3-1998

Nonthyroidal Illness

E. M. Kaptein, M. D.

University of Southern California,
Los Angeles, CA, USA

Contents:

Introduction

 Clinical spectrum

 Prevalence of thyroid hormone alternations due to nonthyroidal illnesses, hypothyroidism and hyperthyroidism in sick patients

 Pathogenesis

 Clinical questions requiring further investigation

 Free T4 and free T3 estimates in nonthyroidal disorders

 Free T4 estimates

 Free T3 estimates

 The hypothalamic-pituitary-thyroid axis "set-point" in nonthyroidal illnesses

 Diagnosing hypothyroidism or hyperthyroidism in sick patients

 Should thyroid hormone therapy be given to sick patients without hypothyroidism

 E. M. Kaptein

 References

 

Introduction

Alterations of serum thyroid hormone levels in nonthyroidal illnesses are the net effect of disturbances in thyroid hormone metabolism and "artifacts" of laboratory measurements which occur in sera from sick patients (1-6). These changes in serum thyroid hormone levels are common, and may mimic or mask biochemical features of hypothyroidism or hyperthyroidism. By definition, nonthyroidal disorders reversibly alter the hypothalamic-pituitary- thyroid axis as well as peripheral thyroid hormone metabolism, in the absence of concurrent hypothalamic, pituitary or thyroid disease. Recognition of changes in serum thyroid hormone levels induced by nonthyroidal disorders is essential to accurately diagnose and treat hypothyroidism and hyperthyroidism in sick patients (3-6).

Clinical spectrum

Disturbances of serum total T4 and T3 concentrations represent a continuum of changes, primarily reflecting the severity of nonthyroidal disorders, with the lowest values occurring in the sickest patients regardless of the etiology of the underlying illnesses (Figure 1) (1,3).
View figure 1

Free T4 values by direct equilibrium dialysis are normal or elevated in the majority of patients with nonthyroidal illnesses in the absence of hypothalamic-pituitary-thyroid disease or suppression by therapeutic administration of dopamine (4,5). Free T4 values determined by tracer equilibrium dialysis underestimate free T4 levels in patients with reduced serum T4 binding, including those with severe nonthyroidal illnesses (4,5). In contrast, changes in total reverse T3 levels are more disease specific being elevated in most patients with non-renal nonthyroidal illnesses, but normal or reduced in those with acute or chronic renal failure, nephrotic syndrome with normal glomerular filtration rates, primary hyperparathyroidism and autoimmune deficiency syndrome (AIDS) (1,2). Serum TSH levels tend to be reduced in severe nonthyroidal illnesses, and elevated during recovery from acute illnesses (1-3). Further, the inverse relationship between serum TSH and free T4 values appears to be altered in patients with nonthyroidal illnesses (Figure 2) (5).

View figure 2

However, many laboratory measurements of serum free T4 do not accurately reflect circulating levels (3-5), and serum TSH assays may measure immunoactive hormone that has varying bioactivity (7,8) , thus, obscuring the TSH-free T4 relationship in sick patients.

 

Prevalence of thyroid hormone alternations due to nonthyroidal illness, hypothyroidism and hyperthyroidism in sick patients

In ambulatory and hospitalized patients, abnormal serum free T4 index and TSH values are more frequently due to nonthyroidal illnesses than to hypothyroidism or hyperthyroidism (3,5). The frequency of altered free T4 and free T3 estimates in sick sera appears to primarily depend upon the assay method (Figures 3 and 4) (5,9).
View figure 3
View figure 4

Serum TSH levels are transiently reduced to below normal values in 13% to 35% of "euthyroid" hospitalized patients with nonthyroidal illnesses (10,11). TSH values were below 0.1 mU/L (second generation assay) in 2% to 10% and below 0.01 mU/L (third generation assay) in up to 4% of these patients in the absence of hyperthyroidism, and did not correlate with alterations in free T4 estimates (10,11).

Although acute nonthyroidal illnesses frequently alter serum free T4 , free T3 and TSH values, these values normalize with recovery in euthyroid patients and are persistently abnormal in patients with concurrent thyroid diseases.

