Thyroid and Psychiatric Illness

The thyroid is one of those glands that is hooked into everything.  Mood, cognition, metabolism, bones, heart, cholesterol… all can be affected by perturbations among the thyroid hormones.

Chris Kresser has done a great series of articles and it's worth a look if you are unfamiliar with the thyroid or if you want a comprehensive refresher.  In fact he did such a good job it seems hardly worth reinventing the wheel, so I thought I would start with psychiatric disease and then address any questions and chase the rabbit holes that will inevitably open up.

My information today comes from a review done for the American Journal of Psychiatry this month and pointed out to me by my fantastic colleague Dr. Hale:  Abnormal Thyroid Function Tests in Psychiatric Patients:  A Red Herring?  This review is conventional to the extreme, but I find it is often very useful to start with conventional reasoning and pick apart where there may be issues or something missing.

I will begin with the briefest of primers so that we are more or less on the same page:

The Killers, All These Things That I've Done (right click to open in new tab)

All right.  As with everything endocrine, there is a feedback loop, but for our purposes we will start with the hypothalamus, which makes a chemical called thyrotropin releasing hormone (TRH).  This chemical toodles on down to the anterior pituitary, which produces thyroid stimulating hormone (TSH).  TSH then directs the thyroid gland itself to release thyroid hormone.  The two major ones are T3 (triiodothyronine) and T4 (thyroxine).  High levels of T3 and T4 should feedback to the pituitary and cause it to decrease the amount of TSH secreted, thus nicely regulating itself.  The other important thing to know is that iodine is essential for the creation of the thyroid hormones, and that T4 is actually a prohormone (and is the major hormone secreted by the thyroid) while T3 is the active hormone.  Selenium is required to turn T4 into T3.

Courtesy Wikipedia Commons

Lots of things can go wrong within this complex system.  For example, a nodule in the thyroid can start producing thyroid hormone like gangbusters and won't respond to feedback inhibition.  This is hyperthyroidism, with classic lab results of high T3, high T4, and typically undetectable TSH.  Symptoms are a racy metabolism, so weight loss, rapid heartbeat, somewhat elevated body temperature, insomnia, anxiety, etc.

In the opposite case, the thyroid can stop responding to the pituitary TSH stimulation and stop producing thyroid hormone.  This is known as hypothyroidism, and will result in low T3 and T4 with a very high TSH.  Symptoms can include weight gain, fatigue, depression, cold intolerance, hair loss, and slow heart rate among others.

These two primary thyroid disorders can cause a host of psychiatric symptoms, including depression, mania, dementia, and even psychosis.   If someone comes to my office with weight gain, cold intolerance, fatigue, and depression, I'd better well check the thyroid and grab a pulse rate while I'm at it.  I'd look pretty silly treating hypothyroidism with therapy or antidepressants.

Besides these basic thyroid problems, thyroid hormones can be off kilter in a variety of other ways.  Chris Kresser's series goes into great detail, but with rare exceptions, these abnormalities are not due to thyroid problems in the actual gland.  Alterations in thyroid function can occur in response to all sorts of systemic illnesses, stress states, and medications, and perturbations in the thyroid function tests not thought to be due to actual hypo or hyperthyroidism is called "nonthyroidal illness." In general, any pattern of tests that don't quite fit the classic  hypo/hyperthyroid patterns will indicate a nonthyroidal illness.

Sepsis, heart attack, major surgery, autoimmune disease… all of these can lead to a characteristic pattern of thyroid tests including low T3, normal to low T4, and high reverse T3.  TSH is often normal, but can be low and later become elevated during recovery.  These abnormalities are extremely common and are seen in about 75% of hospitalized medical patients.  The kicker is these abnormalities seem to be a physiologic response to the illness and they resolve without any intervention in a few months.  The reason these changes happen is (as always) complicated, but inflammatory cytokines such IL-6, IL-1, and TNF alpha are likely involved, altering feedback regulation along the hypothalamic-pituitary-thyroid axis.

