The term nephrosclerosis describes a microstructural biopsy pattern of global glomerulosclerosis, arteriosclerosis, and interstitial fibrosis with tubular atrophy. Nephrosclerosis is notably seen with hypertension but is also described in healthy older kidney donors without hypertension or only mild hypertension. Thus the term “hypertensive nephrosclerosis” may be a misnomer. Nephrosclerosis is thought to be due to increased intimal thickening of small arteries in the kidneys (arteriosclerosis, arteriolosclerosis), leading to glomerular ischemia, with the wrinkling of capillary tufts, thickening of the basement membrane, and progressive pericapsular fibrosis. ^, Simultaneously, proteinaceous material accumulates in the Bowman space, likely due to a combination of a disturbed balance between formation and breakdown of the glomerular extracellular matrix and podocyte depletion, ^{, ,} probably due to perturbations in local mechanical forces. Eventually, ischemic glomerular tufts fully collapse, forming GSG. As a consequence of glomerulosclerosis, accompanying tubules atrophy with an accumulation of surrounding interstitial fibrosis.

All four features of nephrosclerosis—global glomerulosclerosis (GSG), arteriosclerosis, IFTA—increase with older age and, when combined together (at equal weighting) they represent a sclerosis score ( Fig. 21.5 ). If we define nephrosclerosis as the presence of at least two of the previously mentioned abnormalities, then from Fig. 21.5 we can see that the youngest donors (18–19 years of age) do not have detectable nephrosclerosis. The prevalence of nephrosclerosis in 20- to 29-year-old donors is 5% and rises to 69% among the oldest donors.

The nephrosclerosis score increases with age among 1814 living kidney donors from the Aging Kidney Anatomy study (expansion of previously reported findings).

The total nephrosclerosis score is obtained by adding the individual scores of histologic abnormalities: any global glomerulosclerosis, any arteriosclerosis, interstitial fibrosis greater than 5%, and any tubular atrophy. White bars in all age groups represent a score of 0 (no abnormalities present), and a dark purple bar represents a score of 4 (presence of all four pathologic abnormalities). Three intermediate scores are presented with different shades of purple color .

From Rule AD, Amer H, Cornell LD, et al. The association between age and nephrosclerosis on renal biopsy among healthy adults. Ann Intern Med. 2010;152:561–567.

The number and size of the nephrons are the primary determinants of the kidney cortical volume. A combination of age-related glomerulosclerosis with corresponding IFTA is responsible for the observed loss of cortical volume with healthy aging seen in living kidney donors. This nephrosclerosis typically starts in glomeruli in the superficial cortical zones, as evident by the high proportion of global glomerulosclerosis in this region compared with middle and deep regions in older adults. Because the corresponding tubules of these superficial glomeruli contribute more to the cortical volume, their atrophy with age-related nephrosclerosis leads to a decrease in cortical volume with aging. There is less evidence of GSG with age in the deeper cortex among nephrons that descend into the medulla. Despite a 24% to 33% decrease in nephron number from ages 18 to 76 years, there is only a 17% decrease in cortical volume and only a 12% decrease in total kidney volume due to hypertrophy of the distal tubule. ^,

One might expect that with age-related nephrosclerosis, there would be compensatory hypertrophy of the remaining, functional glomeruli. However, studies in living kidney donors have not confirmed this hypothesis. A possible explanation may be that with healthy human aging, there is a decreased metabolic demand for glomerular function, so the loss of glomeruli does not cause the enlargement of the remaining glomeruli. Notably, comorbidities that can become more prevalent in older age, such as obesity and type 2 diabetes with albuminuria, associate with glomerular hypertrophy. Also, low nephron endowment at birth or disorders that accelerate nephron loss with advancing age can contribute to glomerulomegaly observed in adulthood.

Similar to age-based thresholds in glomerulosclerosis, the upper 95th percentile reference limit of IFTA has been described. For example, for the youngest kidney donors (18–29 years of age), the 95th percentile expected for age was 0.18% IFTA, which increased to 6.5% for the oldest donors (70–76 years of age) ( Fig. 21.3 ).

