Skip Navigation



NDT Advance Access published online on February 15, 2008
Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfm949
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/7/2242    most recent
gfm949v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Antunes, V. V. H.
Right arrow Articles by Morales, J. V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Antunes, V. V. H.
Right arrow Articles by Morales, J. V.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author [2008]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org


Diagnostic accuracy of the protein/creatinine ratio in urine samples to estimate 24-h proteinuria in patients with primary glomerulopathies: a longitudinal study

Verônica Verleine Hörbe Antunes, Francisco José Veríssimo Veronese and José Vanildo Morales

Post Graduate Program in Medical Sciences: Nephrology, School of Medicine, Universidade Federal do Rio Grande do Sul and Division of Nephrology, Hospital de Clínicas de Porto Alegre (HCPA), Porto Alegre, RS, Brazil

Correspondence and offprint requests to: José Vanildo Morales, Division of Nephrology, Hospital de Clínicas de Porto Alegre, Rua Ramiro Barcelos, n° 2350, sala 2030, CEP 90035-003, Porto Alegre, RS, Brazil. Tel: +55-51-2101-8295; Fax: +55-51-2101-8121; E-mail: jmorales{at}hcpa.ufrgs.br



   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 References
 
Background and objectives. The protein/creatinine (P/C) ratio in urine samples has been used in the clinical management of patients with glomerular diseases. The aim of this study is to perform a prospective evaluation of the P/C ratio accuracy in determining critical levels of proteinuria in patients with glomerulopathies.

Design, setting, participants and measurements. This is a longitudinal study of 41 adult patients with primary glomerulopathies treated with immunosuppressive drugs or angiotensin-converting enzyme inhibitors in a 6-month follow-up. Correlation and agreement level between P24 and the P/C ratio were evaluated. Kappa statistic was employed to evaluate concordance between the two methods taking into account clinically relevant categories of proteinuria. ANOVA for repeated measures was employed. Diagnostic accuracy of the P/C ratio was evaluated by receiver–operator curves (ROC).

Results. There was a significant correlation between P24 and the P/C ratio during the 6-month period (P < 0.001 in all time points). Mean differences between P24 and P/C ratios at baseline and from the first to the sixth month were 2.00, 1.88, 1.22, 1.07, 0.65, 0.34 and 0.57 respectively. In spite of the lower agreement between P24 and the P/C ratio for higher levels of proteinuria, we found substantial Kappa values for categories of proteinuria in all periods. ROC considering the cut-off levels of 0.20 g and 3.5 g for P24 showed that the P/C ratio had a very good accuracy, with areas under the curve of 0.99 (95% CI: 0.97–1.00) and 0.99 (95% CI: 0.99–1.00), respectively.

Conclusion. This longitudinal analysis corroborates the findings of previous cross-sectional studies, supporting the use of the P/C ratio as an accurate test to define critical levels of proteinuria.

Keywords: 24-h proteinuria; nephrotic proteinuria; primary glomerulopathies; protein/creatinine ratio



   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 References
 
Proteinuria is an early sensitive marker of renal damage, and it is the most important test for both the initial evaluation and follow-up of patients with glomerulopathies. Dipstick tests are not clinically useful, since they have a low specificity and sensitivity for the detection and quantification of proteinuria [1]. Moreover, measurement of protein in 24-h urine is not convenient and is subject to collection errors, which may range from 12% to 35% in previous series [2–4]. Since 1983, several studies have used the protein/creatinine (P/C) ratio in urine samples in different clinical settings to estimate 24-h proteinuria (P24), with a good correlation and agreement of these two methods [4–18]. However, these studies are cross-sectional and did not address the critical issue of the P/C ratio accuracy compared with P24 in individual patients prospectively. Ruggenenti et al. correlated the P/C ratio and P24 in baseline urine measurements of non-diabetic nephropathies, but evaluated only the isolated P/C ratio in the longitudinal study [19].

In patients with glomerulopathies with or without nephrotic syndrome, repeated measures of proteinuria are needed to evaluate the effects of therapeutic interventions and to determine the outcome of the glomerular disease, in order to achieve an appropriate clinical management. Furthermore, in these patients, it is necessary to define a precise level of proteinuria to identify a total or partial response, or resistance to treatment.

The purpose of this study was to evaluate the P/C ratio and P24, concurrently and sequentially, in order to validate the clinical usefulness of the P/C ratio compared with P24, in patients with primary glomerulopathies under therapeutic intervention with immunosuppressants or angiotensin-converting enzyme inhibitors (ACEI).



   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 References
 
Study design
This is a longitudinal study in patients with primary glomerulopathies followed over a 6-month period.

