Persistent Nephrogenic Diabetes Insipidus Following Lithium Therapy

Title: Persistent Nephrogenic Diabetes Insipidus Following Lithium Therapy
Authors: Thompson, Christopher J.; France, A. J.; Baylis, P. H.
Publisher: Scottish Medical Journal
Date Published: February 01, 1997
Reference Number: 173
We report the case of a patient who developed severe hypernatraemic dehydration following a head injury. Ten years previously he had been diagnosed to have lithium-induced nephrogenic diabetes insipidus, and lithium therapy had been discontinued. He remained thirsty and polyuric despite cessation of lithium and investigations on admission showed him to have normal osmoregulated thirst and vasopressin secretion, with clear evidence of nephrogenic diabetes insipidus. Lithium-induced nephrogenic diabetes insipidus is considered to be reversible on cessation of therapy but polyuria persisted in this patient for ten years after lithium was stopped. We discuss the possible renal mechanisms and the implications for management of patients with lithium-induced nephrogenic diabetes insipidus.
The following article is presented with the permission of the publisher for educational use within the NDI community. No part of this article may be reproduced in any way without permission in writing from the publisher.


Nephrogenic diabetes insipidus is a common complication of therapy with lithium carbonate; 12% of patients receiving the drug may develop polyuria,1 and 80% show an asymptomatic reduction in maximum urinary concentration.2 Although remission of polyuria almost always occurs within weeks of cessation of lithium therapy,3 recovery may be slow, and persistent symptoms have been reported up to twenty months after withdrawal of lithium.4 We describe a case of nephrogenic diabetes insipidus which persisted for ten years after discontinuation of lithium carbonate therapy.

Case Report

A 56 year old unemployed butcher was admitted with a two week history of confusion, unsteady gait and worsening polyuria. Two weeks prior to admission he had fallen and sustained a head injury without loss of consciousness. He had a 42 year history of bipolar manic depressive psychosis and had been treated with lithium carbonate between 1968 and 1979. Lithium dosage varied between 1000-1500 mg daily, and serum lithium concentrations were maintained on this dosage within the normal laboratory reference range. He had 14 hospital admissions between 1949 and 1989 for acute hypomania. In 1979 he developed severe polyuria and was diagnosed to have nephrogenic diabetes insipidus after a water deprivation test. Lithium treatment was discontinued but despite persistent polyuria and polydipsia he was discharged from follow-up by the physicians, although he continued to receive psychiatric care with chlorpromazine therapy.

At the time of admission he was receiving chlorpromazine 25 mg tid, frusemide 40 mg daily and potassium chloride 1200 mg daily. He was grossly dehydrated, pulse was 102 beats per minute and blood pressure (mmHg) was 124/74 supine, 92/60 standing. He was fully orientated but had marked ataxia. Fundoscopy was normal.

Investigations showed plasma sodium 181 mmol/l, potassium 3.7 mmol/l, urea 17.6 mmol/l, creatinine 132 umol/l, calcium 2.13 mmol/l, albumin 31 g/l, blood glucose 4.6 mmol/l, plasma osmolality 397 mOsm/kg and urine osmolality 290 mOsm/kg. Twenty four hour urine output was 8,200 ml; there was no proteinuria and no growth on urine culture. Skull radiograph was normal, and computerised tomography of the brain showed cerebral atrophy but no focal lesion. Anterior pituitary function was normal.

A presumptive diagnosis of persistent lithium-induced nephrogenic diabetes insipidus was made, and after rehydration with intravenous fluids, plasma sodium, plasma osmolality and blood urea returned to normal, though severe polyuria persisted. When plasma osmolality was normalised, the patient underwent intravenous infusion of hypertonic (855 mmol/l) sodium chloride at a rate of 0.05 ml/kg/min for two hours. Plasma vasopression concentration rose during this test (figure 1), excluding cranial diabetes insipidus, and the rise in plasma osmolality generated appropriate thirst as measured on a linear analogue scale (figure 1), and normal water intake. Urine osmolality at the end of the infusion was inappropriately low at 182 mOsm/kg (normal response>700 mOsm/kg). Nephrogenic diabetes insipidus was confirmed by demonstration of failure of concentration of urine in response to intramuscular desmopressin at either 2 ug low dose (urine osmolality 205 to 181 mOsm/kg), or 20 ug high dose (urine osmolality 164 to 218 mOsm/kg).

