Although fewer than 0.1% of Canadians have end-stage renal disease (ESRD), the economic burden of ESRD is 16 times higher per patient than the average spending for all other health-care conditions.1 Early detection and prevention of kidney disease is the only way to prevent associated morbidity and premature mortality from cardiovascular disease (CVD) or ESRD2; however, kidney disease is often asymptomatic until advanced stages, making early detection unlikely in the absence of active surveillance, and this limits the opportunity for treating and delaying disease progression.3
Decline in kidney function may be a stronger prognostic indicator than static measures of reduced estimated GFR (eGFR). In particular, patients with rapid kidney function decline (RKFD) are at increased risk for CVD and mortality4–7; even among those with mildly reduced kidney function at baseline.5,7 Dynamic eGFR assessment is necessary to identify this silently progressing, high-risk group; however, serial assessment in the general population is not cost-effective.3,8,9 Similarly, screening for proteinuria to prevent ESRD is not cost-effective unless directed at high-risk populations.10
To date, no studies have investigated the clinical utility of combining risk-factor assessment with routine screening tests to identify those at highest risk of RKFD who would most benefit from serial eGFR assessment and early intervention. In this 7-year prospective cohort study, we identify risk factors for RKFD and evaluate the ability of routine screening tests for urine protein to improve the efficiency of detecting RKFD across a broad range of eGFR values.
RESULTS
Of 2,574 adult participants with at least three annual eGFR measures, median follow-up time was 7 years with seven eGFR measurements. Participants were predominantly Caucasian (99%) and ranged in age from 18 to 92 years with 75% younger than 60 years and 25% younger than 37 years. Mean baseline eGFR was 83.6 ml/min per 1.73 m2 (SD 18.4) with a median absolute annual change in eGFR of −0.78 ml/min per 1.73 m2 per year (interquartile range [IQR], −2.2, 0.5) and a median percentage annual change of −0.97% (IQR, −2.60%, 0.62%). The distribution of the absolute annual change was strongly skewed to the right and susceptible to highly variable rates of annual decline in those with baseline eGFR ≥90 ml/min per 1.73 m2. The distribution of the percentage annual change was more symmetrically balanced around the median with only three outliers (all three had baseline eGFR <60, and eGFR increased over time). When we defined RKFD as an absolute annual decline >3 ml/min per 1.73 m2 per year, 15.6% experienced RKFD compared with 8.5% for RKFD defined as an annual percentage decline >5%. As shown in Figure 1, an absolute annual decline >3ml/min per 1.73 m2 per year most commonly identified those with a baseline eGFR ≥90 ml/min per 1.73 m2 (23%). Conversely, when we defined change in eGFR relative to baseline (decline >5%), RKFD was more common in those with an eGFR <60 ml/min per 1.73 m2 (19%) and was equally distributed (∼8%) across GFR categories ≥60. In addition, annual decline >5% was more strongly associated with the risk for future self-reported CVD (relative risk [RR] = 2.7, 95% confidence interval [CI] = 1.4 to 5.4, P = 0.004) comparyed with annual decline >3 ml/min per 1.73 m2 (RR = 1.4, 95% CI = 0.7 to 2.6, P = 0.35). The remainder of the results will focus on RKFD defined as annual decline of >5%; however, a summary of the results for absolute annual decline are provided in Appendix II.
;)
Figure 1: Indicators of rapid kidney function decline in relation to estimated GFR (eGFR) at baseline: Adults with a baseline eGFR above 90 ml/min per 1.73 m2 were three times more likely to demonstrate an absolute decline >3 ml/min per 1.73 m2 than a relative decline >5%.
