Chapter 1: Geriatric Nephrology Has Come of Age: at Last
Chapter 1: Geriatric Nephrology Has Come of Age: at Last
Chapter 1: Geriatric Nephrology Has Come of Age: at Last
At Last
Dimitrios G. Oreopoulos* and Jocelyn Wiggins†
*Department of Medicine, University of Toronto, Toronto, Canada; and †University of Michigan, Ann Arbor,
Michigan
On May 23–24, 1985, the first International Sym- ease (MDRD) formula has revealed a large number
posium on Geriatric Nephrology was held in To- of patients who have impaired kidney function,
ronto. In his excellent review of the symposium,1 most of whom are elderly. Primary care physicians
Michael Kay commented on the challenges and are inundated with elderly patients with impaired
risks facing aged individuals and, in view of the in- kidney function and, in turn, are flooding the neph-
creased costs of their care, the ethical issues facing rologists with referrals.
the caregiver. His report stressed the fact that “the Nephrologists have had to take a serious look at
degree of humanity in our healthcare world will be the plight of elderly, and probably as a result, three
made evident in the way we treat (or do not treat) important developments have ushered in the new
our minorities, our underprivileged, our poor, our era of Geriatric Nephrology.
mentally infirm, those who have no voice to speak
1. For the first time, the American Society of Nephrology has in-
for themselves, and finally, the aged.” cluded in its annual Renal Week program a 2-d course on geri-
After that successful initial meeting, the Interna- atric nephrology that was sold out and that kept the interest of
tional Society of Geriatric Nephrology was formed, the participants to the end. All these presentations have been
with its own journal—the International Journal of taped and are available, with the accompanying slides, at http://
Geriatric Nephrology and Urology—and five addi- asn-online.org/education_and_meetings/media/geriatrics/. A
similar course has been planned for the 2009 ASN meeting.
tional international meetings of the International
Society were held at Salamanca, Lisbon, Atlanta, 2. Recognizing that geriatric nephrology is now essential to ne-
Thessaloniki, and Antalya. phrology training, the Accreditation Council for Graduate Med-
Despite all these efforts and activities, the interest ical Education (ACGME), in its program requirements for train-
among nephrologists concerning geriatric nephrology ing in nephrology, has mandated that “fellows must have formal
instruction, clinical experience and demonstrate competence in
did not increase and, if anything, was decreasing. the prevention, evaluation and management of geriatric aspects
Membership in the Society and participation to the of nephrology, including disorders of ageing kidney and urinary
meetings were small. Also subscriptions to and sub- tract.” In addition, the ACGME has mandated that “fellows must
mission of articles to the Society’s journal were not receive formal instruction in geriatric medicine, including phys-
sufficient to sustain it. As a result, the publisher de- iology and pathology of the ageing kidney, and drug dosing and
renal toxicity in the elderly patient.”
cided to publish the journal as a section in the journal
International Urology and Nephrology. In response to the above, the Chair of the ASN
All these regrettable circumstances seem to have Training Program Directors Committee (Dr. Donald
changed because of two important factors that have Kohan) invited a group of individuals to form a com-
contributed to a renewed interest in geriatric ne- mittee to design a curriculum by identifying the topics
phrology. First is the amazing increase in the inci- and authors to write the corresponding chapters and
dence of new patients with ESRD over the age of 65; we were honored to be asked to co-chair it. We were
this segment of our population is the fastest grow- impressed by the enthusiasm of all members of the
ing group of patients requiring dialysis, and it con- committee (Table 1). The committee identified 37
tinues to grow. Thus, nephrologists forced to prac-
tice as amateur geriatricians now recognize the need Correspondence: Dimitrios G. Oreopoulos, University Health
to master all aspects of geriatrics. Second, the intro- Network, 399 Bathurst Street, 8E-408, Toronto, Ontario M5T 2S8,
duction and automatic reporting of estimated GFR Canada. E-mail: [email protected]
(eGFR) using the Modification of Diet in Renal Dis- Copyright 䊚 2009 by the American Society of Nephrology
In 2005, ACGME (Accreditation Council for Grad- tients with chronic kidney disease (CKD) usually
uate Medical Education) issued the following state- require complicated medication routines, complex
ment with respect to nephrology fellowship train- dietary restrictions, and frequent medical visits.
ing: Many patients in this age group have lost the ability
“Fellows must have formal instruction, clinical to administer their own pills, to buy and cook their
experience and demonstrate competence in the own groceries, or to drive themselves to office visits
prevention, evaluation and management of geriat- or dialysis units. Many patients, particularly in the
ric aspects of nephrology, including disorders of the diabetic population, have difficulties with basic
aging kidney and urinary tract.” In addition, the mobility. It is essential that the nephrologist be fa-
ACGME mandated that “fellows must receive for- miliar with and able to perform routine functional
mal instruction in geriatric medicine, including assessments of their older patients. This includes
physiology and pathology of the aging kidney; and evaluating cognitive, affective, functional, social,
drug dosing and renal toxicity in the elderly pa- economic, and environmental status. This enables
tient.” them to customize a regimen or direct the patient to
The following curriculum is an attempt to fulfill a living environment where such supportive care is
this mandate and prepare the next generation of available. In a busy practice, some aspects of this
nephrologists for the comprehensive care of the assessment can be allocated to other providers such
older population with kidney disease. as social workers, nurses, nurse practitioners, or
The over 65 population in the United States is physician assistants. The management of those el-
rapidly growing. During the next 20 yr, it is ex- derly who require chronic dialysis is even more
pected to double (Figure 1). This means that during complex. They frequently have more difficulty with
the professional lives of current fellows, they can vascular access; they have more cardiovascular dis-
expect to see an increasing number of older patients ease that leads to arrhythmias and hypotension
in their practice. Average life expectancy in 2004 while on dialysis. Furthermore, traveling to and
was 75.2 yr for men and 80.4 yr for women; by 2015, from the unit is a greater burden to them. Further
it is expected to be 76.2 and 82.2 yr, respectively, improvements in assisted dialysis at home will allow
and to continue growing. During the 1990s, the them to enjoy the benefits of treatment at home and
over 85-yr-old population was the fastest growing is an area that needs further exploration.
group at 38% growth. This older age group is the Patients in this older age group are likely to have
largest consumer of healthcare services. In 2005, multiple comorbidities. The average 75 yr old suf-
only 5% of the over 75-yr population had no health fers from 3.5 chronic diseases.1 Many symptoms in
visits, whereas fully 30% of those with 10 or more older patients are caused by multiple deficits and
visits were in this age group, although they consti- not by a single disease. These diseases and their
tute ⬍10% of the population.1 treatments are likely to interact and complicate one
Why should the aging of the population impact another. Murray2 has reported that up to 70% of
nephrologists? There are currently about 35 million dialysis patients 55 yr of age and older have chronic
people over 65 yr of age in the United States. Forty
percent of this population has some level of disabil- Correspondence: Jocelyn Wiggins, University of Michigan, 1150
ity: sensory, physical, mental, or self-care.1 Once a W. Medical Center Drive, 1560 MSRB II, Ann Arbor, MI 48109.
senior develops disability, it greatly impacts on their E-mail: [email protected]
ability to follow a complex medical regimen. Pa- Copyright 䊚 2009 by the American Society of Nephrology
50 39.4
40 34.4 34.7 Malnutrition/weight loss
30 25.7 Urinary incontinence
16.7
20 9 Balance/gait impairment/falls
10 3.1 4.9
0 Polypharmacy
1900 1920 1940 1960 1980 1999 2000 2010 2020 2030 Cognitive impairment
Year Affective disorders
Functional limitations
Figure 1. Projected population over age 65 yr from the US Lack of social support
Census Bureau. Economic hardship
Home environment/safety
cognitive impairment of a level severe enough to impact on their
compliance and ability to make informed decisions.3,4 Prevalence
of depression is reported to be as high as 45% in the older dialysis were significantly more likely to be older (median, 63.2 versus
population.5–7 Metabolic bone disease is complicated by age-re- 59.3 yr; P ⬍ 0.05), because serum creatinine is a poor predictor
lated osteoporosis. The cardiovascular consequences of CKD are of renal function in the elderly. GFR declines with age in nor-
complicated by structural heart disease such as valvular insuffi- mal individuals; therefore, it can be difficult to distinguish age-
ciency and atrial fibrillation. Neurodegenerative disease impacts related decrease in GFR from CKD in the elderly. Older pa-
on the patient’s mobility and cognitive function. Osteoarthritis tients with mild decreased GFR and low risk for progressive
and neuropathy limit their physical activity. As age and disease decline in GFR need to be distinguished from those with pro-
advance, frailty becomes an issue. All of these things combine to gressive disease, because once identified, they probably do not
make their care much more complex than that of a younger pa- need to be followed by a nephrologist.
tient. Drug interactions and inappropriate dosing becomes an in- In conclusion, older patients will make up a growing pro-
creasing issue as the number of comorbidities and medications portion of the nephrologist’s practice. Thus, nephrologists
increases (Table 1). need to become comfortable with shouldering the full care of
In 1992, Nespor and Holley8 did a small study of in-center this segment of their patient population or work closely with a
hemodialysis patients in Pittsburgh. Eighty percent of these geriatrician and family physicians. As we become more willing
patients did not have a family physician and relied on their to offer life-prolonging technologies in the older age groups,
nephrologist for all of their medical care. Ninety-one percent we need to be willing to deal with the consequences of this
sought treatment from their nephrologist for minor acute ill- decision. Finally, with their elderly patients, nephrologists face
ness. Nephrologists were also providing ongoing treatment for challenging ethical problems, such as whether to withhold or
comorbid chronic illnesses such as diabetes and heart disease. withdraw dialysis. Unless addressed promptly and effectively,
In 1993, they went on to confirm similar statistics in their these ethical issues will greatly increase the stress on both the
chronic peritoneal dialysis patients.9 This would suggest that healthcare provider and family members.
the nephrologist needs to be prepared to take on the full com-
plexity of care for their older patients, particularly their
chronic dialysis population. In older patients, this would in- TAKE HOME POINTS
clude health maintenance screening and immunizations. Al- • A knowledge of geriatric medicine is required by ACGME for training in
though malignancies are more common in both the dialysis nephrology
population and in the posttransplantation population than in • The population over 65 yr of age will double in the next 20 yr
the general population, life expectancy, age, and cost effective- • This older population will bring their problems with them to the neph-
rologists
ness need to be considered by the nephrologists before order- • Dialysis patients rely on their nephrologists for most or all of their care
ing screening tests.
Patients with possible CKD are being referred to nephrolo-
gists in greater numbers since the introduction of formulae for
DISCLOSURES
estimating GFR. Most clinical laboratories supply an eGFR None.
when a serum creatinine is ordered. NHANES data estimates
that approximately 11% of the US population has CKD, and
this may be as high as 30% in the older population.10 A recent REFERENCES
Australian study showed that monthly referrals overall in-
creased by 40% after the introduction of eGFR reporting, and *Key References
this was most marked for the tertiary renal service (52% above 1. US Census Bureau: www.census.gov/
baseline).11 Patients referred after the introduction of eGFR 2. Murray AM: Cognitive impairment in the aging dialysis and chronic
In chapter 1, data were presented about the growth hard to get an accurate measure of the extent of
of the older US population and their special needs. the increase.5,6 Coresh et al.7 have estimated that
In this chapter, we will discuss the epidemiology of the overall prevalence of CKD has increased from
the elderly patients with chronic kidney disease 10 to 13% of the US adult population since 1988.
(CKD) and end-stage kidney disease (ESKD). Because 50% of patients in the United States start
dialysis with no previous nephrology care, over-
referral is probably preferable to underreferral.8
CHRONIC KIDNEY DISEASE Another contentious area of debate is the decline
of renal function with age. It is generally accepted
There is much debate in the literature about that renal function declines about 1 ml/min per
whether the incidence of CKD is actually increas- year after the fourth decade of life, even in the ab-
ing or whether we are measuring changes in the sence of comorbidities such as diabetes and hyper-
way we detect and define CKD. In 1998, a report tension. Many nephrologists regard this as “normal
using NHANES data estimated the prevalence of aging” and do not feel that this constitutes a reason
CKD as about 11% in the US adult population.1 for referral. It is certainly true that only a very small
This estimate was based on routine creatinine fraction of these older patients will progress to end
measurements in a subset of the study popula- stage or die from renal failure. It is likely that they
tion. Since that data were published, several de- are at greater risk from vascular disease because
velopments have altered the way we define CKD. there is a very robust association between decline in
The Modification of Diet in Renal Disease GFR and vascular deaths.9
(MDRD) formula was developed and validated.2 So, which older patients should the busy neph-
The National Kidney Foundation created a panel rologists follow and which should be returned to
of experts, who redefined how CKD was classified the care of their primary care physician after an ini-
and staged.3 There has been a change in way cre- tial evaluation? Clearly signs of ongoing active renal
atinine is measured. Automated reporting of disease such as an active urine sediment or signifi-
eGFR based on MDRD formula was initiated in cant proteinuria are reason for a nephrologist’s
most clinical laboratories across the country. care. eGFR values between 45 and 59 ml/min per
These changes resulted in an apparent “pan- 1.73 m2 in those 70 yr of age and older should be
demic” of CKD. A study from an academic de- interpreted with caution. If other signs of kidney
partment in Australia tracked the level of ne- damage (e.g., proteinuria, hematuria) are not
phrology referrals after the implementation of present, a stable eGFR in this range may be consis-
automated eGFR reporting.4 General referrals in- tent with typical GFR for this age and an absence of
creased by 40%, whereas referrals to the tertiary CKD-related complications. Patients showing the
renal service were 52% above baseline. The pa- complications of decreased renal function such as
tients newly referred were significantly older: anemia, phosphorous retention, and hyperkalemia
63.2 versus 59.3 yr. However, the quality of the need nephrology management.
referrals declined with as many as 35% being in-
appropriate. It seems likely with the increases in Correspondence: Jocelyn Wiggins, University of Michigan, 1150
the incidence of diabetes, vascular disease, and W. Medical Center Drive, 1560 MSRB II, Ann Arbor, MI 48109.
the general aging of the population that the true E-mail: [email protected]
prevalence of CKD has increased, but it is very Copyright 䊚 2009 by the American Society of Nephrology
1400
Rates per million pop
1200
1000
800
600
400
200
END-STAGE KIDNEY DISEASE dialysis. The incident rates have been rising steadily over the
last 25 yr (Figure 2), with a narrowing gap between rates in the
Because Medicare mandates that all US dialysis units receiving 70- to 79-yr-old age group compared with the 80⫹-yr age
Medicare compensation report clinical data to the US Renal Data group. These data reflect numbers of patients who have sur-
System (USRDS), information on patients reaching ESKD is vived at least 90 d on dialysis and do not include those who get
much more robust than the data on CKD. The data for these acute dialysis in the hospital and do not progress to chronic
patients were all taken from the USRDS 2007 annual report.10 maintenance dialysis because of recovery or death.
Data collected by USRDS show that ESKD is a disease of the Rates of morbidity and mortality are higher in the ESKD
older population, with numbers starting to rise significantly population than in the general Medicare population. Hospital
after the age of 50. Mean age at the start of renal replacement admission rates are particularly high in the oldest patients,
therapy is 62.3 yr for men and 63.4 yr for women. Peak incident with cardiovascular disease being by far the most common
counts of treated ESKD occur in the 70- to 79-yr age group at cause for hospitalization. Patient admission rates increase lin-
⬎15,000 patients per year. Peak incident rates of treated ESKD early with age. A 20-yr-old patient with ESKD spends an aver-
occur in the 70- to 79-yr-old age group at 1543 per million age of 9 d per patient year in the hospital compared with 15.5 d
population (Figure 1). for patients over 70 yr of age. Older patients also carry signifi-
This probably reflects both a real increase in the rates of cant burden of disability. Overall, 10% carry a diagnosis of
patients reaching ESKD and an increase in the willingness to dementia, and this rises to 21% among those over 80 yr of age.
offer dialysis, regardless of age or comorbidity. The data show This is almost certainly an underestimate because dementia is
a drop off after 79 yr of age. This probably reflects the tendency often undiagnosed. As many as 20% of older patients with
of older patients, with significant burden of disease refusing ESKD have had a stroke that limits their mobility. These co-
1800
Rate per million population
1600
1400
1200
1000 80+
800 70-79
600
400
Figure 2. Incident rates of treated ESKD per
200 million population from 1980 to 2005. Blue
0 bars represent patients over 80 yr of age. Red
1980 1985 1990 1995 2000 2005 bars represent the 70- to 79-yr-old age
group. Data from the USRDS annual report
Year for 2007.
120
Probability of survival
100
80
1 yr
60
2 yr
40
20
Figure 3. Survival probabilities for patients
on dialysis: year 1 in blue and year 2 in red. 0
Data shown for patients across the life span.
+
4
9
9
9
9
9
4
9
9
10 9
14
All age groups show lower survival on dialysis
0-
80
-1
-2
-3
-4
-5
-6
-6
-7
o
5t
to
15
20
30
40
50
60
65
70
than age-matched controls with normal renal
function. Data from USRDS annual report
2007. Age in years
morbidities impact significantly on a patient’s ability to man- • Rates of treated ESRK are increasing in all older age groups
• ESKD carries a poor prognosis
age a complex medical regimen.
