The decline in glomerular filtration rate (GFR) that occurs with aging has been a source of considerable controversy, with the debate centering on two questions: Is such a decline an integral, predictable, and inevitable part of aging, also called senescence, or is it an unpredictable manifestation of the disease processes that commonly accompany senescence?
In this article, I will attempt to support the hypothesis that the decline in GFR is a normal and expected phenomenon that does not in and of itself confer any selective disadvantage upon the individual unless other diseases are superimposed.
Beginning with the pioneering work of Nathan Shock and his colleagues in 1950,1 many studies of glomerular filtration rate in aging, presumably “normal,” community-living subjects have been conducted.
Not surprisingly, most of these studies have been cross-sectional in character, with only a very few carried out longitudinally in individual aging subjects (both “normal” individuals and those with superimposed disease).
Without exception, these cross-sectional studies have shown measured GFR (mGFR) to decrease with advancing age in “normal” individuals.
The measurement of GFR has been the time-honored method of assessing renal function for at least the past 75 years,2 with inulin clearance regarded as the gold standard for making the measurement.
Unfortunately, though, measuring inulin clearance is too cumbersome for routine clinical use. Many substitute methods have been developed, but their added convenience comes at the expense of some precision.
For example, formulas or equations can give an estimate of measured GFR, but the accuracy and precision of these estimated values are rather poor, especially at higher levels of mGFR.3
Rate of Decline
In most of the cross-sectional studies, the GFR descent began around age 30 to 40. It appears in both males and females.
The average rate of decline varies, but it averages about 0.8 mL/min/1.73 m2/ year after age 30. Some have suggested that the decline accelerates after about age 65 to 70.1,4,5
In terms of GFR ranges for specific ages, an average 85-year-old male would be expected to have a glomerular filtration rate around 55–60 mL/min/1.73 m2, depending on his GFR at age 30.
In the population-based, cross-sectional Nijmegen Biomedical Study, conducted in the eastern part of the Netherlands, the 5th, 50th, and 95th percentiles of eGFR for non-diseased, Caucasian males 85 or older were 35, 65, and 92 mL/min/1.73 m2, respectively, compared with 36, 61, and 78 mL/min/1.73 m2, respectively, in non-diseased, Caucasian females 85 or older.6
GFR decline is independent of hypertension or cardiovascular impairment, occurring even in indigenous native societies where hypertension is completely absent.7
Cross-Sectional Versus Longitudinal
Thus, based on cross-sectional studies in a broad range of ages, it would appear that a decline in glomerular filtration rate is part and parcel of aging and can be considered one of the many biological manifestations of senescence.
However, cross-sectional studies suffer from a bias that is introduced when only highly selected survivors (those who reach old age by virtue of some characteristic [selective mortality] or as a result of some unique environmental influence limited to that age cohort) are studied.
To some extent, these objections to cross-sectional analyses can be overcome by the longitudinal study of serial observations of renal function (creatinine clearance or eGFR) in individual subjects.
The validity of longitudinal studies is frequently compromised, however, by the limited periods of observation and the lack of frequent testing—sometimes time is too short or tests too few to determine whether there are statistically significant changes in GFR over time.
Results of the very few longitudinal studies of glomerular filtration rate with aging have been used to challenge the notion of GFR decline as an inevitable consequence of aging.8,9
The classic studies by Rowe et al10 and Lindeman et al,11 conducted as part of the Baltimore Longitudinal Study of Aging and reported in 1976 and 1985, respectively, are frequently quoted.
Both studies used true creatinine clearance (Ccr) as a surrogate for glomerular filtration rate. Such clearances overestimate GFR as measured by inulin clearance by about 22%, and the degree of overestimation shows a slight, but not significant, change with advancing age.
Rowe et al
Rowe et al conducted both a cross-sectional analysis and a longitudinal study of GFR decline with aging.10 In the cross-sectional study, the average decline in creatinine clearance was about 45 mL/ min/1.73 m2 for individuals between age 30 and 80 (0.9 mL/min/1.73 m2/year).
