American Journal of Epidemiology Advance Access originally published online on August 25, 2006
American Journal of Epidemiology 2006 164(8):725-727; doi:10.1093/aje/kwj271
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Original Contribution |
Evaluation of Measures of Urinary Albumin Excretion
1 Department of Nephrology, University Medical Center, Groningen, the Netherlands
2 Department of Clinical Pharmacology, University Medical Center, Groningen, the Netherlands
Correspondence to Dr. Ronald T. Gansevoort, Department of Nephrology, University Medical Center, P.O. Box 30.001, Groningen 9700 RB, the Netherlands (e-mail: R.T.Gansevoort{at}int.umcg.nl).
Received for publication May 16, 2005. Accepted for publication October 13, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT References |
|---|
Albuminuria has recently drawn much attention as a valuable risk marker for cardiovascular and renal disease progression. Albuminuria can be measured and expressed in several ways: 1) in a spot morning urine sample as urinary albumin concentration (mg/liter) or albumin:creatinine ratio (mg/mmol) and 2) in a 24-hour urine collection as urinary albumin excretion (mg/24 hours). It has not yet been clarified which measure for albuminuria is preferable in clinical practice. One of the points on which a choice should be made is which measure shows the least within-person coefficient of variation. From the perspective of their work in the Prevention of Renal and Vascular Endstage Disease Intervention Trial, 1997–2001, the authors discuss several methodological issues that are important when interpreting studies on this topic. It is argued that fresh urine should be used, since freezing at –20°C results in considerable extra variability in the albumin concentration. Furthermore, it is important to use specifically collected urine samples and not portions of a 24-hour urine sample as a surrogate for a spot morning urine sample. Albuminuria follows a circadian rhythm. Consequently, values for the within-person coefficient of variation will therefore be different when they are measured in a portion of a 24-hour urine collection in comparison with a spot morning urine sample.
albumins; albuminuria; creatinine; epidemiologic methods
Abbreviations: SD, standard deviation
Albuminuria has recently drawn much attention as a risk marker for cardiovascular and renal disease progression (1
–4). The amount of albumin excreted in urine over a 24-hour period is considered the "gold standard" for defining albuminuria. However, collecting urine over 24 hours is a cumbersome procedure for most subjects. Therefore, often untimed random specimens are used. In these samples, the urinary albumin concentration can be taken as an alternative for defining albuminuria. Unfortunately, the urinary albumin concentration is dependent on hydration status. When a subject is well hydrated, the urinary albumin concentration will be low because of dilution and vice versa, whereas the 24-hour excretion of albumin would not have been influenced. Another option is using the albumin:creatinine ratio, which "corrects" for dilution or concentration by taking the urinary creatinine concentration into account. The purpose of the recent study by Dyer et al. (5) was to assess the strengths and weaknesses of the urinary albumin:creatinine ratio and the albumin concentration as alternatives to albumin excretion. Although we greatly appreciate their effort to bring more clarity to the discussion of which measure to use, we have several comments concerning their study.
First, they found that the within-person median coefficients of variation for urinary albumin concentration, albumin:creatinine ratio, and albumin excretion are approximately 30–80 percent, with the coefficient of variation for urinary albumin concentration, on average, 1–5 percent lower than the coefficient of variation of the other two measures (5). A problem is that the coefficients of variation for all the variables are so high that they in fact disqualify the use of urinary albumin measures as a reliable risk marker for future cardiovascular events. However, it may well be that these coefficients of variation for the various albumin measures are unrealistically high.
- Samples had been frozen at –20°C for approximately 3 years before being analyzed. As Dyer et al. (5) themselves caution, this may have affected the within-subject coefficient of variation. In our experience, long-term freezing at –20°C results in falsely low urinary albumin concentrations, especially when thawed samples are not hand inverted or vortex mixed before assessment (6). Unfortunately, information on preassessment sample handling is lacking in the study by Dyer et al. (5). More importantly, we also found that even short-term freezing induced considerable variability for albumin measurements (6) but less for creatinine. Others have also observed an increase in the variability of the albumin concentration when assessment takes place after freezing at –20°C (7–10).