In addition to the significant burden of comorbidity, all-
cause mortality is six times higher in the ESKD population DISCLOSURES
than in the general Medicare population. When discussing di- None.
alysis, patients and families need to understand that, although
renal replacement therapy does prolong life, life expectancy is REFERENCES
very limited in the older population. Average 1-yr survival for *Key References
a 70 to 79 yr old is 70%, and for an 80 yr old is 60%. By 2 yr, 1. Jones CA, McQuillan GM, Kusek JW, Eberhardt MS, Herman WH,
survival drops to 52.7 and 39.7%, respectively (Figure 3). Coresh J, Salive M, Jones CP, Agodoa LY: Serum creatinine levels in
CKD and ESKD are huge financial burdens to our medical the US population: Third National Health and Nutrition Examination
system. In 2005, Medicare costs for CKD were $42 billion and for Survey. Am J Kidney Dis 32: 992–999, 1998*
2. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D: A more
ESKD were $20 billion. The cost of ESKD was one half that of accurate method to estimate glomerular filtration rate from serum
CKD, although only a very small percentage of patients with CKD creatinine: a new prediction equation. Modification of Diet in Renal
progress to ESKD. According to NHANES data, about 11% of the Disease Study Group. Ann Intern Med 130: 461– 470, 1999
US population has CKD, whereas ⬍0.2% of the US population 3. Anonymous: K/DOQI clinical practice guidelines for chronic kidney
has ESKD. Despite this low prevalence, ESKD was responsible for disease: evaluation, classification, and stratification. Kidney Disease
Quality Initiative. Am J Kidney Disease 39: S1–S246, 2002
6.4% of the entire Medicare budget. The annual per person cost 4. Noble E, Johnson DW, Gray N, Hollett P, Hawley CM, Campbell SB, Mudge
for dialysis alone exceeded $65,000 in 2005. If all medical care is DW, Isbel NM: The impact of automated eGFR reporting and education on
included, this figure is even higher. For the 70- to 79-yr-old age nephrology service referrals. Nephrol Dial Transplant 23: 3845–3850, 2008
group, the per person annual cost of dialysis is more than $69,000 5. Glassock RJ, Winearls C: The Global Burden of Kidney Disease: How
and in the 80⫹-yr group is more than $74,000. valid are the estimates? Nephron 110: c39 – c47, 2008
6. Coresh J, Stevens LA, Levey AS: Chronic kidney disease is common:
In conclusion, CKD and ESKD are diseases of the elderly. The what do we do next? Nephrol Dial Transplant 23: 1122–1125, 2008
incidence and prevalence of these conditions are rising, especially in 7. Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, Van
the older age groups. Progressing to ESKD carries a significant bur- Lente F, Levey AS: Prevalence of chronic kidney disease in the US
den of comorbidities and clearly shortens life expectancy. Treating during 1988 –1994 and 1999 –2004. JAMA 298: 2038 –2047, 2007*
patientsforESKDisevenmoreexpensiveintheolderagegroupsthan 8. Obrador GT, Ruthazer R, Arora P, Kausz AT, Pereira BJ: Prevalence of
and factors associated with suboptimal care before initiation of dialysis
for younger patients. Preventing progression of CKD should be an in the United States. J Am Soc Nephrol 10: 1793–1800, 1999
urgent priority for every nephrologist, even in the oldest patients. 9. Weiner DE, Tighiouart H, Elsayed EF, Griffith JL, Salem DN, Levey AS,
Sarnak MJ: Inflammation and cardiovascular events in individuals with
and without chronic kidney disease. Kidney Int. 73: 1406 –1412, 2008
TAKE HOME POINTS 10. US Renal Data System: USRDS 2007 Annual Data Report: Atlas of
• eGFR declines with age but does not necessarily indicate clinically Chronic Kidney Disease and End-Stage Renal Disease in the United
significant CKD States. Bethesda, MD, National Institutes of Health, National Institute
• Peak incidence of treated ESKD is in the 70- to 79-yr age group of Diabetes and Digestive and Kidney Diseases, 2007*
In the United States, the elderly and the very elderly juxta-arcuate glomeruli until age 2. At this time, the
population has largely exceeded that of any other size of all of the glomeruli are the same, and the
age group.1 By 1994, the population of these demo- kidney is functioning at adult capacity.17The num-
graphic groups reached 36.5 million and has con- ber of glomeruli among individuals is quite vari-
tinued to increase over the last decade.1 This growth able, ranging from 247,652 to 1,825,380 per kidney,
parallels the increasing number of elderly persons and decreases with age14,18 at a rate of approxi-
classified with chronic kidney disease (CKD) stages mately 6752 glomeruli/yr after the age of 18.14
III through V. Moreover, an estimated 660,000 per- Renal mass increases from 50 g at birth to ⬎400
sons in the United States will have end-stage kidney g during the third and fourth decades of life before
disease (ESKD) by the year 2010, with the greatest decreasing to ⬍300 g by the ninth decade.5,13,14,18,19
growth rate occurring in the elderly and very elderly The latter decrease correlates with the loss of the
persons.1–3 Unfortunately, understanding of the renal cortex. Radiographically, the size of the kid-
normal biologic progression of renal disease in the ney has been shown to decrease in size by 10%after
elderly persons in the absence of comorbid fac- age 40 to 30% by age 80.20 –23 Using the Xenon
tors4 –9 or the progression of CKD is still not clearly washout technique, Hollenberg et al.22 noted that a
understood.10 decrease in the size of the kidney correlated with a
Cross-sectional and longitudinal studies have decrease in function and in the renal blood flow to
looked at the natural progression of the kidney with the cortex.22
aging.4,7,8,11 A linear relationship between aging and
a decline in the renal function was noted,7,8 but el-
derly persons who had no underlying disease had HISTOLOGY
adequate renal reserve.12–14 The Baltimore Longitu-
dinal study (BLS) from 1958 until 1981 studied a The histologic changes with aging observed in hu-
cohort of individuals for 8 or more yr who had a mans have been obtained from information from
least five 24-h urine collections for creatinine clear- autopsies or nephrectomies14,19,23–26 or studies in-
ance.7,8 There were three groups: group 1, CKD; volving laboratory animals.27,28 With aging, there
group 2, on anti-hypertension medications; group are certain universal findings in the cortex, medulla,
3, healthy patients. The overall rate of decline in and, in most cases, in the interstitium and vessels
creatinine clearance was 0.87 ml/min per year be- (Table 1; Figure 1). These histologic changes corre-
ginning at age 40 and was inversely related to age.7,8 late with functional changes observed with aging,
A rise in mean arterial pressure ⬎107 mmHg was including an inability to concentrate or dilute the
positively correlated with a decline in renal func- urine, an increased propensity toward salt reten-
tion.7 Interestingly, in the BLS, one third of the el- tion, dehydration, and acute kidney injury.
derly population had no decrease in renal function
as measured by creatinine clearance, and a small Glomerulus
segment actually had improvement in their renal With aging, hyaline expansion within the mesan-
function.8 gium results in the obliteration of the glomerular
In humans and some animals,14,15 the number of
glomeruli present in adulthood are predetermined
Correspondence: Lynn Schlanger, Assistant Professor, Emory
between weeks 32 and 36 of gestation,14,16 whereas University and Veterans Affairs Medical Center at Atlanta, Atlanta,
the number of glomeruli continue to increase in GA 30033. Phone: 404-727-2525; Fax: 404-727-3425; E-mail:
rats and mice after gestation.15 In humans, the su- [email protected]
perficial cortex glomeruli differ in size from the Copyright 䊚 2009 by the American Society of Nephrology
loops28,29 and is associated with capillary tuft collapse, intra- ular weight proteins.37 These findings differ in the human
capsular fibrosis, and proteinuria.14 The sclerosis in the glo- GBM, where a decrease in hydroxylysine, 4- hydroxyproline,
meruli is primarily in the superficial cortex with sparse changes and glycosylation of collagen occurs with aging.26 The reason
in medulla.14,16,19 Cortical atrophy and loss of the renal paren- for these differences is not clear.
chyma result.18 One hundred forty-six cadaveric kidneys from
medical examiner offices and autopsies from hospital patients Tubulointerstitium
showed an increase in cortical glomerulosclerosis with age With aging, tubular dilation, intratubular cast formation,
from 5% at age 40 to 10% by the eighth decade.30The degree of thickening and splitting of the basement membrane, and fibro-
sclerosis was found to correlate with the degree of atheroscle- sis of the interstitium occurs.18,35,38 In 24-mo-old rats, scans of
rosis, suggesting a hemodynamic role in the aging process.31 the interstitium showed cellular infiltrates consisting predom-
The remaining glomeruli are enlarged to compensate for the inantly of macrophages and lymphocytes and an increase in
decrease in number of functioning cortex glomeruli.12,18,19,32,33 intracellular adhesion molecule (ICAM)-1, osteopontin, and
Electron scans showed podocyte injury with features that in- collagen IV. Areas were marked by an increase of apoptosis.
cluding hypertrophy, intracellular uptake of protein/absorp- None of these findings were detected in the 3-mo-old pups.38
tive droplets, foot process fusion, and detachment of the podo- After the administration of enalapril to 15-d-old CF1 mice, a
cytes from the glomerular basement membranes (GBMs).29,34 decrease in the peritubular and interstitial sclerosis occurred
by 18 mo of age compared with the control mice or mice
Glomerular Basement Membrane treated with nifedipine. A decrease in expression of SM-actin, a
The GBM increases in width with age.27,29,35 In Sprague-Daw- cytoskeleton protein commonly found in fibrosis and repair,
ley rats, the GBM increased in size from 1300 Å at birth to 4800 was also noted in the enalapril-treated group.38
Å at 24 mo.29 In humans, the basement membrane increases
until age 40, and after age 60, the surface area decreases with Vessels
wrinkling of the basement membrane with deposition of hya- An early angiographic study showed changes in the arteriole-
line.29 The composition of the basement membrane also glomerulus unit with aging.24 In the arterioles, hyaline deposi-
changes with aging.36,37 In older rats, the amino acid composi- tion within the vessels walls leads to obliteration of the lumen
tion shifts to a more collagen-like material marked by an in- and is associated with sclerotic glomeruli primarily in the cor-
crease in hydroxylysine, hydroxyproline, and glycine, and tex.22,24,31 Two structural types associated with the afferent and
more insoluble amino acids with higher content of low molec- efferent arterioles have been described.24 In the first case, oblit-
Lumen obliteration
Hyaline deposits in arterioles
RFB
Fas
HSP47
GH/IGF-1
TIMP-1
HSP47 +apoptosis
Androgens Klotho gene
pINK4a MMP
p53 Aging -sclerosis Estradiol
ROS Kidney Ace inhibitor
PAI-1 +sclerosis ET- inhibitor
COX-1
TGFβ -apoptosis
Restricted calories
SMP-30 gene
Klotho gene
Figure 2. Aging kidney. A schematic outline of the various modulators that may be responsible for damage or reno-protection of the
aging kidney. The restricted caloric intake has a reno-protective effect through modulating various proteins by suppressing GH/IGF-1
activity, Fas, and HSP47. The upregulation of MMP, downregulation of TIMP-1 and ICAM-1, and decrease in oxidative stress results in
a decrease in matrix dysregulation and inflammation. The SMP-30 and klotho genes are anti-apoptotic. The klotho gene seems to be
reno-protective by decreasing sclerosis. The inhibition of ET-1 and angiotensin II are known to decrease sclerosis. Areas of fibrosis are
found to have an increase in PA1–1, COX-1, TGF-, PINK4a, and p53. The hormones estradiol and androgen have opposite effects on
aging.35,40,50 –59
AGE AND RATE OF DECLINE OF RENAL change in eGFR seems to be somewhat complex and
FUNCTION perhaps dependent on baseline level of eGFR.
Among a national cohort of veterans with an eGFR
In cross-sectional studies, levels of renal function
⬍60 ml/min per 1.73 m2, eGFR declined more rapidly
are on average lower in older compared with
for older than for younger patients at higher levels of
younger participants.1–3 However, the extent to
eGFR (i.e., ⱖ45 ml/min per 1.73 m2). However, the
which this phenomenon results from an age-asso-
opposite was true at lower levels of eGFR (i.e., ⬍45
ciated decline in renal function versus a higher prev-
ml/min per 1.73 m2), where eGFR declined more
alence of comorbidities linked to chronic kidney
slowly in older than in younger patients.9 Collectively,
disease (CKD) in the elderly is uncertain. Relatively
these data seem to suggest that, although older pa-
few studies have explicitly examined rates of decline
tients are more likely to develop CKD, those who sur-
in renal function across age groups. Most of what
vive long enough to reach more advanced stages of
we know about longitudinal changes in renal func-
CKD are actually less likely than their younger coun-
tion comes from the Baltimore Longitudinal Study
terparts to experience progressive loss of eGFR.
of Aging (BLSA).4 – 6 A subset of participants in this
study underwent serial creatinine clearance mea-
surements over time. Observations on these pa-
AGE AND RISK OF PROGRESSION TO END-
tients have provided some important insights into
STAGE RENAL DISEASE
the effect of age on change in level of renal function.
First, even in individuals without known comorbid
Studies of rate of change in measured or estimated
conditions and without intrinsic renal disease or
renal function can be difficult to interpret for a va-
proteinuria, level of creatinine clearance declined
riety of reasons: (1) progression may not occur in a
on average by 0.75 ml/min per year.4 Second, renal
predictable and linear fashion; (2) the clinical sig-
function was stable and even improved in some
nificance of changes renal function, particularly
subjects.4 Hemmelgarn et al.7 reported a similar
within the normal range, is uncertain; and (3) it can
phenomenon among community-dwelling elderly
be difficult to account for differences in survival
in Canada followed over a 2-yr period (Figure 1).
and follow-up among participants. Thus, results of
Thus, these reports suggest that, on average, renal
studies reporting change in level of renal function as
function declines with increasing age even in the
an outcome are probably quite sensitive to the an-
absence of comorbidity. At the same time, decline
alytic approach selected. Progression to end-stage
in renal function does not seem to be an inevitable
kidney disease (ESKD) often represents a more
consequence of aging.
meaningful clinical outcome than change in level of
Among participants in the BLSA without CKD,
renal function. This outcome is easily defined and
the rate at which creatinine clearance declined over
identified, and the clinical significance of ESKD (de-
time was greater among older participants.6 Consis-
tent with these results and with prior cross-sec-
tional studies showing lower levels of renal function Correspondence: Ann M. O’Hare, MA, MD, Division of Nephrol-
ogy, VA/Puget Sound Medical Center, Nephrology and Renal
among older people, older age seems to be a risk Dialysis Unit, Building 100, Room 5B113, 1660 S. Columbian Way,
factor for the development of CKD, defined as an Seattle, WA 98108. Phone: 206-277-3192; Fax: 206-764-2022;
estimated GFR (eGFR) ⬍60 ml/min per 1.73 m2.8 E-mail: [email protected]
However, the relationship between age and rate of Copyright 䊚 2009 by the American Society of Nephrology
GFR is the best index available to assess kidney and diet are stable, serum creatinine could be used
function in disease and in health in an individual. It for monitoring GFR more closely. In general, a
is 120 to 130 ml/min per 1.73 m2 in young, healthy change in serum creatinine ⬎15% is likely to indi-
adults, and it decreases by about 0.8 ml/min per cate a significant fall in GFR in an individual patient
1.73 m2 per year after 40 yr of age. However, it is rather than being caused by simple biologic and an-
important to note that, in the Baltimore Longitudi- alytical variations. Table 1 shows other limitations
nal Study on Aging, about one third of the patients of Scr.
that were followed did not have a decrease in GFR
with aging. Creatinine Clearance
The GFR cannot be measured directly in an in- Creatinine clearance as measured from a 24-h urine
dividual. Therefore, it is assessed using either exog- collection can be used to measure GFR, but it is vital
enous markers or endogenous markers in their to remember the high likelihood of inaccurate col-
steady states as shown in Table 1. lection, especially in some elderly people with cog-
nitive impairment or the bed bound. It is impor-
tant, therefore, to check for adequacy of urinary
MEASURING GFR USING EXOGENOUS collection before interpretation of clearance. The
MARKERS collection is said to be adequate if the creatinine
excretion is 20 to 25 mg/kg per day in a young
The direct methods in the general population and healthy man and if it is 15 to 20 mg/kg per day in a
in elderly persons are riddled with a number of young healthy woman. In elderly people, adequacy
problems (Table 1). Therefore, except in rare situ- is similarly checked because it is assumed that the
ations such as in a prospective kidney donor with muscle mass (and hence creatinine generation) and
borderline GFR for eligibility, these methods are renal function decline simultaneously with age.
not used in clinical practice. Caution must, therefore, be used if this assumption
cannot be made in individual instances. If this
method is used in the setting of acute renal failure
METHODS OF GFR ESTIMATION USING or rapidly changing serum creatinine, it is necessary
ENDOGENOUS MARKERS to measure an average from simultaneous serial se-
rum creatinine values during urine collection. Cre-
Serum Creatinine (Scr) atinine clearance systematically overestimates GFR
GFR estimation based on serum creatinine alone is because of tubular secretion of creatinine. The 24-h
not an ideal method, especially in elderly persons urine collection for the estimation of GFR has been
because it is influenced by a number of variables shown by many studies to not be any more reliable,
such as age, gender, muscle mass, diet, and medica-
tions that block creatinine’s tubular secretion. For
Correspondence: Devraj Munikrishnappa, Department of Inter-
example, despite reductions in GFR to ⬍60 ml/min nal Medicine, The Nephrology Division, St. Louis University
per 1.73 m2, there may not be a significant increase School of Medicine, 1402 South Grand, St. Louis, MO 63104.
in creatinine in the elderly persons with decreased E-mail: [email protected]
muscle mass. On the other hand, if the muscle mass Copyright 䊚 2009 by the American Society of Nephrology
MDRD Equation surement, because it only requires serum creatinine, age, gen-
The original equation was derived from a study of 1628 mid- der, and race, but not weight or any urine collections.
dle-aged, nondiabetic, chronic renal insufficiency patients that Differences in calibration of creatinine assays between labora-
used a directly measured GFR by urinary clearance of 125I- tories can lead to differences in GFR estimation and thus is an
Iothalamate.8 It has several advantages over the CG equation important limitation of estimation equations in general. The
including providing an estimate of GFR rather than creatinine four-variable MDRD was therefore re-expressed in 2005 (as
clearance, and in addition, a greater percent of these estimates shown below) for use with creatinine methods calibrated to the
are within the clinically useful range for decision making: 90% reference assay method. It is important to note that laborato-
of the MDRD based estimates were within 30% of the mea- ries without calibrations of their serum creatinine assays cali-
sured GFR compared with about 75% of CG-based estimates. brated to the reference method, the isotope– dilution mass
However, the MDRD equation also has several limitations in- spectrometry (IDMS) method, should report eGFR using the
cluding that it is less accurate at levels above 60 ml/min per 1.73 original four-variable MDRD study equation, recognizing it is
m2. Consequently, it may lead to misdiagnosis and misclassi- less accurate, especially at higher levels of GFR.13
fication of CKD in individuals with mild renal insufficiency.1,9 In recent years, many laboratories in the United States
Table 4 shows additional limitations. started reporting, along with the serum creatinine, MDRD-
There have been some validation studies of the MDRD based eGFR values in routine chemistry laboratories and, in
equation in elderly people concluding that is it is better than some instances, with separate values for African Americans
the CG equation.10 –12 The most widely used form of MDRD in and non-African Americans. However, if the value of eGFR is
elderly people is the four variable version or the version that ⬎60 ml/min per 1.73 m2, no specific values are mentioned but
was abbreviated from the original six variable version (shown reported simply as ⬎60 ml/min per 1.73 m2. As mentioned
below). This is especially advantageous for elderly people com- above, this is because of a lack of precision of estimation at
pared with the CG formula or the creatinine clearance mea- higher levels of GFR. If the specific value ⬎60 ml/min per 1.73
REFERENCES
GFR ESTIMATION FROM SERUM CREATININE AND
SERUM CYSTATIN C–BASED EQUATIONS *Key References
1. Stevens LA, Coresh J, Greene T, Levey AS: Assessing kidney func-
tion—measured and estimated glomerular filtration rate. N Engl
As mentioned above, a recent study involving a pooled analysis J Med 354: 2473–2483, 2006*
of individuals with chronic kidney disease proposed an estima- 2. Stevens LA, Coresh J, Schmid CH, Feldman HI, Froissart M, Kusek J,
tion equation that included serum cystatin C in addition to Rossert J, Van Lente F, Bruce RD 3rd, Zhang YL, Greene T, Levey AS:
serum creatinine, age, sex, and race. The study concluded this Estimating GFR using serum cystatin C alone and in combination with
serum creatinine: a pooled analysis of 3,418 individuals with CKD.
equation provided the most accurate estimates. However, fur-
Am J Kidney Dis 51: 395– 406, 2008
ther studies are needed to confirm this, especially in elderly 3. Shlipak MG: Cystatin C: research priorities targeted to clinical decision
persons.2 making. Am J Kidney Dis 51: 358 –361, 2008
4. Köttgen A, Selvin E, Stevens LA, Levey AS, Van Lente F, Coresh J:
eGFR ⫽ 177.6 ⫻ Scr⫺0.65 ⫻ CysC⫺0.57 ⫻ 共Age兲⫺0.20 Serum cystatin C in the United States: the Third National Health and
Nutrition Examination Survey (NHANES III). Am J Kidney Dis 51:
⫻ 0.82 共if female兲 ⫻ 1.11 共if black兲 385–394, 2008
This chapter will include two large longitudinal, ob- unique for two reasons: (1) both cross-sectional
servational studies of renal function changes in nor- and longitudinal observations were used in the pre-
mal aging, a smaller cross-sectional and interventional diction and (2) only individuals considered healthy,
study focusing on oxidants and inflammation, and a based on strict, standard criteria, were selected for
consideration of the causes of decreased function in these analyses. Reports on the longitudinal studies
aging in both animal models and aging normal hu- only used data from BLSA participants followed for
mans. We chose to focus on oxidant stress (OS) and at least 10 yr and who had five or more serial mea-
inflammation because they increase in aging and are sures. Overall, the longitudinal data analysis con-
thought to both underlie aging-related diseases, in- firmed the cross-sectional observations, although
cluding decreased kidney function. Importantly, it is the accelerated decline of kidney function with age
now possible to reduce OS and inflammation in was more accentuated and statistically significant.
both normal adults and patients with chronic kid- The estimated average annual change in CrCl was
ney disease (CKD). Therefore, if OS and inflamma- ⫺0.26 ml/min per 1.73 m2 in the age group 20 to 39
tion are critical in the pathogenesis of reduced renal yr and became 1.51 ml/min per 1.73 m2 after the age
function in aging, and progression in CKD, it is of 80 yr.3 A study of water restriction confirmed
incumbent on the renal community to recognize/ that older BLSA participants have impaired re-
reduce their levels as a part of normal care or in the sponse of renal tubules to change in plasma osmo-
construction of clinical trials aimed at reducing CKD lality associated with impaired sodium homeostasis,
or cardiovascular disease (CVD). Finally, increased previously reported in smaller studies.3 Although
OS and inflammation may reduce the ability of the these data are a widely cited reference for kidney aging,
aging person to sustain metabolic or physical stress. they have some intrinsic limitations. First, they
were estimated from men only. Second, the number
of very long-lived individuals (⬎85 yr of age) was
LONGITUDINAL, OBSERVATIONAL quite small. Third, while BLSA participants were
STUDIES OF RENAL FUNCTION IN ADULTS selected to be “healthy,” the diagnostic technology
available at that time may have not detected sub-
The Baltimore Longitudinal Study of Aging clinical cardiovascular and kidney disease. This is
The Baltimore Longitudinal Study of Aging important because the distinction between aging-
(BLSA), the first continuing scientific examination related renal changes and progressive renal insuffi-
of human aging, was started in 1958 and has been an ciency are associated with a different prognosis.
important source of information on the aging kid-
ney.1 In 1976, the BLSA confirmed the previously
Correspondence: Helen Vlassara, MD, Division of Experimental
postulated progressive decline of renal function Diabetes and Aging, Department of Geriatrics, Mount Sinai
with aging, using age-adjusted standards for creat- School of Medicine, 1 Gustave Levy Place, New York, NY 10029.
inine clearance (CrCl).2 The BLSA data were Copyright 䊚 2009 by the American Society of Nephrology
100
cipally derived from cellular metabolism, it is now widely ap-
Creatinine Clearance (mL/min)
50
0
200
Uptake, Detoxification, and Elimination of AGEs
150 AGEs react with cell surface receptors, which mediate opposite
100
responses. One receptor (AGER1) lowers OS and inflamma-
50
0
tory reactions,18 –20 whereas another (RAGE) mediates in-
30 40 50 60 70 30 40 50 60 70 80 40 50 60 70 30 40 50 60 70 30 40 50 60 70
creases in these parameters.9,21–23 Both receptors are driven by
Age (yrs)
the ambient levels of ligand (AGEs) in normal, healthy sub-
Figure 1. Representative subjects from the Baltimore Longitudi- jects. Serum AGEs, often bound to small peptides, are filtered
nal Study of Aging, who did not show a significant decline in by the kidneys and AGEs. They can also be metabolized to
measured creatinine clearance values over a period of several years. inactive molecules and excreted in the renal tubules.
