This chapter will review core questions and concepts of chronic kidney disease (CKD) in older adults including physiological changes associated with aging, identification and epidemiology of CKD in older populations, and differences in manifestations and management that directly affect the care of older individuals with abnormal kidney function.
Keywordschronic kidney disease, elderly, geriatric, nephrology
Kidney Structural and Physiologic Changes Associated With Aging, 73
Changes in the Glomerulus, 73
Changes in Renal Vasculature and Renal Tubules, 74
Age-Related Changes in Renal-Active Hormones and Vasodilators, 74
Changes in Glomerular Filtration Rate, 74
Electrolyte Management, 74
Clinical Implications of Age-Associated Physiological Changes, 75
Identification of CKD in Older Adults, 75
Characteristics of Older Adults With CKD, 77
Outcomes Associated With CKD in Older Adults, 77
Management of Specific Comorbidities and Characteristics in Older Adults With CKD, 78
Multimorbidity And Complexity In Older Adults With Chronic Kidney Disease, 81
Management Options in Advanced Renal Disease in Older Adults, 82
Dialysis Initiation in Older Adults, 82
Resources to Inform Productive Discussion, 83
Palliative Support as an Alternative or Adjunct to Dialysis Preparation, 86
This chapter will review core questions and concepts of chronic kidney disease (CKD) in older adults including physiologic changes associated with aging, identification and epidemiology of CKD in older populations, and differences in manifestations and management that directly affect the care of older individuals with abnormal kidney function.
Kidney Structural and Physiologic Changes Associated With Aging
As in all body systems, the kidneys experience significant structural and consequent physiologic changes with advancing age. These changes affect all parts of the kidney, including the glomerulus, tubules, and vasculature with resultant changes in renal function including decreased filtration, decreased vascular responsiveness, and changes in water and sodium balance. Understanding usual, age-related changes in the kidneys can help delineate the difference between usual effects of age versus non–age-related pathology and can support age-appropriate individualization of patient care.
Changes in the Glomerulus
It is long established that prevalence of glomerulosclerosis increases with age, affecting approximately 5% to 10% of nephrons in the fourth decade of life and increasing thereafter. In 1975 Kaplan et al. examined autopsy-obtained renal tissue from people aged <1 to 90 years who died of nonrenal causes (primarily trauma, overdose, and alcohol in younger subjects compared with cancer, liver disease, and pneumonia in older subjects). They observed less than 10% glomerulosclerosis in subjects who died before age 40, compared with an increased but variable rate of sclerosis more than 10% in adults older than age 40 (most marked in adults over age 50). Although the degree of glomerulosclerosis varied over a 10-year age range examined, a linear increase in glomerulosclerosis was generally noted as age increased. Similarly, Rule et al. noted linear increases in the percentage of glomerulosclerosis indicated on biopsy samples from live kidney donors, with as little as 3% in adults under age 30 to as much as 70% in adults over age 70.
The etiology behind increasing glomerulosclerosis in older age is likely multifactorial. Rat models have demonstrated decreased pressure in the afferent arteriole with advancing age, which may result in increased transmission of hydrostatic pressure to the glomerular capsule. Hyalinosis of the afferent arterioles has also been demonstrated in humans. Hill et al. noted that, in kidneys removed due to malignancy, hyalinosis of afferent arterioles not only correlated with increased luminal size but also was strongly associated with hypertrophy and glomerulosclerosis. With progressive loss of functioning glomeruli, compensatory hypertrophy may also occur. Studies have identified a clear relationship between glomerular hypertrophy and the development of glomerulosclerosis.
Concurrent with a decline in glomerular filtration rate (GFR), kidney morphology also changes with increasing age. Like GFR, overall renal mass peaks in the fourth decade of life; at its greatest, renal mass is around 400 grams and declines by up to 25%, or as low as 300 grams in the ninth decade of life. This decline in renal mass is thought to correlate with the overall change in body surface area that occurs with aging and is primarily attributed to a loss of cortical tissue. Overall change in glomerular shape has also been observed in older adults, specifically a more lobular appearance than the spherical glomeruli of younger adults. These increased lobulations may potentially result in less filtration area, contributing to an overall decline in GFR with age.
