SEARCH

SEARCH BY CITATION

Keywords:

  • Cockcroft-Gault formula;
  • glomerular filtration rate;
  • Modification of Diet in Renal Disease formula;
  • serum creatinine;
  • serum cystatin C

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

Objectives.  To estimate the prevalence of decreased kidney function in an elderly population and to evaluate the impact of using alternative markers of glomerular filtration rate (GFR), focusing on serum cystatin C (Cys C) and the Modification of Diet in Renal Disease (MDRD) Study prediction equation.

Design and methods. In a cross-sectional community-based survey renal function was assessed by serum creatinine (SCreat), Cys C and GFR predicted by the Cockcroft-Gault (CG) and the MDRD Study formulae. Associations with age, gender and proteinuria were analysed by linear models.

Subjects.  A total of 1246 elderly residents in Lieto, Finland, 64–100 years of age.

Results.  The prevalence of moderately or severely decreased renal function, estimated by the MDRD Study equation, was 35.7%; the CG formula yielded 58.6%. The profile of Cys C performance, including variation across age groups and level of health status, showed greater similarity to GFR estimated using the MDRD Study equation than to SCreat alone, or GFR estimated using the CG formula. Discordance between high Cys C levels and only mildly decreased GFR estimates was observed in subjects with functional limitations. Microalbuminuria was associated with Cys C levels only (P =0.047).

Conclusion.  Prevalence estimates of decreased renal function amongst the elderly vary considerably depending on prediction formula used. Variation in creatinine metabolism amongst elderly comorbid patients and the critical dependence on the SCreat assay and exact calibration, make the use of creatinine-based formulae to predict GFR questionable in geriatric clinical practice. In this setting, Cys C is a promising alternative.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

As the burden of chronic kidney disease in the elderly is increasing and the importance of implementing preventive measures has been recognized, there has been a growing interest in the epidemiology of renal dysfunction in the aged. However, for clinical purposes, there is no ideal marker of GFR and the limitations of the currently used tests are magnified in the elderly.

The most frequently used laboratory index of renal function, serum creatinine (SCreat), may be misleading in elderly persons, who may have severely compromised renal function despite a SCreat concentration within the normal range [1]. The measurement of creatinine clearance (CCR) requires timed urine collection, which especially in the elderly, has proven laborious and prone to collection failure [2]. Direct techniques for evaluating glomerular filtration rate (GFR), which involve inulin or radiopharmaceuticals, are expensive and impractical in clinical work or for screening large populations.

To overcome these problems, formulae for rapid evaluation of GFR based on SCreat and biometric data have been developed. The most widely used one, the Cockcroft-Gault formula (CG), predicts CCR from SCreat, age, weight and sex [3]. However, several studies have questioned its clinical utility amongst the elderly [4–6]. More recently, a formula, claimed to be more accurate than CCR or the CG formula, was developed from the Modification of Diet in Renal Disease (MDRD) Study data to predict GFR from age, sex, race, SCreat, serum albumin and serum urea [7]. However, it has not been validated in persons without renal disease, patients with serious comorbidities, or the elderly, and the better performance of the MDRD formula, compared with the CG formula, has been questioned in older patients with chronic renal insufficiency [8].

Furthermore, serum cystatin C (Cys C), a low-molecular weight protein steadily produced by all human nucleated cells and easily determined from serum samples, has been claimed to be superior to SCreat as an endogenous marker of GFR, especially in the elderly, because of its lack of dependence on muscle mass and less analytical interference [9, 10]. Glucocorticoid treatment, rheumatoid factor and hyperthyroidism have been recognized as nonrenal Cys C increasing factors [11–13].

The objectives of the present study were to estimate the prevalence of decreased kidney function in the elderly, and to evaluate the impact of using alternative markers of GFR, especially focusing on Cys C and the MDRD Study prediction equation.

Study design and population

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

The study population consisted of 1246 elderly residents (525 men, 721 women; mean age 74 years; range 64–100 years) in Lieto, a semi-industrialized rural municipality in south-western Finland. During 1998–1999, all Lieto residents born in or before 1933 (n = 1519) were invited to participate in a community-based cross-sectional epidemiological survey; 83% of those eligible gave a written consent. Subjects living in nursing homes or other institutions were included and constituted 5% of the study population. Fourteen subjects were excluded because of missing data. Information on chronic health conditions and current medications was obtained through standardized interviews, thorough clinical examinations and review of medical records. ICD-10 diagnosis codes were used for documentation. The Joint Commission of Ethics for the Hospital District of Varsinais-Suomi in south-western Finland approved the study protocol.

