Postpartum Thyroiditis and Psychosis: What Providers Miss

Postpartum Psychosis Is an Endocrine Event

Postpartum psychosis is a psychiatric emergency. It is also an endocrine event, and the standard postpartum workup is not built to see that. Most women get, at best, one TSH at the six-week visit and no plan at all for the steepest hormone drop of their lives. Postpartum thyroiditis alone affects 5 to 10 percent of women in the first year, and a single TSH is not a screening strategy for it.

I want to be precise about what this article argues. It does not argue that hormones replace psychiatric care. A woman in postpartum psychosis needs psychiatric hospitalization, and nothing here changes that. The question is a different one: what could we do in the days, weeks, and months before the emergency arrives?

The case that put this question in front of every provider

In 2023, multiple providers prescribed 13 psychiatric medications to a labor and delivery nurse in Massachusetts over four months. They were treating anxiety, insomnia, and depression in the months after the birth of her third child [1]. The case, Lindsay Clancy’s, ended in tragedy and remains before the courts [2].

I am not going to relitigate the case, and I am not going to second-guess the clinicians in it. None of us were trained to think about this differently. What stays with me is the pattern: four months of escalating psychiatric prescribing. As far as I have seen in any public reporting, no one looked at the endocrine picture underneath. That pattern reaches well beyond one woman, because postpartum care screens for psychiatric illness and rarely looks at hormones.

What does postpartum psychosis look like, and why do we miss it?

Postpartum psychosis is rare, affecting roughly 1 to 2 women per 1,000 births [3]. Many providers will go an entire career without seeing one, which is part of why we miss it.

Onset is fast. Most cases begin within the first two weeks after delivery, often within the first 72 hours [4]. The presentation is not gradual sadness. It looks like confusion, mania or hypomania, disorganized thinking, intrusive thoughts, delusions, and insomnia far out of proportion to a newborn’s schedule.

That is the core distinction from postpartum depression, which is a mood disorder and tends to build slowly. Postpartum psychosis sits much closer to a manic or mixed episode with psychotic features.

Some cases do not follow the textbook. The Clancy case, from what reporters have described, built more slowly and looked like anxiety and depression for long stretches. Atypical presentations are exactly where a physiology-first differential matters most.

Who is at highest risk for postpartum psychosis?

The women at highest risk are well known [4]:

• A personal history of bipolar disorder
• A prior episode of postpartum psychosis
• A first-degree relative with bipolar disorder
• Sleep deprivation and first delivery
• Autoimmune disease, including autoimmune thyroiditis and Hashimoto’s

The autoimmune link is the one I think we underuse. In one cohort of first-time mothers with no psychiatric history, 19 percent of women with postpartum psychosis had autoimmune thyroid disease. Among controls, it was 5 percent [5]. The overlap is well documented. In practice, most of us ignore it. (For how autoimmune thyroid disease fits into the wider picture, see thyroid combination therapy and hormone replacement.)

For many women, a first episode of postpartum psychosis turns out to be the first sign of a bipolar spectrum illness. No one diagnosed it before pregnancy [4]. A clean psychiatric history does not rule anyone out.

What happens to hormones in the first 48 hours after delivery?

Pregnancy is the largest sustained endocrine load a woman will ever carry. By the third trimester, estradiol runs roughly 100 times normal reproductive levels. Progesterone runs around ten times luteal levels, now produced mostly by the placenta [6]. Cortisol is chronically elevated. Thyroid binding globulin rises and pushes total T4 up with it [6, 7].

Then she delivers the placenta, and within about 48 hours progesterone falls from roughly 100 to 150 ng/mL to essentially zero. Estradiol falls on the same timeline [6, 7]. Cortisol drops, after months of running high (more on cortisol and hormones). Prolactin surges in women who are nursing or pumping, and it keeps ovarian function suppressed for months.

The drop is closer to pulling a lamp cord out of the socket: full power to zero, almost at once. It is the fastest hormonal transition in adult human physiology, and most women still move through it without a scratch.