Changes in Renal Vasculature and Renal Tubules
Significant changes in renal vasculature occur with increasing age, which may not only predispose to increased glomerular pressure but also diminish the ability of glomeruli to autoregulate and adjust renal blood flow during acute events. In addition to afferent arteriolar dilation due to hyalinosis, described previously, anastomoses can form between glomerular capillary loops, and the glomerular basement membrane may show increased folding and thickening with eventual hyalinosis and collapse. Juxtamedullary glomeruli have been shown to develop channels between the afferent and efferent arterioles, leading to glomerular ischemia and eventual aglomerular vasculature. These combined changes to the typical and appropriate vasodilatory and vasoconstrictive capacities of the renal vasculature hinder the ability of the kidneys to autoregulate blood flow. This may lead to an inappropriate or exaggerated decline in GFR, for example in lower blood flow states (intravascular depletion, poor cardiac output), or an inappropriate transmission of higher pressures to the glomerulus (sustained systolic hypertension), fostering and exacerbating glomerular hypertrophy and injury. Renal tubules also experience changes with advancing age, including decreased tubule mass, thickening of basement membranes, and increased atrophy and fibrosis.
Age-Related Changes in Renal-Active Hormones and Vasodilators
In addition to changes within the vasculature, hormonal influences on the vasculature also change with age. Both renin and aldosterone levels decline with advancing age, which is thought to primarily occur due to decreased renin production and release in older versus younger adults. It has been found that older adults have relatively higher basal rates of atrial natriuretic peptide (ANP), which can suppress renin release. Decreases in renin and aldosterone levels compromise the ability of the kidneys to autoregulate via efferent vasoconstriction and appropriate sodium conservation in states of decreased renal blood flow.
Changes in nitric oxide may also contribute to differences in vasodilatory capacity of the vasculature in older adults. Nitric oxide is known to decrease with advancing age, particularly in the renal cortex. Animal models suggest that this may be related to increased oxidative stress with age, leaving fewer cofactors for nitric oxide generation, as well as a decline in L-arginine with age, which is a key for nitric oxide production. Decreased asymmetrical dimethyl arginine (ADMA) also occurs with increasing age and may lead to a decline in nitric oxide synthase, directly affecting the body’s ability to produce nitric oxide. Lower levels of nitric oxide may lead to increased renal vasoconstriction, as well as increased matrix production and mesangial fibrosis.
Changes in Glomerular Filtration Rate
GFR is known to decline with age, and many factors that determine GFR can be impacted by age-associated changes ( Fig. 5.1 ). Davies and Shock first described the decline in GFR with age in 1950 in their hallmark study of 70 men between the ages of 24 and 89 years; using directly measured GFR, they described a decline in GFR by 8 mL/min/1.73 m 2 per decade after age 20. Population-level studies have described peak GFR, which occurs around the fourth decade of life, as being 140 mL/min/1.73 m 2 in men and 120 mL/min/1.73m 2 in women, with a decline thereafter to mean levels around 80 mL/min/1.73 m 2 in the ninth decade of life and 60 mL/min/1.73 m 2 in the eighth decade of life for men and women, respectively. Hoang et al. noted decreases in GFR, renal plasma flow, and the filtration coefficient ( K f) in older versus younger adults. GFR declined by 22% and renal plasma flow declined by up to 28% in adults over age 55 compared with adults ages 40 and younger. They also observed a decrease in both two-kidney and single-kidney K f, which was thought to be most likely related to a decrease in filtration surface as well as an observed decline in permeability.
Loss of functioning glomeruli is associated with increased rates of albuminuria. Early animal studies noted an increase in proteinuria related to increased permeability of the glomerular basement membrane. More recent animal models suggest that increased glomerular size is not necessarily accompanied by increases in podocyte size. Increased markers of podocyte stress occur with prolonged hypertrophy followed by eventual widening of foot processes and consequent increases in proteinuria.