Patients
During an 18-month period, patients with primary glomerulopathies followed at the outpatient clinic in Hospital de Clínicas de Porto Alegre were included in the study, according to the following criteria: age >18 years, creatinine clearance ≥15 mL/min/1.73 m2, absence of severe cardiac disease, absence of urinary tract infection, no drug use that could interfere in urinary creatinine excretion and absence of pregnancy. Forty-five patients were initially selected, but four were excluded because they did not fulfil inclusion criteria.

Clinical characteristics of these remaining 41 patients were: 26 patients with nephrotic syndrome (P24 ≥ 3.5 g/ 1.73 m2) treated with immunosuppressants with segmental and focal glomerulosclerosis (FSGS) and 6 with idiopathic membranous nephropathy (IMN), and 15 patients with non-nephrotic proteinuria (P24 < 3.5 g/1.73 m2) using ACEI, other anti-hypertensives and statins as needed. Endogenous creatinine clearance was measured by the standard formula with collected 24-h urine. Both creatinine clearance and 24-h proteinuria were corrected for a body surface area of 1.73 m2. Patients with blood pressure ≥ 140/90 mmHg were considered hypertensive [20].

Treatment protocol
In patients with FSGS, prednisone was used initially at a dose of 1 mg/kg/day, for up to 4–6 months; in patients who had frequent relapses and in those who were steroid dependent, cyclophosphamide was included at a dose of 1.5–2.5 mg/kg/day and in the steroid-resistant group, cyclosporine was introduced at a dose of 4 mg/kg/day. Patients with IMN were treated according to Ponticelli et al. during 6 months, followed by cyclosporine in non-responsive cases (4 mg/kg/day) [21].

Urine samples
Patients were instructed to collect urine during 24-h observing total collection time, followed by a blood sample. After first voiding in the morning, 3 mL of urine was taken out from the second voiding for the P/C ratio measurement. The P/C ratio was calculated dividing urinary protein by urinary creatinine, both in mg/dL. Twenty-four hours of proteinuria was measured in grams and adjusted for a body surface area of 1.73 m2. The specimen was rejected as probably incomplete if creatinine excretion in 24 h (mg/kg) was below the minimum limits for age and sex, as follows: age ≤50 (males ≤18.5 and females ≤16.5), or >50 (males ≤15.7 and females ≤11.8).

Laboratory methods
Serum creatinine and urinary creatinine (mg/dL) concentrations were determined using the modified Jaffé method, in an automated ADVIA 1650/Mega Bayer® apparatus. Urinary protein concentration was determined using the colorimetric method with pyrogalol red using an automated ADVIA 1650/Cobas Mira Plus® apparatus. Coefficients of variation (CV) for serum creatinine, urinary creatinine and urinary protein were 3.23%, 3.77% and 4.08%, respectively.

Statistical analysis
Descriptive statistics are presented as percentages for qualitative data and as mean ± standard deviation (SD) or median with minimum to maximum values for quantitative data in the presence of skewness. The correlation between P24 and the P/C ratio during each period was analysed by the Pearson coefficient, as well as the correlation between the two methods at baseline. The Bland and Altman method was used to evaluate agreement between P24 and the P/C ratio in all periods [22,23]. Kappa statistic was employed to evaluate agreement between the two methods expressed in relative terms (P24 over P/C ratio) and taking into account clinically relevant categories of proteinuria: ≤0.20 g/ 24 h (normal), 0.21–3.5 g/24 h (non-nephrotic) and >3.5 g/ 24 h (nephrotic) [24]. Analysis of variance (ANOVA) for repeated measures (GLM) was employed to evaluate the mean differences (95% CI) between P24 and the P/C ratio. Within subject variation of P24 and the P/C ratio was calculated and compared between the two methods. Accuracy of the P/C ratio in the diagnosis of pathological non-nephrotic and nephrotic range proteinuria was plotted in receiver–operator curves (ROC). Data were processed and analysed in SPSS for Windows software, version 12.0. P values <0.05 were considered statistically significant.

All patients were informed about the study purposes and gave their written consent. The present study was approved by the Committee of Ethics in Research at Hospital de Clínicas de Porto Alegre, with Institutional Review Board (IRB) number 00000921.



   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 References
 
Twenty-six patients with primary nephrotic syndrome under immunosuppressive therapy and 15 patients with non-nephrotic proteinuria treated with ACEI and other drugs were included. Two hundred and forty-six urine samples were used for analysis, with six samples per patient.

