• Open Access

Urinary Markers in Healthy Young and Aged Dogs and Dogs with Chronic Kidney Disease


  • Work was performed at the Department of Small Animal Medicine and Clinical Biology, the Department of Pharmacology, Toxicology and Biochemistry and the Department of Physiology and Biometrics, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium. Part of this work was presented as an abstract at the annual ECVIM-CA congress in Ghent, Belgium, September 4–8, 2008.

Corresponding author: Pascale M. Y. Smets, DVM, Department of Small Animal Medicine and Clinical Biology, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, 9820 Merelbeke, Belgium; e-mail: pascale.smets@Ugent.be.


Background: Blood urea nitrogen and serum creatinine concentrations only detect a decrease of >75% of renal functional mass. Therefore, there is a need for markers that allow early detection and localization of renal damage.

Hypothesis: Urinary albumin (uALB), C-reactive protein (uCRP), retinol binding protein (uRBP), and N-acetyl-β-d-glucosaminidase (uNAG) concentrations are increased in dogs with chronic kidney disease (CKD) compared with healthy controls and in healthy older dogs compared with young dogs.

Animals: Ten dogs with CKD, 10 healthy young dogs (age 1–3 years), and 10 healthy older dogs (age > 7 years) without clinically relevant abnormalities on physical examination, hematology, biochemistry, and urinalysis.

Methods: Urinary markers were determined using an ELISA (uALB, uCRP, and uRBP) or a colorimetric test (uNAG). Results were related to urinary creatinine (c). The fixed effects model or the Wilcoxon rank sum test were used to compare the different groups of dogs.

Results: uALB/c, uRBP/c, and uNAG/c were significantly higher in CKD dogs than in healthy dogs. No significant difference was found for uCRP, which was not detectable in the healthy dogs and only in 3 of the CKD dogs. Between the healthy young and older dogs, no significant difference was detected for any of the markers.

Conclusion: The urinary markers uALB/c, uRBP/c, and uNAG/c were significantly increased in dogs with CKD compared with healthy controls. Additional studies are needed to evaluate the ability of these markers to detect renal disease before the onset of azotemia.


blood urea nitrogen


chronic kidney disease


serum creatinine


urinary albumin-to-creatinine ratio


urinary C-reactive protein-to-creatinine ratio


urinary N-acetyl-β-d-glucosaminidase-to-creatinine ratio


urinary protein-to-creatinine ratio


urinary retinol binding protein-to-creatinine ratio

Chronic kidney disease (CKD) is an important cause of morbidity and mortality in dogs. The prevalence of CKD increases with age, with 15% of dogs over 10 years old being affected.1 Early diagnosis may allow therapeutic intervention that prevents further damage and progressive decline of renal function. However, only a decrease of >75% of renal functional mass will be detected by current diagnostic tests such as blood urea nitrogen (BUN) and serum creatinine (sCr) concentrations.2

Unlike these insensitive serum tests, urinary markers are sensitive indicators of renal injury and also have the potential to reflect the site and severity of damage.3 They include proteins categorized according to their molecular weight: high molecular weight (HMW), intermediate molecular weight (IMW), and low molecular weight (LMW) proteins.4 In general, renal pathologic proteinuria may be because of increased glomerular filtration or tubular dysfunction causing impaired reabsorption of normally filtered protein.2

Glomerular dysfunction leads to higher filtration of IMW proteins such as albumin and in more advanced stages to the presence of HMW proteins in the ultrafiltrate.4 A urinary albumin (uALB) concentration above normal, but below the limit of detection (LOD) of conventional dipstick analysis (ie, between 1 and 30 mg/dL) is defined as microalbuminuria. Several studies suggest that microalbuminuria may be an early indicator of glomerular disease in dogs.a,b An example of a HMW protein is C-reactive protein (CRP), an acute phase protein with an increased serum concentration in many inflammatory diseases.5 To the authors' knowledge, urinary CRP (uCRP) has only recently been evaluated in 1 study, reporting an increased CRP-to-creatinine ratio (uCRP/c) in dogs with renal damage secondary to pyometra.c

