Estimation of glomerular filtration rate by MDRD ... - Springer Link

11 downloads 0 Views 184KB Size Report
Giovanni Lombardi • Patrizia Lanteri •. Giuseppe Banfi. Received: 3 February 2011 / Accepted: 15 April 2011 / Published online: 26 April 2011.

Eur J Appl Physiol (2012) 112:201–206 DOI 10.1007/s00421-011-1969-1


Estimation of glomerular filtration rate by MDRD equation in athletes: role of body surface area Radoje Milic • Alessandra Colombini • Giovanni Lombardi • Patrizia Lanteri • Giuseppe Banfi

Received: 3 February 2011 / Accepted: 15 April 2011 / Published online: 26 April 2011 Ó Springer-Verlag 2011

Abstract Creatinine-based equations to estimate the glomerular filtration rate (GFR) have recently been advocated over serum creatinine values as a valuable tool to more accurately assess kidney function. The Cockcroft– Gault (CG) equation requires a body weight parameter, whereas the Chronic Kidney Disease-Epidemiology Collaboration (CKD-EPI) and Modification of Diet in Renal Disease (MDRD) Study equations do not. In this study we evaluated the effect of the calculated body surface area (BSA) on MDRD values in professional athletes characterized from different body mass index, gender, and sport discipline. Serum creatinine concentration was measured by Jaffe reaction in 17 male rugby players and 28 male and 26 female swimmers, before the start of training and throughout the competitive season. The values of estimated GFR (eGFR) calculated for creatinine determination by means of CG and CDK-EPI with respect to MDRD formula showed a significant difference in different groups of athletes. The statistical significance was confirmed for BSAcorrected MDRD-derived eGFR values in rugby players and in male swimmers, but not in female swimmers, who showed a BSA close to the ‘‘standard’’ value of 1.73 m2 Communicated by Susan A. Ward. R. Milic Faculty of Sport, Institute of Sport, University of Ljubljana, Gortanova 22, 1000 Ljubljana, Slovenia A. Colombini (&)  G. Lombardi  P. Lanteri  G. Banfi Laboratory of Cell Cultures and Molecular Biology, IRCCS Istituto Ortopedico Galeazzi, Via R. Galeazzi 4, 20161 Milan, Italy e-mail: [email protected] G. Banfi School of Medicine, University of Milan, Milan, Italy

traditionally included in MDRD equation. The CG equation can overestimate the eGFR in healthy overweight subjects such as rugby players, whereas the MDRD formula systematically underestimates it. The differences between the two equations increase as a function of BMI, appearing highest in rugby players and lowest in female swimmers. Real BSA correction of the MDRD equation could help to avoid an overestimation of renal excretory function in subjects with increased BSA. Keywords Professional athletes  Creatinine  Glomerular filtration rate  CG equation  MDRD equation  Body surface area

Introduction In sports medicine, serum creatinine is widely used for evaluating general health status of athletes, particularly given the critical nature of fluid and electrolyte balance in this setting. Reference values for any biochemical parameter specific to sportsmen have never been established and, therefore, those used for the general population are routinely applied to athletes. Notably, the values for serum creatinine observed in professional athletes are higher than those found in the general population (sedentary subjects), as demonstrated in a large series of top-level sportsmen competing in eight different sports disciplines (Banfi and Del Fabbro 2006). The serum creatinine concentration in athletes has been shown to relate to their body mass index (BMI) (Banfi et al. 2006). However, some athletes with a low BMI (i.e. cyclists) have serum creatinine concentrations lower than non-physically active subjects (Lippi et al. 2004), whereas, athletes having a high BMI (i.e. rugby players) demonstrate high concentrations of this parameter.