In a series of pioneering studies begun more than 60 years ago, Davies and Shock examined 70 healthy adult men between the ages of 24 and 89 years and, in a cross-sectional analysis, showed (by urinary inulin clearance) a monotonically lower GFR in subjects older than 30 years. The GFR in the oldest group was found to be 46% less as compared with the GFR in the youngest group. Several decades later, Lindeman and colleagues carried out a longitudinal study (the first of its kind) by following 254 mostly healthy adult individuals, although some had diabetes, for up to 14 years. They found that the mean annual decline in GFR, estimated by 24-hour urinary creatinine clearance, was 7.5 mL/min per decade; however, one third of the subjects had no decrease in kidney function and a small subset experienced an increase in GFR. In addition to measurement error explaining this finding, a rise in GFR with age might also be explained by onset of comorbidities (e.g., obesity, type 2 diabetes) that cause hyperfiltration and become more common with aging. This longitudinal rate of GFR decline is similar to the cross-sectional decline of 6.3 mL/min per decade obtained from a study of potential kidney donors (GFR measured by iothalamate clearance). Among 4500 potential kidney donors in the expanded analysis from the Aging Kidney Anatomy study ( Fig. 21.6 ), the average decline per decade for estimated GFR was 7.4 mL/min/1.73 m ² , whereas for measured GFR (iothalamate clearance) it was 6.1 mL/min/1.73 m ² .

Age-related glomerular filtration rate ( GFR; mean values) decline among 4500 potential kidney donors in the Aging Kidney Anatomy study (expansion of previously reported findings).

Decline in the estimated GFR (CKD-EPI study equation; see text) is calculated to be 7.4 mL/min/decade and 6.1 mL/min/decade for the measured GFR.

From Rule AD, Amer H, Cornell LD, et al. The association between age and nephrosclerosis on renal biopsy among healthy adults. Ann Intern Med. 2010;152:561–567.

Although Figure 21.6 shows mean estimated GFR (eGFR) and measured GFR, there is variability within each age group ( Table 21.2 ). It can be appreciated that both upper and lower reference limits also decline with aging. Similar findings in healthy donors have been reported by Pottel and colleagues. Heathy aging is associated with a higher measured GFR compared with unhealthy aging (with comorbidity). Other longitudinal studies of the decline in kidney function (GFR or creatinine clearance) with aging have shown generally similar findings to those of Lindeman and associates, including the relative stability of kidney function for extended periods in some persons who were otherwise experiencing healthy aging. In addition, in these studies, the rate of decline of kidney function appeared to increase with each passing decade in persons beyond the fourth decade.

Among patients who underwent radical or partial nephrectomy for a kidney tumor and who had a minimal amount of nephrosclerosis on biopsy, there was still a substantial decrease in eGFR of about 35 mL/min/1.73 m ² from age 25 years to 75 years. ^, Notably, this decline in eGFR with 5 decades of aging among persons with minimal nephrosclerosis is about double the decrease in eGFR from minimal to moderate nephrosclerosis at the same age (17 mL/min/1.73 m ² ).

Several studies have found that a higher whole-kidney GFR correlates with larger kidneys or larger kidney cortical volume. ^{, ,} The age-related nephron loss due to increased nephrosclerosis is followed by a parallel whole-kidney GFR decline. There is no compensatory increase in single-nephron GFR (snGFR) as nephrons are lost with aging.

