Search tips
Search criteria 


Logo of ndtLink to Publisher's site
Nephrol Dial Transplant. 2009 April; 24(4): 1197–1205.
Published online 2008 November 6. doi:  10.1093/ndt/gfn594
PMCID: PMC2721425

Predictivity of survival according to different equations for estimating renal function in community-dwelling elderly subjects


Background. Detection of subjects with early chronic kidney disease (CKD) is important because some will progress up to stage 5 CKD, and most are at high risk of cardiovascular morbidity and mortality. While validity and precision of estimated glomerular filtration rate (eGFR) equations in tracking true GFR have been repeatedly investigated, their prognostic performance for mortality has not been hitherto compared. This is especially relevant in an elderly population in whom the risk of death is far more common than progression.

Methods. We analysed data of participants in the InCHIANTI study, a community-based cohort study of older adults. Twenty-four-hour creatinine clearance (Ccr), Cockcroft–Gault (C-G) and Modification of Diet in Renal Disease (MDRD)-derived equations (six and four input variables) were calculated at enrolment (1998–2000), and all-cause mortality and cardiovascular mortality were prospectively ascertained by Cox regression over a 6-year follow-up.

Results. Of the 1270 participants, 942 (mean age 75 years) had complete data for this study. The mean renal function ranged from 77 ml/min/1.73 m2 by Ccr to 64 ml/min/1.73 m2 by C-G. Comparisons among equations using K/DOQI staging highlight relevant mismatches, with a prevalence of CKD ranging from 22% (MDRD-4) to 40% (C-G). Reduced renal function was a strong independent predictor of death. In a Cox model–-adjusted for demographics, physical activity, comorbidities, proteinuria and inflammatory parameters—participants with Ccr 60–90 ml/min/1.73 m2 and Ccr <60 ml/min/1.73 m2 were, respectively, 1.70 (95% CI: 1.02–2.83) and 1.91 (95% CI: 1.11–3.29) times more likely to die over the follow-up compared to those with Ccr >90 ml/min/1.73 m2. For the C-G, the group with values <60 ml/min/1.73 m2 had a significant higher all-cause mortality compared to those with values >90 ml/min/1.73 m2 (HR 2.59, 95% CI: 1.13–5.91). The classification based on the MDRD formulae did not provide any significant prognostic information. The adjusted risk of all-cause mortality followed a similar pattern when Ccr and estimating equations were introduced as continuous variables or dichotomized as higher or lower than 60 ml/min. C-G was the best prognostic indicator of cardiovascular mortality. Possibly, Ccr and C-G are better prognostic indicators than MDRD-derived equations because they incorporate a stronger effect of age.

Conclusions. In a South-European elderly population, the prevalence of CKD is high and varies widely according to the method adopted to estimate GFR. Researchers and clinicians who want to capture the prognostic information on mortality related to kidney function should use the Ccr or C-G formula and not MDRD equations. These results highlight the importance of strategies for early detection and clinical management of CKD in elderly subjects.

Keywords: Cockcroft–Gault formula, elderly, MDRD equations, mortality, population-based study


The epidemiology of chronic kidney disease (CKD) is receiving growing attention across medical disciplines [1–6]. This global recognition parallels the growing evidence that mild to moderate renal dysfunction is associated with increased risks of all-cause mortality as well as cardiovascular morbidity and mortality [7–13]. CKD is predominantly a disease of older people [14–18], and age affects outcomes in CKD with older people more likely to die than to develop stage 5 CKD [19,20]. Thus, an accurate and feasible estimate of glomerular filtration rate (eGFR) is critical in the clinical evaluation of older subjects.

Equations widely accepted in the medical community are the Cockcroft and Gault (C-G) formula [21] and the Modification of Diet in Renal Disease (MDRD) formulae [22]. The abbreviated MDRD formula [23] was recommended for classifying patients with CKD into stages established by the National Kidney Foundation kidney Disease Outcomes Quality Initiative (NKF-K/DOQI) [1].

