- Email: [email protected]
Serum cystatin C in the aged: relationships with health status
Serum Cystatin C in the Aged: Relationships With Health Status Elise Wase´n, MD, Raimo Isoaho, MD, PhD, Kari Mattila, PhD, Tero Vahlberg, MSc, Sirkka-Liisa Kivela¨, MD, PhD, and Kerttu Irjala, MD, PhD ● Background: Serum cystatin C (Cys C) is claimed to be superior to serum creatinine (Cr) in estimating glomerular filtration rate, but its utility in assessing renal function in the polymorbid elderly needs to be evaluated. Methods: In a cross-sectional, community-based survey performed in Lieto in southwestern Finland, Cys C, Cr, and urinary albumin-creatinine ratio (ACR) were measured in 1,260 subjects aged 64 to 100 years. Associations of demographic characteristics and health status factors with levels of Cys C, Cr, and ACR were assessed by means of linear models. Results: In men, hypertension, coronary heart disease, urinary infection, rheumatoid arthritis, glucocorticoid treatment, older age, and lower functional status were found to be significant predictors of higher Cys C values, whereas hypertension, coronary heart disease, urinary infection, older age, and increasing body mass index (BMI) significantly predicted higher Cr values. Among women, corresponding factors were hypertension, glucocorticoid treatment, age, functional status, and BMI for Cys C and hypertension, BMI, and age for Cr. Diabetes was significantly associated only with ACR. These factors explained 35% of variation in Cys C values in men and 34.5% in women versus only 14.8% and 11.3% for Cr, respectively. Conclusion: Glucocorticoid treatment was recognized as an independent Cys C–increasing factor, presumably nonglomerular. In comparison with Cys C, a considerably greater proportion of total variation in Cr values seems to be explained by extrarenal factors. Am J Kidney Dis 42:36-43. © 2003 by the National Kidney Foundation, Inc. INDEX WORDS: Cystatin C (Cys C); creatinine (Cr); glomerular filtration rate (GFR); albumin-creatinine ratio (ACR); elderly population; multivariate analysis.
T
HE INCIDENCE OF end-stage renal disease has increased in recent years worldwide, and a growing proportion of patients with chronic kidney diseases are elderly.1,2 Involutional changes render the aging kidney susceptible to impairment by disease and nephrotoxic agents and make timely identification and monitoring of renal dysfunction crucial in geriatric care.3 Because accumulating evidence indicates that renoprotective measures might delay the progression of renal failure, it also is important to identify determinants of early renal dysfunction.4 From the Institute of Clinical Medicine, General Practice; Departments of Biostatistics; and Clinical Chemistry, University of Turku; Central Laboratory, Turku University Central Hospital; Mehila¨inen Corporation, Turku; Salo Health Center, Salo; Ha¨rka¨tie Health Center, Lieto; and Satakunta Central Hospital, Pori, Finland. Received December 30, 2002; accepted in revised form February 28, 2003. 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. Presented in abstract form at the World Congress of Nephrology, Berlin, Germany, June 8-12, 2003. Address reprint requests to Elise Wase´n, MD, Institute of Clinical Medicine, General Practice, University of Turku, Lemminka¨isenkatu 1, FIN-20014, Turku, Finland. E-mail: [email protected] © 2003 by the National Kidney Foundation, Inc. 0272-6386/03/4201-0004$30.00/0 doi:10.1016/S0272-6386(03)00406-2 36
Serum creatinine (Cr) level is used routinely in assessing glomerular filtration rate (GFR), although several widely recognized limitations compromise its usefulness in clinical work. The impact of muscle mass on Cr values makes interpretation difficult, especially in elderly individuals.5 Serum cystatin C (Cys C) is believed to be a more sensitive indicator of GFR in the general population and specific patient categories.6 In our earlier study, we found Cys C to be more reliable than Cr in assessing glomerular function in the healthy elderly.7 In acutely hospitalized elderly patients, Cys C was not found to be superior to Cr level in the detection of renal impairment8,9; however, small sample size and possible interference by medical treatment and pseudochromogens produced in acute illness may have influenced the results. Microalbuminuria has been shown to indicate incipient nephropathy in patients with diabetes and other patients with chronic vascular disease.10 Determinants of elevated Cr levels and microalbuminuria have been reported in populationbased studies including elderly subjects,11,12 whereas Cys C has not been evaluated extensively in this respect in representative elderly populations. In this study, we assessed the associations of demographic and health status characteristics with levels of Cys C and Cr and urinary
American Journal of Kidney Diseases, Vol 42, No 1 (July), 2003: pp 36-43
CYSTATIN C IN THE AGED
albumin-creatinine ratio (ACR) by means of multivariate analyses in an elderly population. The main objective is to evaluate features of CysC versus Cr level in this context. METHODS
Study Population and Data Collection The study population consisted of 1,260 elderly residents (533 men, 727 women; mean age, 74 years; range, 64 to 100 years) in Lieto, a semi-industrialized rural municipality in southwestern Finland. During 1998 to 1999, all Lieto residents born in or before 1933 (n ⫽ 1,519) were invited to participate in a community-based, cross-sectional epidemiological survey; 83% of those eligible gave a written consent and were included. Information on chronic health conditions, current medications, and physical functioning was obtained through standardized interviews, thorough clinical examination, and review of medical records. International Classification of Diseases, Tenth Revision, diagnosis codes were used for documentation. The Joint Commission of Ethics for the Hospital District of Varsinais-Suomi in southwestern Finland approved the study protocol.
Measurements and Definitions Fasting venous blood samples and early-morning spot urine samples were collected and then analyzed in the Central Laboratory of Turku University Central Hospital. Before the laboratory visit, all participants were given verbal and written instructions. 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).13 Cr was measured using the Jaffe´ reaction (Roche Diagnostics, Mannheim, Germany, and Hitachi 917; Hitachi Ltd, Tokyo, Japan). Urinary albumin was analyzed using an immunoturbidimetric method (Optima, Microalbuminuria Kit, Thermo Clinical Labsystem, Helsinki, Finland), and urinary creatinine concentrations were determined from the same sample using the Jaffe´ method. Clinical proteinuria was assessed by means of a urine dipstick test. A 12-lead resting electrocardiogram was recorded and interpreted using the Minnesota Code.14 Hypertension was categorized as present if a diagnosis was documented in the medical records. Diabetes was defined by medical history or fasting glycemia with glucose of 126 mg/dL or greater (ⱖ7 mmol/L), acknowledging that type 2 diabetes may be present years before clinical recognition.15 Coronary heart disease was diagnosed based on medical records or electrocardiogram findings. Diagnosis of urinary infection was based on significant bacterial growth in an early-morning urine specimen or documentation of chronic urinary infection. 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; ie, dressing, eating, bathing, going to bed, and using the toilet. Each had scores of 0 to 3 (0, unable to do; 1, some help needed; 2, with difficulty, but no help needed; 3, no limitations).16 According to the sum score, the functional index
37
was categorized as less than 23, 23 to 26, or 27 (no limitations). Body mass index (BMI), measured as kilograms per square meter, was classified as less than 20, 20 to 24.9, 25 to 29.9, 30 to 34.9, and 35 or greater according to the World Health Organization recommendation, except for the lowest cutoff value, which was placed at 20 instead of 18.5, to ensure sufficient numbers in the lowest class.17
Design and Analysis All statistical analyses were performed using SAS System for Windows, version 8.02 (SAS Institute Inc, Cary, NC). Because of skewed distributions, Cys C and Cr values were inversely transformed and ACR values were log transformed for statistical analysis. Associations of independent variables with the dependent variables Cys C, Cr, and ACR were evaluated using univariate and multivariate linear models. Explanatory variables were selected to multivariate analyses if they reached statistical significance in univariate analysis, but were excluded if they did not remain significant in at least one of the models. The final list of explanatory variables included in all multivariate models consisted of age (continuous variable), hypertension, coronary heart disease, diabetes, urinary infection, rheumatoid arthritis, and oral glucocorticoid treatment (2 categories), as well as BMI (5 categories) and functional index (3 categories). All analyses were stratified by sex. Correlations were calculated using nonparametric Spearman’s rank-order correlation coefficients. P less than 0.05 is considered significant. Missing data and exclusion of subjects with clinical proteinuria (n ⫽ 68) reduced the final study sample to 509 men and 702 women in the Cys C and Cr analyses and 458 men and 668 women in the ACR analyses.