0.50 0.50
0.25 0.25
0.00 0.00
0 1 2 3 4 5 6 0 1 2 3 4 5 6
MDRD-4 MDRD-6
1.00 1.00
Survival Distribution Function
0.50 0.50
0.25 0.25
0.00
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Years Since Enrollment Years Since Enrollment
Figure 2. Comparisons of different methods of calculating GFR in the InCHIANTI study individuals with respect to predictions of
survival. Note that there was considerable variation in the predictive value of the various equations, suggesting that additional factors
may have to be considered in the aging population.
Oxidant Stress and Inflammation in Normal Adults a decrease in eGFR with aging (Figure 3C, inset). Nonetheless,
From Early to Late Adulthood urinary excretion of AGEs was significantly lower in older
Many studies have shown that oxidant stress and inflamma- adults, consistent with a lower intake.
tion generally increase with aging; however, few have consid-
ered adults older than 75 yr old.2,24 When we studied a normal
cohort residing in New York that included older adults, we
found that there was considerable heterogeneity between indi- SOURCES OF PRO-OXIDANTS IN ADULTS
viduals (Figure 3).14 Although, on average, there is an increase
in oxidant stress and inflammation in aging, this was not a Generally speaking, the amount of pro-oxidants in food in-
universal finding in the normal population. For instance, al- creases when food is cooked at high heat and without wa-
though there was an overall increase in carboxymethyl lysine ter.25,26 We find that the way the food is cooked, rather than the
(CML; Figure 3A) and methylglyoxal (MG; Figure 3B), there composition of the diet, is the critical factor in the amount of
was a substantial number of older individuals who had normal oxidants in the food (Table 1). For instance, meat cooked with
levels (upper limit of normal for CML ⫽ 12 to 15 and MG ⫽ water (steamed or boiled) has a much lower AGE content than
1.0). There was an inverse correlation between CML levels and broiled meat. In addition, in food that is cooked in the presence
estimated GFR (eGFR; MDRD; Figure 3C), but, as with the of lipids at high temperature (as in fries), the amount of AGEs
BLSA, there was a number of normal adults that did not show is markedly increased. Additional ways to reduce oxidant for-
10
0 0.0 0
20 40 60 80 100 20 40 60 80 100 0 50 100 150 200 250
age (years) age (years) eGFRCr
Figure 3. Data from a cross-section of subjects without obvious concurrent disease. (A) Serum carboxymethyllysine, (B) serum
methylglyoxal, and (C) serum CML versus eGFRCr (inset, eGFRCr versus age). Note that the slope of the curves in A and B is largely
driven by the large range in CML and MG levels in the aged. The eGFRCr values in the aged also show a large spread.
∆%
MODIFYING OXIDANT INTAKE 20
∆%
intake directly correlates with serum AGEs14(Figure 4A). Al-
though this has not been extensively analyzed in the CKD pop- -40 * * * *
ulation, many of the same principles are likely to apply. * *
Namely, adults have constant habits for cooking and order -60 * *
*
TN -1
foods cooked in similar ways in restaurants. Surprisingly, it
6 hc
1
GE
AM
GE
F␣
so
al
ML
ER
p6 s
appears that this behavior can be readily modified. We enrolled G
dC
dA
8-i
RA
VC
sM
sC
AG
adults in the highest tertile of consumption of AGEs among
normal subjects of all ages (Figure 4B), randomly divided them
into two groups, and followed them for 4 mo.27 One group B CKD Patients on a Low-AGE Diet
40
food that they normally eat so that the formation of AGEs
would be lowered. Two results were noted: first, there was a
20
substantial reduction in the amount of AGEs consumed by the
subjects who modified their food preparation methods. Sec-
0
ond, this intervention was associated with an approximate de-
crease of 30 to 60% in the amount of circulating inflammatory -20
∆%
400 12
-1
6 hc
sM L
F␣
GE
AM
mg/l
so
G
4
al
GE
M
p6 s
TN
dC
sC
dA
8-i
VC
8
RA
200
2
4
0 0 Figure 5. The influence of a reduction of the intake of AGEs on
0 5 10 15 20 <15.4 >23 <15.4 >23
sAGE (U/ml) the blood levels of AGEs, measures of OS and inflammation, and
Dietary AGEs (Eq/day)
AGE receptors. (A) After a 4-mo period on a low-AGE diet (⬃50%
Figure 4. Correlations of serum levels of AGEs and levels of reduction), the levels of serum markers of OS, inflammation, and
oxidants markers of inflammation. (A) Serum levels of AGEs di- AGE receptors were all substantially reduced, whereas calorie
rectly correlate with markers of lipid oxidation (8-isoprostanes). (B) intake remained essentially unchanged. (B) After a 4-wk period on
Those subjects who consumed lower levels of AGEs had lower a low-AGE diet, CKD 2– 4 patients had a similar reduction of
serum levels of hsCRP and TNF␣. hsCRP, high-sensitivity C-reac- markers of OS and inflammation. Note that whereas RAGE levels
tive protein; TNF, tumor necrosis factor; sAGE, serum advanced were reduced by the dietary intervention, AGER1 levels returned
glycation factor. to normal levels.
Diabetes is a major health issue affecting the aging has better accuracy than the Cockroft formula in
US population: epidemiologic studies show an in- moderate and severe kidney function.9 Pathologi-
creased prevalence of diabetes with aging. The re- cally, the aging kidney may be associated with
ported incidence of diabetes in the elderly US pop- changes of basement membrane thickening and
ulation is at least 10 to 17%, caused by factors such mesangial expansion that are also key histologic
as obesity, decreased activity, insulin resistance, and features of diabetic glomerulopathy.1 Global glo-
increased oxidative tissue damage. The prevalence merulosclerosis affecting the kidneys of elderly per-
of metabolic syndrome also increases with age and sons may relate to hyperperfusion, also observed in
is frequently complicated by hypertension and diabetes. However, studies of the diagnosis and
chronic kidney disease (CKD). With further ad- prevalence of diabetic kidney disease in the elderly
vances in the treatment of diabetes, longer lifespan are lacking. Thus, there could be a higher preva-
is leading to more diabetes-related complications. lence of unusual presentations of diabetic kidney
Kidney disease secondary to diabetes has an in- disease (e.g., decreased GFR without albumin-
creased prevalence in the geriatric population,1 uria).10 Nondiabetic glomerular syndromes present
which comprises the fastest-growing subgroup of more commonly in geriatric patients because of
CKD and end-stage kidney disease (ESKD) in the conditions such as vasculitis, amyloidosis, parapro-
United States. About one third of older diabetic in- teinemia, membranous glomerulopathy, and anti-
dividuals have microalbuminuria,2 and an equal glomerular basement membrane (GBM) disease.
fraction have depressed kidney function. However, Another factor that needs to be considered in el-
CKD care of the elderly diabetic patient remains derly persons is the existence of renal artery stenosis
underemphasized, and nephrology consultation re- caused by atherosclerotic disease.
mains underused. Clinical guidelines for type 2 di-
abetes in the elderly do not address CKD, and
guidelines for diabetic CKD have not distinguished TREATMENT
age groups.3
The standard therapy of diabetic kidney disease is
the triad of blood glucose control, BP control, and
KIDNEY FUNCTION IN THE ELDERLY administration of angiotensin converting enzyme
DIABETIC PATIENT inhibitors (ACEIs) or angiotensin receptor blockers
(ARBs). The goals that have been established
Renal blood flow and GFR diminish over time in through many clinical studies are a hemoglobin A1c
elderly persons, minimized by a rise in the filtration of ⬍7%, a BP of ⬍130/80 (with weak data support-
fraction.4 In older diabetic patients, the decrease in ing a lower systolic goal if proteinuria persists), and
kidney mass, particularly from the renal cortex, and reduction of total urine protein to ⬍500 mg/g of
the histologic changes of diabetic nephropathy are creatinine or of urine albumin to ⬍300 mg/g of
compounded by advanced vascular changes.5 The creatinine. Although these goals have been vali-
term “concealed renal failure” has been applied to
elderly patients with normal serum creatinine but
decreasing GFR.6,7 The Modification of Diet in Re-
nal Disease (MDRD) is increasingly used in the Correspondence: Mark Williams, Renal Section, Joslin Diabetes
United States and has been found to be accurate in Center, One Joslin Place, Boston, MA 02215. Phone: 617-732-
diabetic kidney disease.8 A recent study of 160 dia- 2477; E-mail: [email protected]
betic patients reported that the MDRD equation Copyright 䊚 2009 by the American Society of Nephrology
Aging is a natural process of human development day (Figure 3). These numbers are significantly
and is characterized by a progressive loss of physio- higher for patients with CKD, including those un-
logic and reproductive functions.1,2 Despite signif- dergoing kidney dialysis. In the elderly, new-onset
icant advances in the fields of human physiology, adverse drug reactions are commonly mistaken by
pharmacology, and pathology, as well as medical healthcare providers as a new-onset disease or mor-
and clinical interventions, aging continues to be a bidity related to aging.13 Approximately 15 to 45%
significant risk factor and strong predictor of mor- of older adults develop moderate to severe forms of
bidity and mortality. memory impairment from medication use.14 This
Geriatric patients, while accounting for 15% of cognitive impairment may position older adults at a
the population, use 30% of all prescription drugs higher risk of overdose or nonadherence compared
prescribed in the United States. By 2020, ⬎50 mil- with the rest of the population.15 In addition, older
lion people will be over 65 yr of age and account for adults with kidney disease are cared for by a variety
25% of the US population (Figure 1). It is estimated of healthcare providers such as nephrologists, car-
that by 2030, the population of individuals over 65 diologists, general practitioners, and pharmacists
yr of age will increase by four-fold, and this age and may lack good continuity of care. Medication
group is the fastest growing segment of the US pop- reconciliation is still a major problem in this era of
ulation. This increase will add to healthcare costs the electronic medical record. Failure to disclose a
and strengthen the economic tsunami that our complete list of current medications taken to each
country is facing today.3,4 provider because of cognitive impairment may
Chronic kidney disease (CKD) is a common and have a profound impact on potential risk of drug–
progressive condition that continues to rise in the drug interactions or disease– drug interactions.16
United States.5–7 Limited data exist regarding drug The overall incidence of adverse drug reactions
use in elderly patients with CKD. Proven therapies is three- to ten-fold higher in older adults with kid-
are often underused in the geriatric population. For ney disease compared with those without CKD.17
example, recent data indicated that aggressive man- The incidence of adverse drug reactions correlates
agement of hypertension may be beneficial in older exponentially with renal function.18 Most drugs
patients.8,9 However, most physicians hesitate to and/or their metabolites are excreted renally
treat older patients more aggressively. This is, in through glomerular filtration. The overall size,
part, because of a high risk of drug– drug interac- mass, and effective area of filtration decreases with
tions, adverse drug reactions, and lack of clinical increasing age.19,20 These morphologic changes in-
data in this population. Monitoring the medica- crease the risk for drug and/or active metabolite ac-
tions used in older adults and identifying drug in- cumulation in older patients with kidney disease.
teractions and adverse events are crucial. Drug ther- After the age of 50, the number of nephrons pro-
apy management in older adults is challenging, and gressively declines from approximately ⬎1,000,000
many factors related to normal aging, disease states,
and lifestyle should be considered before initiation Correspondence: Ali J. Olyaei, PharmD, BCPS, Associate Profes-
of pharmacotherapy (Figure 2).10,11 sor of Medicine, Division of Nephrology and Hypertension, Ore-
In a recent study, Qato et al.12 documented that gon Health Sciences University, 3314 SW US Veterans Hospital
Road, Portland, OR 97201. Phone: 503-494-8007; Fax: 503-494-
91% of older adults regularly use one prescription 2994; E-mail: [email protected]
drug and ⬎50% use five or more prescriptions per
Copyright 䊚 2009 by the American Society of Nephrology
Age-related
Drug-related Disease-related
81%
57-64 27%
5%
87%
Women 65-74 33%
5%
92%
75-85 37%
4%
0 .2 .4 .6 .8 1
Figure 2. Percent of polypharmacy at different age
1+ 5+ 10+
groups: ⬎1 drug, ⬎5 drugs, and ⬎10 drugs.
justed according to estimated creatinine clearance. There are a both prescription and nonprescription, should be obtained to
number of methods to estimate creatinine clearance; however, identify potential nephrotoxins or interacting medications. A
the most commonly used method is the Cockcroft-Gault thorough medication history should be obtained to identify
method. There are a number of limitations when using the drug allergies or intolerances and previous adverse drug reac-
Cockcroft-Gault method in older patients. The production tions. Body mass index (BMI) and ideal body weight (IBW)
and elimination of creatinine decreases with age.26 This may should be calculated using the following formulas: BMI ⫽
overestimate renal function and mask the early stage of renal weight in kilograms divided by height in meters squared; IB-
dysfunction. In older adults, the use of Modification of Diet in W(men) ⫽ 50.0 kg ⫹ 2.3 kg for every 2.5 cm over 152 cm; and
Renal Disease (MDRD) may provide a better estimate of renal IBW(women) ⫽ 45.5 kg ⫹ 2.3 kg for every 2.5 cm over 152 cm.
function. Some drugs may increase the metabolic load by in- Volume status, both intracellular and extracellular, should
creasing creatinine production and/or urea production (e.g., be assessed frequently. Shifts in extracellular fluid volume may
glucocorticoids and androgens). Some agent may also interfere change the volume of distribution of many drugs. Patients with
with creatinine tubular secretion (e.g., cimetidine and tri- dehydration have a higher predisposition to drug toxicity. In
methoprim). Ketosis, hyperbilirubinemia, and some cephalo- older adults, total body volume decreases by 10 to 15%. Muscle
sporins may influence the measurement of plasma creatinine atrophy, reduced tissue perfusion, and increase in fat content
and cause renal function assessment inaccurate when serum change the volume of distribution of most hydrophilic agents.
creatinine is used. The plasma concentration of drugs with a narrow therapeutic
Considering these limitations, estimating renal function is window and small volume of distribution (aminoglycosides,
difficult in the elderly population. However, the use of the lithium) may alter considerably with any change in extracellu-
Cockcroft-Gault method is the safest and most effective ap- lar fluids. Coexisting hepatic dysfunction may alter protein
proach for dosage adjustments in patients with renal impair- binding, volume of distribution, and intravascular volume and
ment. necessitate further dosage modification.28
APPROACH TO AND UNIQUE FEATURES of the benefit of diagnosis and appropriate inter-
OF GLOMERULONEPHRITIS IN ELDERLY vention.
PERSONS
Pauci-immune, MPO-ANCA positive, crescentic glomerulo- Although GN is relatively uncommon in elderly individuals, it
nephritis is the most common form of GN in elderly persons, does occur. Because treatment response rates and complica-
and its incidence steadily increases with increasing age. Un- tions are comparable for individuals in all age groups, elderly
treated, this form of GN progresses rapidly to ESKD. Differ- persons will benefit from diagnosis and treatment using the
ences in the forms of vasculitis and implications for therapy same criteria for biopsy and intervention as in younger indi-
were recently reviewed by Jennette and Falk.16 Individuals with viduals. This recommendation is particularly justified given
this disease can have a significant and sustained remission with that ANCA-positive, pauci-immune, crescentic GN is the most
appropriate therapy; thus, recognition of this disorder in el- common form of GN in elderly individuals, and left untreated,
derly individuals, with prompt biopsy and treatment, is critical this disease rapidly leads to ESKD. Furthermore, following
The major goals of lowering BP in patients with to lower than usual BP targets is not associated with
chronic kidney disease (CKD) include reduction of an increased risk of adverse outcomes.
mortality, cardiovascular events, and slowing pro-
gression. Key considerations in the management of
hypertension include selection of a target BP and BP TARGETS IN THE ELDERLY
selection of agents used to attain the chosen target.