Measurements and definitions

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

Fasting venous blood samples and early morning spot urine samples were collected and then analysed in the Central Laboratory of Turku University Central Hospital. Cys C concentrations were determined from serum samples using a particle-enhanced nephelometric immunoassay (N Latex Cystatin C, BN II System; Dade Behring, Marburg, Germany) with an upper reference limit 0.95 mg L−1 for adults. The age range-specific reference limits earlier established for the study population were 1.3 (65–74 years) and 1.47 mg L−1 (≥75 years) [14]. Serum and urinary creatinine were measured using the Jaffé reaction (Roche Diagnostics, Mannheim, Germany and Hitachi 917; Hitachi Ltd, Tokyo, Japan). The reference limit was 118 μmol L−1 for men, and 104 μmol L−1 for women according to the reference values earlier determined for the population [14]. Serum albumin was analysed using a nephelometric immunoassay (BN II System), and urinary albumin was analysed using an immunoturbidimetric method (Optima, Microalbuminuria Kit; Thermo Clinical Labsystem, Helsinki, Finland). Serum urea was measured enzymatically (Boehringer, Mannheim, Germany and Hitachi 917, Hitachi Ltd). Clinical proteinuria was assessed by means of a urine dipstick test and microalbuminuria was determined by the urinary albumin/creatinine ratio (ACR).

Body mass index (BMI) was measured as kg per square metre. Body surface area (BSA) was estimated using the DuBois formula [15]. A 12-lead resting electrocardiogram was recorded and interpreted using the Minnesota Code [16]. Physical functioning was assessed using a questionnaire adapted from the protocol of the Eleven Countries Study and included 4 items on mobility (capability to walk outdoors, between rooms, in stairs, or at least 400 m) and 5 items on activities of daily living, i.e. dressing, eating, bathing, going to bed and using the toilet [17]. Each had scores of 0–3 (0, unable to do; 1, some help needed; 2, with difficulty, but no help needed; 3, no limitations). The maximum sum score, indicating no limitations in physical functioning, was 27.

Renal or urogenital disease and hypertension were categorized as being present if there was a documented diagnosis in the medical records. The diagnosis of urinary infection was based on significant bacterial growth in an early morning urine specimen or documentation of chronic urinary infection. Diabetes was defined based on the medical history or fasting glycaemia with glucose values of 7 mmol L−1 or higher. Coronary heart disease was diagnosed based on medical records or electrocardiography.

The GFR was estimated using the CG equation estimate of CCR adjusted for BSA (CG/BSA) and the MDRD Study equation 7 (MDRD7) as follows:

  • 1
    CG/BSA estimate (mL min−1/1.73 m2) = 1.23 × (140 − age) × weight/SCreat × 1.73/BSA (×0.85 if female)
  • 2
    MDRD7 estimate (mL min−1/1.73 m2) = 170 × (SCreat × 0.0113)−0.999 × age−0.176 × (SUrea × 2.8)−0.170 × (SAlbumin × 0.1)+0.318 (×0.762 if female)

where, weight was measured in kg, SCreat in μmol L−1, serum urea in mmol L−1 and serum albumin in g L−1.

Estimated GFR was categorized as normal (≥90 mL min−1/1.73 m2), mildly decreased (60–89 mL min−1/1.73 m2), moderately decreased (30–59 mL min−1/1.73 m2) or severely decreased (<30 mL min−1/1.73 m2) according to the Kidney Disease Outcomes Quality Initiative (K/DOQI) of the National Kidney Foundation guidelines [18].

In order to assess the impact of health status on the performance of markers of renal function, the population was divided into an apparently healthy group with no documentation of renal or urogenital disease, diabetes, hypertension, coronary heart disease, or clinical proteinuria (n = 337; 132 men and 205 women; mean SCreat 88 μmol L−1, mean Cys C 0.95 mg L−1) and a nonhealthy group consisting of the remaining subjects (n = 909; 393 men and 516 women; mean SCreat 96 μmol L−1, mean Cys C 1.13 mg L−1).

