Download Albuminuria vs Urinary Total Protein for

CLIN. CHEM. 37/5, 621-624 (1991)
Albuminuria vs Urinary Total Protein for Detecting Chronic Renal Disorders
Z. K. Shihabi,’
J. C. Konen,2 and M. L. O’Connor’
A literature review and our own data are presented to
demonstrate
that urinary albumin (UA) excretion in-
creases in many renal disorders and that it offers a far
more sensitive indicator than the commonly used urinary
total protein (UTP) for the early detection of renal involvement in many chronic diseases such as diabetes mellitus,
hypertension, and systemic lupus erythematosus. In
many individuals with these disorders, UA increases
severalfold, even while UTP remains within the reference
interval. UA is also more suited than UTP for following
therapeutic responses in these slowly progressive renal
disorders. Increases in UA are associated with increased
mortality. UTP measurements are plagued with many
analytical problems, whereas UA is much easier to standardize. We recommend that both UA and UTP be mea-
sured when quantitative unne protein assays are ordered,
especially when the UTP is <300 mg/g of creatinine.
AddItionalKeyphrasea:diabetes mellitus
systemiclupuseiythematosus
.
hypertension
pre-eclampsia
Urinary total proteins (UTP) are commonly assayed
for diagnosing and monitoring
various primary
and
secondary nephropathies. In many systemic and metabolic disorders, e.g., hypertension,
diabetes mellitus,
and systemic lupus erythematosus, the kidney may
undergo slow but progressive deterioration,
leading
eventually to renal failure. Renal involvement in these
disorders is often first manifested by a gradual increase
in proteinuria, for which sensitive assays are therefore
desirable. Highly sensitive assays specific for urinary
albumin (UA) have also been used, but most attention
has focused on the importance of UA in diabetes. We
will illustrate, using published data as well as our own
experience, that the UA assay is more specific and
sensitive than the commonly used UTP assay for early
detection and subsequent monitoring of chronic renal
disorders in general. In this context, we propose the
routine addition of UA assays to UTP assays even when
UTP is within its reference interval.
ClInIcal Significance
Urinary proteins consist of a variable mixture of an
ultrafiltrate
of serum plus some proteins produced by
the urogenital tract. The UTP reference interval in our
Departments of’ Pathology and2 Family and Community Medicine, Bowman Gray School of Medicine, Wake Forest University,
Winston-Salem, NC 27103.
Received September 19, 1990; accepted February 19, 1991.
laboratory is 0-200 mg/day, or 0-200 mg/g of creatinine.
The major single protein detectable by electrophoresis of
urine of normal individuals is albumin. The central 95%
reference interval for UA in our laboratory is 0.4-2.8
g/mol of creatinine (5-32 mg/g of creatinine). This
reference interval for UA is based on assays of UA in
random (untimed) daytime urine specimens from 80
healthy adults (1) and is in close agreement with the
reference interval of 0.2-2.8 glmol of creatinine reperted
by Silver et al. (2). UA excretion has been also reported
as mg/24 h or g/min. Many workers use 30 mg/24 h as
the upper limit of the UA reference interval (3-6),
although different reference intervals or cutoff values
may be set for special studies. Even with a fivefold
increase in UA, the UTP values may remain <200 mg/g
of creatinine (i.e., within the UTP reference interval)
and be interpreted, falsely, as clinically normal. The
most common proteinuria
screen is the “dipstick,” with
a sensitivity
for urinary protein (mainly albumin) of
150-300 mgfL, which is suitable only for detecting
relatively gross albuminuria.
In 1982 Viberti et al. (7), using a radioimmunoassay,
found that an increased rate of albumin excretion could
predict the onset of nephropathy in insulin-dependent
diabetics, even when the UTP test was within the
reference interval. This observation has been confirmed
by several other groups (4,8-10). Viberti eta!. (7) coined
the term “microalbuminuria”
to refer to the measurement of an increased UA with UTP within the reference
interval or with a negative dipstick result. Microalbuminuria has been defined also as albumin excretion of
30-300 mg/24 h (4, 5). This term has become very
common in the literature and, unfortunately, is likely to
remain with us even though, in traditional protein
nomenclature, the prefix “micro” is used to indicate a
small-molecular-mass
protein such as 32-microglobulin.