After excluding those with RKFD (8.5%), the median change in eGFR was −0.57 ml/min per 1.73 m2 per year. Nearly two-thirds (64.7%) of those with RKFD advanced to a new stage on the basis of GFR classification compared with 19.0% of those without RKFD. Table 1 summarizes baseline characteristics, overall and by future RKFD. RKFD was more common among participants older than 60 years (16% versus 6%); however, there was no difference in the incidence of RKFD between men and women (9.1% versus 8.2%). Although RKFD was twice as common in those with an eGFR <60 versus ≥60 ml/min per 1.73 m2 far more patients with eGFR ≥60 versus <60 presented with RKFD during follow-up (n = 183 versus 38). After adjusting for age, we found that hypertension, diabetes, CVD, obesity, and family history of diabetes were all significantly associated with increased risk for RKFD (Table 2). With the exception of obesity, which lost statistical significance in the multivariable model (P > 0.2), each was retained in the final, fully-adjusted clinical risk model. The greatest risk was seen among those with CVD (RR = 2.22, 95% CI = 1.57 to 3.12), followed by age ≥60, hypertension, and diabetes. Taken together, patients presenting with hypertension, diabetes, a history of CVD, or age ≥60 were nearly three times more likely to develop RKFD (RR = 2.8, 95% CI = 2.1 to 3.8). Among patients with clinical risk factors, 22% of those with a baseline eGFR <60 ml/min per 1.73 m2 developed RKFD compared with 25% with eGFR ≥60 ml/min per 1.73 m2.
Table 1: Participant characteristics: Overall and by percent annual decline in eGFR (ml/min per 1.73 m2 per year), 2002 to 2008
Table 2: Relative risk for rapid kidney function decline (>5% annual decline) among community-dwelling adults
Compared with those with normoalbuminuria, the presence of albuminuria conferred a 53% increased risk for RKFD (95% CI = 1.09 to 2.17) (Figure 2). Participants with trace levels of protein or greater had more than a two-fold increase in the risk of RKFD (95% CI = 1.51 to 3.19), which increased to 3.26 (95% CI = 1.95 to 5.45) for ≥3g/L. The left axis (frequency) in Figure 3 shows the number who screened positive for protein at baseline and the number with RKFD by screening criteria. The right axis shows the percentage of those who screened positive for protein that developed RKFD. Although more participants were identified with albuminuria (n = 253), far fewer developed RKFD (6%) as compared with those identified at various thresholds of dipstick protein, of which 12% with trace or above, 33% with ≥1 g/L, and 40% with ≥3 g/L of protein-developed RKFD.
Figure 2: Dipstick protein was a stronger predictor of rapid kidney function decline (RKFD) than albuminuria. Age-adjusted and fully-adjusted relative risks for RKFD (% annual decline >5% from baseline) by screening test at baseline (albuminuria or dipstick for protein [≥trace; ≥1 g/L; ≥3 g/L]).
Figure 3: More adults screened positive for albuminuria than dipstick protein; however, those with dipstick protein were more likely to experience rapid kidney function decline than those with albuminuria. Number with rapid kidney function decline (#RKFD) relative to the total number who screen positive (# Screen positive) at baseline and RKFD as a percentage of those who screen positive (% RKFD); by screening test (dipstick for protein or albuminuria).
Table 3 summarizes the diagnostic utility of albuminuria and/or thresholds for dipstick protein (≥trace, ≥1 g/L, or ≥3 g/L) at baseline to predict RKFD, overall and in risk subsets of the population. Overall, the probability of developing RKFD before screening was 9%. The greatest increase in the probability of RKFD came after a dipstick of ≥3 g/L. The probability of identifying RKFD from serial eGFR measurements increased from 9 to 41%; an increase in the certainty of the likelihood of disease of 7.8 after a positive test relative to a negative test. Overall, 42 (1.6%) had a dipstick protein ≥3 g/L, and if all were followed with serial eGFR measurements, one patient with RKFD would be identified for every 2.5 patients followed. The smallest increase in the probability of RKFD was seen among those with albuminuria. The probability of RKFD in this group was 20%, an increase in the certainty of the likelihood of disease of 3.2 after a positive test relative to a negative test. Overall, 63 (2.5%) had a dipstick protein value of ≥1 g/L, and if all were followed with serial eGFR measurements, one patient with RKFD would be identified for every 2.6 patients followed. On the basis of the result of a single screening test and using a criteria of ≥1 g/L of protein, 90.8% of patients were correctly identified as either RKFD or nonprogressive, 1.5% were mislabeled as RKFD, and 7.7% with eventual RKFD were missed.
