Elevated relative mortality risk with mild-to-moderate chronic kidney disease decreases with age

Nephrol Dial Transplant (2007) 22: 3214 3220 doi:10.1093/ndt/gfm396 Advance Access publication 12 July 2007 Original Article Elevated relative mortality risk with mild-to-moderate chronic kidney disease decreases with age Neil T. Raymond 1, Daniel Zehnder 2, Stephen C. H. Smith 3, Julie A. Stinson 3, Hendrik Lehnert 2 and Robert M. Higgins 4 1 Health Sciences Research Institute and 2 Clinical Sciences Research Institute, Warwick Medical School, University Of Warwick, Coventry, CV4 7AL, 3 Clinical Biochemistry, University Hospital Coventry and Warwickshire NHS Trust, Clifford Bridge Road, Coventry CV2 2DX and 4 Department of Nephrology, University Hospital Coventry and Warwickshire NHS Trust, Clifford Bridge Road, Coventry CV2 2DX, UK Abstract Background. Renal disease is common in the general population and whilst few people progress to end-stage renal failure, mortality is increased. The aim of this study was to examine all-cause mortality risk in relation to chronic kidney disease (CKD) stages defined by estimated glomerular filtration rate (egfr). Methods. Data were extracted from a computerized central laboratory system for a defined geographical area over a 3-year study period. The egfr was calculated using the four-variable Modification of Diet in Renal Disease (MDRD) formula and aligned to the MDRD laboratory. Average annual mortality and relative risk (RR) of all-cause mortality was determined and compared for defined age and CKD bands. Results. 106 366 participants (55.5% female; 85% White, 13% South Asian, 2% Black and others) were eligible and studied, representing 49% of the Coventry adult population. 12 540 (12%) of the sample had some evidence of decreased kidney function, with an egfr <60 ml/min/1.73 m 2. 7611 (7%) participants died and there were significantly elevated risks of mortality with increasing renal dysfunction; RR ¼ 4.0, 8.3, 16.2 and 43.5 for egfr 45 59, 30 44, 15 29 and <15 ml/min/ 1.73 m 2, respectively. Within age bands, RRs were statistically significantly raised with CKD progression and within CKD stage, RR of death decreased as age increased. Conclusions. CKD prevalence increased with age and absolute and RR of mortality increased with progression of CKD. People aged over 75 years, with Correspondence and offprint requests to: Dr Robert M. Higgins, MD, Consultant Nephrologist, Department of Nephrology, University Hospital Coventry and Warwickshire NHS Trust, Clifford Bridge Road, Coventry, CV2 2DX, UK. Email: robert.higgins@uhcw.nhs.uk The authors wish it to be known that, in their opinion the first two authors contributed equally to this work. mild-to-moderate renal disease, representing 41% of this age group, have no increased RR of mortality. Further study of CKD and mortality, particularly progression over time and with respect to age is needed. Keywords: all-cause mortality; chronic kidney disease; diabetes; South Asian ethnicity Introduction In recent years it has been recognized that renal impairment is very common in the general population and more so in the elderly, with a high prevalence of people with reduced kidney function [1]. Whilst only a minority of people progress to end-stage renal disease, renal impairment is associated with an increased risk of death, often from cardiovascular disease [2 9]. Chronic kidney disease (CKD) is now regarded as an independent and novel cardiovascular risk factor leading to premature death [10]. CKD has been classified by the National Kidney Federation (NKF) Kidney Disease Quality Outcomes Initiative (KDOQI) system, in five stages [11]. Epidemiological studies using this classification have indicated that as many as 10% of the adult population and 20 54% of the elderly population have CKD in stages 3 5 [1,12 16]. In the UK, with an ageing population and increasing prevalence of type-2 diabetes mellitus, CKD is a substantial public health problem requiring urgent development and evaluation of targeted interventions to improve clinical outcomes. Importantly, the KDOQI CKD classification is not staged according to increasing