This chapter outlines key considerations in apply- In applying this recommendation to the elderly, it is
ing current guidelines for the management of BP to worth noting that none of the trials used to support
older patients with CKD. the safety of lower than usual BP targets in patients
with CKD enrolled any participants older than 75
(Table 1). Thus, the safety of treating to a lower than
BP TARGETS usual BP level in older patients with CKD is not
known. Indeed, in the very elderly (e.g., 85 yr or
older), observational data showed that there is a J-
Many clinical practice guidelines recommend a
shaped relationship between BP and survival and
lower than usual BP target for patients with CKD.
that optimal BP may be higher than in younger peo-
For example, the Kidney Disease Outcomes Quality
ple.2,3 Although a number of trials have specifically
Initiative (KDOQI) recommends a target BP of
examined the effect of BP lowering on nonrenal
⬍130/80 mmHg for all patients with CKD, which is
outcomes in the elderly, and in many instances have
defined as an eGFR ⬍60 ml/min per 1.73 m2 or
shown a benefit, these trials have tended to target a
“kidney damage” (specified as microalbuminuria
higher than usual (rather than lower than usual or
or macroalbuminuria for patients with diabetes).
even usual) BP.2,4 –7
Citing KDOQI, the seventh report from the Joint
In pursuing a lower than usual BP target in an
National Committee on Prevention, Detection,
older person, it is important to assess the impor-
Evaluation, and Treatment of High Blood Pressure
tance to that patient of the implicit goals of this
(JNC 7) also recommends a target BP of ⬍130/80
intervention (e.g., slowing progression of CKD and
mmHg for all patients with CKD defined as an
reducing cardiovascular risk) as well as the poten-
eGFR ⬍60 ml/min per 1.73 m2 or protein-to-creat-
tial harms. For a variety of reasons, it is likely that
inine ratio ⱖ200 mg/g.1 A target BP ⬍130/80
for many patients who meet criteria for CKD, the
mmHg is also recommended by the American Dia-
risk-to-benefit ratio of BP lowering may differ from
betes Association (ADA) and by JNC 7 for all pa-
that in younger patients. First, CKD in the elderly is
tients with diabetes.
often slowly progressive or nonprogressive, and the
Despite the consistency of guideline recommen-
risk of progression to end-stage kidney disease
dations for lower than usual BP targets in patients
(ESKD) is lower for older than for younger patients
with CKD, these recommendations are based on
opinion rather than the results of randomized con-
trolled trials. Few trials have shown that treatment Correspondence: Ann M. O’Hare, MA, MD, Division of Nephrol-
ogy, VA/Puget Sound Medical Center, Nephrology and Renal
to lower than usual BP targets slows progression of Dialysis Unit, Building 100, Room 5B113, 1660 S. Columbian Way,
CKD or reduces other clinically significant out- Seattle, WA 98108. Phone: 206-277-3192; Fax: 206-764-2022;
comes in patients with CKD (Table 1). On the other E-mail: [email protected]
hand, available evidence also suggests that treating Copyright 䊚 2009 by the American Society of Nephrology
with similar levels of eGFR.8,9 Thus, there may be less to be nician must consider each individual patient’s likelihood of
gained from slowing progression if this is slow to begin with. It experiencing progressive loss of renal function and mortality
is also not clear that mortality risk for most elderly patients in relation to their age peers and in the context of their risk for
with a low eGFR is any higher than for their age peers with a other (perhaps competing) health outcomes and their risk for
“normal” eGFR.10,11 The majority of older patients who meet adverse events as a result of BP lowering.
criteria for CKD have very moderate reductions in eGFR (e.g.,
45 to 59 ml/min per 1.73 m2). For many of these patients, the
relative and absolute risk of death may be no greater than for CHOICE OF AGENTS
patients of the same age whose eGFR falls in the normal
range.10,11 Thus, it is not even clear that reducing mortality risk In addition to lowering BP, progression of CKD can probably
provides a compelling rationale for a lower than usual BP in also be slowed by reducing proteinuria.14 For this reason, an-
many older patients with an eGFR ⬍60 ml/min per 1.73 m2. giotensin-converting enzyme inhibitor (ACEI) and angioten-
Second, it is not clear that slowing progression of CKD is al- sin II receptor antagonists (ARBs) are considered first-line
ways the most meaningful goal of anti-hypertensive therapy in agents for patients with CKD in a number of different clinical
older patients with a low eGFR.9,12Their risk for other out- practice guidelines. KDOQI recommends that these agents be
comes such as cardiovascular events, disability, and cognitive prescribed for patients with diabetic CKD (defined as an eGFR
insufficiency is often much higher than that for ESKD. CKD in ⬍60 ml/min per 1.73 m2 or micro- or macroalbuminuria) and
the elderly rarely occurs in the absence of other comorbid con- for those with nondiabetic proteinuric CKD, even in the ab-
ditions.9,13 The presence of multiple comorbid conditions in sence of hypertension.1,15,16 Based on recommendations from
older patients with CKD may complicate the management of KDOQI and a subset of trials referenced in KDOQI, JNC 7
CKD by creating potentially conflicting or competing treat- identifies CKD, defined as an eGFR ⬍60 ml/min per 1.73 m2 or
ment goals. Third, the potential harms of BP lowering may be a protein-to-creatinine ratio ⱖ200 mg/g, as a compelling indi-
greater in the elderly. Most elderly patients with CKD have cation for the use of ACEI or ARB.
isolated systolic hypertension. Thus, theoretically, treatment
of their systolic hypertension may have the unintended effect
of lowering diastolic pressure to suboptimal levels, leading to CHOICE OF AGENTS IN THE ELDERLY
impaired perfusion during diastole. Orthostatic hypotension is
also more common in the elderly and may be aggravated by In applying these guidelines to the management of older pa-
treatment to lower than usual BP targets. Finally, elderly pa- tients with CKD, it is important to note that many of the key
tients, particularly those who are frail, may be more likely to studies supporting these recommendations did not include
experience injury as a result of an episode of hypotension. participants older than 70.17–19 Nevertheless, a subgroup anal-
In summary, in deciding whether to target a lower than ysis among participants older than 65 enrolled in the RENAAL
usual BP in older patients who meet criteria for CKD, the cli- trial, a trial among type II diabetics with macroalbuminuria,20
Heart disease constitutes the leading cause of death CKD and 32% for those with both CKD and diabe-
in the United States. Age is an important, albeit tes.8
nonmodifiable, risk factor for cardiovascular dis- Ample information is available on the epidemi-
ease in the general population, as well as in patients ology of cardiovascular disease in older individuals
with chronic kidney disease (CKD). The prevalence and its relationship with kidney function, including
of chronic ischemic heart disease in men and key prospective studies in elderly individuals such
women ⱖ65 yr of age in the United States in 1995 as the Cardiovascular Health Study. In a prospec-
was 83 per 1000 men and 90 per 1000 women. tive study of traditional and novel cardiovascular
Among those ⱖ75 years of age, the prevalences risk factors, diabetes, hypertension, smoking, low
were 217 per 1000 for men and 129 per 1000 for physical activity, left ventricular hypertension, and
women. Increasing evidence has accumulated that nonuse of alcohol were all predictors of subsequent
elderly individuals with cardiovascular disease can cardiovascular mortality, whereas high-density li-
benefit greatly from several aspects of secondary poprotein (HDL)-cholesterol, low-density lipopro-
prevention.1 tein (LDL)-cholesterol, triglycerides, and obesity
Kidney disease has been shown to be an impor- were not associated with such risk.9 None of the
tant determinant of cardiovascular disease,2,3 and novel cardiovascular risk factors that were tested
patients with CKD should be regarded a “highest were independently associated with cardiovascular
risk” group for cardiovascular disease, irrespective mortality, including C-reactive protein and anemia
of levels of traditional cardiovascular disease among others. Of note, homocysteine and phos-
(CVD) risk factors (http://www.kidney.org/profes- phorus were not evaluated in that study, and the
sionals/kdoqi/guidelines_ckd/p7_risk_g15.htm). other negative associations need to be interpreted
Furthermore, several cardiovascular risk factors are in light of the relatively low power of this study.
increasingly prevalent with declining kidney func- Table 1 provides a list of established and novel car-
tion.4 Interestingly, the Framingham Risk Score is diovascular risk factors in patients with CKD.
only poorly predictive for CVD in patients with In contrast, little evidence has been generated on
CKD, and standard factors only account for a small the efficacy and safety of standard curative or pre-
proportion of the observed risk in these patients.5 ventive cardiovascular interventions in patients
Finally, older age is an important determinant of with CKD. Most landmark trials have explicitly ex-
kidney function (as indicated by its representation cluded patients with CKD,10,11 and similarly, older
in the Modification of Diet in Renal Disease subjects were also barred from participation in
(MDRD) estimation equation for GFR6). It has most of these trials. These two independent phe-
been estimated that more than a third of US indi- nomena jointly explain the particular evidence vac-
viduals over age 70 have CKD Stages 3 to 5 and the uum for the population of older adults with CKD.
prevalence is increasing over time.7 One can postu-
late that the older individual with CKD is at the
highest risk of CVD, and even more so if additional Correspondence: Wolfgang C. Winkelmayer, Renal Division and
comorbid conditions including diabetes (DM), hy- Division of Pharmacoepidemiology and Pharmacoeconomics,
Brigham and Women’s Hospital and Harvard Medical School,
pertension, obesity, and other vascular disease are 1620 Tremont Street, Suite 3-030, Boston, MA 02120. Phone:
present. Indeed, among adults over age 67, 2-yr car- 617-278-0036; Fax: 617-232-8602; E-mail: wwinkelmayer@
diovascular mortality was 10% for those without partners.org
diagnosed CKD or diabetes but 30% for those with Copyright 䊚 2009 by the American Society of Nephrology
The fundamental question is whether evidence can be extrap- emia is clearly accepted as an important cardiovascular risk
olated to older patients with CKD from trials that effectively factor, and medical treatment, predominantly with statins, is
excluded those patients or contained only few such patients? well established for both primary and secondary cardiovascu-
Or should we require that specific trials be conducted in this lar prevention. Cardiovascular prevention with statins has also
relatively small segment of the population? Alternatively, been studied in a trial dedicated to the older population. In the
should we require that prespecified and sufficiently powered PROspective Study of Pravastatin in the Elderly at Risk (PROS-
tests for interaction of drug efficacy with age and kidney func- PER) trial, patients aged between 70 and 82 yr were enrolled
tion be planned and conducted? Although it would be desir- and compared with placebo. Treatment with 40 mg of prava-
able to inform more evidence-based practice in geriatric ne- statin per day conferred a 15% reduction in the risk of the
phrology, it is unlikely that such information will become primary combined cardiovascular endpoint (fatal or nonfatal
available on a larger scale anytime soon. Only recently, studies myocardial infarction or stroke) and a 19% reduction in the
were conducted, at the very least, that specifically focused on secondary endpoint of fatal or nonfatal myocardial infarction.
the older population.12 Additional top-level evidence has been These risk reductions were found to be in line with those found
made available from post hoc analyses of individual or pooled in trials of younger patients. In a post hoc analysis of data from
data from randomized trials. The vast majority of the evidence three pravastatin trials, it was found that statins were also effi-
on cardiovascular risk interventions in older patients with cacious in reducing cardiovascular outcomes in patients with
CKD, however, has come from retrospective pharmaco-epide- CKD Stage 3.13 The mean age in this study was 65.7 yr. Al-
miologic studies, often with serious methodological limita- though an interaction test with age was not conducted in this
tions. The following aims to provide evidence on a selected analysis, it is probably safe to assume that lipid-lowering treat-
number of cardiovascular risk factors and interventions in el- ment using statins is also efficacious in older patients with
derly patients with CKD: lipid disorders and lipid-lowering CKD. The optimal dose of specific agents or any preferred lipid
therapy, C-reactive protein and inflammation, homocysteine, targets, however, is not clearly established. Statins were also
as well as hyperphosphatemia and use of phosphate binders. efficacious in reducing cardiovascular events in kidney trans-
Other risk factors are covered in other chapters of this curric- plant patients.14 Whether statins are also efficacious in patients
ulum, notably diabetes and proteinuria, hypertension, and on hemodialysis is unclear. At the very least, chronic dialysis
anemia. patients with diabetes did not benefit from statin treatment in
a large randomized trial.15 Further evidence can be expected in
the near future when the results from the large Study of Heart
CARDIOVASCULAR RISK FACTORS IN CKD and Renal Protection (SHARP) trial will be released.
PHYSIOLOGIC EFFECTS OF AGING ON stiffening of medium and large arteries lined with
BLOOD VESSEL ELASTICITY AND atheromatous plaques. The progressive accumula-
COMPLIANCE tion of atherosclerotic plaque continues with aging
and often remains silent until lesions reach a critical
The aging process is commonly associated with in- stenotic threshold or rupture, leading to dimin-
creased vascular rigidity and decreased vascular ished organ perfusion. Arterial stiffening also leads
compliance. This process reflects the accumulation to an increase in pulse wave velocity and an en-
of smooth muscle cells and connective tissue in the hancement in central aortic systolic pressure.3 The
walls of major blood vessels. Endothelial cells and increased waveform velocity allows the backward
smooth muscle cells constitute most of vessel wall reflective wave to return earlier back to the heart.
cellularity and the remainder of the wall is com- The resulting increases of the left ventricular myo-
posed of extracellular matrix including collagen cardial load and the loss of coronary perfusion at
and elastin. Although aging has minimal effect on the onset of diastole may potentiate myocardial
the muscular tunica media layer thickness, aging ischemia.3
leads to profound progressive thickening of the tu- Common ailments in the elderly including dia-
nica intima layer comprised of endothelial cells and betes, dyslipidemia, hypertension, obesity, and cig-
an extracellular matrix. In addition, with aging, arette smoking can accelerate the process. Although
there is a thinning and separation of individual elas- the exact mechanisms have yet to be elucidated,
tin lamellae, as well as an increase in the collagen multiple factors including genetics, advanced glyca-
matrix.1 tion endproducts, and oxidative stress likely con-
The age-related vascular rigidity and decreased ar- tribute to the pathogenesis of accelerated athero-
terial compliance leads to progressive increase in sys- sclerosis.
tolic BP, with 25% of patients over 75 yr of age suffer-
ing from isolated systolic hypertension.2 Healthy
elderly patients without hypertension also show a EFFECTS OF AGING, ATHEROSCLEROSIS,
modest increase in peripheral vascular resistance and AND HYPERTENSION ON THE KIDNEYS
only a modest related increase in systolic BP.2 Dilation
and stiffening of the proximal aorta and its major Renal blood flow declines with aging at a magnitude
branches including the brachiocephalic, carotid, and of 6 ml/min per year, with a proportionately larger
subclavian arteries occur to a greater extent than the reduction in cortical blood flow than medullary
peripheral arteries with aging. This also blunts the ca- flow.2 GFR also declines with aging albeit at a slower
rotid baroreceptor sensitivity and increase the risk of rate of 1 ml/min per year. It is unclear if this change
end organ damage to the kidneys, heart, and brain.1,2 is caused by the aging process alone or is also attrib-
It is important to keep in mind that the presence of utable to hypertension and arteriosclerosis. It is es-
pre-existing hypertension and accelerated atheroscle- timated that as much as 26% of all end-stage kidney
rosis intensify the vascular pathology discussed to a far disease (ESKD) in the United States is related to
greater extent. hypertensive arteriolar nephrosclerosis. This num-
Atherosclerosis is characterized by the progressive Copyright 䊚 2009 by the American Society of Nephrology
CONCLUSIONS REFERENCES
Aging alters the physiology of arterial vasculature. Arterial cir- *Key References
1. O’Rourke MF: Arterial aging: pathophysiological principles. Vasc Med
culation stiffens and becomes less compliant, proximal arteries (Lond) 12: 329 –341, 2007*
dilate, and carotid baroreceptors become less sensitive. GFR 2. Rowe JW: Clinical consequences of age-related impairments in vas-
also declines with aging. Coexisting hypertension and athero- cular compliance. Am J Cardiol 60: 68G–71G, 1987
sclerotic disease exaggerate the normal vascular aging process, 3. Townsend RR: Analyzing the radial pulse waveform: narrowing the gap
endangering the longevity of end organ function including the between blood pressure and outcomes. Curr Opin Nephrol Hypertens
16: 261–266, 2007
kidneys. Hypertensive and atherosclerotic renal diseases ac- 4. Bauer JH, Reams GP, Wu Z: The aging hypertensive kidney: patho-
count for a large proportion of progressive CKD and ESKD in physiology and therapeutic options. Am J Med 90: 21S–27S, 1991*
the United States. It is important for the nephrologists to focus 5. Mailloux LU, Napolitano B, Bellucci AG, Vernace M, Wilkes BM, Mos-
on the preventative measures of atherosclerosis including an sey RT: Renal vascular disease causing end-stage renal disease, inci-
aggressive treatment of dyslipidemia, diabetes, and hyperten- dence, clinical correlates, and outcomes: a 20-year clinical experience.
Am J Kidney Dis 24: 622– 629, 1994
sion along with cigarette smoking cessation. Furthermore, a 6. Greco BA, Breyer JA: Atherosclerotic ischemic renal disease. Am J
careful evaluation of suspected atherosclerotic disease may be Kidney Dis 29: 167–187, 1997*
warranted. 7. Bloch MJ, Pickering T: Renal vascular disease: medical management,
angioplasty, and stenting. Semin Nephrol 20: 474 – 488, 2000*
8. Bloch MJ, Trost DA, Whitmer J, Pickering TG, Sos TA, August P: Ostial
renal artery stent placement in patients 75 years of age or older. Am J
TAKE HOME POINTS
Hypertens 14: 983–988, 2001
• Effect of aging on baroreceptor reflex may profoundly increase the 9. Bonelli FS, McKusick MA, Textor SC, Kos PB, Stanson AW, Johnson
orthostatic side effect profile of antihypertensive vasodilators CM, Sheedy PF 2nd, Welch TJ, Schirger A: Renal artery angioplasty:
• As much as 26% of all ESKD in the United States is related to hyper- technical results and clinical outcome in 320 patients. Mayo Clin Proc
tensive arteriolar nephrosclerosis 70: 1041–1052, 1995*
• As much as 62% of patients over 70 may suffer from RAS and the risk 10. Ramchandani P, Jones LP, Langer JE, Toririan DA, Perini RF, Divgi CR:
increases with aging Renal imaging. NephSAP 7: 218 –287, 2008
An 81-yr-old African-American woman with end- issue problematic to define. Osteoporosis in CKD is
stage kidney disease (ESKD) who has been on he- only a part of the constellation of metabolic bone
modialysis for 10 yr seeks a consultation with you. problems. Therefore, its diagnosis and manage-
Her primary physician obtained a dual-energy x- ment may differ from general population. Bones
ray absorptiometry (DEXA) scan and asked her to are more severely affected in CKD than that from
discuss the results with her nephrologist for further normal aging. In a patient with renal osteodystro-
management. She is diagnosed with osteoporosis phy, there is the potential for low BMD to coexist
based on her T-scores of ⫺5.1 (AP spine), ⫺4.1 (left with an enormous range of functional abnormali-
femoral neck), and ⫺5.4 (left total hip). Her last ties. These range from high turnover bone lesions in
dialysis laboratory results show serum calcium is patients with uncontrolled hyperparathyroidism to
10.7 mg/dl, intact parathyroid hormone (PTH) is severely reduced bone remodeling activity in pa-
207 pg/ml, phosphorus is 5.7 mg/dl, and alkaline tients with adynamic bone disease. This is in con-
phosphatase (ALP) is 122 IU/L. She receives pari- trast to the non-CKD patient with osteoporosis
calcitol 5 mg intravenously on hemodialysis three where bone remodeling is not severely affected.
times per week.
Chronic kidney disease (CKD)-related bone dis-
ease is known as renal or uremic osteodystrophy. It IMPACT ON QUALITY OF LIFE
is associated with derangements in bone and min-
eral metabolism that leads to abnormal regulation Patients with CKD-MBD and osteoporosis are as-
of calcium, phosphorous, vitamin D, and PTH. It sociated with increased risk of fractures and are at a
encompasses a spectrum of conditions that are clas- high risk of cardiovascular disease.2 The overall in-
sified based on bone biopsy findings including os- cidence of hip fractures among dialysis patients is
teitis fibrosa (high turnover disease), mixed uremic about four-fold higher than that expected for gen-
osteodystrophy, osteomalacia (low turnover dis- eral population. The risk is increased in both men
ease), and adynamic bone disease. KDIGO (kidney and women.3 Fractures may limit ambulation, lead-
disease: improving global outcomes) has proposed ing to loss of independence and chronic pain,
to define CKD-related bone and mineral metabolic thereby decreasing quality of life. Mortality risk in
abnormalities in the context of a systemic disorder dialysis patients with hip fracture is twice that of
called CKD–mineral and bone disorder (CKD- patients without hip fracture.4 Women who are 65
MBD).1 yr of age and older and have moderate renal dys-
Osteoporosis is a condition characterized by low function (eGFR ⬍ 60 ml/min per 1.73 m2) are also
bone mass leading to reduced bone strength and an at an increased risk of hip fractures.5 In addition to
increased risk of fractures. Hip, spine, and wrist are the traditional risk factors, several risk factors for
most commonly affected. The WHO definition of low BMD have been identified in the CKD popula-
osteoporosis is based on bone mineral density tion such as renal osteodystrophy, ethnicity, trans-
(BMD) measurements. The NIH consensus state- plant status, and duration of dialysis.6
ment refers to osteoporosis as a skeletal disorder
characterized by compromised bone strength pre-
Correspondence: Harmeet Singh, Division of Nephrology, Brody
disposing to increased risk of fracture.