RESULTS
Some characteristics of the population and mean ⫾ SD and median values of Cys C, Cr, and ACR according to the explanatory variables are listed in Tables 1 through 4. Age correlated with Cys C level (men, r ⫽ 0.444, P ⬍ 0.001; women, r ⫽ 0.465, P ⬍ 0.001), Cr level (men, r ⫽ 0.157, P ⬍ 0.001; women, r ⫽ 0.164, P ⬍ 0.001), and ACR (men, r ⫽ 0.253, P ⬍ 0.001; women, r ⫽ 0.263, P ⬍ 0.001). In the multivariate linear model, hypertension, coronary heart disease, urinary infection, rheumatoid arthritis, glucocorticoid treatment, older age, and lower functional index were significantly associated with greater Cys C values in men (Table 5). Hypertension, coronary heart disease, urinary infection, advancing age, and increasing BMI were significantly related to greater Cr levels in men. Among women, corresponding factors were hypertension, glucocorticoid treatment, age, functional index, and BMI for Cys C and hypertension, BMI, and age for Cr.
WASE´ N ET AL
38 Table 1.
Age (y) BMI (kg/m2) Serum Cys C (mg/L) Serum Cr (mg/dL) Urinary ACR (mg/g)
Characteristics of the Population
Men (N ⫽ 533)
Women (N ⫽ 727)
All (N ⫽ 1,260)
73.2 ⫾ 6.4 26.8 ⫾ 3.9 1.09 ⫾ 0.30 1.15 ⫾ 0.23 33.8 ⫾ 83.2
74.6 ⫾ 7.1 27.2 ⫾ 5.2 1.09 ⫾ 0.35 1.01 ⫾ 0.21 26.6 ⫾ 64.6
74.0 ⫾ 6.8 27.0 ⫾ 4.7 1.09 ⫾ 0.37 1.07 ⫾ 0.23 29.7 ⫾ 73.2
NOTE. Values expressed as mean ⫾SD. To convert Cr in mg/dL to mol/L, multiply by 88.4; urinary ACR in mg/g to mg/mmol, divide by 8.84.
erably greater in multivariate analyses for Cys C compared with those for Cr (35% versus 14.8% in men and 34.5% versus 11.3% in women). In ACR analyses, corresponding percentages were 17.3% (men) and 17.8% (women). When sex was entered into the full model, R2 changed from 33.0% to 33.4% in Cys C analysis, from 8.7% to 22.9% in Cr analysis, and from 15.3% to 15.9% in ACR analysis.
Diabetes was significantly associated with greater levels of microalbuminuria, measured by ACR, in both sexes (Table 6), but neither was associated with Cys C or Cr level. In men, other variables significantly associated with higher levels of ACR were hypertension and age, whereas in women, urinary infection, age, functional index, and BMI reached significance. Coefficients of determination (R2) were considTable 2. Variable
Hypertension Yes No Coronary heart disease Yes No Diabetes Yes No Urinary infection Yes No Rheumatoid arthritis Yes No Use of glucocorticoids Yes No BMI (kg/m2) ⬍20 20–24.9 25–29.9 30–34.9 ⱖ35 Functional index ⬍23 23–26 27
Serum Cys C and Serum Cr Values According to Health Status Variables in Men No.
Cys C (mg/L)
Cr (mg/dL)
Mean ⫾ SD
Median
Mean ⫾ SD
Median
173 336
1.13 ⫾ 0.39 1.05 ⫾ 0.34
1.03 0.97
1.20 ⫾ 0.26 1.11 ⫾ 0.20
1.15 1.07
212 297
1.15 ⫾ 0.45 1.02 ⫾ 0.27
1.04 0.97
1.18 ⫾ 0.26 1.11 ⫾ 0.19
1.13 1.07
84 425
1.09 ⫾ 0.34 1.07 ⫾ 0.37
0.97 0.99
1.17 ⫾ 0.24 1.14 ⫾ 0.22
1.12 1.09
27 482
1.47 ⫾ 0.50 1.05 ⫾ 0.34
1.37 0.98
1.34 ⫾ 0.33 1.13 ⫾ 0.21
1.31 1.09
22 487
1.28 ⫾ 0.52 1.07 ⫾ 0.35
1.15 0.99
1.20 ⫾ 0.25 1.14 ⫾ 0.22
1.16 1.09
21 488
1.37 ⫾ 0.78 1.06 ⫾ 0.33
1.15 0.99
1.18 ⫾ 0.26 1.14 ⫾ 0.22
1.12 1.09
21 150 235 88 15
1.17 ⫾ 0.35 1.07 ⫾ 0.47 1.07 ⫾ 0.30 1.08 ⫾ 0.30 1.17 ⫾ 0.34
1.09 0.98 0.98 1.00 1.05
1.10 ⫾ 0.23 1.10 ⫾ 0.24 1.16 ⫾ 0.20 1.15 ⫾ 0.22 1.24 ⫾ 0.34
1.02 1.06 1.11 1.13 1.07
81 148 280
1.42 ⫾ 0.64 1.08 ⫾ 0.24 0.98 ⫾ 0.21
1.26 1.03 0.94
1.25 ⫾ 0.37 1.14 ⫾ 0.19 1.11 ⫾ 0.17
1.14 1.12 1.07
NOTE. To convert Cr in mg/dL to mol/L, multiply by 88.4.
CYSTATIN C IN THE AGED Table 3.
39
Serum Cys C and Serum Cr Values According to Health Status Variables in Women
Variable
No.
Hypertension Yes No Coronary heart disease Yes No Diabetes Yes No Urinary infection Yes No Rheumatoid arthritis Yes No Use of glucocorticoids Yes No BMI (kg/m2) ⬍20 20–24.9 25–29.9 30–34.9 ⱖ35 Functional index ⬍23 23–26 27
Cys C (mg/L)
Cr (mg/dL)
Mean ⫾ SD
Median
Mean ⫾ SD
Median
275 427
1.18 ⫾ 0.42 1.02 ⫾ 0.29
1.07 0.96
1.05 ⫾ 0.24 0.98 ⫾ 0.19
1.00 0.95
279 423
1.15 ⫾ 0.38 1.04 ⫾ 0.32
1.05 0.96
1.03 ⫾ 0.24 0.99 ⫾ 0.19
0.98 0.96
99 603
1.22 ⫾ 0.49 1.06 ⫾ 0.32
1.06 0.99
1.08 ⫾ 0.32 1.00 ⫾ 0.18
1.01 0.96
200 502
1.15 ⫾ 0.44 1.06 ⫾ 0.31
1.02 0.99
1.03 ⫾ 0.25 1.00 ⫾ 0.19
1.00 0.96
27 675
1.05 ⫾ 0.20 1.08 ⫾ 0.36
1.00 1.00
0.96 ⫾ 0.16 1.01 ⫾ 0.21
0.94 0.97
28 674
1.25 ⫾ 0.46 1.08 ⫾ 0.35
1.08 1.00
1.04 ⫾ 0.29 1.01 ⫾ 0.21
0.96 0.97
52 181 286 134 49
1.15 ⫾ 0.38 1.04 ⫾ 0.35 1.07 ⫾ 0.34 1.11 ⫾ 0.39 1.14 ⫾ 0.30
1.06 0.96 1.00 1.04 1.04
0.94 ⫾ 0.22 1.00 ⫾ 0.22 1.01 ⫾ 0.20 1.02 ⫾ 0.19 1.03 ⫾ 0.22
0.89 0.96 0.97 0.98 1.01
157 215 330
1.34 ⫾ 0.48 1.06 ⫾ 0.29 0.97 ⫾ 0.24
1.25 1.00 0.93
1.07 ⫾ 0.31 1.01 ⫾ 0.20 0.98 ⫾ 0.13
1.02 0.97 0.96
NOTE. To convert Cr in mg/dL to mol/L, multiply by 88.4.