The nervous system feels it first. Estradiol, progesterone, and thyroid hormone all support mitochondrial energy production in the brain. So the crash is also a sudden drop in the metabolic support the brain has run on for nine months.

How does allopregnanolone connect the progesterone crash to mood?

Progesterone is not only a uterine hormone. Through 5-alpha reductase it converts to allopregnanolone, the most potent endogenous positive allosteric modulator of the GABA-A receptor [8]. That is the pathway behind progesterone’s effects on sleep, stress buffering, and mood. It is also the piece most conventional care leaves out, as I have argued in defense of progesterone. (For more on this pathway, see progesterone and brain health.)

During pregnancy, allopregnanolone floods the brain. One plausible adaptation, and I want to be clear this part is still partly theoretical, is that GABA-A receptors downregulate to avoid constant saturation. The clearest evidence so far comes from animal work showing exactly that downregulation in pregnancy, with a rebound after delivery [9]. Too much allopregnanolone is too much brake on the brain.

Now remove it in two days. The brake is gone, and the receptors have not yet had time to upregulate. The result, in a susceptible woman, is insomnia. In some, it is psychosis.

Estrogen matters here too. Estradiol modulates dopamine, serotonin, and glutamate signaling, and reproductive hormones touch nearly every system implicated in postpartum mood disorders [7]. The leading model for postpartum psychosis is that these changes after birth trigger illness in women who are already susceptible [4]. Progesterone tends to calm the central nervous system. Estradiol tends to excite it. Both disappear together, which is why progesterone and hormone balance matters more in this window than at almost any other time.

What do the brexanolone and zuranolone approvals tell us?

This mechanism is not fringe. The FDA acknowledged it twice.

The FDA approved brexanolone, an intravenous formulation of allopregnanolone, for postpartum depression in 2019 after two phase 3 trials [10]. Women on it improved quickly, with less depression [11]. The FDA withdrew that approval in 2025 [12]. In 2023 it approved zuranolone, an oral neuroactive steroid that acts on the same GABA-A target, for postpartum depression [13].

Neither drug carries an approval for psychosis. But both rest on the same premise: that replacing allopregnanolone’s GABA-A activity after delivery treats the syndrome. The system accepts that premise for a patented molecule. In the same postpartum window, it rarely considers the hormone the body uses to make allopregnanolone in the first place.

Is hormone withdrawal itself the trigger?

The best direct evidence comes from a small NIMH study published in 2000 [14]. Researchers induced a hypogonadal state in 16 euthymic women and added back high doses of estradiol and progesterone for eight weeks to simulate pregnancy. Then they withdrew both under double-blind conditions.

Five of the eight women with a history of postpartum depression developed significant mood symptoms during withdrawal. None of the eight women without that history did.

That tells us two things. The withdrawal, not the low level alone, is the trigger. And some women are differentially sensitive to it. Most women go through the same crash and are fine. A subgroup is not, and that subgroup is identifiable in advance.

How common is postpartum thyroiditis, and why does one TSH miss it?

Postpartum thyroiditis affects roughly 5 to 10 percent of women in the first year after delivery [15]. That is not rare. For comparison, postpartum psychosis affects about 0.1 to 0.2 percent.

It is autoimmune. During pregnancy the body partially suppresses the immune system to tolerate the fetus. After delivery it has to ramp back up, and the thyroid seems to be where that rebound lands. It may not show up in the first days or weeks. It can surface months later, as either a hyperthyroid or a hypothyroid picture.

Both phases, the thyrotoxic and the hypothyroid, can pass for a primary psychiatric problem.

• Thyrotoxic phase: insomnia, anxiety, irritability, racing thoughts, weight loss, tachycardia
• Hypothyroid phase: depression, fatigue, low mood, cognitive slowing

Picture a woman at four to eight weeks postpartum with a suppressed TSH and an elevated free T4. She isn’t sleeping. She’s anxious, losing weight, and her heart rate is up. That is a classic thyrotoxic picture. Without a thyroid panel, she is likely to leave with a benzodiazepine, an SSRI, or both, and the underlying problem continues.