While predictive performance of these equations in estimating true GFR has been extensively exploited [24,25], their prognostic value for mortality has seldom been compared. If one goal of estimating renal function in clinical practice is to obtain estimates of death or cardiovascular risks due to early CKD, it seems logical to use the equation that provides the best prediction of these outcomes. This is especially relevant in an elderly population.

Thus, we examined the relationship between renal function estimated by different equations and all-cause and cardiovascular mortality over a 6-year follow-up among participants in the Invecchiare in Chianti (InCHIANTI) study.

Material and methods

The study participants consisted of men and women aged 65 and older who participated in the InCHIANTI study. The rationale, design and data collection have been described elsewhere [26]. Briefly, a representative sample of 1270 persons was randomly selected from an elderly population living in two small towns in the Chianti area, Tuscany, Italy. Of them, 942 had complete data for the analyses here presented. In comparison with subjects included in the study, those excluded were older and had greater comorbidities, as reported elsewhere [27].

The study was described to each participant and all signed written, informed consent. The participants were evaluated at the study entry (1998–2000), at 3-year and 6-year follow-up visits. The study protocol complied with the Declaration of Helsinki and was approved by the Italian National Institute of Research and Care on Aging Ethical Committee. After 6 years, we collected data on all-cause and cardiovascular mortality for the study entry cohort, using data from the Mortality General Registry maintained by the Tuscany Region and the death certificates that are deposited after the death at the Registry office of the Municipality of residence. We classified cardiovascular deaths with ICD-9-CM codes from 410 to 438.

Blood and urinary parameters

At the time of the home interview, participants were provided a plastic container and received detailed instructions for 24-h urine collection, including advice to take note of the start and end time. Subjects with incomplete times of collection were excluded. Blood samples were obtained after a 12-h fast and after the participants had been resting for at least 15 min. Aliquots of serum and 24-h urine samples were stored at −80°C and were not thawed until analysed. Serum urea nitrogen and serum albumin were measured using commercial kits. Urinary protein excretion was assessed on early morning spot sample by an automated urine test-strip analyser (Aution Max AX-4280, A. Menarini Diagnostics, Florence, Italy) with minimum detecting sensitivity of 5 mg/dl. Serum high sensitivity C-reactive protein (CRP) was measured in duplicate using an enzyme-linked immunoabsorbent assay (ELISA) and colorimetric competitive immunoassay that uses purified protein and polyclonal anti-CRP antibodies; the minimum detectable threshold was 0.03 mg/l, and the inter-assay coefficient of variation was 5%. Serum and urine creatinine were measured by using a compensated modified Jaffe method for Roche/Hitachi analyser (Roche Diagnostics GmbH, Mannheim, Germany). The method has been standardized against isotope dilution mass spectrometry (IDMS) starting with a primary calibrator, e.g. the standard reference material (SRM) 914. Making use of a standardized IDMS-traceable calibrated creatinine assay, we applied the re-expressed MDRD equations [28,29].

The following indexes of renal function were calculated:

  • Twenty-four-hour Ccr: urinary creatinine (mg/dl)* urinary volume (ml)/serum creatinine (mg/dl)*length of collection (min).
  • Cockcroft–Gault (C-G): (140−age)*weight*0.85(if female)/(serum creatinine*72).
  • eGFR from the six-variable MDRD equation (MDRD-6): 161.5*(serum creatinine^−0.999)*(age^ −0.176)*0.762(if female)*(BUN^−0.170)*(albumin^ 0.318).
  • eGFR from the four-variable MDRD equation (MDRD- 4): 175*(serum creatinine^−1.154)*(age^−0.203)* 0.742(if female).

In MDRD equations, we did not apply the multiplier factor for blacks as all individuals were whites.

To make allowance for comparisons with MDRD formulae, Ccr and C-G results were adjusted for 1.73 body surface area (BSA) calculated according to the DuBois and Dubois formula [30]: BSA = 0.007184* (weight^0.425)*(height^0.725) where weight (kg) and height (cm) were measured objectively.

