Diabetes insipidus

Diabetes insipidus is a fairly uncommon disorder, affecting approximately 3 out of 100,000 people. Although it greatly affects quality of life and is potentially life-threatening, it is quite easy to treat, with very satisfying results. Patients with diabetes insipidus present with marked polyuria and polydypsia. On testing, they are found to have inappropriately low urine osmolality and high serum osmolality. Their osmoregulatory system is abnormal, either because of insufficient production of antidiuretic hormone or because of renal insensitivity to antidiuretic hormone. Whether the etiology of diabetes insipidus is neurogenic or nephrogenic, the patient is unable to conserve water when deprived of fluids.

Normal Physiology

In a healthy person, serum osmolality is maintained within the narrow range of 285 to 295 mOsm per kg (285 to 295 mmol per kg). When the person becomes dehydrated, osmoreceptors in the anterior pituitary gland sense the rising serum osmolality and trigger the release of antidiuretic hormone from the posterior pituitary, where it is stored. The antidiuretic hormone binds to the V2 receptors of the renal collecting duct, increasing the duct's permeability to water,[1,2] so that more free water is reabsorbed by the kidneys.

In persons with normal antidiuretic hormone secretion and normal renal concentrating ability, urine osmolality can range from approximately 100 to 1,200 mOsm per kg (100 to 1,200 mmol per kg), depending on the need to excrete or conserve free water. For maximal renal concentrating ability, the medullary gradient must also be intact. To fully restore normal serum tonicity, there must be a concomitant increase in fluid intake; when serum osmolality exceeds 290 mOsm per kg (290 mmol per kg), the person experiences increased thirst. Thus, in the face of water deprivation, the body compensates by concentrating the urine and stimulating the urge to drink.

Although antidiuretic hormone secretion is most sensitive to changes in osmolality, it is also stimulated by hypovolemia, hypoglycemia and nausea.[1] A 10 percent drop in blood volume is required to stimulate antidiuretic hormone release. Drugs that stimulate antidiuretic hormone release are nicotine, carbamazepine (Tegretol) and clofibrate (Atromid-S).[1]

Etiology and Pathophysiology

In persons who secrete suboptimal levels of antidiuretic hormone and in those whose renal tubules do not respond to the hormone, the ability to conserve free water is compromised. Thus, these persons excrete large volumes of dilute urine, necessitating a large intake of dilute solutions to maintain normal fluid and electrolyte balances. Patients with milder forms of diabetes insipidus can partially compensate at the level of the glomerulus and proximal tubule. When fluid is restricted in these patients, the volume of filtrate at the glomerulus decreases, and more water is reabsorbed' et the proximal tubule. In addition, the collecting duct is partially permeable to water at all times, allowing for some reabsorption even without antidiuretic hormone.

As mentioned earlier, there are two types of diabetes insipidus, neurogenic (central) and nephrogenic (peripheral). Neurogenic diabetes insipidus is much more common and is classified as either idiopathic or secondary to an intracranial event. In one case series[3] of 119 patients with neurogenic diabetes insipidus, 25 percent of cases were idiopathic. Thirty percent were related to benign brain tumors, and many of the 119 patients developed diabetes insipidus postoperatively. Less commonly, neurogenic diabetes insipidus was precipitated by blunt head trauma (17 percent), neurosurgery (9 percent), ischemic or toxic brain injury (6 percent) or metastatic cancer (8 percent). Rare causes are histiocytosis and granulomatous disease (tuberculosis, sarcoidosis); there is also an autosomally dominant inherited form (Table 1). In children, central diabetes insipidus is idiopathic in only 8 percent of cases and most often is caused by tumors (craniopharyngiomas), neurosurgery and trauma, or is a sequela of meningitis.[4]

Differential Diagnosis

Included in the differential diagnosis for polyuria and polydypsia are psychogenic polydypsia, various electrolyte disturbances, certain nephropathies and adverse effects of certain drugs. Patients with psychogenic polydypsia (compulsive water drinkers) may drink up to 20 L per day and vary their intake dramatically from day to day. The renal concentrating defect is caused by dilution of the medullary concentrating gradient rather than by suboptimal secretion or response to antidiuretic hormone. Excessive solute load, as in uncontrolled diabetes or salt-wasting nephropathy, will overpower the concentrating ability of the tubules through osmotic diuresis and can mimic diabetes insipidus. Various drugs, such as antipsychotics (i.e., chlorpromazine [Thorazine]) or anticholinergics, stimulate increased water intake (by causing dry mouth), leading to polydypsia and polyuria.