Pathogenesis

Altered serum thyroid hormone concentrations in nonthyroidal disorders reflect the net effect of multiple disturbances of thyroid hormone metabolism. These include transient changes in 1) the hypothalamic-pituitary-thyroid gland axis, 2) hormone binding in serum, 3) hormone uptake into tissues, 4) hormone clearance, production and degradation rates, as well as 5) receptor and post- receptor events (1-3).

Reduced total T4 concentrations are primarily due to impaired T4 binding to serum carrier proteins, presumably by circulating inhibitors (Table 1) (2,3). These inhibitors may also interfere with in vitro free T4 assays, inhibit T4 uptake by rat and human hepatocytes (3), and impair T4 transfer from serum to tissues in vivo in humans (3).
Table 1 Potential serum binding inhibitors in nonthyroidal illnesses 
 
Uremic serum Non-uremic serum Pharmacological agents 
CMPF NEFA/albumin Salsalate/salicylate* 
Indoxyl sulfate* Bilirubin/albumin Furosemide 
Hippuric acid* Oleic acid Heparin/NEFA 
Interleukin-1 beta Interleukin-6 Some NSAIDs 
Tumor necrosis factor Tumor necrosis factor Dilantin 
* Dialyzable in vitro 
NEFA = non-esterified fatty acids  NSAIDs = nonsteroidal anti-inflammatory drugs such as fenclofenac and mefenamic acid 
 

T4 production and degradation rates are within the normal range of values in patients with nonthyroidal illnesses (1-3), consistent with normal TSH values in most patients and normal thyroid gland function1–3. Free T4 values determined by direct equilibrium dialysis are frequently increased in nonthyroidal illnesses without TSH being suppressed (Figure 2). These elevated free T4 values may be secondary to reduced serum clearance rates of T4 (1,2). Compared to healthy euthyroid subjects with reduced serum T4 binding, patients with nonthyroidal disorders and decreased serum T4 binding have impaired fractional transfer rates of T4 from blood to tissues (1-3). Sick sera impairs T4 uptake into perfused rat liver and into rat and human hepatocytes (3). Reduced T4 uptake into tissues may lower T4 clearance rates and, thereby increase free T4 values by direct equilibrium dialysis in nonthyroidal illnesses and with caloric deprivation (1-3). Decreased total and free T3 levels in patients with nonthyroidal illnesses and caloric deprivation may primarily reflect reduced T3 production from T4 in extrathyroidal tissues (1-3). Reduced T3 binding to serum carrier proteins may play a lesser role. Decreased T3 production from T4 may be secondary to impaired T4 uptake into tissues like liver, rather than to decreased 5'-de-iodinase activity (3). Circulating inhibitors in sera from sick patients decrease T4 uptake by rat and human hepatocytes and by perfused rat liver resulting in reduced T4 availability for T3 production (3). In fasted subjects and with fructose administration, reduced T4 uptake into liver may relate to reduced ATP levels which, in turn, may impair the active transport of T4 into tissues (3). Thus, a number of factors may impair T4 uptake into tissues like liver and decrease T4 availability for T3 production in patients with nonthyroidal illnesses (3).

In patients with nonthyroidal disorders, fractional T4 transfer rates from serum to rapidly equilibrating tissues like liver are lowest in the most severely ill patients, who also have the highest free fractions of T4 by equilibrium dialysis (2). In contrast, in healthy euthyroid patients, fractional T4 transfer rates from serum to rapidly equilibrating tissues are highest in those with the highest free fractions of T42. Both T4 binding to serum carrier proteins and T4 fractional transfer rates from serum to rapidly equilibrating tissues may be reduced by circulating inhibitors in nonthyroidal disorders (3).

The effects of reduced serum T4 binding to serum carrier proteins on fractional transfer rates of T4 from serum to rapidly equilibrating tissues may be accounted for by determining the ratio of these two parameters. The ratio of fractional T4 transfer rates from serum to rapidly equilibrating tissues and free fractions of T4 in sera is directly correlated with T3 production rates in nonthyroidal disorders and altered binding states (Figure 5).

View figure 5

In other words, critically ill patients have the highest free fractions of T4, the lowest T4 fractional transfer rates, and the lowest T3 production rates, which may reflect the highest concentration of circulating inhibitors (3). These findings are consistent with in vitro findings that sick sera may impair T4 uptake into liver, which may be rate limiting and reduce T3 production from T4 in nonthyroidal disorders (3).