A variety of exogenous (radioisotopic and non-radioisotopic) and endogenous markers have been used to estimate the glomerular filtration rate (GFR), of these, inulin has long been regarded as the gold standard. However, lack of availability of simple methods of measurement remains an impediment to universal usage and a variety of alternative ‘‘silver’’ standard estimates of GFR are used, including 125 I-iothalamate, 51Cr-ethylenediaminetetraacetic acid (DTPA), 99mTc-DTPA and iohexol. 51Cr-EDTA is preferred to 99mTc-DTPA and 125I-iothalamate since its clearance is considered to be closest to that of inulin and the British Nuclear Medicine Society has recently endorsed its use as the standard measure of GFR (Lamb et al. 2005). Creatinine-based equations to estimate GFR have recently been advocated over mere serum creatinine values as a valuable tool to more accurately assess kidney function (Myers et al. 2006; Lamb et al. 2005). The oldest and most commonly used formula proposed for this purpose is the Cockcroft–Gault (CG) equation (Cockcroft and Gault 1976). Recently, the Modification of Diet in Renal Disease (MDRD) Study Group equation was proposed and recommended by expert working groups for calculating the estimated glomerular filtration rate (eGFR) (Levey et al. 2000). More recently, Levey et al. developed and validated a new Chronic Kidney Disease-Epidemiology Collaboration (CKD-EPI) creatinine equation, more accurate than the MDRD Study equation, especially for higher GFRs, which could replace it for routine clinical use (Levey et al. 2009). The use of equations to assess kidney function, especially the MDRD formula, has some practical advantages as listed in published recommendations (Myers et al. 2006; Lamb et al. 2005). However, the validation of this equation was performed in patients suffering from chronic kidney disease (CKD), but in healthy individuals has not yet been validated or defined. In general, eGFR calculated by means of the MDRD equation will systematically deviate toward mean (regression to the mean) of the population: it was derived from patients with CKD having lower GFR values. Given these constraints inherent to the MDRD formula, values of eGFR should simply be reported as [60 mL/min/1.73 m2 (Myers et al. 2006). Moreover, it should be considered that GFR is proportional to the body mass and it has become standard practice to normalize GFR to the arbitrary value of body surface area (BSA) of 1.73 m2 to enable comparisons of GFR between individuals. The use of this arbitrary value offers theoretical and practical advantages compared with the real BSA value, but, as reviewed by Heaf (2007) it is clearly no longer applicable to Western population, since periodic adjustments of the value would be required (Heaf 2007). Considering BSA is of particular importance, especially for obese subjects, since the use of absolute, noncorrected with the true BSA value GFR may led to overestimation of renal function (Rigalleau et al. 2006).


Eur J Appl Physiol (2012) 112:201–206

The objective of the present study was to define the effect that the higher serum creatinine values, noted in healthy, top-level rugby players during a complete competitive season, would have on corresponding eGFR values, using the CG, CKD-EPI and MDRD equations, the last one both original or corrected for BSA values. The lack of comparison with a reference method is a limitation of the study, but its principal aim was not the validation of equations in a clinical setting, but a determination of possible advantages and/or the pitfalls of using these formulas in the considered athletes, in whom a heavy training workload and psychophysical stress can induce, at times, apparent pathological modifications of serum creatinine.

Materials and methods We measured the serum creatinine concentration in 17 male athletes belonging to the Italian National Rugby team and in 28 male and 26 female athletes belonging to the Slovenian National Swimming team. The mean age was 26.0 ± 4.2 years for rugbyists, 21.9 ± 4.2 years for male swimmers and 18.4 ± 2.2 years for female swimmers. Blood drawings were performed before the start of training; blood was immediately centrifuged and serum stored at -20°C, strictly following preanalytical warnings (Banfi and Dolci 2003). All the subjects recruited for the study were in the fasting state and had rested for a period of 24 h since their last competition or training session. Because serum creatinine was one of the parameters routinely measured for the evaluation of health status in these athletes, no additional blood was drawn for this study. The serum creatinine level was measured by Jaffe reaction in Aeroset c8000 (Abbott, Chicago, USA) for rugby players and in Roche Hitachi 917 (Roche, Basel, Switzerland) for swimmers. The reproducibility of the methods showed a coefficient of variation \2% within-run and 4% between-run. BMI in athletes was calculated as weight/(height)2; the BMI values reported were registered before the start of training and competition season. However, they were reconfirmed during the season for all the athletes and demonstrated no significant differences from the basal value. The body surface area (BSA) was calculated by using both the historical formula of DuBois and DuBois [BSA (m2) = [weight (kg)0.425] 9 [height (cm)0.725] 9 0.007184] and the formula of Haycock and Schwarz [BSA (m2) = [weight (kg)0.5378] 9 [height (cm)0.3964] 9 0.024265] as recommended by the British Nuclear Medicine Society (Lamb et al. 2005). Diet was controlled by team physicians and it was unchanged during the whole season. Creatine supplementation was never prescribed to athletes by the team physicians,