Diagnosis

The water deprivation test is used to confirm the presence of diabetes insipidus and to differentiate between central and nephrogenie forms.[2,11,12] In a controlled environment, a mild state of dehydration is induced, with hourly monitoring of urine osmolality, urine specific gravity, serum Na+ and serum osmolality. Adequate dehydration is documented by a weight loss of 1.35 to 2.25 kg (3 to 5 lb) or two sequential urine osmolality values that differ by less than 30 mOsm per kg (30 mmol per kg), indicating that a steady state of maximum concentration has been achieved. Serum antidiuretic hormone level is then determined, and the patient is given 5 U of subcutaneous antidiuretic hormone. Urine and serum osmolality values are determined several times at hourly intervals.

The key to diagnosing diabetes insipidus and distinguishing between nephrogenic and central diabetes insipidus lies not in the absolute test results but in the association between urine and serum osmolality, and antidiuretic hormone level in response to dehydration (Table 2). The following parameters are only guidelines.

ADH = antidiuretic hormone.

(*)--Dehydration test: no fluid intake for 4 to 18 hours; hourly measurement of urine osmolality, urine specific gravity, serum Na+ and serum osmolality; weigh patient (or urine) frequently to avoid loss of more than 3 percent of body weight (if more than 3 percent of body weight is lost, measure plasma and urine osmolality and terminate test to avoid hypotension). After two successive hourly urine osmolality determinations differ by less than 30 mOsm per kg (30 mmol per kg), indicating that a plateau has been reached, administer 5 U of ADH subcutaneously and, 60 minutes later, measure urine osmolality.

[dagger]--One hour after subcutaneous injection of 5 U of aqueous ADH. Plasma ADH levels must always be interpreted relative to plasma osmolality.

[double dagger]--Useful to distinguish partial from complete central diabetes insipidus.

[subsections]--Differs from central diabetes insipidus.

Adapted from Wallach J. Interpretation of diagnostic tests: a synopsis of laboratory medicine. 4th ed. Boston: Little, Brown, 1986:511. Used with permission.

Urine osmolality values that remain inappropriately low as the serum osmolality rises (i.e., a urine-to-serum osmolality ratio of less than 1) suggests diabetes insipidus. The patient's response to the administered antidiuretic hormone is quite helpful. In patients with complete central diabetes insipidus, a 50 percent rise in urine osmolality occurs, along with a fall in urine output, serum sodium level and serum osmolality value. In partial or milder forms of central diabetes insipidus, the rise in urine osmolality typically ranges from 10 to 50 percent. In nephrogenic diabetes insipidus, urine osmolality does not increase, and in primary polydypsia, it rises less than 9 percent.[2,12]

Measured antidiuretic hormone levels after water deprivation are inappropriately high in patients with nephrogenic diabetes insipidus and often below 1.0 pg per mL (1.0 ng per L) in patients with neurogenic diabetes insipidus. Since antidiuretic hormone deficiency covers a spectrum, at times the water deprivation test may not be diagnostic. In cases where the serum osmolality does not reach 300 mOsm per kg (300 mmol per kg), implying insufficient dehydration, hypertonic saline may be infused, and antidiuretic hormone levels measured again.[13]

In all cases, the test must be aborted in patients who demonstrate vascular compromise or a hyperosmolar state. During pregnancy, the water deprivation test criteria may be modified to reflect the patient's lower resting serum osmolality, and some clinicians recommend that the fetus be monitored because of the risk of maternal dehydration.[14] Patients with confusing test results should be referred.

If a patient is diagnosed with central diabetes insipidus, the underlying cause must be determined (Figure 1). In a large majority of patients, the cause will be evident from the history and physical examination (i.e., history of trauma, neurosurgery, known tumor). Magnetic resonance imaging (MRI) of the head is indicated in patients who are otherwise asymptomatic, to rule out a hypothalamic tumor. MRI is thought to be 80 to 95 percent sensitive in detecting microadenomas, with computed tomographic (CT) scan only slightly less sensitive.[15] In neurogenic diabetes insipidus, MRI is quite specific, since the normal "bright spot" of a functioning posterior pituitary gland will be absent. In persons with partial diabetes insipidus, this may not be the case. However, 10 to 20 percent of normal patients do not have a "bright spot."[15]

[Figure 1 ILLUSTRATION OMITTED]

Treatment

As mentioned, medical treatment is indicated in patients whose symptoms are debilitating. Patients with central diabetes insipidus are treated with nasal desmopressin (DDAVP), a synthetic analog of vasopressin without pressor or uterine effects. DDAVP is the formulation of antidiuretic hormone that is best tolerated, easiest to administer and most consistent in its antidiuretic effect. The duration of relief ranges from eight to 20 hours and varies between patients but remains constant for individual patients.

The starting dose of 10 [Micro]g at night is usually sufficient to relieve nocturia, and a morning dose may be added if symptoms persist during the day. The actual dose delivered can be titrated by using a special nasal catheter. DDAVP is expensive. One bottle (50 doses of 10 [Micro]g each) typically costs $100. All patients are encouraged to undertreat themselves, to guard against volume overload and hyponatremia. The shorter acting lysine vasopressin may be useful as a supplement.