The longitudinal study included 293 normal, nondiabetic men age 17 to 96. Three or more serial true creatinine clearance determinations were made over a mean interval of six years, which clearly was inadequate to estimate the true value for the slope of the creatinine clearance over time.
The authors posited that to define an individual's slope of creatinine clearance over time with “minimally acceptable accuracy,” annual testing would have to be performed over 18 years in normal subjects.
Nevertheless, the limited longitudinal data available showed good concordance with the cross-sectional data.
Rowe et al concluded that the age-related decline in creatinine clearance was likely to be a physiologic and pathologic consequence of senescence.
Lindeman et al
Nine years later, Lindeman and colleagues11 published their longitudinal studies of renal function and aging, using many of the same subjects studied by Rowe and colleagues.
Endogenous true creatinine clearance was studied serially in 254 “normal” men age 22 to 97. Patients were defined as normal if they did not have hypertension or receive treatment for hypertension; did not have an edematous disorder, including congestive heart failure; and did not have known renal disease or urinary infection.
Patients with benign prostatic hypertrophy or prostate cancer were included in the normal group if there were no signs of obstruction. Subjects with overt diabetes or abnormal glucose tolerance tests were also included in the normal group if there was no proteinuria.
Longitudinal studies of creatinine clearance were carried out at 12- to 60-month intervals. Analysis was limited to those who had five or more serial creatinine clearance studies (maximum of 14) over periods ranging from eight to 14 years.
Cross-sectional analysis of these normal subjects demonstrated the now familiar decline in creatinine clearance with aging, which averaged 0.75 mL/min/ 1.73 m2/year.
Most noteworthy in the longitudinal studies was that the frequency distribution of the slope of creatinine clearance versus time was normal, suggesting that the decrease is due to a “progressive involutional process” affecting all individuals rather than one due to a “superimposed but undetected” disease affecting a subset of individuals.
Among the 254 “normal” subjects, 92 (36%) showed an increase in creatinine clearance over time of up to 5 mL/ min/year, but this rise was only statistically significant in two individuals.
On the other hand, creatinine clearance decreased over time at a rate of up to 8 mL/min/year in 162 subjects, with 35 (21.6%) individuals having a statistically significant decline. Hypertension did not influence the rate of decline.
If only statistically significant slopes are examined, one could conclude from this study that stable or increasing creatinine clearance with aging is uncommon in normal subjects, including those with diabetes or abnormal glucose tolerance. If all slopes are examined, one could conclude that stable or increasing creatinine clearance is more common.
This study is confounded by the inclusion of individuals with diabetes in the normal group, as hyperglycemia is known to contribute to an increase in glomerular filtration rate (hyperfiltration), especially in subjects without proteinuria.
The analysis is further confounded by the fact that few, if any, normal subjects had annual determinations of their creatinine clearance over the 18-year period suggested by Rowe et al as the minimum duration needed to examine statistically significant changes in creatinine clearance with aging.
It seems unlikely that a changing tubular handling of true creatinine with aging influences the decline.
In my opinion, the study by Lindeman et al11 does not conclusively demonstrate that a decline in glomerular filtration rate with aging is avoidable. The results are also consistent with the view that a decline in GFR with aging is part of the normal biological process of senescence.
While the actual rate of GFR decline can be influenced by superimposed disease processes, such as atherosclerosis, diabetes, or congestive heart failure—most certainly these diseases have effects on GFR that sometimes can be profound—even in the absence of these superimposed diseases, GFR relentlessly declines with aging.
The decline of GFR with aging is accompanied by morphological changes in the kidneys that appear to be universal.
These changes include loss of cortical and interstitial volume, global glomerulosclerosis, glomerular involution and disappearance (absorption), and loss of filtration surface area.12,13 Pathologic changes may differ according to the presence and type of superimposed diseases.
Implications for Diagnosis
The concept that a loss of glomerular filtration rate is a normal phenomenon closely connected with physiologic organ and cellular senescence has important consequences for diagnosis and treatment.
