Original Article

Ann Clin Biochem 1992; 29: 523-528

Reference values of serum and urine creatinine, and of creatinine clearance by a new enzymatic method

a Sugita,

K Uchiyama, T Yamada, T Satol, M Okada and K Takeuchi-

From the Department of Laboratory Medicine, Niigata University School of Medicine and 'Section of Clinical Chemistry, Niigata Cancer Center Hospital and 2Research and Development Division of Denka Seiken Co., Ltd, Niigata, Japan

A new, totally enzymatic procedure for the determination of creatinine in serum and urine, using creatinine amidohydrolase, creatine amidinohydrolase, sarcosine oxidase and formaldehyde dehydrogenase is described. The assay was adapted to a discontinuous analyser with each analysis requiring only 20 J.lL of serum or 3 J.lL of urine. Analytical recovery of creatinine in serum and urine averaged 100· 61170. Within-run and between-run precision studies gave coefficients of variation of 1·1 % and I· 8%, respectively, for a serum with mean values of 83 J.lmollL (9' 4 mg/L) creatinine. Creatinine concentrations in serum and urine were measured by this procedure, in Japanese children and adults. The reference intervals for serum creatinine concentrations in adults 'were 55-96J.lmol/L (6'2-10'9mg/L) in men and 40-66J.lmollL (4'5-7'5mg/L) in women, and for urine, 9'46-19'0Immollday (1070-2150 mg/day) in men and 6,75-10,61 mmollday (764-1200 mg/day) in women. The reference intervals of creatinine clearance were 88· 0-176' 4 mL/min in men and 75'7-173'OmL/min in women.

SUMMARY.

Additional key phrases: creatine and creatinine amidohydrolases; interference with enzymatic method Numerous methods have been described for determining creatinine in serum and urine. The most widely used procedures, such as the Folin method, \ are based on the Jaffe alkaline picrate procedure.? which is not specific and is subject to interference by a large number of substances.v? Many investigators have attempted to improve the procedure by minimizing the effect of interfering substances present in the sample.! Consequently, alternative methods using high performance liquid chromatography-" or enzymatic assays have been investigated with a view to solving these problems.Pv'? In the present work, we shall describe a new, automated, totally enzymatic assay for creatinine in serum and urine, which offers high specificity and rapid throughput. We determined the reference values of creatinine in serum and urine, and of creatinine clearance, by this method.

Using our procedure, creatine is produced from creatinine by creatinine amidohydrolase (EC 3.5.2.10) and degraded to sarcosine and urea by creatine amidinohydrolase (BC 3.5.3.3). The sarcosine formed then reacts with sarcosine oxidase (BC 1.5.3.1), in the presence of formaldehyde dehydrogenase (BC 1.2.1.1) and NAD. The NADH + H+ produced is measured at 340 nm. The principle of this enzyme reaction has already been reported for determining creatine.P'!? The creatinine concentration can be calculated directly from the rate of increase in the absorbance of NADH generated in the reaction, or from creatinine standard solutions. The principle of these reactions may be summarized as follows: Creatinine + H 20

creatinine arnidohydrolase »0

Creatine

creatine amidinohydrolasc »0

Sarcosine + Urea

-

sarcosine oxidase

Correspondence: Dr 0 Sugita, Department of Laboratory Medicine, Niigata University School of Medicine, l-Asahirnachi Niigata 951, Japan.

Sarcosine + H 20 + 02

Glycine + H,O, + HCHO

formaldehyde

dehydrogenase

HCHO+NAD+ +H,O

~

HCOOH + NADH + H +

523

524

Sugita et al.