Statistical analyses

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

All statistical analyses were performed using SAS System for Windows, version 8.02 (SAS Institute Inc, Cary, NC, USA). Because of skewed distributions, Cys C and SCreat values were inversely transformed and ACR values were log-transformed for statistical analysis. Correlations were calculated using Pearson correlation coefficients. MDRD7, CG and CG/BSA estimates were compared using a paired t-test. Multivariate linear models were used to assess associations of age, gender and proteinuria with Cys C, SCreat and GFR estimates included as continuous dependent variables. Box-and-whisker plots were generated to show distribution of markers of renal function across subgroups based on age and health status. Analyses were adjusted for potential confounding factors. P < 0.05 was considered significant.

Results

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

Descriptive population data are given in Table 1. The MDRD7 formula yielded significantly higher estimates than the CG (P < 0.001) or the CG/BSA formula (P < 0.001), but the between-gender difference, with higher estimates in men, was less pronounced when the CG/BSA formula was used (Fig. 1). After correcting for age, diabetes, hypertension, coronary heart disease and urinary infection, the between-gender difference remained highly significant for both formulae (P < 0.001). The level of kidney function modified the between-gender difference of MDRD7 estimates, with a proportionally smaller difference at lower levels. After adjustment for age and BMI, the interaction between gender and level of kidney function was significant when decreased kidney function was categorized as Cys C >1.3 mg L−1 (P = 0.017) or SCreat >118 μmol L−1 (men) and >104 μmol L−1 (women) (P = 0.049) [14].

Table 1.  Population characteristics by gender
CharacteristicsMen (n = 525)Women (n = 721)All (n = 1246)
  1. Values are given as mean (SD). MDRD7, Modification of Diet in Renal Disease Study, equation 7; GFR, glomerular filtration rate; CG, Cockcroft-Gault formula; CCR, creatinine clearance; BSA, body surface area.

Age (year)73.2 (6.4)74.5 (7.0)74.0 (6.8)
Body mass index (kg m−2)26.8 (3.9)27.2 (5.2)27.0 (4.7)
Serum cystatin C (mg L−1)1.09 (0.38)1.08 (0.35)1.08 (0.36)
Serum creatinine (μmol L−1)101.2 (20.2)89.1 (18.4)94.2 (20.1)
Serum albumin (g L−1)41.3 (3.8)41.1 (3.9)41.2 (3.9)
Serum urea (mmol L−1)6.99 (2.43)6.57 (2.36)6.75 (2.40)
MDRD7 GFR (mL min−1/1.73 m2)69.1 (13.2)60.3 (11.7)64.0 (13.1)
CG CCR (mL min−1)67.8 (17.5)55.1 (15.8)60.4 (17.7)
CG/BSA CCR (mL min−1/1.73 m2)59.7 (12.6)55.1 (12.7)57.0 (12.9)
image

Figure 1. (a) Distribution of estimated creatinine clearance (CCR) calculated by the Cockcroft-Gault equation adjusted for body surface area (CG/BSA) by gender (n = 1246). (b) Distribution of estimated glomerular filtration rate (GFR) calculated by the Modification of Diet in Renal Disease Study equation 7 (MDRD7) by gender (n = 1246).

Download figure to PowerPoint

The Cys C correlated significantly with SCreat (r = 0.61), CG estimates (r = 0.45), CG/BSA estimates (r = 0.61) and MDRD7 estimates (r = 0.68). P was <0.001 in all cases.

Prevalence of decreased kidney function and distribution of GFR categories

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

The prevalence of normal kidney function estimated using the MDRD7 equation was 2.2% and the prevalence of mildly, moderately and severely decreased function was 62.1%, 34.4% and 1.4%, respectively. When the CG/BSA formula was used the corresponding proportions were 0.6, 40.8, 56.1 and 2.5%.