In diabetes, increased UA excretion is associatedwith
renal glomerular hyperfiltration; it occurs early in the
course of the disease (11), before renal histological
or
structural
changes are detectable (12, 13). Among the
Pima Indians, an inbred tribe with a very high prevalence of diabetes, increased UA excretion can be detected in some individuals before the onset of hyperglycemia (5). This finding is important because, in the early
stages of diabetes, increased UA and renal hyperfiltration may be reversed by rigorous treatment (8, 14).
Early detection and intensified treatment may thus be
able to delay, if not completely prevent, the progression
of proteinuria and renal failure (8).
CLINICAL CHEMISTRY, Vol. 37, No. 5, 1991 621
Other studies have shown that increased UA excretion, while UTP remains within its reference interval,
occurs in many other slowly progressive disorders affecting the kidney. About one-fourth of hypertensive
patients have increased concentrations
of UA (15, 16)
(Tables 1 and 2) that decrease in response to medications (17). Many patients with systemic lupus erythematosus show increased albumin excretion before developing structural renal changes (18); again, the albumin
excretion decreases in response to cortisol treatment
(18). Increased UA excretion occurs in pre-eclampsia
(19). We have found increased UA in two patients with
rheumatoid
arthritis.
About one-third of type II diabetics exhibit increased
UA, which correlates with glycemia, blood pressure, and
duration of illness (20). However, UA in type II diabetes
correlates best with cardiovascular
disease (21), a major
cause of death in this group of patients (22). Even in
nondiabetic patients, albuminuria may signify an increased risk of cardiovascular
disease (23). We have
been studying a group of patients older than 65 years as
part of a special cardiovascular
risk study; many of
these individuals showed increased IJA excretion (Tables 1 and 2). Although UA increases have been detected in some febrile patients with immune complexes
(9,24), the significance of UA in these disorders has not
been adequately investigated.
Increased UA occursin response to several physiological factors: catecholaniines
and growth hormone (11,
25), increased systemic and transglomerular
pressure
(25), changes in blood rheology (26), and changes in
charge and pore size of glomerular membranes (11,27).
According to the Steno hypothesis (27), UA excretion in
diabetes reflects widespread vascular damage, leading
to increased transcapillary
escape of proteins.
Although albuminuria
is generally considered a gbmerular proteinuria,
this is not always the case. About
90% of the filtered albumin is reabsorbed by the tubular
cells (28). Thus, in tubular proteinuria, a modest UTP as
great as 500 mg/24 h can represent a 15-fold increase in
UA (29). Tubular proteinuria
is also often combined
with glomerular proteinuria. The effect of tubular proteinuria on UA has not been studied adequately with
the new sensitive UA immunoassay.
We have seen a
nearly eightfold
increase in UA above the reference
interval in six hospital patients taking gentamicin
or
tobramycin,
drugs known to cause tubular proteinuria.
Tubular proteinuria is, however, detected more specifically by increases in urine excretion of low-molecular-
Table 1. Percentage of Patlents wIth Increased
Albuminurla (>32 mglg of Creatlnlne) but Normal UTP
Type I diabetics
Type Ii diabetics
Hypertensives
n
81
271
131
% above
normal
11.3
19.3
18.9
afold
Increasea
1.6
7.2
2.2
4.4
294
8.6
Seen at our family practice clinic. b Mean of the above-normal values/
upper reference range.
Geriatrics
a
622 CLINICALCHEMISTRY,Vol.37, No.5, 1991
Table 2. Percentage of Total Patients with
Above-Normal Valuesa
a-fold Incrs.seb
% above normai
Group
Type I diabetic
Type Ii diabetic
Hypertensive
Geriatric
Bence
Jones
Myoglobinuna
n
81
271
131
294
11
12
Aibumin
LJTP
Dipstick
39.5
33.2
25.2
13.9
100.0
92.6
34.6
23.7
8.7
22.5
6.9
14.6
100.0
92.6
Albumin
9.7
15.6
6.2
2.9
2.3
5.5
-
3.2
-
7.3
-
UTP
38.1
15.1
1.6
13.3
22.7
UT!’ >200 m/g of creatinine;albumin >32 mg/g of creatinine, or dipstick
positive b Meanof the above-normalvalues/upper reference range.
a
mass proteins and someenzymes, e.g., 32-microglobulin,
retinol-binding protein, N-acetyl-/3-glucoseaminidase,
and lysozyme (30).
