 mIU/L). Women already on LT4 should self-increase dose by ~30% (two extra tablets per week) at the first positive pregnancy test, then test TSH within 4 weeks Alexander et al. 2017. Typical starting dose for new-diagnosis overt hypothyroidism: 1.6 µg/kg/day, adjusted by 25–50 µg every 4 weeks based on TSH. Take on empty stomach 30–60 min before food, ≥4 hours from iron / calcium / prenatal vitamins.

Antithyroid drugs for Graves. PTU 50–300 mg/day in divided doses during the first trimester; switch to methimazole 5–20 mg/day at the start of the second to limit cumulative PTU hepatotoxicity exposure. Goal: maintain FT4 at or just above the upper limit of normal — over-treatment crosses the placenta and causes fetal hypothyroidism / goiter Cooper & Laurberg 2013. TRAb (TSH-receptor antibody) titer measured in 2nd–3rd trimester predicts fetal/neonatal thyrotoxicosis risk.

Iodine. WHO and ATA: 150 µg/day prenatal supplement (in addition to dietary iodine) preconception through pregnancy and lactation; total daily intake target 250 µg, ceiling 500 µg Pearce et al. 2016 Alexander et al. 2017. Many US prenatal vitamins do not contain iodine; checking the label is the actionable step.

Postpartum monitoring. Symptomatic women, plus all TPO+ women, get TSH at 3 and 6 months postpartum. The hyperthyroid phase of PPT is usually self-limited; symptomatic management with low-dose propranolol; antithyroid drugs are contraindicated in PPT because the gland is leaking stored hormone, not over-producing. Hypothyroid phase is treated with LT4; trial of withdrawal at 6–12 months to test whether the gland recovers Stagnaro-Green 2012.

contraindications

The entry itself is about pregnancy, so the contraindication framing is different: it is methimazole (specifically in the first trimester) and radioactive iodine (absolutely contraindicated in pregnancy and lactation; deferred until ≥6 months after weaning if used at all) that are contraindicated. Antithyroid drugs at over-treatment doses cross the placenta and cause fetal goiter/hypothyroidism — under-treat slightly rather than over-treat Cooper & Laurberg 2013. Beta-blockers (propranolol) are acceptable short-term but raise neonatal hypoglycemia/bradycardia risk if continued near term.

misconceptions

"Pregnancy fatigue is just pregnancy." Half of pregnant women with overt or subclinical hypothyroidism are symptom-attributing — their fatigue, constipation, cold intolerance, weight changes, and low mood are routinely written off as normal pregnancy ACOG 2020. The blood test is the only reliable discriminator.

"My TSH is in the normal range." Non-pregnant lab reference ranges (typically 0.4–4.5 or 0.4–5.0 mIU/L) overcall normal in pregnancy. Trimester-specific ranges shift the upper bound to ~4.0 and the lower bound to ~0.1 in the first trimester Stricker et al. 2007.

"Suppressed TSH in early pregnancy means Graves." hCG-driven physiologic TSH suppression is normal at 8–14 weeks; FT4 distinguishes (mildly elevated and self-resolving in gestational transient thyrotoxicosis; persistently elevated with TRAb positivity in Graves) Cooper & Laurberg 2013.

"Postpartum thyroiditis is just postpartum depression." The mood and fatigue symptoms of the hypothyroid phase of PPT overlap with PPD; TSH is the cheap discriminator and is not part of routine PPD workup. PPT is missed in an unknown but substantial fraction of postpartum mood presentations Stagnaro-Green 2012.

"Treating subclinical hypothyroidism makes the baby smarter." The two largest RCTs found no IQ benefit when treatment started after 12 weeks Lazarus et al. 2012 Casey et al. 2017. The leverage if it exists is pre-12-week treatment, which the RCTs did not test.

"My prenatal vitamin has iodine." US prenatal vitamins are highly variable; ATA-recommended 150 µg iodide is not in every product. Label-check is the actionable step Pearce et al. 2016.

stakes

Untreated overt maternal hypothyroidism: ~60% chance of pregnancy complication including pregnancy loss, gestational hypertension, preterm birth, placental abruption; 7-point average IQ deficit and 4× rate of IQ ≤85 in offspring Haddow et al. 1999. Untreated overt hyperthyroidism: pre-eclampsia in ~15–20%, preterm birth in ~25–50%, congestive heart failure 5–7%, thyroid storm in 1–2%, ~25% perinatal mortality in the most severe untreated series Cooper & Laurberg 2013. Subclinical hypothyroidism: ~2× miscarriage, ~30% higher preterm birth Maraka et al. 2016. TPO+ euthyroid: ~3× miscarriage, ~2× preterm birth Thangaratinam et al. 2011. Permanent postpartum hypothyroidism in ~25–50% of women who develop the hypothyroid phase of PPT, often masked for years as fatigue / mood / weight problems Stagnaro-Green 2012.