age, meaning that a glomerular filtration rate (GFR) of 50 ml/min/1.73 m 2 is regarded as indicative of moderate CKD (stage 3) whatever the age of the subject, even though this GFR falls well within the normal range in the elderly [11]. ß The Author [2007]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Mortality risk in chronic kidney disease 3215 It has been recognized for many years that GFR declines with age at a rate of 1.0 ml/min/1.73 m 2 /year [1,17,18]. The significance of this decline is uncertain and it is possible that CKD stage 3 (GFR 30 59 ml/ min/1.73 m 2 ) could be normal in individuals at the extremes of age, in vegetarians or after unilateral nephrectomy [11]. The aim of this study was to explore the clinical significance of CKD stage in the population, by measuring its effect on mortality. Subjects and methods Setting The study took place in the city of Coventry with an adult population of 218 990 (female 112 041, 51.2%) aged 20 years or older (Census 2001)[19]. Coventry has an ethnically diverse population with 87% White, 10% South Asian, 2% Black and 2% other ethnic groups, 2.5 times more South Asian than the UK average. Participants After local Ethical Committee approval, data were extracted from the computerized central laboratory system for Coventry. Participants were included if they were aged 20 years or older between 1 June 2000 and 31 May 2003, if they were Coventry residents and had one or more serum creatinine measurements. Demographic data were recorded (age, sex and ethnicity). Ethnicity was self-reported, combined with Nam Pehchan software to assign South Asian ethnicity on the basis of names [20]; this software has been reported to have a sensitivity of over 90% [21]. Death year was recorded, using the hospital patient management system and the National Strategic Tracing Service (NSTS) [22]. For reasons of confidentiality, patients identified solely through the Genito-urinary medicine department were excluded. Patients were excluded if they had neither a Coventry post code nor registration with a Coventry General Practitioner. Duplicated records were amalgamated as appropriate. Participants seen by a nephrologist were identified (nephrologists have supported acute general medicine and have admitted approximately 2000 patients/year). Prior to analysis, data were fully anonymized in the University Hospital Coventry and Warwickshire (UHCW) information department. Creatinine measurements and estimated GFR (egfr) Serum creatinine was measured with the Roche Modular P-Unit by the O Leary modification of picrate method of Jaffe [23] (Synermed Europe Ltd, Sussex, UK). The between-day CVs were <2.0% at concentrations of 100 and 485 mmol/l. The reference range of serum creatinine is 70 120 mmol/l for men and 50 105 mmol/l for women. The egfr was estimated using the four-variable abbreviated Modification of Diet in Renal Disease (4vMDRD) formula [24]. Specifically, the modified MDRD study equation with serum creatinine concentration traceable to the isotope dilution mass spectrometry (ID-MS) reference method, as suggested by Levey et al. [25]: egfr ¼ 175 (serum creatinine in mg/dl) 1.154 age 0.203 (0.742 if female) (1.21 if ethnic group ¼ black). Our serum creatinine values were corrected according to the Roche enzymatic method. This enzymatic assay method is traceable to the ID-MS reference method for serum creatinine measurement. Further, calibration to allow for drift in the assay was applied resulting in a new formula: egfr ¼ 175 {[1.1203(serum creatinine serum creatinine 0.028) 34.66]/88.4} 1.154 age 0.203 0.742 (female) 1.212 (if ethnic group ¼ black). These serum creatinine adjustments align the results from this study with the creatinine assay used in the development of the modified 4vMDRD equation and allow direct comparison of our results with previous work [25]. We used a modified NKF classification of CKD [11] to classify egfr in the following ranges: at least 60 ml/min/1.73 m 2 (stages 1 and 2), 45 59 ml/ min/1.73 m 2 (stage 3A), 30 44 ml/min/1.73 m 2 (stage 3B), 15 29 ml/min/1.73 m 2 (stage 4) and <15 ml/min/1.73 m 2 (stage 5). Since proteinuria was not routinely measured, CKD