Caution must be used in interpreting T3 production and degradation rates determined from serum kinetic studies in vivo (12). Since T3 is primarily produced and degraded in extrathyroidal tissues, T3 can be produced from T4 and degraded without entering the circulation. Thus, total body production and degradation rates of T3 may be underestimated when determined from serum tracer studies in normal subjects as well as in patients with nonthyroidal illnesses or with caloric deprivation.

Elevated total and free reverse T3 concentrations in patients with nonthyroidal illnesses and with caloric deprivation are primarily due to impaired reverse T3 clearance rates, since reverse T3 production rates from T4 appear to be normal (1-3). Decreased reverse T3 serum clearance rates in nonthyroidal illnesses appear to reflect the net effect of impaired reverse T3 fractional transfer rates from serum to tissue sites of degradation such as liver, and decreased reverse T3 binding by serum carrier proteins (1-3). Normal reverse T3 production rates from T4 in nonthyroidal disorders, suggests that free T4 availability to tissues producing reverse T3 is normal (1-3). However, as with T3 , reverse T3 production and degradation primarily occurs in tissues and, thus, total body reverse T3 production and degradation rates may be underestimated by serum tracer kinetic studies in vivo.

If T4 entry into tissues is rate limiting for reverse T3 production, as it may be for T3 production, normal reverse T3 production rates in nonthyroidal disorders suggest normal T4 entry into reverse T3 producing tissues, which therefore, may differ from T3 producing tissues, and may be extrahepatic1–3. Normal reverse T3 concentrations in patients with acute or chronic renal failure may relate to normal clearance and production rates, while reduced reverse T3 levels in patients with nephrotic syndrome appear to be secondary to impaired reverse T3 production rates (2).

In patients dying of severe nonthyroidal illnesses, tissue levels of total T3 were normal in heart and skeletal muscle but reduced in cerebral cortex, hypothalamus, pituitary, liver, kidney, and lung, while total T4 levels were normal in cerebral cortex but reduced in liver, compared to tissues of trauma victims (13). However, total tissue T3 and T4 content may not reflect thyroid hormone availability to nuclear sites of action any more than total serum thyroid hormone levels reflect free hormone available for uptake into tissues. Tissue effects of T3 are mediated by T3 nuclear receptor proteins, which appear to be normal in liver tissues from patients with chronic liver diseases, as are the expression of thyroid hormone-regulated mRNAs encoding T4-binding globulin, sex hormone-binding globulin, cortisol-binding globulin, and transthyretin (14). These findings indicate that free T4 and T3 availability to tissues may be normal and that patients with nonthyroidal illnesses may indeed be euthyroid.

Serum TSH levels may transiently decrease during the first few days of critical illness (8). Serum TSH bioactivity was increased in 64% of critically ill euthyroid patients (7), and may relate to transiently decreased TSH sialylation (8). TSH bioactivity correlated inversely with total T4, total T3, and Gammacoat free T4 (Travenol, USA) but not with serum TSH levels, suggesting that more severely ill patients with lower total T4 and total T3 values, may have more bioactive TSH7. Interestingly, free T4 estimates by direct equilibrium dialysis are typically normal or elevated in critically ill patients, and have a direct, rather than an inverse relationship, with serum TSH concentrations (15). These latter findings could also be interpreted to indicate a state of T4 resistance in severe nonthyroidal illnesses (1,2).

In patients with chronic renal failure, 10% of serum TSH values may be above the normal range in the absence of hypothyroidism (16). TSH carbohydrate branching and sialylation may be altered in these patients, while bioactivity may be normal (16). As patients begin to recover from acute nonthyroidal illnesses, serum TSH levels may rise, in some instances, to hypothyroid levels, as total T4, free T4, and total T3 levels return to normal (2,3). These findings may reflect normalization of T4 binding to serum carrier proteins resulting in transient decreases in ambient free T4 levels, increased T4 transport into tissues for T3 deiodination, perhaps transiently increased sialylation and reduced bioactivity of TSH, as well as decreased hypothalamic-pituitary suppression with caloric replacement and decreased glucocorticoid responses to stress (1-3). Additional alterations of TSH in nonthyroidal illnesses include blunted TSH responses to TRH, and altered diurnal rhythms (1-3,16). The clinical significance of these findings remains to be elucidated.