payoff

Levothyroxine-treated overt hypothyroidism brings miscarriage, preterm birth, and offspring IQ rates back to near-population baseline Alexander et al. 2017. Felt effect within 4–8 weeks: normalized energy, lifted brain fog, regained warmth tolerance, regular bowel function, mood lift — particularly notable in women whose pre-pregnancy hypothyroidism was missed and is being treated for the first time. Postpartum TSH check that catches PPT in the hypothyroid phase recovers the mother who was mistakenly being treated for postpartum depression. Pre-conception identification of TPO antibodies + family history triggers monitoring that catches the gradual axis failure during pregnancy before it crosses into overt disease.

audience

Highest-leverage subgroups: women trying to conceive with any risk factor (test pre-conception, not at 8 weeks gestation); women with previously diagnosed hypothyroidism on LT4 (need dose increase the day pregnancy is confirmed); women with prior miscarriage or preterm birth (TPO + TSH workup); women with type 1 diabetes or other autoimmune disease (3× baseline thyroid autoimmunity risk); women with hyperemesis gravidarum (rule out gestational thyrotoxicosis vs early Graves); women with postpartum fatigue or mood symptoms at 1–8 months (rule out PPT before treating for PPD) Alexander et al. 2017.

practicalities

TSH ± FT4 ± TPO antibodies: a single venous draw at the first prenatal visit, or pre-conception. Cost in the US is ~$20–60 cash through direct-to-consumer labs; covered by most insurance when ordered with a pregnancy-related diagnosis code. Levothyroxine is a generic, ~$4–10/month at most pharmacies, ~$30–60/year. Methimazole and PTU are generic, similarly cheap. Monitoring during pregnancy: TSH every 4 weeks through mid-pregnancy in any treated woman, then every 6–8 weeks. Postpartum: TSH at 3 and 6 months for TPO+ women or anyone with prior PPT.

history

The link between maternal hypothyroidism and offspring cretinism dates to 19th-century iodine-deficiency endemic-goiter regions (Switzerland, Andes). The modern subclinical/cognitive thread opened with Haddow et al. 1999 and Pop et al. 1999, which together moved the field from "treat the overt cases" to "consider the subclinical axis." The pendulum swung toward universal screening through the 2000s, then back when CATS and Casey trials failed to show child-IQ benefit from mid-pregnancy treatment. The current synthesis: aggressive treatment of overt disease, careful management of women already on LT4, risk-factor-based testing for SCH, and conservative treatment when SCH is found with TPO antibodies — with the residual epistemic hole being whether pre-12-week treatment of SCH would have shown the benefit the mid-pregnancy RCTs missed.

failure-modes

Delayed dose increase in known hypothyroidism. A woman on stable LT4 who waits for her 8-week prenatal visit to recheck TSH may spend 4–6 weeks of the critical first trimester in relative hypothyroxinemia. Self-increasing dose at the first positive pregnancy test (two extra tablets per week) is the protective move Alexander et al. 2017.

Applying non-pregnant TSH ranges. A TSH of 3.8 reported as "normal" by lab reference ranges is borderline-elevated for pregnancy; the lab range and the pregnancy range disagree.

Misdiagnosing PPT as PPD. Standard PPD screening doesn't include TSH; the hypothyroid phase of PPT can drive depression-like symptoms at 4–8 months postpartum and is treated entirely differently.

Methimazole in the first trimester. Switching to PTU after the positive pregnancy test is the standing recommendation; women diagnosed with Graves before pregnancy who don't know to switch carry the embryopathy risk Andersen et al. 2013.

Iodine omission. A US prenatal vitamin without iodine, combined with a low-dairy / low-seafood diet, can leave a woman mildly iodine-deficient during the period of highest demand Pearce et al. 2016.

alternatives

For overt hypothyroidism, levothyroxine is the only first-line treatment in pregnancy. Liothyronine (T3) and desiccated thyroid are not recommended — T3 does not cross the placenta efficiently, so combined T4/T3 preparations leave the fetus relatively T4-deficient even when the mother's labs look normal Alexander et al. 2017. For Graves, definitive treatment (radioactive iodine ablation or thyroidectomy) is the standing alternative outside pregnancy; in pregnancy, antithyroid drugs are the only option (RAI is fetal-toxic; surgery is reserved for drug failure or compression).

out-of-scope

Thyroid nodules in pregnancy (different workup, biopsy decisions). Congenital hypothyroidism in the newborn (a fetal axis issue; universal newborn screening covers it). Hyperemesis gravidarum (often co-presents with biochemical thyrotoxicosis but is a distinct clinical entity). Iodine deficiency outside pregnancy. Hashimoto's and Graves disease outside pregnancy. Subfertility and IVF cycling thresholds (an adjacent topic with overlapping evidence).