Both sodium and potassium homeostasis have been observed to change with age. Much of this change is hypothesized to stem from differences in the decreased production of and reduced sensitivity to renin and aldosterone in older adults. Overall sodium homeostasis is less reliable in older adults. Epstein described excess sodium excretion when conservation would be appropriate, such as during a state of volume depletion. Conversely, the rate of sodium excretion after a large sodium load may also be slowed, increasing the risk of volume overload. This observation is thought to result from a decreased responsiveness to aldosterone, as well as decreased overall GFR in the setting of fewer functioning nephrons. Potassium excretion is also relatively decreased in older versus younger adults and, like sodium, is hypothesized to stem from a decreased aldosterone response to a potassium load.
Clinical Implications of Age-Associated Physiological Changes
Despite significant changes in renal structure and function, clinically meaningful changes in renal function are not uniform or consistent across older adults. The documented loss of 8 mL/min/m 2 per decade noted by Davies and Shock, starting from GFR of ∼140 mL/min/1.73 m 2 , would still result in a GFR greater than 90 mL/min/1.73 m 2 by the ninth decade, well above the threshold for the current CKD definition. Population studies of GFR over time do show significant decline between the fourth and ninth decades, but the mean GFR even in the oldest age group is ≥60 mL/min/1.73 m 2 (with stage 3 CKD beginning at ≤59 mL/min/1.73 m 2 ). Fliser and colleagues notably described intact renal reserve (defined as increased GFR in response to an amino acid load), even among the adults as old as 90 years, in both those who were healthy and those who were chronically ill. Moreover, in otherwise healthy older adults, hemoglobin and erythropoietin levels typically remain normal, whereas lower erythropoietin levels are a hallmark of CKD in younger adults with abnormal kidney function. Perhaps most striking, in patients included in the study by Rule and colleagues, which demonstrated as much as 73% prevalence of nephrosclerosis in adults aged 70 to 77 years (the oldest age group examined), all had eGFR >60 mL/min/1.73 m 2 because all participants were active kidney donors at the time of this study. The mean eGFR in the oldest age group was 70 ± 7 mL/min/1.73 m 2 . This does not negate the described and documented changes in kidney structure and function with age, but it does place perspective on the clinical ramifications of these changes, particularly in comparison to more severe decrements in GFR that may satisfy the formal definition of CKD.
Identification of CKD in Older Adults
The definition of CKD is age neutral, with the same criteria applied to both older and younger adults. The Kidney Disease Improving Global Outcomes (KDIGO) work group defines CKD as decreased GFR (<60 mL/min/1.73 m 2 ) or other markers of kidney damage (albuminuria >30 mg/day, history of kidney transplantation, urine findings associated with renal disease including glomerular hematuria, or abnormalities on renal imaging studies or pathology); any one criterion present for >3 months establishes the diagnosis of CKD.
The prevalence of CKD is notably high among older adults and increases with age when using the current KDIGO definition of CKD. Twenty-five to thirty percent of adults aged 65 to 79 years meet the criteria for CKD diagnosis, and this increases to as much as 40% to 50% in the oldest age group (adults over age 80). It should be noted that although there was an increase in the reported prevalence of CKD between the late 1980s and the early 2000s, these data come from population level studies often using diagnostic codes. As such, this trend could represent greater CKD recognition rather than a true change in disease frequency, because the time period over which this increase in CKD prevalence occurred coincides with the development of the first GFR estimating equation and institution of automated eGFR reporting by laboratories.
A point of some debate related to CKD prevalence is the severity of identified CKD. The majority all adults over age 18 with CKD have modest reductions in GFR, a pattern that holds true for both older and younger adults. In older individuals, however, the diagnosis of CKD is more likely to be established based on eGFR alone rather than alternative or additional criteria, including albuminuria. In a population of veterans aged 18 years and older, 72% of adults aged 65 years and younger meeting the definition of CKD did so with some contribution from albuminuria, compared with just under one-third of adults over age 65. In that same cohort, 44% of adults over age 60 met CKD criteria by GFR alone, versus less than 25% of adults younger than age 60. Even when present, proteinuria in older adults is less likely to be found in larger amounts (200 mg/g) compared with their younger counterparts. In the National Health and Nutrition Examination Survey (NHANES) cohort, the rate of albuminuria >200 mg/g declined with advancing age, from 38% in adults aged 20 to 54 years, 18% in adults aged 55 to 70, and only 12.9% in adults older than age 70. The relatively lower incidence of proteinuria and comparatively larger role of GFR in identifying CKD among older adults highlights the importance that estimating GFR plays in determining CKD incidence and prevalence for older adults.