Comorbidity and other variables

All participants were examined at home by an experienced clinician. Diseases were ascertained according to pre-established criteria that combine information from self-reported physician diagnoses, current pharmacological treatment, medical records, clinical examinations and previous blood tests. Diseases included in the current analysis were myocardial infarction, angina, chronic heart failure (CHF), stroke, diabetes and hypertension. Criteria for defining hypertension were previous positive history and/or a measured blood pressure >140/90 mmHg. Smoking history was determined from self-report and dichotomized in the analysis as ‘current smoking’ versus ‘ever smoked’ and ‘never smoked’. Physical activity in the year before the interview was coded as (1) sedentary, (2) moderate and (3) high. Alcohol intake (grams/day) was estimated by the European Prospective Investigation into Cancer and Nutrition food frequency questionnaire [31].

Statistical analysis

Variables are reported as mean values ± standard deviations (SD), medians and inter-quartile ranges or percentages. Comparisons between participants who died and those who remained alive were performed by age-adjusted linear regression models (ANCOVA) and Mantel–Hansel chi-square. Agreement between indexes of renal function was explored by scatter plots with lines drawn at the 90 and the 60 ml/min/1.73 m2 values as well as by Pearson's correlations. The relationship between measures of renal function and all-cause and cardiovascular mortality was analysed by the Cox proportional hazard model. Potential confounders of the relationship between renal function and mortality were those significantly associated with preliminary univariate analysis. We included age and gender among the potential confounders for Ccr but not for MDRD and C-G, as those variables were already incorporated in the estimate equations. Moreover, we did not compute as potential confounders the cardiovascular events that have occurred after the first wave of InCHIANTI, e.g. after study entry, since they may be involved in the process by which reduced renal function may increase the risk of death. Ccr and different estimate equations were introduced in the model according to the K/DOQI staging [1] or as continuous variables or dichotomized for values above or below 60 ml/min/ 1.73 m2.

Correlations were performed also to evaluate the relationship between age and different estimate equations. We tested non-linearity of this relationship using the quadratic term of the age.

We performed all analyses using SAS (version 8.2, SAS Institute, Inc., Cary, NC, USA) assuming a statistical significance level set at P < 0.05.


Characteristics of the study population at enrolment according to vital status at the end of the follow-up are reported in Table Table1.1. Only 1.6% males and 1.3% females had serum creatinine values >1.5 mg/dl. Mean levels of renal function ranged from 77 ml/min/1.73 m2 by Ccr and 64 ml/min/ 1.73 m2 by MDRD-6 and C-G, with MDRD-4 values lying in between. Percentages of comorbidities are in line with figures found in the elderly Italian population [32]. Spot protein urinary excretion was absent in the overwhelming majority of this cohort and very mild when detectable; only 17 subjects had values ≥20 mg/dl.

Table 1
Characteristics of the study population

After a 6-year follow-up, 171 of the 942 participants enrolled at baseline had died, with an annual mortality rate of ~3%. Compared to participants who were still alive at the end of the follow-up, those who died had significantly higher serum levels of creatinine, glucose and CRP and significantly lower total cholesterol serum levels. Those who died were older and more likely to be male, sedentary and to be affected by hypertension, stroke and CHF. Interestingly enough, proteinuria did not predict mortality (Table (Table11).

Renal function was categorized according to K/DOQI staging [1]. Due to their limited number (4, 7, 9 and 11 subjects for MDRD-4, MDRD-6, C-G and Ccr, respectively), the participants with eGFR <30 ml/min/1.73 m2 were included in the eGFR <60 ml/min/1.73 m2 group. Scatter plots between various indexes of renal function are shown in Figure Figure1.1. The lines drawn at 60 and 90 ml/min/ 1.73 m2 disclosed substantial mismatches. As a consequence of this, the prevalence of an eGFR <60 ml/min/1.73 m2 varied widely depending on the equation used: from 22% (MDRD-4) to 40% (C-G) (Figure (Figure2).2). Correlation coefficients (not shown in Figure Figure1)1) indicate good agreement between the two MDRD formulae (r = 0.97). Correlations between MDRD-4 or MDRD-6 and C-G were lower (r = 0.82 for both comparisons) and even worst when estimate equations were compared with Ccr (r values between 0.51 and 0.53); all these comparisons were highly significant (P < 0.0001).