Eur J Appl Physiol (2012) 112:201–206

and none of the athletes reported personal intake of creatine throughout the time considered period. Paired Student’s t test was used to statistically evaluate the difference between eGFR values obtained with the CG equation [mL/min = [(140 - age (years)) 9 weight (kg)/(0.814 9 serum creatinine (lmol/L))] 9 (0.85 if female)], CKD-EPI equation [GFR = 141 9 min(Scr/j, 1)a 9 max(Scr/j, 1)-1.209 9 0.993Age 9 1.018 [if female] 9 1.159 [if black], where Scr is serum creatinine (mg/dL), j is 0.7 for females and 0.9 for males, a is -0.329 for females and -0.411 for males, min indicates the minimum of Scr/j or 1, and max indicates the maximum of Scr/j or 1] and MDRD equation [mL/min/1.73 m2 = 186 9 [serum creatinine (lmol/L) 9 0.011312]-1.154 9 [age]-0.203 9 [1.212 if black] 9 [0.742 if female]] (Lamb et al. 2005), calculated according to the creatinine measurement traceable to the international reference standard, and between BSA values as estimated with two equations. For each group of athletes, eGFR obtained from MDRD equation was corrected with the corresponding BSA value according this formula GFRcorrected = GFRmea2 sured 9 (1.73/BSA m ) (Lamb et al. 2005) and paired Student’s t test was used to statistically evaluate the difference from the correspondent original MDRD values. Unpaired Student’s t test was used to statistically evaluate the differences between the three groups of athletes. Pearson correlation coefficient was used for regression analysis. The probability value of p = 0.05 was considered as the significance threshold.

Results The anthropometrical characteristics of the study participants are provided in Table 1. There was a statistically significant difference (p \ 0.01) between the BSA calculated by using the two different equations in rugby players. A correlation was found between serum creatinine values and BMI in overall group of athletes (r = 0.82). This correlation was confirmed for all the swimmers (y = 55.5 ? 1.25x, r2 = 0.06) and rugby players (y = 144.27 – 0.98x, r2 = 0.04), suggesting that the anthropometric characteristics of athletes belonging to different sport disciplines are influencing the creatinine concentrations. The values of creatinine measurements and the corresponding eGFR calculated by means of CG, CKD-EPI and Table 1 Anthropometrical characteristics of the athletes

* Significantly different


MDRD formula as well as of MDRD formulas corrected by BSA are listed in Table 2. In all cases values obtained using CG and CKD-EPI equations were significantly higher than those calculated using MDRD equation (paired t test, p \ 0.01 for male athletes, and p \ 0.05 for female ones for GC values and p \ 0.01 for both sexes for CKDEPI values). It is evident that the differences between CG and MDRD values increase as a function of BMI, inducing higher disequilibrium in male athletes than in female ones. There is significant difference (p \ 0.05) between the MDRD values of rugby players and male swimmers and for both CG and MDRD values between male and female swimmers. These differences are mainly due to different BMI and BSA. The correction by BSA of MDRD values is statistically significant for male athletes, but not for female ones. The correlation between serum creatinine values and eGFR values obtained by CG and MDRD and by MDRD corrected by real BSA was not significant. No significance was found by comparing age and eGFR values in overall athletes, but also in the different groups.