MATERIALS AND METHODS Reagents Two enzyme solutions were prepared for the creatinine assay. As reagent-L, 15 U/mL of sarcosine oxidase, O' 5 U/mL of formaldehyde dehydrogenase, 50 U/mL of creatine amidinohydrolase (these enzymes were obtained from Toyobo Co., Ltd, Osaka, Japan) and I .0 mmol/L of NAD (Boehringer Mannheim Yamanouchi, Tokyo, Japan) were dissolved in 100 mmol/L phosphate buffer (pH 7· 8) containing O' 5 mmol/L of EDTA, O· 111I0 of Triton X-100 and 10 mmol/L of Oxamic acid. Reagent-2 for serum and urinary creatinine measurements contained 130 U/mL, and 65 UI mL of creatinine amidohydrolase (Toyobo) in the buffer, respectively. Standard solutions were prepared by dissolving creatinine (Wako Pure Chemical Industries, Tokyo, Japan) 442/o1mol/L for serum and 8· 84 mmol/L for urine in 0·1 N HCI solution. Procedure The procedure was adapted for a Hitachi 736 autoanalyser (Hitachi, Tokyo, Japan). Samples of 20 iLL of serum or 3 /oiL of urine were transfered into the instrument and, 240/olL of reagent-l (for serum) or 300 iLL (for urine) were added. The mixture was incubated for 3 min at 37°C. Endogenous creatine in the sample was degraded by the reaction at this stage. Then 60 iLL (for serum) or 75 iLL (for urine) of reagent-2 was added. After 1· 6 min, the rate of increase in absorbance at 340 nm at 37°C was recorded for l : 6 min. The creatinine concentration was determined from a standard curve using 442/o1mol/L creatinine for serum, and results were printed 7· 5 min after sampling. In the manual version of this assay, sample and reagent volumes were 10 times higher than those employed in the automated method described above and the rate of absorbance at 340 nm at 37°C was recorded for 2·Omin. Comparison of the enzymatic method with other methods To assess the validity of our enzymatic technique, comparative studies were undertaken using the diagnostic kit 'creatinine Jaffe-rate' from Denka Seiken Co., Ltd (Niigata, Japan), an enzymatic Kodak single-slide method (EKTACHEM, USA)lS and an HPLC method described elsewhere."

Interference studies To examine whether interfering substances in the sample would affect results, potential interfering anticoagulants, metabolites and drugs were added to a serum pool containing a known amount of creatinine. Aliquots of the serum were then analysed and the results compared with those for the interferent-free pool. If the mean difference was greater than 2 SD of the mean for the untreated samples, the added compound was considered to be an interferent. Reference values of creatinine Our reference group included 1307 adult blood donors and hospital staff volunteers, aged between 20 and 65 years. In order to ensure the validity of our results, the individuals involved were screened, to exclude a history of renal, or related disease, drug abuse or other dependency, such as the daily consumption of more than 50 mL of alcohol. Subjects with abnormal values of urea, uric acid, glucose, bilirubin, AST (EC 2.6.1.1), ALT (EC 2.6.1.2), CK (EC 2.7.3.2) and 'Y-GT (EC 2.3.2.2) in serum were eliminated from the study. More than half these individuals were considered to be alcoholic. The sera of the remaining 235 men and 355 women were used for calculating reference values of serum creatinine in adults. In the same manner, 24 h urinary samples from 123 men and 91 women were used for the calculation of reference values of urinary creatinine in adults. Both the enzymatic method and the Jaffe-rate method were used for these calculations. The reference values of creatinine concentrations were calculated by a parametric method. The samples used for calculating reference values of serum creatinine in children were selected from 455 subjects who had been admitted to Niigata University Hospital. We made every effort to exclude children over 10 years old, or whose clinical laboratory tests showed abnormal values. Sera taken from the remaining 100 children were used to determine the reference values of serum creatinine in that age group. The reference values for creatinine concentrations by the enzymatic method were calculated by a nonparametric method, to be within 9511I0 of the samples. Reference values of creatinine clearance Serum creatinine and 24 h urinary creatinine was measured in 330 healthy subjects (161 men aged 20-81 years, 169 women aged 20-77 years) of Denka Seiken Co., Ltd, and local area volunteers.