A single TSH at the six-week visit catches whichever phase happens to be present that day, if it is abnormal enough to flag. It says nothing about antibodies, conversion, or where her thyroid is going. Free hormones matter even inside the normal range. Women with lower free T3 and free T4 before delivery had more postpartum mood disturbance, even at normal levels [16]. (For the full interpretation framework, see how to read thyroid labs like a pro.)

Which labs catch postpartum thyroiditis early?

My minimum for any postpartum woman with mood symptoms is the full panel below.

| Marker | What it adds | Why TSH alone misses it |

| — | — | — |

| TSH | Pituitary signal | Lags behind rapid postpartum shifts |

| Free T4 | Circulating thyroid hormone | Distinguishes the thyrotoxic and hypothyroid phases |

| Free T3 | The active hormone at the cell | Poor conversion can hide behind a normal TSH |

| TPO antibodies | Autoimmune thyroid disease | TSH can be normal while antibodies are rising |

TPO antibodies deserve special attention. TPO antibody positivity in pregnancy predicts both postpartum thyroiditis and postpartum depression [15]. In a prospective cohort of 1,075 women, those with TPO antibodies in early pregnancy had higher odds of first-onset depression at four months postpartum. The adjusted odds ratio was 3.8 [17]. A 2023 review reached the same conclusion, with third-trimester TPO antibodies the most promising marker [18].

In my own practice, I check this panel quarterly for at least the first year postpartum. I think there is a case for going longer. That is my clinical approach, and it is not a guideline. Four panels cost little next to a missed thyroid in a woman we are already treating for anxiety. If the panel comes back abnormal, thyroid dosing becomes part of the mood plan. Thyroid hormone drives mitochondrial output in every cell, including neurons. It often falls first, like the opening domino in a chain.

Estradiol and progesterone levels are not useful to draw in the first weeks postpartum, since we already know they will be near zero. The thyroid is the part of the endocrine picture we can see.

Can progesterone help prevent the postpartum crash?

This is where the evidence gets thinner, and I want to say so plainly. Progesterone-based prevention of the postpartum crash is mechanism-based and off-label. There are no controlled trials behind it yet. I am not describing treatment for a woman in active psychosis. I am describing prevention in women we already know are at risk.

Four things support the approach, starting with how it holds up across cases.

• A clear mechanism. Progesterone is the precursor to allopregnanolone.
• Pharmaceutical validation. Two FDA approvals built on replacing allopregnanolone activity after delivery [10, 13].
• A withdrawal model. Rapid steroid withdrawal reproduces symptoms in sensitive women [14].
• Clinical observation. Many providers, myself included, have used bioidentical progesterone in the postpartum window with good results. That is anecdote. I treat it as a signal worth studying.

What we do not have is a trial of bioidentical progesterone for postpartum prevention. Studies of progesterone in postpartum depression are few and inconclusive. Researchers are still testing whether oral progesterone raises allopregnanolone in the brain [19]. A Cochrane review found no randomized placebo-controlled trial of natural progesterone for postpartum depression. The trial it did find used a synthetic progestin, norethisterone, given within 48 hours of delivery, and it raised the risk of postpartum depression. The review’s conclusion was direct: “Synthetic progestogens should be used with significant caution in the postpartum period” [20].

That result is a warning about progestins, not a verdict on progesterone. Synthetic progestins cannot convert to allopregnanolone, and their chemical structure blocks the neuroactive metabolites that make progesterone calming [21]. They are a different molecule with a different receptor profile, which is the same reason progesterone and medroxyprogesterone acetate are not interchangeable. A negative progestin trial does not answer the progesterone question.

