INTRODUCTION — The aging kidney is a topic of great interest in geriatric medicine and clinical nephrology. In 1999, glomerular filtration rate (GFR)-estimating equations started to replace serum creatinine for the evaluation of kidney function. Since that time, more and more older adults have been identified and labeled as having chronic kidney disease (CKD). Approximately one-half of adults over the age of 70 years have a measured or estimated GFR (eGFR) <60 mL/min/1.73 m2 [1], an advocated threshold to diagnose CKD [2]. Thus CKD, diagnosed by examination of eGFR, is largely a problem of older adults, and populations with higher prevalence of the aged will experience a higher burden of CKD [3].
This higher prevalence of diagnosed CKD is not simply due to increased recognition of diseases that tend to cluster in older adults, such as antineutrophil cytoplasmic antibody (ANCA)-positive small-vessel vasculitis, amyloidosis, diabetic nephropathy, nonamyloid monoclonal immunoglobulin deposition diseases, membranous nephropathy, and chronic tubulointerstitial disorders. Rather, much of the increased rate of CKD diagnoses in the older adult population results from the interpretation that normal structural and functional changes that occur in the kidney with aging are a manifestation of a "chronic disease of the kidneys" [4], an interpretation with which we disagree. Many have argued that this increased recognition of CKD is a positive development and leads to better care in older adult populations [5,6]. Others have argued that this is a harmful development that has led to unnecessary labeling of far too many older patients as "diseased" without any proven clinical benefit [7-11].
This review describes the structural and functional changes in the kidney with normal aging and the clinical significance of these changes. Issues related to estimation of GFR, the definition of CKD, and evaluation of patients with CKD are discussed elsewhere:
●(See "Assessment of kidney function".)
●(See "Definition and staging of chronic kidney disease in adults".)
●(See "Diagnostic approach to adult patients with subacute kidney injury in an outpatient setting".)
NORMAL AGING VERSUS CHRONIC DISEASE — Aging is a natural and inevitable biological process that results in structural and functional changes in many organ systems. The kidney systematically loses function (eg, glomerular filtration rate [GFR]) and undergoes anatomical changes (senescence) with age. In addition to specific kidney diseases that are common in older adults, such as diabetic nephropathy, physiological senescence of the kidney occurs, even with healthy aging [4]. A similar process occurs in the lung, which systematically loses function (ie, forced expiratory volume in one second), even among "healthy" adults [12]. (See "Selecting reference values for pulmonary function tests".)
It can sometimes be difficult to distinguish the structural and functional changes of a kidney affected by a specific preventable or treatable disease from those of a kidney undergoing the inevitable consequences of aging. However, even if the reduction in function is not preventable or treatable, senescent changes in the kidney are relevant and important to managing older patients. Specifically, loss of kidney function reserve with aging has the following clinical significance (see 'Clinical significance of the aging kidney' below):
●More advanced disease if a new specific nephropathy, such as diabetic nephropathy or vasculitis, develops
●Increased susceptibility to acute kidney injury
●Toxic accumulation of medications cleared by the kidney
●The need for age-appropriate criteria in the selection of living kidney donors
Living kidney donors as a resource to examine normal aging — Living kidney donors are a unique and important resource for understanding the structural and functional changes of normal aging. Kidney donors are systematically and meticulously evaluated to confirm "health" or at least "healthy for age" as currently understood. This provides the opportunity to characterize the physiological changes in the kidney with optimal aging. Kidney donors undergo kidney function testing, urinalysis, and contrast computed tomography (CT) angiograms as part of their evaluation. In addition, some transplant centers obtain implantation kidney biopsies (ie, "time zero" kidney biopsies) at the time of surgical donation. Autopsy studies of the general population (identified by sudden deaths) can also characterize structural changes in the kidney with aging, but they lack simultaneous kidney function and other clinical evaluations [13]. By contrast, general population studies of the aging kidney in living humans rely upon nonspecific biomarkers to evaluate the kidney since structural characteristics of the kidney are not typically available.
Many examples of studies from living kidney donors are presented in the ensuing discussions of structural and functional changes with aging. (See 'Structural changes of the kidneys with aging' below and 'Functional changes of the kidneys with aging' below.)
STRUCTURAL CHANGES OF THE KIDNEYS WITH AGING — Various structural changes occur with aging, including microanatomic changes, such as nephrosclerosis and a decline in nephron number, and macroanatomic changes, such as decreased kidney cortical volume and the development of kidney cysts.
Microanatomical changes of the kidneys with aging — The major microanatomical changes that occur with aging are nephrosclerosis and nephron hypertrophy.
Nephrosclerosis — The primary structural finding of the aging kidney on light microscopy is nephrosclerosis, characterized by an increased detection of focal and global glomerulosclerosis, tubular atrophy, interstitial fibrosis, and fibrointimal hyperplasia (arteriosclerosis). In a study of kidney transplant donors, nephrosclerosis, defined as the combination of two or more of glomerulosclerosis (global not segmental), tubular atrophy, interstitial fibrosis, or arteriosclerosis, increased from 2.7 percent in kidneys from 18- to 29-year-olds to 73 percent in kidneys from 70- to 77-year-olds. The difference in global nephrosclerosis comparing two kidneys separated in age by eight years was roughly equivalent to the difference in nephrosclerosis comparing a kidney from a hypertensive donor with a similarly aged nonhypertensive donor [14].
Nephrosclerosis is likely due to an ischemic injury of the nephrons that is thought to result from arteriosclerosis and hyalinosis of small arteries. With ischemic injury, the glomerulus develops wrinkled capillary tufts, thickened basement membranes, and pericapsular fibrosis, eventually leading to collapse of the tuft and a diminutive, globally sclerosed glomerulus with collagen deposition filling Bowman's space [15-18]. Some of these obsolescent glomeruli may subsequently undergo complete absorption [19,20]. The tubule associated with the sclerotic glomerulus also atrophies, and the interstitium surrounding the atrophic tubule undergoes fibrosis [14,21,22]. Glomerulosclerosis, tubular atrophy, and interstitial fibrosis on biopsy of kidney donors correlate with the presence of arteriosclerosis [14].
