Although incomplete, the link between thyroid function, bipolar affective disorder, and lithium has been acknowledged for many years. This article provides an overview of the relationship and recent literature.
The exact therapeutic mechanism of lithium in bipolar disorder (BD) remains unclear, due in part to its many actions on the cells of the central nervous system. The same complexity applies to the interactions between lithium and the thyroid. Lithium is widely known to impair the function of the thyroid through a variety of mechanisms.
In 1998, Lazarus listed impaired thyroidal uptake of iodine, impaired iodination of tyrosine, altered thyroglobulin structure and impaired release of thyroxine from the thyroid gland as the mechanisms by which lithium induces hypothyroidism. Impaired release of thyroxine is considered the most clinically significant, and this phenomenon has been used to enhance the effectiveness of radioactive iodine when treating thyrotoxicosis (Bogazzi et al., 1999). Lithium has also been used, with reported success, to treat refractory hyperthyroidism (Dickstein et al., 1997).
Elevation of thyrotropin in response to lowered circulating thyroxine is probably the main stimulant for goiter formation, which has been reported with incidences of 3% to 60% in lithium-treated patients; much of this variation is related to varying definitions of goiter and differences in the methods used to measure thyroid size (Lazarus, 1998).
Thyroxine is secreted as T4 and then metabolized to its active form, T3, by the enzyme 5'deiodinase. Lithium appears to impair the process of deiodination of T4 peripherally (deiodinase I) and within some cells (deiodinase III) (Terao et al., 1995). Eravci et al. (2000) found varying effects of lithium on different isoenzymes of deiodinase and noted that it appeared to enhance the activity of deiodinase II present in rat frontal lobes. This effect may contribute to alterations in cellular responsiveness to thyroxine.
It is widely known that patients receiving lithium therapy commonly develop hypothyroidism. Johnston and Eagles (1999) found hypothyroidism in 10.4% of 718 patients treated with lithium. Women suffered a higher risk of developing hypothyroidism within two years of therapy, compared to the general population of the geographical area studied. Hypothyroidism occurred in 14% of women, significantly more frequently than in men (4.5%), however, both sexes had a significantly increased risk. Women who were 40 to 59 years of age had the highest risk, with an incidence of 20%.
An even larger number of patients appear to develop subclinical hypothyroidism. Deodhar et al. (1999) reviewed 132 outpatients receiving lithium therapy and found 39% had an elevated thyroid stimulating hormone (TSH), low T4 and normal T3 levels. This fits well with older series documenting elevated TSH and goiter without overt hypothyroidism in similar numbers of patients taking lithium. It remains to be elucidated why some individuals and not others develop clinically significant hypothyroidism, but it seems likely that some other predisposing factor is required.
For instance, in the Deodhar et al. study (1999), a large number of cases were from an area where dietary deficiency of iodine is common; and Kusalic and Engelsmann (1999), prospectively studying a group of patients with BD treated with lithium, found that cases with a family history of thyroid disease had an earlier onset of hypothyroidism than those who did not.
The association between hyperthyroidism and lithium is less widely recognized than that of hypothyroidism. However, Barclay et al. (1994) documented a statistically significant increase in the incidence of hyperthyroidism in lithium-treated patients. The mechanism of this association remains unclear, especially in view of the many ways lithium impairs thyroid function. Enhanced immune stimulation is an unsatisfactory explanation since the bulk of cases were due to toxic multi-nodular goiter, not Graves' disease or other immune conditions (Barclay et al., 1994). Nonetheless, this relationship appears to be both real and relevant.
Responsiveness to Thyroxine
An increasing body of research supports the hypothesis that lithium alters cellular responsiveness to thyroxine in addition to inducing significant changes in the function of the thyroid gland. The induction of cellular unresponsiveness to thyroxine may account for the apparent efficacy of lithium in treating thyrotoxicosis. My colleagues and I reported two cases in which cessation of lithium appeared to precipitate a thyroid crisis, presumably because the presence of lithium prevented clinical manifestations of thyroid excess (Oakley et al., 2000).
Bolaris et al. (1995) documented altered binding of T3 in the CNS of rats, implying that a state of cellular hypothyroidism developed. Hahn et al. (1999b) studied the effect of lithium on gene expression in response to T3 in different cells. In some cell lines, lithium reduced the transcription of mRNA in response to T3, while other cell lines were unaffected. This effect changed with duration of therapy and was deemed to be time-dependent and cell-line specific. Thus, the effect of lithium to alter cellular responsiveness to thyroxine is not uniform for all cells and may change with duration of lithium therapy.
The same group studied the effect of lithium on the expression of different subtypes of thyroid hormone receptors in rat brains (Hahn et al., 1999a). These studies showed that transcription of messenger RNA induced by thyroxine was actually enhanced by the presence of lithium in some cells and reduced in others, depending on the receptor subtype. As different receptor subtypes are distributed differently throughout the CNS, cellular responsiveness to thyroxine is enhanced by the presence of lithium in some areas of the brain and impaired in others. In this study, gene expression was enhanced by the presence of lithium in the cortex and reduced by its presence in the hypothalamus.