DISCUSSION
In healthy elderly subjects, Cys C levels increase with age, and age range–specific reference intervals have been proposed, whereas no difference between sexes has been found.7 The dependence of Cr level on muscle mass is the most important factor compromising the utility of Cr measurements in geriatric care. In this study, we used a multivariate sex-specific approach to control for the impact of age and sex and identify the clinically important independent determinants of GFR variation and microalbuminuria in a polymorbid elderly population. Information on lean body mass was lacking. Age was by far the most powerful predictor of variation in Cys C values, but age also was significantly associated with Cr and ACR levels in both sexes. Sex was the most powerful predictor of variation in Cr values, whereas the contribution of sex to the variation in Cys C and ACR
values was marginal. Urinary ACR was assessed from a single early-morning urine specimen that captured both persistent and transient microalbuminuria, and results must be interpreted accordingly. Diagnoses were based mainly on earlier extensively available documentation. In addition to underlying diseases, pharmacological treatment in itself may have an impact on renal function or laboratory outcomes. However, the multiple antihypertensive, cardiac, and anti-inflammatory medications documented in this study were not included in the analyses as independent factors, and their impact hence cannot be evaluated per se. Glucocorticoid treatment and associated diagnoses were easily defined and subsequently included. Hypertension was a determinant of elevated renal values in both sexes, and coronary heart disease was found to be significantly influential
WASE´ N ET AL
40 Table 4.
Urinary ACR Values According to Health Status Variables by Sex
Variable
Hypertension Yes No Coronary heart disease Yes No Diabetes Yes No Urinary infection Yes No Rheumatoid arthritis Yes No Use of glucocorticoids Yes No BMI (kg/m2) ⬍20 20–24.9 25–29.9 30–34.9 ⱖ35 Functional index ⬍23 23–26 27
Men
Women
No.
Mean ⫾ SD
Median
No.
Mean ⫾ SD
Median
154 304
25.6 ⫾ 34.8 18.1 ⫾ 33.6
13.3 10.6
261 407
24.3 ⫾ 47.9 22.7 ⫾ 63.7
9.7 10.6
182 276
21.4 ⫾ 28.9 20.1 ⫾ 37.3
12.4 10.6
260 408
24.2 ⫾ 46.5 22.7 ⫾ 64.4
12.4 8.8
69 389
31.0 ⫾ 42.7 18.8 ⫾ 32.1
16.8 10.6
93 575
37.0 ⫾ 69.6 21.1 ⫾ 55.7
15.0 9.7
16 442
52.0 ⫾ 70.4 19.5 ⫾ 31.7
30.5 10.6
181 487
34.5 ⫾ 86.0 19.1 ⫾ 42.7
14.1 9.7
19 439
27.7 ⫾ 65.5 20.3 ⫾ 32.2
8.0 11.5
24 644
18.8 ⫾ 24.5 23.5 ⫾ 59.0
8.4 10.2
19 439
19.1 ⫾ 14.5 20.7 ⫾ 34.8
15.0 10.6
27 641
37.6 ⫾ 70.6 22.7 ⫾ 57.4
15.0 9.7
19 137 214 75 13
36.9 ⫾ 53.1 18.5 ⫾ 36.2 20.4 ⫾ 30.4 20.4 ⫾ 35.8 23.9 ⫾ 23.0
14.1 10.6 11.1 11.5 10.6
43 172 278 127 48
47.1 ⫾ 99.1 22.0 ⫾ 36.5 23.0 ⫾ 70.9 17.9 ⫾ 22.9 23.0 ⫾ 50.3
23.0 11.5 8.8 9.7 9.7
66 135 257
33.2 ⫾ 49.2 25.2 ⫾ 46.3 14.9 ⫾ 16.2
15.9 11.5 9.7
137 210 321
47.9 ⫾ 86.5 17.5 ⫾ 17.8 16.6 ⫾ 57.5
18.6 10.6 8.8
NOTE. ACR values expressed as mg/g. To convert urinary ACR in mg/g to mg/mmol, divide by 8.84.
in men, with Cys C and Cr analyses showing similar results. Atherosclerotic vascular disease, including hypertension, is a known risk factor for renal impairment in the elderly.3,18 Previous findings also have shown an elevated Cr level to be associated with hypertension and history of heart attack in both sexes in an aged population.11 In both Cys C and Cr analysis, the effect of urinary infection was significant only among men. Because of anatomy, urinary infection in men signals underlying urological and renal problems more distinctly than in women. Even if participants were instructed to obtain midstream urine samples, the possibility of misleading results because of sampling errors cannot completely be ruled out, particularly in women. Oral glucocorticoid treatment was significantly associated with greater Cys C level in both sexes, but in Cr analyses, no significant effect was found. Because chronic lung disease and
rheumatoid arthritis were the 2 main diagnoses documented in conjunction with glucocorticoid therapy, they also were tested as explanatory variables. Chronic lung disease showed no significant effect, whereas rheumatoid arthritis was significantly associated with greater Cys C values in men. Patients with prolonged rheumatoid arthritis previously have been reported to show elevated Cys C levels.19 Greater Cys C values have been found in asthmatic patients after the use of methylprednisolone,20 and in adult renal transplant recipients, glucocorticoid medication has been associated with elevated Cys C values.21 A promoter-mediated dexamethasoneinduced increase in Cys C production by HeLa cells has been described.22 Our results support previous results recognizing the use of glucocorticoids as an independent factor increasing Cys C levels. Diabetes defined by medical history or glyce-
CYSTATIN C IN THE AGED Table 5.
41
Associations of Demographic and Health Status Variables With Serum Cys C and Cr Values in Multivariate Linear Models in Men and Women
Variable
Men
Women
Cys C*
Hypertension Coronary heart disease Diabetes Urinary infection Rheumatoid arthritis Use of glucocorticoids Age BMI Functional index
Cr*
Cys C*
Cr*
F Statistic
P
F Statistic
P
F Statistic
P
F Statistic
P
17.73 4.54 2.94 7.50 5.15 5.52 66.51 1.52 9.50
⬍0.001 0.034 NS (0.09) 0.006 0.024 0.019 ⬍0.001 NS (0.20) ⬍0.001
16.09 7.44 0.71 9.61 1.07 0.82 12.21 3.63 0.12
⬍0.001 0.007 NS (0.40) 0.002 NS (0.30) NS (0.37) ⬍0.001 0.006 NS (0.88)
28.99 0.41 0.31 0.01 0.00 5.33 110.45 7.97 10.36
⬍0.001 NS (0.52) NS (0.58) NS (0.94) NS (0.99) 0.021 ⬍0.001 ⬍0.001 ⬍0.001
12.80 1.21 2.22 1.74 2.64 0.15 17.63 8.33 0.42
⬍0.001 NS (0.27) NS (0.14) NS (0.19) NS (0.11) NS (0.70) ⬍0.001 ⬍0.001 NS (0.66)
NOTE. Men, N ⫽ 509; women, N ⫽ 702. Abbreviation: NS, not significant. *Cys C and Cr were inversely transformed for statistical analysis.