Tubular dysfunction is reflected by urinary loss of LMW proteins or urinary enzymes.6 Retinol binding protein (RBP) is a LMW protein associated with tubular impairment in humans and dogs with renal disease and in hyperthyroid cats.7–9 Excessive amounts of urinary enzymes appear in urine because leakage from damaged tubular cells.10 One such enzyme is N-acetyl-β-d-glucosaminidase (NAG), a lysosomal glycosidase present in proximal tubular cells. Urinary NAG (uNAG) has proven to be a useful tool in early detection of renal injury in various human diseases as reviewed by Skalova.11 In veterinary medicine, an increased uNAG-to-creatinine ratio (uNAG/c) has been observed in dogs in various stages of CKD,12,13 leishmaniasis,14 and pyometra13,15 and in dogs treated with nephrotoxic antibiotics,16 NSAIDS, or glucocorticoids.17

Most studies in dogs have focused on 1 urinary marker, whereas reports on the combined measurement of glomerular and tubular markers in healthy and CKD dogs are scarce. Data about urinary marker assay validation likewise are scarce, especially for NAG.18 Finally, little information exists on the urinary concentration of these markers in aging dogs.d Therefore, the aims of this study were 1st to validate a commercial canine albumin ELISA and NAG colorimetric test. Secondly, we wanted to compare 2 glomerular markers, uALB and uCRP, and 2 tubular markers, urinary RBP (uRBP) and uNAG between healthy and CKD dogs. A 3rd objective was to detect a possible age-related difference in these markers between healthy young and older dogs.

Material and Methods


The current study was performed at the Faculty of Veterinary Medicine, Ghent University, after approval by the Local Ethical Committee. All owners agreed to participate in the study and signed an informed consent form.

In this cross-sectional study, 20 staff- and student-owned healthy dogs were recruited and divided into 2 groups according to their age: a group of young dogs (age 1–3 years) and a group of healthy older dogs (≥7 years). Dogs were judged healthy based on history, physical examination, CBC, biochemistry profile, and urinalysis [negative bacterial culture and urine protein-to-creatinine ratio (UPC) < 0.5. Dogs were excluded when medication possibly influencing renal function had recently been administered. Descriptive statistics for both groups of healthy dogs are presented in Table 1.

Table 1.   Group descriptive statistics (median, range) for the healthy and chronic kidney disease (CKD) dogs.
VariableYoung DogsHealthy Dogs
Older Dogs
TotalCKD Dogs
  • **

    Significant difference (P < .0001).

  • M, male intact; F, female intact; MN, male neutered; FN, female neutered; sCr, serum creatinine; BUN, blood urea nitrogen; UPC, urine protein-to-creatinine ratio; USG, urine specific gravity.

Number of dogs10102010
Age (years)2.4 (1.1–2.9)**8.3 (7–10.9)**5 (1.1–10.9)4.7 (0.5–10.5)
Sex2 M, 5 F, 1 MN, 2 FN2 M, 4 F, 2 MN, 2 FN4 M, 9 F, 3 MN, 4 FN2 M, 4 F, 2 MN, 2 FN
Body weight (kg)23.8 (8.8–36.0)24.6 (8.7–39)24.5 (8.7–39)17.4 (4.2–40.1)
sCr (mg/dL)0.99 (0.69–1.24)0.92 (0.42–1.10)0.94 (0.42–1.24)**5.51 (1.37–15.94)**
BUN (mg/dL)33.99 (26.01–43.00)26.13 (18.98–43.00)29.52 (18.98–43.00)**362.52 (95.98–689.01)**
UPC0.08 (0.06–0.11)0.1 (0.06–0.47)0.08 (0.06–0.47)**1.9 (0.63–3.39)**
USG1.040 (1.008–1.050)1.030 (1.016–1.049)1.035 (1.008–1.050)**1.010 (1.008–1.017)**

Ten privately owned dogs with CKD of all ages, breeds, and both sexes were included. Diagnosis was based on clinical signs compatible with CKD (eg, polyuria, polydipsia, weight loss, inappetence, vomiting), laboratory findings such as anemia, azotemia, electrolyte disturbances, and urine specific gravity (USG) < 1.030. Dogs with CKD were classified according to the International Renal Interest Society (IRIS) into stages I to IV.19 Exclusion criteria were the presence of concurrent infectious, neoplastic, or endocrine diseases. Descriptive statistics for the CKD dogs are presented in Table 1.

