Werner & Ingbar's The Thyroid: A Fundamental & Clinical Text, 9th Edition

65.Myxedema Coma

Leonard Wartofsky

Myxedema coma is severe, life-threatening hypothyroidism. It was probably first reported in 1879 by Ord from the St. Thomas Hospital, London. Two of 12 patients with fatal hypothyroidism described in a report of the Clinical Society of London in 1888 appeared to have died in coma (1). Remarkably, the next cases did not appear in the literature until 1953 (2,3), and about 200 cases have been reported subsequently. Additional references may be obtained from earlier reviews (4,5).

Most patients with myxedema coma are elderly women with long-standing hypothyroidism and who therefore usually have myxedema and other characteristic clinical manifestations of hypothyroidism. Once considered, the diagnosis should be easy to establish on the basis of both clinical and laboratory findings; despite vigorous therapy, the mortality rate may be high.


Myxedema coma is more likely to occur in the winter than in the summer, suggesting that low environmental temperature may somehow precipitate the syndrome, and hypothermia is a cardinal clinical finding. Other events that may trigger the onset of myxedema coma include pulmonary and other infections, cerebrovascular accidents, and congestive heart failure (Table 65.1). Pulmonary infection may occur as a secondary event because of hypoventilation due to somnolence, as can aspiration pneumonia. Similarly, it may be difficult to determine whether other abnormalities, such as hypoglycemia, hyponatremia, hypoxemia, and hypercapnia, which often are present in patients with myxedema coma, contributed to the onset of the coma or are consequences of it. In other patients, sedative, analgesic, antidepressant, hypnotic, and anesthetic drugs have been incriminated as precipitating or exacerbating myxedema coma because of their ability to depress respiration. Drug-induced myxedema coma is particularly likely to occur in hospitalized patients because these types of drugs are more likely to be dispensed there than outside of the hospital, and the patients are typically not known to have hypothyroidism.




Cerebrovascular accidents

Congestive heart failure








Gastrointestinal bleeding

Metabolic disturbances exacerbating myxedema coma








Most patients with myxedema coma have had symptoms of hypothyroidism for many months, and the onset of stupor or coma is precipitated by cold exposure, infection, or other systemic disease or by drugs, as noted above. There may be a past history of antecedent thyroid disease, thyroid hormone therapy that was discontinued for no apparent reason, or radioiodine therapy. Examination of the neck may reveal a surgical scar and no palpable thyroid tissue or a goiter. Approximately 5% to 10% of patients have hypothalamic or pituitary disease as the cause of their hypothyroidism.

A survey of hospitals in Germany (1993–1995) identified 24 patients with myxedema coma, probably the largest number ever analyzed together, although the authors reclassified 12 patients as having severe hypothyroidism but not coma (6). The mean age of the 24 patients (20 women, 4 men) was 73 years. Twenty-three had primary hypothyroidism and one central hypothyroidism; nine were known to have hypothyroidism before the episode of coma. With respect to clinical findings, 80% had hypoxemia, 54% had hypercapnia, and 21 had a temperature less than 94°F. All received some thyroid hormone therapy, and six (25%) died.

The course often is one of lethargy progressing to stupor and then coma, with respiratory failure and hypothermia, all of which may be hastened by the administration of drugs that depress respiration and other central nervous system (CNS) functions. Most patients have the characteristic features of severe hypothyroidism, such as dry, coarse, and scaly skin; sparse or coarse hair; nonpitting edema (myxedema) of the periorbital regions, hands, and feet; macroglossia; hoarseness; and delayed deep-tendon reflexes. Moderate to profound hypothermia is common.