Table 1 shows demographic and laboratory characteristics of the patients. Medians (minimum to maximum values) of P24 and the P/C ratio were 6.7 g (0.5 to 28.0) and 5.0 (0.35 to 13.0), respectively. All patients had a creatinine clearance ≥30 mL/min (K-DOQI stages I–III). In patients with nephrotic syndrome (n = 26), median of P24 (minimum to maximum values) was 8.9 g (3.7 to 28.0) and median of the P/C ratio was 6.7 g (3.9 to 13.0). In patients with non-nephrotic proteinuria (n = 15), medians (minimum to maximum values) of P24 and the P/C ratio were 2.0 g (0.5 to 3.3) and 1.6 (0.35 to 2.6), respectively.


View this table:
[in this window]
[in a new window]

  		Table 1 Demographic and laboratory data at presentation

 
Figure 1 presents median and range of 24-h proteinuria and the P/C ratio in the six collection time points. As expected in patients treated with immunosuppressants or ACEI, there was a reduction in 24-h proteinuria levels and the P/C ratio over the 6-month period.


Figure 1
View larger version (16K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 Median and range of 24-h proteinuria and P/C ratio measurements in the baseline (month 0) and six time points of observation. P/C: protein/creatinine, P24: 24-h proteinuria.

 
At baseline the correlation between the mean values of the two methods was r = 0.90, P < 0.001. There was a significant correlation between P24 and the P/C ratio throughout follow-up, with the following r values: 0.90; 0.92; 0.89; 0.91; 0.93 and 0.94, respectively (P < 0.001, for all correlations).

Figure 2 shows the means and limits of agreement (95% CI) of the differences between P24 and the P/C ratio during the period of observation (Bland and Altman method). During the first 3 months, when urinary protein excretion was very high, the agreement between the two methods was poor; however, a good agreement was found in lower levels of proteinuria. When analysis was performed by the Bland and Altman method and GLM, statistical significance was found in the limits of agreement of the two methods (95% CI) (P = 0.01). Intra-individual coefficients of variation for P24 and the P/C ratio were 24% and 25%, respectively. In Table 2 we show that agreement expressed in relative terms had a large variation (14–41%), but the Kappa coefficient applied to the three different categories of proteinuria (normal, non-nephrotic and nephrotic) was substantial or almost perfect in all time points.


Figure 2
View larger version (16K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 Bland and Altman representation of the limits of agreement between P24 – P/C ratio difference with 95% CI at baseline (month 0) and during the six time points of collection.

 

View this table:
[in this window]
[in a new window]

  		Table 2 Levels of agreement expressed in relative terms (P24 over P/C) and Kappa coefficients calculated for three categories of proteinuria*

 
Figure 3 presents the ROC showing diagnostic accuracy to determine pathological (P24 > 0.20 g) and nephrotic (P24 > 3.5 g) levels of proteinuria. Both areas under the curve (AUC) were equal to 0.99 with a 95% CI of 0.97–1.00 and 0.99–1.00, respectively.


Figure 3
View larger version (9K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3 ROC showing diagnostic accuracy of P/C to determine pathological [P24 > 0.20 (A)] and nephrotic [P24 > 3.50 (B)] levels of proteinuria.

 


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
References
 
In clinical practice, a morning urine sample can be used to detect and monitor urinary protein excretion. Although the albumin/creatinine ratio in urine samples is more reliable, limitations due to cost and technical difficulties make its routine use impractical. The P/C ratio is an acceptable alternative, especially to monitor protein excretion under conditions where repeated measurements are mandatory for a successful clinical management. The recommendations for its use were published in the NKF K/DOQI, Guideline 5 [25].

Some features that might influence P/C ratio values, such as age, sex, muscle mass and level of renal function, must be taken into account when following patients with nephropathies. It is well known that the P/C ratio underestimates proteinuria in men and overestimates its value in women, in the elderly and in individuals with tiny muscle mass. The level of renal function has been considered another important issue that can influence interpretation of the P/C ratio. However, Morales et al. in a recent study demonstrated that this variable does not interfere significantly with the results [18].

Within subject variation of P24 and the P/C ratio found in the present study did not differ from those reported by Chitalia et al. [5]. These authors collected two urine samples to measure P/C ratios and urine to measure P24 on the same day. Intra-individual coefficients of variation were 28% and 32% for the two P/C ratios and 29% for P24.

When the correlation between P24 and the P/C ratio is evaluated by the Pearson coefficient and the agreement by the Bland and Altman method, we and others [7,9,18] found that the higher the level of proteinuria, the smaller the correlation and agreement between the methods.