stages 1 and 2 could not be separated. The reference group for risk estimation used in this study comprised all participants in CKD stages 1 and 2 and included all with egfr 60 ml/min/1.73 m 2, including those with no CKD. Definition of diabetes mellitus Diabetes was defined as random serum glucose 11.1 mmol/l or glycosylated haemoglobin >6.8%. Statistics Analyses were performed using the SAS statistical analysis package. Estimates of period prevalence (3 years) of CKD stages for the age banded population were made based on egfr calculated using the modified 4vMDRD, with the mean of all serum creatinine measurements available (1 3 measurements). Since the egfr formula used includes a term for gender, 1998 people identified in the study with no recorded gender were excluded from all CKD stages prevalence and relative risk (RR) estimates, leaving a total of 106 366. For analysis of RR of death, a conservative approach was adopted, using the minimum creatinine measured for each subject during the 3-year study period to calculate egfr and CKD stage. To eliminate confounding effects of age, an age band stratified analysis was performed. Within age bands, the RR of death for people in CKD stages 3(A and B), 4 and 5 was estimated using participants in CKD stages 1 and 2 (egfr 60 ml/min/1.73 m 2 ) as the reference group. Annual percent mortality was estimated by dividing deaths by population in age and CKD bands and further dividing by 3 years of study. Results Over the 3 years studied, a total of 108 364 participants with a mean age of 57.7 19.1 years were identified, representing 49% of the Coventry population aged 20 years or older. Gender was missing for 1998 (2%) who were excluded from all analyses. With increasing age the proportion of the population identified as

3216 N. T. Raymond et al. Age bands (years) 85+ 75 to 84 65 to 74 55 to 64 45 to 54 35 to 44 20 to 34 Females Males 40000 30000 20000 10000 0 10000 20000 30000 40000 Number of participants Study sample Coventry population Fig. 1. Coventry population and sample distribution in age bands. Table 2. Prevalence of CKD stages, based on mean creatinine measurement, showing proportions of South Asian ethnicity and participants with diabetes CKD stage Total (106 366) n (%) South Asian ethnicity group (13 471) People with diabetes (10 286) 1 and 2 (egfr 60) 93 826 (88) 12 658 (94) 7755 (75) 3A (egfr 45 59) 6831 (6) 399 (3) 1184 (12) 3B (egfr 30 44) 3365 (3) 200 (2) 710 (7) 4 (egfr 15 29) 1666 (2) 98 (1) 434 (4) 5 (egfr <15) 678 (0.7) 116 (0.9) 203 (2) Table 1. Sample characteristics of 106 366 adults from the city of Coventry with measured serum creatinine by gender [n (%)] Characteristic Females (n ¼ 58 993) Males (n ¼ 47 373) Total Age band 20 34 10 933 (19) 7248 (15) 18 181 (17) 35 44 8594 (15) 6796 (14) 15 390 (14) 45 54 8383 (14) 7413 (16) 15 796 (15) 55 64 9224 (16) 8810 (19) 18 034 (17) 65 74 9097 (15) 8542 (18) 17 639 (17) 75 84 8779 (15) 6580 (14) 15 359 (14) 85þ 3983 (7) 1984 (4) 5967 (6) Ethnicity White 50 492 (86) 40 426 (85) 90 918 (85) South Asian 7423 (13) 6048 (13) 13 471 (13) Black and Other 1078 (2) 899 (2) 1977 (2) Seen Nephrologist Yes 3388 (6) 3427 (7) 6815 (6) No 55 605 (94) 43 946 (93) 99 551 (94) Diabetes Yes 5101 (9) 5448 (12) 10 549 (10) No 53 892 (91) 41 925 (89) 95 817 (90) Died Yes 3895 (7) 3716 (8) 7611 (7) No 55 098 (93) 43 657 (92) 98 755 (93) participants increased; 31% at ages 20 44 rising to 100% of the age band 85þ years (Figure 1). Almost the entire population aged 65 years and older were included. Table 1 shows basic characteristics of the sample. Ethnic group distribution of the sample was the same as for Coventry. The overall 3-year period prevalence in the study sample is shown in Table 2. The total number in the sample with any CKD (egfr <60 ml/min/1.73 m 2 ) was 12 540 (12%). Assuming there were no further cases this would be equivalent to a Coventry adult population period prevalence of 5.7%. 