A modest increase in albumin excretion also occursin
overflow proteinuria, e.g., Bence Jones proteinuria,
myoglobinuria, and hemoglobinuria (Table 2). For
proper diagnosis of these disorders, specific protein assays (immunoassays and electrophoresis) are more suitable. A greater increase in UTP than in UA suggestsan
overflow proteinuria.
Benign proteinuria
(mainly albuminuria) will be detected by either UA or UTP assays.
Acute renal failure is usually associated with a rapid
onset of clinical symptoms and a greater increase of
UTP, with albumin the major protein. In these cases,
UA assays yield the same information as UTP assays.
As renal disease progresses in diabetes and hypertension, the UA/IJTP fraction is quite variable but generally increases. For example, when UTP is less than 100
mg/g of creatinine, UA.ItJTP is about 20%. As UTP
approaches 200 mg/g of creatinine, the UA!UTP
increases to about 50%. With some exceptions, when the
UTP exceeds 1000 mg/g of creatinine, UA/UTP approaches 80% (1). Thus, a UTP of 1000 mgfL may
represent only a fivefold increase over the reference
range for IJTP but a 25-fold increase for UA. For this
reason, UA will be increased more than UTP and also in
a greater percentage of the patients, especially when
IJTP is only marginally increased (Table 2). The UA/
UTP ratio may prove useful for evaluating renal status.
In diabetics, the filtration of some proteins such as
transferrin
(31) and free sclight chains (32,33) increases
more than that of albumin because of their higher
isoelectric point (31). Measurement of urinary excretion
of these proteins has, as yet, no demonstrated advantage
over UA determination. Compared with IJA, these proteins are more difficult and thus less convenient to
assay. Furthermore,
our preliminary data show that,
whereas albumin excretion is increased in essential
hypertension in humans, transferrin excretion is not.
AnalytIcal Aspects
The assay of UTP is plagued by several major problems: many interfering substances, poor reproducibility
at low ranges, and different sensitivities to different
proteins by various assay techniques. Many colorimetric
are used to assay UTP in addition to several
common semiquantitative screening methods, e.g., dipsticks, precipitation, and heat and acid tests. The lack of
sensitivity of the dipsticks for UTP is evident from the
data in Table 2. A recent survey of the College of
American
Pathologists showed a CV of 21.5% among
the means of various methods used to assay a sample
with an above-normal concentration of protein (mean
1434 mgfL). The CV among laboratories for one method
for that sample was as high as 33.8%. Most imprecision
has to do with standardization of the test. A similar
survey for urine protein in the United Kingdom also
showed poor performance among the participant laboratories (34). The UTP assay is considered one of the
worst analyses performed by the laboratories (34).
These problems diminish the clinical significance of
UTP values obtained for samples with borderline increases and have led to proposals to replace UTP by UA
(34, 35).
Analytically,
it is easier to measure one well-defined
protein than a complex mixture of several proteins. For
UA determinations,
the concentration
of albumin standards can be assigned by mass or by molar absorptivity
at 280 nm (5). For many years, UA has been estimated
by electrophoretic analysis after concentration of the
sample. However, this technique is not sensitive enough
to detect the low concentrations present in normal
subjects and in the early stagesof diabetes. Radioimmunoassays have been used to measure low concentrations
of UA (2, 8). As the importance of measuring these low
concentrations has become established, simpler UA immunoassays have been introduced. Nephebometric (36),
turbidimetric
(1, 37, 38), and latex particle (39) imniunoassays have been described for UA determination,
involving the use of polyethylene and latex particles to
enhance the sensitivity and the speed of the reaction.
Radial immunodiffusion (1), fluoroimmunoassay (2,40),
and enzyme immunoassay (41) methods have also been
described for determining
UA. The advantages and
disadvantages of these various methods have been discussed (5, 42).
