Nondiabetic Ketoacidosis Caused by Severe Hyperthyroidism

Emily T. Wood; William B. Kinlaw


Thyroid. 2004;14(8) 

In This Article


This case did not fit the traditional differential diagnosis of ketoacidosis. Diabetic ketoacidosis, alcoholic ketoacidosis,uremia, lactic acidosis and toxic ingestion were not present. Starvation ketosis caused by vomiting was a possible contributor, however, the degree of our patient's ketosis is similar to that seen in prolonged complete fasting. β-OHB levels climb to 2 mmol/L in the first days of complete fasting, and reach levels of 5–6 mmol/L with prolonged (>3 week)starvation.[4,5] Although our patient had decreased oral intake, she was not completely fasting, and thus her β-OHB level of 5.2 mmol/L cannot be explained by starvation alone. Previous studies have shown that hyperthyroid patients develop higher degrees of ketosis when fasting than do euthyroid patients.[6,7] Therefore, the development of ketosisin our patient was probably magnified by decreased nutrition.

Genetic disorders of ketone metabolism such as succinyl-CoA:3-oxoacid CoA transferase deficiency or β-thioketolase deficiency were not present as these are usually detected in childhood. We have not excluded a heterozygous form of aketolytic enzyme defect, however, the concomitant resolution of ketoacidosis and hyperthyroidism demonstrates that excess thyroid hormone (TH) was a major contributor to the ketoacidosis.

A major component of ketoacidosis is the presentation of excess fatty acids to the liver. In adipocytes, β- and α2-adrenergic receptors are coupled by Gs and Gi proteins to adenylate cyclase. Adenylate cyclase produces cyclic adenosine monophosphate (cAMP), regulating hormone sensitive lipase (HSL) and thus the release of fatty acids from adipocytes. β-adrenergic stimulation activates adenylate cyclase thereby promoting lipolysis while α2-adrenergic stimulation inhibits adenylate cyclase thereby inhibiting it. An important player in the lipolytic pathway is the enzyme phosphodiesterase (PDE) which degrades cAMP and is therefore antilipolytic.

Thyroid hormone, TH effects several of these mechanisms to increase lipolysis. Disabling the TH receptor decreases adrenergically mediated lipolysis in mouse adipocytes, illustrating the overall effect of TH on adipocytes.[8] Adipocytes from hyperthyroid patients have increased numbers of α2-adrenergic receptors and increased lipolytic response to α-agonists.[9] A recent study using in vivo microdialysis techniques showed that hyperthyroid patients have elevated norepinephrine (NE) and glycerol levels in subcutaneous adipose tissue compared to controls, suggesting that thyroid hormone may increase lipolysis by increasing the local release of NE.[10] Triiodothyronine (T3) has been shown to downregulate α2receptors in adipocytes.[11] TH also increases cAMP by downregulating G-protein i2 in human adipocytes.[11] Not only do hyperthyroid adipocytes produce more cAMP, but the metabolism of cAMP in hyperthyroid adipocytes is decreased because of low PDE activity. Adipocytes from hyperthyroid patients have lower PDE activity than controls,[12] and TH has been shown to downregulate adipocyte PDE mRNA.[11]

In the liver, long-chain fatty acids are transported by the carnitine shuttle into the mitochondria for β-oxidation. TH may also act on this pathway to increase ketogenesis. Hyperthyroidism is associated with increased total carnitine in rat liver,[13] and urinary carnitine excretion is elevated in hyperthyroid patients.[14] Carnitine palmitoyltransferase-I(CPT-1) facilitates the carnitine shuttle, and its expression is upregulated by T3in hepatocytes.[15]

Interestingly, our patient's acidosis worsened after the initiation of treatment for hyperthyroidism. In light of the above mechanisms, why wasn't the acidosis most severe when plasma T3 was at its peak? The level of ketones in the plasma depends on both ketone production and clearance. It takes some 24 hours to turn off gluconeogenesis and ketogenesis,thus ketone production continues after the initiation of treatment. Ketone clearance is linear with ketone concentration,so as ketone levels fall, clearance also falls. There is also evidence that hyperthyroidism increases clearance of ketones.[16] We speculate that with the initiation of thyroxine-lowering treatment, ketone clearance decreased more rapidly than production, allowing ketones to transiently accumulate in the plasma.

In conclusion, TH promotes lipolysis in adipocytes by several mechanisms and it may increase ketogenesis in the liver as well. Thus hyperthyroidism may promote ketoacidosis in the face of normal insulin action. Why is this so rarely described in hyperthyroid patients? In moderate hyperthyroidism, the ketogenic effects of thyroid hormone may be masked by basal insulin secretion.[17] We propose that severe hyperthyroidism is required for ketoacidosis to occur, and reduced carbohydrate intake, and to a lesser degree protein intake, may potentiate it. Mechanistic considerations aside, severe hyperthyroidism should be considered in the differential diagnosis of ketoacidosis.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.