9927 (79%) participants in CKD stages 3 5 had not seen any nephrologist, including 13% in stages 4 and 5 (advanced CKD). However, more than 97% of these had a serum creatinine <130 mmol/l in accordance with previous referral guidelines. Figure 2 shows the relationship between age and CKD; deteriorating renal function is shown by the shift of egfr into stages 3 and 4 with increasing age. % CKD stages 50% 40% 30% 20% 10% Changing egfr bands with increasing age Stage 1/2 Stage 3a Stage 3b Stage 4 Stage 5 0% 20 to 34 35 to 44 45 to 54 55 to 64 65 to 74 75 to 84 85+ Age bands (years) Fig. 2. Distribution of KDOQI CKD stages based on mean creatinine measurement, adjusted for laboratory assay and drift over time (egfr bands <15, 15 29, 30 44, 45 59 and 60 ml/min/ 1.73 m 2 ), by age band (n ¼ 106 366). Note: Y axis includes 0 50% only; the other 50% of the study sample are included in CKD stages 1 & 2 and are not shown in the figure. In adults younger than 45 years, <1% had any CKD (egfr < 60 ml/min/1.73 m 2 ), increasing to 30 and 50% by ages 75 and 85 years, respectively. This trend with age was similar for men and women separately. In the South Asian ethnicity subgroup, there were higher proportions in CKD stage 5 at all ages, but an overall significantly lower prevalence of any other CKD stages (Table 2). Participants with diabetes experienced considerably higher prevalence of all CKD stages 3 5 (Table 2) but increasing prevalence with age was still evident (data not shown). A total of 7611 (7%) participants died during the 3 years, representing around 90% of total Coventry adult deaths. Within the cohort, average annual mortality increased with progression of renal dysfunction: 1.9, 6.4, 11.1, 16.5 and 20.6% in CKD stages 1 and 2, 3A, 3B, 4 and 5, respectively. In Figure 3, annual average mortality by age band and CKD stagedefined subgroups are shown. The age and CKD stage progressions are illustrated clearly, with the youngest participants having lowest mortality which increased as egfr declined: from 0.2% with egfr 60 up to 5% with egfr <15 ml/min/1.73 m 2. Annual mortality percentage increased with age within CKD-defined bands.

Mortality risk in chronic kidney disease 3217 There were significantly elevated RRs of mortality with deteriorating renal function as compared with CKD stages 1 and 2 combined. RRs within age bands increased as CKD stage progressed, and within CKD stage decreased as age increased (Table 3). RR increased substantially with deteriorating kidney function as indicated by progression from stages 1 and 2 through to 5, with statistically significant differences in mortality compared with the reference group. At ages 75þ in CKD stage 3A, RRs were not significantly elevated compared with the reference group; RR (95% CI) for age band 75 84 was 1.2 (1.0 1.3) and 0.9 for ages 85þ (0.8 1.0). These groups with no significant increase in RR comprised 41% of the study population with egfr <60 ml/min/1.73 m 2 and 14% of all participants aged 75þ. Similar trends in RR in CKD stages and by age were observed for sub-groups defined by diabetes mellitus, ethnic group and gender (data not shown). Figure 4 shows mean egfr as determined for the study sample, plus mean egfr 1 and 2 SDs. These are based on the first creatinine measurement for the whole study sample and illustrate clearly decreasing egfr with advancing age. The overlaid reference lines show upper egfr bounds for CKD stages 3 5. The intersections of the reference lines and egfr slopes indicate that older population groups classified as having mild/moderate renal dysfunction using the 4vMDRD and KDOQI CKD stages; have mean egfr well within 1 and 2 SDs from the mean. Discussion Participants were included in this study on the basis of serum creatinine measurement alone using routinely collected laboratory data to capture the sample with no need to obtain consent to enter a cohort study. The clear advantages of this method were the lack of refusals and the inclusion of all data in the study; on the other hand only limited clinical data were available. Almost 50% of the entire Coventry adult population and higher proportions of the elderly were included in the study. This is substantially higher 35 30 25 20 % mortality 15 10 5 0 Stage 5 CKD stage Stage 4 Stage 3b Stage 3a Stages 1&2 20-44 45-54 55-64 65-74 75-84 85+ 55-64 Age bands 45-54 20-44 85+ 75-84 65-74 