Discussion The common reference range for creatinine in the general population is 76–115 lmol/L (0.9–1.3 mg/dL) for adult males, by using Jaffe method in automated systems (Banfi et al. 2009). In a report based on the results of a wide population from 102 laboratories, a range of 57–95 lmol/L (0.68–1.13 mg/dL) was determined for adult males, narrower than the above one (Rustad et al. 2004). However, no discernment is made in these ranges for age, race or health status. We previously studied a large number (n = 220) of e´lite male athletes who participated in one of eight sport disciplines, each characterized by different aerobic/anaerobic activities, competitive seasons, training, and anthropometric values. We found that serum creatinine values were higher in these sportsmen than in age-matched sedentary subjects (Banfi and Del Fabbro 2006). Recently, the use of serum creatinine to assess renal function has been criticized because it is affected by many factors independent of GFR including age, gender, body size, diet, drug consumption, and analytical method of measurement (Myers et al. 2006). Estimating equations that combine serum creatinine concentrations with known external factors affecting creatinine measurement have

Parameter (mean ± SD)

Rugby players

Male swimmers

Female swimmers

Body mass index (kg/m2)

28.5 ± 2.3

22.9 ± 1.6

20.7 ± 2.2

Body surface area (DuBois and DuBois; m2)

2.25 ± 0.14*

2.02 ± 0.16

1.72 ± 0.15

Body surface area (Haycock and Schwartz; m2)

2.29 ± 0.15*

2.00 ± 0.15

1.72 ± 0.15



Eur J Appl Physiol (2012) 112:201–206

Table 2 Mean values ± SD of serum creatinine and eGFR calculated by means of Cockcroft and Gault (CG), CDK-EPI and MDRD equations Serum creatinine; lmol/L (mean ± SD)

eGFR from CG; mL/min (mean ± SD)

115.8 ± 10.6

123.1 ± 16.8

75.7 ± 8.6

Male swimmers

93.3 ± 7.6

120.9 ± 15.8

Female swimmers

77.5 ± 6.5

98.9 ± 10.3

Rugby players

eGFR from CKDEPI; mL/min/ 1.73 m2 (mean ± SD)

eGFR from MDRD; eGFR from MDRD BSA eGFR from MDRD BSA corrected; mL/min/m2 corrected; mL/min/m2 mL/min/1.73 m2 (mean ± SD) (mean ± SD)a (mean ± SD)b 71.4 ± 7.7§,#

55.1 ± 7.4*

54.0 ± 7.2*

101.3 ± 10.8

95.8 ± 10.5§,#,^

84.2 ± 13.6*

84.8 ± 14.0*

96.9 ± 11.3

89.9 ± 10.6§,#

89.2 ± 11.9

89.2 ± 11.9


Significantly different from CG values: p \ 0.05 for females, p \ 0.01 for males; # significantly different from CKD-EPI values: p \ 0.01 for females and males; ^ significantly different from value of rugby players; * significantly different from correspondent original MDRD: p \ 0.01


BSA is calculated on the basis of equations by DuBois and DuBois


BSA is calculated on the basis of equations by Haycock and Schwartz

been proposed as more accurate means to evaluate GFR. However, estimating equations are recognized as being potentially inaccurate in populations different from those in whom the equations were developed (i.e. patients with chronic kidney disease). In a healthy population of 365 potential kidney donors (205 women), these equations were found to underestimate GFR by 14 mL/min (CG) and by 29 mL/min/1.73 m2 (MDRD) when compared with nonradiolabeled iothalamate reference method (Rule et al. 2004). It was found that lower values of eGFR (i.e. \60 mL/min/1.73 m2) more accurately approximate actual GFR values (Myers et al. 2006). The literature is replete with evidence demonstrating the limitations for the use of both equations. When obesity is present (body mass index [ 30 kg/m2), no reliable eGFR can be obtained using either equation (Verhave et al. 2005). Moreover, the gross overestimation in obese subjects by using the CG equation suggests an urgent need for further validation of GFR equations in the obese, as stated by Lamb et al. (2005). Verhave et al. (2005) found that the CG equation markedly underestimated eGFR in lean subjects and clearly overestimated eGFR in obese subjects. Conversely, better accuracy might be achieved in matching the GFR formula to the population surveyed. The accuracy of CDK-EPI is claimed better than this of MDRD (Levey et al. 2009) in clinical setting. However, we can remark that the limits of MDRD on overweight and healthy subjects are not solved by the newly proposed equation. Some limitations on the use of eGFR equations, and especially MDRD, have been reported also in athletes (Banfi et al. 2009). The values of eGFR calculated by MDRD in cyclists at rest appeared higher than those of sedentary people, since the results of this equation could be unreliable when the athletes are unaccustomed to the training load (Lippi et al. 2008a); the equation, however,