Clinical signs were polyuria and polydipsia (8/10), lethargy (6/10), weight loss (5/10), decreased appetite or anorexia (5/10), vomiting (5/10) and diarrhea (2/10). Fecal examination of 1 dog with diarrhea indicated a Giardia and Strongyloides stercoralis infection. Two dogs had positive urine cultures (Escherichia coli). CKD dogs were IRIS stage I (n= 2), III (n= 2), or IV (n= 6) and all were proteinuric (UPC > 0.5). The 2 dogs in stage I had sCr concentrations of 1.39 and 1.37 mg/dL and UPC ratios of 2.33 and 3.06, respectively. Ultrasound findings were hyperechoic renal cortices and a renal cortical cyst of 3 mm diameter in 1 dog and decreased kidney size (3.8 cm), hyperechoic parenchyma, and decreased corticomedullary demarcation in the other dog. Juvenile renal disease was found as underlying cause for CKD in 4 dogs: a Leonberger, a Boxer, a Bull Mastiff, and a Schapendoes (ages 2.3, 0.7, 2.6, and 0.5 years, respectively). The Boxer had a littermate that had been euthanized because of end-stage renal disease and the Schapendoes was diagnosed with polycystic kidney disease.

Concurrent heart disease was present in 1 dog with aortic and pulmonary valve stenosis (Cairn Terrier) and in 2 dogs with mitral valve disease, International Small Animal Cardiac Health Council (ISACHC) class Ia (Maltese) and class II (Cavalier King Charles). Nine dogs were newly diagnosed and referred to our clinic for further diagnostic work-up. In 1 dog, the diagnosis of CKD had been made 13 months earlier but the owner had stopped feeding the renal diet and stopped the angiotensin-converting enzyme (ACE) inhibitor therapy for several months. Six dogs had not yet received any treatment and 4 dogs had already been treated with antiemetics, gastroprokinetics, antibiotics, or fluid therapy by their referring veterinarians. The dog with ISACHC class II mitral valve disease was the only dog on an ACE inhibitor (benazepril) for 12 months.

Sample Collection

Morning urine samples were collected by cystocentesis (10 mL, 22 G needle). Urinalysis consisting of dipstick analysis,e USG, UPC, sediment analysis, and bacterial culture, was performed. After centrifugation (3 minutes at 447 ×g), the supernatant was divided into aliquots of 0.5 mL and stored at −80°C until analysis of each marker after thawing on ice.

Laboratory Methods

Because spot urine samples were used, urinary concentration of each biomarker was related to urinary creatinine (c) and expressed as a ratio.16 Urinary creatinine concentration was measured by the modified Jaffé reaction using picric acid.20

Validation of the Albumin ELISA and NAG Colorimetric Assay

As an indication of assay precision, within-day coefficient of variation (CV) was calculated from 30 duplicate samples assayed on the same day and between-day CV from 15 samples assayed on separate days, respectively. In addition, between-day variation was assessed using Lin's sample concordance correlation coefficient (Lin's CCC, ρc), which indicates the degree of deviation from total agreement between 2 repetitive measurements of the same sample.21

Accuracy of the assays was assessed by evaluation of linearity under sequential dilution.22 Urinary concentration in diluted samples was plotted against dilution factor and linear regression analysis was performed.