The credibility range

The optimist case

The strongest case for an aggressive, universal-screening, low-threshold-treat approach: maternal T4 is the sole driver of fetal brain T3 in the first trimester, so even small deficits during organogenesis carry permanent neurodevelopmental cost; the Haddow IQ data are causal; observational signals for SCH and TPO autoimmunity match a real underlying biology; the CATS and Casey trials failed because they treated too late (median 13–17 weeks gestation, past the window of leverage), and Negro 2010 showed benefit in the high-risk subgroup. Levothyroxine is cheap, safe, and easy to titrate; the cost of universal early screening (a single TSH) is trivial against the lifetime cost of a preventable IQ deficit or a preterm birth. ACOG's official position not to universally screen is regulatory caution; many practicing endocrinologists screen universally anyway. Postpartum, the case is even stronger — PPT is common, often missed, often misdiagnosed as PPD, and the test that distinguishes is a single TSH.

The skeptic case

Two large, well-conducted RCTs (CATS, Casey) failed to show child-cognition benefit from antenatal SCH treatment. The TABLET trial showed no live-birth benefit from levothyroxine in euthyroid TPO+ women with prior miscarriage. Observational TPO/SCH–outcome links are confounded (women with autoimmunity tend to have other autoimmune conditions, fertility issues, older maternal age). Universal screening would diagnose many borderline-TSH women whose pregnancies would have been fine and commit them to monitoring/treatment with no demonstrated outcome benefit. The IQ effects in Haddow are real for overt disease — but overt disease is rare and is caught by symptom-based testing. The pendulum-swing history shows the field has overcorrected on screening before. The cheap-and-safe argument cuts both ways: if treatment is null at scale, "cheap" doesn't justify a clinical workflow change.

The author's call

The article lands closer to the optimist side, but disciplined. Test every pregnant or pre-conception woman at any of ATA's risk factors — which is most women — and TPO-antibody-test women with prior miscarriage, preterm birth, or known autoimmunity. Treat overt disease aggressively and pre-emptively; raise the bar for treating SCH, especially after week 12, where two large RCTs failed to show benefit. Counsel TPO+ women on monitoring and PPT risk; do not treat euthyroid TPO+ women with levothyroxine (TABLET was clean). Postpartum, screen TSH at 3 and 6 months in TPO+ women and in any woman with symptoms. Score: meta evidence 4 (settled for overt disease, two clean RCTs for subclinical), meta controversy 3 (universal-screening and SCH-treatment debates live). The largest residual uncertainty: did the negative trials miss a pre-12-week window of benefit?

Stakeholder + incentive map

ACOG / Endocrine Society / ATA — historically more conservative on universal screening (ACOG, Endocrine Society) vs more permissive on universal screening at the margins (ATA). The guideline tension is the main professional dispute.
Practicing obstetricians — vary widely; many test TSH universally as part of first-trimester labs whether or not the patient meets formal risk criteria.
Reproductive endocrinologists / infertility clinics — much more aggressive on testing and treating SCH and TPO+ women (commercial incentive interacts with patient population that has prior loss/infertility, where the optimist case is strongest).
Patient advocacy (thyroid patient groups, postpartum thyroiditis awareness) — push for universal screening; argue PPT is dramatically under-recognized.
Skeptic / RCT-purist — academic emergency-medicine and family-medicine literature emphasizes the negative RCTs and resists test/treatment expansion absent positive trials.

Population variability

Pre-existing hypothyroid women on LT4. Highest-leverage subgroup: requires dose increase the day pregnancy is confirmed, not at the first prenatal visit. ATA's two-extra-tablets-per-week heuristic is designed for this group.

Women with Type 1 diabetes or other autoimmune disease. 3× baseline thyroid autoimmunity rate; routine TPO + TSH screening warranted.

Iodine-deficient regions. WHO maps inland Europe, parts of Africa, parts of Asia as still marginal. US is generally iodine-replete via dairy iodophors and bread, but a low-dairy / no-iodized-salt diet can leave a woman marginal.

Hyperemesis gravidarum. 30–60% have biochemical thyrotoxicosis from hCG excess; almost all resolve by 20 weeks without treatment. The discrimination from Graves: TRAb negative, no goiter, no pre-pregnancy hyperthyroid history.

Twin / molar pregnancies. Supraphysiologic hCG → higher rates of gestational transient thyrotoxicosis; rarely needs treatment but flags monitoring.