The patient was discharged on chlorpromazine 25 mg tid and advice on fluid intake but as polyuria persisted (24 hour urine volume 6,800 ml) he was commenced on indomethacin 25 mg tid with a fall in urine output (24 hour urine volume 3,800 ml) and subjective symptomatic improvement.

In June, 1989 he was admitted to a medical ward with a cal vein thrombosis which was treated with warfarin. After a fall sustaining a blow to his right leg he developed a painful haematoma with subsequent immobilisation. His recovery was halted by the development of bilateral bronchopneumonia complicated by acute renal failure which led to his death in July, 1989. He remained markedly polyuric until his death, ten years after ceasing lithium therapy.


Our patient presented with severe hypernatraemic dehydration due to nephrogenic diabetes insipidus, with normal osmoregulated thirst and vasopressin release. The temporal relationship of his nephrogenic diabetes insipidus to his previous lithium therapy and absence of other metabolic or renal causes of nephrogenic diabetes insipidus strongly suggest a persistent lithium-induced aetiology. Because of the nature of his presentation following a head injury, it was necessary to exclude cranial diabetes insipidus as a cause for his polyuria, and the results of the hypertonic saline infusion confirmed that vaspressin secretion was entirely normal.

Fig. 1  Responses of plasma vasopressin (pAVP, left panel) and thirst (right panel) to hypertonic saline infusion . The shaded area represents the normal range in 40 healthy individuals. LD = limit of detection of the assay (0.3 pmol/l).

It is unusual for patients with either cranial or nephrogenic diabetes insipidus to develop hypernatraemia, as the thirst mechanism is intact, and fluid intake is therefore appropriate to urinary losses. The degree of hypernatraemia in this patient therefore raised the possibility of hypodipsia in conjunction with nephrogenic diabetes insipidus, an association which has previously been reported in a patient treated with lithium.5 Thirst ratings during the hypertonic saline infusion were, however, entirely normal and on recovery from the head injury fluid intake remained appropriate to his urine output. We would therefore suggest that the severe hypernatraemia resulted from inadequate fluid intake during a period of post-head injury confusion, with continuing polyuria due to nephrogenic diabetes insipidus.

Lithium induced nephrogenic diabetes insipidus is usually reversible within weeks of cessation of therapy,3 though persistent polyuria twenty months after stopping the drug has been reported.4,6 It is now well established that lithium can produce chronic renal lesions in both man7 and experimental animal models;8 the animal data shows that lithium induces a variety of histological abnormalities including interstitial fibrosis, tubular atrophy and glomerular sclerosis.10 We propose that lithium therapy induced irreversible renal tubular damage in this man, leading to permanent nephrogenic diabetes insipidus.

Although the long duration of diabetes insipidus following cessation of lithium therapy in this patient is highly unusual, the implications of the case are extremely important, as the failure to maintain fluid intake following head injury led to the development of severe hypernatraemic dehydration in this man and his continued polyuria during his fatal bronchopneumonia led to acute renal failure. Lithium is widely prescribed for the treatment of mania and manic-depressive psychosis and nephogenic diabetes insipidus is undoubtedly a common side effect of the drug.2 As lithium can induce chronic renal lesions7,8 and led to prolonged diabetes insipidus in our patient, we suggest that those patients who develop polyuria during treatment with the drug should continue to have their fluid balance monitored not only during treatment with lithium, but also after cessation of therapy, until urine output returns to normal. This is particularly important as recent evidence suggests that persistent polyuria may respond to treatment with indomethacin, which can reduce urine output in patients with lithium-induced nephrogenic diabetes insipidus.10 This case is a reminder that the renal complications of lithium therapy do not always reverse on cessation of therapy, and emphasises the necessity of continued monitoring of patients who develop polyuria.