Table 3: Comparison of the diagnostic utility of thresholds for dipstick protein and albuminuria at baseline to identify patients with future rapid kidney function decline: Overall and by risk subsets of the population
Similarly, among those with risk factors (history of CVD, diabetes, or age ≥60 years), using a criteria of a dipstick protein ≥1 g/L resulted in an increase in the probability of RKFD from 13 to 44%. Although albuminuria had greater sensitivity, particularly among diabetics, a higher false-positive rate resulted in a greater number needed to follow. In fact, among patients with no pre-existing risk factors, using albuminuria as a screening criteria resulted in only a slight increase in the certainty of the likelihood of disease of 1.2 after a positive test relative to a negative test, and the post-test probability of RKFD increased by <1%.
DISCUSSION
In this 7-year follow-up of a community-based cohort, dipstick proteinuria had better diagnostic utility for identifying patients at risk for RKFD than albuminuria. A dipstick for protein ≥1 g/L correctly identified 91% of those screened as either RKFD or nonprogressive. If all of the positive test results were followed with serial eGFR assessment, one patient with RKFD would be identified for every 2.6 patients followed. This decreased to 2.3 among those with CVD, diabetes, hypertension, or age ≥60 years.
Approximately 60 million individuals globally have some degree of chronic kidney disease, and early detection and prevention of kidney disease may be the only way to prevent death from CVD or ESRD.2 Hyperfiltration after renal injury and the inability to distinguish age-related decline from pathologic renal decline pose significant barriers to early diagnosis and treatment when using static tests of kidney function. Strategies that focus on identifying progressive renal disease in those with eGFR <60 ml/min per 1.73 m2 identify patients later in their disease, and the vast majority die of CVD before developing ESRD.11 Our analysis focused on all adults, including those with eGFR >60 ml/min per 1.73 m2 who may otherwise go undetected and yet are likely to experience greater therapeutic benefit if identified at an earlier stage.12 More than 80% of those with RKFD in our community-based cohort had an eGFR >60 ml/min per 1.73 m2. Importantly, Matsushita et al.5 observed an increased risk for CVD and all-cause mortality from RKFD in those with eGFR 60 to 90 ml/min per 1.73 m2 at baseline, and other studies of the general population show an increased risk for adverse outcomes from RKFD or proteinuria independently of baseline eGFR.4,6,13,14 Although albuminuria has greater sensitivity at low levels and is the recommended method for monitoring proteinuria in diabetics,15–17 we show that dipstick urinalysis, an inexpensive test that is already a common feature of primary care, has better diagnostic utility for identifying RKFD.