- The fact that the two 24-hour urine collections in the study of Dyer et al. (5) were collected at a specific point in time, being 3–6 weeks apart, may have influenced results. It may well be that the coefficient of variability for the various measures of urinary albumin excretion varies over time, and that the variation over time is different for the various albumin measures.
To investigate the possible relevance of these issues, we analyzed results from the Prevention of Renal and Vascular Endstage Disease Intervention Trial (PREVEND-IT study) (11). In this double-blinded intervention study, 864 subjects with persistent high-normal or microalbuminuria were selected at random according to a 2 x 2 factorial design to fosinopril/placebo and to pravastatin/placebo. Persistent high-normal or microalbuminuria was defined as a urinary albumin concentration greater than 10 mg/liter in a spot morning urine sample (first morning void) and a mean urinary albumin excretion of 15–300 mg/24 hours in two subsequent 24-hour urine samples. Other inclusion criteria were the following: blood pressure less than 160/100 mmHg, no use of antihypertensives, total cholesterol less than 8.0 mmol/liter (or <5.0 mmol/liter in the case of a previous myocardial infarction), and no use of lipid-lowering medication. We used the data of all 216 subjects receiving placebo-placebo, since with these subjects the coefficient of variation of albumin measures could be ascertained without the interfering influence of medication. These subjects collected 24-hour urine samples at day 1, day 2, and months 3, 6, and 15. Albumin was measured in fresh urine by immunonephelometry (Dade Behring Diagnostics, Marburg, Germany), with intra- and interassay coefficients of variation of 2.7 percent and 4.5 percent, respectively, and creatinine was measured by dry chemistry (Eastman Kodak, Rochester, New York), with intra- and interassay coefficients of variation of 0.9 percent and 2.9 percent, respectively. Intraindividual variability in albumin excretion was assessed by means of the within-person coefficient of variation, which is calculated from multiple measurements made from the same individual as 100 x standard deviation/mean. To estimate the daily variability for each urinary albumin measure, we calculated each individual's coefficient of variation for the first collection and the repeat collection and then used the median values among participants to summarize daily within-person variability. This procedure was chosen in analogy with the study by Dyer et al. (5).
The baseline characteristics of the subjects under study were as follows: age, 49 (standard deviation (SD): 11) years; male, 59 percent; systolic blood pressure, 127 (SD: 15) mmHg; diastolic blood pressure, 74 (SD: 9) mmHg; total cholesterol, 5.6 (SD: 1.0) mmol/liter; glucose, 4.9 (SD: 1.0) mmol/liter; diabetes mellitus, 3.3 percent; serum creatinine, 83 (SD: 13) µmol/liter; and albuminuria, 22.5 (range: 8.6–125.7) mg/24 hours. Results are shown in table 1.
|
These data show that the within-person coefficients of variation of all albumin measurements are considerably less than the values obtained by Dyer et al. (5
). This may, of course, be due to differences between the subjects included or the techniques applied. However, we hypothesize that the fact that we measured albumin and creatinine in fresh urine instead of in urine that has been frozen at –20°C is more important. Of course, it is logistically difficult to conduct epidemiologic studies on fresh urine samples. It might be that storage at –80°C will be suitable, as suggested by Schultz et al. (7). However, these latter data need confirmation. In our experience, the urinary albumin concentration performs worse than the two other measures. Furthermore, our data show that the within-person coefficient of variation is time dependent. When two urine samples are collected within a short period, the within-person coefficient of variation is superior to when the time between two urine collections is longer. This information is important for the correct interpretation of comparisons between studies reporting on coefficients of variation for the various albumin measures.