The two most commonly utilized GFR estimating equations are the Modification of Diet in Renal Disease (MDRD) study equation and the Chronic Kidney Disease Epidemiology (CKD-EPI) equation. Both equations include an age variable to model the decline in creatinine generation that accompanies normal aging (associated with decreased muscle mass), but the cohorts from which these equations were validated had relatively few individuals >65 years.
Multiple studies have compared the performance of the MDRD and CKD-EPI equations at estimating GFR in older adults, and the CKD-EPI equation generally produces more accurate estimates, including at correctly estimating CKD stage compared with the gold standard of measuring GFR ( Table 5.1 ). The improved diagnostic accuracy of the CKD-EPI equation has important clinical implications; the CKD-EPI equation better predicts mortality and end-stage renal disease (ESRD) compared with MDRD and other equations in adult populations, and subgroup analysis indicates this is true for adults <65 and ≥65 years alike.
|Study||Population||Age (years)||Gold Standard||Outcome of Interest||Result|
|Kilbride et al., 2013||394 community-dwelling older adults in England; all of European ancestry; median measured GFR 53.4 mL/min||Median age 80 (range 74–97)||Iohexol clearance||Accuracy of estimating equation (defined as estimate within 30% of measured GFR)||Accuracy for MDRD, CKD-EPI(cr), CKD-EPI(cys), and CKD-EPI(cr-cys) was 81%, 83%, 86%, and 86%, respectively (MDRD inferior to CKD-EPI[cr] with P =.004)|
|David-Neto et al., 2016||70 older adults with a functioning kidney transplant in Brazil; mean measured GFR 47 mL/min||Mean age 65 ± 4||Creatinine EDTA clearance||Bias and accuracy (defined as estimate within 30% and, separately, 10% of measured GFR)||CKD-EPI(cr) had the best 30% and 10% accuracy at 74% and 34%, respectively ( P < .04 compared with MDRD, BIS, and CG)|
|Lopes et al., 2013||95 community-dwelling older adults in Brazil; mean measured GFR 55 mL/min||Mean age 85 (range 80–97)||Iohexol clearance||Accuracy of estimating equation (defined as estimate within 30% of measured GFR)||Compared with MDRD, CKD-EPI(cr-cys) had greater accuracy at 85.3% vs. 70.5% ( P < .01)|
|Liu et al., 2013||431 older adults in China; mean measured GFR 53.4 mL/min||Mean age 69.9 ± 6.8||99m Tc-DTPA||Median absolute difference between estimating equation and measured GFR||CKD-EPI(cr-cys) had best accuracy with median absolute difference of 10.5 mL/min ( P < .05 compared with CG and MDRD); 30% accuracy not statistically significant between equations|
|Bevc et al., 2011||317 Caucasian older adults; mean measured GFR 34.5 mL/min||Mean age 72.7 ± 5.1||Creatinine EDTA clearance||Diagnostic accuracy as defined by area under the ROC curve||No statistically significant difference in diagnostic accuracy between MDRD and CKD-EPI at measured GFR ≤60 mL/min; CKD-EPI (cr-cys) performed better than MDRD or CKD-EPI(cr) at measured GFR ≤45 mL/min ( P < .013)|
|Stevens et al., 2010||Pooled dataset of 3896 people across 16 studies; mean measured GFR 68 mL/min||Mean age 50 ± 15; 15% of sample >65 ( n =568)||A variety of exogenous filtration markers||Bias (reported as measured GFR – eGFR)||Though bias was significantly improved for the whole population, subgroup analysis for participations >65 years demonstrated no significant improvement in bias using CKD-EPI(cr) compared with MDRD; cystatin C was not examined|
|Fan et al., 2015||805 older adults enrolled in the community-based Age, Gene/Environment Susceptibility (AGES)-Reykjavik Study; mean measured GFR 64 mL/min||Mean age 80.3 ± 4.0||Iohexol clearance||24 comparisons of measured bias, precision and accuracy between three CKD-EPI equations, BIS equation, and CAPA equation||CKD-EPI equations performed better than other equations in 9 metrics, similar in 13 and worse in 2; among CKD-EPI equations, the combined CKD-EPI equation performed better than eGFR cr in all 4 metrics examined, better than eGFR cys in 2 and similar to eGFR cys in 2|