Assay sensitivity was determined using the mean and standard deviation (SD) of blank samples to define the lowest concentration of albumin and NAG that can be reliably distinguished from a blank measurement. The LOD was calculated as mean blank result + 2.6 × SD.23

Albumin, CRP, and RBP ELISA

uALB and uCRP concentration were measured with commercial canine-specific sandwich ELISA kitsf and uRBP with a human ELISA kit,f previously validated in our laboratory for use with canine urine.24 A microtiter plate precoated with affinity-purified antibodies for canine albumin, CRP, or human RBP was used. Wells were filled with 100 μL of diluted urine samples or standards (albumin, 12.5–400 ng/mL; CRP, 3.125–200 ng/mL; RBP, 7.8–250 ng/mL) and incubated. After several wash steps, peroxidase-labeled albumin, CRP, or RBP antibody conjugate was added. After incubation and wash steps, each well was filled with 100 μL of tetramethylbenzidine substrate and incubated. Addition of sulfuric acid stopped the colorimetric reaction and absorbance was measured with an ELISA plate readerg at a wavelength of 450 nm with 650 nm as a reference. A 4 parameter logistic curve-fitting programh was used to generate the standard curve and calculate albumin, CRP, or RBP concentrations in the urine samples.

NAG Colorimetric Assay

uNAG activity was calculated with a colorimetric assay.i The NAG enzyme hydrolyzes the 4-nitrophenyl-NAG substrate and releases p-nitrophenol (PNP). Addition of the basic stop solution causes ionization of the PNP to the p-nitrophenylate ion. The absorbance of the latter ion was measured at 405 nm with a plate reader.g NAG activity was calculated using a standard formula and divided by urinary creatinine concentration to determine the uNAG/c (U/g) = NAG activity (U/L) / urinary creatinine (g/L).13,16

Statistical Analysis

Analyses were performed with a commercial software program.j For normally distributed data, the linear model with normally distributed random error term and dog group (healthy and CKD) and age category (young and older) as categorical fixed effects was fitted. An F-test was used to compare the 2 dog groups. When the normal distribution assumption did not hold, the Wilcoxon rank sum test stratified for age category was performed to compare the CKD dogs with the healthy dogs.

In a 2nd analysis, the young healthy dogs were compared with the old healthy dogs using the same analysis techniques as above, but only for the healthy dogs and without including age category as an adjusting covariate.

A global significance level of 5% was used to test, leading to a Bonferroni-adjusted significance level of 2.5% for each of the 2 comparisons above.

Correlations between different urinary markers and between the markers and other variables (sCr, BUN, UPC, and IRIS classification) were determined using Pearson's correlation coefficient for normally distributed data or Kendall's τ for the other data.


Validation of the Albumin and NAG Assays

Precision and sensitivity of the albumin ELISA and NAG colorimetric assay are presented in Table 2. Both assays had satisfactory variability with within-day CV<10% and between-day CV<15%. Moreover, a Lin's concordance correlation coefficient of >0.95 indicated substantial agreement between repetitive measurements of samples on separate days. Regression analysis after serial dilution of urine samples showed a linear relationship between albumin concentration or NAG activity and dilution factor (Fig 1).

Table 2.   Assay characteristics for canine albumin ELISA and NAG colorimetric assay.
Assay CharacteristicALBNAG
  1. ALB, albumin ELISA; NAG, N-acetyl-β-d-glucosaminidase colorimetric assay; LOD, limit of detection; CV, coefficient of variation; ρc, Lin's concordance correlation coefficient.

Sensitivity: LOD15.1 ng/mL0.84 U/L
 Within-day CV (%)5.24.9
 Between-day CV (%)12.08.1
Figure 1.

 Sequential dilution of a urine sample from a dog with CKD. (A) Urinary albumin concentration (ng/mL) plotted against dilution factor (× 10−6). The linear regression equation was y= 0.8812x− 0.3507 (r2= 0.9995). (B) Urinary NAG activity (U/L) plotted against dilution factor (× 10−1). The linear regression equation was y= 1.0296x+ 0.3107 (r2= 0.995). CKD, chronic kidney disease; NAG, N-acetyl-β-glucosaminidase.