The reasons why nephrosclerosis occurs with aging are not fully understood. Increased DNA methylation, a biomarker of biological aging, is evident on tissue biopsies from older kidneys. However, DNA methylation does not associate with the nephrosclerosis of aging [23].
Glomerulosclerosis — The focal and global glomerulosclerosis of aging has been described in a variety of populations, including living kidney donors [14,18,20,22,24-26]. There is reduction in podocyte density throughout life that may also cause glomerular tuft collapse and contribute to age-related glomerulosclerosis via a nonischemic pathway [27]. The upper reference limit (95th percentile) for the number of globally sclerotic glomeruli expected on biopsy according to age was determined in 1847 normotensive living kidney donors [28]. In a kidney biopsy showing 20 glomeruli, for example, up to 1 glomerulus would be expected to be globally sclerotic in a 25-year-old, but up to 6 would be expected to be globally sclerotic in a 75-year-old. Among patients with nephrotic syndrome who undergo kidney biopsy and those with other kidney diseases, the risk for progressive chronic kidney disease (CKD) is best identified using these age-specific reference limits for global glomerulosclerosis [29,30].
The global glomerulosclerosis associated with older age occurs almost entirely in the superficial cortex, whereas global glomerulosclerosis associated with lower glomerular filtration rate (GFR), proteinuria, hypertension, interstitial fibrosis, and diabetes mellitus occurs more diffusely or in the deeper cortex [31]. Superficial glomeruli are the most distal glomeruli perfused by the renal artery, and age-related glomerulosclerosis is consistent with an ischemic process from arteriosclerosis [31].
By contrast, a lesion of focal segmental glomerulosclerosis (FSGS) is not observed as a component of normal aging in human kidneys and, when present, should be considered pathological. (See "Focal segmental glomerulosclerosis: Clinical features and diagnosis".)
Nephron number in aging kidneys — The number of functional nephrons with which a person is born (ie, nephron endowment) varies between 700,000 and 1.8 million per kidney, and the number progressively declines with aging due to nephrosclerosis [32-34]. The total number of nephrons (nonsclerosed glomeruli) in living kidney donors decreases from 990,000 per kidney for ages 18 to 29 years to 520,000 per kidney for ages 70 to 75 years [35] or approximately 50 nonsclerosed glomeruli lost per day. This loss of nonsclerosed glomeruli with aging is not simply explained by an increasing fraction of globally sclerosed glomeruli. As glomeruli sclerose with aging, there appears to be eventual absorption of these sclerosed glomeruli, and the total number of glomeruli (sclerosed and nonsclerosed) declines with age [19,20,35].
There is progressive atrophy of the corresponding tubule attached to the sclerosed glomeruli. The atrophic tubules that accumulate with age progressively atrophy and create a scattered pattern with multiple small foci of interstitial fibrosis and tubular atrophy [36]. Over time these small "scars" may gradually be absorbed. This process of atrophy and absorption of age-related nephrosclerosis also explains the small amount of interstitial fibrosis and tubular atrophy evident in the cortex with healthy aging despite the substantial nephron loss since birth [14].
Low nephron endowment associated with low birth weight may accelerate the age-related decline in functional nephrons in adulthood [37-40]. Thus, the evaluation of nephron loss with aging must take into account the number of nephrons present at birth.
Nephron hypertrophy — Nephron size (both glomerulus and tubule) increases in a variety of metabolic risk states, including obesity and diabetes, due to cellular hypertrophy and hyperplasia [41-43]. In parallel with age-related glomerulosclerosis [44,45], hypertrophy of the remaining functional nephrons also occurs [46,47]. This appears to be due primarily to hypertrophy of renal tubules rather than the glomeruli [48]. Nephron (tubular) hypertrophy leads to a decrease in nephron density (glomerular number per mm2 of a histologic section) due to spreading of the space between glomeruli. Glomerular enlargement is seen with aging in the presence of age-related comorbidities [27,49]. However, healthy aging in older kidney donors is not associated with an increase in glomerular volume [47,48]. Consistent with this finding, neither glomerular filtration capacity [33] nor single-nephron GFR [50] increases with healthy aging [48]. Any glomerular hypertrophy observed in older subjects is more likely due to comorbid factors, such as obesity, diabetes, or loss of nephrons from disease. When the supply of nephrons is inadequate (such as from low nephron endowment at birth), even for the reduced metabolic needs of older individuals, this can lead to nephron overload and progressive CKD [51].
Other microanatomical changes — Microdissection of nephrons from autopsied kidneys shows a strong association between aging and frequency of tubular diverticula [26].
Macroanatomical changes of the kidneys with aging — The major macroanatomical changes that occur with aging include a reduction in cortical volume and the appearance of kidney cysts and tumors.
Kidney volume — Radiographic imaging studies have elucidated the age-related changes in the kidney volume. A decline in total kidney volume with aging was noted in clinical populations using ultrasound or computerized tomography (CT) [52-56]. In one study, the rate of age-related decline was significantly faster in patients with underlying atherosclerosis (estimated from carotid intima-media thickness) [55]. In the general population, total kidney volume by magnetic resonance imaging (MRI) declined with older age but not after adjusting for age differences in eGFR and comorbidities [57].
Some studies in living kidney donors failed to detect a decline in kidney volume with age, but these studies had small sample sizes and included few older donors [45,58,59].
In a large study of 1344 potential kidney donors, kidney parenchymal volume was stable before age 50 years but declined after age 50 years [60]. This study also separately measured the kidney cortical and medullary volumes. Cortical volume declined throughout the adult age spectrum in both males and females, but with an accelerated rate of decline after age 50 years [60]. By contrast, medullary volume increased with age in both males and females until age 50 years, after which it declined in females and remained static in males [60]. Renal sinus fat increases with age; this could have masked some of the decline in kidney parenchymal volume with aging in studies that included sinus fat in kidney volume or weight measurements [35,52,53].