One concern raised by creatinine-based estimating equations of particular relevance to older adults is the dependence of creatinine on muscle mass. Because there is a predictable decline in muscle mass with age, and because older adults are more likely to have comorbid conditions that contribute to decreased muscle mass, the generation of creatinine can be quite heterogeneous across individuals and demographic characteristics. Cystatin C is an alternative endogenous biomarker filtered freely at the glomerulus that is thought to be less dependent on muscle mass and demographic characteristics, thus providing an opportunity for improved GFR estimation. Indeed, compared with creatinine, cystatin C is a stronger predictor of risk of death and cardiovascular events in older adults. Modifications of the original CKD-EPI equation that use cystatin C alone or in combination with creatinine have been developed for clinical practice and have been shown to provide more accurate estimates of renal function. Fan and colleagues examined the performance of the three CKD-EPI equations (creatinine alone, cystatin C alone, or the two in combination) in comparison to measured GFR using plasma clearance of iohexol among 805 older adults with a mean age of 80.3 years. The combined CKD-EPI equation performed better than the creatinine-based CKD-EPI equation in all four metrics examined; better than the cystatin C-based CKD-EPI equation in two metrics; and similar to the cystatin C-based CKD-EPI equation in two metrics. Given the wide clinical use of creatinine, it remains the recommended tool for estimating GFR, but KDIGO suggests the addition of cystatin C to creatinine when there is uncertainty about the diagnosis and need for confirmation of CKD.
Recently the Berlin Initiative Study (BIS) developed two novel equations utilizing creatinine (BIS-1) or the combination of creatinine and cystatin C (BIS-2) for estimating GFR that, in contrast to other equations, were specifically derived from a cohort of older adults ( n = 610, mean age 78.5 years). Despite internal validation that performed well compared with measured GFR and resulted in less CKD-stage misclassification than the CKD-EPI equation, the BIS equations demonstrated increased bias at age <80 years and higher GFR when externally validated in two cohorts. Currently, KDIGO only recognizes the MDRD and CKD-EPI equations as recommended tools for estimating GFR, though it must be noted that the BIS equations were derived after the most recent KDIGO Practice Guideline on this topic. The Cockcroft-Gault formula has not been validated for use with standardized creatinine assays and, as for all adults, its use in estimating GFR is not recommended for older adults.
Characteristics of Older Adults With CKD
Older adults with CKD are likely to experience a high burden of additional comorbid diseases. Approximately 75% of older adults with CKD have two or more comorbid conditions; in one study, 74% of older adults with CKD had four or more comorbidities. Hypertension is the most common condition associated with CKD in older adults (prevalence ∼90%), followed by dyslipidemia (39% to 50%), diabetes mellitus (21% to 45%) and coronary artery disease (21% to 27%). The prevalence of coronary artery disease is almost three times higher in CKD patients ≥75 years versus those ≤45 years, consistent with the linear relationship observed between aging and cardiovascular disease. Frailty, a state of non–disease-specific decline characterized by the presence of any three of five measures (weight loss, self-reported exhaustion, poor grip strength, slow walking speed, and low physical activity), is three times as likely among older adults with CKD as it is among those with normal renal function. This association remains significant even after adjusting for other comorbidities and demographic factors. The prevalence of anemia is increased in community-dwelling older adults with severe CKD (eGFR <30 mL/min/1.73 m 2 ) compared with older adults with normal renal function, and associates with decreased quality of life. CKD is also associated with cognitive impairment as determined by performance on the Mini-Mental State Examination in older adults. Taken together, these findings underscore the importance of recognizing the presence of CKD in older adults and, when appropriate and in line with goals of care, comprehensively evaluating for and managing associated conditions.