Patients with partial pituitary diabetes insipidus may respond to the hypoglycemic drug chlorpropamide (Diabinese), which stimulates antidiuretic hormone secretion and potentiates its effect on the kidneys. Hypoglycemia, however, often limits its use. Clofibrate and carbamazepine may also stimulate antidiuretic hormone secretion and may be used as treatment in patients with partial diabetes insipidus.

DDAVP is safe and the only commercial form of antidiuretic hormone that is effective for use in pregnant women. Women typically require higher dosages while pregnant.[16]

In patients with nephrogenic diabetes insipidus, antidiuretic hormone replacement is ineffective. Certain drugs do, however, improve polyuria and thus improve the polydypsia. Diuretics, such as thiazides, amiloride (Midamor), or both, function by depleting total body salt, thus increasing the isotonic absorption of water in the proximal tubule. However, this effect can be countered by a diet heavy in salt. Most patients respond to a daily dosage of 50 to 100 mg of hydrochlorothiazide (Esidrix, Hydrodiuril) but run the risk of developing hypokalemia. Again, patients must be instructed to have fluids available at all times.

Final Comment

Although diabetes insipidus is not a common illness, most family physicians occasionally encounter a patient with polyuria and polydypsia and should be able to determine when to consider this potentially serious illness in a differential diagnosis. Patients often tolerate the inconveniences of this illness for several years, and physicians can easily overlook it. Treatment is simple yet often has a dramatic effect on quality of life.

REFERENCES

[1.] Seckl JR, Dunger DB. Diabetes insipidus. Current treatment recommendations. Drugs 1992,44:216-24.

[2.] Reeves WB, Andreoli TE. The posterior pituitary and water metabolism. In: Wilson JD, Foster DW, eds. Williams' Textbook of endocrinology. 8th ed. Philadelphia: Saunders, 1992:311-56.

[3.] Moses AM. Clinical and laboratory observations in the adult with diabetes insipidus and related syndromes. Front Horm Res 1985;13:156-75.

[4.] Greger NG, Kirkland RT, Clayton GW, Kirkland JL. Central diabetes insipidus. 22 years' experience. Am J Dis Child 1986;140:551-4.

[5.] Grof P, MacCrimmon DJ, Smith EK, Daigle L, Saxena B, Varma R, et al. Long-term lithium treatment and the kidney. Interim report on fifty patients. Can J Psychiatry 1980;25:535-44.

[6.] Bucht G, Wahlin A. Renal concentrating capacity in long-term lithium treatment and after withdrawal of lithium. Acta Med Scand 1980;207:309-14.

[7.] Williams DJ, Metcalfe KA, Skingle L, Stock AI, Beedham T, Monson JR Pathophysiology of transient cranial diabetes insipidus during pregnancy. Clin Endocrinol 1993;38:595-600.

[8.] Iwasaki Y, Oiso Y, Kondo K, Takagi S, Takatsuki K, Hasegawa H, et al. Aggravation of subclinical diabetes insipidus during pregnancy. N Engl J Med 1991;324:522-6.

[9.] Robinson AG, Amico JA. "Non-sweet" diabetes of pregnancy [Editorial]. N Engl J Med 1991;324: 556-8.

[10.] Durr JA. Diabetes insipidus in pregnancy. Am J Kidney Dis 1987,9:276-83.

[11.] Robertson GL. Differential diagnosis of polyuria. Annu Rev Med 1988;39:425-42.

[12.] Blevins LS Jr., Wand GS. Diabetes insipidus. Crit Care Med 1992;20:69-79.

[13.] Zerbe RL, Robertson GL. A comparison of plasma vasopressin measurements with a standard indirect test in the differential diagnosis of polyuria. N Engl J Med 1981;305:1539-46.

[14.] Krege J, Katz VL, Bowes WA Jr. Transient diabetes insipidus of pregnancy. Obstet Gynecol Surv 1989;44:789-95.

[15.] Elster AD. Imaging of the sella: anatomy and pathology. Semin Ultrasound CT MR 1993;14:182-94.

[16.] Kallen BA, Carlsson SS, Bengtsson BK. Diabetes insipidus and use of desmopressin (Minirin) during pregnancy. Eur J Endocrinol 1995;132(2):144-6.

PATRICIA ADAM, M.D., M.S.P.H. is an assistant professor at the Riverside Unit of the University of Minnesota's Department of Family Practice and Community Health in Minneapolis. Dr. Adam graduated from the University of Virginia School of Medicine, Charlottesville, and completed an academic fellowship at the University of Missouri-Columbia School of Medicine, where she also completed a residency in family medicine.

Address correspondence to Patricia Adam, M.D., Smiley's Clinic, 2615 E. Franklin Ave., Minneapolis, MN 55406.

COPYRIGHT 1997 American Academy of Family Physicians
COPYRIGHT 2004 Gale Group