The distribution of Cys C concentrations, indicating proportions of MDRD7 estimated GFR categories at each Cys C level, is shown in Fig. 2. All MDRD7 estimates higher than 80 mL min−1/1.73 m2 occur at Cys C concentrations below 1.26 mg L−1, and all estimates higher than 90 mL min−1/1.73 m2 at Cys C concentrations below 1.03 mg L−1. Severely decreased MDRD7 estimates are not found at Cys C concentrations below 1.66 mg L−1. Small proportions of moderately decreased kidney function are found in a wide range of Cys C concentrations (0.7–3.5 mg L−1), but from 1.3 mg L−1 onwards, the majority of MDRD7 estimates are classified as either moderately or severely decreased. All of nine subjects classified to have mildly decreased kidney function, in spite of high Cys C levels (1.62–2.98 mg L−1), experienced limitations regarding physical functioning [functional index questionnaire: mean sum score 16.6 (SD: 9.3) vs. 24.0 (SD: 5.4) for the whole population]. Rheumatoid factor was not measured, but only one of these nine subjects was diagnosed to have rheumatoid arthritis. No one was treated with glucocorticoids and no one was in a hyperthyroid state.

image

Figure 2. Distribution of serum cystatin C (Cys C) levels in the study population (n = 1246). The proportion of individuals at each serum Cys C level having a glomerular filtration rate (GFR) in the ranges indicated is calculated using the Modification of Diet in Renal Disease Study equation 7 and denoted within each bar.

Download figure to PowerPoint

When the gender-specific SCreat reference limits earlier established for the same population were applied, 11.9% (N = 148) of the subjects had elevated levels. When the adult reference limit for Cys C (0.95 mg L−1) was applied, 58.1% (N = 724) had elevated levels, whereas the age-specific regression-based reference limits for Cys C yielded 14% (N = 175) elevated concentrations [14].

Impact of age and health status on distribution of markers of renal function

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

In box-and-whisker plots across groups based on age and health status, the Cys C concentrations and MDRD7 estimates showed similar profiles with gradually increasing/decreasing levels at older age (Fig. 3). However, in the healthy group, the impact of age was more pronounced for Cys C. The SCreat concentrations showed little variation with age, especially in the healthy group, and the CG/BSA estimates showed a steep decline with age in both groups. Corrected for gender, BMI, hypertension, diabetes, urinary infection, coronary heart disease and use of oral glucocorticoids, the difference between age groups was significant in all cases in the Cys C and CG/BSA analyses, whereas for MDRD7 estimates, the differences were significant in all cases in the nonhealthy group only. In SCreat analysis, the differences between the youngest and the other two age groups, respectively, were significant in the nonhealthy group.

image

Figure 3. Box-and-whisker plots showing distribution of serum cystatin C (Cys C), serum creatinine (SCreat), glomerular filtration rate estimated using the Modification of Diet in Renal Disease Study equation 7 (MDRD7) and the Cockcroft-Gault equation normalized for body surface area (CG/BSA) in three age groups, separately for the nonhealthy group (64–74 years, n = 529; 75–85 years, n = 293; >85 years, n = 87) and the healthy group (64–74 years, n = 255; 75–85 years, n = 66; >85 years, n = 16). 25, 50 and 75 percentiles, whiskers at ±1.5 time the interquartile range and outliers are denoted. P-values indicate statistical significance between groups after adjustment for gender, body mass index, hypertension, diabetes, urinary infection, coronary heart disease and use of oral glucocorticoids.

Download figure to PowerPoint

Association between markers of renal function and proteinuria

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

After adjustment for age, gender, and urinary infection, clinical proteinuria was significantly associated with Cys C concentrations (P = 0.045) and MDRD7 estimates (P = 0.014), but not with SCreat concentrations (P = 0.08) or CG/BSA estimates (P = 0.16). Urinary ACR, included as a continuous variable, excluding 67 subjects with clinical proteinuria, was significantly associated with Cys C concentrations only (P = 0.047). When urinary ACR was treated as a dichotomous variable with 2 mg mmol−1 as cut point, significance was reached in Cys C analysis if subjects with clinical proteinuria were included (P = 0.033).

Discussion

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

Community-based large-scale studies evaluating the prevalence of decreased kidney function amongst the elderly are scarce. Most studies have used SCreat, CCR, or SCreat-based prediction equations for evaluation of glomerular function; representative studies applying direct gold standard laboratory techniques are not available. There are no previous studies evaluating the use of Cys C versus the MDRD Study prediction equation in a large representative elderly population.