Urinary Markers

Results for uALB/c, uRBP/c, and uNAG/c in healthy and CKD dogs are presented in Figure 2. There was a highly significant difference between healthy and CKD dogs for uALB/c, uRBP/c, and uNAG/c (P < .0001), but not for uCRP (P= .07). In the healthy dogs, median uALB/c was 7.3 mg/g (range, 1.5–296.5) and in the dogs with CKD uALB/c was 1868.6 (range, 12.7–4594.6). None of the healthy dogs had any detectable uCRP, whereas 3 of the CKD dogs did, with uCRP/c of 84.3, 179.8, and 236.9 μg/g. uRBP/c was 0.1 mg/g (range, 0–0.9) in the healthy dogs whereas dogs with CKD had a median ratio of 53.4 (range, 6.8–1372.2). The healthy dogs had a median uNAG/c of 2.2 U/g (range, 1.2–5.5) and the CKD dogs had a uNAG/c of 5.7 U/g (range, 1.5–9.5).

Figure 2.

 Scatter plot of uALB/c (mg/g) (A), uRBP/c (mg/g) (B), and uNAG/c (U/g) (C) in healthy and CKD dogs. All 3 markers were significantly different (P < .0001) between groups. uALB/c, urinary albumin-to-creatinine ratio; uRBP/c: urinary retinol binding protein-to-creatinine ratio; uNAG/c, urinary N-acetyl-β-glucosaminidase-to-creatinine ratio; CKD, chronic kidney disease.

When comparing the healthy young with the healthy older dogs, no significant differences were found for uALB/c, uCRP/c, uRBP/c, or uNAG/c (P > .025). uALB/c in the young dogs was 4.7 mg/g (range, 1.5–46.3) and in the older dogs 17.8 (range, 3.3–296.5). In the young dogs, median uRBP/c was 0.1 mg/g (range, 0–0.2) and in the older dogs also 0.1 (range, 0–0.9). Median uNAG/c was 2.5 U/g (range, 1.6–5.5) in the young dogs and in the older dogs it was 2 (range, 1.2–5.2).

Correlations between different urinary markers and between the markers and other variables are shown in Table 3. The highest correlations were found between uALB/c and UPC (r= 0.96, P < .0001), uALB/c and uRBP/c (r= 0.65, P < .0001), uRBP/c and UPC (r= 0.74, P < .0001), and uRBP/c and IRIS stage (r= 0.72, P < .0001). uRBP/c also showed moderate correlations to sCr (r= 0.51, P < .0001) and BUN (r= 0.56, P < .0001), and uALB/c to IRIS stage (r= 0.53, P < .025), but not to sCr and BUN. UCRP/c was weakly to moderately associated with sCr (r= 0.38, P < .025), IRIS stage (r= 0.52, P < .025), and BUN (r= 0.41, P < .025), but not with UPC (P > .025). UNAG/c was moderately correlated to IRIS stage (r= 0.49, P < .025), BUN (r= 0.42, P < .025), and UPC (r= 0.51, P < .025), but not to sCr.

Table 3.   Correlations (r) for the urinary markers, sCr, IRIS stage, BUN, and UPC.
  • *

    Significant correlation (P < .025).

  • **

    Significant correlation (P < .0001).

  • sCr, serum creatinine; BUN, blood urea nitrogen; IRIS, International Renal Interest Society; UPC, urine protein-to-creatinine ratio; uALB/c, urinary albumin-to-creatinine ratio; uCRP/c, urinary CRP-to-creatinine ratio; uRBP/c, urinary retinol binding protein-to-creatinine ratio; uNAG/c, urinary N-acetyl-β-d-glucosaminidase-to-creatinine ratio.