Fig. 1
Scatter plots comparing (A) MDRD-6 versus MDRD-4; (B) MDRD-6 versus C-G; (C) MDRD-4 versus C-G; (D) Ccr versus C-G; (E) Ccr versus MDRD-4 and (F) Ccr versus MDRD-6. Each point represents one subject. Continuous lines are drawn at 60 and 90 ml/min/1.73 ...
Fig. 2
Prevalence of CKD, defined as renal function <60 ml/min, according to Ccr and each estimate equation.

When in the Cox model renal function was categorized according to K/DOQI staging, participants with Ccr 60–90 ml/min/1.73 m2 and with Ccr <60 ml/min/1.73 m2 were more likely to die during follow-up than those with Ccr >90 ml/min/1.73 m2 (Table (Table2,2, Figure Figure3A).3A). For the C-G, the group with Ccr <60 ml/min/1.73 m2 had significantly higher mortality compared to the reference group (Table (Table2,2, Figure Figure3B).3B). Per contra, both MDRD-4 and MDRD-6 formulae did not provide any significant prognostic information (Table (Table2).2). When in the Cox model indexes of renal function were introduced as continuous variables, the adjusted risk of death was highly significant for Ccr (HR 1.01, CI: 1.00–1.02, P < 0.02) and even more so for C-G (Table (Table3).3). C-G was also the best predictor when renal function was dichotomized as higher or lower than 60 ml/min/1.73 m2 (Table (Table3).3). Once again both MDRD equations were not associated with death neither when introduced as continuous variable nor when dichotomized as higher or lower than 60 ml/min/1.73 m2. Stroke, CHF, level of physical activity, CRP and cholesterol were independent predictors of all-cause mortality in the many Cox models tested.

Fig. 3
Cox's survival curves during the 6-year follow-up according to renal function strata for Ccr (A) and C-G (B).
Table 2
Cox model for all-cause mortality with K/DOQI staging estimated by different methods
Table 3
Full Cox model for 6-year all-cause mortality with renal function estimated by the C-G equation

During the 6-year follow-up, death from cardiovascular causes occurred in 67 subjects, e.g. 7.1% of the baseline cohort. The adjusted hazard ratio for cardiovascular deaths increased inversely with renal function when it was estimated by C-G (Table (Table4),4), whereas both MDRD formulae did not provide any significant prognostic information.

Table 4
Full Cox model for 6-year cardiovascular mortality with renal function estimated by the C-G equation

Although not significant in preliminary univariate analysis, we forced proteinuria into the many Cox models tested but results did not change.

Correlations between indexes of renal function and quadratic term of age were not significant; therefore we assumed linear associations. Linear correlations between formulae and age are presented in Figure Figure4.4. Slopes for Ccr, C-G, MDRD-6 and MDRD-4 resulted −1.26, −1.40, −0.54 and −0.55, respectively. That is to say that, for every year exceeding 65, the loss of renal function detected by Ccr and C-G was double that estimated by MDRD.

Fig. 4
Correlations between Ccr or estimate renal function equations and age. For details see the text.


In this study, we evaluated whether eGFR equations and calculated Ccr from 24-h urine collections in an elderly European population provide similar informations. We found that K/DOQI staging, and therefore the CKD prevalence, varied widely according to the assessment method adopted to estimate GFR. The main result of our study is that graded reduction in renal function, assessed by both Ccr and C-G but not by MDRD-derived equations, was a highly significant independent predictor of all-cause mortality as well as cardiovascular mortality after adjustment for major confounding baseline clinical characteristics, including proteinuria.

Inferences from epidemiologic studies are often difficult to extrapolate to the general population, due to the lack of detailed descriptions of the underlying reference population. This is particularly problematic in studying chronic diseases because chronically ill patients may be more or less likely to participate in screening programmes [33]. In contrast, the data reported here are from a population sample randomly selected from a well-defined reference population, with an extremely high participation rate. Moreover, our reference population is representative of the Italian population, with both percentages of elderly subjects being in the region of 20% [32]. The mean age for our community-based cohort was 75 years. It has been recently shown that, at variance with younger subjects, patients ≥75 years were far more likely to die than to progress to stage 5 CKD, age being an important effect modifier in CKD [19,20]. This not only might imply the need to develop separate risk scores for different clinical outcomes [20] but also underlines the critical need to identify the renal function estimate equation that provides the best prediction of these outcomes.

