Renal diluting ability is also impaired as a function of aging (Crowe et al., 1987; Epstein, 1985; Lindeman et al., 1966). In water-diuresing subjects as a result of water loading, minimal urine osmolality was significantly higher: 92 mOsmol/kg in the elderly subjects (aged 77 to 88 years) when compared with 52 mOsmol/kg in the young subjects (aged 17 to 40 years). Free water clearance was also decreased: 5.9 mL/minute in the elderly subjects compared with 16.2 mL/ minute in the young subjects (Lindeman et al., 1966). While the impairment is largely due to the decrease in GFR, when free water clearance is factored for GFR, the ratio of free water clearance to GFR is, however, still decreased in the older subjects (Crowe et al., 1987; Lindeman et al., 1966). Mechanisms of the impaired diluting ability in the elderly have not been well studied. In addition to the major role of impaired GFR, inadequate suppression of arginine vasopressin release or impaired solute transport in the ascending loop of Henle may also play a role.

The age-related impairments in renal-concentrating and sodium-conserving ability are associated with an increased incidence of volume depletion and hypernatremia in the elderly (Snyder et al., 1987). Under normal physiological conditions, increased thirst and fluid intake are natural defense mechanisms against volume depletion and hypernatremia. A deficit in thirst and regulation of fluid intake in the elderly, however, may further contribute to the increased incidence of dehydration and hypernatremia.

Several studies confirm the long-held clinical observation that thirst and fluid intake are impaired in the elderly (Fish et al., 1985; Miller et al., 1982; Murphy et al., 1988; Phillips et al., 1984). In a series of studies the osmotic threshold for thirst during hypertonic saline infusion has been found to be much higher in healthy elderly subjects than in their younger counterparts, with many apparently healthy elders not reporting thirst despite elevations of plasma osmolality to levels over 300 mOsmol/kg (Fish et al., 1985). In studies of water ingestion after intravenously induced hyperosmolality, elderly individuals demonstrated marked reductions in their water intake and rate of return of plasma osmolality to baseline when compared with the younger group (Murphy et al., 1988). The influence of free access to water on prevention of serum osmolality increases during hypertonic saline infusion was also investigated (McAloon-