Although in vivo measurements of snGFR through glomerular micropuncture are possible and have been done in animal studies, ^, a direct measure of snGFR (by micropuncture) is not feasible or safe in humans. However, dividing the whole-kidney GFR by the number of functional (nonsclerosed) glomeruli in both kidneys allows for an estimation of the mean snGFR. The snGFR remains stable with age because the loss of nephrons parallels the loss in total GFR. Thus the effect of age on snGFR is minimal, consistent with other physiologic characteristics, including height and sex, whereas obesity, family history of ESKD, and nephrosclerosis exceeding that expected for age are associated with an increased snGFR. ^,

Although other studies of snGFR in living human subjects are lacking, some studies have indirectly assessed the single-nephron ultrafiltration coefficient (snK _f ) in individuals similar to living kidney donors. ^{, ,} The snK _f represents the filtering capacity of the glomerulus, as determined by the surface area and permeability of the glomerular filtration barrier. In these studies, snK _f was estimated from electron microscopy measures of the glomeruli on kidney biopsy. The snGFR could be calculated by multiplying snK _f by the perfusion pressure across the glomerular filtration barrier. Thus it is possible that the relationship of clinical characteristics to snGFR may be similar to that of snK _f . Consistent with the findings of a relatively stable snGFR across the age spectrum described earlier, snK _f did not show significant differences between younger and older living kidney donors. ^, Adaptive hyperfiltration occurs in the aging kidney following unilateral nephrectomy ; however, either due to an increase in Kf, glomerular capillary filtration pressure, or increased glomerular plasma flow or some combination of the three, the increase in snGFR (and whole-kidney GFR [wkGFR]) is reduced compared with younger persons. The definition of hyperfiltration based on measurements of the whole-kidney GFR is best determined by an age-adapted measured GFR (not eGFR), uncorrected for body surface area. However, such definitions do not distinguish between a high snGFR a or high nephron number.

Renal plasma flow (RPF) represents the volume of plasma delivered to both kidneys per unit of time. The RPF can be calculated from the amount of plasma that is cleared of para-aminohippurate (PAH) per unit of time. PAH is a compound with a nearly 100% extraction through the kidneys. A Swedish study of 122 potential kidney donors, ages 21 to 67, has shown that similar to the GFR, the renal plasma flow (RPF) declines with older age. Older age influences the degree of renal blood flow change as a response to exercise ^, and attenuates the renal hemodynamic response to atrial natriuretic peptide and the degree of vasodilation. One study has postulated that higher levels of the endogenous nitric oxide (NO) inhibitor asymmetric dimethylarginine (ADMA) lead to a reduced RPF and hypertension with older age. Animal studies have provided some evidence that ADMA is associated with tubulointerstitial ischemia and fibrosis, glomerular capillary loss, and glomerulosclerosis. In a study of healthy individuals, infusion of dopamine and amino acids was used to test the effects of maximal vasodilation on RPF and NO. Whereas both RPF and NO levels increased in young and middle-aged individuals, they did not change in older persons. This suggests that due to advanced age-related vascular changes in older adults, renal blood vessels are less responsive to maximal vasodilation.

The filtration fraction (FF) represents the proportion of fluid entering the kidneys that reaches into the renal tubules—that is, it is a ratio between the GFR and RPF. Normally, the value for FF is about 20% and remains relatively constant in older age. However, in very old age, atherosclerosis and arteriosclerosis of the arteries and arterioles reduce blood flow to kidneys, thereby increasing FF.

The phenomenon of renal reserve refers to an acute increase in GFR (usually by 20% or more from basal values) when human subjects (or experimental animals) are fed oral protein loads (typically cooked red meat) or infused intravenously with certain amino acids. The mechanisms underlying these physiologic phenomena are not fully understood but appear to be a consequence of vasodilatation, an increase in glomerular plasma flow, and an increase in the snGFR (see also Chapter 3). Growth hormone may be involved, but this is controversial. The efficiency of renal reserve is blunted to some degree by aging per se, most often in the very elderly. Functional reserve is well maintained up to about age 80 years in otherwise healthy men and women. Diminished levels of sex hormones may play a role in the diminished renal reserve seen with aging (see later). The diurnal variation in GFR is also blunted with aging.

Aging in indigenous Kuna Indians (Panama), who subsist on a low-protein, high–dark chocolate diet and remain largely free of hypertension and cardiovascular disease (CVD), is associated with a steady decline in GFR and RPF, similar to that observed in westernized countries. These findings suggest that hypertension is a consequence of CKD and/or nephron underendowment, not the cause of CKD with aging.