Microanatomical connections to kidney volume — The age-related decline in kidney cortical volume is largely due to underlying nephrosclerosis (senescent global glomerulosclerosis and atrophy of tubules are lesions that decrease the volume occupied by nephrons) [48]. Some of the preservation of kidney parenchymal volume with aging can be explained by compensatory hypertrophy of the remaining functional tubules [20,44,46,48,49,61]. Increased medullary volume with aging can be explained by hypertrophy of tubules attached to juxtamedullary glomeruli in compensation for atrophy (glomerulosclerosis and nephrosclerosis) of the more superficial nephrons in the cortex [62]. There appears to be a more accelerated loss of total kidney parenchymal volume after middle age (approximately 50 years) [3,60,63-65]. This may be the point at which hypertrophy of functional nephrons no longer compensates for the volume lost from sclerosed and atrophied nephrons [46]. This compensatory hypertrophy of healthy tubules for age-related nephrosclerosis makes total kidney volume and kidney cortical volume poor markers of age-related nephron loss [35].
Kidney cysts and tumors — Kidney parenchymal cysts become more common, more numerous, and larger with aging [66,67]. Age-related diverticula in renal tubules are the precursor lesion to these kidney cysts [68]. Because of this age-related increase in kidney cysts, the diagnosis of autosomal dominant polycystic kidney disease (ADPKD) with kidney ultrasound imaging requires age-specific criteria (Ravine criteria) [69]. CT scans have higher resolution and can detect smaller cysts not seen by ultrasound. The figure shows the upper limit of normal (97.5th percentile) for the number of cortical and medullary cysts detected by CT according to age and sex (table 1) [67]. Although kidney cysts that are not associated with malignancy or a polycystic disorder are often called "benign" or "simple," they frequently reflect chronic underlying parenchymal injury of the aging kidney and are associated with increased urine albumin excretion. In addition to parenchymal simple kidney cysts, other kidney cysts are also more common with aging, including parapelvic kidney cysts (due to lymphatic dilation near the renal sinus), hyperdense cysts, angiomyolipomas, and malignant cysts or tumors [67].
Other macroanatomical changes — Atherosclerosis of renal arteries (with or without significant stenosis) is more prevalent in older individuals, ranging from 0.4 percent in 18- to 29-year-olds to 25 percent in 60- to 75-year-olds [70]. Fibromuscular dysplasia (FMD) also becomes more prevalent with aging, though it is much more common in females [70]. Compared with the smooth surface of younger kidneys, older kidneys have a more irregular, rough surface consistent with the effects of nephrosclerosis [48]. In addition, parenchymal calcifications and cortical scars (a focal atrophic region of cortex) are more common with aging [70].
FUNCTIONAL CHANGES OF THE KIDNEYS WITH AGING — Glomerular filtration rate (GFR) is the primary tool for assessing age-related functional changes in the kidney and is typically the only tool for examining age-related changes since kidney biopsy and kidney imaging are not typically justified in routine patient care.
GFR declines with normal aging — Rates of glomerular filtration rate (GFR) decline with aging follow a fairly normal (Gaussian) distribution, suggesting that it is primarily driven by a physiological process [71]. Not only does the lower reference limit for GFR decrease with age, but the upper reference limit also decreases with age (table 2) [72-76]. If GFR is preserved in some persons with healthy aging, then the upper limit of normal (95th or 97.5th percentile) among healthy adults should be stable or only slightly decrease with aging. Some studies with GFR measurements among healthy adults across the age spectrum have refuted this hypothesis. As an example, one study found that the 95th percentile for GFR decreased by 38 mL/min/1.73 m2 from age 18 to 70 years [75]. Another study found the 97.5th percentile for GFR decreased by 30 mL/min/1.73 m2 from age 50 to 90 years, and the 2.5th percentile for healthy 50 year olds was similar to the 97th percentile for healthy 95 year olds [76]. Furthermore, population level variability in GFR did not increase with age as would be expected if a subset of older healthy persons had unrecognized kidney disease that contributed to their GFR decline [76]. Because of this universal or near-universal decline in GFR with aging, healthy persons over the age of 70 years are misidentified as having "chronic kidney disease (CKD)" based on a GFR <60 mL/min/1.73 m2 [76].
A longitudinal study of 254 males followed for 23 years found that the average GFR (estimated serially by urinary creatinine clearance) declined by 7.5 mL/min per decade of life [71]. Some of these individuals had diabetes or obesity, which could have impacted the results, and in fact, one-third of the cohort had an increase in their creatinine clearance over time, which likely reflected both hyperfiltration (seen with obesity, early diabetes, and subclinical cardiovascular disease [77]) and imprecision with urinary creatinine clearance measurements. (See "Calculation of the creatinine clearance" and "Mechanisms of glomerular hyperfiltration in diabetes mellitus".)
Some investigators have attributed this GFR decline with normal aging to a high-sodium diet or some other pervasive influences of a "modern" or "Western" lifestyle. However, GFR decline with normal aging is also observed in populations unaffected by Western lifestyle factors. In an isolated population of Kuna Indians in Panama, for example, GFR (measured by inulin clearance) declined by 9.5 mL/min per decade of life, despite not having a Western lifestyle and not having an age-related increase in blood pressure that is typical of a Western lifestyle [78-80]. Thus, the decline in GFR with aging can be disassociated from the effects of increasing blood pressure.
Trajectory of decline in GFR in advanced age — Data on the evolution of GFR in advanced age (over 70 to 75 years) are limited and largely cross-sectional. Individuals of advanced age often have associated comorbidity that further compounds the evaluation of physiologic versus disease-related changes. A comprehensive evaluation of the evolution of eGFR among older adults (mean age 80 years, range 70 to 99 years) was reported from the Berlin Initiative Study [81]. This longitudinal study included community-living subjects over 70 years of age, 91 percent of whom had at least one comorbidity. Diabetes was present in 26 percent, hypertension in 79 percent, heart failure in 29 percent, cancer in 23 percent, and approximately 26 percent had increased urinary albumin excretion; thus, this is not a study of "healthy" aging.