Respiratory System

Respiratory depression, common in patients with myxedema coma, is caused by decreased hypoxic respiratory drive and a decreased ventilatory response to hypercapnia (7) (see Chapter 54). The resulting alveolar hypoventilation leads to progressive hypoxemia, and ultimately to COnarcosis and coma. Impaired respiratory muscle function may exacerbate the hypoventilation (8,9). Obesity also can exacerbate hypoventilation, although its importance has been questioned (10). With respect to coma, the principal factor seems to be a depressed respiratory center response to CO2 (11,12). Whether attributable to central respiratory depression or to respiratory muscle dysfunction, hypoventilation is sufficiently severe in most patients with myxedema coma to require mechanically assisted ventilation. Other factors that may impair respiration in severely hypothyroid patients include pleural effusions or ascites, by reducing lung volume, and macroglossia and myxedema of the nasopharynx and larynx by reducing the lumen of the airway. Recovery from respiratory failure may be slow (13) despite apparently adequate therapy.

Cardiovascular Manifestations

The findings considered typical of hypothyroid heart disease also are found in myxedema coma and include cardiac enlargement, bradycardia, decreased cardiac contractility, and nonspecific electrocardiographic abnormalities (see Chapter 53). The cardiac enlargement may be due to either ventricular dilatation or pericardial effusion. The decrease in cardiac contractility results in a low stroke volume and low cardiac output, but frank congestive heart failure is rare.

Patients with myxedema coma may have hypotension because of decreased intravascular volume, and cardiovascular collapse and shock may occur late in the course. If at all, the latter characteristically responds only when both a vasopressor drug such as dopamine and thyroid hormone are given.

Gastrointestinal Manifestations

A neurogenic oropharyngeal dysphagia has been described that is associated with delayed swallowing, aspiration, and risk of aspiration pneumonia (14). Decreased intestinal motility is common in hypothyroidism, and its most severe manifestations, paralytic ileus and megacolon, can occur in patients with myxedema coma. Thus, many patients have anorexia, nausea, abdominal pain, constipation, and a distended, quiet abdomen. Gastric atony also occurs and can be a particularly troublesome problem because it may serve to reduce absorption of oral medications.

Renal and Electrolyte Manifestations

Hyponatremia and a decreased glomerular filtration rate are rather consistent findings among patients with myxedema coma. The hyponatremia is due to the inability to excrete a water load, caused both by decreased delivery of water to the distal nephron (15) and excess vasopressin secretion (16) (see Chapter 55). Urinary sodium excretion is normal or increased, urinary osmolality is high relative to plasma osmolality, and there may be bladder atony, with urinary retention. Hyponatremia alone may cause lethargy and confusion, and it undoubtedly contributes further to these problems in patients who have them as a result of hypothyroidism.

Neuropsychiatric Manifestations

Just as is seen in other patients with hypothyroidism (see Chapter 64). Patients with myxedema coma may have a history of lethargy, slowed mentation, poor memory, cognitive dysfunction, depression, or even psychosis. They do not complain of these symptoms, however, because of their impaired state of consciousness. Up to 25% of patients with myxedema coma have focal or generalized seizures caused by CNS dysfunction per se, hyponatremia, or hypoxemia resulting from decreased cerebral blood flow (17).


The high mortality rate of myxedema coma may be due, at least in part, to infection. Complicating our ability to detect infection is the hypothermia commonly present in hypothyroidism, and even if the patient is normothermic, the temperature may not be increased much by an infection. Thus, infection may be difficult to recognize or even overlooked entirely and antibiotic treatment consequently delayed or not given at all. Similarly, the patients often perspire little, and their pulse rate tends to be slow, both of which may result in the likelihood that the physician will overlook the possibility of an acute infection. Pulmonary infections may aggravate or even cause hypoventilation, and susceptibility to these infections may be increased because of aspiration caused by neurogenic dysphagia, semicoma, or seizures (14,17).

These considerations led to the suggestion that all patients with myxedema coma be given antibiotic therapy (18). A more prudent approach is to search vigorously for infection but to initiate antibiotic therapy only when some evidence of infection is present.


