Introduction

Kidney anatomy and function change with increasing age. A large cohort trial noted the decline in creatinine clearance of 1 mL/min/year beginning in the fifth decade of life.[1] These changes may begin to occur at the age of 40, if not before.[2] Pathologically, glomerulosclerosis of cortical glomeruli is a common finding in elderly subjects.[3] Clinically, elderly subjects develop a number of alterations in renal function related to change in renal blood flow, loss of glomeruli and changes in tubular function. Such changes in renal function include a reduced ability to raise glomerular filtration rate (GFR) during times of stress and a reduction in renal dilution and concentration abilities.[4–7] These changes have the potential to increase the risk of acute kidney injury and electrolyte imbalances as well as lead to long-term abnormalities in kidney function. To what extent long-term abnormalities in kidney function represent normal aging or a disease state is an area of debate.

Chronic kidney disease

Chronic kidney disease (CKD) is a five-tier disease state based on abnormalities in kidney function and/or other urinary abnormalities such as microscopic hematuria or proteinuria for three or more months. An equation designed from the Modification of Diet in Renal Disease Study (MDRD) uses age, sex, gender, race, and serum creatinine to estimate GFR.[8] CKD affects 10%–13% of the adult US population.[9] A large number of elderly subjects fall into stage 3 CKD because of the impact of their age, and this may label them with an inappropriate disease state.[10, 11] The level of proteinuria becomes very important in determining prognosis. Finally, muscle mass declines in older age and creatinine would be expected to decrease. Practically speaking then, healthcare providers need to recognize how different levels of kidney function can be associated with a creatinine of 1.2 mg/dL. Table 25.1 illustrates this point.

Most CKD patients do not progress to end-stage renal disease (ESRD) because of intrinsic CKD stability as well as the high rate of cardiovascular and noncardiovascular mortality in this population.[12–15] Proposals exist to more clearly distinguish elderly subjects with stage 3 CKD who are more likely to have progressive disease from those whose renal dysfunction appears less progressive and more age related.[11] Consequently, a growing body of literature supports the use of stage 3 A and stage 3B to define those patients who have an eGFR of 45–60 and 30–44 mL/min, respectively. In the absence of proteinuria or other signs of inflammatory kidney disease, it is not expected that patients with stage 3 A disease will progress.[16] The CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) formula and the MDRD equation both estimate GFR, but the CKD-EPI formula classifies fewer individuals as having kidney disease.[17] This formula may be more specific for identifying renal dysfunction in the geriatric population, although studies are ongoing.

A broad variety of disease states can lead to chronic kidney disease. As much as possible, delineating the cause of disease is important in order to predict which patients are at higher risk of progression. Table 25.2 lists causes of abnormal kidney function commonly seen in elderly patients and the associated urinary findings seen for each. Other diseases that patients have experienced in the past may result in abnormal kidney function. One classic example is reflux nephropathy as a child manifesting as subtle renal dysfunction in an adult. The pathologic finding that differentiates advancing kidney disease from previous renal injury is glomerulosclerosis. In examples such as this, particularly with limited proteinuria, a biopsy is usually not pursued if it is not felt to alter therapy. As a general rule, kidney disease progression is less likely to occur in diseases without inflammation (hematuria) or significant glomerular disease (proteinuria).

Determining the cause of CKD in elderly subjects is no different than nonelderly subjects and involves a thorough medical history, physical examination, and evaluation of the urine. Important considerations include a history of frequent urinary infections (particularly as a child) suggestive of reflux nephropathy, symptoms and signs of a connective tissue disease or vasculitis, use of NSAIDs (which may lead to temporary renal dysfunction due to reduced blood flow or chronic interstitial nephritis), or suggestions of significant peripheral vascular disease. The latter may be suggestive of atherosclerotic renal vascular disease. As noted in Table 25.2, a UA without blood or protein rules out most glomerular disease except those related to myeloma. For this reason, all elderly subjects with unexplained kidney disease should have at least one spot urine protein (not albumin) to urine creatinine ratio in order to exclude proteinuria. Spot urine protein/creatinine ratios provide enough information to assess proteinuria levels and, as a result, a 24-hour urine collection is no longer required. However, the urine protein to creatinine ratio cannot be assessed with a traditional office dipstick for protein, as the office dipstick only reacts to urinary albumin.