UA assays are adaptable to the automated instruments used in the routine laboratory with a reagent cost
per test for UA similar to that for UTP. Some UA
immunoassays
are, however, subject to problems from
the prozone effect in the presence of antigen excess.
Because the UA concentration can range over several
orders of magnitude, special care should be taken to
select methods that can detect antigen excess. Specific
colorimetric and fluorometric assayswould be desirable.
In the near future, automated capillary electrophoresis
may prove suitable for UA measurement;
it has the
sensitivity
to measure low values and can incorporate
methods
internal
standards.
UA excretion is quite variable in the individual, with
reported CVs ranging from 45% (3, 9) to 100% (5).
Random (untimed) urine samples have more variability
than do 24-h collections, but indexing the UA of untimed samples to creatinine excretion reduces some of
the variability. Because of this increased variability, the
reference interval for untimed samples is slightly wider
than that based on 24-h collections (5). From a practical
point of view, untimed samples are more convenient for
the patient and thus are recommended by someworkers
(5); some workers prefer a morning sample (43). Exercise (4, 9, 31), posture (44), and diuresis (45) affect UA.
The effects of diet (e.g., salt or high protein) on UA have
not been studied adequately. Because UA changes over
a very wide range (up to 100-fold), a slight increase in an
untimed sample might not be significant. Given the
variability in UA excretion, some workers use two to
three samples (3, 9) to verify above-normal values. For
slowly progressive disorders, the change in UA over
time may be more important than an isolated single
value (9).
Conclusion
UA is a more specific and sensitive indicator of a wide
variety of renal disorders than is UTP. In the Framingham study (46), individuals with persistent proteinuria
(mainly UA) and increased risk for cardiovascular disease had a higher mortality
rate. A recent editorial in
Diabetic Medicine (47) stated that measurements of UA
and glycohemoglobin were the two major advances of
the 1980s in care for diabetes patients. Because of the
lack of studies of UA as a marker of renal disorders
other than in diabetes, UA has not been widely used in
general clinical medicine. Yet, in our experience, about
20% of the hospital patients’ samples requested for UTP
that yield values between 100 and 300 mg/g of creatinine will have about a threefold increase above the
upper limit of the reference interval for UA.
Answers to several practical and basic questions also
must precede the widespread use of this test. What is
the best sample to collect? How long can samples be
stored?What medication and dietary and physiological
factors affect the excretion of albumin? What is the basic
mechanism for increased excretion in different disorders? Additionally, it is important to investigate the
various renal disorders, including tubular and acute
renal disorders, that might cause increases in UA and to
assess the clinical value of this information for patient
treatment. These questions should not be discouraging,
but rather should be seen as an exciting new opportunity for the clinician, pathologist, and clinical chemist to
pursue knowledge to improve patient care.
Thus, we advocate the addition of a UA assay to the
measurement
of UTP for patients presenting with normal UTP or a borderline increase in UTP (100-300 mg/g
of creatinine) and for patients who might be at risk of
developing chronic renal disease. This practice will
detect these disorderswith better specificity and greater
sensitivity. Because UA increases in a wide variety of
renal disorders, UA concentrations can be used to follow
the course of disease and signal the necessity for early
treatment when the diseasemight be reversible. Results
must be interpreted in light of the patient’s history.
However, UA is most useful in following the course of
slowly progressive disorders. The ratio of UA/UTP adds
CLINICALCHEMISTRY,Vol.37, No.5, 1991 623
for patient diagnosis. The addition of
UA testing to samples with normal UTP values would
augment the clinical significance of UTP and enhance
the detection of chronic renal disorders.
extra information
We thank Drs. R. W. Prichard,
S. Iskandar, R. Appel, and R.
Bloomfield for reviewing this manuscript before submission. This
study was supported in part by a grant from the Centers for
Disease Control (U32CCU-403318).
References
1. McCormick
and
CP, Konen JC, Shihabi ZK.
albuminuria:
enhanced immunoassay.
Microtransferrinuria
Ann Clin Lab Sci
1989;19:444-51.
2. Silver A, Dawnay A, Landon J, Cattell WR. Inimunoassays
for
low concentrations
of albumin in urine. Clin Chem 1986;32:
1303-6.
3. Watts GF, Morris RW, Khan K, Polak A. Urinary albumin
excretion in healthy adult subjects: reference values and some
factors affecting
their interpretation.