Prevalence estimates of stages of decreased kidney function, as predicted by the MDRD Study formula, vary across studies, depending on differences in recruitment, age of subjects, and calibration of the SCreat assay. Based on the Third National Health and Nutrition Examination Survey (NHANES III), using the abbreviated MDRD Study equation, excluding serum albumin and serum urea as variables, and SCreat concentrations recalibrated to the laboratory involved in the MDRD Study, Coresh et al. reported a prevalence of 20.6% of moderately or severely decreased kidney function amongst noninstitutionalized subjects older than 65 years in the United States [19]. Another study using the MDRD7 equation and SCreat concentrations, which were not recalibrated, reported a considerably higher prevalence of moderately or severely decreased kidney function in the white elderly nondiabetic participants in the NHANES III; 42.4% for men aged 70–79 years, and 57% for similarly aged women [20]. In the Cardiovascular Health Study (CHS) where indirect recalibration was used and ≈10 μmol L−1 (0.11 mg dL−1) was subtracted from the original SCreat concentrations, 21% of the subjects 65 years and older had a normal predicted GFR, whilst 23.4% had moderately or severely decreased estimates at baseline [21]. In the present study, which was characterized by a high participation rate and a wide age range, with no exclusions based on residential or health status, the proportion of persons categorized as having moderately or severely decreased function was higher (35.7%), and the proportion of those with normal kidney function was considerably lower (2.2%) than reported by Coresh et al. or the CHS. In the present study, disabled and polymorbid persons and institutionalized residents were actively recruited and, hence, nonparticipants were judged not to represent the most diseased part of the population. As far as can be judged from available data, the population was older and less selected for excellent health compared with the studies quoted above.

The variable prevalence estimates and between-gender differences found across studies also indicate calibration bias, accounting for SCreat differences up to 20 μmol L−1 across laboratories, with a substantial impact in the normal-near normal range of SCreat levels and, therefore, a proportionally greater impact in women with lower mean SCreat concentrations [22]. In the present study, recalibration of the SCreat measurements to results obtained at the laboratory used in the validation of the MDRD Study equation was not possible [23], but based on a comparison of laboratory techniques used and results obtained in other studies using the MDRD Study formula, they were judged to be consistently higher. Hence, calibration bias was thought to be the main reason for the low MDRD7 estimates observed in this study. For better comparability between studies, a specific correction factor should be derived for each equation in each laboratory, but outside the university clinics this is obviously not convenient.

Another possible reason for the observed differences between studies involves the question of reproducibility of creatinine-based equations outside the population in which they were developed, as SCreat generation varies across age, gender and ethnicity [1]. The MDRD Study population was recruited amongst persons <70 years of age diagnosed with chronic kidney disease (mean SCreat 200 μmol L−1), and the equation has not been validated in a community-based elderly population. Black ethnicity has been recognized as a risk factor for altered creatinine generation [7, 24], but the possible differences between individuals of variable decent within the white race, have not been explored. Aside from hereditary differences in body composition, cultural and environmental differences may affect dietary protein intake, and, especially in elderly populations, creatinine metabolism may deviate from the average because of malnourishment, disability and comorbidity. In the CHS, 196 subjects, predominantly women, of 4893, had estimated GFR between 130 and 200 mL min−1/1.73 m2, thought to be falsely elevated because of low muscle mass [21]. In the present study, a majority of subjects having high Cys C levels in spite of only mildly decreased GFR predicted by the MDRD7 equation, had several functional limitations, probably indicating decreased muscle mass and, hence, low SCreat level despite altered renal function.

Prediction equations also assume gender differences in creatinine production to be unaffected by age and variation in health status. After adjustment for age and morbidity related to kidney dysfunction, a significant between-gender difference in MDRD7 estimates was observed. Coresh et al. reported no gender difference after adjustment for age [19]. In the MDRD Study population, older women, compared with younger, had higher weight and body mass index, whereas no such difference was seen in men [25]. This could imply that the female coefficient, developed in the MDRD Study population, may perform differently in an aged population. We also found the between-gender difference to be significantly modified by the level of kidney function, probably indicating that the female coefficient performs differently in a population not selected for decreased kidney function.

Prediction equations are less precise in the higher range of GFR, where the role of the kidney in determining SCreat is of comparable magnitude with the variation in creatinine metabolism and analytical interference. Other markers, such as proteinuria, are recommended to identify early decline in renal function, especially in the elderly, as there is no general agreement defining the lower limit of normal GFR in this group [18]. Clinical proteinuria was significantly associated with Cys C and MDRD7 estimates, whereas microalbuminuria was associated with Cys C only. This indicates Cys C to be sensitive in identifying early kidney damage. However, urinary ACR was assessed from a single early morning urine specimen that captured both persistent and transient microalbuminuria, and results must be interpreted accordingly.