Clinical questions requiring further investigation

  1. Which, if any, free T4 and free T3 methods reflect circulating hormone levels in nonthyroidal disorders without concurrent hypothyroidism or hyperthyroidism?

    2.  
  2. Is the "set-point" for the hypothalamic-pituitary-thyroid axis altered in nonthyroidal illnesses? Do available in vitro assays for TSH and free T4 allow us to assess this axis?

    3.  
  3. How can changes in serum thyroid hormone levels in sick patients without hypothalamic, pituitary or thyroid disease be distinguished from those with concurrent hypothyroidism or hyperthyroidism?

    4.  
  4. Should thyroid hormone therapy be given to sick patients without hypothyroidism?
 

Free T4 and free T3 estimates in nonthyroidal disorders

Free T4 and free T3 assay methods typically provide appropriate values in sera from otherwise well patients with hypothyroidism or hyperthyroidism, and in sera of euthyroid subjects with moderate alterations of serum binding protein concentrations (1-3). However, free T4 and free T3 estimates are highly method dependent in sera from "euthyroid" patients with nonthyroidal illnesses who by definition do not have concurrent hypothalamic, pituitary or thyroid diseases (Figures 3 and 4) (5,9,17).

Problems with free T4 and free T3 assay methods are multiple and may include:

1) serum dilution, which may weaken effects of low affinity binding inhibitors, 2) sequestration of hormone, 3) artefactual interactions between assay components such as tracers and binding proteins, and 4) alterations of protein binding properties by temperature variations or by buffer pH and composition (9,15,17-19). In addition, in equilibrium dialysis methods, dialyzable contaminating substances in tracer, such as free iodide, can be mistaken for free hormone (19,20), and dialyzable inhibitors of hormone binding to serum carrier proteins may be diluted in dialysate (18). In both equilibrium dialysis and ultrafiltration methods, movement of small quantities of proteins across membranes may increase free hormone estimates (20). Ultrafiltration methods using undiluted serum should avoid in vitro artifacts induced by dilution of serum and by dialyzable inhibitors in serum (18,19). Use of radioimmunoassays which measure free hormone concentrations directly in ultrafiltrate (20), or chromatography to identify the hormone (19) should avoid problems of tracer impurities being mistaken for free hormone (19).

Free T4 estimates

Severely ill "euthyroid" patients with low total serum T4 concentrations, usually have normal or high free T4 estimates by direct equilibrium dialysis, Amerlite MAB, and SPAC-ET methods while free T4 values are frequently reduced by other methods (Figure 3) (3-5). Tracer equilibrium dialysis underestimates free T4 values in severely ill patients with reduced serum hormone binding (4,5) (Figure 1). By tracer ultrafiltration methods, free T4 values were decreased (21), normal (22) or elevated (19,23) in sick sera in which total T4 concentrations were reduced to 41% to 73% of normal mean values, presumably due to methodological variations.

Despite normal or elevated free T4 levels by direct equilibrium dialysis in the majority of patients with nonthyroidal illnesses (4,5) (Figure 1), 31% of 175 patients with chronic renal failure had reduced free T4 levels by direct equilibrium dialysis (24). In addition, 80% of these patients also had reduced free T4 values by the Amerlex-MAB method (24). For both methods, serum free T4 correlated inversely with serum creatinine (24), suggesting a relationship to severity of illness. In another study, free T4 estimates by tracer ultrafiltration were within the normal range in 100% of critically ill and 100% of hemodialysis patients although mean values were reduced (21), while by tracer equilibrium dialysis 11% of hemodialysis patients had elevated values and 5.6% had reduced values (21). This variation is not readily explained.

Dialyzable inhibitor(s) of T4 binding in uremic serum could explain inappropriately low free T4 levels by direct equilibrium dialysis in chronic renal failure patients (Personal communication Dr. JC Nelson, Loma Linda University, California) (18) (Table 1). As depicted in Figure 6, a dialyzable inhibitor in uremic serum would cross the dialysis membrane and be diluted in dialysate, normalizing T4 binding to serum carrier proteins, thereby, reducing free T4 levels in serum and dialysate. Indeed, serum concentrations of 3- carboxy-4-methyl-5-propyl-2-furan propanoic acid (CMPF), indoxyl sulfate and hippuric acid are markedly elevated in uremic serum, and indoxyl sulfate and hippuric acid are dialyzable in vitro (25).