In a series of studies, mainly focused on murine species, Baylis and coworkers developed the concept of sexual dimorphism in renal aging (also see review by Gava and associates ). Women develop less age-related decline in kidney function, perhaps due to a renoprotective action of estrogens. In contrast, men have more pronounced loss of kidney function with advancing age, perhaps due to the adverse effects of androgens. The mechanisms involved are complex but seem to implicate disturbances of NO synthesis pathways, with a lower rate of NO production potentially influenced by sex steroids. ^, A role for the renin-angiotensin-aldosterone system (RAAS) has also been described. Interestingly, in aging experimental animals, there was no observed rise in glomerular capillary pressure or glomerulomegaly in males or females, castrated or not. These findings suggest that the age-dependent sexually dimorphic loss of kidney function does not appear to result in compensatory hemodynamic changes or glomerular hypertrophy, as is observed in humans. Age-dependent sexual dimorphic effects on kidney function might be related to increased body size in males and therefore may be conditional on the balance between energy demand and supply. These findings in murine species have been recapitulated in humans by Melsom and colleagues in a longitudinal study of measured GFR in men and women aged 50 to 62. The measured GFR in healthy women declined more slowly than in men.

Studies in healthy human subjects clearly demonstrate that kidney aging reduces the capacity of the tubules to conserve filtered sodium, probably by effects in the distal nephron, previously discussed in Chapter 13.

Sodium homeostasis can also be influenced by age-related changes in levels and responses to renin and aldosterone, hormones that regulate sodium conservation. Although plasma renin activity and aldosterone levels are generally lower in older adults, they are not associated with changes in fluid or electrolyte metabolism. These changes increase older adults’ susceptibility to sodium retention in disease states or with the use of medications that alter renin release.

It has also been hypothesized that the higher prevalence of hypertension in older persons may be due in part to the attenuated natriuretic response with aging. There is also a reduction in the natriuretic response to saline loading in older versus younger kidney donors after donation.

Atrial natriuretic peptide (ANP) is a hormone secreted by atrial myocytes, and one of its functions is to control sodium excretion at the tubular level. With older age, the tubular response to ANP is reduced. ANP inhibits sodium reabsorption from the luminal side and induces hyperfiltration and suppression of renin release, which together lead to natriuresis, diuresis, and lowering of blood pressure. Studies have shown that the plasma ANP concentrations are several times higher in older individuals, and this may be a compensatory effect of the reduced response at the receptor level. This hypothesis is in agreement with two studies in which sodium excretion plateaued after an ANP infusion in older subjects, ^, whereas in younger subjects sodium excretion continued to increase with increasing ANP dose. Of note, it appears that ANP production does not change with older age; therefore the observed increase of ANP in older adults results from reduced metabolic clearance. ^,

The role of altered renal sodium handling in the pathogenesis of age-related elevation of systemic arterial pressure is a topic of great contemporary interest (see also Chapter 46; reviewed by Frame and Wainford ). Vascular factors, such as reduced compliance of major arterial vessels, and neurohumoral alterations, such as increased sympathetic nervous system activity, play important roles in the elevation of systolic blood pressure with aging. The sensitivity of blood pressure changes consequent to salt administration increases with older age.

The expression of the epithelial sodium channel (ENaC) can be reduced in aging rodents. These studies collectively indicate that dysregulation of sodium handling by the aging nephron can be involved in age-related blood pressure elevation. Integration of this knowledge into the complex array of biologic factors in systems affecting blood pressure (e.g., aortic compliance, sympathetic nervous system activity, RAAS) will be challenging. Moreover, the role of reduced nephron mass that accompanies physiologic aging cannot be ignored.

Potassium homeostasis is regulated by several different mechanisms; specifically, in the kidney, the potassium excretion rate is altered depending on the intake (discussed further in Chapter 16).

Total body potassium (TBK) declines with older age with a reduction in intracellular stores that accompany the decline in muscle mass (sarcopenia) seen with normal aging. The observed decline in TBK may also be explained by the lower renin and aldosterone levels, as well as the blunted response to aldosterone in older adults (see later). Alternatively, older adults may express impaired ability to excrete acute potassium loads. Insulin-mediated potassium homeostasis in human subjects is not influenced by age.