mia with glucose of 126 mg/dL or greater (ⱖ7 mmol/L) was not associated with Cys C or Cr level. When the definition was narrowed down to include only pharmaceutically treated diabetes, significance still was not reached. The association between diabetes and ACR level was highly significant in both sexes. Diabetes is associated with a high risk for nephropathy.23 In an aged population, a significant relation between diabetes and elevated Cr levels has been reported in women.11 The result of the present cross-sectional study may be explained by selective cardiovascular mortality among persons with diabetes with advanced renal insufficiency.24,25 Because information on glycemic control and duration of Table 6.
the disease is lacking, definite conclusions cannot be drawn. BMI was associated with Cr level in both sexes and Cys C level in women. Mean Cr values showed a gradual increase with increasing BMI, probably reflecting growing muscle mass. Mean Cys C values were greatest for BMIs less than 20 and 35 or greater, presumably signaling the greater general morbidity associated with extreme body weights.26 Without a more exact measurement of muscle mass, all interpretations must be performed with caution. The functional index was constructed to reflect a summation of somatic illness and physical aging processes, rather than dependency in daily
Associations of Demographic and Health Status Variables With Urinary ACR in Multivariate Linear Models in Men and Women
Variable
Hypertension Coronary heart disease Diabetes Urinary infection Rheumatoid arthritis Use of glucocorticoids Age BMI Functional index
Men
Women
F Statistic
P
F Statistic
P
15.31 0.21 13.83 3.26 0.57 0.42 19.82 1.29 1.70
⬍0.001 NS (0.65) ⬍0.001 NS (0.07) NS (0.45) NS (0.52) ⬍0.001 NS (0.27) NS (0.18)
0.02 1.33 11.77 14.70 0.99 3.39 4.09 2.69 10.96
NS (0.90) NS (0.25) ⬍0.001 ⬍0.001 NS (0.32) NS (0.07) 0.044 0.031 ⬍0.001
NOTE. Urinary ACR was log transformed for statistical analysis. Men, N ⫽ 458; women, N ⫽ 668. Abbreviation: NS, not significant.
WASE´ N ET AL
42
living, and scoring was adjusted accordingly. Variation in Cys C values was significantly predicted by the functional index, whereas no association between Cr values and functional index was found. The greatest Cys C mean values were seen in individuals characterized by limited functional status. The diminished muscle mass associated with functional limitations might explain the smaller increase in Cr values, whereas Cys C, as a more sensitive marker of glomerular function, presumably reflects nephropathy associated with undefined concomitant disease. Salive et al11 reported that persons with one or more activity of daily living limitations had slightly greater mean Cr values, but found no gradient with the number of limitations. Finney et al27 found greater Cys C values in institutionalized versus free-living elderly subjects, but no significant difference in Cr values. Variables included in multivariate models explained a considerably greater proportion of variation in Cys C compared with Cr values. This could be explained because factors not specified in this study, ie, muscle mass, dietary intake, and drug interaction, might influence the production and laboratory analysis of Cr and cause variation not explained by changes in GFR. From this point of view, CysC seems to be a more useful tool than Cr in evaluating glomerular function in a geriatric setting. In addition to known risk factors, such as atherosclerotic disease and diabetes, older age, functional limitations, and altered body composition should be recognized as possible signs of renal impairment in elderly patients. ACKNOWLEDGMENT The authors thank Dade Behring, Marburg, Germany, for providing kits for the measurement of Cys C.
REFERENCES 1. Ruggenenti P, Schieppati A, Remuzzi G: Progression, remission, regression of chronic renal diseases. Lancet 357: 1601-1608, 2001 2. Berthoux F, Gellert R, Jones E, et al: Epidemiology and demography of treated end-stage renal failure in the elderly: From the European Renal Association (ERAEDTA) Registry. Nephrol Dial Transplant 13:S65-S68, 1998 (suppl 7) 3. 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 51:1196-1204, 1997 4. Parving HH, Hovind P, Rossing K, Andersen S: Evolv-
ing strategies for renoprotection: Diabetic nephropathy. Curr Opin Nephrol Hypertens 10:515-522, 2001 5. Perrone RD, Madias NE, Levey AS: Serum creatinine as an index of renal function: New insights into old concepts. Clin Chem 38:1933-1953, 1992 6. 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 40:221-226, 2002 7. Wasen E, Suominen P, Isoaho R, et al: Serum cystatin C as a marker of kidney dysfunction in an elderly population. Clin Chem 48:1138-1140, 2002 8. Van Den Noortgate NJ, Janssens WH, Delanghe JR, Afschrift MB, Lameire NH: Serum cystatin C concentration compared with other markers of glomerular filtration rate in the old old. J Am Geriatr Soc 50:1278-1282, 2002 9. Burkhardt H, Bojarsky G, Gladisch R: Diagnostic efficiency of cystatin C and serum creatinine as markers of reduced glomerular filtration rate in the elderly. Clin Chem Lab Med 40:1135-1138, 2002 10. Viberti GC, Hill RD, Jarrett RJ, Argyropoulos A, Mahmud U, Keen H: Microalbuminuria as a predictor of clinical nephropathy in insulin-dependent diabetes mellitus. Lancet 1:1430-1432, 1982 11. Salive ME, Jones CA, Guralnik JM, Agodoa LY, Pahor M, Wallace RB: Serum creatinine levels in older adults: Relationship with health status and medications. Age Ageing 24:142-150, 1995 12. Jones CA, Francis ME, Eberhardt MS, et al: Microalbuminuria in the US population: Third National Health and Nutrition Examination Survey. Am J Kidney Dis 39:445459, 2002 13. Finney H, Newman DJ, Gruber W, Merle P, Price CP: Initial evaluation of cystatin C measurement by particleenhanced immunonephelometry on the Behring nephelometer systems (BNA, BN II). Clin Chem 43:1016-1022, 1997 14. Rose GA, Blackburn H, Gillum RF, Prineas RJ: Cardiovascular Survey Methods. Geneva, Switzerland, World Health Organization Monograph Series no. 56 (ed 2), 1982 15. Alberti KGMM, Zimmer PZ: Definition, diagnosis and classification of diabetes mellitus and its complications: Part I. Diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 15:539-553, 1998 16. Heikkinen E, Waters WE, Brzezinski ZJ (eds): The elderly in eleven countries. A sociomedical survey, in Public Health in Europe 21. Copenhagen, Denmark, World Health Organization, 1983, pp 161-231 17. World Health Organization: Obesity: Preventing and Managing the Global Epidemic. Report of a Who Consultation presented at the World Health Organization, June 3-5, 1997, Geneva, Switzerland, Publication WHO/NUT/NCD/ 98.1 18. Kasiske BL: Relationship between vascular disease and age-associated changes in the human kidney. Kidney Int 31:1153-1159, 1987 19. Mangge H, Liebmann P, Tanil H, et al: Cystatin C, an early indicator for incipient renal disease in rheumatoid arthritis. Clin Chim Acta 300:195-202, 2000 20. Cimerman N, Brguljan PM, Krasovec M, Suskovic S, Kos J: Serum cystatin C, a potent inhibitor of cysteine
CYSTATIN C IN THE AGED
proteinases, is elevated in asthmatic patients. Clin Chim Acta 300:83-95, 2000 21. Risch L, Herklotz R, Blumberg A, Huber AR: Effects of glucocorticoid immunosuppression on serum cystatin C concentrations in renal transplant patients. Clin Chem 47: 2055-2059, 2001 22. Bjarnadottir M, Grubb A, Olafsson I: Promotermediated, dexamethasone-induced increase in cystatin C production by HeLa cells. Scand J Clin Lab Invest 55:617623, 1995 23. Selby JV, FitzSimmons SC, Newman JM, Katz PP, Sepe S, Showstack J: The natural history and epidemiology of diabetic nephropathy: Implications for prevention and control. JAMA 263:1954-1960, 1990 24. Shlipak MG, Heidenreich PA, Noguchi H, Chertow
43