Nevertheless, there was an annual decline in eGFR of -1.67 to -0.99 mL/min/1.73 m2 for males and -1.52 to -0.97 mL/min/1.73 m2 for females. These rates of decline are higher than those seen in adults less than 70 years of age. Of interest, there appeared to be a slowing of the rate of decline with aging in adults over 85 years old. It is possible that at late age an increase in single nephron GFR is a factor in this phenomenon. This study adds to our knowledge concerning the trajectory of eGFR in community-living (but largely unhealthy) older adults. However, it is not known whether this late aging attenuation of eGFR decline is due to GFR or non-GFR determinants, due to the lack of measured GFR in the study.
Mechanistic basis for GFR decline with aging — With age-related loss of nephrons, it is reasonable to ask whether the remaining glomeruli compensate to maintain glomerular filtration rate (GFR) [82]. Single-nephron GFR reflects the average function of remaining nephrons and does not increase but remains stable at 80 nL/min despite the substantial loss of nephrons with healthy aging [50]. This finding is further supported by the stable glomerular volume [48] and glomerular filtration capacity [33] with healthy aging. Single-nephron GFR does increase with factors that reflect increased metabolic demand such as very tall height or obesity [50], but the lack of a compensatory increase in single-nephron GFR for the age-related loss of nephrons suggests a declining metabolic demand for GFR with aging. Indeed, the GFR decline with aging has been linked to both lower urea generation and the lower metabolic rate with aging [83,84].
Nephrosclerosis is associated with an increase in single-nephron GFR but only when nephrosclerosis exceeds that expected for age [50]. Consistent with nephron loss but a stable single-nephron GFR with aging, cortical atrophy [60] and nephrosclerosis [48] with aging appear to follow the same biological pathway as the GFR decline with aging.
High proportion of older adults with reduced GFR — The Kidney Disease Improving Global Outcomes (KDIGO) 2012 guidelines define and stage CKD by levels of glomerular filtration rate (GFR) and albuminuria (table 3). (See "Definition and staging of chronic kidney disease in adults".)
Since nearly all healthy young adults have a GFR ≥90 mL/min/1.73 m2, this GFR range has been defined as "optimal" or "normal." A GFR of 60 to 89 mL/min/1.73 m2 has been defined as "mildly decreased relative to young adult levels." Regardless of age, any GFR ≥60 mL/min/1.73 m2 is not considered to be CKD unless some other criteria of "kidney damage" are met (eg, albuminuria) [3]. A study of 610 community-dwelling adults older than 70 years found that half had "CKD" as determined by a measured GFR <60 mL/min/1.73 m2 [85]. Often, a specific disease process for a GFR reduction to <60 mL/min/1.73 m2 in older adults cannot be identified (other than normal aging). One study found the lower reference limit for measured GFR to be 49 mL/min/1.73 m2 at the age of 75 years among healthy potential kidney donors [86].
Estimated glomerular filtration rate (eGFR) — Direct measurement of GFR is not available in most clinical settings. Instead, GFR is usually estimated from serum creatinine (eGFRcreatinine) using the 2009 or 2021 Chronic Kidney Disease Epidemiology (CKD-EPI) equation [87]. As with measured GFR, approximately one-half of adults over the age of 70 years have CKD as determined by an eGFRcreatinine <60 mL/min/1.73 m2 [1]. Creatinine-based GFR-estimating equations include an age variable to model the decline in creatinine generation (due to decreased muscle mass) with normal aging [88]. This age variable unmasks the age-related decline in GFR that is not detected by serum creatinine alone. Interestingly, the age-related decline in GFR is approximately canceled out by the age-related decline in muscle mass, such that serum creatinine remains stable with healthy aging [89], until late in life (over 70 years) when a small increase may occur [90]. The only way a patient can maintain a stable eGFRcreatinine with aging is to have a progressive decrease in his or her serum creatinine level. However, a declining serum creatinine level can represent loss of muscle mass from a disease process (or treatments that lead to muscle wasting, such as glucocorticoids). A patient with a stable serum creatinine level who "ages into CKD" based upon eGFRcreatinine <60 mL/min/1.73 m2 could have experienced nothing more than the expected age-related decline in GFR. Ideally, patients would lose neither GFR nor muscle mass with aging, and serum creatinine levels would still remain stable with aging [91].
When and how to confirm CKD in older adults — Uncertainty regarding a reduction in GFR when using eGFRcreatinine may exist when other evidence of chronic kidney disease (CKD) such as albuminuria is lacking, or when the muscle mass is more than average for the patient's age. When such uncertainty exists regarding the accuracy of eGFRcreatinine, a urinary creatinine clearance or direct GFR measurement can further assess the kidney function of older adult patients. This may be particularly important when an accurate GFR assessment is needed for a patient who wants to donate a kidney. (See 'Living kidney donor evaluation' below.)
Older patients in good health and with higher-than-average muscle mass may meet the definition of CKD by eGFRcreatinine <60 mL/min/1.73 m2 but may actually have a GFR ≥60 mL/min/1.73 m2 using a confirmatory test. This can be reassuring to some older patients who are concerned about being diagnosed with CKD. However, if an older patient's urinary creatinine clearance or direct GFR measurement is still modestly reduced (eg, 45 to 59 mL/min/1.73 m2), it is still possible that this could be nothing more than a reduced GFR from normal aging. (See 'What to do with isolated mild reductions in eGFR in older adults' below.)