Reference values of serum and urine creatinine The criteria used in the selection of subjects were the same as those employed previously. Creatinine clearances were calculated by use of the following formula: CC = U x V x i - 731 (P x A), where CC is creatinine clearance (mLI min); U = concentration of urinary creatinine C/LmollL); P = concentration of serum creatinine (umul/L); Y = (24 h urinary volume)/1440 (mL/min); I' 73 = standard body surface area; A = individual body surface area extrapolated from weight and height from an established nomogram. 16 RESULTS Calibration curve and recovery studies The calibration curve was linear up to I' 77 mmol/L of creatinine for serum and 26' 52 mmol/L for unne. Analytical recovery of creatinine added to both pooled human serum and urine was investigated (Table 1). The average recovery of creatinine in serum was 102'81170, and in urine the average was 98'4%. . Precision studies Within-run precision studies were carried out by assaying fifty replicates each of pooled serum and urine samples. Using the enzymatic method mean creatinine values of 82· 8 j.tmoliL and 322'1 umol/L for pooled serum were found, with SDs of 0·9 and 2'0, and CV of 1·09% and O' 62%, respectively. Mean urinary creatinine values of 3'76 mmol/L and 9· 93 mmol/L were found by this method, with SDs of 35' 4 and 74· 3 j.tmol/L, and CVs of 0·94% and 0,75%, respectively. Between-run precision studies on the enzymatic method were conducted by performing daily creatinine assays on serum and urine pools, over a period of 30 days. Mean values of 82· 9 j.tmoliL and 294· 4 j.tmol/L were found for pooled serum, with SD=I'4 and 5,2, and CV=I'69% and 1'77%, respectively. For urine, mean values of 6· 17 mmol/L and 12· 09 mrnol/L were found by the method, with SD=65'4 and 122'0 and CV = 1. 06% and 1·01 %, respectively. Comparison of the enzymatic method with other methods Five hundred and ninety-one samples for serum and 126 samples for urinary creatinine were used to obtain correlations between the enzymatic method and the Jaffe-rate method. The correlation coefficients and regression equations were

1. method

TABLE

525

Recovery experiments with the enzymatic

Added creatinine S {J.tmoIlL) 44

88 133 177

Serum or urine values

Creatinine found

Added creatinine recovered (0J0)

71 71 71 71

118 162 202 252

106·8 103·4 98'5 102·3

5·40 5·40 5·40 5·40

6·20 9·00 10'71 14·31

90·9 101'7 100·2 100·8

U (mmoIlL)

0·88 3'54 5·30 8·84

S = Pooled human serum. U = Pooled human urine.

as follows: serum-r=0'969, Y=0'947X-14; urine-r = O' 999, Y = O' 994X + 19, where X C/Lmol/L) indicates the Jaffe-rate method, and Y (j.tmol/L) indicates the enzymatic method. In the same way, 33 serum and 32 urine samples were used to obtain correlations between the enzymatic method and the HPLC method. The correlation coefficients and regression equations were as follows: serum-r=0'997, Y=0'989X-4; urine-r=0'999, Y=0'945X+ 368, where X (j.tmol/L) indicates the HPLC method, and Y C/Lmol/L) indicates the enzymatic method. A further, 50 samples of serum were used to obtain correlations between the single-slide method (X: j.tmollL) and the proposed enzymatic method (Y: j.tmollL). The results were as follows: r=O'999, Y= l'OO6X -19. The results of the enzymatic creatinine method for serum correlated well with the other three methods. However, the serum creatinine concentrations tended to be somewhat lower by the enzymatic method. The mean bias of the enzymatic method to the HPLC method was approximately - 4 j.tmol/L, to the Jaffe-rate method was - 14 j.tmol/L and to the single-slide method was - 19 j.tmoliL. Interference studies Several substances seemed to show little or no interference with the method (Table 2). The addition of 684 j.tmoliL bilirubin raised serum creatinine concentrations by 44 j.tmol/L, but had no effect when added at concentrations under 171 j.tmol/L. Heparin itself had no noticeable effect on the results, but when it contained 1% benzyl alcohol, the apparent creatinine

526

Sugita et al.

2. Effect of substances added to serum on determination of creatinine

TABLE

No interference Substances tested Anticoagulants Benzyl alcohol (in heparin) Disodium EDTA Heparin Potassium oxalate Sodium citrate Sodium fluoride Metabolites Albumin Ammonia Bilirubin Creatine Formic acid Glucose Glycine Hemoglobin Lactic acid Pyruvic acid (at lactate dehydrogenase 452 U/L) Triglyceride Drugs Aminophylline Ampicillin Ascorbic acid Atropine sulphate Bromhexine HCl Butylscopolamine bromide Carbenicillin sodium Carnigen Cefazolin sodium Citicoline Cyclophosphamide Cytarabine Deslanoside Ephedrine HCI Flavin adenine dinucleotide Fluorouracil Fosfestrol Fursultiamine Glutathione Lidocaine Mitomycin C Pyridoxal phosphate Retinol palmitate Sulpyrine Vincristine sulphate