Older age is often accompanied by comorbid conditions for which medications that interfere with potassium excretion might be prescribed. Examples include potassium-sparing diuretics, angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers, heparin, calcineurin inhibitors, sodium channel blockers, and nonsteroidal antiinflammatory drugs (NSAIDs). Care is needed when starting these medications for older patients, given the risk of hyperkalemia.

It is estimated that about 99% of magnesium is stored in cells, predominantly in bones and muscles, and only about 1% is in extracellular fluid including plasma. Normally, the kidneys reabsorb 96% of filtered magnesium and regulate magnesium homeostasis (see Chapter 17).

Because plasma and intracellular magnesium homeostasis are finely regulated, total plasma magnesium is maintained remarkably stable with increasing age in healthy individuals. ^, However, subclinical deficits of magnesium not detected by serum magnesium concentrations can occur in older individuals.

Changes in magnesium homeostasis seen in the elderly are likely due to diseases that frequently occur in older age, use of therapeutic agents that alter its metabolism, and an age-related decline in kidney function. On the other hand, hypermagnesemia occurs rarely, even in the disease setting, including in those with severe acute kidney injury (AKI), CKD, or ESKD. Interestingly, low dietary magnesium intake in older adults may also contribute to a faster rate of decline in GFR, potentially exacerbating a proinflammatory milieu, endothelial dysfunction, and promotion of vascular calcification.

In normal adults, approximately 10 g of calcium is filtered through the glomeruli each day, and 98% to 99% is reabsorbed. Net calcium excretion during a 24-hour period ranges from 100 to 200 mg. Most filtered calcium reabsorption (≈60%–70%) occurs in the proximal tubules, around 20% in the loop of Henle (thick ascending limb), around 5% to 10% in the distal convoluted tubule, and only around 5% in the collecting ducts (see also Chapter 17).

Calcium homeostasis with aging has been extensively studied in man and experimental animals. Some animal and human studies have shown that there are age-related changes in renal calcium homeostasis, such as decreased tubular reabsorption and/or the renal response to parathyroid hormone (PTH), despite relative preservation of renal calcium excretion with aging. Renal and intestinal calbindins are proteins with a critical role in active transcellular calcium transport, and their expression decreases with age. There is also an age-related decrease in the capacity of 1,25-dihydroxyvitamin D ₃ (1,25-[OH] ₂ D ₃ ) to stimulate calcium reabsorption, indirectly increasing PTH concentrations in both rats and humans. ^, Deficits in renal and duodenal calcium intracellular transporters can lead to impaired renal and duodenal calcium reabsorption, consistent with the increased calciuria observed in older mice. Alpha-Klotho (α-Klotho) is a gene associated with accelerated aging when mutated. It encodes a protein that is predominantly expressed in tissues involved with calcium homeostasis. A study in mice has demonstrated that alpha-klotho binds to a subunit of sodium-potassium adenosine triphosphatase (Na ⁺ -K ⁺ -ATPase), suggesting that the age-related decline in Klotho protein could contribute to reduced sodium-potassium ATPase sensing of low calcium levels. Serum-soluble α-Klotho (sKlotho) levels decline with advancing age in humans, independent of the GFR.

There is also an age-related decrease in intestinal calcium reabsorption, which is linked with reduced 1α-hydroxylase activity, lower serum 1,25-(OH) ₂ D ₃ concentrations, and higher serum PTH concentrations. Moreover, older men also had a higher threshold for PTH release; however, this was not associated with a decrease in blood ionized calcium or 1,25-(OH) ₂ D ₃ . This finding suggests that the relationship between PTH and calcium in older age is changed so that irrespective of calcium concentrations, serum PTH concentrations are higher. An interesting study has compared healthy older individuals (≥75 years of age) to patients with CKD and a similar GFR and found that patients with CKD have higher fractional excretion of calcium, suggesting more pronounced calcium wasting in these patients. Concentrations of vitamin D–dependent calcium-binding proteins also decrease with older age, which is associated with the change in intestinal calcium absorption. ^, Another study found that the renal response to PTH infusion and renal vitamin D levels were equivalent in both young and older adults, with the only difference being that the production of vitamin D is somewhat delayed in older adults.