Hypothermia is present in about three fourths of patients. It may be dramatic (< 80°F), and it may be the first clinical clue to the diagnosis of myxedema coma. Conversely, the diagnosis should be considered seriously in any unconscious patient with an infection who does not have fever. There is a correlation of survival with presenting body temperature, with patients having core temperatures below 90°F having the poorest prognosis. Hypoglycemia may decrease body temperature further. It is fortunate that old-style mercury thermometers are rarely used today because the degree of hypothermia may be underestimated if the mercury column has not been lowered to well below normal before the patient's temperature is measured. Moreover, these thermometers do not record below 94°F, and therefore the degree of hypothermia may be underestimated. Electronic thermometers record temperature over a much wider range.


The presence of marked stupor, confusion, or coma in a patient with a history and physical findings of hypothyroidism, especially if the patient has hypothermia, should suggest the diagnosis of myxedema coma. Treatment should be initiated immediately, without awaiting the results of measurements of serum thyrotropin (TSH) and thyroxine (T4). Marked fatigue and somnolence often occur in patients with severe hypothyroidism, and a clear distinction between them and coma may be neither practical nor sensible. Because of the risks for high-dose thyroid hormone therapy in elderly patients, therapy should not be undertaken lightly. Therefore, in a lethargic or somnolent but not comatose patient, the presence of abnormalities often present in myxedema coma, such as hypothermia, hypoventilation, and hyponatremia, should be used to determine whether or not to initiate high-dose thyroid hormone therapy.

As mentioned, the typical patient is an older woman who is brought to the hospital during the winter in a severely obtunded state, if not comatose. If a history can be elicited, there often is a history of thyroid disease. Physical examination may reveal bradycardia and the skin and other abnormalities described earlier, in addition to hypoventilation and hypothermia. Routine laboratory evaluation may reveal anemia, hyponatremia, hypercholesterolemia, and high serum lactate dehydrogenase and creatine kinase concentrations.

In many patients, the clinical features may be sufficiently clear to make measurements of serum TSH and T4 necessary only for confirmation of the diagnosis. Nonetheless, in many hospitals, both can be measured in several hours on an emergency basis. Although a high serum TSH concentration is the most important laboratory evidence of the diagnosis, severe systemic illness can lower TSH secretion in hypothyroid patients, and therefore the values may not be notably high (19,20). Also, as noted, a few patients with myxedema coma have central hypothyroidism, with normal or low serum TSH concentrations (see Chapter 51). Nearly all patients with myxedema coma have very low serum total and free T4 and triiodothyronine (T3) concentrations, and associated nonthyroidal illness will contribute to the observed reduction in serum T3 concentration.


In view of the high mortality rate among patients with myxedema coma, treatment should be instituted promptly as soon as the diagnosis is strongly suspected. Treatment with thyroid hormone alone without addressing all of the physiologic and metabolic derangements described herein is inadequate therapy, and will likely contribute to a poor prognosis. All patients should be admitted to an intensive care unit so that their pulmonary and cardiac status can be monitored continuously. A central venous pressure line should be used to monitor volume repletion therapy, and placement of a Swan-Ganz catheter is justifiable in patients with cardiac disease.

Ventilatory Support

Hypoventilation is an important component of myxedema coma and is a common cause of death in these patients. Evaluation of respiratory function should include assessment of not only pulmonary function (blood gas measurements), but also the possibility of pulmonary infection and airway obstruction by macroglossia or myxedema of the larynx. Thus, in addition to ventilatory support to relieve or prevent hypoxemia and hypercapnia, antibiotic therapy and steps to maintain an open upper airway may be indicated. Most patients require mechanical ventilatory support for 24 to 48 hours, especially those in whom the hypoventilation was caused by drug-related respiratory depression, and some may require it for several weeks (13).

During the period of ventilatory support, arterial blood gases should be measured frequently, and it may be necessary to insert an endotracheal tube or even perform tracheostomy to ensure adequate oxygenation. The tube should not be removed until the patient is fully conscious and there is evidence that the removal will be successful. Very rarely, a patient with myxedema coma may require emergency surgery and management of these patients should adhere to the same general principles (21).