Despite the high prevalence of kidney disease in the United States, only a small number of patients progress to ESRD. This relates to the stability of most cases of CKD as well as the large burden of cardiovascular disease in this population.[18] Additional complications related to kidney disease include anemia, metabolic acidosis, hypertension, and CKD mineral-bone disorder (MBD) manifested as hyperparathyroidism and hyperphosphatemia. Prospective well-designed studies support the treatment of metabolic acidosis and hypertension, although the traditional goal blood pressure of 130/80 mm Hg has been replaced with 140/90 mm Hg.[19, 20] The treatment of anemia and CKD mineral bone disorder is based on guidelines with less prospective based evidence.[21, 22] A workshop recently reviewed the challenges to treating elderly subjects with CKD and areas in which focus should occur.[23] Major challenges identified by this group included best methods to differentiate age-related declines in kidney function from more aggressive disease states, the lack of prospective studies in the large population of elderly patients with CKD, and unique situation of elderly patients with CKD. For example, elderly CKD patients with less aggressive kidney disease, as shown by a lack of proteinuria, may not benefit from earlier referral to a nephrologist as would a younger patient.[24]

Acute kidney injury

Acute kidney injury (AKI) commonly occurs in elderly patients. Although the same differential diagnosis applies to patients of all ages, elderly patients are more prone to AKI due to several characteristics of the aging kidney. As mentioned earlier, anatomic and physiologic changes in the kidneys of elderly patients can have important consequences. Anatomically, kidney size diminishes with age.[25] This change is associated with microscopic findings that show decreased glomeruli number, glomerulosclerosis, tubulointerstitial fibrosis, and intimal thickening of arteries and arterioles.[26–28] With aging, muscle mass decreases. A creatinine in the normal reference range may be inappropriately high and could be indicative of AKI. Age-associated renal atrophy results in diminished capability to withstand pathologic insults and increased likelihood of developing AKI.

In addition, renal blood flow can decrease by 50% by the age of 80.[29] This decrease may be due, in part, to a reduction in prostaglandin and dopamine levels.[30] Nitric oxide (NO), a vasodilator that balances decreases of renal perfusion may be reduced in older adults.[31] In animal models of AKI, aged rats appear to be more sensitive to lowering of NO levels.[32] Overall, mechanisms needed to maintain renal blood flow to prevent renal injury may be significantly reduced with aging.[33] Additionally, NSAIDs, which are commonly administered to elderly patients, may further counteract these vasodilatory intrinsic compensatory processes.[34]

The algorithm for diagnosis of AKI begins by differentiating between prerenal, renal, and postrenal causes. Renal, or intrinsic renal causes, are divided in to acute tubular necrosis (ATN), AIN, glomerulopathies, vasculitis, and intratubular obstruction that can be seen, for instance, in uric acid nephropathy. In the inpatient setting, prerenal reasons can cause up to 30% of cases of AKI.[35, 36] Factors complicating AKI in the elderly patient include decrease in thirst sensation, impairment of urinary concentrating ability and concomitant diuretic usage.[37–39] Physiologic changes associated with the aging kidney may contribute to the development of AKI. For instance, (1) sodium reabsorption in the loop of Henle decreases with aging and (2) plasma renin and aldosterone levels also decrease.[33] This latter change causes less collecting tubule sodium reabsorption. With sodium loss from the body, hypovolemia may develop and could lead to AKI. Consistent with this pathophysiologic development, Macias et al. showed that when placed on low-sodium diets, elderly patients have higher urinary sodium levels compared to younger patients.[40]

A more common cause of AKI in the hospitalized setting is acute tubular necrosis (ATN), which occurs in up to 48% of cases in patients 65 years old and greater.[36] Sepsis is the most frequent cause of ATN.[41] ATN can also be caused by contrast dye and medications. In particular, NSAIDs are frequently recognized as the source of ATN.[34] Of note, contrast nephropathy can occur in 17% of elderly patients administered contrast dye during cardiac catheterization as compared to in 9.1% of the general population receiving contrast dye during cardiac catheterization.[42, 43] ATN due to contrast nephropathy is more common in the setting of CKD, diabetic nephropathy, hypovolemia, and use of ace inhibitors or angiotensin receptor blockers.[44, 45] In general, a less common cause of AKI is rapidly progressive glomerulonephritis. However, it occurs most in those 65 and older. The incidence is 60 cases per million in patients 65–74 years old.[46] Advancing age is not a contraindication for immunosuppression therapy.[47–50] Other less common causes of intrinsic renal disease to be considered as causes of AKI are interstitial nephritis and other glomerulopathies.

