Increased susceptibility to thiazide-induced hyponatremia in the elderly

Hyponatremia is a common cause of morbidity in the elderly, and thiazide diuretics are often implicated. Eleven healthy young volunteers, eight healthy old volunteers, and five elderly patients with a history of thiazide-induced hyponatremia were studied to determine susceptibility to thiazide-induced hypoosmolality in age. Each of the healthy subjects ingested a water load (20 mL/kg) after 3 days of hydrochlorothiazide (HCTZ) (100 mg/day) or placebo. Although there were no differences in minimum Uosm between young and old, the healthy old had lower hourly free water clearances (CH₂O) as compared with the young and a greater decline in serum osmolality in response to water loading (P < 0.05). HCTZ impaired minimum urine osmolality and CH₂O and delayed recovery of serum osmolality after the water load in both healthy young and old (P < 0.005, placebo versus HCTZ), but the impairment in the latter two parameters was greater in the healthy elderly (P < 0.05, young versus old). Vasopressin levels were not different between healthy young and old (1.9 ± 0.3 versus 2.0 ± 1.0 pm with placebo; 3.0 ± 0.7 versus 4.4 ± 1.0 with HCTZ). Five of the young subjects were restudied after the addition of ibuprofen (400 mg thrice daily) to the thiazide and placebo regimens. Creatinine clearance was not changed, but free water clearance and serum osmolality after water loading were significantly reduced to a degree similar to that seen in the elderly subjects on the thiazide regimen (P < 0.05), suggesting an important role for renal prostaglandins in the defense against hyponatremia. Five elderly subjects with a prior history of thiazide-induced hyponatremia were also challenged with a water load. Compared with the healthy elderly subjects, these patients had an impairment in the ability to dilute the urine maximally as well as an impairment in free water excretion and a greater reduction in serum osmolality (P < 0.05). Vasopressin levels were not elevated in these patients. In summary, HCTZ produces defects in water excretion after a water load (a decrease in CH₂O and a decline in serum osmolality) that are greater in healthy elderly than young subjects and that are unrelated to differences in vasopressin or minimum Uosm. Prostaglandin inhibition in young subjects together with thiazide further decreased CH₂O and compromised the recovery of serum osmolality in a manner similar to that observed in the elderly. The elderly patients with a history of thiazide-induced hyponatremia had defects in urinary dilution and an even greater susceptibility to hyponatremia. In conclusion, alterations in the distal tubular delivery of water, possibly related to lower prostaglandin production, may contribute to susceptibility to thiazide-induced hyponatremia in the elderly. Elderly subjects with unsuspected underlying defects in urinary dilution may present with clinically significant hyponatremia in the setting of thiazide use and water ingestion.