Hyponatremia undoubtedly contributes to the mental status changes in patients with myxedema coma, especially in patients with serum sodium concentrations of less than 120 mEq/L. Therefore, although comatose patients must be given some saline (and glucose) intravenously to replace daily losses, the volume should be limited in those with mild to moderate hyponatremia (serum sodium concentrations of 120–130 mEq/L) such that all water lost is not replaced. In patients with serum sodium concentrations of less than 120 mEq/L, it may be appropriate to administer a small amount of hypertonic saline (50–100 mL of 3% sodium chloride), followed by an intravenous bolus dose of 40 to 120 mg of furosemide to promote a water diuresis (22).


Thyroid hormone must be given to restore body temperature to normal, but its action is slow. Therefore, warming of patients with hypothermia and maintaining normal temperature in the others with an electric blanket is advisable, but should be done cautiously so as not to cause vasodilatation with a decrease in peripheral vascular resistance and hypotension.


Hypotension should be corrected by judicious administration of intravenous fluid, initially 5% to 10% glucose in half-normal sodium chloride or as isotonic sodium chloride if hyponatremia is present. Administration of hydrocortisone (100 mg intravenously every 8 hours) is indicated if there is any suspicion of adrenal insufficiency. Rare patients may require vasopressor drug therapy to maintain a blood pressure sufficient to sustain adequate perfusion until thyroid hormone action begins. Although an adverse interaction between administered vasopressor drug and thyroid hormone is possible, this risk is counterbalanced by the high mortality rate among patients with myxedema coma who have hypotension unresponsive to initial fluid therapy.

Glucocorticoid Therapy

The few patients with myxedema coma who have central hypothyroidism are likely to have corticotropin (ACTH) deficiency as well as TSH deficiency. Some patients with primary hypothyroidism also have primary adrenal insufficiency (Schmidt's syndrome) (see Chapter 59). The coexistence of adrenal insufficiency in patients with myxedema coma may be suggested by the presence of hypotension, hypoglycemia, hyponatremia, hyperkalemia, and azotemia; however, most patients with myxedema coma have normal serum cortisol concentrations, although their ACTH and cortisol responses to stress may be slightly impaired.

It is prudent to administer hydrocortisone or another glucocorticoid to patients with myxedema coma based on the occasional presence of definite adrenal insufficiency, the possibility of impaired ACTH and cortisol responses to stress, and the possibility that thyroid hormone therapy may increase cortisol clearance. Moreover, short-term glucocorticoid therapy is safe and can be discontinued when the patient has improved and pituitary–adrenal function has been assessed to be adequate. Hydrocortisone usually is given intravenously in a dosage of 50 to 100 mg every 6 to 8 hours for several days, after which it is tapered and discontinued on the basis of clinical response and plans for further diagnostic evaluation.

Thyroid Hormone Therapy

The most controversial aspect of the treatment of patients with myxedema coma concerns what may be the most appropriate method for restoring to normal the low serum and tissue thyroid hormone concentrations. The main controversy concerns whether to administer T4 alone, with conversion to T3 being dependent on the deiodinase activity in the patient, or to directly administer T3 itself. Secondary concerns include dose, frequency, and route of administration (of either hormone). Different approaches to treatment are based on balancing concerns for the high mortality of untreated myxedema coma and the obvious need for attaining effective thyroid hormone concentrations in different tissues fairly rapidly, against the risks of high-dose thyroid hormone therapy, which may include atrial tachyarrhythmias or myocardial infarction. Because of the rarity of myxedema coma and the paucity of studies of the effects of treatment, the optimal therapy remains uncertain and several different approaches have been used.

Advocates of administering T4 alone point out that it provides a steady and smooth, but rather slow, onset of action with low risk for adverse effects. Conversely, the onset of action of T3 is quicker, and its serum (and probably tissue) concentrations fluctuate more between doses. Another argument for using T4 is because it is somewhat easier to measure than is serum T3, and the results are easier to interpret because the values vary less between doses. In either case, serum TSH values should provide information about the impact of treatment at the tissue level, with the caveat that the sick patient may exhibit TSH suppression as part of the nonthyroidal illness syndrome.