View figure 6

Further, salicylates and salsalate are dialyzable, and salicylate ingestion results in underestimation of serum free T4 values by direct equilibrium dialysis compared to a tracer ultrafiltration method using minimally diluted sera (18). Other weak inhibitors of T4 and T3 binding to serum carrier proteins such as furosemide, dilantin and some nonsteroidal agents (Table 1) are diluted by some methods and consequently may also cause underestimation of free hormone values. In contrast, heparin therapy increases the free fraction of T4 when measured by equilibrium dialysis more than by ultrafiltration, due to liberation of free fatty acids during the in vitro dialysis procedure, which then displace T4 from binding proteins (18,20).

Circulating free T4 concentrations would appear to be normal or elevated in the majority of patients with nonthyroidal illnesses, with exceptions being patients receiving pressor doses of dopamine, and those with concurrent hypothalamic, pituitary or thyroid disease (1-5). Since free T4 is the primary substrate for reverse T3 production, normal reverse T3 production rates in critically ill patients with reduced total T4 concentrations indicate normal free T4 availability in vivo (1-3). Thus, free T4 methods which provide normal or increased free T4 estimates in sera from sick patients without concurrent hypothyroidism would appear to most accurately reflect the in vivo state.

Free T3 estimates

Free T3 values were variably reduced in sick sera by the seven methods used by Sapin et al (9) (Figure 4). Free T3 estimates were lowest in patients with the most reduced total T4 and total T3 concentrations by all methods (9), presumably reflecting the greater severity of illness. Free T3 values determined by tracer or direct equilibrium dialysis, tracer ultrafiltration, and direct Sephadex LH- 20 method in sick sera did not correlate with total T3 concentrations, and were reduced in all studies (1,9,19,21,23,26) except three (22,27,28).

Faber et al. (22,27) used minimally diluted sera but did not demonstrate that radioactivity in ultrafiltrates was T3 rather than labeled contaminants; mean free T3 values in sick sera were 104% and 92% of normal. The direct equilibrium dialysis method of Chopra et al. (26,28) required addition of unlabelled T3 to sick serum samples (exogenous to endogenous T3 concentration ratio of 4:1), to allow detectable levels of T3 to be measured in dialysate. Using this method, mean free T3 estimates varied from 46% to 96% of normal in sick sera in two different reports (26,28). Surks et al (19) avoided the most obvious problems, by using undiluted sera at pH 7.4 and 37 degrees Celsius, and identified T3 in ultrafiltrates by paper chromatography, and observed free T3 values in critically ill patients to be modestly reduced to 80% of normal values. Indeed, free T3 availability to tissue sites of action may be normal in nonthyroidal disorders as indicated by normal T3 nuclear receptor proteins and normal thyroid hormone-regulated mRNAs (14).

The hypothylamic-pituitary-thyroid Axis "set-point" in nonthyroidal illnesses

Central TSH suppression has been postulated to occur in patients with acute critical illnesses due to endogenous glucocorticoid elevations and caloric deprivation (1-3). T4 and reverse T3 production rates are within the normal range even in the low total T4 state of critical nonthyroidal illness and free T4 values are normal or elevated by direct equilibrium dialysis in the majority of patients without concurrent hypothalamic-pituitary-thyroid disease (1-5). However, elucidation of the pituitary-thyroid axis in nonthyroidal disorders requires simultaneous assessment of bioactive TSH and free T4 by a method which provides credible values in sick sera.

Diagnosing hypothyroidism or hyperthyroidism in sick patients

Patterns of changes in serum thyroid hormone and TSH levels due to nonthyroidal illnesses tend to differ from those in sick patients with concurrent hypothyroidism or hyperthyroidism (Figure 2). For example, in "euthyroid" patients with reduced TSH values due to nonthyroidal illnesses, total T4 and free T4 index values are typically normal or reduced, while in those with elevated TSH values, total T4 and free T4 index values are usually normal (3,10). In patients with hypothyroidism, the presence of thyroid autoantibodies may indicate autoimmune hypothyroidism. Serum reverse T3 concentrations do not reliably differentiate hypothyroid from "euthyroid" sick patients, and have questionable clinical utility (29). In patients with Graves' disease, thyroid stimulating immunoglobulin titers and thyroidal radioiodide uptake may be elevated, confirming the diagnosis. A radioiodide thyroid scan may differentiate between other causes of hyperthyroidism. Algorithms may be useful (6).