uALB/cr= 0.27r= 0.53*r= 0.35r= 0.96**r= 1r= 0.27r= 0.65**r= 0.56*
uCRP/cr= 0.38*r= 0.52*r= 0.41*r= 0.32r= 0.27r= 1r= 0.43*r=−0.02
uRBP/cr= 0.51**r= 0.72**r= 0.56**r= 0.74**r= 0.65**r= 0.43*r= 1r= 0.49*
uNAG/cr= 0.33r= 0.49*r= 0.42*r= 0.51*r= 0.56*r=−0.02r= 0.49*r= 1


Results of the present study indicate satisfactory assay characteristics of the quantitative canine albumin ELISA and NAG colorimetric assay when applied for evaluation of urine. Furthermore, uALB, uCRP, uRBP concentrations, and uNAG activity were found to be significantly higher in dogs with CKD compared with healthy controls. No age-related difference in urinary markers was detected between young and older healthy dogs.

The assays used in this study were able to measure canine uALB and uNAG in a linear manner with within- and between-run imprecision of 5.2 and 12%, and 4.9 and 8.1%, respectively. As a comparison, for uALB ELISA's, other studies have reported within- and between-run imprecision ranging between 2–17 and 4.5–21.7%, respectively.25,26 For the NAG colorimetric assay, within- and between-run CVs of 2.8 and 11.9% have been reported.27 In research settings, where samples frequently are analyzed in the same run, these levels of imprecision are considered acceptable, but for diagnostic purposes steps to decrease between-run variability should be taken.

In the current study, both quantitative as well as the qualitative aspects of proteinuria were evaluated by measuring several proteins as candidate urinary markers for glomerular and tubular damage. Because albuminuria is the earliest detectable form of proteinuria and all dogs with CKD had a UPC > 0.5, the strong correlation between uALB/c and UPC was expected.28 This strong correlation also was found in a study in cats with CKD, in which UPC also was predictive for survival.25

The use of microalbuminuria as a marker of early CKD is supported by several studies in dogs predisposed to glomerular disease because of a genetic cause or a heartworm infection.a,b In these dogs, microalbuminuria reflected the onset of disease more rapidly than the UPC. However, in the present study, most of the dogs had advanced renal disease and it was not the objective to prospectively evaluate uALB/c as an early marker of CKD.

This study is the first to describe uCRP concentrations in dogs with CKD. To the authors' knowledge, increased uCRP/c ratios only have been described in dogs with pyometra, and uCRP has never been evaluated in humans.c UCRP was not detected in the healthy dogs and increased uCRP/c ratios were present in 3 out of 10 CKD dogs. No increase in uCRP/c was detected in the 2 dogs with positive urine cultures or in the dog with a positive fecal examination. Although no statistically significant difference could be detected between CKD and control dogs, the presence of uCRP in dogs with CKD still is an interesting finding. For CRP to appear in urine, its plasma concentration must be increased and the glomerular barrier must be sufficiently damaged to allow HMW protein filtration. Thus, 1 possible hypothesis is that the increased uCRP/c in these 3 CKD dogs reflects an inflammatory response leading to increased plasma concentrations and subsequent leakage of CRP through the damaged glomerular barrier.

In humans, inflammation and oxidative stress start early in the process of failing kidney function and mild increases in CRP concentration are present even in patients with moderate renal impairment.29,30 Among other causes, decreased renal clearance of CRP, proinflammatory cytokines such as interleukin-6 or both, and uremia are suggested reasons for chronic inflammation in these human patients.31 This might also have been the cause in 1 dog with increased uCRP/c, which was in an advanced staged of CKD (IRIS stage IV) with severe clinical signs (eg, vomiting, hemorrhagic diarrhea). In humans, serum CRP concentration also is a predictor of cardiovascular disease, such as congestive heart failure.32 Interestingly, 2 of the dogs with increased uCRP/c were older dogs that also had mitral valve disease. One dog with aortic and pulmonary valve stenosis had a normal uCRP/c. To obtain better insight into the role of inflammation in canine CKD, plasma and urinary concentrations of CRP, and other proinflammatory mediators should be evaluated in a larger number of dogs.