Outcomes Associated With CKD in Older Adults
The relatively modest reductions in estimated eGFR as well as less frequent proteinuria have stirred debate about the accuracy and clinical relevance of the current standard for CKD definition among older individuals. Differences in outcomes associated with abnormal renal function in older versus younger adults have further stirred that controversy.
Older adults with CKD have significantly increased risk of morbidity and mortality. In a community-based prospective cohort study of 4893 adults ≥65 years, Manjunath and colleagues identified an inverse linear relationship between eGFR and incident cardiovascular disease (myocardial infarction, coronary angioplasty, coronary artery bypass graft, angina pectoris, heart failure, cardiac death, peripheral vascular disease, stroke, and transient ischemic attack); for every 10 mL/min/m reduction in eGFR the risk of cardiovascular disease increased 5%. Risk of hospitalization is also increased among older adults with severe CKD (eGFR <30 mL/min/m 2 ); older individuals with CKD are 60% more likely than those with normal renal function to be hospitalized at least twice over a 2-year period, even after adjusting for additional cardiovascular disease risk factors. Although the risk of AKI is increased among those with CKD compared with those without CKD, this risk does not appear to be different for older versus younger adults.
Older adults represent nearly 40% of all prevalent ESRD patients in the United States and nearly half of all new ESRD cases each year. This presumably stems from the profound prevalence of CKD in older adults. However, older adults with abnormal renal function actually experience a decreased risk of progression to ESRD than do their younger counterparts with comparable levels of eGFR. Data suggest this difference may stem more from an age-dependent increase in mortality, compared with a true difference in CKD pathology in older versus younger adults. For instance, the Cardiovascular Health Study (CHS) determined that, in a cohort of community-dwelling adults ages 65 years and older, participants were 13 times more likely to die of any cause than progress to ESRD and six times more likely to die of cardiovascular causes than progress to ESRD over almost 10 years of follow up. The importance of these competing risks in older adults highlights the need to explore patient goals and preferences to individualize CKD-related care in the context of each patient’s preferences and expectations.
Older adults experience an increased risk of death compared with their age-matched peers with normal renal function. Data from the United States Renal Data System (USRDS) indicate that the risk of death is more than twice as high among older adults with CKD than those without CKD after adjusting for age, sex, and race (111.2 deaths compared with 45.2 deaths per 1000 patient-years at risk among older adults with and without CKD, respectively). The risk of death associated with CKD varies, however, by severity of renal dysfunction. In particular, the risk of death is attenuated among those older adults with moderate reductions in GFR (eGFR 45 to 59 mL/min/1.73 m 2 ). In older individuals, the risk of death with moderate reductions in GFR is only slightly greater than that in their age-matched counterparts with GFR >60 mL/min/1.73 m 2 . The risk of death associated with CKD also varies by CKD trajectory (the rate of renal function decline); older adults with a steeper decline in renal function (more rapidly progressive CKD) experience an increased risk of death compared with their peers with more stable (less progressive) renal disease. The trajectory of renal function decline can help identify older adults who may benefit from more frequent follow up and discussion regarding goals of care.
Management of Specific Comorbidities and Characteristics in Older Adults With CKD
Ever since publication of the landmark Veterans Affairs health studies in the 1960s, evidence from randomized controlled trials has highlighted the effect that treating high blood pressure has on reducing the risk of cardiovascular complications and death. Although these data have been reliably replicated across more diverse populations, this cornerstone concept of hypertension management was not immediately applied to the care of older adults because the populations examined in these early trials were not elderly (mean ages ranged 50 to 52). Moreover, long established dogma held to the notion that higher blood pressures, particularly systolic blood pressures (SBPs), were acceptable and perhaps even preferable for perfusion of older organs. As a result, customary practice was to adopt the tenet that SBP values up to “100 + age”—that is, 100 mmHg plus 1 mmHg per year alive—were clinically acceptable. This premise was disproven, however, by the end of the 20th century when randomized trials demonstrated that a reduction in both systolic and diastolic blood pressure decreased the risk of cardiovascular disease and stroke even in adults over the age of 60. Further, in 2008, investigators from the Hypertension in the Very Elderly Trial (HYVET) demonstrated that treatment of blood pressure, compared with no treatment, led to lower risk of cardiovascular disease, stroke, and death even among the oldest age group (veterans aged 80 and older).