Although T4 (or T3) tablets can be given by nasogastric tube, administration by this route has the risks for aspiration and uncertain absorption, particularly in patients with gastric atony. Preparations of T4 for parenteral administration are available in vials containing 100 and 500 mg. A high dose, given as a single intravenous bolus dose, was popularized by a report suggesting that replacement of the entire extrathyroidal pool of T4 (usually 300–600 mg) was desirable to restore near-normal hormonal status as rapidly as possible, with the pool then maintained by administration of 50 to 100 mg daily given either intravenously or orally (23). With this regimen, serum T4 concentrations increase abruptly to supranormal values and decrease to within normal range in 24 hours as the administered T4 is distributed throughout the extracellular and then intracellular spaces, serum T3 concentrations increase slightly, and serum TSH concentrations decrease substantially (24). Although the importance of extrathyroidal T4 conversion to T3 was not known when this regimen was proposed, the approach proved effective. The 24 patients with myxedema coma or severe hypothyroidism reported from Germany (described earlier) all were treated initially with T4, in doses ranging from 25 to 500 µg, with six deaths (6).

The important potential drawback to total reliance on generation of T3 from T4 is that the rate of extrathyroidal conversion of T4 to T3 is reduced in patients with virtually all illnesses, including hypothyroidism (20) (see section on nonthyroidal illness in Chapter 11D). Furthermore, the onset of action of T3 is considerably more rapid than is that of T4, which some argue may be crucial to increasing survival (22), and in baboons it crosses the blood–brain barrier more readily than does T4 (25).

For intravenous administration, T3 is available in vials containing 10 µg. When given alone, the usual dose is 10 to 20 µg followed by 10 µg every 4 hours for the first 24 hours, then 10 µg every 6 hours for a day or two, after which oral administration should be feasible. Increases in body temperature and oxygen consumption may occur 2 to 3 hours after intravenous administration of T3, compared with 8 to 14 hours after intravenous administration of T4. Significant clinical improvement may be seen within 24 hours with T3 (26), but the more rapid action of T3 may be associated with a higher risk of adverse cardiovascular actions, and in one report high serum T3 concentrations during treatment with T4 alone were associated with fatal outcome in several patients (27). In another series of eight patients, two of three patients treated with high-dose T4 died, whereas the other five who were treated with smaller doses of T4 or T4 survived (28). These researchers reviewed 82 cases from the literature and found that advanced age, high-dose T4 therapy, and cardiac complications were associated with mortality. They concluded that a 500 µg dose of T4 should be safe in younger patients, but lower doses are safer in the elderly.

Consequently, our approach to therapy is to prudently administer both T4 and T3. The T4 is given intravenously in a dose of 4 µg/kg lean body weight (or about 200–250 µg), followed by 100 µg 24 hours later and then 50 µg daily, either intravenously or orally, as appropriate. The dose subsequently is adjusted on the basis of clinical and laboratory results, as in any other hypothyroid patient. With respect to T3, the initial intravenous dose is 10 µg, and the same dose is given every 8 to 12 hours until the patient can take maintenance oral doses of T4.

No general guide to treatment can take into account all the factors that might affect sensitivity to thyroid hormone, such as age, intrinsic cardiovascular function, and neuropsychiatric status. Hence, patients should be monitored closely before each dose of thyroid hormone is administered. In addition to the specific treatments considered above, attention should be given to any comorbid conditions, recognizing that drug dosages may need to be modified because hypothyroidism can result in altered drug distribution and metabolism.

With aggressive comprehensive treatment, most patients with myxedema coma should recover. Better, however, that it be prevented, by recognition of hypothyroidism earlier and by ensuring that patients with hypothyroidism do not discontinue therapy because they feel well.


1. Report of a committee of the Clinical Society of London to investigate the subject of myxedema. Trans Clin Soc (Lond) (Suppl) 1888:21.