could be applied to athletes when acute variations of GFR during intense physical exercise are studied (Lippi et al. 2008b). The MDRD equation does not require body weight because it normalizes the eGFR to a standard body surface area of 1.73 m2, calculated on the basis of DuBois and DuBois formula (Lamb et al. 2005). The significant differences among adjusted and unadjusted MDRD we found, by using DuBois and DuBois and Haycock and Schwarz, more modern and presumably more accurate, do not appear to be dependent upon which of the two equations was used in calculating BSA. This begs for definition of a consensus regarding which of these two formulas should be routinely used for BSA determinations. It should be remarked that the CG and MDRD values are often compared after a modification of CG data, by dividing by 1.73 m2, as reported in wide cohorts of subjects from general population (Wetmore et al. 2010) or in patients suffering from kidney diseases (Zhang et al. 2010). The use of a ‘‘standard’’ value of 1.73 m2 is generally accepted for evaluating and comparing eGFR equations, actually omitting a reliable and effective evaluation of the impact of BSA on calculations. However, the concept that GFR should be dependent on BSA is contentious (Lamb et al. 2005) with some studies demonstrating that the relationship between GFR and BSAis poor (Dooley and Poole 2000). The mean BSA of the patients participating in the MDRD study was 1.91 ± 0.23 m2 (Levey et al. 2000), higher than the currently used value of 1.73 m2, likely due to the increasing prevalence of overweight subjects in the general population. The standard BSA of 1.73 m2 represents the body surface of an average size young adult. However, it is also lower than the BSA seen in our male athletes; in particular, for purposes of eGFR calculation, rugby players can be

Eur J Appl Physiol (2012) 112:201–206

considered similar to the overweight subjects, in whom the correction for BSA of inulin-based GFR has been shown to result in a significant ([20 mL/min/1.73 m2) underestimation of GFR (Delanaye et al. 2005). The BSA values of female swimmers is close to the ‘‘standard’’ value applied to MDRD formula: it is evident that the gender adjustment already included in the equation could not be sufficient to describe correctly the eGFR in male and female athletes performing the same sport discipline and should always be corrected also by BSA. Nevertheless, the range of eGFR is relatively small and conclusion about potential correlation between eGFR and BSA is hazardous. The Australasian Creatinine Consensus Group guidelines stated that ‘‘adjusted’’ GFR estimates are adequate except in patients with a body size that is very different from average (Mathew and Australasian Creatinine Consensus Working Group 2005). It should be noted that this is true not only for obese subjects (BMI [ 30 kg/m2), but also for healthy subjects with BMI values in the range 25–30 kg/m2. This is of practical importance because lean body mass, a finding particularly prevalent among athletes, has been shown to be independent of serum creatinine values in a large population study of healthy individuals (Swaminathan et al. 1986). Multiple variables undoubtedly contribute to this finding and include not only the production and excretion of creatinine, but its volume of distribution in total body water. The latter is related to fat-free or lean body mass. Our data provide further confirmation that ‘‘adjusted’’ MDRD for a BSA of 1.73 m2 can result in an overestimation of renal excretory function in the setting of an increased BSA. While our study demonstrates this in athletes, previous studies have shown similar findings in more sedentary, ‘‘normal’’ populations with increased BSA (Delanaye et al. 2005). In conclusion, we demonstrate that differences do exist between BSA (1.73 m2) and real BSA, but there is no proof that using actual BSA will improve the equations accuracy. The use of equations to estimate renal excretory function should be recommended in chronic kidney disease, but not in the general population. This is particularly important for professional athletes that show a wide range of serum creatinine concentrations, sometimes higher than reference intervals, and with marked gender differences, even in the same sport discipline.