Fig. 3. Average annual mortality by age and estimated glomerular filtration rate (egfr) bands. egfr (ml/min/1.73 m 2 ) 140 120 100 80 60 40 20 0 20-34 35-44 45-54 55-64 65-74 75-84 85+ Age bands Mean egfr (Coventry popn) Mean egfr - 1sd Mean egfr - 2sd Fig. 4. Declining egfr with age (based on mean creatinine measurement) and CKD stage egfr reference lines. Grid lines show cut-offs for CKD stages 5 (<15 ml/min/1.73 m 2 ), 4 (15 29 ml/min/1.73 m 2 ) and 3 (30 59 ml/min/1.73 m 2 ). Table 3. RR (95% CI) of death during 3 years from any cause in adults among 106 366 ambulatory adults according to their egfr calculated with lowest recorded creatinine using the MDRD formula, adjusted for laboratory assay and drift over time CKD Stage (egfr) Age bands (years) ml/min/1.73 m 2 20 44 45 54 55 64 65 74 75 84 85þ CKD stages 1 and 2 (egfr 60) 1.0 1.0 1.0 1.0 1.0 1.0 Reference category CKD stage 3A (egfr 45 59) 13.6 (6.2 29.8) 7.5 (4.4 12.6) 3.0 (2.2 4.1) 1.8 (1.5 2.1) 1.2 (1.0 1.3) 0.9 (0.8 1.0) CKD stage 3B (egfr 30 44) 12.1 (4.0 36.5) 13.6 (7.5 24.7) 5.9 (3.9 8.9) 3.2 (2.6 3.9) 1.9 (1.7 2.1) 1.3 (1.2 1.5) CKD stage 4 (egfr 15 29) 17.4 (5.9 51.4) 4.6 (1.2 17.4) 9.3 (6.1 14.2) 5.2 (4.1 6.5) 3.3 (2.9 3.8) 1.8 (1.7 2.0) CKD stage 5 (egfr <15) 26.1 (9.1 74.8) 28.6 (17.4 47.2) 18.2 (13.9 23.9) 7.6 (5.7 10.1) 4.4 (3.7 5.3) 2.5 (2.3 2.8) Participants n 33 571 15 796 18 034 17 639 15,359 5967 egfr <60 n (%) 170 (0.5) 201 (1.3) 549 (3.0) 1540 (8.7) 3061 (19.9) 1882 (31.5) Died n (%) 207 (0.6) 278 (1.8) 681 (3.8) 1438 (8.2) 2695 (17.6) 2312 (38.8)

3218 N. T. Raymond et al. than recent primary care-based data reported in the NeoErica Study [26] but similar to a study of US Veterans [27] and probably reflects our inclusion of cases from both primary and secondary care sources. The sample studied reflected variation in GP referral patterns, with fewer young and more older people included. The estimates of CKD prevalence may be inflated at younger ages as healthy individuals are unlikely to be tested, but the proportion of those aged 60 and older tested was over 80%, sufficient for us to feel confident in the estimates and risks evaluated. In contrast to other studies with insured populations [2,5,6,27], our study included a whole community population, in effect a cohort defined by their exposure to serum creatinine measurement. We excluded 1998 people with no gender recorded, but these did not differ markedly from the included group and an overall missing data rate of 2% for gender from combined primary and secondary care referrals seems acceptable and unlikely to have much impact on CKD stage or mortality risk estimation. We found an overall period prevalence of CKD among the adult Coventry population of 6%, on the basis of decreased kidney function (egfr <60 ml/min/ 1.73 m 2, CKD stages 3 5) calculated using mean creatinine measurement during the study period. This estimate is conservative, since it assumes that no cases of CKD were missed. Whilst we had access to all data recorded, there are likely to be some missed cases. However, our population coverage was extremely high in the older age bands; those at highest risk of CKD. After alignment to the MDRD laboratory, the NHANES III survey found a CKD prevalence of 4.7% in the adult American population [1]. Without alignment to the MDRD laboratory, or using the Cockroft Gault GFR estimation method, prevalence rates between 4.9% and 13% have been reported [14 16,26]. In comparing these data it is important to remember that our estimates are based on 3-year period prevalence rather than a point prevalence. There are no reported prevalence data for CKD within the South Asian community. Our overall CKD prevalence of 3.9% in those of South Asian ethnicity was lower than the 6.2% for the White population. This may in part be explained by a 10 years younger mean age for the South Asian group compared with the studied White population. The South Asian ethnicity group did have a slightly higher CKD stage 5 proportion, probably reflecting previously reported over-representation of dialysis treatment within this community [28]. It is important to be aware that the MDRD formula has not been