School of Medicine, 2355 W. Arlington Boulevard, Greenville, NC
Osteoporosis and renal osteodystrophy may co- 27834. E-mail: [email protected]
exist in elderly patients with CKD, which makes the
Copyright 䊚 2009 by the American Society of Nephrology
Anemia is a common sequela of chronic kidney dis- lower Hg levels must always be considered when mak-
ease (CKD). As GFR declines, the risk of anemia ing the diagnosis of anemia. These include altitude of
increases, particularly when the GFR is ⬍60 ml/min residence, smoking status, normal fluctuations in
per 1.73 m2.1 Anemia is also common in the geriat- plasma volume, and other patient-centered factors.
ric population, with a prevalence of ⬎10% in pa- These are not routinely considered when Hg concen-
tients greater than 65 yr old (Figure 1).2 This poses a tration is reported by clinical laboratories, which typ-
significant challenge in the diagnosis and manage- ically only correct for age and sex. In addition, refer-
ment of elderly patients with anemia who have con- ence ranges for Hg may vary from laboratory to
current CKD. The following review discusses the laboratory across all subjects. Thus, results of Hg test-
definition, diagnosis, management, and outcomes ing must always be interpreted in the full context of
of anemia in elderly patients with CKD. the individual patient.
Defining normal hemoglobin levels in the elderly Because of the increased prevalence of lower Hg levels
population is important for purposes of establish- in the geriatric population, as well as increased risk of
ing a diagnosis and monitoring treatment effects comorbidities leading to anemia, determining the
and outcomes. In 1968, an expert panel from the cause of anemia can be challenging, especially in the
World Health Organization recommended normal face of CKD. Determining whether CKD is the cause
hemoglobin (Hg) levels be ⬎13.0 g/dl for men and of anemia, whether there is another explanation, or
⬎12.0 g/dl for nonpregnant women.3 Although the perhaps whether there are several etiologies of anemia
accuracy of these definitions has been called into has important implications for treatment and thus
question, more recent population-based studies, should be investigated rigorously before simply begin-
including NHANES III and the Kaiser-Scripps da- ning therapy with an erythropoietic stimulating agent
tabase, have been relatively consistent with the pro- (ESA). Below is a discussion of the diagnosis of anemia
posed levels.2,4 in the elderly with CKD.
As previously mentioned, anemia is quite common
in the elderly population. Using the WHO definition
of anemia, NHANES III data have shown that the CKD AND ERYTHROPOIETIN PRODUCTION
prevalence of anemia in subjects ⱖ65 yr old is
⬎10.6%, significantly higher than that of the general Anemia attributable to CKD was suspected in ap-
population. However, when present in the elderly, proximately 8% of all anemia cases in the elderly,
anemia tends to be mild. From the NHANES III data, according to NHANES III. Furthermore, some
⬍3% of subjects 65 yr of age and older had an Hg studies have shown lower than expected erythro-
⬍11.0 g/dl. Whether these modestly lower values of poietin (Epo) levels for the degree of anemia in ge-
Hg correlate with poorer outcomes is not known.
Attention should be paid to racial and ethnic dif-
ferences and other patient-centered factors in anemia Correspondence: Julio C. Vijil Jr, MD, MPH, Department of
Medicine, University of Illinois at Chicago, Section of Nephrology,
prevalence in the elderly as well. Data from NHANES MC 793, 820 South Wood Street, Room 417W, Chicago, IL
III showed that, among individuals 65 yr of age and 60612. Phone: 312-996-6736; Fax: 312-355-0418; E-mail: jvijil1@
older, African Americans have a higher prevalence of uic.edu
anemia (Figure 2).2 Other factors that contribute to Copyright 䊚 2009 by the American Society of Nephrology
riatric patients. Some have theorized that the elderly have a eGFR alone cannot be used reliably to exclude anemia of CKD
decreased ability to produce Epo in response to hypoxia or that in the geriatric population.
there is decreased hypoxia-sensing in the kidney with increas-
ing age and thus a higher Epo requirement to maintain the
same Hg level.5,6 However, one study has shown that normal,
NUTRITIONAL ANEMIA
healthy elderly individuals have the capacity to produce ade-
quate Epo in response to phlebotomy.7 Whether this translates
Inadequate nutrition is a common problem in the elderly. Iron
into an ability to maintain Epo production and Hg levels in the
deficiency, both nutritional and because of chronic blood loss,
setting of chronic blood loss or other perturbations in Hg con-
accounts for most cases of nutritional anemia, representing
centration is not known. Thus, to date, there is no solid evi-
about one third of anemia cases in the elderly in the NHANES
dence that the elderly have a decreased ability to produce Epo
III cohort. Diagnosis of iron deficiency should be made with
when kidney function is normal.
free iron, total iron binding capacity, and ferritin levels. As is
One possible explanation for the observation of lower than
the case in the general CKD population with iron deficiency,
expected Epo levels is that CKD may be present to some degree
ferritin levels are not reliably low because of chronic inflam-
in some elderly subjects despite the results of GFR estimating
mation and thus should not be used to exclude iron deficiency
equations. This may be because of the fact that the equations to
anemia. Similarly, mean corpuscular volume (MCV) is not
estimate GFR are suspect in the elderly as the population of
always low in CKD, despite iron deficiency anemia, so a nor-
subjects in the studies used to create these equations did not
mal MCV should not exclude the diagnosis. Based on early
include very elderly patients. Second, the correlation of esti-
studies, an iron/total iron binding capacity (TIBC) ratio ⱕ16%
mated GFR to renal endocrine function is not as well studied in
or a serum ferritin ⬍12 ng/ml is necessary to make the diag-
the elderly. For the general population, anemia caused by CKD
nosis of iron deficiency anemia.8,9
can present as early as eGFR of about 60 ml/min per 1.73 m2.
Other causes of nutritional anemia must be investigated in
Whether this relationship between eGFR and Epo production
the elderly also. Particularly, folate and B12 deficiencies are
is similar for the elderly population is not known. Therefore,
more frequently present than in the general population. One
study found that ⬎10% of the elderly had borderline or low B12
levels.10 As was the case with iron deficiency, one must be care-
ful not to exclude the diagnosis of B12 or folate deficiency solely
by MCV when concurrent CKD is present.
MALIGNANCY
epidemiologic information and special diagnostic consider- • ESAs and iron supplementation are effective in treating anemia of CKD
in this population
ations must be taken into account since the elderly are more • Hg should be targeted to a level between 10.0 and 12.0 g/dl
likely to have a concomitant cause of anemia in the face of
CKD. Treatment with ESAs is effective, and iron supplemen-
tation, initially oral but intravenous if ineffective, should be
DISCLOSURES
considered in the regimen given the frequency of iron defi-
None.
ciency in these individuals. When choosing a target Hg level, it
is important to remember that higher Hg levels may be associ-
ated with increased cardiovascular and thrombotic events. The
REFERENCES
elderly are at risk for conditions that raise the risk of these
events, so anemia must be treated judiciously. A target Hg of *Key References
10.0 to 12.0 g/dl is an acceptable target based on available evi- 1. Nurko S: Anemia in chronic kidney disease: causes, diagnosis, treat-
dence. ment. Cleve Clin J Med 73: 289 –297, 2006*
2. Guralnik JM, Eisenstaedt RS, Ferrucci L, Klein HG, Woodman RC:
Prevalence of anemia in persons 65 years and older in the United
States: evidence for a high rate of unexplained anemia. Blood 104:
TAKE HOME POINTS
2263–2268, 2004
• Anemia in the elderly with CKD is often multifactorial, with iron defi- 3. World Health Organization: Nutritional anemia: report of a WHO
ciency being common Scientific Group. World Health Organ Tech Rep Ser 405: 5–37, 1968
• Special consideration must be given to diagnoses common in this 4. Beutler E, Waalen J: The definition of anemia: what is the lower limit
population, including GI blood loss and hematologic causes (MDS, of normal of the blood hemoglobin concentration? Blood 107: 1747–
multiple myeloma) 1750, 2006
with endocrine abnormalities, and medication his- Copyright 䊚 2009 by the American Society of Nephrology
Confirm Hypoosmolality
Urine Na Urine Na
>30 mEq/L <20 mEq/L
Urine Na Urine Na Congestive Heart Failure
<20-30 mEq/L >20-30 mEq/L SIADH
Cirrhosis
Diuretic Use - Water Replacement
Vomiting Addison’s Disease Nephrotic Syndrome
Endocrine Deficiency
Diarrhea Polycystic Kidneys Hypoalbuminemia
Pancreatitis Bicarbonaturia
Diuretic Use Figure 1. Hyponatremia (differential diagnosis).
be considered indicative of some degree of conservation of generally treated by addressing the primary underlying abnor-
sodium. These urine sodium levels will often be associated with mality and using diuretic regimens.
urine osmolality values at least 1.5 times that of the plasma, In elderly hyponatremic patients who appear euvolemic
suggesting attempts at water conservation in response to fluid and have elevated urine sodium concentrations (⬎20 or 30
deficits. Such patients generally respond to the replacement of mEq/L) and elevated urine osmolalities suggesting inappropri-
intravascular volume with normal saline. Patients who appear ate water retention, a diagnosis of SIADH is often made.6 Here
dehydrated but who have elevated urine sodium levels (⬎30 it is important to rule out endocrine abnormalities such as
mEq/L) often have urine osmolalities closer to that of the hypoadrenalism and hypothyroidism. In addition, a careful
plasma levels and should be considered to have renal salt wast- search for drugs that stimulate AVP or facilitate the effects of
ing in the face of intravascular volume contraction. They AVP in the kidney should be undertaken. If these consider-
should also be treated with normal saline while attempts to ations are eliminated from diagnostic possibilities, the patient
diagnose the underlying abnormality are made. Various for- should be evaluated for other causes of SIADH. Table 1 lists
mulas such as the Adrogué-Madias formula (Figure 2) have some of the commonly associated central nervous system dis-
been used in an attempt to predict the increase in serum so- orders, tumors, and drugs associated with inappropriate AVP
dium that can occur when various concentrations of sodium secretion. Certain drug therapies such as SSRIs may pose a
replacement therapy are employed.5 If more or less than 1 L of special risk for the development of hyponatremia in the el-
replacement fluid is used, the change in serum sodium will derly, especially those who are older and smaller in body size.7
vary directly in proportion to the amount of fluid adminis-
tered. Table 1. Common causes of SIADH
When patients appear to be fluid overloaded or edematous, Central nervous system disease
they may have one of several edema-forming states, including Trauma
congestive heart failure, nephrotic syndrome, and cirrhosis. Stroke
Infection
The poor renal perfusion associated with any of these states is
Tumor
generally associated with urine sodium concentrations in the
Intracranial bleeding
lower ranges (⬍20 to 30 mEq/L) and a tendency to urine os- Neoplasms
molalities closer to that seen in the plasma. These disorders are Lung
Pancreas
ADROGUÉ-MADIAS FORMULA Prostate
Throat
Lymphoma
Infusate sodium concentration – Drugs
Patient serum concentration
Change in serum sodium concentration with 1L of infusate = SSRIs
Total body water + 1L Carbamazepine
Figure 2. The Adrogué-Madias Formula for the prediction of the Opiates
change in serum sodium that can occur following intravenous Cyclophosphamide
sodium replacement therapy. Mirtazapine
(70 kg Patient x 50% Total Body Water Factor for Older Patient) = 35L
Disorders of fluid balance are common in the el- duction to the nighttime period becomes evident.
derly and often are caused by age-related alterations With further increase in age, there is often reduc-
in urinary tract function, which can present clini- tion in the ratio of day to night urine flow to the
cally as urinary frequency, nocturia, and inconti- point that nighttime flow rates become equal to or
nence. Among the factors predisposing to the de- exceed daytime rates. Despite the change in circa-
velopment of these clinical disturbances are aging dian pattern of urine excretion, total urine produc-
changes in the renal and hormonal systems that tion per 24 h is not affected.3
control water and sodium excretion along with
changes in the reservoir function of the bladder1
(Table 1). This chapter will examine the effect of ARGININE VASOPRESSIN SECRETION
aging on the systems involved in urine formation
and consider how these changes interact with those Arginine vasopressin [AVP; antidiuretic hormone
of bladder function and lead to urinary frequency, (ADH)] is the major hormone responsible for the
nocturia, and incontinence. regulation of urine formation. The magnocellular
In young, healthy persons there is a circadian supraoptic and paraventricular nuclei of the hypo-
pattern to urine production in which nighttime thalamus where AVP is synthesized do not seem to
urine flow rate is less than daytime flow rate. In undergo age-related degenerative changes with ei-
association with the normal aging process, there is ther morphologic features of cell destruction or de-
an increase in nocturnal production of urine, so cline in cell hormone synthetic ability or hormone
that nighttime urine flow rate equals or exceeds content.
daytime production rate.2 When of sufficient mag- There is some controversy regarding the influ-
nitude to result in nocturia, this change has been ence of normal aging on daytime blood AVP levels.
termed nocturnal polyuria syndrome. A number of studies have indicated that, under
The establishment of a circadian rhythm of urine basal conditions, daytime plasma AVP concentra-
flow takes place during childhood, generally be- tion is not affected by increasing age. In contrast,
tween the ages of 2 to 5 yr. In normal children with several other studies have reported increased basal
a mean age of 7 yr, daytime rate of urine production plasma AVP levels in healthy elderly persons. Fur-
is two to three times that of the nighttime period. ther adding to controversy are reports that healthy
Delay or failure of the circadian rhythm to develop elderly subjects have daytime plasma AVP concen-
during childhood is associated with the presence of trations that are significantly lower than in young
nocturnal enuresis. In adulthood, the ratio of day- subjects.
time to nighttime urine production is usually In healthy adults, there is a diurnal release of
greater than 2:1, so that about 25% or less of daily AVP into the circulation with peak blood concen-
urine output occurs during sleep. Typical urine tration occurring during the hours of sleep. This
production rates are approximately 70 to 80 ml/h rhythm seems to be linked to the wake-sleep cycle
during the waking period and 30 to 40 ml/h during
sleep. This circadian pattern seems to be linked to
the day-night sleep pattern. The circadian pattern Correspondence: Myron Miller, MD, Director, Division of Geriat-
ric Medicine, Department of Medicine, Sinai Hospital of Balti-
of urine flow is paralleled by similar rhythms of re- more, 2401 West Belvedere Avenue, Baltimore, MD 21215.
nal plasma flow and GFR. The circadian pattern is Phone: 410-601-6852; Fax: 410-601-9146; E-mail: myrmiller@pol.
maintained in healthy persons until around age 60 net; [email protected]
yr when a shift to a greater proportion of urine pro- Copyright 䊚 2009 by the American Society of Nephrology
In healthy adults, 24-h urine production is usually in the range Nocturnal Polyuria in Multiple System Atrophy
of 1000 to 1500 ml and bladder capacity generally ranges from Multiple system atrophy (MSA) is a central nervous system
approximately 400 to 750 ml. This is associated with a daytime degenerative disease that most commonly occurs in older
voiding frequency of four to five times in men and five to six persons. It affects many areas of the central nervous system
times in women and with rare voiding during normal periods and has central autonomic insufficiency or Shy-Drager syn-
of sleep. There is a suggestion that voiding frequency increases drome as one its components. Patients with central auto-
with age, especially in men. With advancing age, there is a nomic insufficiency have been observed to excrete large
progressive increase in the prevalence of urgency, most com- amounts of urine when recumbent at night, and this noc-
monly caused by detrusor instability. The prevalence of ur-
Table 3. Association between nocturnal urine flow rate,
Table 2. Typical parameters related to circadian water bladder volume, and nocturnal urinary frequency in healthy
excretion in healthy young and elderly adults* young and elderly adults*
Young Elderly Young Elderly
Day Night Day Night Nighttime urine flow rate (ml/h) 35 70
Urine volume (ml/h) 75 35 50 70 Time in bed (h) 8 8
Urine osmolality (mosm/kg) 700 830 510 450 Nighttime urine production (ml) 280 560
Plasma AVP (pg/ml) 1.1 2.0 1.9 1.3 Bladder capacity (ml) 400 200
Plasma ANH (pg/ml) 19 17 40 55 Number of voids during sleep period 0 2
*Data show typical values based on literature and/or author’s experience. *Data show typical values based on literature and/or author’s experience.
Acute kidney injury (AKI), as defined by the precip- proposed that parenchymal loss in the aging kidney
itous decline in GFR, is frequently encountered in directly confers a higher susceptibility to acute
the elderly. The effect of advancing age in decreas- damage, this is not supported by experimental data
ing renal reserve and the associated comorbid con- in which the reduction in renal mass surprisingly
ditions of elderly patients increase the risk for the protected against ischemia/reperfusion injury in a
development of AKI. Although studies describing 5/6 nephrectomy model.7 Thus, cellular and molec-
the incidence of AKI in this population are difficult ular alterations that occur with aging may be more
to compare because the definitions of AKI vary dra- important than simply a loss in nephron numbers.
matically from study to study, it is clear that the Lameire et al.8 showed that, in combination with
elderly are at the highest risk for the development of dehydration, a disturbance in autoregulatory de-
AKI. Indeed, Feest et al.1 showed that there is a fense mechanisms that would normally preserve
three- to eight-fold, progressive, age-dependent in- GFR and renal blood flow can, in the elderly kidney,
crease in the frequency of development of commu- lead to ischemia and AKI. One hypothesis links
nity-acquired AKI in patients older than 60 yr of blunted nitric oxide (NO) production in the elderly
age. The mean age of patients with AKI has in- kidney to an increased risk for AKI.9 For example,
creased by 5 to 15 yr over the past 25 yr.2 Groen- studies in a rat model of gentamicin-induced AKI
eveld et al.3 showed that the age-related yearly inci- show that an increased glomerular NO production
dence of AKI rose from 17 per million in adults seems to protect renal function through its vasodi-
under age 50 yr to 949 per million in the 80 to 89 yr latory effects.10 When old and young animals are
age group. Although all causes of AKI are encoun- treated with equivalent doses of gentamicin, older
tered in this age group, the frequency of prerenal animals show more severe AKI that correlates with
and postrenal etiologies is especially prevalent in a blunted stimulation in NO production.10 Miura et
the elderly.4 Furthermore, elderly patients are more al.11 have also hypothesized that, in addition to al-
frequently subjected to invasive procedures and ex- teration autoregulation, aging tubular cells may be
posure to multiple (and possibly nephrotoxic) more vulnerable to ischemic damage because cellu-
medications and to radiocontrast agents, all of lar antioxidant defenses decline with age as well as
which increase the risk for AKI. the fact that tubular cells have alterations in metab-
olism that render them more susceptible to injury
(such as an accelerated rate of ATP depletion caused
STRUCTURAL AND FUNCTIONAL by mitochondrial alterations).