Accurate GFR measurements using inulin or iothalamate infusions are impractical for large-scale application, while Ccr is made complex by the need of obtaining accurate 24-h urine collection. To minimize this last limitation, we instructed subjects to take note of the start and end time of urine collection to allow for computation of the length of collection. The safeguard adopted probably minimizes the source of error deriving from inaccurate urine collection, but does not preserve from other critical factors such as creatinine metabolism and renal handling.

In the last decades, several equations to estimate GFR have been developed. The C-G equation has been validated against Ccr [21], whereas MDRD equations have been validated against iothalamate clearance [22]. Hence, C-G and MDRD are two distinct variables, the former being an estimate of Ccr, the latter of GFR. Although these equations are very attractive due to their simplicity and ease of use, their prediction is highly influenced by factors that influence creatinine metabolism [24]. Scr is produced by muscle catabolism and, therefore, is strongly influenced by muscle mass. Formulae use age, sex, race and weight to estimate muscle mass and to improve the prediction of renal function. This approach is based on the assumption that the elderly, women, whites, leaner and smaller subjects have lower muscle mass and, therefore, produce less creatinine.

Unfortunately, while most of the estimating equations show a very good fit in the population where they were estimated, their performance in other population with different characteristics is usually much lower and they may even produce biased results [34–44]. For example, studies of eGFR equations in healthy potential kidney donors reported underestimation of GFR by 29% for the MDRD equation and 27% for C-G [41], a less negative bias when using the C-G formula in comparison to the MDRD equation [42] or the opposite [37]. Moreover, also racial differences may well impact on equation performance. For instance, MDRD-4 in China and Japan yielded conflicting results with underestimation or overestimation of the measured GFR, respectively [43,44]. The InCHIANTI study population is a Caucasian cohort representative of the general Italian elderly population with more participants having ‘normal’ creatinine and low prevalence of proteinuria and of comorbidities. Per contra, the participants of the MDRD study were blacks (12%), younger, heavier and had serum creatinine 2.5 times higher than in InCHIANTI [22]. Similarly, the C-G formula was developed from a primarily male (209 of 236) Canadian inpatient population [21]. Hence, it comes as no surprise how in our population K/DOQI CKD staging differed according to the method adopted to estimate renal function and how, more importantly, the prevalence of CKD ranged from 22% by MDRD-4 to almost double, 40%, by C-G. In other elderly populations, one can find quite different figures. The US NHANES population over 70s had a CKD prevalence of 38% by MDRD-4 [16]. In Jerusalem community, people aged 70 years and older had a CKD prevalence of 34% by MDRD-4 and 51% by C-G [17]. UK subjects over 80s recruited from residential care homes had a CKD prevalence ranging from 83% to 97% according to whether MDRD-4 or C-G was adopted [18]. These widespread differences may reflect true differences in population or may be due to different selection criteria. Moreover, the use of a single serum creatinine value may have led to an overestimation of the prevalence of CKD in some of these surveys, including ours. It has actually been estimated that 29% of subjects initially categorized as stage 3 CKD would no longer fall into this category upon having a second test after 3 months or more [45]. Therefore, we acknowledge the lack of estimation of chronicity as the main limitation of our study.

To gain insight into the validity of the prediction provided by the different assessment methods, we examined how well the different formulae predict mortality. Mortality is an important reference parameter, especially considering the growing evidence of association between renal damage and increased morbidity and mortality [7–13]. Our analysis was adjusted for a wide panel of potential confounders, including traditional risk factors identified in the Framingham Heart study and non-traditional factors. As far as we know, measurements of sophisticated parameters, such as high-sensitivity CRP associated with measured renal function by 24-h creatinine clearance, are unique features of this cohort compared to other population-based studies.