validated for those of South Asian ethnicity. It has been recognized that the physiological ageing process of the kidney leads to a deterioration of kidney function of about 1 ml/min/1.73 m 2 per year [1,17,18]. Our study appears to confirm this, although our egfr estimates for age bands are based on a cross-section of groups of patients rather than coming from a longitudinal study following a cohort over time (Figure 4). Since our sample includes only those referred for creatinine measurement rather than a random population-based sample, the actual mean egfr shown may be lower than for the total adult population. However, at ages 60 years and older our sample includes 80% of the total population, making the estimated mean egfr more accurate in the older age groups in whom we are particularly interested. Not surprisingly, more severe CKD, indicated by reduced egfr, was associated with significantly increased RR of death, with higher RRs at advanced CKD stages in all age bands. Within CKD stage bands, the RR decreased with increasing age band, with younger adults in CKD stages 3 5 having higher RRs than older adults. Indeed adults aged 75þ years, in CKD stage 3A had elevated RRs which were not statistically significant. The RRs of death observed were clearly dependent on renal function and confirm previous large population and cohort studies [2,3,5,6,8]. A new and potentially important finding is that within the CKD stages a clear age effect was observed. In all CKD stages, the magnitude of elevated RRs declined as age increased and there was some suggestion that at older ages, moderate-to-mild declines in renal function may confer little or no excess mortality risk [27]. A similar observation has been made for end-stage renal disease patients [29]. In interpreting these data it is important to consider possible selection and survivorship biases. Those at older ages appear to be tested almost routinely. Since we had no clinical data we were unable to investigate possible influences of hypertension or other comorbidities which are likely to have more effect at older ages with longer duration of exposure. Similarly we were unable to investigate possible differential dialysis and transplantation rate effects within age bands. It would be expected that younger patients undergoing transplantation would be fitter and our estimates of RR in the egfr<15 band might therefore be elevated at younger ages. The higher absolute mortality of older patients is in no doubt and gradients in average annual mortality reflect this, but our data do indicate that risk associated with mild renal dysfunction varies by age. Previous investigations of impaired renal function at CKD stage 3 and all-cause mortality have found conflicting results [30]. Some studies that reported increased mortality analysed together both CKD stage 3 and more severe renal disease; all GFR <59 ml/min [3], all patients coded as chronic kidney disease [31], the top decile of serum creatinine levels [32] or all serum creatinine >133 mmol/l [33]. Analysis of all those aged 65 years plus as a single group tends to bias results in favour of an effect of CKD, as renal function deteriorates with age [31]. In a Framingham study, where those with severe renal failure were excluded, no increased mortality with moderate renal impairment was shown [34]. Conversely, in the Cardiovascular Health study [3] and in a pooled analysis of several trials [9], excluding patients with egfr<15 ml/ min/1.73 m 2 and taking account of age, increased mortality with egfr levels of 15 60 ml/min/1.73 m 2