Available hypertension guidelines through the seventh issue of The Joint National Committee (JNC-7) recommended an age-neutral blood pressure target of less than 140/90 mmHg, with a lower target for those patients who also experienced diabetes or kidney disease (<130/80 mmHg). The generalizability of these lower targets to older adults, however, remains unclear. Studies involving adults over age 60 who demonstrated reduced rates of stroke and major cardiovascular events in the treatment arm generally achieved SBPs greater than 140 mmHg and less than 160 mmHg (compared with higher values in the usual care or higher SBP goal arms). In contrast, studies of older adults who achieved SBPs less than 140 mmHg compared with higher values reported more mixed results without consistent across-study reduction in stroke or major cardiovascular events. Historically, data from trials of younger adults were extrapolated to older individuals, and hypertension management guidelines remained age-neutral. The JNC’s eighth guideline was the first to recommend a higher blood pressure target (<150/90 mmHg) for adults over age 60, and a lower target (<140/90 mmHg) for younger adults, based on available trial data specific to adults over age 60. These recommendations were echoed in subsequent recommendations by the American College of Physicians and American Academy of Family Physicians, which were based on a large evidence review and metaanalysis of data in older adults.
The applicability of these guidelines to older adults with CKD remains unclear. The vast majority of trials of hypertension management in older adults have explicitly excluded individuals with moderate to severe kidney disease. Importantly, most trials used creatinine rather than eGFR to identify CKD and exclude study participants. Because relatively “normal” creatinine values (correlating to eGFR closer to or >60 mL/min/1.73 m 2 ) are more likely to correlate with mildly abnormal eGFR in older adults with lower muscle mass, adults with mild CKD (stage 3, eGFR 45 to 59 mL/min/1.73 m 2 ) were less often explicitly excluded. Adults with eGFR less than 45 mL/min/1.73 m 2 , however, and certainly less than 30 mL/min/1.73 m 2 in men and less than 25 mL/min/1.73 m 2 in women, were not included in most hallmark trials of hypertension management in older adults. For this reason, hypertension targets for older adults with CKD have been generalized from trials of younger CKD populations. Studies by Shulman et al. and Walker et al. in 1989 and 1992, respectively, were akin to the early Veterans Affairs Cooperative trials in that these studies demonstrated that compared with no control, “some” blood pressure control could slow renal function decline ( Table 5.2 ). Subsequent studies evaluating lower blood pressure targets (SBP ≤130 mmHg) compared with higher targets (SBP >130 mmHg) did not show a consistent benefit of more intensive blood pressure control in slowing kidney disease progression. One exception to these findings was the subgroup of patients in the MDRD study who had more than 1 gram of proteinuria per day, among whom blood pressure less than 130/80 mmHg slowed eGFR decline. The mean age across all of these studies, however, was 46 to 54 years, and no patients over age 70 were included.