Decline in the GFR with older age leads to reduced levels of non–protein-bound (free) serum calcium but increased levels of serum phosphate. In response to these changes, PTH synthesis increases, which in turn decreases the number of sodium phosphate cotransporters in the proximal tubules, thus leading to increased phosphaturia. Elevated serum phosphate concentrations also stimulate the production of FGF23, which regulates serum phosphate concentrations in the setting of impaired kidney function and suppresses production of 1,25 (OH) ₂ D ₃ . FGF23 concentrations have been found to increase with diminished GFR but not with age per se.

In addition to the hormonal regulation of serum phosphate concentrations in human aging, animal studies have demonstrated a PTH-independent reduction in the intrinsic tubular capacity to reabsorb phosphate. Others have demonstrated that kidneys in aged animals show impaired phosphate transport in renal tubules in response to a low-phosphate diet. ^, It has been proposed that the reason for this impaired phosphate transport is decreased fluidity of a luminal brush border membrane due to the age-related increase in cholesterol and sphingomyelin, which then diminishes sodium phosphate cotransporter activity directly. Consistent with in vivo studies, an in vitro study of cultured renal tubular cells harvested from young and adult rats showed reduced phosphate uptake and adaptation to a phosphate-free medium only in the adult kidney cells. Another study revealed that the reason for the age-related decline in tubular phosphate reabsorption and tubular adaptation to a low dietary phosphate is lower expression of renal sodium phosphate cotransporters on the apical brush border membrane. As mentioned earlier, age-related reduction of α-klotho (a receptor for FGF23) may also be involved in the changes in phosphate hemostasis seen with aging.

Older persons commonly exhibit vitamin D insufficiency or deficiency, often in association with a subtle state of chronic inflammation. Serum concentrations of 25-hydroxyvitamin D (25- [OH] D) are inversely associated with biomarkers of inflammation in older adults, such as C-reactive protein and interleukin 6 (IL-6). It is not known whether these changes are due to primary changes in vitamin D metabolism or a secondary effect of inflammation. Vitamin D absorption from dietary sources may not be impaired with aging, but reduced dietary intake can contribute to low serum 25-(OH)D levels in aging individuals. Reduced production of 7-dehydrocholesterol in the aging epidermis with ultraviolet light exposure is likely involved in the decreased production of pre–vitamin D ₃ (cholecalciferol) in older adults. Serum 1,25-(OH) ₂ D (calcitriol) concentrations decline in aging, at least in part due to the decline in the GFR and impaired hydroxylation of 25-(OH)D mentioned earlier. Changes in vitamin D sufficiency or insufficiency in older persons can modulate left ventricular geometry and modulate the risk of left ventricular hypertrophy in hypertensive older subjects.

It is important to keep in mind the age-related changes of 1,25-(OH) ₂ D metabolism on intestinal phosphate absorption (described earlier) because several studies have shown that dietary vitamin D supplementation improves renal and intestinal phosphate absorption in vitamin D–deficient animals.

Thus aging has diverse effects on calcium, phosphorus, and vitamin D metabolism. These include a decrease in intestinal absorption of calcium, impaired vitamin D production in the skin, dietary vitamin D deficiency, impaired production of 1,25-(OH) ₂ D ₃ , increased PTH secretion, lower serum Klotho concentrations, higher serum FGF23 concentrations, and alterations in renal tubular handling of calcium and phosphorus. Interestingly, low serum Klotho concentrations also appear to increase the risk of hyperuricemia.