Second, the conclusions by Dyer et al. (5) are based on a comparison of the urinary albumin concentration and the albumin:creatinine ratio in relation to a reference standard (i.e., 24-hour albumin excretion). In daily practice, the urinary albumin concentration and the albumin:creatinine ratio are assessed from spot morning urine samples. However, in the study by Dyer et al. (5), the same 24-hour urine sample was used for all measurements. It is generally acknowledged that urinary albumin and creatinine excretions follow a circadian rhythm (12, 13). This rhythm is dependent on posture, exercise, and dietary factors, such as protein intake, fluid intake, and so on. This is not a theoretical issue. In the Prevention of Renal and Vascular Endstage Disease Intervention Trial, study, subjects collected a spot morning urine sample (first morning void) 3 months before entering the trial. In this sample, the urinary albumin concentration was 18.2 mg/liter, the urinary creatinine concentration was 12.3 mmol/liter, and the albumin:creatinine ratio was 1.64 mg/mmol. These values differ considerably from the results obtained in the subsequent 24-hour urine collections, especially with regard to the urinary creatinine concentration (8.7 mmol/liter; p < 0.001 for comparison with the spot morning sample) and the albumin:creatinine ratio (1.91 mg/mmol; p < 0.001), and less for the urinary albumin concentration (17.4 mg/liter; p = 0.03). This phenomenon is likely to be of importance. It is to be expected that values for the within-person coefficient of variation will be quite different when they are measured from spot morning urine samples instead of 24-hour urine collections. It is also to be expected that the urinary albumin concentration and especially the albumin:creatinine ratio cutoff values indicating microalbuminuria will be different when these variables are measured in a spot morning sample instead of a 24-hour urine collection. Of course, we fully acknowledge that this drawback also concerns our own data shown in table 1.
In conclusion, we propose that carefully designed prospective studies should be performed to clarify the issue of which urinary albumin measure to use in patient care and in epidemiologic studies. In such studies, urinary albumin and creatinine should be measured preferably in fresh urine. Another option may be measurement in urine that has been stored at –80°C. When interpreting studies on which urinary albumin measure to use, investigators should take into account the time dependency of the within-person coefficient of variation for the various albumin measures. Finally, portions of a 24-hour urine collection should not be used as a surrogate for an untimed random sample. Until such data are available and confirmed, the 24-hour urinary albumin excretion should in our opinion remain the gold standard.
|
ACKNOWLEDGMENTS |
|---|
Conflict of interest: none declared.
|
NOTES |
|---|
Editor's note: An invited commentary on this article appears on page 728, and the authors' response appears on page 731.
|
References |
|---|
|
TOPABSTRACTReferences |
|---|
- Hillege HL, Fidler V, Diercks GF, et al. Urinary albumin excretion predicts cardiovascular and noncardiovascular mortality in general population. Circulation 2002;106:1777–82.
[Abstract/Free Full Text] - Verhave JC, Gansevoort RT, Hillege HL, et al. An elevated urinary albumin excretion predicts de novo development of renal function impairment in the general population. Kidney Int Suppl 2004;(92):S18–21.
- Lane JT. Microalbuminuria as a marker of cardiovascular and renal risk in type 2 diabetes mellitus: a temporal perspective. Am J Physiol Renal Physiol 2004;286:F442–50.
[Abstract/Free Full Text] - Gerstein HC, Mann JF, Yi Q, et al. Albuminuria and risk of cardiovascular events, death, and heart failure in diabetic and nondiabetic individuals. JAMA 2001;286:421–6.
[Abstract/Free Full Text] - Dyer AR, Greenland P, Elliott P, et al. Evaluation of measures of urinary albumin excretion in epidemiologic studies. Am J Epidemiol 2004;160:1122–31.
[Abstract/Free Full Text] - Brinkman JW, de Zeeuw D, Duker JJ, et al. Falsely low urinary albumin concentrations after prolonged frozen storage of urine samples. Clin Chem 2005;51:2181–3.
[Free Full Text] - Schultz CJ, Dalton RN, Turner C, et al. Freezing method affects the concentration and variability of urine proteins and the interpretation of data on microalbuminuria. The Oxford Regional Prospective Study Group. Diabet Med 2000;17:7–14.[CrossRef][Web of Science][Medline]
- Elving LD, Bakkeren JA, Jansen MJ, et al. Screening for microalbuminuria in patients with diabetes mellitus: frozen storage of urine samples decreases their albumin content. Clin Chem 1989;35:308–10.