In spite of the low mortality in our population, creatinine clearance measured with a 24-h urine collection or estimated with the C-G formula was a significant and independent predictor of all-cause as well as cardiovascular mortality, while MDRD-derived equations were not. This conclusion holds true either when renal function estimates were categorized according to K/DOQI CKD staging, or dichotomized as lower or higher than 60 ml/min/1.73 m2, or when challenged in the Cox model as continuous variables. We analysed estimate equations in different ways because of their established inaccuracies for values >60 ml/min/1.73 m2 [46,47]. Actually K/DOQI classification brings about an inherent potential limitation in ‘defining different segments of a continuous distribution as separate stages of a disease process’ [48].

Our results are in line with those recently achieved in different clinical settings. Perkovich et al. in subjects with cerebrovascular disease found that C-G is a better predictor of major clinical events than serum creatinine or MDRD-4 [49]. That MDRD-4 cannot be the gold standard worldwide is reinforced by data from Israel, where Maaravi et al. found that in an elderly general population the new Mayo Clinic equation was a stronger predictor of mortality than MDRD or C-G [17].

The way Ccr and estimate formulae describe the decline in GFR by age might explain the different performance. According to our data, slopes of Ccr and C-G against age are much steeper than those of MDRD. Similar results have been already described by Cirillo et al. in Italian individuals without CKD [50]. It is tempting to speculate that calculated or estimated creatinine clearance is a better prognostic indicator than eGFR by MDRD because it incorporates a stronger effect of age. However, estimating equations C-G and MDRD already incorporate age in the formula and this might beset validity of their correlation with age.

The limitations related to the measurements and interpretation of CKD tests are actively being addressed. International organizations have developed a process to standardize the serum creatinine assay to a high-quality reference standard with the goal of implementing assay standardization in all clinical laboratories [51]. This may attenuate the dispute on the performance of various eGFR equations. Until that time, researchers and clinicians who want to capture the link between CKD and all-cause as well as cardiovascular mortality should measure 24-h creatinine clearance or estimate it by the C-G formula. MDRD equations do not predict mortality. This holds true at least in a general elderly population with characteristics similar to InCHIANTI.

According to our and other reports, there appears to be little question that the prevalence of CKD is high in elderly people and is strongly associated with risk of death. These results may impact on designing and implementing strategies for early detection and clinical management of CKD in elderly subjects.


This work was supported by National Institute on Aging Contracts N01-AG-916413, N01-AG-821336, N01-AG-5-0002, NIA Grant R01 AG027012, and supported in part by the Intramural Research Program, National Institute on Aging, NIH.

Conflict of interest statement. None declared.