Some women carry a personal history of bipolar disorder, a prior postpartum psychosis, or a strong family history. For them, starting bioidentical progesterone in the first days postpartum to blunt the allopregnanolone withdrawal curve is, in my view, a reasonable conversation to have. How much to give and for how long are real clinical decisions, beyond the scope of this piece. (For the broader case, see the benefits of progesterone.)

Here is the comparison I keep coming back to. We taper SSRIs slowly to prevent discontinuation syndrome. Yet we let the largest hormone withdrawal of a woman’s life happen with no consideration of a smoother landing. That is less a gap in evidence than a gap in imagination.

What does a prevention plan look like before delivery?

The patients who teach me the most are the ones who plan ahead. A woman tells me that with her first baby she had postpartum depression, and that it was awful. Now she is pregnant again and scared of what is coming. With her, we map it out before delivery. Progesterone starts immediately postpartum. Thyroid labs happen as soon as she is home and settled, and close follow-up runs from there. The difference that planning makes is hard to overstate.

That approach generalizes. Risk stratification belongs before delivery, not after:

1. Screen for risk in pregnancy. Personal or family history of bipolar disorder, prior postpartum psychosis or depression, autoimmune thyroid disease. If she is high risk, she needs to know it, her partner needs to know it, and her OB needs to know it.
2. Coordinate with psychiatry for high-risk women. Not every case calls for it, but for women with these risk factors it makes sense, and we should not work in a silo. Hormone and thyroid work belongs alongside psychiatric care, not in place of it.
3. Run a full thyroid panel each trimester and again postpartum. TSH, free T4, free T3, and TPO antibodies, so you have a baseline to compare against.
4. Protect sleep. In high-risk women, around six hours of sleep in the first weeks should not be optional. Build the support to make it happen. Progesterone helps here as well.
5. Set a check-in cadence. Contact at 72 hours, one week, two weeks, four weeks, and eight weeks, then monthly through six months. Most of these can be a five-minute call.
6. Discuss progesterone as a mechanism-based, off-label option for the highest-risk women, with the evidence limits stated clearly.

The OB remains the quarterback. Providers who understand women’s hormone physiology can cover the areas that currently fall between visits.

Key takeaways

• Postpartum psychosis is a psychiatric emergency that requires hospitalization, and it is also an endocrine event.
• Progesterone falls from roughly 100 to 150 ng/mL to near zero within about 48 hours of delivery, taking allopregnanolone and its GABA-A activity with it.
• The FDA approved brexanolone (2019, since withdrawn) and zuranolone (2023) for postpartum depression on that same mechanism.
• Postpartum thyroiditis affects 5 to 10 percent of women, and both of its phases can look psychiatric.
• A single TSH at six weeks is not a screen. Order TSH, free T4, free T3, and TPO antibodies.
• TPO antibodies predict postpartum thyroiditis and postpartum depression, and lower free T3 and free T4, even inside the normal range, track with early postpartum mood symptoms.
• Progesterone for postpartum prevention is mechanism-based and off-label, with no controlled trials yet. Synthetic progestin trials do not answer the question.
• The plan starts before delivery: risk screening, psychiatric coordination, thyroid baselines, sleep protection, and frequent check-ins.

The bottom line for prescribers

We do not need to wait for a news story to take the postpartum endocrine picture seriously. The mechanism is understood, the thyroid labs are cheap, and we can identify the highest-risk women before they deliver. Physiology will not replace psychiatric care in postpartum psychosis. It can give us a better chance of reaching women before the crisis does.

[Podcast embed: Ep 111, Postpartum Psychosis Is an Endocrine Event]

Frequently asked questions

What is postpartum thyroiditis?

Postpartum thyroiditis is autoimmune inflammation of the thyroid in the year after delivery. It affects roughly 5 to 10 percent of women. It can present as a hyperthyroid or hypothyroid picture, sometimes months after delivery.

Can postpartum thyroiditis cause anxiety or depression?