When treating patients with glomerulopathies, with or without nephrotic syndrome, the clinical goal is to normalize or at least to reduce proteinuria. Therefore, in clinical practice the absolute level of proteinuria in individual measurements is less important than its modification and reproducibility over time as a result of therapeutic interventions. In this context, assuming that reproducibility of the P/C ratio since initial diagnosis is adequate, it can be said that significant reduction in the P/C ratio means reduction in protein excretion, even if absolute values cannot be estimated with optimal accuracy [25]. This is demonstrated in Table 2, in which, in spite of the large variation of the P24/P/C ratio when expressed in relative terms, there was an excellent agreement when the Kappa coefficient was calculated to clinically relevant categories of proteinuria [normal (≤0.2 g), non-nephrotic (0.21–3.5 g) and nephrotic (>3.5 g)]. This information is crucial to the clinician to optimize the therapeutic approach of patients with nephrotic syndrome under immunosuppressive protocols. In other words, the decision to change, maintain or withdraw the drug regimen depends on these critical levels of proteinuria.

For detection of these levels of proteinuria, the P/C ratio presented a high level of accuracy. Two previous studies [4,6] used the P/C ratio cut-off values of 0.2 and 3.5 in patients with various nephropathies and stable renal function to establish the diagnosis of pathologic proteinuria (P24 ≥ 0.2 g) and nephrotic range proteinuria (P24 ≥ 3.5 g), respectively. In a study with pregnant women, Ramos et al. reported that the best cut-off to define normal protein excretion (<0.3 g) was a P/C ratio of 0.5 [10]. Two other recent studies [5,18] with a large number of patients also adopted a value ≥0.2 in the definition of pathologic proteinuria and values ≥3.5 for nephrotic range proteinuria. Chitalia et al. described the P/C values of 0.26 and 3.2 as the best cut-off points to establish critical levels of proteinuria [5].

Morales et al. reported in patients with primary glomerular diseases a sensitivity of 93% and a specificity of 100% for P/C values ≥0.3, in relation to P24 as the reference standard. In this study, the best cut-off to define nephrotic range proteinuria was 3.0 (90% sensitivity and 97% specificity) [18]. Using the likelihood ratio (LR) to evaluate the accuracy of the P/C ratio in higher levels of proteinuria, values <2.5 exclude the possibility of nephrotic range proteinuria (LR = 0.0), while values >3.5 confirm this condition (LR = 53.39). For P/C values between 2.51 and 3.5, LR ranged from 1.74 to 4.64, suggesting that interpretation of the P/C ratio in that range should be based on clinical data such as weight, gender, race and body surface area for individual patients (unpublished data).

To our knowledge, this is the first longitudinal study comparing 24-h proteinuria and urinary P/C ratio in patients with primary glomerulopathies. This longitudinal analysis corroborates the findings of previous cross-sectional studies, supporting the use of the P/C ratio in the clinical setting by its simplicity and low cost. In spite of the lower agreement at higher levels of proteinuria in absolute terms (Bland and Altman), the P/C ratio is still adequate to categorize patients as nephrotic. Most important, the P/C ratio reflects with high accuracy the changes in proteinuria over time in patients with primary glomerular diseases under treatment.



   Acknowledgments
 
We would like to thank the Research Incentive Fund (FIPE) of Hospital de Clínicas de Porto Alegre for their financial support.

Conflict of interest statement. None declared.