ALTERATIONS IN THE AGED KIDNEY
In the absence of a specific disease, the kidney un- CAUSES OF AKI IN THE ELDERLY
dergoes age-dependent structural and functional
alterations leading to a significant decrease in renal In the elderly, AKI is often iatrogenic and multifac-
mass, functioning nephron numbers, and baseline torial. Elderly patients show the same spectrum for
kidney function.5 Under normal conditions, these the causes of AKI as the general population. How-
changes can be functionally compensated for by ad-
aptations in renal hemodynamics to maintain a suf-
Correspondence: Mitchell H. Rosner, Division of Nephrology,
ficient GFR. However, in the setting of pathophys- University of Virginia Health System, Box 800133, Charlottesville,
iologic challenges, the older kidney lacks sufficient VA 22908. Phone: 434-924-2187; Fax: 434-924-5848; E-mail:
functional reserve and is more likely to develop [email protected]
clinically relevant damage.6 Although it has been Copyright 䊚 2009 by the American Society of Nephrology
Nocturia is defined by the International Conti- Conditions such as congestive heart failure, ne-
nence Society as the interruption of sleep one or phrotic syndrome, autonomic neuropathy, and ve-
more times at night to void.1 Although nocturia is nous insufficiency lead to interstitial edema forma-
relatively uncommon among younger adults, by 80 tion during the day. Mobilization of the
yr of age, the prevalence rises to 80 to 90% in both accumulated interstitial fluid while recumbent re-
men and women.2 The presence of nocturia dis- sults in nocturia. Obstructive sleep apnea is associ-
rupts sleep, leading to daytime somnolence, depres- ated with excessive atrial natriuretic peptide pro-
sive symptoms, cognitive dysfunction, and a re- duction. Neurologic dieases, such as Alzheimer’s
duced sense of well being and quality of life.3 disease and Parkinson’s disease, are associated with
Moreover, nocturia is associated with an increased alterations in the diurnal secretory pattern of neu-
risk morbidity and even mortality.4,5 rohormones, such as natriuretic peptides and anti-
diuretic hormone. Patients with chronic kidney dis-
ease are unable to maximally concentrate their
PATHYPHYSIOLOGY urine and often must void at night.
In many cases, the cause of nocturnal polyuria is
Although it is commonly assumed that nocturia in undefined. In idiopathic nocturnal polyuria, As-
the elderly is primarily a urologic problem, such plund and Aberg8 suggested that anti-diuretic hor-
thinking is inaccurate. The pathophysiology of noc- mone (ADH) levels, which are typically elevated
turia in the elderly involves the complex interplay of during sleep, are abnormally low in these individu-
several factors.6 Age-related changes in the urinary als. This finding is not universally seen, however,
system and in renal function occur. Sleep itself has particularly among women.9 Furthermore, a rela-
effects on renal function. Sleeping patterns and tive nocturnal deficiency of ADH fails to explain the
sleep architecture change with aging. Finally, dis- altered diurnal excretion patterns of sodium and
ease states and medications may affect the urinary nonelectrolyte solutes that occur among these indi-
system, sleep architecture, and renal function. viduals. In some individuals with nocturnal poly-
Common causes of nocturia in the elderly are listed uria, diurnal variation in GFR is absent or even re-
in Table 1. versed, such that creatinine and sodium excretion
rates are higher at night than during the day.10 Some
investigators suggest that these increases are associ-
SYNDROME OF NOCTURNAL POLYURIA ated with higher night-time BP or the “nondip-
ping” phenotype.11
Nocturnal polyuria is a syndrome where the usual
day to night ratio of urine production is altered.7 In
patients with nocturnal polyuria, ⬎33% of the total
daily urine output occurs at night, although the
daily total urine output remains normal. A careful Correspondence: Dean A. Kujubu, Program Director, Nephrol-
voiding diary, incorporating measurements of ogy Fellowship, Clinical Assistant Professor of Medicine, UCLA
voided volumes, is essential to make the diagnosis. School of Medicine, Kaiser Permanente Los Angeles Medical
Center, 4700 Sunset Boulevard, 2nd Floor, Los Angeles, CA
Common causes of nocturnal polyuria are listed in 90027. E-mail: [email protected]
Table 2.
Copyright 䊚 2009 by the American Society of Nephrology
When dialysis was a scarce resource worldwide, pa- WHAT IS THE LIFE EXPECTANCY OF
tients of advanced age were often excluded. How- ELDERLY HD PATIENTS?
ever, this has changed dramatically. Hemodialysis
(HD) of elderly patients has become routine, just as Unsurprisingly, mortality is considerably higher for
selected nonagenarians and centenarians now un- elderly patients on dialysis than for elderly patients
dergo invasive procedures such as coronary bypass who are not. For example, the actuarial life expect-
or valve repair. As a consequence of the aging of the ancy of a 75-yr-old patient on dialysis is approxi-
dialysis population, new issues have emerged. HD mately 3 yr, as opposed to 11 yr for one not on
in elderly and very elderly patients has important dialysis.1 A Canadian database study from the late
differences from younger patients, and an under- 1990s found that patients older than 75 had survival
standing of these issues is critical in effectively guid- at 1 and 5 yr of 69.0 and 20.3%, respectively, after
ing their care. HD initiation.4 For the very old, such as those 90 yr
Before summarizing the available data, there is and older when starting dialysis, survival is ⬍50%
an important caution that applies to many studies at 1 yr.3
describing elderly patients. Until recently, many However, it is important to note that survival is
studies used a cut-off of 65 yr old (or even younger) influenced strongly by comorbidities such as vascu-
as a definition of elderly, grouping all patients over lar disease and cardiac disease. Once these are taken
that age. However, the medical and social issues are into account, the age of elderly patients is not an
clearly very different for a patient in her late 60s in independent risk factor for increased mortality in
comparison to another in her late 90s. Attention to some models.5 This is consistent with the approach
the age of the study group is important when con- that the characteristics of an individual patient are
sidering how study results might apply to an indi- more important than numerical age. Dialysis
vidual patient. should not be withheld on the basis of age alone if
otherwise appropriate.
All types of hemodialysis access are potentially that promote fistulas—the rate of fistula placement
problematic in elderly persons. Fistulas or grafts in the elderly has been rising. It is uncertain whether
can be difficult to initiate because of atherosclerosis this approach will improve long-term patient sur-
and prior vascular damage and additionally pose a vival or morbidity. There is some evidence that
risk of causing distal ischemia. In patients with more aggressive fistula placement is being accom-
heart failure, the extra cardiac load of an arterio- panied by a greater failure rate.2 Despite this general
venous shunt might raise concern. Last, in some trend, it is important to keep in mind that KDOQI
cases, the maturity time for a fistula may exceed the also advises tailoring the vascular access options for
lifespan of the patient, and therefore, placing one the individual patient.3 As will be detailed below,
exposes the patient to a procedure for no benefit. this is particularly important for the elderly patient
Although catheters are simpler to initiate, elderly in whom the risk/benefit balance of interventions
patients are also at higher risk of death from infec- may be less well defined.
tion or other complications associated with them Before exploring the issues of vascular access for
than are younger patients. elderly patients in more detail, there is one impor-
tant caveat. The definition of “elderly” is not con-
sistent among various studies, making comparisons
difficult. Most studies have used 65 yr old as a cut-
EPIDEMIOLOGY AND BACKGROUND
off. However, this is somewhat unsatisfying, be-
cause the issues facing a patient in his or her 60s may
Despite these disadvantages, most elderly patients
be quite different from those facing an octogenar-
treated for end-stage kidney disease (ESKD) in the
ian. Only recently have some studies begun to ex-
United States are on hemodialysis and therefore re-
amine differences specifically in the “older old.”
quire one of these options. This situation has arisen
because there are limited alternatives. For one, renal
transplantation in patients over 65 is relatively un-
WHAT IS THE PREFERRED ACCESS IN AN
common. Furthermore, peritoneal dialysis is rarely
OLDER PATIENT?
used in the elderly in the United States, representing
only about 4% of the dialysis patients over 70. This
There are no randomized trials to answer this ques-
tiny rate is far less than in other countries and likely
tion definitively. Older patients have shorter life ex-
represents underuse, but it is nevertheless continu-
pectancies and more comorbidities, so one might
ing to decrease.
speculate that there is less opportunity to realize a
The vast majority of these accesses are central
benefit from a fistula. However, the limited data
venous catheters: for patients over 75, the rate of
available from observational studies show a mortal-
catheter use is nearly 85% at the time of dialysis
ity pattern that favors fistulas in the elderly, just as
initiation. One study examining dialysis patients 65
in younger patients. Specifically, in patients over 67
to 75 and ⬎75 yr of age found that even 3 mo after
yr old, fistulas are associated with lower death rates
starting dialysis, two thirds were still using a cathe-
ter.1
KDOQI recommendations for vascular access Correspondence: Seth Wright, Beth Israel Deaconess Medical
do not differentiate between younger and older pa- Center, 330 Brookline Avenue, DA 517, Boston, MA 02215. E-
tients, so under the current “fistula first” ap- mail: [email protected]
proach—including the advent of quality measures Copyright 䊚 2009 by the American Society of Nephrology
The issue of whether—and how—to implement of patients older than 65 yr. However, if there was
peritoneal dialysis (PD) in elderly patients is in- not a contraindication, one third of elderly patients
creasingly important given the rapid growth of this elected to start PD rather than HD.4 Elderly patients
dialysis population. PD has some particular advan- on PD can do quite well: the 2- and 5-yr survival of
tages and disadvantages in the elderly. Further- patients over 65 yr of age in Hong Kong was re-
more, these advantages and disadvantages are not ported to be 88 and 56%, respectively.5
always fully understood by medical providers. Not In comparing PD and HD, one should keep in
only is a better understanding of PD in elderly pa- mind that data quality is limited by the inability to
tients relevant for patient autonomy, medical out- randomize patients. The larger prospective cohort
come, and comfort, but there are systemic implica- studies such as the NTDS have subject numbers in
tions for cost and education as well. the hundreds, whereas registry studies are larger but
presumably are confounded by selection bias. Fur-
thermore, all comparisons are complicated by vary-
BACKGROUND AND EPIDEMIOLOGY ing definitions of what age constitutes “elderly.”
Most use ⬎65 as a cut-off, whereas others use 70 or
In the United States, PD is used less frequently in higher.
elderly patients than in younger patients, and the
rate is declining. In recent USRDS data, 12% of pa-
tients ages 20 to 55 were on PD, whereas only 4% of PHYSIOLOGY OF PD IN ELDERLY
dialysis patients ⬎75 yr of age used this modality.1 PATIENTS
This contrasts strikingly with other countries. In
France, PD is dominant in elderly patients, with The fundamental physiology of PD is not age de-
more than one half of all PD patients being ⬎70 yr pendent; a rich capillary plexus brings blood into
old. In Hong Kong, 80% of all dialysis patients are the peritoneum and filtrate flows across the perito-
on PD, with a median age of 62. The United King- neal membrane into the dialysate. However, there
dom and Canada are intermediate, with 17 and are several physiologic considerations unique to el-
12% of incident elderly dialysis patients treated derly patients that may affect clinical outcomes.
with PD.2 Emerging data suggest that peritoneal mesothelial
The cost of PD is generally less than that of he- cells change during the aging process and may be
modialysis (HD).3 Because elderly patients are the more prone to inflammation. Whether this ob-
fastest-growing segment of the dialysis population, served proinflammatory profile in elderly patients
their relatively infrequent use of PD has financial actually has clinical significance remains untested at
implications. this point.
The reasons for the wide variation in use of PD in In addition to possible age-related changes in the
elderly patients are multifactorial, including finan- peritoneal membrane, elderly patients have a
cial, resource availability, and cultural issues. How- higher incidence of intestinal pathology, including
ever, a particular concern is that unfamiliarity of diverticulosis, bowel perforation, and constipation.
providers with the use of PD in elderly patients All of these can affect the underlying physiology of
leads to a self-perpetuating cycle of underuse. This
is especially of note because, given the opportunity, Correspondence: Seth Wright, Beth Israel Deaconess Medical
many elderly would elect PD. It is not always an Center, 330 Brookline Avenue, DA 517, Boston, MA 02215. E-
option; in one study, it was considered contraindi- mail: [email protected]
cated for medical or social reasons in about one half Copyright 䊚 2009 by the American Society of Nephrology
Peritoneal dialysis (PD) utilization is on the decline perience, 8% of patients were assisted by family
in many regions around the world. There are mul- members and 14% received regular visits by home
tiple potential contributors, but the role of barriers care nurses. For the remainder of this chapter, we
to self-care PD in an aging population is likely an will focus on the role of assistance provided by vis-
important factor to consider. Providing home care iting home care nurses in the management of the
assistance to support elderly persons on PD may elderly persons with end-stage kidney disease
help to overcome these barriers and increase the (ESKD). This specific form of assistance is
number of individuals that can be safely treated in HCAPD.2
the home. This chapter will cover the following is- HCAPD can be a valuable form of therapy for a
sues: (1) what is home care assisted PD (HCAPD); number of reasons. First, the offer of home care
(2) who is a candidate for HCAPD; (3) what are the support can provide reassurance to elderly persons
logistical considerations when providing HCAPD; and their families when considering PD as a treat-
(4) how do patient outcomes on HCAPD compare ment option. In this way, assistance may act as an
with other dialysis modalities; and (5) is HCAPD a incentive to some individuals to choose PD regard-
cost-effective therapy? less of whether it is actually required. Second,
HCAPD is a form of ongoing training or mentoring
that allows patients to gain confidence performing
DEFINING ASSISTANCE PD-related tasks in a supervised setting. Cognitive
impairment is common in the elderly dialysis pop-
Providing assistance to PD patients involves identi- ulation and may make it difficult for patients to
fying and training an individual other than the pa- learn what they need to know to perform PD inde-
tient to perform dialysis-related tasks. These tasks pendently during a traditional training program.
may include connecting the patient to a cycler, set- With ongoing home care support and education by
ting up the cycler, disconnecting from a cycler, or nurses, some patients eventually reach a point over
performing continuous ambulatory peritoneal di- weeks to months to feel comfortable enough to
alysis (CAPD) exchanges. Spouses or other family graduate to self-care PD. Third, HCAPD can act as
members, paid caregivers, or visiting health care a bridge therapy in individuals performing self-care
professionals in the home can provide assistance. PD who develop an intervening illness that makes it
Assisted PD can also be provided by staff in other temporarily difficult to continue on PD. These in-
settings including rehabilitation centers, retirement dividuals can be provided with home care support
homes, nursing homes, and complex continuing for as long as is required. The support can then be
care facilities. The reliance of PD patients on others withdrawn as indicated when the patient recovers.
for assistance in the home is often underappreci- In those that develop permanent barriers to self-
ated and is borne out by studies showing that mar-
ried individuals are more likely to receive PD and
that patients who live alone are less likely to be Correspondence: Matthew Oliver, MD, MHS, FRCPC, Sunny-
treated by this modality. Evidence from the French brook Health Sciences Centre, Room A239, 2075 Bayview Ave-
nue, Toronto, Ontario M4N 3M5, Canada. Phone: 416-480-4755;
PD Registry suggested that one in five patients re- Fax: 416-480-4245; E-mail: [email protected]
quired some form of assistance.1 In the French ex-
Copyright 䊚 2009 by the American Society of Nephrology
alysis, may become more available in the United States in the Integration of the comprehensive geriatric assessment in a
future, as it is practiced in other countries. It would be a very dialysis patient’s records is important for coordination of care.
viable option for patients with self-care limitations who wish to
continue with home dialysis. Decision About Withdrawal of Dialysis
Discontinuing dialysis and entering hospice care is usually a
Transplantation Referral joint decision between nephrologist and geriatrician. Initial
Age is no longer a single limiting factor for kidney transplantation, conversation with the family and patient is best held outside of
and reasons for a nonreferral must be documented by the IDT for the dialysis facility, with the focus on the overall prognosis and
every dialysis patient. In most elderly patients with significant co- palliative care options.
morbidities, nonreferral is quite obviously a correct choice. The nephrologist should reach out to geriatricians early on,
Even a highly functioning older dialysis patient would ben- if an elderly patient is developing increasing difficulties and
efit from a comprehensive geriatric assessment as a part of the discomfort during dialysis treatments, because it takes collab-
transplant evaluation. Functional abilities can dramatically de- orative effort to make a difficult decision to stop a life-sustain-
crease after initiation of dialysis and timely interventions by ing therapy. Institution of time-limited trials may be appropri-
geriatricians could help maintain physical and mental capabil- ate for some patients.
ities of waitlisted older patients.
Elderly persons have special needs and problems that dialysis are optimal choices for periodic geriatric assess-
ment.
often escape recognition and treatment until late in • A comprehensive geriatric assessment provides a systematic
their course. When considering the older individual, it approach to the collection of patient data; allows a patient’s
is important to be able to determine what findings health status to be evaluated for existing and also potential
constitute normal age-related changes compared with problems; emphasizes functional capacity; and does not aim
age-prevalent illness or an atypical and nonspecific to cure all problems but rather identifies and suggests ways
to maximize quality life years.
presentation of disease. Because there are often several • The primary goal is to help the patient regain lost
reasons for a given finding, one should not be too function and maintain as much independence as possi-
quick to conclude the “reason” for a specific issue; ble. It is important to note that physical and mental
there may also be more than one reason for a particu- illnesses affecting the elderly often interact and result in
lar presenting problem. a loss of functional ability much more than any one
problem in itself. The elderly are particularly vulnerable
A well-formulated problem list is the best way to as they have less “reserve capacity.”
unlock the possible causes of each finding whether
it is a key historical fact, physical finding, laboratory There is a great variation in the way that geriatric as-
test abnormality, or some diagnostic test result. A sessments are conducted from single physician evalu-
“change” in what has been a stable problem or an ations with referral to other skilled professionals as
acute functional decline should also alert the physi- necessary to full teams of professionals evaluating all
cian. Because many older persons have multiple in- patients. In the latter case, a team meeting is usually
teracting problems, a team of skilled professionals is held to discuss individual findings with the end prod-
often needed to provide a comprehensive evalua- uct being a unified team care plan. The team may in-
tion to address current and future needs. The neph- clude representatives from medicine, nursing, den-
rologist has the advantage of already having experi- tistry, dietary, social work, audiology/speech, physical
ence caring for persons with chronic illness and and occupational medicine, and psychology/psychia-
multiple interacting problems. Dialysis also in- try, among others. In the case of persons with renal
volves working with a team of skilled professionals, failure who may or may not already be receiving dial-
much like the model used in geriatric medicine. The ysis with a carefully designed care plan, care plans de-
following is a summary of key issues relating to veloped after the Geriatric Assessment must be care-
comprehensive geriatric assessment and its possible fully coordinated to avoid conflicting plans, messages,
benefits.
• The periodic comprehensive geriatric assessment should
be part of a coordinated care plan with the patient’s other
health care providers. Correspondence: Steven R. Gambert, Department of Medicine,
• Individuals of significantly advanced age, those suffer- Johns Hopkins University School of Medicine and University of
Maryland School of Medicine, Baltimore, MD 21201. E-mail:
ing from multiple and interacting medical problems,
[email protected]
those on multiple medications, and individuals meeting
criteria for “frailty” benefit most; elderly persons on Copyright 䊚 2009 by the American Society of Nephrology
Based on recent data, Canadians starting dialysis rehabilitation involves a process by which patients
between the ages of 75 and 79 yr will have an average both learn new ways to restore function but also meth-
life expectancy of 3.2 yr.1 In the United States, pa- ods which help them adapt to the new disability.
tients 65 to 79 yr of age starting dialysis have a re- Geriatric rehabilitation depends highly on a
ported life expectancy of 25 mo.2 During this time, model of interdisciplinary care. In this model, dif-
patients may experience transient or permanent ferent team members have both overlapping and
loss of personal independence, which, in turn, has a complementary skills. Unlike acute care units
negative impact on their quality of life and financial where the physician often heads the team, rehabili-
situation and a significant impact on health care tation teams do not depend on leadership from one
utilization.3 In general, older patients have complex particular discipline. Rather, treatment decisions
medical histories and a higher incidence of chronic are led by the team member most involved with the
ailments such as arthritic pain, vision loss, or fa- patient. Team members include physicians, nurses,
tigue. Often patients adapt by walking more slowly social workers, occupational therapists, physiother-
or taking more rests and developing fixed routines apists, speech therapists, psychologists, and phar-
with little variability or limiting activities. Over macists. Nurses play a key role because they spend
time, these symptoms and adaptations have a cu- the most time with the patient. Through their
mulative effect on functional status, possibly lead- daily interactions with patients, nurses empower
ing to dependency. In the renal literature, prevalent patients to assume self-care and responsibility
patients on hemodialysis seem to have high levels of and evaluate their psychosocial needs. Often
functional loss.4 Preliminary studies show that this nurses help reduce muscle deconditioning by en-
is exacerbated by acute hospitalization.5 The impact couraging patients to perform self-care activities
of a lower functional status may be reduced by of- outside of their formal therapy sessions. After
fering rehabilitation to dialysis patients. In the non- multiple team members assess the impact of dis-
dialysis literature such programs are common and ease on functionality from a variety of perspec-
seem to limit the impact of functional disability on tives, they identify, through discussion, which of
patients, their families, and the healthcare system.6,7 the disciplines is best suited to developing solu-
In this chapter, I will review the benefits and con- tions before working in a collaborative manner.
cepts of geriatric rehabilitation, the role of the As an example, one patient with difficulty walk-
nephrologists, and highlight some common com- ing indoors may work with a physiotherapist to
plications. increase muscle strength, whereas another may
benefit from occupational therapy sessions to
learn to overcome visual limitations.