GM, Browner WS, McClellan MB: Association of renal insufficiency with treatment and outcomes after myocardial infarction in elderly patients. Ann Intern Med 137:555-562, 2002 25. Haffner SM, Lehto S, Ronnemaa T, Pyorala K, Laakso M: Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med 339:229-234, 1998 26. Harris T, Cook EF, Garrison R, Higgins M, Kannel W, Goldman L: Body mass index and mortality among nonsmoking older persons: The Framingham Heart Study. JAMA 259:1520-1524, 1988 27. Finney H, Bates CJ, Price CP: Plasma cystatin C determinations in a healthy elderly population. Arch Gerontol Geriatr 29:75-94, 1999
T
HE INCIDENCE OF end-stage renal disease has increased in recent years worldwide, and a growing proportion of patients with chronic kidney diseases are elderly.1,2 Involutional changes render the aging kidney susceptible to impairment by disease and nephrotoxic agents and make timely identification and monitoring of renal dysfunction crucial in geriatric care.3 Because accumulating evidence indicates that renoprotective measures might delay the progression of renal failure, it also is important to identify determinants of early renal dysfunction.4 From the Institute of Clinical Medicine, General Practice; Departments of Biostatistics; and Clinical Chemistry, University of Turku; Central Laboratory, Turku University Central Hospital; Mehila¨inen Corporation, Turku; Salo Health Center, Salo; Ha¨rka¨tie Health Center, Lieto; and Satakunta Central Hospital, Pori, Finland. Received December 30, 2002; accepted in revised form February 28, 2003. 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. Presented in abstract form at the World Congress of Nephrology, Berlin, Germany, June 8-12, 2003. Address reprint requests to Elise Wase´n, MD, Institute of Clinical Medicine, General Practice, University of Turku, Lemminka¨isenkatu 1, FIN-20014, Turku, Finland. E-mail: [email protected] © 2003 by the National Kidney Foundation, Inc. 0272-6386/03/4201-0004$30.00/0 doi:10.1016/S0272-6386(03)00406-2 36
Serum creatinine (Cr) level is used routinely in assessing glomerular filtration rate (GFR), although several widely recognized limitations compromise its usefulness in clinical work. The impact of muscle mass on Cr values makes interpretation difficult, especially in elderly individuals.5 Serum cystatin C (Cys C) is believed to be a more sensitive indicator of GFR in the general population and specific patient categories.6 In our earlier study, we found Cys C to be more reliable than Cr in assessing glomerular function in the healthy elderly.7 In acutely hospitalized elderly patients, Cys C was not found to be superior to Cr level in the detection of renal impairment8,9; however, small sample size and possible interference by medical treatment and pseudochromogens produced in acute illness may have influenced the results. Microalbuminuria has been shown to indicate incipient nephropathy in patients with diabetes and other patients with chronic vascular disease.10 Determinants of elevated Cr levels and microalbuminuria have been reported in populationbased studies including elderly subjects,11,12 whereas Cys C has not been evaluated extensively in this respect in representative elderly populations. In this study, we assessed the associations of demographic and health status characteristics with levels of Cys C and Cr and urinary
American Journal of Kidney Diseases, Vol 42, No 1 (July), 2003: pp 36-43
CYSTATIN C IN THE AGED
albumin-creatinine ratio (ACR) by means of multivariate analyses in an elderly population. The main objective is to evaluate features of CysC versus Cr level in this context. METHODS
Study Population and Data Collection The study population consisted of 1,260 elderly residents (533 men, 727 women; mean age, 74 years; range, 64 to 100 years) in Lieto, a semi-industrialized rural municipality in southwestern Finland. During 1998 to 1999, all Lieto residents born in or before 1933 (n ⫽ 1,519) were invited to participate in a community-based, cross-sectional epidemiological survey; 83% of those eligible gave a written consent and were included. Information on chronic health conditions, current medications, and physical functioning was obtained through standardized interviews, thorough clinical examination, and review of medical records. International Classification of Diseases, Tenth Revision, diagnosis codes were used for documentation. The Joint Commission of Ethics for the Hospital District of Varsinais-Suomi in southwestern Finland approved the study protocol.
Measurements and Definitions Fasting venous blood samples and early-morning spot urine samples were collected and then analyzed in the Central Laboratory of Turku University Central Hospital. Before the laboratory visit, all participants were given verbal and written instructions. 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).13 Cr was measured using the Jaffe´ reaction (Roche Diagnostics, Mannheim, Germany, and Hitachi 917; Hitachi Ltd, Tokyo, Japan). Urinary albumin was analyzed using an immunoturbidimetric method (Optima, Microalbuminuria Kit, Thermo Clinical Labsystem, Helsinki, Finland), and urinary creatinine concentrations were determined from the same sample using the Jaffe´ method. Clinical proteinuria was assessed by means of a urine dipstick test. A 12-lead resting electrocardiogram was recorded and interpreted using the Minnesota Code.14 Hypertension was categorized as present if a diagnosis was documented in the medical records. Diabetes was defined by medical history or fasting glycemia with glucose of 126 mg/dL or greater (ⱖ7 mmol/L), acknowledging that type 2 diabetes may be present years before clinical recognition.15 Coronary heart disease was diagnosed based on medical records or electrocardiogram findings. Diagnosis of urinary infection was based on significant bacterial growth in an early-morning urine specimen or documentation of chronic urinary infection. 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; ie, dressing, eating, bathing, going to bed, and using the toilet. Each had scores of 0 to 3 (0, unable to do; 1, some help needed; 2, with difficulty, but no help needed; 3, no limitations).16 According to the sum score, the functional index
37
was categorized as less than 23, 23 to 26, or 27 (no limitations). Body mass index (BMI), measured as kilograms per square meter, was classified as less than 20, 20 to 24.9, 25 to 29.9, 30 to 34.9, and 35 or greater according to the World Health Organization recommendation, except for the lowest cutoff value, which was placed at 20 instead of 18.5, to ensure sufficient numbers in the lowest class.17
Design and Analysis All statistical analyses were performed using SAS System for Windows, version 8.02 (SAS Institute Inc, Cary, NC). Because of skewed distributions, Cys C and Cr values were inversely transformed and ACR values were log transformed for statistical analysis. Associations of independent variables with the dependent variables Cys C, Cr, and ACR were evaluated using univariate and multivariate linear models. Explanatory variables were selected to multivariate analyses if they reached statistical significance in univariate analysis, but were excluded if they did not remain significant in at least one of the models. The final list of explanatory variables included in all multivariate models consisted of age (continuous variable), hypertension, coronary heart disease, diabetes, urinary infection, rheumatoid arthritis, and oral glucocorticoid treatment (2 categories), as well as BMI (5 categories) and functional index (3 categories). All analyses were stratified by sex. Correlations were calculated using nonparametric Spearman’s rank-order correlation coefficients. P less than 0.05 is considered significant. Missing data and exclusion of subjects with clinical proteinuria (n ⫽ 68) reduced the final study sample to 509 men and 702 women in the Cys C and Cr analyses and 458 men and 668 women in the ACR analyses.