up to 9·3,..moIlL 6'OmmollL 1(Yl UlL

21'7 mmollL 34'OmmollL 476· 2 mmollL 1·7 mmollL 5'9mmollL 171·0,..moIlL 2'9mmollL 10·9 mmollL 333·0 mmollL 13·3 mmollL 58·8,..moIlL 2'2mmollL 908 . 5 ,..mollL 2·0 giL 13'9mmo1lL 4·1 mmollL 5'7 mmollL 72·0,..moIlL 474·4,..moIlL 4'5 mmollL 47'4mmo1lL 275' 6 ,..mollL 21'OmmollL 10·2 mmollL 7·2 mmollL 8·2 mmollL 42·4,..moIlL 19·8 mmollL 120·6,..moIlL 3·8 mmollL 11·7 mmollL 114·9,..mol/L 65·I,..moIlL 1'1 mmollL 598 .0 ,..mollL 113·I,..moIlL 6·3,..moIlL 711' 5 ,..mol/L 10·8,..moIlL

concentrations increased. Benzyl alcohol, when employed in concentrations of up to 9 /LmollL (heparin 100 U/L) had no effect. However, at 92/Lmo1lL (heparin 1000 U/L) which is the standard concentration, the serum creatinine concentration increased by Ll umol/L. When pyruvate was added (454/LmoIlL; at activities of

2840 lUlL of lactate dehydrogenase), a drop of 18/Lmo1/L was noted, but it had no effect, when lactate dehydrogenase activity was 1520 lUlL (upper reference limit of lactate dehydrogenase in serum in our laboratory is 434 lUlL). There was no visible difference in the effect of any of the drugs tested, when at therapeutic concentrations. Reference values of creatinine and creatinine clearance The calculation of reference values of creatinine was carried out for both the enzymatic and the Jaffe-rate methods (Table 3). The reference intervals for serum creatinine in adults by this procedure were 55-96/LmoIlL in men and 4O-66/LmoIlL in women and for urine were 9· 4619·01 mmollday in men and 6'75-10'61 mmoll day in women. Adult reference values in serum, as determined by the enzymatic method, were lower by about 16/Lmo1lL than those by the Jaffe-rate method, and the serum creatinine concentrations for men were higher by about 23 /LmollL than those for women. Similar differences were obtained when comparing the urinary reference values of men and women. Those of men were higher by about 5· 57 mmollday than those of women. Creatinine concentrations by the enzymatic method in the serum of children under one year of age averaged 16/LmoIlL, and then rose gradually with age, reaching 36/-Lmo1lL at 10 years. The reference values in children, as determined by the enzymatic method were lower by approximately 20/Lmo1lL than those by the Jaffe-rate method in any age group. The calculation of reference values of creatinine clearance was carried out for the enzymatic method. The results were stratified by age and sex (Table 4). The reference intervals of creatinine clearance averaged 88· 0-176' 4 mllmin in men and 75'7-173'Omllmin in women. DISCUSSION

We used the principle of enzymatic determination of creatine.Pv" in order to measure creatinine concentrations in serum and urine. The enzymatic method was specific for determining creatine, and was in no way interfered with by substances normally present in the urine. 14 The enzymatic method described here is sensitive, does not require sample dilution, and is simple to perform. Generally, it is necessary to dilute the urine samples, because creatinine concentrations in

Reference values of serum and urine creatinine

527

TABLE 3. Reference intervals for creatinine as determined by the enzymatic and the Jafferate methods

Enzymatic

Age (years)

Jaffe-rate

n

Interval

Average

Interval

Average

0-1 2-5 6-9 10-

18 26 41 15

3·5-29·2· 3'5-39·8· 17·7-46·0· 19·4-52·2·

15·9· 22·1· 31·8· 36·2·

18·6-55·7· 31·8-53·9· 37'1-66' 3· 39·8-66·3·

37·1· 42·4· 52·2· 53·0·

S Adult M F

235 355

54·8-96·4· 39·8-66·3·

76·0· 53·0·

73·4-109·6· 60·1-79·6·

91·9· 69·8·

U Adult M F

123 91

9'46-19'801 6'63-12'55 1

14'23 t 8'66 t

9'46-19'0J! 6'75-1O'6J!