Despite the strong association between RKFD and adverse outcomes,4–7 population-wide screening for eGFR is not cost-effective.3,8,9 Our results support an alternative sequential screening strategy in which patients with dipstick protein ≥1 g/L are followed with serial eGFR assessment. Boulware et al.10 demonstrated that screening for dipstick proteinuria to slow renal progression was cost-effective in high-risk populations (age >60, hypertension, and diabetes). Although cost-effectiveness was not demonstrated in low-risk groups, the major cost in their model resulted from specialist evaluation of all patients with eGFR <90. Our analysis suggests that monitoring for RKFD before specialist referral would reduce the referral rate by 80% in those without risk factors and by 56% in those with risk factors (Table 3). Moreover, the comparator group in Boulware's analysis was “no screening”; however, general practitioners routinely perform dipstick urinalysis as part of an annual physical exam. Finally, prevention of CVD morbidity should also be accounted for, because a greater number patients with chronic kidney disease will experience CVD than ESRD.11,18
A standardized approach to treating patients with RKFD has yet to be developed; however, evidence that RKFD predicts CVD and mortality, even when kidney function is only mildly reduced (eGFR, 60 to 90),5 suggests that patients with dipstick protein ≥1 g/L and eGFR <90 ml/min per 1.73 m2 should be followed with serial eGFR assessment, and early intervention may be warranted for those with RKFD. In contrast, the meaning of RKFD at eGFR >90 ml/min per 1.73 m2 is less clear because there is no body of evidence documenting adverse outcomes in this group.19 Although previous research has established that proteinuria in the presence of eGFR <60 ml/min per 1.73 m2 is a stronger risk factor for CVD and ESRD than either indicator alone,13,20 serial eGFR assessment would provide additional information about a patient's risk trajectory. For example, in elderly patients with reduced but stable kidney function, a low eGFR (30 to 60 ml/min per 1.73 m2) may be an indicator of age-related comorbidity rather than vascular pathology, the latter being responsive to treatment with angiotensin-receptor blockers if caused by proteinuria.21,22 Depending on their cardiovascular risk profile and level of proteinuria, elderly patients with stable kidney function could be monitored in primary care,23–25 whereas those with RKFD are more likely to benefit from specialist referral because they are at greater risk for adverse outcomes.4–6 However, future research must assess the effect and cost-effectiveness of different follow-up strategies.
Although eGFR is the most readily available indicator of kidney function, there are limitations associated with its use, namely fluctuation in the assay of serum creatinine and increased measurement error associated with longitudinal tracking of eGFR. The median change in eGFR was −0.78 ml/min per 1.73 m2 per year with 16% experiencing an absolute annual decline >3 ml/min per 1.73 m2 per year and 9% experiencing a percentage annual decline >5%. The high variability in absolute change among those with baseline eGFR ≥90 ml/min per 1.73 m2 made percentage change a better measure of renal decline in this general population sample.5 In addition, percentage annual change was a stronger indicator of incident CVD than absolute change. We also conducted several sensitivity analyses to test the robustness of our findings and account for the imprecision of the Modification of Diet in Renal Disease (MDRD) GFR estimating equation at higher levels of GFR (Appendix I). Overall, dipstick proteinuria remained a stronger predictor of RKFD than albuminuria, regardless of how RKFD was defined or whether we excluded those with eGFR >90 ml/min per 1.73 m2 or those with highly variable rates of kidney function decline.
Our study has many strengths including routine collection of serum creatinine within a nonreferred community-based sample. We measured change in kidney function across a broad age range over a 7-year time span. Our study sample included individuals with a minimum of three annual serum-creatinine measurements, and seven measures were available for more than half of the sample. Selected participants were more likely to be older or obese, with a family history of hypertension or diabetes. No differences in baseline BP, hypertension, history of CVD, or family history of kidney disease were evident; however, selected participants had a lower baseline eGFR than the original cohort. It is possible that individuals concerned with their kidney function were more likely to join the study and participate longer. Most importantly, our sample had sufficient variability in both initial GFR and change over time to allow the estimation of risk of RKFD across the spectrum of eGFR categories.
CONCLUSION
Dipstick urinalysis, an inexpensive test that is already a common feature of primary care, has better diagnostic utility for identifying patients at risk for RKFD than albuminuria. Screening for dipstick proteinuria ≥1 g/L in patients with or without risk factors would allow clinicians to follow fewer patients with serial eGFR assessment to identify those with RKFD. This novel strategy, although not identifying all patients with silent RKFD, addresses the shortcomings of many prior studies by changing the focus from static eGFR assessment among those with an eGFR below 60 ml/min per 1.73 m2 to dynamic assessment of those with an eGFR both above and below 60 ml/min per 1.73 m2. This strategy will enable earlier identification of many patients with RKFD in the general population, of whom the large majority have eGFR above 60 ml/min per 1.73 m2. At the population level, this screening strategy may enable more appropriate targeting of resources. However, future research must assess the effect and cost-effectiveness of different follow-up strategies in independent cohorts.