The design of this large epidemiological study did not include gold standard techniques to compare the diagnostic accuracy between markers of GFR. However, at the population level, the profile of performance of Cys C, including variation across age groups and levels of health status and proteinuria, showed greater similarity to the MDRD7 estimates than to SCreat alone or the CG formula. There was a good agreement between Cys C concentrations and MDRD7 estimates at high and low levels, but in the area of mildly/moderately reduced kidney function, the pattern was less distinct. CG/BSA estimates were significantly lower than MDRD7 estimates and showed a steeper decline with age. Earlier studies have concluded that the CG formula underestimates GFR in the elderly, especially in the upper range [4–6]. The K/DOQI guidelines recommend that GFR should preferably be predicted by the MDRD Study formula or the CG formula. In the present study, these formulae yielded considerably diverging prevalence estimates of decreased renal function amongst the elderly.

Regardless of which prediction equation was used, the proportion of normal kidney function was surprisingly small and the proportion of mildly decreased kidney function was substantial. To what extent mildly reduced renal function represents pathology and to what extent normal ageing, is unclear. The Baltimore Longitudinal Study of Ageing showed that loss of renal function with ageing may not be inevitable in healthy older people [26, 27]. Establishing reference intervals for laboratory markers of GFR in older people is complicated because the impact of subclinical pathology is difficult to completely eliminate by conventional procedures, and health outcomes of decreased GFR in the elderly, with or without chronic kidney disease, are not known [18, 28]. The level of decreased renal function, at which dose-adjustment of drugs excreted by the kidney is needed, may differ from the level at which other clinical interventions are judged to be beneficial. However, according to the K/DOQI guidelines, a clinical action plan based on the level of GFR requires knowledge of age-associated normal values [18]. The clear age-dependent rise in Cys C concentrations, observed in carefully examined apparently healthy elderly subjects, probably reflects the impact of age per se [14]. When the adult reference limit for Cys C was applied in the present study, 58.1% of the concentrations were elevated, as opposed to only 14% when age-adjusted limits were used. This may indicate that Cys C offers a possibility to estimate the age-dependent as well as the disease-related declines of renal function and, hence, a possibility to develop guidelines for screening and optimal treatment of older people with mildly decreased renal function in geriatric and primary care settings. However, further research by means of longitudinal studies, employing different laboratory techniques, is still needed to clarify the health outcomes of decreased renal function in the elderly.

Prevalence estimates of decreased renal function amongst the elderly vary considerably depending on which marker of GFR is used. The use of Cr-based equations to estimate GFR relies on the calibration of SCreat concentrations, proper validation in the population, and the individual having an average body composition, age, gender, and ethnicity taken into account. Despite recommendations, exact calibration of the SCreat assay is not easily implemented in the clinical setting. The superiority of Cr-based formulae over SCreat alone cannot be questioned in the elderly, but the clinical utility of these formulae in elderly subjects with altered creatinine metabolism caused by disability and comorbidity, can be disputed. At least in these cases, Cys C is a promising alternative.

Acknowledgements

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References

The study was supported in part by grants from The Red Feather Campaign of the Nordic Lions Clubs and the February 19th Foundation of the Finnish Heart Association.

The authors thank Dade Behring, Marburg, Germany, for providing kits for the measurement of Cys C.