14'67 1 9'641

·I"mol/L. 'mmol/day. M = Male subjects; F = female subjects; S = serum; U = urine. 4. Reference intervals for creatinine clearance, according to age and sex by the enzymatic method. The values were calculated by standard body surface area of 1· 73 m'

TABLE

Men Age (years)

20-29 30-39 40-49 50-59 60-69 70Total

Women

n

Interval (mLimin)

Average (mLimin)

n

Interval (mLimin)

Average (mLimin)

34 54 42 15 8 8 161

87'1-174·5 94,9-175,3 91,3-176,9 73,1-186,2 44,4-170,7 45,1-131,6 88,0-176,4

130·8 135'1 134·1 129·6 107·5 88·4 132·2

30 48 21 6 48 16 169

101,6-172,2 86,9-195,6 86,3-183'5 99,7-138,0 73,8-139'7 47,8-153,9 75-7-173'0

136·9 141·0 134·9 118·9 106'7 100·9 124·4

urine are higher than those in serum. In the course of developing our method, we sought for a means of eliminating the need to dilute the urine samples. This was achieved by reducing the volume of urine samples and establishing different reagent conditions from those used for the serum assay. The reagent mixture was stable for a period of 9 days at 4 DC, making it ideal for routine use in clinical laboratories. Recovery studies, within-run CV and betweenrun CV compared satisfactorily with data published for other methods.v-" Although our method was relatively free of interference from picrate-reactive substances.v" a few could possibly interfere. Negative interference was observed with pyruvate, when associated with high lactate dehydrogenase activity. This phenomenon indicated that oxarnic acid totally suppressed lactate dehydrogenase activity with a lactate substrate, but only partially suppresses its activity with pyruvate, the remaining lactate dehydrogenase reacting with pyruvate and thereafter consuming NADH.

However, excess oxamic acid suppressed creatinine amidohydrolase activity. There was positive interference with highly concentrated (over 171,..mol/L) bilirubin with this enzymatic method. Bilirubin also interferes, at lower concentrations, in the enzymatic method using peroxidase." Among anticoagulants, only benzyl alcohol with added heparin interfered with the reaction. The interference of various other factors, such as high creatine content or lidocaine treatment had a direct effect on the Kodak single-slide creatinine method.P-" but not on our enzymatic method. The creatinine values obtained by the enzymatic method agreed well with those achieved by the HPLC method. The bias of the single-slide method is presumably related to differences in calibration. There are some differences between reference values of enzymatic methods and the Jaffe reaction-based methods. Reference values of serum creatinine by the enzymatic method are approximately 18 ILmol/L lower.

528

Sugita et al.

As recently as 1990, Savory-? examined a large number of children with regard to establishing reference values for serum creatinine by the Jaffe method. Serum creatinine concentrations in children by the enzymatic method were found to be lower than those cited in Savory's data, the average disparity being 27 JLmollL. Likewise, the data gained from the enzymatic method were lower than those from our Jaffe-rate method. The average disparity here was 20 JLmollL. Serum creatinine concentrations in children by both the above methods increased with age (Table 3). The reference values for creatinine clearance by the proposed method were higher than those by the Jaffe-rate method. These results were almost the same as those obtained for other enzymatic methods.vand may indicate that the Jaffe method measures other substances in the serum as well as creatinine. All the creatinine clearance values in Table 4 were expressed, as per 1'73 m 2 standard body surface area. Average intervals for creatinine clearance in men were 88· 0176'4mL/min, and in women were 75'7173· 0 mL/min. When those values were calculated by the Japanese standard body surface area of l ' 48 m-, those values were reduced to 75' 3150· 9 mL/rnin in men, and 64' 8- 148· 0 mLimin in women. In 1976, Rowe et al. 22 measured the creatinine clearance in 548 men. These clearance levels averaged 132· 6 mL/min in subjects ranging in age from 35-44 years. For men of 30-39 years of age, the clearance was 135· 1 mL/min. The creatinine clearances obtained by our enzymatic method agreed well with those of Rowe. We considered that this agreement with Rowe's method seems to be ascribable to the removal of noncreatinine chromogens by Lloyd's reagent. REFERENCES I Folin O. Beitrag zur Chemie des Kreatinins und Kreatins in Harne. Z Physiol Chem 1904; 41: 223-42 2 Jaffe M. Ueber den Niederschlag, welchen Pikrinsaure in normalem Ham erzeugt und iiber eine neue Reaction des Kreatinins. Z Physiol Chem 1886; 10: 391-400 3 Watkins Pl. The effect of ketone bodies on the determination of creatinine. Clin Chim Acta 1967; 18: 191-6 4 Cook lGH. Creatinine assay in the presence of protein. Clin Chim Acta 1971; 32: 485-6 5 Mascioli SR, Bantle lP, Freier EF, Hoogwerf Bl. Artifactual elevation of serum creatinine level due to fasting. Arch Intern Med 1984; 144: 1575-6 6 Rockerbie RA, Rasmussen KL. Rapid determination of serum creatinine by an ion-exchange technique. Clin Chim Acta 1967; 15: 475-9