Clin Chim Acta 1988;
172:191-8.
4. Mogensen CK, Schmitt 0. The diabetic kidney: from hyperfiltration and microalbuminuria to end-stage renal failure [Review].
Med Clm North Am 1988;72:1465-92.
5. Rowe DJF, Dawnay A, Watts GF. Microalbuniinuria
in diabetes mellitus: review and recommendations
for measurement of
albumin in urine. Ann Clin Biochem 1990;27:297-312.
6. Nelson RG, Bennett PH. Diabetic renal disease in Pima mdians. Transplant Proc 1989;21:3913-5.
7. Viberti CC, Hill RI), Jarrett R.J, et al. Microalbuminuria
as a
predictor of clinical nephropathy in insulin-dependent diabetes
mellitus. Lancet 1982;i:1430-2.
8. Viberti GC. Recent advances in understanding mechanisms
and natural history of diabetic disease [Review]. Diabetes Care
1988;11(suppl 1):3-9.
9. Mogensen CE. Microalbuminuria
as a predictor of clinical
diabetic nephropathy [Nephrology Forum]. Kidney mt 1987;31:
673-89.
10. Parving
HI!, Hommel E, Mathiesen E, et al. Prevalence of
microalbuminuria,
arterial hypertension, retinopathy, neuropathy
in patients
with
insulin-dependent diabetes. Br Med J
1988;296:156-60.
11. Viberti CC. Early functional and morphological changes in
diabetic nephropathy [Editorial]. Clin Nephrol 1979;12:47-53.
12. Chavers BM, Bilous RW, Ellis EN, et al. Glomerular lesions
and urinary albumin excretion in type I diabetes without overt
proteinuria.
N EngI J Med 320;1989:966-70.
13. L.orentz WB Jr, Shihabi ZK, Widner N. Galactose nephropathy in the rat. Clin Physiol Biochem 1987;5:261-7.
14. Weiseman MJ, Saunders AJ, Keen H, et al. Effect of blood
glucose control on increased glomerular filtration rate and kidney
size in insulin-dependent
diabetes. N EngI J Med 1985;312:617-
21.
HH, Mogensen CE, Jensen HAE, et al. Increased
albumin excretion in benign essential hypertension. Lan-
15. Paving
urinary
21. Jarrett RJ, Viberti GC, Argyropoulos A, et al. Microalbuminin non-insulin dependent diabetes. Diabetic Med 1984;1:17-9.
22. Nosadini R, Vigorito C, Crepaldi G. Diabetic cardiopathy. In:
Andereani D, Crepaldi G, Mario UD, Pozza G, eds. Diabetic
complications: early diagnosis and treatment. New York: John
Wiley and Sons, 1987:37-42.
23. Haffner SM, Stern MP, Gruber KK, et al. Microalbuminuria.
Potential marker for increased cardiovascular risk in nondiabetic
uria predicts mortality
subjects? Arteriosclerosis 1990;10:727-31.
24. Solling J, Solling K, Mogensen CE. Patterns of proteinuria
and circulating immune complexes in febrile patients. Acts Med
Scand 1982;212:167-70.
25. Viberti GC, Keen H, The pattern of proteinuria in diabetes
mellitus. Relevance to pathogenesis and prevention of diabetic
nephropathy
[Review]. Diabetes 1984;33:686-92.
26. Bran JF, Criqui C, Carttry E, Fedou C, Osertti A. Microalbuminuria in diabetics. Possible involvement of hemorheologic factors in baseline excretion but not in exercise-induced rise. Clin
Hemorheol 1989;9:351-9.
27. Deckert T, Feldt-Rasmussen
B, Borch-,Jenson K, et al. Albuminuria reflects widespread vascular damage. The Steno hypothesis. Dibetologia 1989;32:219-26.
28. Bourdeau JE, Carone FA. Protein handling by the renal
tubule. Nephron 1974;13:22-34.
29. Abram CK. Diabetic proteinuria
[Editorial]. Am J Nephrol
1984;4:37-46.
30. Walker Ky, Ward KM, Mahan JD, Wismatt DK. Current
concepts in proteinuria [Review]. Clin Chem 1989;35:755-65.