Should thyroid hormone therapy be given to sick patients without hypothyroidism

In some instances, short-term T3 administration to patients undergoing cardiopulmonary bypass surgery or coronary bypass procedures may have short-term beneficial cardiovascular effects (30-32). However, in the absence of hypothyroidism, no beneficial effect of T4 or T3 replacement therapy has been demonstrated in other "euthyroid" sick patients, and, therefore, can not be recommended (3,16,30-32).

Elaine M. Kaptein

Elaine M. Kaptein, M.D., is a Professor of Medicine at the University of Southern California in Los Angeles, California. She is board certified in Endocrinology and Nephrology. Her main research interest is the study of altered thyroid hormone metabolism in patients with nonthyroidal illnesses and altered nutritional states. She has authored about 100 scientific papers. She is actively involved in the American Thyroid Association, and is currently on the Program Committee.

References

1. Kaptein EM: Thyroid hormone metabolism in illness. In Hennemann G (ed): Thyroid Hormone Metabolism, Basic and Clinical Endocrinology Series, Volume 8, New York, Marcel-Dekker Inc, 1986, p 297-333.

2. Kaptein EM. The effects of systemic illness on thyroid hormone metabolism. In: Wu SY, ed. Thyroid hormone metabolism. Oxford: Blackwell 1991; 211-237.

3. Docter R, Krenning EP, de Jong M, Hennemann G. The sick euthyroid syndrome: Changes in thyroid hormone serum parameters and hormone metabolism. Clinical Endocrinology 1993; 39: 499-518.

4. Kaptein EM. Clinical application of free thyroxine determinations. Clin Lab Med 1993; 13: 653-672.

5. Kaptein EM. Thyroid in vitro testing in non-thyroidal illness. Exp Clin Endocrinol 1994; 102: 92-101.

6. Kaptein EM. Clinical relevance of thyroid hormone alterations in nonthyroidal illnesses. Thyroid International 1977; 4: 22-25.

7. Kakezono F, Yamashita S, Yokoyama N, Morita S, Kubo I, Ohtakara S, Okamoto S, Morimoto I, Ikari N, Izumi M, Nagataki S. Stimulation of thyroid adenylate cyclase activity by sera from patients with non-thyroidal illness. Acta Endocrinol 1986; 113: 340-345.

8. Magner J, Roy P, Fainter L, Barnard V, Fletcher P. Transiently decreased sialylation of thyrotropin (TSH) in a patient with euthyroid sick syndrome. Thyroid 1997; 7: 55-61.

9. Sapin R, Schlienger J-L, Kaltenbach G, Gasser F, Christofides N, Roul G, Gervais A, Petitjean P, Chambon J. Determination of free triiodothyronine by six different methods in patients with non-thyroidal illness and in patients treated with amiodarone. Ann Clin Biochem 1995; 32: 314-324.

10. Spencer CA, Eigen A, Shen D, et al: Specificity of sensitive assays of thyrotropin (TSH) used to screen for thyroid disease in hospitalized patients. Clin Chem 1987; 33: 1391-1396.

11. Franklyn JA, Black EG, Betteridge J, Sheppard MC. Comparison of second and third generation methods for measurement of serum thyrotropin in patients with overt hyperthyroidism, patients receiving thyroxine therapy, and those with nonthyroidal illness. J Clin Endocrinol Metab 1994; 78: 1368-1371.

12. DiStefano JJ, Jang M, Malone TK, Broutman M. Comprehensive kinetics of triiodothyronine production and metabolism in blood and tissue pools of the rat using optimized blood-sampling protocols. Endocrinology 1982; 110: 198-213.

13. Arem R, Wiener GJ, Kaplan SG, Kim H-S, Reichlin S, Kaplan MM. Reduced tissue thyroid hormone levels in fatal illness. Metabolism 1993; 42: 1102- 1108.