When confirmation of CKD is desired, the KDIGO guidelines suggest using GFR-estimating equations that are based upon cystatin C (eGFRcystatin C) or upon both creatinine and cystatin C (eGFRcreatinine-cystatin C) [92]. Specifically, they suggest these calculations be made in patients with an eGFRcreatinine of 45 to 59 mL/min/1.73 m2 and no other markers of kidney damage [3]. There is improved discrimination of outcomes (end-stage kidney disease [ESKD] and mortality) with eGFRcystatin C or eGFRcreatinine-cystatin C compared with eGFRcreatinine [3]. However, the optimal GFR-estimating equation for predicting such outcomes does not necessarily reflect the true association between GFR and outcomes [93]. Unlike an elevated serum creatinine, elevated cystatin C occurs with other chronic diseases besides CKD. Cystatin C is a protease inhibitor that plays a role in atherosclerosis, obesity, and inflammation [94-96], and serum cystatin C associates with both CKD risk factors and the development of ESKD, independent of the patient's underlying GFR (figure 1) [97-100]. Thus, although patients with elevated serum cystatin C have a higher risk for ESKD, cardiovascular disease, and mortality, this risk may be mediated by factors other than kidney function (ie, GFR). As a result, we do not recommend use of eGFRcystatin C or eGFRcreatinine-cystatin C to confirm a diagnosis of CKD in older patients with eGFRcreatinine of 45 to 59 mL/min/1.73 m2.
Mortality and kidney failure risk with CKD from aging — It is not clearly evident that the relationships between eGFR and death or ESKD support the use of a common threshold for chronic kidney disease (CKD) in all age groups. A meta-analysis by the CKD Prognosis Consortium studied 46 cohorts with 2,051,244 patients and found that a low eGFRcreatinine was associated with mortality and ESKD in all age groups [101]. However, several features of this meta-analysis argue that a separate GFR threshold is needed for older adults:
●Among individuals without albuminuria, an isolated mild reduction in eGFRcreatinine of 45 to 59 mL/min/1.73 m2 (compared with 75 to 89 mL/min/1.73 m2) was associated with a 20 percent higher relative mortality in those 75 years and older, but a 179 percent higher relative mortality in those aged 18 to 54 years. Other studies have not found an increased risk of mortality in older adults with eGFR 45 to 59 mL/min/1.73 m2 in the absence of albuminuria [102,103].
●The lowest risk for mortality in adults aged 18 to 54 years was with an eGFRcreatinine ≥105 mL/min/1.73 m2. However, in adults aged 75 years and older, those with an eGFRcreatinine in this range had a 60 percent higher relative mortality. If the separation between "normal" and "disease" is based upon risk, as suggested by KDIGO, then age-specific ranges are justified (figure 2).
●In the KDIGO guidelines, an eGFRcreatinine ≥90 mL/min/1.73 m2 is used to define "optimal" or "normal" [3]. However, the meta-analysis performed by the CKD Prognosis Consortium to evaluate risks associated with isolated mild reductions in eGFR did not use this "normal" range as the reference group, thereby limiting the use of these data.
This meta-analysis was also not limited to representative samples of the general population as many high-risk cohorts were included [101]. Furthermore, serum creatinine levels were not consistently standardized, and the eGFR categories were based upon a single serum creatinine level rather than serum creatinine levels over at least three months to establish chronicity [104]. These limitations were addressed in two subsequent reports:
●A nationwide study in Iceland studied 218,437 adults and used the full KDIGO definition of CKD that required at least two outpatient determinations of eGFRcreatinine obtained three months apart with standardized serum creatinine levels [104]. By these criteria, younger adults (ages 18 to 45 years) with an eGFRcreatinine of 60 to 75 mL/min/1.73 m2 and no proteinuria or specific kidney disease had an increased risk of mortality (standardized mortality ratio 3.0, 95% CI 1.1-5.9) despite not meeting criteria for CKD by an eGFRcreatinine <60 mL/min/1.73 m2. Alternatively, older adults (age >65 years) with an eGFRcreatinine of 45 to 59 mL/min/1.73 m2 and no proteinuria or specific kidney disease did not have an increased risk of mortality (standardized mortality ratio 1.01, 95% CI 0.97-1.05) despite meeting criteria for CKD by an eGFRcreatinine <60 mL/min/1.73 m2.
●A population-based study of 127,132 adults in Alberta, Canada required sustained reductions in eGFRcreatinine for at least three months to define CKD [105]. Among older persons (age >65 years) with normal or mild albuminuria, the five-year risk of kidney failure with an eGFRcreatinine of 45 to 59 mL/min/1.73 m2 was very low (<0.12 percent). The five-year risk of mortality with eGFRcreatinine of 45 to 59 mL/min/1.73 m2 was also consistent in magnitude with that of persons with an eGFRcreatinine of 60 to 89 mL/min/1.73 m2. There was no significantly increased risk of mortality with eGFRcreatinine of 45 to 59 mL/min/1.73 m2 for ages 75 to 79 years (mortality hazard ratio [HR] 1.0, 95% CI 1.0-1.1) or for ages 80 years and older (mortality HR 0.9, 95% CI 0.8-0.9).
Other functional changes — The kidneys of older patients demonstrate impaired kidney conservation of sodium in response to an acute reduction of sodium intake, as well as impaired ability to rapidly excrete a large sodium load. Animal studies show a decreased abundance of sodium chloride transporters and epithelial sodium channels with aging [106]. There is also loss of both the maximum urine concentrating ability (from approximately 1200 to 800 mosm/L) and the maximum urine diluting capacity (from approximately 50 to 100 mosm/L) [107-110]. The expression of aquaporins is reduced in the medulla with aging in animal models [111]. The reduced diluting capacity of the kidney increases the risk of hyponatremia in older patients, particularly those with a low-protein "tea and toast" diet that is consumed by some older adult patients. Changes in the diluting capacity also contributes to thiazide-associated hyponatremia, which is more common in older adults [112]. There is also a tendency toward hyperkalemia in older adults from hyporeninemic hypoaldosteronism [113].
●(See "Causes of hypotonic hyponatremia in adults", section on 'Low dietary solute intake'.)