1. National Kidney Foundation K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;39:S1–S266. [PubMed]
2. Levey AS, Coresh J, Balk E, et al. National Kidney Foundation practice guidelines for chronic kidney disease: evaluation, classification and stratification. Ann Intern Med. 2003;139:137–147. [PubMed]
3. Levey AS, Eckardt KU, Tsukamoto Y, et al. Definition and classification of chronic kidney disease: a position statement from Kidney Disease Improving Global Outcomes (KDIGO) Kidney Int. 2005;67:2089–2100. [PubMed]
4. Coresh J, Byrd-Holt D, Astor BC, et al. Chronic kidney disease awareness, prevalence and trends among US adults, 1999 to 2000. J Am Soc Nephrol. 2005;16:180–188. [PubMed]
5. El Nahas AM, Bello AK. Chronic kidney disease: the global challenge. Lancet. 2005;365:331–340. [PubMed]
6. Stevens LA, Coresh J, Greene T, et al. Assessing kidney function: measured and estimated glomerular filtration rate. N Engl J Med. 2006;354:2473–2483. [PubMed]
7. Ruilope L, Antonio S, Jamerson K, et al. Renal function and intensive lowering of blood pressure in hypertensive participants of the hypertension optimal treatment (HOT) study. J Am Soc Nephrol. 2001;12:218–225. [PubMed]
8. Henry RMA, Kostense PJ, Bos G, et al. Mild renal insufficiency is associated with increased cardiovascular mortality: The Hoorn Study. Kidney Int. 2002;62:1402–1407. [PubMed]
9. Sarnak MJ, Levey AS, Schoolwerth AC, et al. Kidney disease as a risk factor for development of cardiovascular disease. Circulation. 2003;108:2154–2169. [PubMed]
10. Go AS, Chertow GM, Fan D, et al. Chronic kidney disease and the risk of death, cardiovascular events and hospitalization. N Engl J Med. 2004;351:1296–1305. [PubMed]
11. Tonelli M, Wiebe N, Culleton B, et al. Chronic kidney disease and mortality risk: a systematic review. J Am Soc Nephrol. 2006;17:2034–2047. [PubMed]
12. Hillege HL, Nitsch D, Pfeffer MA, et al. Renal function as a predictor of outcome in a broad spectrum of patients with heart failure. Circulation. 2006;113:671–678. [PubMed]
13. Kottegen A, Russel SD, Loehr LR, et al. Reduced kidney function as a risk factor for incident heart failure: the Atherosclerotic Risk in Communities (ARIC) study. J Am Soc Nephrol. 2007;18:1307–1315. [PubMed]
14. Jungers P, Chauveau P, Descamps-Latscha B, et al. Age and gender-related incidence of chronic renal failure in a French urban area: a prospective epidemiologic study. Nephrol Dial Transplant. 1996;11:1542–1546. [PubMed]
15. Magnason RL, Indridason OS, Sigvaldason H, et al. Prevalence and progression of CRF in Iceland: a population-based study. Am J Kidney Dis. 2002;40:955–963. [PubMed]
16. Coresh J, Selvin E, Stevens LA, et al. Prevalence of chronic kidney disease in the United States. JAMA. 2007;298:2038–2047. [PubMed]
17. Maaravi Y, Bursztyn M, Hammerman-Rozember R, et al. Glomerular filtration rate estimation and mortality in an elderly population. Q J Med. 2007;100:441–449. [PubMed]
18. Carter JL, O’Riordan SE, Eaglestone GL, et al. Chronic kidney disease prevalence in a UK residential care home population. Nephrol Dial Transplant. 2008;23:1257–1264. [PubMed]
19. O’Hare AM, Choi AI, Bertenthal D, et al. Age affects outcomes in chronic kidney disease. J Am Soc Nephrol. 2007;18:2758–2765. [PubMed]
20. Johnson ES, Thorp ML, Yang X, et al. Predicting renal replacement therapy and mortality in CKD. Am J Kidney Dis. 2007;50:559–565. [PubMed]
21. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16:31–41. [PubMed]
22. Levey AS, Bosch JP, Lewis JB, et al. Modification of Diet Renal Disease Study Group A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med. 1999;130:461–470. [PubMed]
23. Levey AS, Greene T, Kusek JW, et al. A simplified equation to predict glomerular filtration rate from serum creatinine [Abstract] J Am Soc Nephrol. 2000;11:155A.
24. Levey AS. Measurement of renal function in chronic renal disease. Kidney Int. 1990;38:167–184. [PubMed]
25. Coresh J, Astor BC, McQuillan G, et al. Calibration and random variation of the serum creatinine assay as critical elements of using equations to estimate glomerular filtration rate. Am J Kidney Dis. 2002;39:920–929. [PubMed]
26. Ferrucci L, Bandinelli S, Benvenuti E, et al. Subsystems contributing to the decline in ability to walk: bridging the gap between epidemiology and geriatric practice in the InCHIANTI study. J Am Geriatr Soc. 2000;48:1618–1625. [PubMed]