7 Brown ND, Sing HC, Neeley WE, Koetitz SE. Determination of "true" serum creatinine by highperformance liquid chromatography combined with a continuous-flow microanalyzer. Clin Chem 1977; 23: 1281-3 8 Ambrose RT, Ketchum DF, Smith lW, Creatinine determined by "high-performance" liquid chromatography. Clin Chem 1983; 29: 256-9 9 Moss GA, Bonder RlL, Buzzelli DM. Kinetic enzymatic method for determining serum creatinine. Clin Chem 1975; 21: 1422-6 10 Tanganelli E, Prencipe L, Bassi D, Cambiaghi S, Murador E. Enzymic assay of creatinine in serum and urine with creatinine iminohydrolase and glutamate dehydrogenase. Clin Chem 1982; 28: 1461-4 11 Toffaletti r, Blosser N, Hall T, Smith S, Tompkins D. An automated dry-slide enzymatic method evaluated for measurement of creatinine in serum. Clin Chem 1983; 29: 684-7 12 Fossati P, Prencipe L, Berti G. Enzymic creatinine assay: A new colorimetric method based on hydrogen peroxide measurement. Clin Cbem 1983; 29: 1494-6 13 Yasuhara M, Fujita S, Furukawa 1, Arisue K, Kohda K, Hayashi C. Continuous-flow enzymic determination of creatine in urine. Clin Chem 1981; 27: 1888-91 14 Yasuhara M, Fujita S, Arisue K, Kohda K, Hayashi C. A new enzymatic method to determine creatine. Clin Chim Acta 1982; 122: 181-8 15 Mauck lC, Mauck L, Novros J, Norton GE. Development of a single slide Kodak Ektachem't thin-film assay for serum and urine creatinine [Abstract]. Clin Chem 1986; 32: 1197-8 16 Bell GH, Davidson IN, Scarborough LH. Textbook of Physiology and Biochemistry, 6th edn. Edinburgh: E & S Livingstone, 1965 17 Goren MP, Osborne S, Wright RK. A peroxidasecoupled kinetic enzymatic procedure evaluated for measuring serum and urinary creatinine. Clin Chem 1986; 32: 548-51 18 Sena SF, Syed D, McComb RB. Effect of high creatine content on the Kodak single-slide method for creatinine. Clin Chem 1988; 34: 594-5 19 Roberts RT, Alexander NM, Kelner Ml. Definitive liquid-chromatographic demonstration that Nethylglycine is the metabolite of lidocaine that interferes in the Kodak sarcosine oxidase-coupled method for creatinine. Clin Chem 1988; 34: 2569-72 20 Savory Dl. Reference ranges for serum creatinine in infants, children and adolescents. Ann Clin Biochem 1990; 27: 99-101 21 Apple F, Bandt C, Prosch A, Erlandson G, Holmstrom Y, Scholen J, Googins MK. Creatinine clearance: Enzymatic vs 1affe determinations of creatinine in plasma and urine. Clin Chem 1986; 32: 388-90 22 Rowe lW, Andres R, Tobin lD, Norris AH, Shock NW. The effect of age on creatinine clearance in men: A cross-sectional and longitudinal study. J Gerontol 1976; 31: 155-63