●(See "Causes of hypotonic hyponatremia in adults", section on 'Diuretic-induced hyponatremia'.)
CLINICAL SIGNIFICANCE OF THE AGING KIDNEY — There are several important issues in the management of patients with age-related changes in kidney structure and function. These include:
●There are no proven therapies to halt or reverse age-related declines in glomerular filtration rate (GFR). Any therapy aimed at raising the GFR by causing the remaining functional nephrons to filter more may actually be harmful rather than beneficial to the kidney, by inducing pathologic hyperfiltration per nephron. Hypertrophy and hyperfiltration of functional glomeruli in animals, for example, can lead to further glomerulosclerosis, probably mediated by podocyte stress or injury. (See "Secondary factors and progression of chronic kidney disease", section on 'Intraglomerular hypertension and glomerular hypertrophy'.)
This is a key difference distinguishing age-related "chronic kidney disease (CKD)" from other age-related chronic disease such as hypertension or hyperlipidemia. Age-related hypertension and hyperlipidemia can be reversed, and reducing the blood pressure and cholesterol has clinical benefit in older adult patients. By contrast, generic therapies for CKD (such as angiotensin inhibitors and protein-restricted diets) do not reverse age-related GFR decline, and it is not clear that these therapies are of any benefit to an older adult with only age-related "CKD." A simulation study suggests that angiotensin inhibitors will have marginal benefit at most for age-related "CKD" [114].
●Patients with age-related reductions in GFR may require medication dose adjustments, for water-soluble drugs cleared mainly by GFR. Lipid-soluble drugs and those metabolized to nontoxic end products do not need dose adjustment in the presence of low GFR. (See 'Medication dosing' below.)
●The loss of functioning nephrons and resultant lower GFR associated with normal aging put older adults at higher risk for acute kidney injury. (See 'Acute kidney injury' below.)
●A reduction in estimated GFR (eGFR) with normal aging (sometimes to the range of 45 to 59 mL/min/1.73 m2 does not substantially improve the assessment of cardiovascular risk over conventional predictors and does not indicate a high risk of end-stage kidney disease (ESKD) at least when abnormal albuminuria is not present. (See 'Cardiovascular morbidity and mortality risk' below and 'End-stage kidney disease' below.)
●Age-related declines in GFR do not typically prevent older adults from becoming living kidney donors. (See 'Living kidney donor evaluation' below.)
●Overall, the typical age-related declines in GFR (between 45 and 59 mL/min/1.73 m2) have little, if any, effect on life expectancy, and this point is important in discussions with older adult patients. (See 'Counseling patients with CKD due to normal age-related GFR declines' below.)
What to do with isolated mild reductions in eGFR in older adults — There is controversy about whether or not to diagnose older adults with CKD based upon an eGFRcreatinine of 45 to 59 mL/min/1.73 m2 in the absence other evidence of CKD (such as albuminuria). An eGFR of 45 to 59 mL/min/1.73 m2 has a modest though statistically significant association with adverse events such as death and ESKD in older adults, leading many experts and the Kidney Disease Improving Global Outcomes (KDIGO) group to endorse an eGFR <60 mL/min/1.73 m2 for diagnosing CKD regardless of age. (See "Definition and staging of chronic kidney disease in adults".)
By contrast, many authorities caution against labeling older patients with a stable and normal serum creatinine as having CKD solely based upon an eGFRcreatinine between 45 and 59 mL/min/1.73 m2. The major reasons for this caution are as follows:
●Estimation of GFR is imprecise and can be inaccurate. A true decrease in GFR in someone with CKD, for example, can be partially cancelled out by a parallel decrease in creatinine generation [97]. In addition, the age variable that is included in the estimating equations can inflate the adverse risks attributed to reduced GFR, particularly in older adults [88]. Many problems associated with estimating GFR using estimation equations apply both to younger and older populations [74].
●Because of a decline in GFR with healthy aging, the association of a given eGFR with mortality and ESKD is not identical across the age spectrum (figure 2). The relative risk of mortality is substantially less in older compared with younger patients with an eGFR of 45 to 59 mL/min/1.73 m2. The development of ESKD is much rarer event in older compared with younger patients with an eGFR of 30 to 59 mL/min/1.73 m2 and no proteinuria [115]. This issue is discussed above in detail. (See 'Mortality and kidney failure risk with CKD from aging' above.)
●The underlying cause of an age-related decline in eGFR is loss of nephrons from global glomerulosclerosis. While glomerulosclerosis exceeding that expected for age is an important predictor of progressive CKD, age-related global glomerulosclerosis does not appear to contribute to progressive CKD [29].
There is concern for both over-diagnosis of CKD in older patients with an eGFR below 60 mL/min/1.73 m2 and under-diagnosis of CKD in younger patients with eGFR above 60 mL/min/1.73 m2. Thus, some have suggested using the upper reference limits for serum creatinine (1.26 mg/dL in males and 1.05 mg/dL in females [111.6 and 93.0 micromol/L, respectively]) to identify those patients who have CKD resulting from GFR reductions that exceed the expected age-related decline. These reference limits are derived using healthy adults in the general population and in healthy kidney donors [116,117]. The upper reference limits for serum creatinine may be slightly higher in African American individuals and slightly lower in Asian individuals, although few studies have assessed reference limits in these populations [117].
Some have also suggested reporting age-specific lower reference limits for eGFRcreatinine to prevent the overdiagnosis of CKD in older adults and the underdiagnosis of CKD in young adults [118,119]. These lower reference limits can be calculated from the upper reference limits for serum creatinine (table 2) [72,73]; however, few studies have assessed reference limits for serum creatinine in older adult populations (>70 years) [90]. In an older adult patient with an eGFR of 45 to 59 mL/min/1.73 m2 and no albuminuria, we advocate caution in diagnosing CKD (rather than an age-related GFR decline). Approximately one-half of older patients will have an eGFR <60 mL/min/1.73 m2, and the relative risks of ESKD and mortality associated with lower eGFR values are substantially less than in younger patients (figure 2) [4]. A consortium of several investigators has called for an age-adapted approach to the definition of CKD, as present systems of categorizing CKD do not take into account the age-related decline in GFR with normal, healthy aging [7].