27. Marsh AP, Miller ME, Saikin AM, et al. Lower extremity strength and power are associated with 400-meter walk time in older adults: The InCHIANTI study. J Gerontol A Biol Sci Med Sci. 2006;61:1186–1193. [PMC free article] [PubMed]
28. Levey AS, Coresh J, Greene T, et al. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145:247–254. [PubMed]
29. Levey AS, Coresh J, Greene T, et al. Expressing the modification of diet in renal study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clin Chem. 2007;53:766–772. [PubMed]
30. DuBois D, DuBois EF. A formula to estimate the approximate surface area if height and weight be known. Arch Int Med. 1916;17:863–871. [PubMed]
31. Pisani P, Faggiano F, Krogh V, et al. Relative validity and reproducibility of a food frequency dietary questionnaire for use in the Italian EPIC centres. Int J Epidemiol. 1997;26(Suppl 1):S152–S160. [PubMed]
33. McMurdo ME, Witham MD, Gillespie ND. Including older people in clinical research. BMJ. 2005;331:1036–1037. [PMC free article] [PubMed]
34. Lin J, Knight EL, Hogan ML, et al. A comparison of prediction equations for estimating glomerular filtration rate in adults without kidney disease. J Am Soc Nephrol. 2003;14:2573–2580. [PubMed]
35. Coresh J, Astor BC, Greene T, et al. Prevalence of chronic kidney disease and decreased kidney function in the adult US population: third National Health and Nutrition Examination survey. Am J Kidney Dis. 2003;41:1–12. [PubMed]
36. Verhave JC, Gansevoort RT, Hillege HL, et al. The use of indirect estimates of renal function to evaluate the effect of risk factors on renal function. J Am Soc Nephrol. 2004;15:1316–1322. [PubMed]
37. Froissart M, Rossert J, Jacquot C, et al. Predictive performance of the modification of diet in renal disease and Cockroft–Gault equations for estimating renal function. J Am Soc Nephrol. 2005;16:763–773. [PubMed]
38. Poggio ED, Nef PC, Xuelei W, et al. Performance of the Cockcroft–Gault and Modification of Diet in Renal Disease equations in estimating GFR in ill hospitalized patients. Am J Kidney Dis. 2005;46:242–252. [PubMed]
39. Verhave JC, Fesler P, Ribstein J, et al. Estimation of renal function in subjects with normal serum creatinine levels: influence of age and body mass index. Am J Kidney Dis. 2005;46:233–241. [PubMed]
40. de Jong PE, Halbesma N, Gansevoort RT. Screening for early kidney disease—what method fits best? Nephrol Dial Transplant. 2006;21:2358–2361. [PubMed]
41. Rule AD, Gussak HM, Pond GR, et al. Measured and estimated GFR in healthy potential kidney donors. Am J Kidney Dis. 2004;43:112–119. [PubMed]
42. Poggio ED, Wang X, Greene T, et al. Performance of the modification of diet in renal disease and Cockroft–Gault equations in the estimation of GFR in health and in chronic kidney disease. J Am Soc Nephrol. 2005;16:459–466. [PubMed]
43. Ma YC, Zuo L, Chen JH, et al. Modified glomerular filtration rate estimating equation for Chinese patients with chronic kidney disease. J Am Soc Nephrol. 2006;17:2937–2944. [PubMed]
44. Imai E, Horio M, Nitta K, et al. Modification of the Modification of Diet in Renal Disease (MDRD) study equation for Japan. Am J Kidney Dis. 2007;50:927–937. [PubMed]
45. Quinn MP, Rainey A, Cairns KJ, et al. The practical implications of using standardized estimation equations in calculating the prevalence of chronic kidney disease. Nephrol Dial Transplant. 2008;23:542–548. [PubMed]
46. Glassock RJ, Winearls C. An epidemic of chronic kidney disease: fact or fiction? Nephrol Dial Transplant. 2008;23:1117–1121. [PubMed]
47. Chronic Kidney Disease in Adults UK Guidelines for Identification, Management and Referral. . [PubMed]
48. Couser WG. Chronic kidney disease—the promise and the perils. J Am Soc Nephrol. 2007;18:2803–2805. [PubMed]
49. Perkovic V, Algert C, Arima H, et al. Predictive ability of different measures of kidney function: data from the PROGRESS study. J Am Soc Nephrol. 2006;17:401A.
50. Cirillo M, Anastasio P, De Santo NG. Relationship of gender, age and body mass index to errors in predicted kidney function. Nephrol Dial Transplant. 2005;20:1791–1798. [PubMed]
51. Vassallotti JA, Stevens LA, Levey AS. Testing for chronic kidney disease: a position statement from the national kidney foundation. Am J kidney Dis. 2007;50:169–180. [PubMed]

Articles from Nephrology Dialysis Transplantation are provided here courtesy of Oxford University Press