[Abstract/Free Full Text] - Osberg I, Chase HP, Garg SK, et al. Effects of storage time and temperature on measurement of small concentrations of albumin in urine. Clin Chem 1990;36:1428–30.
[Abstract/Free Full Text] - Innanen VT, Groom BM, de Campos FM. Microalbumin and freezing. Clin Chem 1997;43:1093–4.
[Free Full Text] - Asselbergs FW, Diercks GFH, Hillege HL, et al. Effects of fosinopril and pravastatin on cardiovascular events in subjects with microalbuminuria. Circulation 2004;110:2809–16.
[Abstract/Free Full Text] - Hansen HP, Hovind P, Jensen BR, et al. Diurnal variations of glomerular filtration rate and albuminuria in diabetic nephropathy. Kidney Int 2002;61:163–8.[CrossRef][Web of Science][Medline]
- Van Acker BA, Stroomer MK, Gosselink MA, et al. Urinary protein excretion in normal individuals: diurnal changes, influence of orthostasis and relationship to the renin-angiotensin system. Contrib Nephrol 1993;101:143–50.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
Related articles in Am. J. Epidemiol.:
- Invited Commentary: Evaluation of Measures of Urinary Albumin Excretion in Epidemiologic Studies
- Alan R. Dyer
Am. J. Epidemiol. 2006 164: 728-730.[Extract] [FREE Full Text] - Gansevoort et al. Respond to "Using Measures of Albumin Excretion"
- Ronald T. Gansevoort, Jacoline Brinkman, Stephan J. L. Bakker, Paul E. De Jong, and Dick de Zeeuw
Am. J. Epidemiol. 2006 164: 731-732.[Extract] [FREE Full Text]
This article has been cited by other articles:
|
|
 |
|
  E. L. Schopick, N. D. Fisher, J. Lin, J. P. Forman, and G. C. Curhan Post-menopausal hormone use and albuminuria Nephrol. Dial. Transplant., December 1, 2009; 24(12): 3739 - 3744. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. J. Lambers Heerspink, A. H. Brantsma, D. de Zeeuw, S. J. L. Bakker, P. E. de Jong, R. T. Gansevoort, and for the PREVEND Study Group Albuminuria Assessed From First-Morning-Void Urine Samples Versus 24-Hour Urine Collections as a Predictor of Cardiovascular Morbidity and Mortality Am. J. Epidemiol., October 15, 2008; 168(8): 897 - 905. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. M. Aakre, G. Thue, S. Subramaniam-Haavik, T. Bukve, H. Morris, M. Muller, M. V. Lovrencic, I. Plum, K. Kallion, A. Aab, et al. Postanalytical External Quality Assessment of Urine Albumin in Primary Health Care: An International Survey Clin. Chem., October 1, 2008; 54(10): 1630 - 1636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M S O'Neill, A V Diez-Roux, A H Auchincloss, T G Franklin, D R Jacobs Jr, B C Astor, J T Dvonch, and J Kaufman Airborne particulate matter exposure and urinary albumin excretion: the Multi-Ethnic Study of Atherosclerosis Occup. Environ. Med., August 1, 2008; 65(8): 534 - 540. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. T. Gansevoort, H. Lambers, and E. C. Witte Methodology of screening for albuminuria Nephrol. Dial. Transplant., August 1, 2007; 22(8): 2109 - 2111. [Full Text] [PDF]
|
|
|
|
|
|
R. T. Gansevoort, J. Brinkman, S. J. L. Bakker, P. E. De Jong, and D. de Zeeuw Gansevoort et al. Respond to "Using Measures of Albumin Excretion" Am. J. Epidemiol., October 15, 2006; 164(8): 731 - 732. [Full Text] [PDF]
|
|
|
|
|
|
A. R. Dyer Invited Commentary: Evaluation of Measures of Urinary Albumin Excretion in Epidemiologic Studies Am. J. Epidemiol., October 15, 2006; 164(8): 728 - 730. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||