Postpartum. TPO+ women have ~33–50% PPT risk; type 1 diabetics ~25%; general population ~5–10%.

Knowledge gaps

Whether pre-12-week treatment of subclinical hypothyroidism would show the IQ benefit the post-12-week trials missed (CATS and Casey both started past the window of maternal-T4 dominance; a pre-conception start trial in TPO+ SCH women would be the cleanest test, ethically difficult to power because of small effect sizes and large sample needs). Whether mild iodine deficiency in the US population has measurable cognitive cost in offspring (ongoing observational studies, no large interventional trials in marginally deficient populations). Long-term cardiovascular and neurodevelopmental outcomes in offspring of treated vs untreated SCH and TPO+ pregnancies (current follow-up is to age 5–9). Optimal TSH targets in pregnancy at the per-individual level (current upper-bound of 4.0 is population-average; whether tighter targets matter in specific subgroups is unresolved).

Scope coverage vs brief. The brief named miscarriage and preterm birth risk, fetal brain development, maternal energy and mood, postpartum thyroid function, and shifting trimester targets. All five are covered end-to-end. The brief did not separately call out thyroid autoimmunity in euthyroid women (TPO antibodies), but the literature treats it as inseparable from the miscarriage/preterm story and from postpartum thyroiditis prediction, so it is woven in.

The pre-12-week framing is a judgment call. CATS and Casey are clean negative trials for treating subclinical disease starting at 13–17 weeks. The article leans on the pre-12-week window as the live-leverage hypothesis. This is mainstream in the ATA-aligned community but not a settled trial result — the residual epistemic uncertainty is acknowledged in the body's evidence section and in the research dossier's credibility range.

Universal vs targeted screening. The article frames the asymmetry of harms in favour of testing — risk-factor lists capture most pregnant women in the US, the blood draw is cheap, the miss cost is permanent. ACOG's formal position is targeted screening; ATA calls universal screening acceptable. The article does not invent a "universal screening" recommendation; it points the reader at asking for the test, which is action they control regardless of formal guideline.

Action vs decide. Marked action: test because the central reader move is the blood test. Treatment decisions land downstream with a clinician and would warrant decide, but the reader-facing leverage is upstream of that.

Applicability call. Scored 3 on the women's-health-decision-audience rule (per meta §6 emergency-recognition/decision-audience guidance, broadly analogous). Roughly 80% of women have at least one pregnancy; the entry is also relevant to women planning pregnancy and to all postpartum women (PPT affects 5–10% regardless of antenatal thyroid status). Did not score higher — it remains a women's-only entry and the addressable population is bounded by reproductive timing.

Beauty scores zero, deliberately. Maternal hypothyroidism affects hair (telogen effluvium, brittle), skin (dry), and face (puffiness/myxedema in overt cases) — but these are pregnancy-and-postpartum effects, transient, and not the felt-experience anchors for the entry's pitch. Scoring beauty non-zero would dilute the dimension's signal across the catalogue without earning the reader's attention here.

Dream narrative written by choice. Computed overall score is ~28, below the 40 obligation. Wrote one anyway because the entry's hook is genuinely a relief lever (per dream-narrative.md §3) — the missed-this-could-have-been version of the story — and the dek and tagline benefit from carrying it. Floor rule observed: dek and tagline land sharper than a straight version would, never vaguer.

Future-link candidates. Iodine in pregnancy as a standalone entry (currently a paragraph here, would warrant its own); postpartum depression vs postpartum thyroiditis as a discriminator entry (currently a misconceptions paragraph); infertility workup including thyroid; Hashimoto's disease and Graves disease as the parent thyroid-condition entries this one cross-references.

Separate-entry candidates flagged but not split. Postpartum thyroiditis is borderline-substantial enough to warrant its own entry once the catalogue has the surface for it — currently kept here because the test that catches it (TSH at 3 and 6 months postpartum) is the same workflow as the rest of the entry, and splitting would fragment the reader's mental model.

Contraindications field left empty. The condition this entry is about is pregnancy, so the schema's pregnancy/breastfeeding contraindication tokens don't fit. The drug-specific contraindications (methimazole 1st trimester, radioactive iodine, T3 preparations) are covered in the body's protocol and failure-modes sections.

Rating difficulty: longevity. Scored 1 — for the mother, the longevity effect of catching previously-undiagnosed Hashimoto's is real but indirect. The dominant longevity story is offspring (Haddow's IQ data), which is one generation downstream and not what this dimension scores. Considered scoring 2 to reflect offspring; decided the holistic score should match what the substance does for the rated person, with the offspring framing carried by the pitch.