The most appropriate diagnostic test to perform next is a water restriction test to evaluate for diabetes insipidus (DI). DI is caused by either an absence of antidiuretic hormone (ADH) secretion (central DI) or renal resistance to ADH (nephrogenic DI), which results in an inability to appropriately concentrate the urine in response to an increase in plasma osmolality. This patient's symptoms of polyuria and polydipsia, in association with long-standing lithium therapy, are suggestive of nephrogenic DI. Lithium is one of the most common causes of nephrogenic DI in adults. On laboratory testing, serum sodium and plasma osmolality are usually high normal or slightly elevated, whereas urine osmolality is lower than plasma osmolality. Like the patient described, most patients with DI do not usually have frank hypernatremia because increased thirst stimulates oral consumption of fluids, which maintains the serum sodium near the upper normal range as long as access to fluids is not impaired. In a water restriction (or deprivation) test, urine volume, urine osmolality, and plasma sodium concentration are measured hourly after complete water restriction. A normal urine osmolality response (usually defined as an increase in urine osmolality above 600 mOsm/kg H2O) indicates that ADH release and corresponding renal response to ADH are intact. A failure of the urine osmolality to rise despite rising plasma osmolality suggests either central or nephrogenic DI. Desmopressin is then administered. Patients with central DI will respond with increased urine osmolality, whereas in patients with nephrogenic DI (as is likely in this patient), desmopressin will not result in increased urine osmolality after water restriction, confirming the diagnosis.
The cosyntropin stimulation test is used to diagnose adrenal insufficiency, which is manifested by hyponatremia, decreased plasma osmolality, and increased urine osmolality, none of which is seen in this patient.
Hypothyroidism, which is diagnosed by the finding of a high thyroid-stimulating hormone level, can be associated with hyponatremia. In contrast to this patient with DI, patients with hyponatremia secondary to hypothyroidism present with hyponatremia, decreased plasma osmolality, and increased urine osmolality.
A urine sodium measurement is useful in the evaluation of patients with suspected urinary salt wasting, such as those with adrenal insufficiency. Patients with DI have a deficit of free water rather than urinary salt wasting, and as such, urine sodium measurement is not useful in diagnosing DI.