Lastly, from an intrinsic renal disease perspective, atherosclerotic vascular renal disease occurs in 42% of elderly patients.[51] Atheroembolic disease as a cause of AKI should be thought of in cases of aortic procedures, angiography, thrombolysis, and anticoagulation.[52–55] The remainder of AKI cases is due to post-renal causes. Post-renal causes include intrinsic ureteral obstruction (i.e., calculi, blood clots), extrinsic ureteral obstruction (i.e., malignancy), bladder outlet obstruction (i.e., prostatic hypertrophy), and urethral obstruction (i.e., stricture).[33]

As part of the AKI evaluation, urine indices including a random urine sodium are often obtained. Owing to increases in urinary sodium, in early stages of prerenal azotemia due to hypovolemia, urine sodium may be higher than the classical threshold of 20 meq/L.[33] Renal biopsy can be beneficial when intrinsic renal disease is suspected and the specific cause remains uncertain. Findings may alter management as histologic findings may indicate a disease different from the one suspected based on clinical grounds;[56] renal biopsy may directly influence treatment.[57]

Consistent with the anatomic and physiologic issues mentioned earlier, age has been independently associated with AKI.[58] After developing AKI, renal recovery has been reported in 43%–67% of patients.[41] The need for chronic dialysis has ranged from 2.7%–6% [59]. In another study, dialysis dependence was studied in 425 patients who needed hemodialysis after developing AKI due to ATN. Fifty-three percent of patients survived, of which 57% had complete recovery of renal function, 33% had creatinine levels >1.3 and <3 mg/dL, and 10% continued to have creatinine levels >3 mg/dL.[60] Although degree of illness, history of chronic kidney disease, and cause of CKD can affect the occurrence of renal recovery in the setting of AKI, the study of age as a prognostic factor has produced contradictory results.[41, 61–63] In terms of mortality, age has not been shown to be an associated factor in elderly patients.[62]

The preceding studies detail AKI in elderly inpatients. Feest et al. reported AKI risk in the outpatient setting.[64] In a large, community-based study, the incidence of AKI rose substantially with age. The annual incidence of AKI was 17 per million in adults <50 years old and increased with age to a yearly incidence of 949 per million in adults 80–89 years old. Medical causes included hypovolemia (34% of cases), and second most common was prostatic obstruction (25% of cases).

Hypertension in elderly subjects

A number of mechanisms lead to increased blood pressure in elderly subjects. First, systolic blood pressure increases with age, independent of CKD status, due to stiffening of vessels over time. A consequence of this is the frequent finding of isolated systolic hypertension and a large pulse pressure. In addition, CKD leads to elevations in sympathetic tone and hormones of the renin-angiotensin activating system. The major clinical consequence is increased vasoconstriction and increased sodium absorption in the kidney. Key medication classes directed at these dysfunctional physiologic systems include diuretics, ace-inhibitors, aldosterone antagonists, and beta blockers. Although beta blockers reduce renin levels, they are also associated with significant fatigue and, in the absence of clear cardiac indications, should not be first-line agents for blood pressure control.[20]

Recent blood pressure guidelines suggest that in the absence of chronic disease, blood pressure treatment targets in subjects over the age of 60 should be <150/90 mm Hg. This is largely based on the HYVET trial; importantly, nearly 30% of subjects in a subset of this trial had white coat HTN.[65] Consequently, a low threshold should be used to investigate for white coat hypertension by home BP monitoring or ambulatory blood pressure monitoring. Additional factors to consider in the elderly population include the potential need to reduce BP medications over time due to changes in diet and weight loss with advancing age, the need to monitor for orthostatic hypotension regularly, and the challenge of treating patients with isolated high systolic blood pressures.[66] Although patients with isolated high systolic blood pressures are at high risk for vascular events, lowering DBP to less than 55 mg Hg may lead to symptoms of cerebral, renal, or peripheral underperfusion.[67]

Fluid and electrolyte disorders associated with the aging kidney

Hyponatremia has been reported to occur in 20% of elderly patients, compared to 5% in all patients.[68] Hyponatremia risk increases with age.[69, 70] The syndrome of inappropriate antidiuretic hormone (SIADH) was found to be the etiology in 73.6% of normovolemic elderly patients.[71] Elderly patients treated with thiazide diuretics, bupropion, and SSRIs are more prone to become hyponatremic.[72–74] Consequences of hyponatremia may lead to gait disturbance, secondary falls and fracture risk.[75, 76] Intracellular sodium in bone can serve as a sodium reservoir and can be released in hyponatremia; resultant osteoporosis may further increase risk for fractures.[77] Also, regression analysis revealed that anemia is independently associated with hyponatremia and consideration should be given to monitoring hemoglobin levels in patients with this electrolyte abnormality.[78]