31. McCormick CP, Konen JC, Shihabi ZK. Microtransferrinuria
and albuminuria: I. In the diabetic human. Clin Physiol Biochem
1990;8:53-8.
32. Dyer R, Shihabi Z, Konen J. Increased kappa chain excretion
in diabetes [Abstract]. Clin Chem 1990;36:1168.
33. Groop L, Makiperna A, Stemnan S, et al. Urinary excretion of
kappa light chains in patients with diabetes mellitus. Kidney mt
1990;37:1120-5.
34. Chambers RE, Bullock DG, Wincher JT. Urinary total protein
estimation-fact
or fiction? Nephron 1989;53:33-6.
35. Barratt M. Proteinuria. Br Med J 1983;287:1489-90.
36. Marre M, Claudel J, Ciret P. et al. Laser immunonephelometry of urinary albumin excretion. Clin Chem 1987;33:209-13.
37. Harmoinen A, Vuorinen P, Jokela H. Turbidixnetric measurement of microalbuminuria. Clin Chim Acts 1987;166:85-9.
38. Bakker AJ. Immunoturbidity of urinary albumin: prevention
of adsorption of albumin; influence of other urinary constituents.
Clin Chem 1988;34:82-6.
39. Cambiaso CL, Collet-Cassart
D, Lievens M. Immunoassay of
low concentrations of albumin in urine by latex particle counting.
Clin Chem 1988;34:416-8.
40 Chavers BM, Simonson J, Micheal AF. A solid phase fluorescent immunoassay
for the measurement of human urinary albumin [Tech Note]. Kidney Int 1984;25:576-8.
41. Niwa T, Katsuzaki T, Tatemicichi N, et al. Enzyme immunoassay for urinary albumin at low concentration in diabetes
mellitus. Clin Chim Acts 1990;186:391-6.
42. Watts GF, Bennett JE, Rowe DJ, et al. Assessment of immu-
cet 1974;i:1190-2.
16. Giaconi S, Levanti C, Fommei E, et al. Microalbuminuria and
casual ambulatory
blood pressure monitoring in normotensives
and in patients with borderline and mild essential hypertension.
Am J Hypertens 1989;2:259.-61.
17. Pedersen EB, Mogensen CE. Effect of antihypertensive treatment on urinary albumin excretion, glomerular filtration rate and
renal plasma flow in patients with essential hypertension. Scand J
Clin Lab Invest 1976;36:231-7.
18. Terai C, Nojima Y, Takano K, et al. Determination
of urinary
albumin excretion by radioimmunoassay
in patients with subclinical lupus nephritis. Clin Nephrol 1987;27:79-83.
19. Pedersen EB, Rasmussen AB, Johannesen
F, et al. Urinary
excretion of albumin, beta-2-microglobulin
and light chains in
pre-eclampsia, essential hypertension in pregnancy and normotensive pregnant control
subjects. Scand J Clin Lab Invest
Med 1985;2:451-5.
44. Robinson RR, Glenn WG. Fixed and reproducible orthostatic
proteinuria.
IV. Urinary albumin excretion in healthy human
subjects in recumbent and upright postures. J Lab Clin Med
1964;64717-23.
45. Viberti GC, Mogensen CE, Keen H, et al. Urinary excretion of
albumin in normal man: the effect of water loading. Scand J Clin
Lab Invest 1982;42:147-51.
46. Kannel WB, Stampfer MJ, Castelli WP, Verter J. The prognostic significance of proteinuria: the Framingham study. Am
1981;41:777-84.
Heart J 1984;108:1347-52.
20. Konen JC, Curtis LG, Shihabi ZK, Dignan MB. Screening
diabetics for microalbuminuria. J Farn Pract 1990;31:505-10.
47. Homes P. Diabetes, the 1980’s and beyond [EditonalJ. Diabetic
Med 1990;7:97-8.
624
CLINICAL CHEMISTRY, Vol. 37, No. 5, 1991
nochemical methods for determining low concentrations of albumin in urine. Clin Chem 1986;32:1544-8.
43. Gatling W, Knight C, Hill RD. Screening for early diabetic
nephropathy: which sample to detect microalbuminuria? Diabetic