WHAT IS GERIATRIC REHABILITATION?
in those with a recent loss of function and, second, that Copyright 䊚 2009 by the American Society of Nephrology
PUBLISHED BENEFITS OF REHABILITATION IN and high levels of functional dependence. In our unit, dialysis
DIALYSIS CARE is offered on site on a daily dialysis schedule. Staffing levels
have been increased in both the dialysis suite and on the ward
Numerous programs have reported using rehabilitation in to accommodate the higher burden of comorbidity and higher
their dialysis units. Most develop exercise programs designed dependency levels.24
to build muscle strength through exercise regimens during di-
alysis or within the home.9 –14 Such programs have been largely
successful but are often confounded by high drop-out rates
ISSUES UNIQUE TO REHABILITATION OF ELDERLY
and the high costs of providing staff to supervise the exercises.
DIALYSIS PATIENTS
Cardiac rehabilitation is recommended for dialysis patients
who have recently survived myocardial infarction, had bypass
Comorbidity Burden
surgery, or those with chronic stable angina. Dialysis patients
Elderly dialysis patients have a high incidence of comorbidity
have been shown to benefit from actively participating in car-
with recent studies in prevalent hemodialysis patients showing
diac rehab, with outcome studies showing a 35% reduced risk
the mean number of medical conditions to be around 10.5 ⫾
for cardiac mortality.15 However, cardiac rehab seems to be
3.5.25 The high number of comorbid illnesses place a heavy
underused in dialysis patients, with only 10% of dialysis pa-
medical burden on the rehabilitation unit staff and the phar-
tients, compared with 23% of the general population, under-
macy. Elderly dialysis patients have a high rate of transfer out
going cardiac rehabilitation after coronary artery bypass graft
to acute care for new or recurrent acute illnesses. Of those who
(CABG) surgery.15 Reasons behind this apparent paradox have
not been clearly identified.
In the older individual, building muscle can be challenging, Number of patients
Success rates, as measured by the proportion of patients re- Figure 1. Graph showing discharge disposition of patients ad-
turning home, vary from 20 to 100%. In the largest and most mitted to the Toronto Dialysis Geriatric Rehabilitation Program.
recent report, approximately 70% of patients23,24 met their Patients who did not return to rehabilitation after an acute inter-
personalized goals and returned home (Figure 1). At the time current illness or who transferred to palliative care are shown as
of admission, patients had a significant burden of comorbidity having been discharged to Other.
Nephrologists worldwide are increasingly taking illness. Nephrologists were also providing ongoing
care of older patients. Mean age at the start of renal treatment for comorbid chronic illnesses such as
replacement therapy is 62.3 yr for men and 63.4 yr diabetes and heart disease. In 1993, they went on to
for women. Peak incident counts of treated end- confirm similar statistics in their chronic peritoneal
stage kidney disease (ESKD) occur in the 70 to 79 dialysis patients.10 This would suggest that the
age group at ⬎15,000 patients per year. Peak inci- nephrologist needs to be prepared to take on the full
dent rates of treated ESKD occur in the 70- to complexity of care for their older patients, particu-
79-yr-old age group at 1543 per million popula- larly their chronic dialysis population. In older pa-
tion.1 Patients in this older age group are likely to tients, this would include health maintenance
have multiple comorbidities. The average 75 yr old screening and immunizations. Although malignan-
suffers from 3.5 chronic diseases.2 Chronic kidney cies are more common in both the dialysis popula-
disease (CKD) in the elderly rarely occurs in isola- tion and in the posttransplantation population
tion from other chronic medical conditions and is than in the general population, life expectancy, age,
often a marker for those conditions. Many symp- and cost effectiveness need to be considered by the
toms in older patients are caused by multiple defi- nephrologists before ordering screening tests.
cits and not by a single disease. These diseases and What can the geriatrician offer this very sick
their treatments are likely to interact and compli- population? Geriatricians are trained to perform
cate one another. Murray3 has reported that up to comprehensive geriatric assessment. This involves
70% of dialysis patients 55 yr of age and older have evaluation of all medical problems, but also covers
chronic cognitive impairment of a level severe several other domains— cognition, affect, func-
enough to impact on their compliance and ability to tional level, sensory, socio-economic needs, and en-
make informed decisions.4,5 Prevalence of depres- vironmental needs. Most geriatricians practice in a
sion is reported to be as high as 45% in the older multidisciplinary group that includes social work-
dialysis population.6 – 8 Metabolic bone disease is ers, pharmacists, physical therapists, and nurse
complicated by age-related osteoporosis. The car- practitioners.
diovascular consequences of CKD are complicated This group can help the patients with organizing their
by structural heart disease such as valvular insuffi- medical regimen and transportation to the dialysis unit.
ciency and atrial fibrillation. Neurodegenerative They can provide support and counseling to the patient
disease impacts on the patient’s mobility and cog- and family for coping with chronic disease.
They can advise on cognitive function and whether the
nitive function. Osteoarthritis and neuropathy patient should still be signing their own consents.
limit their physical activity. As age and disease ad- They can take care of the intercurrent illnesses and chronic
vance, frailty becomes an issue. All of these things conditions.
combine to make the care of the older dialysis pa- They can review and manage polypharmacy.
tient much more complex than that of a younger They can help manage pain.
They can coordinate the other specialty care.
individual. Drug interactions and inappropriate Many practices now provide a transitional care clinic for
dosing becomes an increasing issue as the number rapid follow-up on discharge to prevent bounce back to
of comorbidities and medications increases. the hospital.
In 1992, Nespor and Holley9 did a small study of
in-center hemodialysis patients in Pittsburgh.
Eighty percent of these patients did not have a fam- Correspondence: Jocelyn Wiggins, University of Michigan, 1150
ily physician and relied on their nephrologist for all W. Medical Center Drive, 1560 MSRB II, Ann Arbor, MI 48109.
of their medical care. Ninety-one percent sought E-mail: [email protected]
treatment from their nephrologist for minor acute Copyright 䊚 2009 by the American Society of Nephrology
Both older persons and persons with kidney failure 3. Overall, the best treatments are exercise (espe-
are highly prone to develop nutritional deficiencies cially resistance exercise), anabolic hormones (e.g.,
(Table 1). The major nutritional problem experi- testosterone and selective androgen receptor mole-
enced by both groups is weight loss associated with cules), and vitamin D replacement. Creatine to-
protein energy malnutrition. Renal failure patients gether with exercise improves muscle strength in
with low body mass index, weight loss, low albu- older persons. It also reduces cramps. There are no
min, and low cholesterol all have increased morbid- data in renal failure.
ity and mortality.1 These are classical components The decreased testosterone and vitamin D asso-
of the malnutrition, inflammation, and atheroscle- ciated with kidney disease make sarcopenia more
rosis (MIA) syndrome in ESRD. Similarly, in older likely to occur in renal failure. Insulin resistance in
persons, weight loss is associated with increased renal failure further decreases muscle anabolism
mortality. and increases fat accumulation in muscle. With ag-
ing, there is a physiologic anorexia of aging with
older males reducing their caloric intake by a third
WEIGHT LOSS and females by a quarter over their lifespan.3 There
are multiple causes of this physiologic anorexia. Ag-
The causes of weight loss in older persons are as ing is associated with a decline in taste and olfac-
follows2: cachexia, anorexia, sarcopenia, and dehy- tion. With aging there is a decrease in adaptive re-
dration. laxation of the fundus of the stomach, resulting of a
Sarcopenia is the loss of muscle mass that occurs quicker filling of the antrum and early satiation.
with aging. It is associated with varying degrees of This is associated with slower gastric emptying that
muscle power (dynapenia). Severe sarcopenia (de- occurs with large gastric volumes in older persons.
fined as appendicular skeleton lean mass corrected The satiation hormone, cholecystokinin, is in-
for height that is 2 SD below the normal value for creased with aging, and it is more effective at reduc-
young persons) occurs in 5 to 13% of persons over ing food intake in older persons. The reduction in
the age of 70 yr. Sarcopenia is associated with in- testosterone in older males leads to an increase in
creased disability, and its medical costs have been leptin that can reduce food intake and increase en-
calculated to be $18.4 billion per year in the United ergy metabolism.
States. On the whole, fat older persons who main- Renal failure is classically associated with an-
tain muscle mass do fairly well, but those who are orexia because of circulating uremic toxins. In ad-
fat but have lost muscle mass (the “fat frail” or sar- dition, urea in the mouth produces gingivitis, de-
copenic obese) have worse outcomes as far as dis- creasing the enjoyment of eating. Male kidney
ability and mortality than do the thin sarcopenic. failure patients have low testosterone, increasing
There are many causes of sarcopenia (Table 2). the potential of higher leptin levels increasing an-
These include genetic factors, weight at birth, poor orexia.
energy and protein intake, low levels of activity, de-
creased motor units, insulin resistance, decreased
anabolic hormones, low vitamin D, increased cyto- Correspondence: John E. Morley, MB, BCh, Division of Geriatric
Medicine, St. Louis University School of Medicine, 1402 S. Grand
kines, and peripheral vascular disease. The poten- Boulevard, M238, St. Louis, MO 63104. E-mail: [email protected]
tial treatments for sarcopenia are outlined in Table
Copyright 䊚 2009 by the American Society of Nephrology
As our population ages, the number of patients pre- uroflow/urodynamic studies, and cystoscopy. Com-
senting to their primary care physicians with uro- mon transurethral treatment modalities include re-
logic problems is significantly increasing. Urologic section, laser ablation, and microwave or radiofre-
issues are the third most common type of complaint quency therapy.
in patients 65 yr of age or older and account for at There are two major approaches of medical ther-
least a part of 47% of office visits.1 One of the most apy for prostatic outflow obstruction: relaxing the
predominant urologic problems in elderly persons, prostate smooth muscle tissue or decreasing glan-
and the focus of this chapter, is lower urinary tract dular volume. ␣1-adrenergic blockade relaxes the
symptoms (LUTS). There are several disease pro- smooth muscle fibers of the prostatic stroma and
cesses that can lead to LUTS, as well as a number of can significantly improve urine flow. Because ␣
consequences. In this chapter, we will give a brief blockade can also have significant cardiovascular
overview of the major issues as they relate to elderly side effects, ␣1 selective medications were devel-
persons. oped to specifically target the urinary system. Com-
mon nonselective agents include terazosin and dox-
azosin; selective medications are tamsulosin and
BENIGN PROSTATIC HYPERPLASIA AND alfuzosin. 5-␣ reductase inhibitors block the con-
LUTS version of testosterone 3 DHT, which is a potent
stimulator of prostatic glandular tissue. This reduc-
The prostate surrounds the male urethra between tion in local androgen stimulation results in a pro-
the bladder neck and urinary sphincter like a gressive decrease in prostatic volume over a period
doughnut. When the doughnut enlarges, the of 6 mo to 1 yr. There is also a concomitant decrease
doughnut hole can close off and create an outflow in prostate-specific antigen (PSA) level by approx-
obstruction and/or irritative voiding symptoms. imately 50%, necessitating a doubling of the post-
Benign prostatic hyperplasia (BPH) is a condition treatment PSA to compare it with the pretreatment
that affects the majority of elderly men.2 Not all level. Common agents include finasteride and
cases of BPH need treatment. LUTS are assessed dutasteride. The combination of an ␣-blocker and a
with both subjective and objective studies. The 5␣-reductase inhibitor may work synergistically, al-
American Urological Association BPH symptom beit expensively, to improve LUTS.
score was designed to evaluate subjective com-
plaints. Patients are asked a series of questions re-
garding their urination in addition to a “bother PROSTATE CANCER
score.”3 Low scoring patients are advised of helpful
lifestyle changes, median range patients are given With increasing age, clinical prostate cancer be-
the option of medication, and patients with high comes more prevalent. It is estimated that about
scores (or those patients who are very bothered by 10% of patients ⬎65 yr of age have been diagnosed
their symptoms) are offered medication or trans- with prostate cancer. On autopsy studies, the inci-
urethral surgery. Surgery may be indicated for pa-
tients with recurrent/persistent infection, hematu-
Correspondence: Damon Dyche, William Beaumont Hospital,
ria, bladder stones, hydronephrosis, progressive
Department of Urology, 3601 West 13 Mile Road, Royal Oak, MI
renal failure, or acute urinary retention. Urologic 48073. E-mail: [email protected]
work-up is available and includes postvoid residual,
Copyright 䊚 2009 by the American Society of Nephrology
Urinary incontinence is not a normal part of aging. It can be The bladder’s basic line of defense against infection is a healthy
caused by a number of factors including medications, medical mucosa with low pressure storage and complete emptying of
comorbidities, and urologic pathology. A helpful mnemonic urine. Bacteria that may be introduced by poor hygiene, sexual
for the differential diagnosis of temporary or reversible incon- activity, and catheterization are normally flushed out of the
tinence is “DIAPERS”: delirium, infection/inflammation, bladder. When there is bladder obstruction or poor emptying,
atrophic vaginitis, polypharmacy, endocrine (diabetes), re- this defense mechanism is less effective and can lead to coloni-
stricted mobility, and stool impaction. There are four main zation or infection. The most common cause of urinary tract
types of urinary incontinence including urge, stress, overflow, infections (UTIs) in elderly patients are gram-negative organ-
and mixed. Urgency is a subjective feeling of a sudden need to isms. Residents of nursing facilities or patients who have had a
void, and it can be associated with incontinence. Causes in- long hospitalization are at an increased risk for multidrug-
clude infection, stones, medications, tumors, and neurologic resistant organisms such as pseudomonas and MRSA.
pathology. Stress incontinence is the failure of the sphincter to Bacteriuria can be either symptomatic or asymptomatic. All
remain closed during urine storage, because of an increase in patients with symptomatic bacteriuria should be treated with
sence of urinary tract–specific symptoms).4 For those Copyright 䊚 2009 by the American Society of Nephrology
Falls and associated fragility fractures are a major brain, e.g., pulmonary embolus, myocardial infarc-
cause of morbidity and mortality in older persons tion, anemia, stroke, seizures, dehydration, meta-
with kidney disease. In a longitudinal study from bolic abnormalities, and subdural hematoma (Ta-
one dialysis center for a median of 468 d, 47% fell.1 ble 4). Problems with lower limb strength and
The fall incident rate was 1.60 falls per year. Overall balance disorders are common in older dialysis pa-
studies suggest that the fall rate is much greater in tients and treatable with physical therapy. Drugs
dialysis patients than in the general population. In associated with falling are listed in Table 6. Studies
the general population over 75 yr of age, 30% of in older diabetics suggest that falls are reduced
persons fall each year, with one in five having an when HbA1C is not lowered below 7%.
injury. Hip fractures in persons on dialysis occur
three to four times more commonly than in the
general population. One-year mortality in dialysis BP ABNORMALITIES AND FALLS
patients who have a hip fracture is two to three
times of that in older community-dwelling persons Postural hypotension is a major cause of falls. It can
who have a fracture. A single fall in a dialysis patient occur without any dizziness. For this reason, BP
over 65 yr increases the risk of death after adjust- needs to be regularly measured in the standing po-
ment for comorbidities.2 In the first 2 yr after dial- sition. Orthostatic hypotension occurs more com-
ysis, renal transplant patients have a higher risk of monly in the morning, and in an individual with
fracture than patients on dialysis.3 Table 1 lists side severe orthostasis, it may only be present on one
effects of falls. half of the BP measurements. In a group of 23 el-
A community study of fall prevention in Con- derly on dialysis, orthostasis was present in 8 pa-
necticut showed that a simple education program tients before dialysis and 16 of 23 after dialysis.5
(focused on medication reduction and balance and Besides falls, orthostatic hypotension can lead to
gait training) could reduce falls and the need for syncope, myocardial infarction, stroke, and death.
fall-related medical services4 (www.fallsprevention. Causes of orthostatic hypotension include anticho-
org). Falls can be either caused by extrinsic (envi- linergic medications, anemia, prolonged recum-
ronmental) or intrinsic factors. Environmental fac- bency, dehydration, inadequate salt intake, protein
tors include wet, slippery floors, poor lighting, energy malnutrition, adrenal insufficiency, diabetic
uneven surfaces, and stairs. Descending stairs is a autonomic neuropathy, Parkinson’s disease, and
particular risk factor. There are multiple causes of multiple system atrophy (Shy-Drager syndrome).
falls caused by intrinsic factors as shown in Tables 2 Postprandial hypotension (a fall in BP of ⬎20
and 3.4 Specific dialysis-related causes of delirium mmHg) occurs in up to 25% of older persons and
include uremic encephalopathy, dialysis dementia, persons with diabetes. Its nadir is reached 1 to 2 h
Wernicke’s encephalopathy, and dialysis dysequi- after a meal. It is not necessarily associated with
librium. orthostasis. It has also been shown to be present
New onset falls are often caused by delirium. De- during dialysis in nondiabetic patients.6 Postpran-
lirium can present as purely the inability to pay at-
tention. Delirium should be considered as a cause
Correspondence: John E. Morley, MB, BCh, Division of Geriatric
of falling in any patient on dialysis who suddenly
Medicine, St. Louis University School of Medicine, 1402 S. Grand
starts falling. Delirium has multiple causes such as Boulevard, M238, St. Louis, MO 63104. E-mail: [email protected]
drugs, infection, active decrease in oxygenation to
Copyright 䊚 2009 by the American Society of Nephrology
LOSS OF CONSCIOUSNESS AND FALLS Many persons who fall or have disequilibrium develop a “fear
of falling.” Studies in older persons suggest that fear of falling
Persons who fall with loss of consciousness either have seizures puts persons at a marked increased risk of falls and other ad-
or syncope. More than one half of older persons with seizures verse outcomes.