|Trial||Population||Age (years)||BP Target by Treatment Group (mmHg)||Medications Used||(mmHg)||CKD||Outcomes|
|Shulman et al., 1989||10,940 patients with hypertension||Range 30–69 |
|Stepped vs. referred care |
Goal: DBP <90 or 10 below baseline (if 90 at baseline)
|Not reported||Subgroup: patients with creatinine >1.5 and >1.7 mg/dL (∼200 patients)||More rapid decline in renal function for those with higher vs. lower BPs|
|Walker et al., |
|5524 men with DBP >90 mmHg at baseline||Range 35–57 |
|Usual BP care vs. “special” care (stepped approach)||Diuretic |
|<140 mmHg in special care |
Remainder 150–159 mmHg and >160 mmHg
|Unclear: excluded if creatinine >2||More rapid decline in renal function for those with higher vs. lower BPs|
|Klahr et al., |
|840 pts total; 585 patients with elevated creatinine in BP arm||Range 18–70 |
|Low BP (<125/75) vs. usual care (<140/90)||Any allowed.Encouraged ACE inhibitor as first line, CCB as second line||MAP 92 low BP group vs. ∼98 usual care||All had CKD ||Lower blood pressure significantly slowed GFR decline ONLY in subgroup of those with proteinuria >1 g/d|
|Wright et al., 2002 |
|African Americans with DBP ≥95, GFR 20–65||Range 18–70 |
|BP usual care (MAP 102–107) vs. low BP (MAP ≤92)||Metoprolol |
|BP 128/78 mean low BP group vs. 141/85 usual BP group||All had CKD||No significant difference GFR slope or composite of GFR decline/ESRD/death|
|Men and women with nondiabetic nephropathy and persistent proteinuria||Range 37–69 |
|<130/80 vs. diastolic <90||Ramipril |
|130/80 vs. 134/82||All had CKD (nondiabetic nephropathy with proteinuria)||No significant difference in ESR or change in eGFR|
Observational data of older adults with CKD have consistently demonstrated a U-shaped association between SBP and mortality, with risk of death increased at very low (SBP <120 mmHg) and very high (SBP >160 mmHg) thresholds. The risk of residual confounding persists in these studies, particularly as relates to impact of other diseases, which both lower blood pressure and increase risk of death (e.g., heart failure, cirrhosis, severe malnutrition), and causation cannot be established from this work. More recently, the Systolic Blood Pressure Intervention Trial (SPRINT) has added to these data. SPRINT evaluated the effect of very low (SBP <120 mmHg) versus more moderate (SBP 140 mmHg) blood pressure control on risk of cardiovascular events, stroke, and mortality. SPRINT was terminated early when participants in the SBP less than 120 mmHg arm were found to have a 25% reduction in all-cause mortality and 30% reduction in cardiovascular events compared with participants in the higher blood pressure arm. SPRINT oversampled for both older adults and adults with CKD, and found no difference in the reduction in mortality or cardiovascular events for their CKD subgroup or adults over age 75. Individuals in the more intensive treatment arm did experience an increased risk of adverse events and severe adverse events, including hypotension, syncope, and electrolyte abnormalities (categorized as severe when emergency care was needed). The overall population of SPRINT notably did not reach required numbers to power for statistical significance due to the earlier than anticipated stop date, and subgroup analyses were not of sufficient size for powered analyses. The population included in SPRINT was also specific to adults with increased cardiovascular risk (defined by age >75, presence of clinical or subclinical cardiovascular disease, presence of CKD, or Framingham score 10-year risk of cardiovascular disease of at least 15%). Overall, these data add to our understanding of the effect of low versus moderate blood pressure targets in this population, and suggest that lower blood pressure targets even in older adults with CKD may result in benefit in terms of reduced cardiovascular disease and death, with the potential countering cost of increased adverse effects.
Perhaps the greatest challenge in examining available trial data on hypertension management in older adults is how to apply those data to individuals who would have been excluded from those trials. Adults with certain comorbid conditions are often excluded from clinical trials due to the concern that those additional comorbidities may cofound interpretation of the efficacy of the therapy or management strategy being evaluated. Older adults with dementia, poor functional status, and those who reside in skilled nursing facilities are explicitly excluded from many studies of hypertension management in older adults, presumably due to the concern that these conditions would directly interfere with an individual’s ability to adhere to trial regimens and trial-related appointments ( Table 5.3 ). The end result is the absence of trial data specific to those older adults with CKD and hypertension who providers may see as most vulnerable, or the most likely to have competing risks associated with more aggressive blood pressure control. Moreover, for patients who experience multiple comorbidities and potentially significant symptom burden from multiple comorbidities, outcomes including quality of life, independence, and functional status may hold greater import than traditional outcomes in trials of hypertension management such as mortality and composite cardiovascular events. These less traditional and less disease-specific outcomes have not been the primary focus of major hypertension studies in older adults ( Fig. 5.2 ). For those older adults with CKD who would not have qualified for inclusion in clinical trials based on frailty, poor functional status, or other markers of fragility, providers must incorporate and interpret available trial data in concert with specific patient characteristics and goals to provide a more tailored and individualized approach to blood pressure management.