CONCISE METHODS
Participants and Design
Data for this analysis were obtained from The Walkerton Health Study (2002 to 2008), a prospective cohort study designed to evaluate the long-term health sequelae after exposure to bacterially contaminated water.26 The study sample has previously been shown to be representative of the target population, with a slight over-representation of women and a slight under-representation of the very elderly.27 Ethics approval was obtained from the University of Western Ontario's Research Ethics Board for Health Sciences. Of 3,371 participants ≥18 years, we excluded 797, including those who developed hemolytic uremic syndrome as a result of the outbreak (n = 2) and those with fewer than three annual eGFR measurements (n = 791) or missing baseline data for random urine protein (n = 4).
Measures
Participants completed a baseline assessment that included questions on risk factors, family history, and doctor-diagnosed health conditions (following the format of the U.S. Third National Health and Nutrition Examination Survey). We measured height, weight, and BP using standardized protocols, and blood and urine samples were collected annually.26 We used a urine dipstick for protein (Bayer 8SG Multistix) to measure protein from a random spot urine sample, and the IMAGE Beckman Coulter immunoassay (Fullerton, CA) to measure the albumin:creatinine ratio. Serum creatinine was measured by the modified kinetic method of JAFFE using a Vitros 950 autoanalyzer, with an interassay coefficient of 4% (reference normal range, 59 to 117 μmol/L for men and 51 to 95 μmol/L for women). We calculated eGFR using the abbreviated MDRD equation.28
To calculate the rate of change in eGFR over time, we fitted an ordinary least-squares regression line to all eGFR measures for each participant. The slope of the regression line describes the rate of change in kidney function over time. On the basis of previous research, we defined RKFD as both an absolute annual decline >3 ml/min per 1.73 m2 per year4 and as a percentage annual decline >5% (slope/baseline eGFR*100).5
We defined CVD as a self-reported, doctor-diagnosed heart attack, stroke, or congestive heart failure and recorded the date of diagnosis. We defined diabetes mellitus by the current diagnostic criteria29 on the basis of random or fasting plasma glucose, or 2-hour oral glucose tolerance test or current medication (oral hypoglycaemic agents or insulin). We classified participants taking anti-hypertensive medication or presenting with a systolic/diastolic BP ≥140/90 mmHg (or ≥130/80 mmHg in the presence of diabetes or eGFR <60 ml/min per 1.73 m2)30 as hypertensive. We examined two random urine screening tests for proteinuria obtained at baseline: the dipstick for protein (≥trace, ≥1 g/L, and ≥3 g/L) and albuminuria (albumin:creatinine ratio >2.0 mg/mmol for men or >2.8 mg/mmol for women).31,32 We chose additional covariates on the basis of their biologic plausibility or prior studies, including: male gender, obesity (body mass index of ≥30 kg/m2), and family history of diabetes, kidney disease, or hypertension. We distinguished between recognized clinical risk factors for RKFD (age ≥60, CVD, diabetes, and hypertension) and screening tests for urine protein to improve prediction of RKFD.
Analyses
We used SAS 9.2 (SAS Institute, Inc., Cary, NC) for all analyses and summarized normally distributed data by the mean and SD and skewed distributions by the median and IQR. We tested the associations between RKFD and baseline characteristics and screening tests using t tests and χ2 tests as appropriate. We used a log-binomial model to estimate the adjusted RR for RKFD and tested covariates for inclusion at α = 0.20 and interactions among clinical risk factors, as well as interactions between clinical risk factors and baseline eGFR <60 at α = 0.05. We estimated the adjusted RR for RKFD at three separate cut-points for dipstick protein (≥trace, ≥1 g/L, and ≥3 g/L) and albuminuria at baseline by adding them separately to the clinical risk model. We assessed the diagnostic utility of thresholds for dipstick protein and albuminuria to identify RKFD, both overall and within high-risk subgroups (definitions in Appendix II).
Disclosures
None.
An abstract of this work was presented at the World Congress of Nephrology (Vancouver, April 8 to 12, 2011).