RESULTS
Some characteristics of the population and mean ⫾ SD and median values of Cys C, Cr, and ACR according to the explanatory variables are listed in Tables 1 through 4. Age correlated with Cys C level (men, r ⫽ 0.444, P ⬍ 0.001; women, r ⫽ 0.465, P ⬍ 0.001), Cr level (men, r ⫽ 0.157, P ⬍ 0.001; women, r ⫽ 0.164, P ⬍ 0.001), and ACR (men, r ⫽ 0.253, P ⬍ 0.001; women, r ⫽ 0.263, P ⬍ 0.001). In the multivariate linear model, hypertension, coronary heart disease, urinary infection, rheumatoid arthritis, glucocorticoid treatment, older age, and lower functional index were significantly associated with greater Cys C values in men (Table 5). Hypertension, coronary heart disease, urinary infection, advancing age, and increasing BMI were significantly related to greater Cr levels in men. Among women, corresponding factors were hypertension, glucocorticoid treatment, age, functional index, and BMI for Cys C and hypertension, BMI, and age for Cr.
WASE´ N ET AL
38 Table 1.
Age (y) BMI (kg/m2) Serum Cys C (mg/L) Serum Cr (mg/dL) Urinary ACR (mg/g)
Characteristics of the Population
Men (N ⫽ 533)
Women (N ⫽ 727)
All (N ⫽ 1,260)
73.2 ⫾ 6.4 26.8 ⫾ 3.9 1.09 ⫾ 0.30 1.15 ⫾ 0.23 33.8 ⫾ 83.2
74.6 ⫾ 7.1 27.2 ⫾ 5.2 1.09 ⫾ 0.35 1.01 ⫾ 0.21 26.6 ⫾ 64.6
74.0 ⫾ 6.8 27.0 ⫾ 4.7 1.09 ⫾ 0.37 1.07 ⫾ 0.23 29.7 ⫾ 73.2
NOTE. Values expressed as mean ⫾SD. To convert Cr in mg/dL to mol/L, multiply by 88.4; urinary ACR in mg/g to mg/mmol, divide by 8.84.
erably greater in multivariate analyses for Cys C compared with those for Cr (35% versus 14.8% in men and 34.5% versus 11.3% in women). In ACR analyses, corresponding percentages were 17.3% (men) and 17.8% (women). When sex was entered into the full model, R2 changed from 33.0% to 33.4% in Cys C analysis, from 8.7% to 22.9% in Cr analysis, and from 15.3% to 15.9% in ACR analysis.
Diabetes was significantly associated with greater levels of microalbuminuria, measured by ACR, in both sexes (Table 6), but neither was associated with Cys C or Cr level. In men, other variables significantly associated with higher levels of ACR were hypertension and age, whereas in women, urinary infection, age, functional index, and BMI reached significance. Coefficients of determination (R2) were considTable 2. Variable
Hypertension Yes No Coronary heart disease Yes No Diabetes Yes No Urinary infection Yes No Rheumatoid arthritis Yes No Use of glucocorticoids Yes No BMI (kg/m2) ⬍20 20–24.9 25–29.9 30–34.9 ⱖ35 Functional index ⬍23 23–26 27
Serum Cys C and Serum Cr Values According to Health Status Variables in Men No.
Cys C (mg/L)
Cr (mg/dL)
Mean ⫾ SD
Median
Mean ⫾ SD
Median
173 336
1.13 ⫾ 0.39 1.05 ⫾ 0.34
1.03 0.97
1.20 ⫾ 0.26 1.11 ⫾ 0.20
1.15 1.07
212 297
1.15 ⫾ 0.45 1.02 ⫾ 0.27
1.04 0.97
1.18 ⫾ 0.26 1.11 ⫾ 0.19
1.13 1.07
84 425
1.09 ⫾ 0.34 1.07 ⫾ 0.37
0.97 0.99
1.17 ⫾ 0.24 1.14 ⫾ 0.22
1.12 1.09
27 482
1.47 ⫾ 0.50 1.05 ⫾ 0.34
1.37 0.98
1.34 ⫾ 0.33 1.13 ⫾ 0.21
1.31 1.09
22 487
1.28 ⫾ 0.52 1.07 ⫾ 0.35
1.15 0.99
1.20 ⫾ 0.25 1.14 ⫾ 0.22
1.16 1.09
21 488
1.37 ⫾ 0.78 1.06 ⫾ 0.33
1.15 0.99
1.18 ⫾ 0.26 1.14 ⫾ 0.22
1.12 1.09
21 150 235 88 15
1.17 ⫾ 0.35 1.07 ⫾ 0.47 1.07 ⫾ 0.30 1.08 ⫾ 0.30 1.17 ⫾ 0.34
1.09 0.98 0.98 1.00 1.05
1.10 ⫾ 0.23 1.10 ⫾ 0.24 1.16 ⫾ 0.20 1.15 ⫾ 0.22 1.24 ⫾ 0.34
1.02 1.06 1.11 1.13 1.07
81 148 280
1.42 ⫾ 0.64 1.08 ⫾ 0.24 0.98 ⫾ 0.21
1.26 1.03 0.94
1.25 ⫾ 0.37 1.14 ⫾ 0.19 1.11 ⫾ 0.17
1.14 1.12 1.07
NOTE. To convert Cr in mg/dL to mol/L, multiply by 88.4.
CYSTATIN C IN THE AGED Table 3.
39
Serum Cys C and Serum Cr Values According to Health Status Variables in Women
Variable
No.
Hypertension Yes No Coronary heart disease Yes No Diabetes Yes No Urinary infection Yes No Rheumatoid arthritis Yes No Use of glucocorticoids Yes No BMI (kg/m2) ⬍20 20–24.9 25–29.9 30–34.9 ⱖ35 Functional index ⬍23 23–26 27
Cys C (mg/L)
Cr (mg/dL)
Mean ⫾ SD
Median
Mean ⫾ SD
Median
275 427
1.18 ⫾ 0.42 1.02 ⫾ 0.29
1.07 0.96
1.05 ⫾ 0.24 0.98 ⫾ 0.19
1.00 0.95
279 423
1.15 ⫾ 0.38 1.04 ⫾ 0.32
1.05 0.96
1.03 ⫾ 0.24 0.99 ⫾ 0.19
0.98 0.96
99 603
1.22 ⫾ 0.49 1.06 ⫾ 0.32
1.06 0.99
1.08 ⫾ 0.32 1.00 ⫾ 0.18
1.01 0.96
200 502
1.15 ⫾ 0.44 1.06 ⫾ 0.31
1.02 0.99
1.03 ⫾ 0.25 1.00 ⫾ 0.19
1.00 0.96
27 675
1.05 ⫾ 0.20 1.08 ⫾ 0.36
1.00 1.00
0.96 ⫾ 0.16 1.01 ⫾ 0.21
0.94 0.97
28 674
1.25 ⫾ 0.46 1.08 ⫾ 0.35
1.08 1.00
1.04 ⫾ 0.29 1.01 ⫾ 0.21
0.96 0.97
52 181 286 134 49
1.15 ⫾ 0.38 1.04 ⫾ 0.35 1.07 ⫾ 0.34 1.11 ⫾ 0.39 1.14 ⫾ 0.30
1.06 0.96 1.00 1.04 1.04
0.94 ⫾ 0.22 1.00 ⫾ 0.22 1.01 ⫾ 0.20 1.02 ⫾ 0.19 1.03 ⫾ 0.22
0.89 0.96 0.97 0.98 1.01
157 215 330
1.34 ⫾ 0.48 1.06 ⫾ 0.29 0.97 ⫾ 0.24
1.25 1.00 0.93
1.07 ⫾ 0.31 1.01 ⫾ 0.20 0.98 ⫾ 0.13
1.02 0.97 0.96
NOTE. To convert Cr in mg/dL to mol/L, multiply by 88.4.