Medication dosing — Dosing with medications that are excreted by glomerular filtration (typically water-soluble medications) should be adjusted to the patient's GFR as indicated by medication dosing guidelines (typically provided by the manufacturer). Historically, most medication dosing guidelines have been developed using the Cockcroft-Gault equation (calculator 1 and calculator 2) [120]. The Cockcroft-Gault equation is not as accurate as the Modification of Diet in Renal Disease (MDRD) study or Chronic Kidney Disease Epidemiology (CKD-EPI) equations for estimating GFR. If GFR is estimated by the MDRD study or CKD-EPI equations, the result must be converted to mL/min by multiplying by the patient's body surface area (calculator 3) and dividing by 1.73. Issues related to estimating GFR as it pertains to drug dosing are discussed elsewhere in detail. (See "Assessment of kidney function", section on 'Glomerular filtration rate for drug dosing'.)
Acute kidney injury — Loss of kidney function reserves (functioning nephrons) makes older patients more susceptible to acute kidney injury [121,122]. Medications that have potential kidney toxicity, such as nonsteroidal antiinflammatory drugs (NSAIDs) and radiocontrast agents, should be used cautiously in older patients. (See "NSAIDs: Acute kidney injury", section on 'Risk factors'.)
Cardiovascular morbidity and mortality risk — Age is already included in models that estimate cardiovascular risk. (See "Cardiovascular disease risk assessment for primary prevention: Risk calculators".)
Addition of eGFRcreatinine to cardiovascular risk models does not meaningfully improve risk prediction over conventional risk stratification models [123-125]. In addition, the inclusion of cystatin C with conventional factors in cardiovascular risk assessment only minimally improves prediction over conventional risk factors alone [126]. Cystatin C might contribute to the prediction of cardiovascular outcomes along non-GFR biological pathways, such as inflammation and atherosclerosis [94-96]. The issue of CKD as a risk factor for cardiovascular disease is presented in detail elsewhere. (See "Chronic kidney disease and coronary heart disease" and "Chronic kidney disease and coronary heart disease", section on 'CKD as a CHD risk equivalent'.)
End-stage kidney disease — CKD from aging alone does not, by itself, progress to kidney failure, even in very old persons [127]. A validated calculator has been developed to estimate the risk of kidney failure [128,129]. Like most prior kidney failure risk models, both age and eGFRcreatinine were included in the prediction model. While lower eGFRcreatinine is associated with increased risk of kidney failure, older age is actually associated with a lower risk of kidney failure. This occurs because, at the same eGFRcreatinine, older patients as compared with younger patients are more likely to die before developing ESKD [130]. Kidney failure prediction models can help inform decisions about the need to prepare for dialysis in older adult patients (eg, placement of an arteriovenous fistula). These models can also reassure older patients with only age-related reductions in eGFRcreatinine that their risk of progression to ESKD is very low. As an example, a 70-year-old female who has been diagnosed with "CKD" based on an isolated eGFR of 50 mL/min/1.73 m2 will have <1 percent risk of ESKD in the next five years.
Living kidney donor evaluation — There are substantial changes in the microanatomy (glomerulosclerosis), macroanatomy (cortical volume loss), and kidney function (GFR decline) with older age, even among healthy adults. Even a kidney donor in his or her thirties will have more nephrosclerosis on kidney biopsy than a kidney donor in his or her twenties [14]. Living donors have been reported to have similar long-term mortality and kidney failure risk compared with age-matched controls who do not donate a kidney, despite the fact that a postdonation eGFR <60 mL/min/1.73 m2 occurs in more than 60 percent of donors older than 50 years [131-133]. In particular, a study of 219 living donors aged ≥70 years found a lower risk of mortality compared with age- and comorbidity-matched controls [133]. Another study of 3368 older living donors (mean age 59 years) found no increased risk of death compared with older healthy matched controls [134]. Conversely, a study of 1901 living donors in Norway found an increased long-term risk for ESKD and mortality with kidney donation, possibly because the age-matched controls in that study were healthier than age-matched controls in prior studies [135]. Whether this risk they detected differs between older and younger donors was not reported.
Many transplant programs only require the predonation GFR to be appropriate for the donor's age (table 2) and permit certain CKD risk factors (such as mild hypertension) to be present in older donors without rejecting them as candidates [136]. Another approach is to use a calculator to estimate the risk of ESKD in potential donors; older donors have a much lower risk of ESKD than younger donors at the same eGFR [137]. From the recipient's perspective, an older allograft (donated kidney) is associated with somewhat higher rates of graft loss [133], and this may be related to the increased nephrosclerosis present with aging with a resultant lower nephron number in the allograft (see 'Structural changes of the kidneys with aging' above). The use of an age-calibrated measured GFR may improve the living donor-selection process [138,139].
Counseling patients with CKD due to normal age-related GFR declines — In older adults who are diagnosed with age-related "chronic kidney disease (CKD)," a careful explanation is warranted to reassure the patient that a decline in GFR within the range expected for normal aging will have little, if any, effect on his or her life expectancy. Although they may be at higher risk for ESKD because of lower functional reserve (fewer functioning nephrons), this is still a relatively rare event, particularly in older adults with an eGFR of 45 to 59 mL/min/1.73 m2 and normal albuminuria (CKD stage G3a/a1 in the KDIGO classification system) [140]. Context is particularly relevant, as there are no known therapies that can reverse the age-related decline in GFR. It may also be helpful to discuss that senescence occurs in many other organs such as the skin and the lungs. (See "Overview of benign lesions of the skin" and "Selecting reference values for pulmonary function tests".)