uRBP was measured as a 1st marker of tubular dysfunction. In most mammals, this LMW protein circulates in plasma complexed with a 2nd protein (transthyretin) and binds vitamin A, which prevents RBP excretion. However, dogs have high concentrations of transthyretin-uncomplexed RBP that is filtered by the glomeruli. Under physiologic conditions, the filtered RBP is almost completely reabsorbed by megalin-mediated endocytosis in the proximal tubular cells, but tubular dysfunction leads to excessive amounts of uRBP.8,33 In our study, uRBP/c was significantly higher in dogs with CKD compared with healthy controls and highly correlated with sCr, BUN, UPC, and IRIS-stage, which corroborates results from previous studies documenting increased uRBP/c ratios in CKD dogs.8,34

As a 2nd marker for tubular dysfunction, uNAG was determined. Indeed, in humans increased uNAG/c in renal disease does not originate from filtration because glomerular damage but from tubular epithelial cells.35,36 This enzyme was found to be significantly higher in the CKD than in the healthy dogs, although there was large overlap in uNAG/c ratios between the 2 groups. In the present study, uNAG/c in CKD dogs (median, 5.5 U/g; range, 1.5–9.5 U/g) was lower than observed in 2 previous reports in 9 CKD dogs (mean ± SD, 17.1 ± 7.9 U/g) and 7 CKD dogs (median, 25.4 U/g; range, 15.7–136.8 U/g), respectively.12,13 Possible explanations for this difference include variation in laboratory techniques for NAG analysis and dog-related factors, such as stage of CKD. Enzymuria may be a marker of active disease damaging the renal cells. When the primary cause of the damage has disappeared, enzymuria might become minimal despite pronounced structural damage and permanent loss of tubular cells.35 To further investigate this hypothesis, additional studies are needed to determine uNAG/c in dogs with acute renal failure as well as in dogs with various stages of CKD.

Urinary tract infection (UTI) was an exclusion criterion for the healthy dogs, but 2 of the CKD dogs had positive urine cultures. The effect of lower UTI on microalbuminuria and uNAG/c seems to be minimal in dogs, but an increased uNAG/c ratio is reported in the presence of pyelonephritis.13,37 In these 2 dogs, no ultrasonographic signs of pyelonephritis were detected. To the authors' knowledge, the effect of UTI on uRBP and uCRP is unknown in dogs. In humans, microalbuminuria, and uCRP, uRBP, and uNAG concentrations are increased in patients with upper UTI, but not in patients with cystitis or asymptomatic bacteriuria.38–41

The degree of correlation between the urinary markers and routine variables such as sCr, BUN, and UPC differed for each urinary marker. This is not unexpected, because these markers might be more sensitive indicators of renal damage and some may reflect renal dysfunction at another site rather than serving as an indirect glomerular marker as does sCr. Different markers each may reflect different pathophysiologic processes. UALB/c was not correlated with sCr and BUN whereas uRBP/c was strongly correlated with both. Progression of CKD in humans and dogs, respectively, may be initiated by glomerular changes subsequently leading to interstitial injury.4,42,43 It is mainly the tubulointerstitial inflammatory process that precedes renal scarring and is associated with a decreased renal function. In this more advanced stage, large amounts of LMW proteins such as RBP appear in urine. Negatively charged IMW proteins such as albumin are a sign of mildly altered glomerular permeability and already appear in urine in an early stage. Although uCRP/c was moderately to weakly correlated to sCr, BUN, and UPC, this finding needs to be interpreted with caution because only 3 dogs had positive results. As in the present study, a report on urinary enzymes in humans with glomerulonephritis also demonstrated a correlation between uNAG/c and proteinuria and the absence of an association between uNAG/c and sCr.44 The current hypothesis is that urinary enzymes correlate better with tubulointerstitial damage than with an indirect measurement of glomerular filtration rate such as sCr.