On the other side of the spectrum, hypernatremia also increases with age, with the most common cause due to bacterial infections.[79, 80] Decreased thirst response in elderly patients is a contributing factor. Azotemia may further exacerbate hypernatremia, as blood urea nitrogen (BUN) may act as an osmotic agent to increase free water urinary losses. Assuring a patient’s access to water and/or administering free water may be necessary to avoid hypernatremia in clinical situations when dehydration can occur.

Hyperkalemia in the elderly is exacerbated by medication use. Risk of hyperkalemia increases in patients treated with ace inhibitors. Furthermore, aldosterone antagonist use is another cause. This risk was studied in patients already on an ace inhibitor for treatment of congestive heart failure. In patients greater than 75 years of age, with a mean creatinine clearance of 39 mL/min, the addition of this medication class was associated with a 36% incidence of a potassium level greater than 5.5 meq/L and 11% incidence of a potassium level greater than 6 meq/L.[81] Furthermore, hypokalemia can also frequently occur with diuretic use.[82] In elderly patients with a mean age of 68 years, the hazard ratio for all-cause mortality in patients with hypokalemia, compared to normokalemia, was 1.56 (CI 1.25–1.95, p <0.001).[83]

End-stage renal disease

Most patients with ESRD in the United States receive dialysis in the form of in-center hemodialysis, while peritoneal dialysis is used far less frequently. Nonetheless, if patients can follow and comply with necessary instructions, home renal replacement is the preferred modality either in the form of peritoneal dialysis or home-hemodialysis. This view is driven by improved health outcomes in patients on home modalities and potential cost savings. However, such outcomes have not been demonstrated in older populations.

Older adults comprise a significant percentage of those on dialysis, as the mean age for all patients initiating dialysis in 2011 was 62.7 years.[84] The ideal time to initiate dialysis is not clear. One randomized trial, involving over 800 patients with a mean age of 60, compared initiation of dialysis at a creatinine clearance of 5–7 mL/min to a creatinine clearance of 10–14 mL/min, as determined by the Cockcroft Gault equation. This trial found that late initiation of dialysis did not lead to worse outcomes.[85] However, most patients in the late-initiation group actually started dialysis earlier than planned due to the development of ESRD-related symptoms. Ultimately, this study suggests that dialysis initiation should be based on an assessment of the patients’ symptoms, not only their creatinine or eGFR.

In frail elderly patients, dialysis initiation may lead to a downward spiral in health quality and unnecessary pain and suffering.[86] Defining expectations and goals of dialysis before initiation is critical in the frail elderly patient with advanced CKD. A model predicting mortality in the six-month period following dialysis initiation, based on a large group of elderly French patients, has been prospectively validated in a similar population and may be useful in discussion of imminent mortality.[87] Interestingly, increasing age was not associated with a worse outcome in this model though other known CV risk factors were. Guidelines from the nephrology community exist in regards to the importance of shared decision making about dialysis initiation and, at times, withdrawal.[88] Generally, this cites substantial neurologic injury without hopeful recovery as an example of when dialysis would be inappropriate or when withdrawal from dialysis should be considered.

Access for dialysis is essential in any patient with ESRD; all subjects with advanced, progressive CKD who may be candidates for dialysis should discuss preferred dialysis modalities with their nephrology care providers. Nonetheless, the elderly represent a population where the commonly prescribed “fistula first” philosophy of obtaining dialysis access prior to dialysis initiation may not be ideal for all patients, especially those with high surgical risk.[89, 90] In this setting, tunneled dialysis catheters may have a role despite their known increased risk of infection. It is now recognized that brachial and subclavian accesses should be avoided in any patients who may subsequently need an arteriovenous fistula or graft (AVF or AVG). Venous stenosis has been reported to occur in upward of 30% of patients receiving a PICC line.[91]

Age over 65 years is not a contraindication to transplantation, and experience is growing with transplantation in the elderly population. Between 2002 and 2012, the percentage of patients on the kidney transplant wait list between the age of 65 and 74 increased from 11.1% to 18.1%.[92] From a practical perspective, elderly dialysis patients who have a potential living related donor are much more likely to complete the transplant process before associated comorbid conditions disqualify them from transplantation.