have partial complex seizures (unusual behavior not necessar-
ily associated with toxic clonic seizures) explaining why it can
take nearly 2 yr to diagnose seizures in older persons. Older VITAMIN D
persons with syncope need to undergo carotid sinus massage
because those who develop bradycardia need a pacemaker. A 25(OH) vitamin D (calcidiol) level ⬍75 nmol/L (30 ng/ml)
has been identified as a cause of falls that responds to treatment
Table 2. Causes of falls with a reduction in falls.7 25(OH) Vitamin D deficiency is very
Extrinsic (environmental) common in renal failure patients.8 There is some evidence sug-
Uneven pavement gesting calcidiol is more effective than calcetriol.9
Poor lighting
No grab bar in toilet
Stairs cluttered FALLS IN DIALYSIS
Intrinsic
Low levels of 25(OH) vitamin D There is a paucity of studies examining factors associated with
Orthostasis
falls in dialysis patients. Cook et al.10 found that age, comor-
Postprandial hypotension
bidity, lower predialysis systolic BP, and a history of falls rep-
Medications
Poor vision
Poor balance Table 4. DELIRIUMS mnemonic for multiple causes of
Muscle weakness delirium
Gait problems Drugs/dialysis disequilibrium syndrome
Dementia (poor ability to “dual-task”) Emotional (depression and psychosis)
Depression Low PO2 states (pulmonary embolus, myocardial infarction, anemia,
Loss of consciousness and stroke)
Syncope Infection
Carotid sinus massage Retention of urine and feces
Event monitor Ictal or rejection
Seizures Uremic encephalopathy
Grand mal Metabolic (vitamin B12 deficiency, hypothyroidism, thiamine
Petit mal deficiency)
Partial complex Subdural hematoma
The incidence and prevalence of chronic kidney portant concept because it is believed that interven-
disease (CKD) is increasing worldwide and more so tion at that stage of prefrailty may reverse the frailty
in the elderly persons. Among individuals 70 yr of syndrome.9
age or older, the prevalence of CKD increased from Regardless of the definition, it is important to
38% in 1988 to 1994 to 47% in 1999 to 2004.1 This realize two main things. The first is that frailty de-
noted increase in the number of elderly patients velops slowly and that it may be initiated by differ-
with CKD has been associated with new challenges ent factors such as lack of activity, inadequate nu-
where CKD has become increasingly recognized as tritional intake, stress, or triggered by a disease
an important comorbid condition in elderly indi- process such as CKD. The second is that the pres-
viduals leading to death, cardiovascular events, and ence of frailty is associated with several adverse
hospitalizations.2 health outcomes such as functional decline [wors-
Recently, less recognized challenges, such as ening mobility, activities of daily living (ADL) dis-
functional impairment and frailty, became more ability, recurrent falls, hip and nonspine fractures],
apparent as factors that affect the quality of life and hospitalization, and death.8,10 –12 The recognition of
outcomes in the elderly patients with CKD.3– 6 Fried these two issues highlight the importance of early
et al.5 described an association between CKD on identification of the problem, hopefully at the pre-
one hand, and the development of functional im- frailty period, and intervening in such a way to min-
pairment on the other. Roderick et al.7 further imize the poor outcomes that are associated with
showed that the strength of the association with the frailty syndrome.
measures of morbidity and functional impairment
increase as renal function worsened. Identifying the CKD and Frailty
predisposing factors of frailty in these elderly pa- Background.
tients with CKD, to help prevent and manage it, is CKD has long been recognized to be associated with
being recognized more as crucial to improve their various comorbidities. Rocco et al.13showed more
quality of life and their longevity. than a decade ago that patients with moderate to
advanced CKD have a reduced quality of life and an
increased frequency and severity of poor symptoms
FRAILTY such as tiring easily, weakness, lack of energy, diffi-
culty sleeping, muscle cramps, and easy bruising, as
well as psychologic distress. Subsequently, Shidler
Definition
et al.14 showed higher negative perception of illness
Frailty is a term that has been used for some time
to be associated with higher depression scores and
with variable definitions. Earlier definitions were
lower quality of life, even at early stages of CKD.
vague and described an elderly patient who is lack-
Shlipak et al.3 expanded further on that concept
ing general strength and is susceptible to diseases.
to show that CKD is also associated with poor func-
Fried et al.8 suggested a more precise and standard
definition in which three or more of five compo-
nents (Table 1) would define frailty, whereas the
Correspondence: Emaad Abdel-Rahman, Department of Internal
presence of one or two of these frailty components Medicine, Division of Nephrology, University of Virginia, Box
would be termed “prefrailty.” Prefrail elderly per- 800133, Charlottesville, VA 22908. Phone: 434-243-2671; Fax:
sons are more likely than nonfrail elderly persons to 434-924-5848; E-mail: [email protected]
develop the whole frailty syndrome.8 This is an im- Copyright 䊚 2009 by the American Society of Nephrology
Elderly patients and their caregivers are faced with a ments allow for broad evaluation of overall health
wide variety of complex treatment decisions rang- across many different domains. In addition, be-
ing from the risks and benefits of antihypertensive cause generic assessments are geared toward the
therapy to the decision to pursue renal replacement general population, they allow for comparisons
therapy versus palliative care. There is a great deal of with different groups of patients and interventions.
heterogeneity in health status among elderly pa- However, generic assessments may inadvertently
tients with chronic kidney disease (CKD) and end- bias results toward or against subsets of the general
stage kidney disease (ESKD), and age alone cannot population. For example, a generic assessment with
reliably predict outcomes. In addition to measures an emphasis on physical functioning may suggest
of disease severity and psychosocial function (dis- poorer results for elderly persons or patients with
cussed elsewhere), measures of quality of life (QOL) mobility-limiting conditions such as rheumatoid
and functional status may provide useful informa- arthritis.
tion to aid in prognostic stratification and help In contrast to generic assessments, disease-spe-
guide treatment decisions. QOL and functional sta- cific assessments are designed for patients with a
tus themselves are important outcomes that need to specific disease or undergoing a particular interven-
be carefully considered along with survival when tion. However, disease-specific assessments may
evaluating treatment options. In this context, we not accurately reflect QOL if they lack items per-
review the methods to assess QOL and functional taining to dimensions of the disease that affects
status, and their applications in clinical care, focus- QOL (e.g., an ESKD scale that does not assess pru-
ing on ESKD-related interventions and outcomes. ritus). Another limitation of disease-specific assess-
ments is that because they are developed specifically
for use in a particular disease state or for assessment
HEALTH-RELATED QUALITY OF LIFE of a particular intervention, it is difficult to com-
pare the results to a different population.
Quality of life (QOL) is a concept that most people
intuitively understand, yet is difficult to define pre-
cisely. Most definitions of QOL are centered on the METHODS TO ASSESS HRQOL
notion of health put forth by the World Health Or-
ganization as “a state of complete, physical, mental There are many different types of QOL assessments
and social well-being, and not merely the absence of available. It is important when choosing a HRQOL
disease or infirmity.”1 Health-related QOL assessment tool to understand the context and pa-
(HRQOL) focuses specifically on the influence of tient population in which the tool was developed,
health, illness, and medical treatment on QOL. because this affects the reliability and validity of the
types: generic and disease specific. Generic assess- Copyright 䊚 2009 by the American Society of Nephrology
used in conjunction with the ADL scale.9 The IADL scale eval- ESKD. For example, in one study of ⬎17,000 hemodialysis
uates skills necessary to live independently, including using the patients, a 10-point lower physical component summary
telephone, food preparation, handling finances, and taking (PCS) score on the KDQOL-SF was associated with a 25%
medications. Compared with Katz’s ADL scale, which assesses increased risk of death, similar to the predictive ability of low
basic functions, it is probably more sensitive to early changes in serum albumin.12 In another study of 1000 hemodialysis pa-
functional status. The Rosow-Breslau Health scale was devel- tients, patients with a PCS score below the median (⬍34) were
oped to assess functional status in the elderly.10 The scale as- twice as likely to die as those patients with PCS scores at or
sesses ability to perform physical tasks requiring mobility and above the median.13
strength, such as walking half a mile, climbing up stairs, and Information gathered by HRQOL and functional status in-
doing heavy housework. Like the Lawton-Brody scale, it is struments, combined with other clinical data, can be used to
most appropriate for community-dwelling elderly patients. estimate prognosis. For example, in one study of 292 patients
The Nagi scale evaluates four types of physical activity: pushing starting dialysis, a risk stratification score based on age, func-
or pulling large objects, stooping, crouching, or kneeling, tional status, comorbidity, and planned versus unplanned di-
reaching or extending arms above shoulder level, and writing alysis initiation was used to identify groups with low, medium,
or handling small objects.11 As opposed to the ADL and IADL and high mortality risk.7 Another study of 146 octogenarians
scales, the latter two methods assess specific physical activities starting dialysis categorized patients into risk groups based on
and therefore may be more useful for identifying areas for in- body mass index, functional status, and early versus late refer-
tervention. ral. Patients with a body mass index ⱕ18 kg/m2, a KPS ⱕ40,
and referred ⬍4 mo before starting dialysis had an estimated
83% risk of mortality in the first year, whereas patients meeting
APPLICATIONS IN CLINICAL CARE none of these criteria had an estimated 1-yr mortality risk as
low as 15%.14
Risk Stratification Estimates of prognosis thus facilitate informed decision
Both HRQOL and functional status are strong predictors of making and advanced care planning. Ideally, when discussing
adverse outcomes among patients with incident and prevalent dialysis decision making, it is useful to provide information
Neuropsychiatric disorders such as delirium, demen- high risk. Several ESKD-specific syndromes of delir-
tia, and depression are common yet poorly recognized ium deserve special mention:
causes of morbidity and mortality among elderly per-
sons with chronic kidney disease (CKD) including Uremic Encephalopathy.
end-stage kidney disease (ESKD). Patients with neu- Uremic encephalopathy is a syndrome of delirium
ropsychiatric disorders are at higher risk for death, seen in untreated ESKD. It is characterized by lethargy
hospitalization, and dialysis withdrawal. These disor- and confusion in early stages and may progress to sei-
ders are also likely to reduce quality of life and hinder zures and/or coma. It may be accompanied by other
adherence with the complex dietary and medication neurologic signs, such as tremor, myoclonus, or as-
regimens prescribed to patients with CKD. This chap- terixis. Although rarely used for diagnostic purposes,
ter will review the evaluation and management of de- the EEG shows a characteristic pattern in patients with
lirium, dementia, and depression among persons with uremic encephalopathy.2 The syndrome is rapidly re-
CKD and ESKD. versed with dialysis or kidney transplantation.
Dialysis Dysequilibrium.
DELIRIUM This syndrome of delirium is seen during or after
the first several dialysis treatments. It is most likely
Delirium is an acute confusional state characterized to occur in elderly patients with severe azotemia
by a recent onset of fluctuating awareness, impair- undergoing high efficiency hemodialysis; however,
ment of memory and attention, and disorganized it has also been reported in patients undergoing
thinking that can be attributable to a medical con- peritoneal dialysis and long-term hemodialysis.3
dition, intoxication, or medication side effects. A The syndrome is characterized by symptoms of
diagnostic algorithm based on Diagnostic and Sta- headache, visual disturbance, nausea, or agitation,
tistical Manual of Mental Disorders, fourth edition and in severe cases, delirium, lethargy, seizures, and
(DSM-IV) criteria, the Confusion Assessment even coma. The incidence and severity of this syn-
Method, has a sensitivity and specificity for delir- drome are felt to be declining because of earlier ini-
ium detection ⬎90% (Table 1).1 tiation of dialysis and institution of preventative
measures in high-risk patients (Table 2).
Risk Factors
Delirium is typically precipitated by an acute or Prevention and Management of Delirium
subacute event such as a neurologic disorder In hospitalized patients, preventative measures can
(stroke, subdural hematoma, hypertensive encepha- reduce incidence and costs associated with delirium
lopathy), infection, electrolyte disorder (hypoglyce- (Table 3).4 Once identified, management of delirium
mia, hyponatremia, hypernatremia, hypercalcemia), is aimed at identification and treatment of precipitat-
intoxication (alcohol, drugs, star fruit), or sleep disor- ing factors and management of behavioral symptoms.
der. Elderly patients with cognitive or sensory impair- Pharmacologic therapy is indicated only when delir-
ment or those taking multiple medications are
thought to be most vulnerable for delirium. Typically, Correspondence: Manjula Kurella Tamura, Division of Nephrol-
one or more precipitants exist in a vulnerable patient. ogy, Stanford University School of Medicine, 780 Welch Road,
Elderly patients with multiple chronic diseases, such Suite 106, Palo Alto, CA 94304. E-mail: [email protected]
as elderly patients with ESKD, are thus at especially Copyright 䊚 2009 by the American Society of Nephrology
Evaluation
History taking (ideally from the patient and caregiver) should
focus on the duration and severity of cognitive and behavioral
deficits, as well as use of medications that might interfere with
cognitive function such as antihistamines, antipsychotics, and
Additional Resources
http://www.hospitalmedicine.org/geriresource/toolbox Assessment tools for delirium, dementia, and depression
http://www.americangeriatrics.org/education/depression.shtml Assessment and management tools for depression
http://elderlife.med.yale.edu/public/public-main.php Management tools for hospitalized patients with delirium
http://www.psychiatryonline.com/pracGuide/pracGuideTopic_2.aspx Delirium practice guidelines
Index), functional status (Karnofsky scale, Katz and Copyright 䊚 2009 by the American Society of Nephrology
Figure 1. Dialysis patient disease trajectory. Adapted from ref. 18 (reprinted with permission).
such as frailty, functional disability, and serious comorbidity pairments and geriatric syndromes, decisional incapacity, and
(at least three to four conditions) is a suboptimal dialysis can- no stated prior wishes, health care proxy, or surrogates.
didate because these factors are already significant determi- The reader is referred to the Renal Physicians Association—
nants of mortality in patients on dialysis. Conversely, those American Society of Nephrology (RPA-ASN) shared decision-
patients who are more independent, less frail, and more cog- making guidelines for the accepted nephrology community
nitively intact will have a better prognosis, although still com- standard of care format in dialysis decision making.34 Because
promised in the long term (Tables 2– 4 ). geriatric patients with CKD are a heterogeneous group, the
patient should be stratified into healthy, vulnerable, or frail
based on a geriatric assessment to define and document the
METHOD FOR EVALUATING DIALYSIS DECISIONS IN functional status, presence of geriatric syndromes, comorbid-
THE ELDERLY ity, psycho-social status, and home support system, because
these will impact on prognosis and the ability to process infor-
The elderly patient with CKD being evaluated for dialysis may mation. The “Stages of Aging”1 paradigm described above is
range from a “healthy” cognitively intact and interactive per- comprehensive. For the nongeriatrician nephrologist meeting
son who can verbalize his or her preferences to a “frail” nursing the patient for the first time, the Get Up And Go Test (from a
home patient with multiple chronic conditions, functional im- sitting position, stand without using arms for support and walk
10 ft/3 m and back as quickly as possible) and the Rapid Chair
Rise (stand up from a seated position in a hardback chair with
Table 2. Geriatric susceptibility factors associated with
death in the general dialysis population arms folded) can be considered. In this case, physical frailty is
defined as scoring ⬎10 s for the go test and/or an inability to
Diabetic Nondiabetic
rise from the chair without using the arms. A moderately frail
Age (yr) patient would be unable to complete either test, whereas se-
45–64 1.43 (1.28,1.59)* 1.55 (1.40,1.72)*
verely frail is defined as the inability to complete both.35The
65–74 2.23 (2.00,2.48)* 2.65 (2.39,2.93)*
optimal situation would be a CGA done by the geriatric team
75⫹ 3.10 (2.77,3.47)* 3.91 (3.54,4.32)*
Frailty
and reviewed with the nephrologist.
BMI (kg/m2) The four topics method summarized below is a useful tem-
⬍20 1.38 (1.27,1.49)* 1.30 (1.23,1.38)* plate to address the main components of a dialysis discussion
25 to ⬍30 0.85 (0.80,0.90)* 0.84 (0.79,0.89)* in the elderly (Tables 5 and 6). Each topic is framed by under-
30⫹ 0.80 (0.75,0.85)* 0.81 (0.75,0.87)* lying ethical principles and their associated clinical counter-
Albumin 0.81 (0.78,0.84)* 0.77(0.75,0.80)* parts. Although topic 2 takes precedence, the more topics that
Functional disability can be fully explored and discussed, the better informed and
Inability to ambulate 1.36 (1.22,1.52)* 1.53 (1.36,1.72)* shared decisions will be. Two starter questions to help direct
Inability to transfer 1.46 (1.23,1.74)* 1.15 (0.95, 1.39) the initial flow of information are as follows: (1) what does the
Relative risk with 95% confidence interval).
patient/family need to know from the nephrologist to facilitate
Adapted from USRDS 2002.33
*P ⬍ 0.001. decision making (topics 1 and 3) and (2) what does the neph-
*Comorbid conditions: albumin concentration ⬍35 g/L, anemia, underweight, CHF, diabetes, ischemic heart disease, COPD, cancer,
cerebrovascular disease, and PVD.
rologist need to know from the patient/family to frame possi- method, it is decided to forego renal replacement therapy, a
ble scenarios (topics 2– 4)? detailed and consistent plan of action must be imple-
mented. This approach will minimize feelings of abandon-
ment and hopelessness and foster a safe environment. Fam-
NONDIALYTIC TREATMENT/RENAL PALLIATIVE CARE ilies must feel supported, especially if their loved ones are
cognitively impaired. The decision to forego life-sustaining
A conceptual understanding of extending life versus prolong- therapy has powerful symbolic meaning. Although it might
ing the dying process and the ability to explore this with the make sense medically, families may still struggle as the ure-
elderly patient or decision maker(s) is important for shared mic process unfolds. Consistent demonstration of caring,
decision making. This is the palliative care concept of the “big respect, and concern by the team will soften this challenging
picture.” There is a growing literature on nondialytic treat- process and allow an acceptable end of life experience. The
ment (NDT),17,37–39 suggesting that survival may not be signif- medically treated patient with ESRD will require increasing
icantly different in selected subgroups between those on home services and transition into a hospice system (home,
chronic dialysis compared with patients with stage 5 CKD nursing home with hospice, or inpatient hospice depending
treated without renal replacement therapy. Less hospitaliza- on goals of care and adequacy of symptom control). This is
tions and more patient deaths at home may be possible in those best accomplished through a palliative care plan whose phi-
treated medically and using a multidisciplinary team ap- losophy incorporates a patient- and family-oriented ap-
proach.37 This may provide a more humane and dignified end proach that helps achieve their goals.
of life experience for the frail geriatric patient and their family. Renal palliative care40 – 43 is an agreed on management plan
In an important retrospective study, two groups of elderly pa- to optimize QOL and relieve suffering (pain and symptom
tients with CKD stage 5 predicted to need dialysis within 18 mo management); offered simultaneously with all other appropri-
were followed in a multidisciplinary predialysis care clinic with ate medical therapy; not synonymous with end-of-life or hos-
one group initiating renal replacement (RRT) and the other pice care; not just the absence of dialysis provision; suitable in
treated medically. Although the RRT group had a longer overall dialysis patients also; appropriate in all patients with serious
survival, when the groups were stratified into those with high co- illness (high symptom burden, shortened survival, significant
morbidity (more than three comorbid conditions) or ischemic comorbidity) and includes the following: advance care plan-
heart disease versus not, the survival curves were comparable.17 ning (ACP), a process of ongoing communication to update
If through shared decision making using the principles of prognosis/goals of care/preferences as trajectory of decline
the RPA-ASN guidelines framed within the four topics progresses and end-of-life issues become more prominent; pa-
tient and family support to create a sense of control over pa- regardless of which decision is taken. If the family decides on
tient’s healthcare, relieve potential burdens on loved ones, and dialysis, time-limited or otherwise, ongoing geriatric assess-
strengthen interfamily relationships; and hospice referral ment and palliative care protocols must be put in place. If an
when appropriate (⬍6 mo estimated survival). NDT approach is taken, follow-up for meticulous renal medi-
Renal palliative care incorporates geriatric principles of the cal management and geriatric-palliative care is imperative. In
interdisciplinary team and holistic approach. Active medical both cases, given the disease trajectory, cooperation among the
treatment of renal complications (e.g., fluid/electrolyte disor- renal and geriatric-palliative teams is essential for the best pa-
ders, renal anemia, fluid overload, CKD mineral bone disease) tient and family outcomes. A geriatric CKD action plan pro-
is continued simultaneously with evaluation and treatment of vides a framework for evaluation and management (Table 6).
geriatric syndromes and symptoms (e.g., pain, depression, fa-
tigue, insomnia, pruritus, constipation) to maximize function
and QOL, avoid unnecessary hospitalizations, and hopefully
allow a dignified death at home. TAKE HOME POINTS
• ESRD/dialysis can be viewed as a form of accelerated aging and has the
attributes of a serious progressive illness
CONCLUSION • The elderly patient with CKD facing dialysis decisions should have a
geriatric assessment to stage the functional age and geriatric syndrome
burden
An effective dialysis decision-making approach involves not • In patients with serious comorbidity, functional impairment, and frailty,
only the process leading up to and including the decision (yes, dialysis may not prolong life but might increase symptom burden and
no, maybe, time-limited) but also a clear postdecision plan ultimately suffering