Yes. The thyrotoxic phase of postpartum thyroiditis commonly causes anxiety, and the hypothyroid phase causes fatigue and depressed mood. Both are easy to mistake for a primary psychiatric condition, which is why a thyroid panel belongs in any postpartum mood workup.

Which labs should I order for postpartum mood symptoms?

At minimum, TSH, free T4, free T3, and TPO antibodies. A TSH alone can miss early autoimmune disease, conversion problems, and phase changes.

How soon after delivery does postpartum psychosis start?

Most cases begin within the first two weeks, often within the first 72 hours. Atypical cases can build more slowly.

What is the difference between postpartum psychosis and postpartum depression?

Postpartum depression is a mood disorder that usually develops gradually. Postpartum psychosis is closer to a manic or mixed episode with psychotic features: confusion, disorganized thinking, mania, and sometimes delusions, with rapid onset.

Does postpartum psychosis require hospitalization?

Yes. It is a psychiatric emergency, and most women need psychiatric hospitalization.

What is allopregnanolone?

Allopregnanolone is a neurosteroid made from progesterone. It is a potent positive allosteric modulator of the GABA-A receptor, which is why progesterone affects sleep, anxiety, and mood.

Are brexanolone and zuranolone approved for postpartum psychosis?

No. The FDA approved both for postpartum depression, and it has since withdrawn brexanolone’s approval. Zuranolone is an oral neuroactive steroid.

Is progesterone used to prevent postpartum psychosis?

It is not an established treatment. Some providers use bioidentical progesterone in the postpartum window for high-risk women on a mechanism-based, off-label basis. There are no controlled trials yet, and it does not replace psychiatric prophylaxis.

Why did the progestin trial make depression worse?

The trial that exists used a synthetic progestin, which cannot convert to allopregnanolone, and it raised the risk of postpartum depression. No one has yet tested natural progesterone for this in a randomized trial.

How often should I check thyroid labs after delivery?

My own practice is a full panel quarterly for at least the first year postpartum in women with mood symptoms or risk factors.

Advance how you reason through cases like this

The Master Course teaches exactly this kind of reasoning. It covers how progesterone, estradiol, and thyroid interact, and how to read a thyroid panel for function instead of range. It also covers how to think through a case when the standard workup is not giving you the answer. Six clinical modules, 30 CE hours including 12 in pharmacology, jointly accredited through Pinnacle Conference, LLC.