The kidneys play a major role in the regulation of acid-base homeostasis. Normally, during a 24-hour period, kidneys reabsorb around 4500 mmol of filtered bicarbonate. Kidneys have a large capacity to excrete the endogenously generated proton (H ⁺ ) by generating ammonia (NH ₃ ) and its excretion (NH ₄ ⁺ ; see Chapter 15). Older adults are more prone to develop acid-base dysregulation for several reasons. First, age-related structural and functional changes in the kidney reduce the adaptive responsiveness to dietary and/or environmental changes. Second, the age-related decline in GFR reduces the capacity of the kidney to buffer metabolic changes and excrete the excess H ⁺ load. Along with age-related decline in kidney function, there is also a reduced capacity for bicarbonate conservation and generation, which, combined with the stable endogenous H ⁺ production, may lead to metabolic acidosis. This process is enhanced in patients with concomitant CKD. A community-based observational study of healthy persons aged 6 to 75 years has found that the capacity for H ⁺ excretion is similar from childhood to young adulthood but significantly reduced in older age.

Ammonium excretion also decreases with older age. Taken together, it has been suggested that the age-related decrease in H ⁺ excretion is due to reduced renal tubular mass rather than tubular dysfunction. Overall acid production comes from two sources, namely the oxidation of sulfur atoms on amino acids and metabolically produced organic acid ions. In a study of these parameters in aging, Huo and colleagues showed that acid production rates increase with age.

Overall, these studies suggest that older age appears not to alter the fundamental physiologic regulation of acid-base homeostasis but does lead to reduced and diminished responses to acid-base challenge. Consistent with this, a study of older patients with CKD has found that oral sodium bicarbonate supplementation corrects metabolic acidosis, increases serum albumin concentrations, and decreases whole-body protein degradation. Potassium bicarbonate supplementation in postmenopausal women neutralized the endogenous acid, improved calcium and phosphate balance, and reduced bone reabsorption. This was confirmed in a double-blind study of older men and women—in which particularly bicarbonate and not potassium—had a protective effect on bone reabsorption and calcium excretion. Taken together, the results of several studies are consistent with the hypothesis that a higher alkaline residue (vegetarian) diet prevents bone loss in older individuals.

Water homeostasis is regulated by a high-gain feedback mechanism that involves the hypothalamus, neurohypophysis, and kidneys. ^, In the kidneys, water and sodium from the glomerular filtrate are reabsorbed in tubules through water channel aquaporins (AQPs) and sodium cotransporters. The tubular reabsorption of water depends primarily on the driving force (high interstitial osmolality in the deeper medullary zones) and osmotic equilibrium of water across the tubular epithelia (high osmotic water permeability of the membrane). The large majority of glomerular filtrate is reabsorbed in the proximal tubules and descending thin limbs of the loop of Henle. ^, The next tubular segments (thin and thick ascending limbs and distal convoluted tubules) are relatively water impermeable. ^, Finally, the main parts of the nephron where vasopressin-based regulation of body water homeostasis occurs are the connecting tubule and collecting duct (see also Chapter 14).

Maximum urinary concentration and diluting capacity are decreased with aging, leading to a higher risk for hypernatremia and hyponatremia. Findley has proposed an age-related dysfunction of the hypothalamic-renal axis based on clinical observations of increased vasopressin secretion in older age. With older age, maximal urine-concentrating ability is particularly reduced. Compared with younger individuals, older adults have about a 50% reduced capacity to conserve water and solutes. One of the major changes in body composition with aging is an increase in total fraction of body fat by 5% to 10% and an equivalent decrease in total body water, so an older man on average has 7 to 8 fewer L of total body water than a young man with the same body weight. The obvious consequence is that in the event of acute loss, or overload of body water, a more severe change in serum osmolality will occur in older individuals. A study comparing plasma osmolality in older and younger individuals before and after a similar extent of water deprivation has confirmed this presumption. Alterations in water and sodium balance frequently lead to hyponatremia or hypernatremia associated with hypovolemia or hypervolemia in older adults. ^, In addition, dysfunction in water homeostasis may also occur due to abnormal expression and trafficking of AQPs and solute transporters. For example, an animal study has shown that the AQP2 transporter involved in urine-concentrating ability is downregulated in the medulla of older rats. This molecular mechanism is consistent with a reduced urine-concentrating ability in older adults. Water homeostasis is also influenced by the previously described microstructural changes that occur with the aging kidney. In the event of stress, acute disease, volume load, or dehydration, the combined effects of age-related loss of renal mass (i.e., functional reserve) and the change in body composition may cause a significant disruption in water and solute homeostasis. ^,