The funding source for this study was the Ontario Ministry of Health and Long-term Care. The funding source had no role in design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.
All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare that (1) no authors have received support from any companies for the submitted work; (2) no authors have any relationships with any companies that might have an interest in the submitted work in the previous 3 years; (3) their spouses, partners, or children have no financial relationships that may be relevant to the submitted work; and (4) no authors have nonfinancial interests that may be relevant to the submitted work.
Published online ahead of print. Publication date available at www.jasn.org.
See related editorial, “Screening for Proteinuria: A Tool for Predicting Rapid Declines in Kidney Function?,” on pages 1580–1583.
Supplemental information for this article is available online at http://www.jasn.org/.
REFERENCES
1. Zelmer JL: The economic burden of end-stage renal disease in Canada. Kidney Int 72: 1122–1129, 2007
2. Remuzzi G, Weening JJ: Albuminuria as early test for vascular disease. Lancet 365: 556–557, 2005
3. James MT, Hemmelgarn BR, Tonelli M: Early recognition and prevention of chronic kidney disease. Lancet 375: 1296–1309, 2010
4. Rifkin DE, Shlipak MG, Katz R, Fried LF, Siscovick D, Chonchol M, Newman AB, Sarnak MJ: Rapid kidney function decline and mortality risk in older adults. Arch Intern Med 168: 2212–2218, 2008
5. Matsushita K, Selvin E, Bash LD, Franceschini N, Astor BC, Coresh J: Change in estimated GFR associates with coronary heart disease and mortality. J Am Soc Nephrol 20: 2617–2624, 2009
6. Shlipak MG, Katz R, Kestenbaum B, Siscovick D, Fried L, Newman A, Rifkin D, Sarnak MJ: Rapid decline of kidney function increases cardiovascular risk in the elderly. J Am Soc Nephrol 20: 2625–2630, 2009
7. Al-Aly Z, Zeringue A, Fu J, Rauchman MI, McDonald JR, El-Achkar TM, Balasubramanian S, Nurutdinova D, Xian H, Stroupe K, Abbott KC, Eisen S: Rate of kidney function decline associates with mortality. J Am Soc Nephrol 21: 1961–1969, 2010
8. Clase CM: Glomerular filtration rate: Screening cannot be recommended on the basis of current knowledge. BMJ 333: 1030–1031, 2006
9. Glassock RJ, Winearls C: Screening for CKD with eGFR: Doubts and dangers. Clin J Am Soc Nephrol 3: 1563–1568, 2008
10. Boulware LE, Jaar BG, Tarver-Carr ME, Brancati FL, Powe NR: Screening for proteinuria in US adults: A cost-effectiveness analysis. JAMA 290: 3101–3114, 2003
11. Keith DS, Nichols GA, Gullion CM, Brown JB, Smith DH: Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med 164: 659–663, 2004
12. De Jong PE, Curhan GC: Screening, monitoring, and treatment of albuminuria: Public health perspectives. J Am Soc Nephrol 17: 2120–2126, 2006
13. Hallan SI, Ritz E, Lydersen S, Romundstad S, Kvenild K, Orth SR: Combining GFR and albuminuria to classify CKD improves prediction of ESRD. J Am Soc Nephrol 20: 1069–1077, 2009
14. Hemmelgarn B, Manns B, Lloyd A, James M, Klarenbach S, Quinn R, Wiebe N, Tonelli M: Relation between Kidney function, proteinuria, and adverse outcomes. JAMA 303: 423–429, 2010
15. National Kidney Foundation: KDOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Am J Kidney Dis 39: S1–S266, 2002
16. National Institute for Health and Clinical Excellence. Chronic kidney disease: Early identification and management of chronic kidney disease in adults in primary and secondary care. Available at
http://www.nice.org.uk/nicemedia/live/12069/42117/42117.pdf. Accessed July 1, 2011