ALBUMINURIA — Although urine albumin excretion increases with age in the general population [141], 24-hour urine albumin excretion does not inevitably increase with age in healthy adults, provided few or no comorbidities are present [14]. Thus, albuminuria may be a useful marker for distinguishing normal aging from another etiology of chronic kidney disease (CKD).
However, the random first morning urine albumin-to-creatinine ratio (UACR), which is more widely used in clinical practice than a 24-hour urine collection, is more likely to give a falsely elevated result in an aged population. Unlike the 24-hour urine albumin excretion, UACR usually increases with healthy aging due in part to age-related reductions in muscle mass that result in reduced creatinine generation and excretion [88,142,143]. If a random or morning UACR is used to assess albuminuria in older adults, multiplying the UACR by the expected 24-hour urine creatinine (according to age, sex, body weight, and race) may improve the accuracy of the test for detecting an abnormally high albumin excretion. The expected 24-hour urine creatinine can be calculated from one of several formulas [144,145].
SUMMARY AND RECOMMENDATIONS
●Overview – In 1999, glomerular filtration rate (GFR)-estimating equations started to replace serum creatinine for the evaluation of kidney function. Since that time, more and more older adults have been identified as having acute or chronic kidney disease (CKD), and the prevalence of diagnosed kidney disease in this population has increased. Approximately one-half of adults over the age of 70 years now have a measured or estimated GFR (eGFR) <60 mL/min/1.73 m2, a threshold often used to diagnose CKD. This higher prevalence of diagnosed CKD is not simply due to increased recognition of diseases that tend to cluster in older adults. Rather, much of the increased rate of CKD diagnoses in the older adult population results from the normal structural and functional changes that occur in the kidney with aging. (See 'Introduction' above.)
●Normal aging versus chronic disease – It can sometimes be difficult to distinguish the structural and functional changes of a kidney affected by a specific preventable or treatable disease from those of a kidney undergoing the inevitable consequences of aging. However, even if the reduction in function is not preventable or treatable, senescent changes in the kidney are relevant and important to managing older patients. Specifically, loss of kidney function reserves with aging has the following clinical significance (see 'Normal aging versus chronic disease' above):
•More advanced disease if a new specific nephropathy, such as diabetic nephropathy or vasculitis, develops
•Increased susceptibility to acute kidney injury
•Toxic accumulation of medications cleared by the kidney
•The need for age-appropriate criteria in the selection of living kidney donors
●Structural changes with aging – Various structural changes occur with aging, including (see 'Structural changes of the kidneys with aging' above):
•Nephrosclerosis (see 'Nephrosclerosis' above)
•Decline in nephron number (see 'Nephron number in aging kidneys' above)
•Decreased kidney cortical volume (see 'Kidney volume' above)
•Development of kidney cysts (see 'Kidney cysts and tumors' above)
•Glomerular volume and single-nephron GFR remain stable (see 'Mechanistic basis for GFR decline with aging' above)
●Functional changes with aging – Rates of GFR decline with aging follow a fairly normal (Gaussian) distribution, suggesting that it is primarily driven by a physiological process. Not only does the lower reference limit for GFR decrease with age, but the upper reference limit also decreases with age, consistent with a universal or near-universal decline in kidney function (table 2). (See 'GFR declines with normal aging' above.)
The Kidney Disease Improving Global Outcomes (KDIGO) 2012 guidelines define and stage CKD by levels of GFR and albuminuria (table 3). Approximately half of adults older than 70 years are found to have "CKD" as determined by a measured or estimated GFR <60 mL/min/1.73 m2. Often, a specific disease process for a GFR reduction to <60 mL/min/1.73 m2 in older adults cannot be identified (other than normal aging). (See 'High proportion of older adults with reduced GFR' above.)
●Confirming CKD in older adults – Uncertainty regarding the diagnosis of CKD in older adults may exist, particularly when the patient's eGFRcreatinine is 45 to 59 mL/min/1.73 m2, when other evidence of CKD such as albuminuria is lacking, or when the muscle mass is higher than average for the patient's age. When such uncertainty exists regarding the diagnosis of CKD based upon eGFRcreatinine, a urinary creatinine clearance or direct GFR measurement can further assess the kidney function of older adult patients. This may be particularly important when an accurate GFR assessment is needed for an older patient who wants to donate a kidney. In contrast to KDIGO suggestions, we do not recommend use of cystatin C to confirm a diagnosis of CKD in older adults. (See 'When and how to confirm CKD in older adults' above.)
●Clinical significance of the aging kidney – There are several important issues in the management of patients with age-related changes in kidney structure and function. These include (see 'Clinical significance of the aging kidney' above):
•There are no proven therapies to halt or reverse age-related declines in GFR. Any therapy aimed at raising the GFR by causing the remaining functional nephrons to filter more may actually be harmful rather than beneficial to the kidney. Hypertrophy and hyperfiltration of functional glomeruli in animals, for example, can lead to further glomerulosclerosis. (See "Secondary factors and progression of chronic kidney disease" and "Secondary factors and progression of chronic kidney disease", section on 'Intraglomerular hypertension and glomerular hypertrophy'.)
•Patients with age-related reductions in GFR require medication dose adjustments. (See 'Medication dosing' above.)
•The loss of functioning nephrons associated with normal aging puts older adults at higher risk for acute kidney injury. (See 'Acute kidney injury' above.)
•A mild reduction in eGFR with aging does not substantially improve the assessment of cardiovascular risk over conventional predictors and does not indicate a high risk of end-stage kidney disease (ESKD) (figure 2). (See 'Cardiovascular morbidity and mortality risk' above and 'End-stage kidney disease' above.)
•Age-related declines in GFR do not typically prevent older adults from becoming living kidney donors. (See 'Living kidney donor evaluation' above.)
•Overall, the typical age-related declines in GFR have little, if any, effect on life expectancy or the future need for dialysis, and this point is important in discussions with older adult patients. (See 'Counseling patients with CKD due to normal age-related GFR declines' above.)
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