In the healthy dogs, a median uALB/c of 7.3 mg/g (range, 1.5–296.5) was measured. Cut-off values for uALB/c ratios remain to be properly established in a large number of healthy dogs. Usually, a normal range is calculated as mean uALB/c + 2 SD of the healthy control group.45 In 2 previous reports, uALB/c either ranged from 5 to 82 mg/g (n= 6) or mean uALB/c ± SD was 40 ± 80 mg/g (n= 10).45,k Three of 20 healthy dogs in the current study had uALB/c ratios above the upper limit of 42 mg/g (mean + 2 SD of the healthy young dogs). Interestingly, 2 of these 3 dogs were older dogs with uALB/c ratios well above the cut-off (152.6 and 296.5 mg/g) and another 2 older dogs had uALB/c ratios near the cut-off (40.4 and 41.5 mg/g). Previous data indicated a higher prevalence of microalbuminuria with increasing age.d This finding might be related to the higher prevalence of glomerular lesions and CKD in aging dogs.1,46 However, in the present study no renal biopsies were taken because ethical considerations. As for the CKD dogs, knowledge of the histopathologic lesions would unlikely have therapeutic or prognostic implications in advanced disease stages.

uNAG/c ratios in the healthy dogs (median, 2.2 U/g; range, 1.2–5.5 U/g) were in agreement with previous reports.13,27,47 The present study is the first to report uNAG/c ratios in older healthy dogs. No significant difference was found between the young and older healthy dogs. Nevertheless, the true effect of age is best assessed in a longitudinal follow-up of the same individuals. In humans, uNAG is higher in children than in adults.48 This is mainly because changes in muscle mass and consequently in creatinine excretion.49 Therefore, in human medicine each laboratory establishes its own reference ranges for different age categories and patients then are compared against age-matched controls.48,50

In conclusion, combined use of glomerular and tubular markers in conjunction with traditional tests may provide more detailed information about the extent and location of renal damage. The urinary markers discussed in this study appear to be promising as noninvasive tools for the diagnosis of CKD in dogs. Age-related differences seem to be minimal between young adult and older healthy dogs. These results should encourage in-depth investigation of urinary markers in larger numbers of dogs classified according to their etiology and stage of renal disease.


a Grauer GF, Oberhauser EB, Basaraba RJ, et al. Development of microalbuminuria in dogs with heartworm disease. J Vet Intern Med 2002;16:352 [abstract]

b Lees GE, Jensen WA, Simpson DF, Kashtan CE. Persistent albuminuria precedes the onset of overt proteinuria in male dogs with X-linked heriditary nephropaty. J Vet Intern Med 2002;16:353 [abstract]

c Maddens BEJ, Daminet S, Smets P, et al. Urinary immunoglobulin G, C-reactive protein and retinol-binding protein as candidate early biomarkers for renal dysfunction in dogs with pyometra. J Vet Intern Med 2008;22:1473 [abstract]

d Radecki S, Donnelly R, Jensen WA, Stinchcomb DT. Effect of age and breed on the prevalence of microalbuminuria in dogs. J Vet Intern Med 2003;17:406 [abstract]

e Combur stick, Roche Diagnostics, Mannheim, Germany

f Immunology Consultants Laboratory, Newberg, OR

g Multiskan MS, Labsystems Thermo Fisher Scientific, Waltham, MA

h Deltasoft JV, BioMetallics Incorporated, Princeton, NJ

i Sigma-Aldrich, St Louis, MO

j SAS version 9.1, SAS Institute. Inc, Cary, NC

k Mazzi A, Fracassi F, Gentilini F. Urinary protein to creatinine ratio and albumin to creatinine ratio in dogs with diabetes mellitus and pituitary dependent hyperadrenocorticism. Congress Proceedings of the 16th ECVIM-CA Congress, Amsterdam, the Netherlands, September 14–16, 2006 [abstract]


This study was funded by a grant from the “Fonds voor Wetenschappelijk Onderzoek (FWO) Vlaanderen.” The authors thank Kristel Demeyere for her technical assistance.