DISCUSSION
In healthy elderly subjects, Cys C levels increase with age, and age range–specific reference intervals have been proposed, whereas no difference between sexes has been found.7 The dependence of Cr level on muscle mass is the most important factor compromising the utility of Cr measurements in geriatric care. In this study, we used a multivariate sex-specific approach to control for the impact of age and sex and identify the clinically important independent determinants of GFR variation and microalbuminuria in a polymorbid elderly population. Information on lean body mass was lacking. Age was by far the most powerful predictor of variation in Cys C values, but age also was significantly associated with Cr and ACR levels in both sexes. Sex was the most powerful predictor of variation in Cr values, whereas the contribution of sex to the variation in Cys C and ACR
values was marginal. Urinary ACR was assessed from a single early-morning urine specimen that captured both persistent and transient microalbuminuria, and results must be interpreted accordingly. Diagnoses were based mainly on earlier extensively available documentation. In addition to underlying diseases, pharmacological treatment in itself may have an impact on renal function or laboratory outcomes. However, the multiple antihypertensive, cardiac, and anti-inflammatory medications documented in this study were not included in the analyses as independent factors, and their impact hence cannot be evaluated per se. Glucocorticoid treatment and associated diagnoses were easily defined and subsequently included. Hypertension was a determinant of elevated renal values in both sexes, and coronary heart disease was found to be significantly influential
WASE´ N ET AL
40 Table 4.
Urinary ACR Values According to Health Status Variables by Sex
Variable
Hypertension Yes No Coronary heart disease Yes No Diabetes Yes No Urinary infection Yes No Rheumatoid arthritis Yes No Use of glucocorticoids Yes No BMI (kg/m2) ⬍20 20–24.9 25–29.9 30–34.9 ⱖ35 Functional index ⬍23 23–26 27
Men
Women
No.
Mean ⫾ SD
Median
No.
Mean ⫾ SD
Median
154 304
25.6 ⫾ 34.8 18.1 ⫾ 33.6
13.3 10.6
261 407
24.3 ⫾ 47.9 22.7 ⫾ 63.7
9.7 10.6
182 276
21.4 ⫾ 28.9 20.1 ⫾ 37.3
12.4 10.6
260 408
24.2 ⫾ 46.5 22.7 ⫾ 64.4
12.4 8.8
69 389
31.0 ⫾ 42.7 18.8 ⫾ 32.1
16.8 10.6
93 575
37.0 ⫾ 69.6 21.1 ⫾ 55.7
15.0 9.7
16 442
52.0 ⫾ 70.4 19.5 ⫾ 31.7
30.5 10.6
181 487
34.5 ⫾ 86.0 19.1 ⫾ 42.7
14.1 9.7
19 439
27.7 ⫾ 65.5 20.3 ⫾ 32.2
8.0 11.5
24 644
18.8 ⫾ 24.5 23.5 ⫾ 59.0
8.4 10.2
19 439
19.1 ⫾ 14.5 20.7 ⫾ 34.8
15.0 10.6
27 641
37.6 ⫾ 70.6 22.7 ⫾ 57.4
15.0 9.7
19 137 214 75 13
36.9 ⫾ 53.1 18.5 ⫾ 36.2 20.4 ⫾ 30.4 20.4 ⫾ 35.8 23.9 ⫾ 23.0
14.1 10.6 11.1 11.5 10.6
43 172 278 127 48
47.1 ⫾ 99.1 22.0 ⫾ 36.5 23.0 ⫾ 70.9 17.9 ⫾ 22.9 23.0 ⫾ 50.3
23.0 11.5 8.8 9.7 9.7
66 135 257
33.2 ⫾ 49.2 25.2 ⫾ 46.3 14.9 ⫾ 16.2
15.9 11.5 9.7
137 210 321
47.9 ⫾ 86.5 17.5 ⫾ 17.8 16.6 ⫾ 57.5
18.6 10.6 8.8
NOTE. ACR values expressed as mg/g. To convert urinary ACR in mg/g to mg/mmol, divide by 8.84.
in men, with Cys C and Cr analyses showing similar results. Atherosclerotic vascular disease, including hypertension, is a known risk factor for renal impairment in the elderly.3,18 Previous findings also have shown an elevated Cr level to be associated with hypertension and history of heart attack in both sexes in an aged population.11 In both Cys C and Cr analysis, the effect of urinary infection was significant only among men. Because of anatomy, urinary infection in men signals underlying urological and renal problems more distinctly than in women. Even if participants were instructed to obtain midstream urine samples, the possibility of misleading results because of sampling errors cannot completely be ruled out, particularly in women. Oral glucocorticoid treatment was significantly associated with greater Cys C level in both sexes, but in Cr analyses, no significant effect was found. Because chronic lung disease and
rheumatoid arthritis were the 2 main diagnoses documented in conjunction with glucocorticoid therapy, they also were tested as explanatory variables. Chronic lung disease showed no significant effect, whereas rheumatoid arthritis was significantly associated with greater Cys C values in men. Patients with prolonged rheumatoid arthritis previously have been reported to show elevated Cys C levels.19 Greater Cys C values have been found in asthmatic patients after the use of methylprednisolone,20 and in adult renal transplant recipients, glucocorticoid medication has been associated with elevated Cys C values.21 A promoter-mediated dexamethasoneinduced increase in Cys C production by HeLa cells has been described.22 Our results support previous results recognizing the use of glucocorticoids as an independent factor increasing Cys C levels. Diabetes defined by medical history or glyce-
CYSTATIN C IN THE AGED Table 5.
41
Associations of Demographic and Health Status Variables With Serum Cys C and Cr Values in Multivariate Linear Models in Men and Women
Variable
Men
Women
Cys C*
Hypertension Coronary heart disease Diabetes Urinary infection Rheumatoid arthritis Use of glucocorticoids Age BMI Functional index
Cr*
Cys C*
Cr*
F Statistic
P
F Statistic
P
F Statistic
P
F Statistic
P
17.73 4.54 2.94 7.50 5.15 5.52 66.51 1.52 9.50
⬍0.001 0.034 NS (0.09) 0.006 0.024 0.019 ⬍0.001 NS (0.20) ⬍0.001
16.09 7.44 0.71 9.61 1.07 0.82 12.21 3.63 0.12
⬍0.001 0.007 NS (0.40) 0.002 NS (0.30) NS (0.37) ⬍0.001 0.006 NS (0.88)
28.99 0.41 0.31 0.01 0.00 5.33 110.45 7.97 10.36
⬍0.001 NS (0.52) NS (0.58) NS (0.94) NS (0.99) 0.021 ⬍0.001 ⬍0.001 ⬍0.001
12.80 1.21 2.22 1.74 2.64 0.15 17.63 8.33 0.42
⬍0.001 NS (0.27) NS (0.14) NS (0.19) NS (0.11) NS (0.70) ⬍0.001 ⬍0.001 NS (0.66)
NOTE. Men, N ⫽ 509; women, N ⫽ 702. Abbreviation: NS, not significant. *Cys C and Cr were inversely transformed for statistical analysis.
mia with glucose of 126 mg/dL or greater (ⱖ7 mmol/L) was not associated with Cys C or Cr level. When the definition was narrowed down to include only pharmaceutically treated diabetes, significance still was not reached. The association between diabetes and ACR level was highly significant in both sexes. Diabetes is associated with a high risk for nephropathy.23 In an aged population, a significant relation between diabetes and elevated Cr levels has been reported in women.11 The result of the present cross-sectional study may be explained by selective cardiovascular mortality among persons with diabetes with advanced renal insufficiency.24,25 Because information on glycemic control and duration of Table 6.
the disease is lacking, definite conclusions cannot be drawn. BMI was associated with Cr level in both sexes and Cys C level in women. Mean Cr values showed a gradual increase with increasing BMI, probably reflecting growing muscle mass. Mean Cys C values were greatest for BMIs less than 20 and 35 or greater, presumably signaling the greater general morbidity associated with extreme body weights.26 Without a more exact measurement of muscle mass, all interpretations must be performed with caution. The functional index was constructed to reflect a summation of somatic illness and physical aging processes, rather than dependency in daily
Associations of Demographic and Health Status Variables With Urinary ACR in Multivariate Linear Models in Men and Women
Variable
Hypertension Coronary heart disease Diabetes Urinary infection Rheumatoid arthritis Use of glucocorticoids Age BMI Functional index
Men
Women
F Statistic
P
F Statistic
P
15.31 0.21 13.83 3.26 0.57 0.42 19.82 1.29 1.70
⬍0.001 NS (0.65) ⬍0.001 NS (0.07) NS (0.45) NS (0.52) ⬍0.001 NS (0.27) NS (0.18)
0.02 1.33 11.77 14.70 0.99 3.39 4.09 2.69 10.96
NS (0.90) NS (0.25) ⬍0.001 ⬍0.001 NS (0.32) NS (0.07) 0.044 0.031 ⬍0.001
NOTE. Urinary ACR was log transformed for statistical analysis. Men, N ⫽ 458; women, N ⫽ 668. Abbreviation: NS, not significant.