Starvation, fasting, or a very low carb diet can tend to lead to low TSH, normal or slightly elevated free T4, and low T3.  There is nothing wrong with the thyroid and this is not "hypothyroidism" per se, but a normal physiologic response to perceived starvation, and it should resolve without other intervention once someone stops fasting or increases carbohydrate intake.  Sepsis (severe infection) will often present with low TSH, normal free T4, and low free T3, a similar pattern.  Again, once the infection is cleared, the abnormalities will resolve on their own.  This "low T3" syndrome is a bad predictor in the case of cardiovascular disease (1), but that doesn't mean that treating with T3 would be smart.  There is some reason to suspect that "low T3" may be a maladaptive response and T3 after a heart attack may be a good idea after all (2), but I think the proof will be in future research (3).

As we noted before, primary thyroid problems can cause psychiatric symptoms.  However, the reverse is also true, and up to 33% of psychiatric inpatients will have abnormal thyroid tests.  As in the case of the medical "nonthyroidal illness," these abnormalities will tend to be characteristic of the disorder and will also resolve on their own within a few months.  In acute psychosis, for example, TSH is usually normal while total T4 is often elevated.  In mania, total T4 and free T4 will be elevated.  In depression, TSH may be a little low or high, with a high free T4 and low or high total T3.

In general, it is recommended that nonthyroidal illness is NOT treated with thyroid hormone.  In nonthyroidal illness, the low T3 or other abnormality may be part of the adaptive inflammatory response (I've linked some thoughts about possible exceptions above).  This reasoning is why conventional medicine suggests testing TSH alone, and checking other hormone levels only if the TSH is out of whack.  Without multiple findings indicative of thyroid problems (a typical hyper or hypothyroid) symptom complex, it is unlikely to be a primary thyroid issue, except in the elderly, who can sometimes show few symptoms.    Other folks with a history or family history of autoimmune disease, treatment with lithium, radiation exposure, goiter, etc. are obviously at higher risk for primary thyroid disease and one should have higher suspicion.

"Subclinical hypothyroidism" is a bit of a gray area between true thyroid illness and the nonthyroidal problems.  TSH will be high, while free T4 will be low or normal. Usually multiple hypothyroid symptoms will indicate a thyroid problem, while no symptoms will indicate nonthyroidal illness and will resolve on its own, which should show up in serial lab tests over months.

From my perspective as a physician, I tend to rely on the TSH and not aggressively pursue mildly abnormal T4s or T3s, particularly if there are no other symptoms.  However, I think low grade iodine and selenium deficiencies are rarely thought of by a conventional physician and may lead to thyroid symptoms and subclinical or confusing lab results. ("Are you eating too much millet? Is not a typical question in the standard medical interview). In addition, there is some controversy as to the appropriate range of TSH considered normal.  In the past it was around 0.8 or 1.0-6.  Now many labs have narrowed the threshold by reducing the upper limit to 3.5 or 4.  However, in most true hyperthyroidism (99%) there will not be much of a question -- TSH will be undetectable.  In true hypothyroidism, it is not unusual to see levels higher than 20.

TSH levels >2.5 are associated with a greater risk of cardiovascular disease.  But is that due to other factors, such as systemic inflammation?  Does it make sense to use thyroid hormone to treat a mere number?  How is that different from treating a cholesterol number with a medicine for the specific purpose of getting the number into a "better" range?  I think we are still coming to a consensus as to the right thing to do in the case of "subclinical hypothyroidism."

As we discussed in the comments of a previous post, psychiatry is pretty much the only specialty where we have a large amount of data and clinical reason for using T3 hormone to treat depression symptoms.  Most primary care doctors and endocrinologists will use T4 exclusively.  In general, T4 the prohormone is safer, and the body should be able to make about as much T3 as it needs (provided selenium levels and vitamin levels are okay).  Conversely, being too aggressive with T3 can lead to death.

However, as I noted, I've found T3 to be generally ineffective or uncomfortable for most, and the only folks who seem to respond well already have diagnosed hypothyroidism and are on synthetic T4.  I've also found that the recommended psychiatry doses (25 to 50 mcg T3) are way too high, and lead to mild hyperthyroidism symptoms within several months.  I've had better luck with lower doses of T3 (around 5 mcg) combined with lowering the dose of T4.  My anecdotes are hardly data, but I think I might have caught some poor converters from T4 to T3, or maybe they are selenium deficient.  In conventional endocrinology these folks don't really exist, but I'm usually able to get away with treatment if I document "for psychiatric indication."