2. LeMarquand HS, Hausmann W, Hemstead EH. Myxedema as a cause of death. BMJ 1953;1:704.

3. Summers VK. Myxedema coma. BMJ 1953;2:336.

4. Ringel MD. Management of hypothyroidism and hyperthyroidism in the intensive care unit. Crit Care Clin 2001;17:59–74.

5. Nicoloff JT, LoPresti JS. Myxedema coma: a form of decompensated hypothyroidism. Endocrinol Metab Clin North Am 1993; 22:279.

6. Reinhardt W, Mann K. Incidence, clinical picture, and treatment of hypothyroid coma: results of a survey. Med Klin 1997; 92:521.

7. Zwillich CW, Pierson DJ, Hofeldt FD, et al. Ventilatory control in myxedema and hypothyroidism. N Engl J Med 1975;292: 662.

8. Martinez FJ, Bermudez-Gomez M, Celli BR. Hypothyroidism: a reversible cause of diaphragmatic dysfunction. Chest 1989;96: 1059.

9. Wilson WR, Bedell GM. The pulmonary abnormalities in myxedema. J Clin Invest 1960;39:42.

10. Massumi RA, Winnacker JL. Severe depression of the respiratory center in myxedema. Am J Med 1964;36:876.

11. Ladenson PW, Goldenheim PD, Ridgway EC. Prediction of reversal of blunted respiratory responsiveness in patients with hypothyroidism. Am J Med 1988;84:877.

12. Domm BB, Vassallo CL. Myxedema coma with respiratory failure. Am Rev Respir Dis 1973;107:842.

13. Yamamoto T. Delayed respiratory failure during the treatment of myxedema coma. Endocrinol Jpn 1984;31:769.

14. Urquhart AD, Rea IM, Lawson LT, Skipper M. A new complication of hypothyroid coma: neurogenic dysphagia: presentation, diagnosis, and treatment. Thyroid 2001;11:595–598.

15. DeRubertis FR Jr, Michelis MF, Bloom MG, et al. Impaired water excretion in myxedema. Am J Med 1971;51:41.

16. Skowsky RW, Kikuchi TA. The role of vasopressin in the impaired water excretion of myxedema. Am J Med 1978;64:613.

17. Sanders V. Neurologic manifestations of myxedema. N Engl J Med 1962;266:547.

18. Lindberger K. Myxoedema coma. Acta Med Scand 1975;198:87.

19. Hooper MJ. Diminished TSH secretion during acute non-thyroidal illness in untreated primary hypothyroidism. Lancet 1976;1:48.

20. Wartofsky L, Burman KD. Alterations in thyroid function in patients with systemic illness: the euthyroid sick syndrome. Endocr Rev 1982;3:164.

21. Mathes DD Treatment of myxedema coma for emergency surgery. Anesth Analg 1997;85:30–36, and 1998;86:450–451.

22. Pereira VG, Haron ES, Lima-Neto N, et al. Management of myxedema coma: report on three successfully treated cases with nasogastric or intravenous administration of triiodothyronine. J Endocrinol Invest 1982;5:331.

23. Holvey DN, Goodner CJ, Nicoloff JT, et al. Treatment of myxedema coma with intravenous thyroxine. Arch Intern Med 1964; 113:89.

24. Ridgway EC, McCammon JA, Benotti J, et al. Metabolic responses of patients with myxedema to large doses of intravenous L-thyroxine. Ann Intern Med 1972;77:549.

25. Chernow B, Burman KD, Johnson DL, et al. T3 may be a better agent than T4 in the critically ill hypothyroid patient: evaluation of transport across the blood–brain barrier in a primate model. Crit Care Med 1983;11:99.

26. MacKerrow SD, Osborn LA, Levy H, Eaton RP, Economou P. Myxedema-associated cardiogenic shock treated with intravenous triiodothyronine. Ann Int Med 1992;117:1014–1015.

27. Hylander B, Rosenqvist U. Treatment of myxoedema coma: factors associated with fatal outcome. Acta Endocrinol 1985; 108: 65.

28. Yamamoto T, Fukuyama J, Fujiyoshi A. Factors associated with mortality of myxedema coma: report of eight cases and literature survey. Thyroid 1999;9:1167–1174.