WASE´ N ET AL
42
living, and scoring was adjusted accordingly. Variation in Cys C values was significantly predicted by the functional index, whereas no association between Cr values and functional index was found. The greatest Cys C mean values were seen in individuals characterized by limited functional status. The diminished muscle mass associated with functional limitations might explain the smaller increase in Cr values, whereas Cys C, as a more sensitive marker of glomerular function, presumably reflects nephropathy associated with undefined concomitant disease. Salive et al11 reported that persons with one or more activity of daily living limitations had slightly greater mean Cr values, but found no gradient with the number of limitations. Finney et al27 found greater Cys C values in institutionalized versus free-living elderly subjects, but no significant difference in Cr values. Variables included in multivariate models explained a considerably greater proportion of variation in Cys C compared with Cr values. This could be explained because factors not specified in this study, ie, muscle mass, dietary intake, and drug interaction, might influence the production and laboratory analysis of Cr and cause variation not explained by changes in GFR. From this point of view, CysC seems to be a more useful tool than Cr in evaluating glomerular function in a geriatric setting. In addition to known risk factors, such as atherosclerotic disease and diabetes, older age, functional limitations, and altered body composition should be recognized as possible signs of renal impairment in elderly patients. ACKNOWLEDGMENT The authors thank Dade Behring, Marburg, Germany, for providing kits for the measurement of Cys C.
REFERENCES 1. Ruggenenti P, Schieppati A, Remuzzi G: Progression, remission, regression of chronic renal diseases. Lancet 357: 1601-1608, 2001 2. Berthoux F, Gellert R, Jones E, et al: Epidemiology and demography of treated end-stage renal failure in the elderly: From the European Renal Association (ERAEDTA) Registry. Nephrol Dial Transplant 13:S65-S68, 1998 (suppl 7) 3. 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 51:1196-1204, 1997 4. Parving HH, Hovind P, Rossing K, Andersen S: Evolv-
ing strategies for renoprotection: Diabetic nephropathy. Curr Opin Nephrol Hypertens 10:515-522, 2001 5. Perrone RD, Madias NE, Levey AS: Serum creatinine as an index of renal function: New insights into old concepts. Clin Chem 38:1933-1953, 1992 6. 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 40:221-226, 2002 7. Wasen E, Suominen P, Isoaho R, et al: Serum cystatin C as a marker of kidney dysfunction in an elderly population. Clin Chem 48:1138-1140, 2002 8. Van Den Noortgate NJ, Janssens WH, Delanghe JR, Afschrift MB, Lameire NH: Serum cystatin C concentration compared with other markers of glomerular filtration rate in the old old. J Am Geriatr Soc 50:1278-1282, 2002 9. Burkhardt H, Bojarsky G, Gladisch R: Diagnostic efficiency of cystatin C and serum creatinine as markers of reduced glomerular filtration rate in the elderly. Clin Chem Lab Med 40:1135-1138, 2002 10. Viberti GC, Hill RD, Jarrett RJ, Argyropoulos A, Mahmud U, Keen H: Microalbuminuria as a predictor of clinical nephropathy in insulin-dependent diabetes mellitus. Lancet 1:1430-1432, 1982 11. Salive ME, Jones CA, Guralnik JM, Agodoa LY, Pahor M, Wallace RB: Serum creatinine levels in older adults: Relationship with health status and medications. Age Ageing 24:142-150, 1995 12. Jones CA, Francis ME, Eberhardt MS, et al: Microalbuminuria in the US population: Third National Health and Nutrition Examination Survey. Am J Kidney Dis 39:445459, 2002 13. Finney H, Newman DJ, Gruber W, Merle P, Price CP: Initial evaluation of cystatin C measurement by particleenhanced immunonephelometry on the Behring nephelometer systems (BNA, BN II). Clin Chem 43:1016-1022, 1997 14. Rose GA, Blackburn H, Gillum RF, Prineas RJ: Cardiovascular Survey Methods. Geneva, Switzerland, World Health Organization Monograph Series no. 56 (ed 2), 1982 15. Alberti KGMM, Zimmer PZ: Definition, diagnosis and classification of diabetes mellitus and its complications: Part I. Diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 15:539-553, 1998 16. Heikkinen E, Waters WE, Brzezinski ZJ (eds): The elderly in eleven countries. A sociomedical survey, in Public Health in Europe 21. Copenhagen, Denmark, World Health Organization, 1983, pp 161-231 17. World Health Organization: Obesity: Preventing and Managing the Global Epidemic. Report of a Who Consultation presented at the World Health Organization, June 3-5, 1997, Geneva, Switzerland, Publication WHO/NUT/NCD/ 98.1 18. Kasiske BL: Relationship between vascular disease and age-associated changes in the human kidney. Kidney Int 31:1153-1159, 1987 19. Mangge H, Liebmann P, Tanil H, et al: Cystatin C, an early indicator for incipient renal disease in rheumatoid arthritis. Clin Chim Acta 300:195-202, 2000 20. Cimerman N, Brguljan PM, Krasovec M, Suskovic S, Kos J: Serum cystatin C, a potent inhibitor of cysteine
CYSTATIN C IN THE AGED
proteinases, is elevated in asthmatic patients. Clin Chim Acta 300:83-95, 2000 21. Risch L, Herklotz R, Blumberg A, Huber AR: Effects of glucocorticoid immunosuppression on serum cystatin C concentrations in renal transplant patients. Clin Chem 47: 2055-2059, 2001 22. Bjarnadottir M, Grubb A, Olafsson I: Promotermediated, dexamethasone-induced increase in cystatin C production by HeLa cells. Scand J Clin Lab Invest 55:617623, 1995 23. Selby JV, FitzSimmons SC, Newman JM, Katz PP, Sepe S, Showstack J: The natural history and epidemiology of diabetic nephropathy: Implications for prevention and control. JAMA 263:1954-1960, 1990 24. Shlipak MG, Heidenreich PA, Noguchi H, Chertow
43
GM, Browner WS, McClellan MB: Association of renal insufficiency with treatment and outcomes after myocardial infarction in elderly patients. Ann Intern Med 137:555-562, 2002 25. Haffner SM, Lehto S, Ronnemaa T, Pyorala K, Laakso M: Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med 339:229-234, 1998 26. Harris T, Cook EF, Garrison R, Higgins M, Kannel W, Goldman L: Body mass index and mortality among nonsmoking older persons: The Framingham Heart Study. JAMA 259:1520-1524, 1988 27. Finney H, Bates CJ, Price CP: Plasma cystatin C determinations in a healthy elderly population. Arch Gerontol Geriatr 29:75-94, 1999