Chronic kidney disease (CKD) is often diagnosed and staged using the National Kidney Foundation Kidney Disease Outcomes Quality Initiative (NKF KDOQI) guidelines, published in 2002, which are based on eGFR and evidence of kidney damage (e.g., proteinuria, glomerular hematuria, abnormal imaging, abnormal renal biopsy) lasting for three months or longer.14
Stage 3 CKD is currently defined simply as an eGFR between 30 and 59 mL/min/1.73 m2, with no other criteria required.
Thus, many individuals over age 65 may be labeled as having CKD even though their GFR is within the normal range for their age and gender. Most of these individuals will have eGFR values between 45 and 59 mL/min/1.73 m2 (also now known as Stage 3A CKD in the United Kingdom).
Any individual, regardless of age, who has an eGFR less than 30 mL/min/1.73 m2 almost certainly has some form of kidney damage, and, in the absence of effective therapy, is at high risk for progression to end-stage renal disease (ESRD) and the complications of cardiovascular disease.
In addition, a decline in GFR with aging results in altered handling of drugs that are eliminated by the kidney. This needs to be taken into account in drug dosing in the elderly.
It is my opinion that normal physiological senescence is associated with a steady and probably inexorable decline in GFR. The mechanisms underlying this change seem to operate at the capillary level. Recognition of this decline is vital for accurate identification of CKD by measured or estimated glomerular filtration rate.
1. Davies DF, Shock NW. Age changes in glomerular filtration rate, effective renal plasma flow, and tubular excretory capacity in adult males. J Clin Invest
2. Smith HW. The Kidney: Structure and Function in Health and Disease
. New York: Oxford University Press; 1951:49.
3. Botev R, Mallié JP, Couchoud C, et al. Estimating glomerular filtration rate: Cockcroft-Gault and Modification of Diet in Renal Disease formulas compared to renal inulin clearance. Clin J Am Soc Nephrol
4. Wesson LG Jr. Renal hemodynamics in physiological states. In Physiology of the Human Kidney
. New York: Grune and Stratton; 1969:96–108.
5. Macías-Núñez JF, López-Novoa JM. Physiology of the healthy aging kidney. In: Macías-Núñez JF, Cameron JS, Oreopoulos DM, eds. The Aging Kidney in Health and Disease
. New York: Springer; 2008:93–112.
6. Wetzels JF, Kiemeney LA, Swinkels DW, Willems HL, den Heijer M. Age- and gender-specific reference values of estimated GFR in Caucasians: the Nijmegen Biomedical Study. Kidney Int
7. Hollenberg NK, Rivera A, Meinking T, et al. Age, renal perfusion, and function in island-dwelling indigenous Kuna Amerinds of Panama. Nephron
8. Fliser D, Franek E, Ritz E. Renal function in the elderly—is the dogma of an inexorable decline of renal function correct? Nephrol Dial Transplantation
9. Fliser D. Ren sanus in corpore sano: the myth of the inexorable decline of renal function with senescence. Nephrol Dial Transplantation
10. Rowe JW, Andres R, Tobin JD, Norris AH, Shock NW. The effect of age on creatinine clearance in men: a cross-sectional and longitudinal study. J Gerontol
11. Lindeman RD, Tobin J, Shock NW. Longitudinal studies on the rate of decline in renal function with age. J Am Geriatr Soc
12. Darmady EM, Offer J, Woodhouse MA. The parameters of the aging kidney. J Pathol
13. Hoang K, Tan JC, Derby G, et al. Determinants of glomerular hypofiltration in aging humans. Kidney Int
14. National Kidney Foundation Kidney Disease Outcomes Quality Initiative. Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification. Am J Kidney Dis
2002(suppl 1);39: s1–s266.
WRITE TO US!
We'd like to hear what you think of our publication. Letters will be published as space permits and may be edited to fit available space. Please e-mail letters to NephrologyTimes@wolterskluwer.com and include your full name, title, address, daytime phone number, and e-mail address.