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and methods
  5. Study design and population
  6. Measurements and definitions
  7. Statistical analyses
  8. Results
  9. Prevalence of decreased kidney function and distribution of GFR categories
  10. Impact of age and health status on distribution of markers of renal function
  11. Association between markers of renal function and proteinuria
  12. Discussion
  13. Conflict of interest statement
  14. Acknowledgements
  15. References
  • 1
    Perrone RD, Madias NE, Levey AS. Serum creatinine as an index of renal function: new insights into old concepts. Clin Chem 1992; 38: 193353.
  • 2
    Goldberg TH, Finkelstein MS. Difficulties in estimating glomerular filtration rate in the elderly. Arch Intern Med 1987; 147: 14303.
  • 3
    Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976; 16: 3141.
  • 4
    Fliser D, Franek E, Joest M, Block S, Mutschler E, Ritz E. Renal function in the elderly: impact of hypertension and cardiac function. Kidney Int 1997; 51: 1196204.
  • 5
    Baracskay D, Jarjoura D, Cugino A, Blend D, Rutecki GW, Whittier FC. Geriatric renal function: estimating glomerular filtration in an ambulatory elderly population. Clin Nephrol 1997; 47: 2228.
  • 6
    Verhave JC, Balje-Volkers CP, Hillege HL, de Zeeuw D, de Jong PE. The reliability of different formulae to predict creatinine clearance. J Intern Med 2003; 253: 56373.
  • 7
    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: 46170.
  • 8
    Lamb EJ, Webb MC, Simpson DE, Coakley AJ, Newman DJ, O'Riordan SE. Estimation of glomerular filtration rate in older patients with chronic renal insufficiency: is the modification of diet in renal disease formula an improvement? J Am Geriatr Soc 2003; 51: 10127.
  • 9
    Dharnidharka VR, Kwon C, Stevens G. Serum cystatin C is superior to serum creatinine as a marker of kidney function: a meta-analysis. Am J Kidney Dis 2002; 40: 2216.
  • 10
    Fliser D, Ritz E. Serum cystatin C concentration as a marker of renal dysfunction in the elderly. Am J Kidney Dis 2001; 37: 7983.
  • 11
    Wasen E, Isoaho R, Mattila K, Vahlberg T, Kivela SL, Irjala K. Serum cystatin C in the aged: relationships with health status. Am J Kidney Dis 2003; 42: 3643.
  • 12
    Lamb E, Stowe H. Rheumatoid factor can interfere with cystatin C measurement. Ann Clin Biochem 2003; 40: 1956.
  • 13
    Jayagopal V, Keevil BG, Atkin SL, Jennings PE, Kilpatrick ES. Paradoxical changes in cystatin C and serum creatinine in patients with hypo- and hyperthyroidism. Clin Chem 2003; 49: 6801.
  • 14
    Wasen E, Suominen P, Isoaho R et al. Serum cystatin C as a marker of kidney dysfunction in an elderly population. Clin Chem 2002; 48: 113840.
  • 15
    DuBois D, DuBois E. A formula to estimate the approximate surface area if height and weight be known. Arch Intern Med 1916; 17: 86371.
  • 16
    Rose G, Blackburn H, Gillum R, Prineas R. Cardiovascular Survey Methods, 2nd edn. Geneva, Switzerland: World Health Organization Monograph Series no. 56, 1982.
  • 17
    HeikkinenE, WatersWE, BrzezinskiZJ. (eds) The elderly in eleven countries. A sociomedical survey. In: Public Health in Europe 21. Copenhagen, Denmark: World Health Organization, 1983; 161231.
  • 18
    K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Kidney Disease Outcome Quality Initiative. Am J Kidney Dis 2002; 39: S1246.
  • 19
    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: 112.
  • 20
    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: 133849.
  • 21
    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: 11219.
  • 22
    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: 9209.
  • 23
    Lupovitch A. More accurate alternatives to serum creatinine for evaluating glomerular filtration rate. Clin Chem 2002; 48: 22978 (author reply).
  • 24
    Lewis J, Agodoa L, Cheek D et al. Comparison of cross-sectional renal function measurements in African-Americans with hypertensive nephrosclerosis and of primary formulas to estimate glomerular filtration rate. Am J Kidney Dis 2001; 38: 74453.
  • 25
    Coggins CH, Lewis JB, Caggiula AW, Castaldo LS, Klahr S, Wang SR. Differences between women and men with chronic renal disease. Nephrol Dial Transplant 1998; 13: 14307.
  • 26
    Lindeman RD, Tobin J, Shock NW. Longitudinal studies on the rate of decline in renal function with age. J Am Geriatr Soc 1985; 33: 27885.
  • 27
    Lindeman RD. Is the decline in renal function with normal aging inevitable? Geriatr Nephrol Urol 1998; 8: 79.
  • 28
    Lamb EJ, O'Riordan SE, Delaney MP. Kidney function in older people: pathology, assessment and management. Clin Chim Acta 2003; 334: 2540.