Explore the master course

References

1. Boston Globe. Timeline: Lindsay Clancy was prescribed 13 different medications for symptoms including depression and psychosis. July 27, 2026. https://www.bostonglobe.com/2026/07/27/metro/lindsay-clancy-timeline-medications-13/
2. ABC News. Lindsay Clancy trial declared a mistrial. What happens next? September 2026. https://abcnews.com/GMA/News/lindsay-clancy-jury-deadlocked/story?id=136153308
3. VanderKruik R, Barreix M, Chou D, et al. The global prevalence of postpartum psychosis: a systematic review. BMC Psychiatry. 2017;17:272. https://doi.org/10.1186/s12888-017-1427-7
4. Bergink V, Rasgon N, Wisner KL. Postpartum psychosis: madness, mania, and melancholia in motherhood. Am J Psychiatry. 2016;173(12):1179-1188. https://doi.org/10.1176/appi.ajp.2016.16040454
5. Bergink V, Kushner SA, Pop V, et al. Prevalence of autoimmune thyroid dysfunction in postpartum psychosis. Br J Psychiatry. 2011;198(4):264-268. https://doi.org/10.1192/bjp.bp.110.082990
6. Tulchinsky D, Hobel CJ, Yeager E, Marshall JR. Plasma estrone, estradiol, estriol, progesterone, and 17-hydroxyprogesterone in human pregnancy. I. Normal pregnancy. Am J Obstet Gynecol. 1972;112(8):1095-1100. https://doi.org/10.1016/0002-9378(72)90185-8
7. Schiller CE, Meltzer-Brody S, Rubinow DR. The role of reproductive hormones in postpartum depression. CNS Spectr. 2015;20(1):48-59. https://doi.org/10.1017/S1092852914000480
8. Melcangi RC, Panzica GC. Allopregnanolone: state of the art. Prog Neurobiol. 2014;113:1-5. https://doi.org/10.1016/j.pneurobio.2013.09.005
9. Maguire J, Mody I. GABA(A)R plasticity during pregnancy: relevance to postpartum depression. Neuron. 2008;59(2):207-213. https://doi.org/10.1016/j.neuron.2008.06.019
10. Meltzer-Brody S, Colquhoun H, Riesenberg R, et al. Brexanolone injection in post-partum depression: two multicentre, double-blind, randomised, placebo-controlled, phase 3 trials. Lancet. 2018;392(10152):1058-1070. https://doi.org/10.1016/S0140-6736(18)31551-4
11. Epperson CN, Rubinow DR, Meltzer-Brody S, et al. Effect of brexanolone on depressive symptoms, anxiety, and insomnia in women with postpartum depression: pooled analyses from 3 double-blind, randomized, placebo-controlled clinical trials in the HUMMINGBIRD clinical program. J Affect Disord. 2023;320:353-359. https://doi.org/10.1016/j.jad.2022.09.143
12. Food and Drug Administration. Sage Therapeutics, Inc.; Withdrawal of Approval of a New Drug Application for ZULRESSO (brexanolone). Federal Register. March 14, 2025. https://www.federalregister.gov/documents/2025/03/14/2025-04101/sage-therapeutics-inc-withdrawal-of-approval-of-a-new-drug-application-for-zulresso-brexanolone
13. Deligiannidis KM, Meltzer-Brody S, Maximos B, et al. Zuranolone for the treatment of postpartum depression. Am J Psychiatry. 2023;180(9):668-675. https://doi.org/10.1176/appi.ajp.20220785
14. Bloch M, Schmidt PJ, Danaceau M, et al. Effects of gonadal steroids in women with a history of postpartum depression. Am J Psychiatry. 2000;157(6):924-930. https://doi.org/10.1176/appi.ajp.157.6.924
15. Alexander EK, Pearce EN, Brent GA, et al. 2017 Guidelines of the American Thyroid Association for the diagnosis and management of thyroid disease during pregnancy and the postpartum. Thyroid. 2017;27(3):315-389. https://doi.org/10.1089/thy.2016.0457
16. Lambrinoudaki I, Rizos D, Armeni E, et al. Thyroid function and postpartum mood disturbances in Greek women. J Affect Disord. 2010;121(3):278-282. https://doi.org/10.1016/j.jad.2009.07.001
17. Wesseloo R, Kamperman AM, Bergink V, Pop VJM. Thyroid peroxidase antibodies during early gestation and the subsequent risk of first-onset postpartum depression: a prospective cohort study. J Affect Disord. 2018;225:399-403. https://doi.org/10.1016/j.jad.2017.08.058
18. Backer S, Yancheva J, Garcia C, Khanna D. Thyroid predictors of postpartum mood disorders. Cureus. 2023;15(9):e45554. https://doi.org/10.7759/cureus.45554
19. Barak Y, Glue P. Progesterone loading as a strategy for treating postpartum depression. Hum Psychopharmacol. 2020;35(3):e2731. https://doi.org/10.1002/hup.2731
20. Dennis CL, Ross LE, Herxheimer A. Oestrogens and progestins for preventing and treating postpartum depression. Cochrane Database Syst Rev. 2008;(4):CD001690. https://doi.org/10.1002/14651858.CD001690.pub2
21. Stefaniak M, Dmoch-Gajzlerska E, Jankowska K, et al. Progesterone and its metabolites play a beneficial role in affect regulation in the female brain. Pharmaceuticals (Basel). 2023;16(4):520. https://doi.org/10.3390/ph16040520

This article reflects the Ep 111 recording, made before the September 4, 2026 mistrial in the Clancy case.