  1. Forrest JN, Cohen AD, Torretti J et al. On the mechanism of lithium-induced diabetes insipidus in man and the rat. Journal of Clinical Investigation 1974; 53:1115-23.
  2. Waller DG, Edwards JG, Papasthatis-Papayanni S. A longitudinal assessment of renal function during treatment with lithium. Quarterly Journal of Medicine 1988; 255:553-58.
  3. Baldessarini RJ, Lipinski JF. Lithium salts. Ann Int Med 1975; 83:527-33
  4. Simon NM, Garber E, Arieff AJ. Persistent nephrogenic diabetes insipidus after lithium carbonate. Ann Int Med 1977; 86:446-47.
  5. Thompson CJ, Freeman J, Record CO et al. Hypernatraemia due to a reset osmostat for vaspressin release and thirst, complicated by nephrogenic diabetes insipidus. Postgrad Med J 1987; 63:979-82.
  6. Neithercut WD, Spooner RJ, Hendry A, Dagg JH. Persistent nephrogenic diabetes insipidus, tubular proteinuria, aminoaciduia [sic] and parathyroid hormone resistance following long term lithium administration. Postgrad Med J 1990; 66:479-82.
  7. Hestbech J, Hansen HE, Amdisen A et al. Chronic renal lesions following long-term treatment with lithium. Kidney International 1977; 12:205-13.
  8. Walker RG, Escott M, Birchall I et al. Chronic progressive renal lesions induced by lithium. Kidney International 1986, 29:875-81.
  9. Allen HM, Jackson RL, Winchester MD et al. Indomethacin in the treatment of lithium induced nephrogenic diabetes insipidus. Archives of Internal Medicine 1989; 149 1123-26.

    Department of Medicine, Ninewells Hospital and Medical School, Dundee DD1 9SY, Scotland, U.K. (C.J. Thompson, A.J. France), Department of Medicine, Royal Victoria Infirmary Newcastle-upon-Tyne (P.H. Baylis)

    1997, February

    Address correspondence and reprint requests to Dr. Ian H. McKee, Hermiston Publications Ltd, 9 Stonelaws, Whitekirk, East Lothian EH40 3DX.

This translation by the NDI Foundation is to assist the lay reader. To provide a clear, accessible interpretation of the original article, we eliminated or simplified some technical detail and complicated scientific language. We concentrated our translation on those aspects of the article dealing directly with NDI. The NDI Foundation thanks the researchers for their work toward understanding and more effectively treating this disorder.
© Copyright NDI Foundation 2007 (JC)

Thompson, et al., describe a 56-year-old male patient who was still manifesting lithium-induced nephrogenic diabetes insipidus ten years after being taken off lithium therapy. NDI is a common complication of lithium therapy, and although symptoms of NDI generally disappear within weeks of ending lithium use, recovery can be slow, up to 20 months in some cases. And in the present case, much longer than that.

The patient was admitted into the hospital because of complications of a head injury. Since he exhibited extensive polyuria (the chronic passage of large amounts of urine) and had high levels of blood sodium, he underwent tests to find out why. No other cause for his NDI was found: his secretion of the antidiuretic hormone, vasopressin (VP) was normal; his thirst sensors were functional. And it was determined that his excessive blood sodium was due to the patient being too disoriented after his head injury to drink adequately.

It is well-established that lithium can produce chronic kidney lesions in animals and humans. The authors propose that lithium therapy induced irreversible damage to the many vital little tubes in their patient's kidneys, leading to permanent NDI. The authors advise that those patients who develop polyuria during lithium treatment should be monitored regularly until their urine output returns to normal. The monitoring is important especially in the light that the drug indomethacin can reduce urine output in patients with NDI. If this happens, then the patient might continue with lithium even though unbeknownst to him or his doctor, he may be incurring lesions on his kidney.