References Banfi G, Del Fabbro M (2006) Serum creatinine values in elite athletes competing in 8 different sports: comparison with sedentary people. Clin Chem 52:330–331

205 Banfi G, Dolci A (2003) Preanalytical phase of sport biochemistry and haematology. J Sports Med and Phys Fitness 43:223–230 Banfi G, Del Fabbro M, Lippi G (2006) Relation between serum creatinine and body mass index in elite athletes of different sport disciplines. Br J Sports Med 40:675–678 Banfi G, Del Fabbro M, Lippi G (2009) Serum creatinine concentration and creatinine-based estimation of glomerular filtration rate in athletes. Sports Med 39:331–337 Cockcroft DW, Gault MH (1976) Prediction of creatinine clearance from serum creatinine. Nephron 16:31–41 Delanaye P, Radermecker RP, Rorive M, Depas G, Krzesinski JM (2005) Indexing glomerular filtration rate for body surface area in obese patients in misleading: concept and example. Nephrol Dial Transplant 20:2024–2028 Dooley MJ, Poole SG (2000) Poor correlation between body surface area and glomerular filtration rate. Cancer Chemother Pharmacol 46:523–526 Heaf JG (2007) The origin of the 1 9 73–m2 body surface area normalization: problems and implications. Clin Physiol Funct Imaging 27:135–137 Lamb EJ, Tomson CR, Roderick PJ (2005) Estimating kidney functions in adults using formulae. Ann Clin Biochem 42:321–345 Levey AS, Greene T, Kusek JW, Beck GL (2000) A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 11(Suppl):A0828 Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J, CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) (2009) A new equation to estimate glomerular filtration rate. Ann Intern Med 150:604–612 Lippi G, Brocco G, Franchini M, Schena F, Guidi GC (2004) Comparison of serum creatinine, uric acid, albumin and glucose in male professional endurance athletes compared with healthy controls. Clin Chem Lab Med 42:644–647 Lippi G, Banfi G, Salvagno GL, Montagnana M, Franchini M, Guidi GC (2008a) Comparison of creatinine-based estimations of glomerular filtration rate in endurance athletes at rest. Clin Chem Lab Med 46:235–239 Lippi G, Schena F, Salvagno GL, Tarperi C, Montagnana M, Gelati M, Banfi G, Guidi GC (2008b) Acute variation of estimated glomerular filtration rate following a half-marathon run. Int J Sports Med 29:948–951 Mathew TH, Australasian Creatinine Consensus Working Group (2005) Chronic kidney disease and automatic reporting of estimated glomerular filtration rate: a position statement. Med J Aust 183:138–141 Myers GL, Miller WG, Coresh J, Fleming J, Greenberg N, Greene T, Hostetter T, Levey AS, Panteghini M, Welch M, Eckfeldt JH (2006) Recommendations for improving serum creatinine measurement: a report from the Laboratory Working Group of the National Kidney Disease Education Program. Clin Chem 52:5–18 Rigalleau V, Chauveau P, Lasseur C, Perlemoine C, Barthe N, Raffaitin C, Combe C, Gin H (2006) Indexing glomerular filtration rate for body surface area is useful in obese subjects. Nephrol Dial Transplant 21:821 Rule AD, Gussak HM, Pond GR, Bergstralh EJ, Stegall MD, Cosio FG, Larson TS (2004) Measured and estimated GFR in healthy potential kidney donors. Am J Kidney Dis 43:112–119 Rustad P, Felding P, Franzson L, Kairisto V, Lahti A, Ma˚rtensson A, Hyltoft Petersen P, Simonsson P, Steensland H, Uldall A (2004) The Nordic Reference Interval Project 2000: recommended reference intervals for 25 common biochemical properties. Scand J Clin Lab Invest 64:271–284


206 Swaminathan R, Ho CS, Chu LM, Donnan S (1986) Relation between plasma creatinine and body size. Clin Chem 32:371–373 Verhave JC, Fesler P, Ribstein J, du Cailar G, Mimran A (2005) Estimation of renal function in subjects with normal serum creatinine levels: influence of age and body mass index. Am J Kidney Dis 46:233–241 Wetmore JB, Palsson R, Belmont JM, Sigurdsson G, Franzson L, Indridason OS (2010) Discrepancies between creatinine- and


Eur J Appl Physiol (2012) 112:201–206 cystatin C-based equations: implications for identification of chronic kidney disease in the general population. Scand J Urol Nephrol 44:242–250 Zhang QL, Brenner H, Koenig W, Rothenbacher D (2010) Prognostic value of chronic kidney disease in patients with coronary heart disease: role of estimating equations. Atherosclerosis 211:342–347

Suggest Documents