   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 

  1. Caring for Australians with Renal Impairment (CARI). The CARI guidelines. Urine protein as diagnostic test: performance characteristics of tests used in the initial evaluation of patients at risk of renal disease. Nephrology (Carlton) (2004) 9(Suppl):S8–S14.[CrossRef]
  2. Shaw AB, Risdon P, Lewis-Jackson JD. Protein creatinine index and Albustix in assessment of proteinuria. Br Med J (1983) 287:929–932.[Abstract/Free Full Text]
  3. Gaspari F, Perico N, Remuzzi G. Timed urine collections are not needed to measure urine protein excretion in clinical practice. Am J Kidney Dis (2006) 47:1–7.[Web of Science][Medline]
  4. Ginsberg JM, Chang BS, Matarese RA, et al. Use of single voided urine samples to estimate quantitative proteinuria. N Engl J Med (1983) 309:1543–1546.[Abstract]
  5. Chitalia VC, Kothari J, Wells EJ, et al. Cost-benefit analysis and prediction of 24-hour proteinuria from the spot urine protein-creatinine ratio. Clin Nephrol (2001) 55:436–447.[Web of Science][Medline]
  6. Kristal B, Shasha SM, Labin L, et al. Estimation of quantitative proteinuria by using the protein/creatinine ratio in random urine samples. Am J Nephrol (1988) 8:198–203.[Web of Science][Medline]
  7. Boler L, Zbella EA, Gleicher N. Quantitation of proteinuria in pregnancy by the use of single voided urine samples. Obstet Gynecol (1987) 70:99–100.[Web of Science][Medline]
  8. Combs CA, Wheeler BC, Kitzmiller JL. Urinary protein/creatinine ratio before and during pregnancy in women with diabetes mellitus. Am J Obstet Gynecol (1991) 165:920–923.[Web of Science][Medline]
  9. Saudan PJ, Brown MA, Farrell T, et al. Improved methods of assessing proteinuria in hypertensive pregnancy. Br J Obstet Gynecol (1997) 104:1159–1164.[Web of Science][Medline]
  10. Ramos JG, Martins-Costa SH, Mathias MM, et al. Urinary protein/creatinine ratio in hypertensive pregnant women. Hypertens Pregnancy (1999) 18:209–218.[Web of Science][Medline]
  11. Zelmanovitz T, Gross JL, Oliveira JR, et al. The receiver operating characteristics curve in the evaluation of a random urine specimen as a screening test for diabetic nephropathy. Diabetes Care (1997) 20:516–519.[Abstract]
  12. Dyson EE, Will EJ, Davison AM, et al. Use of the urinary protein creatinine index to assess proteinuria in renal transplant patients. Nephrol Dial Transplant (1992) 7:450–452.[Abstract/Free Full Text]
  13. Steinhauslin F, Wauters JP. Quantitation of proteinuria in kidney transplant patients: accuracy of the urinary protein/creatinine ratio. Clin Nephrol (1995) 43:110–115.[Web of Science][Medline]
  14. Abitol C, Zilleruelo G, Freundlich M, et al. Quantification of proteinuria with urinary protein/creatinine ratio and random testing with dipstick in nephritic children. J Pediatr (1990) 116:243–247.[CrossRef][Web of Science][Medline]
  15. Iyer RS, Shailaja SN, Bhaskaranand N, et al. Quantitation of proteinuria using protein-creatinine ratio in random urine samples. Indian Pediatr (1991) 28:463–467.[Medline]
  16. Sato M, Haizuka H, Asakura H, et al. Quantitation of proteinuria by the use of protein-to-creatinine ratios in random urine samples. Nippon Jinzo Gakkai Shi (1996) 36:8–12.
  17. Schwab SJ, Christensen RL, Dougherty K, et al. Quantitation of proteinuria by the use of protein-to-creatinine ratios in single urine samples. Arch Intern Med (1987) 147:943–944.[Abstract/Free Full Text]
  18. Morales JV, Weber R, Wagner MB, et al. Is morning urinary protein/creatinine ratio a reliable estimator of 24-hour proteinuria in patients with glomerulonephritis and different levels of renal function. J Nephrol (2004) 17:666–672.[Web of Science][Medline]
  19. Ruggenenti P, Gaspari F, Perna A, et al. Cross sectional longitudinal study of spot morning urine protein:creatinine ratio, 24 hour urine protein excretion rate, glomerular filtration rate, and end stage renal failure in chronic renal disease in patients without diabetes. BMJ (1998) 316:504–509.[Abstract/Free Full Text]
  20. Chobanian AV, Bakris GL, Black HR, et al. The Seventh Report of the Joint National Committee on Prevention, Evaluation and Treatment of High Blood Pressure. The JNC 7 report. JAMA (2003) 289:2560–2572.[Abstract/Free Full Text]
  21. Ponticelli C, Zuchelli P, Passerini P, et al. A 10-year follow-up of a randomized study with methylprednisolone and chlorambucil in membranous nephropathy. Kidney Int (1995) 48:1600–1604.[Web of Science][Medline]
  22. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet (1986) 1:307–310.[CrossRef][Web of Science][Medline]
  23. Bland JM, Altman DG. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet (1995) 346:1085–1087.[CrossRef][Web of Science][Medline]
  24. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics (1977) 33:159–174.[CrossRef][Web of Science][Medline]
  25. National Kidney Foundation. K/DOQI clinical practice guideline for chronic disease: evaluation, classification and stratification. Am J Kidney Dis (2002) 39:S1–S266.[CrossRef][Web of Science][Medline]
Received for publication: 3. 3.07
Accepted in revised form: 21.12.07


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Nephrol Dial TransplantHome page
E.-Y. Park and T.-Y. Kim
How to interpret the protein/creatinine ratio in patients with low GFR
Nephrol. Dial. Transplant., December 1, 2009; 24(12): 3892 - 3893.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/7/2242    most recent
gfm949v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Antunes, V. V. H.
Right arrow Articles by Morales, J. V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Antunes, V. V. H.
Right arrow Articles by Morales, J. V.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?