Overall, the picture is far more complicated than the initial proposal of cellular hypothyroidism induced by lithium, but the effect is likely to be clinically significant. How much this contributes to the therapeutic effect of lithium in treating BD is difficult to estimate, as lithium significantly alters the metabolism and effects of many neurotransmitters.
Some alterations in thyroid function have been shown to significantly alter lithium excretion. Hyperthyroidism induces a reduction in renal lithium clearance of 10% to 15% (Owada et al., 1993). This is due to enhanced resorption of lithium by the proximal convoluted tubule and occurs despite the increases in renal blood flow and glomerular filtration rate (GFR) that occur in hyperthyroidism. Thus, the development of hyperthyroidism or overzealous use of thyroxine as augmentation for lithium may result in lithium toxicity.
Hypothyroidism has not been reported to alter lithium excretion or result in cases of lithium toxicity. It is conceivable that reduced GFR in hypothyroidism may reduce lithium excretion, but concomitant changes in tubular function may attenuate these changes. In view of the high incidence of hypothyroidism in lithium-treated patients and the lack of any reports of an interaction with lithium, it seems less likely that hypothyroidism precipitates lithium toxicity.
In a recent study of 97 patients with lithium toxicity, a significant association was found between abnormalities of the thyroid axis and the risk of chronic toxicity (Oakley et al., in press). Other factors found to be associated with chronic toxicity were age over 50 years, presence of nephrogenic diabetes insipidus and renal impairment. This study focused on the important contribution of chronic medical conditions, including endocrine disease induced by lithium, to the development of lithium poisoning and toxicity.
While lithium generally impairs thyroid function and has the potential to precipitate hypothyroidism, chronic therapy with this agent appears to increase the risk of thyrotoxicosis as well. In addition to affecting thyroid gland function, lithium alters the bioactivation of secreted thyroxine and alters the responsiveness of cells to thyroxine, enhancing it in some areas of the CNS and impairing it in others. Furthermore, alterations in the thyroid axis can substantially alter the pharmacokinetics of lithium and lead to toxicity.
Clinicians who prescribe lithium to their patients should be aware of these potential deletenous effects, at least in principle, and should regularly evaluate the thyroid health status of all of their patients receiving lithium.
Barclay ML, Brownlie BE, Turner JG, Wells JE (1994), Lithium associated thyrotoxicosis: a report of 14 cases, with statistical analysis of incidence. Clin Endocrinol (Oxf) 40(6):759-764.
Bogazzi F, Bartalena L, Brogioni S et al. (1999), Comparison of radioiodine with radioiodine plus lithium in the treatment of Graves' hyperthyroidism. J Clin Endocrinol Metab 84(2):499-503.
Bolaris S, Margarity M, Valcana T (1995), Effects of LiCl on triiodothyronine (T3) binding to nuclei from rat cerebral hemispheres. Biol Psychiatry 37(2):106-111.
Deodhar SD, Singh B, Pathak CM et al. (1999), Thyroid functions in lithium-treated psychiatric patients: a cross-sectional study. Biol Trace Elem Res 67(2):151-163.
Dickstein G, Shechner C, Adawi F et al. (1997), Lithium treatment in amiodarone-induced thyrotoxicosis. Am J Med 102(5):454-458.
Eravci M, Pinna G, Meinhold H, Baumgartner A (2000), Effects of pharmacological and nonpharmalogical treatments on thyroid hormone metabolism and concentrations in rat brain. Endocrinology 141(3):1027-1040.
Hahn CG, Pawlyk AC, Whybrow PC et al. (1999a), Lithium administration affects gene expression of thyroid hormone receptors in rat brain. Life Sci 64(20):1793-1802.
Hahn CG, Pawlyk AC, Whybrow PC, Tejani-Butt SM (1999b), Differential expression of thyroid hormone receptor isoforms by thyroid hormone and lithium in rat GH3 and B103 cells. Biol Psychiatry 45(8):1004-1012.
Johnston AM, Eagles JM (1999), Lithium-associated clinical hypothyroidism. Prevalence and risk factors. Br J Psychiatry 175:336-339 [see comments].
Kusalic M, Engelsmann F (1999), Effect of lithium maintenance therapy on thyroid and parathyroid function. J Psychiatry Neurosci 24(3):227-233.
Lazarus JH (1998), The effects of lithium therapy on thyroid and thyrotrophin-releasing hormone. Thyroid 8(10):909-913.
Oakley PW, Carter GL, Whyte IM (in press), Lithium toxicity: an iatrogenic problem in susceptible individuals. Aust N Z J Psychiatry.
Oakley PW, Dawson AH, Whyte IM (2000), Lithium: thyroid effects and altered renal handling. J Toxicol Clin Toxicol 38(3):333-337.
Owada A, Tomita K, Ujiie K et al. (1993), Decreased lithium clearance in patients with hyperthyroidism. Nephron 64(1):37-41.
Terao T, Oga T, Nozaki S et al. (1995), Possible inhibitory effect of lithium on peripheral conversion of thyroxine to triiodothyronine: a prospective study. Int Clin Psychopharmacol 10(2):103-105.