Mortality risk in chronic kidney disease 3219 was reported. A more recent cohort study using egfr and age bands observed an age-related attenuation of the association of egfr with RR of death [27]. Equations based on serum creatinine level, age, sex and other variables, such as the MDRD formula perform much better at predicting GFR than serum creatinine level alone [24]. However, any equation used is critically dependent on the calibration and reproducibility of the serum creatinine assay [35,36]. Interand intra-laboratory agreement deteriorate as serum creatinine concentration nears the reference interval; imprecision at lower creatinine concentration contributes to greater error in GFR estimation than imprecision at higher creatinine concentration. In the current study we attempted to minimize errors introduced by imprecision and bias of creatinine measurements. Study samples were measured in one laboratory over the entire 3 years, using the same assay method and instrument. The Roche Enzymatic assay, shown to be very close to the gold standard ID-MS method of measuring serum creatinine was used to estimate the correction needed and to provide the adjustment to the MDRD laboratory [25]. The egfr from the resulting formula allows direct comparison with previously published results and may be used as a starting point in establishing an international standard for the calibration of serum creatinine as previously suggested [25,36]. In the UK, new recommendations promote the disclosure of egfr to General Practitioners rather than simply serum creatinine; the effects of this change in practice are still to be observed and monitoring referrals to nephrology departments should be a high priority. Our data demonstrate that a substantial proportion of the population may be in CKD stages 3A and 3B, particularly at older ages. We used the 4vMDRD equation with recalibration and allowing for assay drift. If some similar correction is not used in other UK laboratories, the proportions of people in CKD stage 3 with no evidence for increased RR of mortality are likely to be significantly higher. This has been acknowledged and the UK National Quality Assessment Scheme (UKNEQAS) have been working to produce method-specific correction factors to minimize the problem of bias. Early identification of individuals and the prevention of progressive CKD are key factors in reducing CKD-associated mortality and alleviating the future burden of ESRD. More research is required to investigate the relationship between CKD, its progression over time and mortality, especially in the elderly. Acknowledgements. The authors would like to thank J Coresh for useful discussion and comment on the recalibration and alignment of creatinine measures to the MDRD laboratory standard. Thanks are also due to Karl O Sullivan and Sharon Williams who were instrumental in extracting data from the laboratory computer systems. Conflict of interest statement. None declared. References 1. Coresh J, Astor BC, Greene T, Eknoyan G, Levey AS. 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 2. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular eventsand hospitalization. N Engl J Med 2004; 351: 1296 1305 3. Manjunath G, Tighiouart H, Coresh J et al. Level of kidney function as a risk factor for cardiovascular outcomes in the elderly. Kidney Int 2003; 63: 1121 1129 4. Shlipak MG, Sarnak MJ, Katz R et al. Cystatin C and the risk of death and cardiovascular events among elderly persons. N Engl J Med 2005; 352: 2049 2060 5. Foley RN, Murray AM, Li S et al. Chronic kidney disease and the risk for cardiovascular disease, renal replacementand death in the United States Medicare population, 1998 to 1999. JAm Soc Nephrol 2005; 16: 489 495 6. Muntner P, He J, Hamm L, Loria C, Whelton PK. Renal insufficiency and subsequent death resulting from cardiovascular disease in the United States. J Am Soc Nephrol 2002; 13: 745 753 7. Henry RM, Kostense PJ, Bos G et al. Mild renal insufficiency is associated with increased cardiovascular mortality: The Hoorn Study. Kidney Int 2002; 62: 1402 1407 8. Manjunath G, Tighiouart H, Ibrahim H et al. Level of kidney function as a risk factor for atherosclerotic cardiovascular outcomes in the community. J Am Coll Cardiol 2003; 41: 47 55 9. Weiner DE, Tighiouart H, Amin MG et al. Chronic kidney disease as a risk factor for cardiovascular disease and all-cause mortality: a pooled analysis of community-based studies. JAm Soc Nephrol 2004; 15: 1307 1315 10. Sarnak MJ, Levey AS, Schoolwerth AC et al. Kidney disease as a risk factor for development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiologyand Epidemiology and Prevention. Circulation 2003; 108: 2154 2169 11. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification and stratification. Am J Kidney Dis 2002; 39[2 Suppl 1]: S46 S169 12. Wasen E, Isoaho R, Mattila K, Vahlberg T, Kivela SL, Irjala K. Estimation of glomerular filtration rate in the elderly: a comparison of creatinine-based formulae with serum cystatin C. J Intern Med 2004; 256: 70 78 13. Garg AX, Papaioannou A, Ferko N, Campbell G, Clarke JA, Ray JG. Estimating the prevalence of renal insufficiency in seniors requiring long-term care. Kidney Int 2004; 65: 649 653 14. Chadban SJ, Briganti EM, Kerr PG et al. Prevalence of kidney damage in Australian adults: The AusDiab kidney study. JAm Soc Nephrol 2003; 14[7 Suppl 2]: S131 S138 15. Viktorsdottir O, Palsson R, Andresdottir MB, Aspelund T, Gudnason V, Indridason OS. Prevalence of chronic kidney disease based on estimated glomerular filtration rate and proteinuria in Icelandic adults. Nephrol Dial Transplant 2005; 20: 1799 1807 16. Clase CM, Garg AX, Kiberd BA. Prevalence of low glomerular filtration rate in nondiabetic Americans: Third National Health and Nutrition Examination Survey (NHANES III). J Am Soc Nephrol 2002; 13: 1338 1349 17. Davies DF, Shock NW. Age changes in glomerular filtration rate, effective renal plasma flowand tubular excretory capacity in adult males. J Clin Invest 1950; 29: 496 507 18. 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 19. Coventry City Council: 2001 Census Statistics. http://www. coventry.gov.uk/ccm/content/city-development-directorate/ coventrystatistics/census-key-statistics.en (Accessed 28 July 2005) 2001

3220 N. T. Raymond et al. 20. South Asian Names Analysis. 1996 Version 1.1, Windows Version 3.15, Bradford Health Authority & City of Bradford Metropolitan District Council 21. Cummins C, Winter H, Cheng KK, Maric R, Silcocks P, Varghese C. An assessment of the Nam Pehchan computer program for the identification of names of South Asian ethnic origin. J Public Health Med 1999; 21: 401 406 22. NHS Strategic Tracing Service http://www.connectingforhealth. nhs.uk/nsts (Accessed 2 August 2006) 23. O Leary N, Pembroke A, Duggan PF. A simplified procedure for eliminating the negative interference of bilirubin in the Jaffe reaction for creatinine. Clin Chem 1992; 38: 1749 1751 24. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130: 461 470 25. Levey AS, Coresh J, Green T et al. Expressing the MDRD study equation for estimated GFR with IDMS traceable (Gold standard) serum creatinine values. J Am Soc Nephrol 2005; 16: A69 26. de Lusignan S, Chan T, Stevens P et al. Identifying patients with chronic kidney disease from general practice computer records. Fam Pract 2005; 22: 234 241 27. O Hare AM, Bertenthal D, Covinsky KE et al. Mortality risk stratification in chronic kidney disease: one size for all ages? J Am Soc Nephrol 2006; 17: 846 853 28. Clark TJ, Richards NT, Adu D, Michael J. Increased prevalence of dialysis-dependent renal failure in ethnic minorities in the west Midlands. Nephrol Dial Transplant 1993; 8: 146 148 29. Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 1998; 32[5 Suppl 3]: S112 S119 30. K/DOQI clinical practice guidelines on hypertension and antihypertensive agents in chronic kidney disease. Am J Kidney Dis 2004; 43[Suppl 1]: S42 S64 31. Collins AJ, Li S, Gilbertson DT, Liu J, Chen SC, Herzog CA. Chronic kidney disease and cardiovascular disease in the Medicare population. Kidney Int Suppl 2003; S24 S31 32. Wannamethee SG, Shaper AG, Perry IJ. Serum creatinine concentration and risk of cardiovascular disease: a possible marker for increased risk of stroke. Stroke 1997; 28: 557 563 33. Fried LP, Kronmal RA, Newman AB et al. Risk factors for 5-year mortality in older adults: the Cardiovascular Health Study. JAMA 1998; 279: 585 592 34. Culleton BF, Larson MG, Wilson PW, Evans JC, Parfrey PS, Levy D. Cardiovascular disease and mortality in a communitybased cohort with mild renal insufficiency. Kidney Int 1999; 56: 2214 2219 35. 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 36. Lamb EJ, Tomson CR, Roderick PJ. Estimating kidney function in adults using formulae. Ann Clin Biochem 2005; 42[Pt 5]: 321 345 Received for publication: 2.1.07 Accepted in revised form: 29.5.07