One of the consequences of the age-related GFR decline is an increase in filtrate reabsorption in the proximal tubules and decreased fluid delivery to the distal diluting tubules, resulting in a reduced diluting capacity of the kidney. This is evidenced by a reduced ability to excrete a free water load or reduced maximal free water clearance in older adults. In addition to the reduced diluting capacity, older kidneys also lose the capacity to conserve water during states of dehydration. Taken together, therefore, these age-related changes may have significant clinical implications, such as worsening of dehydration in older adults in the setting of vomiting, diarrhea, or reduced water and food intake.

Vasopressin secretion, the renal response to vasopressin, and thirst control are also affected by aging. Vasopressin secretion in the hypothalamus is under the delicate control of osmoreceptors in and around the organum vasculosum of the lamina terminalis and the anterior wall of the third ventricle in the brain. Most, but not all, studies have found that basal vasopressin concentrations in the healthy older are usually higher than in younger controls. ^{, ,} Nevertheless, most studies about water homeostasis in aging have shown that compared with younger individuals, older adults have a greater increase of vasopressin secretion per unit change in plasma osmolality, and this is consistent with the higher osmoreceptor sensitivity in older adults.

Vasopressin regulates renal water excretion by controlling the abundance of the water channel AQP2 and its insertion into the apical membrane of epithelial cells in the distal nephron and collecting tubules. In the membrane, AQP2 proteins form channels that allow reabsorption of water molecules from the lumen of the collecting ducts into the medullary interstitium driven by the medullary osmotic gradient. Given that vasopressin levels are mostly found to be higher in older adults, a potential secretory defect in the pituitary gland is improbable and cannot explain the age-related reduced renal response to vasopressin. Animal studies have offered potential explanations for the reduced renal response to vasopressin, which include reduced expression of vasopressin receptors in the collecting ducts and impaired second messenger response to vasopressin receptor signaling. ^,

Normally, stimulation of thirst osmoreceptors in the hypothalamus signals the higher cerebral cortex to develop a conscious perception of thirst- and water-seeking behavior. Aging also affects thirst control. It has been proposed that older adults may have a higher osmolal set point for thirst—that is, for a given plasma osmolality, older adults have a reduced degree of perceived thirst, thus leading to less water intake.

Thus aging undoubtedly influences water homeostasis in many ways. Water conservation (urinary concentrating ability) is primarily affected, although a mild form of impaired diluting capacity may also be present. This is important to keep in mind during the diagnostic process and while taking care of older adults and planning for different clinical, surgical, or pharmacologic interventions. Alterations in solute intake (e.g., protein or sodium) in older adults can also have profound effects on water homeostasis.

Serum levels of erythropoietin (EPO) tend to increase with age, even in nonanemic subjects. ^, This might be in compensation for increased (subclinical) blood loss, accelerated red blood cell turnover (decreased erythrocyte half-life), or increased resistance of red cell precursors to the effect of EPO. Testosterone deficiency might be involved in the latter phenomenon. Mild anemia (hemoglobin <12 g/dL) is fairly common in older persons but varies by geography. In about one third of persons with anemia, a nutritional deficiency (e.g., iron, folate, vitamin B ₁₂ ) was responsible; in one third, a concomitant chronic disease was the culprit (e.g., CKD, cancer, and infection); and in one third, no precise cause could be determined. Independent of anemia, high spontaneous EPO levels in older adults are associated with an increased risk of congestive heart failure but not myocardial infarction or progressive CKD. The mechanisms responsible for this association are not well understood but could be due to tissue hypoxia in low-perfusion states associated with heart failure.