14. Chamba A, Neuberger J, Strain A, Hopkins J, Sheppard MC, Franklyn JA. Expression and function of thyroid hormone receptor variants in normal and chronically diseased human liver. J Clin Endocrinol Metab 1996; 81: 360-367.

15. Wilcox RB, Nelson J, Tomei RT. Heterogeneity in affinities of serum proteins for thyroxine among patients with non-thyroidal illness as indicated by the serum free thyroxine response to serum dilution. Eur J Endocrinol 1994; 131: 9-13.

16. Kaptein EM. Thyroid hormone metabolism and thyroid diseases in chronic renal failure. Endocrine Reviews 1996; 17: 45-63.

17. Nelson JC, Wilcox RB. Analytical performance of free and total thyroxine assays. Clin Chem 1996; 42: 146-154.

18. Faber J, Waetjen I, Siersbaek-Nielsen K. Free thyroxine measured in undiluted serum by dialysis and ultrafiltration: Effects of non-thyroidal illness, and an acute load of salicylate or heparin. Clin Chim Acta 1993; 223: 159-167.

19. Surks MI, Hupart KH, Pan C, et al: Normal free thyroxine in critical nonthyroidal illnesses measured by ultrafiltration of undiluted serum and equilibrium dialysis. J Clin Endocrinol Metab 1988; 67: 1031-1039.

20. Tikanoja SH, Liewendahl BK. New ultrafiltration method for free thyroxin compared with equilibrium dialysis in patients with thyroid dysfunction and nonthyroidal illness. Clin Chem 1990; 36: 800-804.

21. Wang Y-S, Hershman JM, Pekary AE. Improved ultrafiltration method for simultaneous measurement of free thyroxin and free triiodothyronine in serum. Clin Chem 1985;31:517-522.

22. Faber J, Kirkegaard C, Rasmussen B, Westh H, Busch-Sorensen M, Jensen IW. Pituitary-thyroid axis in critical illness. J Clin Endocrinol Metab 1987; 65: 315-320.

23. Lee LA, Mooney RA, Woolf PD. Clinical utility of measuring free thyroxin and free triiodothyronine in serum of critically ill patients with ultrafiltration. Clin Chem 1986; 32: 797-800.

24. Okabayashi T, Takeda K, Kawada M, Kubo Y, Nakamura S, Chikamori K, Terao N, Hashimoto K. Free thyroxine concentrations in serum measured by equilibrium dialysis in chronic renal failure. Clin Chem 1996; 42: 1616-1620.

25. Lim C-F, Bernard BF, de Jong M, Docter R, Krenning EP, Hennemann G.A furan fatty acid and indoxyl sulfate are the putative inhibitors of thyroxine hepatocyte transport in uremia. J Clin Endocrinol Metab 1993; 76: 318-324.

26. Chopra IJ, Taing R, Mikus L. Direct determination of free triiodothyronine (T3) in undiluted serum by equilibrium dialysis/radioimmunoassay (RIA). Thyroid 1996; 6: 255-259.

27. Faber J, Siersbaek-Nielsen. Serum free 3,5,3'-triiodothyronine (T3) in non-thyroidal somatic illness, as measured by ultrafiltration and immunoextraction. Clin Chim Acta 1996; 256: 115-123.

28. Chopra IJ, Trong UT. Simultaneous measurement of free thyroxine (free T4 ) and free 3,5,3'-triiodothyronine (T3 ) in undiluted serum by direct equilibrium dialysis/radioimmunoassay (RIA): Evidence that free T3 (and free) T4 are normal in majority of patients with the low T3 syndrome. Thyroid 1997; 7: S- 123.

29. Burmeister LA. Reverse T3 does not reliably differentiate hypothyroid sick syndrome from euthyroid sick syndrome. Thyroid 1995; 5: 435-441.

30. Wartofsky L. The low T3 or "sick euthyroid syndrome". Update 1994. Endocrine Reviews 1994; 3: 248-251.

31. McIver B, Gorman CA. Euthyroid sick syndrome: An overview. Thyroid 1997; 7: 125-132.

32. Chopra IJ. Euthyroid sick syndrome: Is it a misnomer? J Clin Endocrinol Metab 1997; 82: 329-334.