17. Lambers Heerspink HJ, Gansevoort RT, Brenner BM, Cooper ME, Parving HH, Shahinfar S, De ZD: Comparison of different measures of urinary protein excretion for prediction of renal events. J Am Soc Nephrol 21: 1355–1360, 2010
18. Foley RN, Murray AM, Li S, Herzog CA, McBean AM, Eggers PW, Collins AJ: Chronic kidney disease and the risk for cardiovascular disease, renal replacement, and death in the United States Medicare population, 1998 to 1999. J Am Soc Nephrol 16: 489–495, 2005
19. Bauer C, Melamed ML, Hostetter TH: Staging of chronic kidney disease: Time for a course correction. J Am Soc Nephrol 19: 844–846, 2008
20. Ishani A, Grandits GA, Grimm RH, Svendsen KH, Collins AJ, Prineas RJ, Neaton JD: Association of single measurements of dipstick proteinuria, estimated glomerular filtration rate, and hematocrit with 25-year incidence of end-stage renal disease in the multiple risk factor intervention trial. J Am Soc Nephrol 17: 1444–1452, 2006
21. GISEN Group: Randomised placebo-controlled trial of effect of ramipril on decline in glomerular filtration rate and risk of terminal renal failure in proteinuric, non-diabetic nephropathy. The GISEN Group (Gruppo Italiano di Studi Epidemiologici in Nefrologia). Lancet 349: 1857–1863, 1997
22. Jafar TH, Stark PC, Schmid CH, Landa M, Maschio G, Marcantoni C, de Jong PE, de Zeeuw D, Shahinfar S, Ruggenenti P, Remuzzi G, Levey AS: Proteinuria as a modifiable risk factor for the progression of non-diabetic renal disease. Kidney Int 60: 1131–1140, 2001
23. Levin A, Djurdjev O, Beaulieu M, Er L: Variability and risk factors for kidney disease progression and death following attainment of stage 4 CKD in a referred cohort. Am J Kidney Dis 52: 661–671, 2008
24. Conway B, Webster A, Ramsay G, Morgan N, Neary J, Whitworth C, Harty J: Predicting mortality and uptake of renal replacement therapy in patients with stage 4 chronic kidney disease. Nephrol Dial Transplant 24: 1930–1937, 2009
25. Campbell KH, O'Hare AM: Kidney disease in the elderly: Update on recent literature. Curr Opin Nephrol Hypertens 17: 298–303, 2008
26. Clark W, Sontrop J, Macnab J, Salvadori M, Moist L, Suri R, Garg A: Long-term risk for hypertension, renal impairment and cardiovascular disease after gastroenteritis from drinking water contaminated with
E.coli O157:H7: Findings from a prospective cohort study. BMJ 341: c6020, 2010
27. Garg AX, Macnab J, Clark W, Ray JG, Marshall JK, Suri RS, Devereaux PJ, Haynes B: Long-term health sequelae following
E. coli and
Campylobacter contamination of municipal water: Population sampling and assessing non-participation biases. Can J Public Health 96: 125–130, 2005
28. Levey AS, Greene T, Kusek JW, Beck GJ: A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 11: 155A, 2000
29. Expert Committee on the Diagnosis and Classification of Diabetes Mellitus: Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 26[Suppl 1]: S5–S20, 2003
30. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr, Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella EJ: National High Blood Pressure Education Program Coordinating Committee: Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension 42: 1206–1252, 2003
31. Whaley-Connell AT, Sowers JR, Stevens LA, McFarlane SI, Shlipak MG, Norris KC, Chen SC, Qiu Y, Wang C, Li S, Vassalotti JA, Collins AJ: CKD in the United States: Kidney Early Evaluation Program (KEEP) and National Health and Nutrition Examination Survey (NHANES) 1999–2004. Am J Kidney Dis 51: S13–S20, 2008
32. Vassalotti JA, Stevens LA, Levey AS: Testing for chronic kidney disease: A position statement from the National Kidney Foundation. Am J Kidney Dis 50: 169–180, 2007
