Comparison between different methods of urine collection for estimation of albumin-creatinine ratio in patients with type-2 diabetes mellitus

In patients with diabetes mellitus, urinary albumincreatinine-ratio (ACR) predicts progressive kidney disease. In order to determine the better urine sample for detecting ACR, we estimated ACR in three modes of urine sample. Two hundred patients of uncontrolled diabetes mellitus with proteinuria irrespective of age and sex were studied for urinary ACR in all the three types of samples over a period of 2 years. The statistical analysis showed that first morning ACR (r = 0.999, p > 0.001) of the subjects was more significantly correlated with their 24 hour urine sample ACR than spot urine (r = 0.995, p < 0.001). We conclude that early morning specimens should be used instead of spot and 24 hour sample.


INTRODUCTION
Diabetic nephropathy (DN) is the most frequent single cause of end-stage renal disease in many countries [1].Hyperfiltration and microalbuminuria characterize the clinical stages of DN [2].Increased levels of albumin in the urine have been clearly established as an important determinant for renal complication of diabetes [3][4][5].Screening for increased albumin excretion has therefore been advocated to identify individual at risk for renal disease progression in a timely manner.However, there is still continuing uncertainty as to how urine should be collected and which urinary proteins should be specifically measured for prediction of renal events [6].So, the present study therefore is to investigate which urine sample such as 24 hour, first morning void or spot urine sample is to detect actual value of microalbuminuria for DN.But measuring only urinary albumin excretion (UAE) is not sufficient as it is influenced by intra-individual variations in urinary volume.That can be overcome by using albumincreatinine ratio (ACR) that is dividing UAE with urinary creatinine excretion as creatinine excretion in urine has relative constancy over time [7].So, ACR is used as a parameter to detect albuminuria in the present study.

Selection of Subjects
The present study was conducted in the department of Biochemistry of Burdwan Medical College, Burdwan, West Bengal.One thousand patients having age > 30 years and suffering from uncontrolled diabetes mellitus (plasma glucose > 110 mg/dl, HbA1C > 8% of total Hb) with proteinuria tested by Dipstick technique were selected.The selected patients were from Burdwan district and adjoining areas.Patients having any concomitant infection and recent history of taking drugs (e.g.NSAIDS, ACE inhibitors) were not included in the study.Then two hundred subjects were selected by a simple random method for this study.

Collection of Samples
After noting the age and sex, three types of urine samples (24 hour, first morning voided and random) were collected from all subjects.After collection all the tests were done immediately.

Parameters Assay
Albuminuria was measured by immunoturbidimetric method [8,9] using semiautoanalyser (Chem 5v 2 plus).Intra assay CV% was 2.3 and intra assay CV was 3.1 for this method.Urinary creatinine concentration was assayed by Jaffe' method [10] using auto analyser (Transasia, XL-600).Intra-assay CV% was 1.8 and inter-assay CV% was 2.9.Then urinary ACR was calculated.Using comercially available Hemoglobin A 1C kit supplied by Siemens Company did Hemoglobin A 1C test.It implies the principle of turbidimetric inhibition immunoassay (TINIA).This company also supplied total Hb kit for estimation total Hb by alkaline hematin method.Concentration of plasma glucose of the subjects was done by Glucose-Oxidase and Peroxiase method.

Statistical Analysis
The data for biochemical analysis was subjected to standard statistical analysis using the Statistical Package for Social Science (SPSS) 11.5 software for windows.

RESULT
The personal profiles and clinical parameters of all the subjects under study are shown in Table 1.
Mean value of 24 hour urine ACR (152.38 μg/mg) are more close to first void urine (151.11μg/mg) than spot urine sample (154.85 μg/mg).Another observation is that spot urine ACR value has higher mean, SD and has wider confidence interval than other two samples as shown in Table 2.    Regression analysis is performed to evaluate which method of urine collection and which sample shows best correlate with 24 hour sample.Over the whole range of 24 hour sample ACR, the first morning sample shows a higher r value (r = 0.999, p < 0.0001) and spot urine a lower r value (r = 0.995, p < 0.0001) of ACR.That signifies that first morning ACR of the subjects is more significantly correlated with their 24 hour urine sample than spot urine ACR as shown in Table 3 and Figure 1.
The area under the curve (AUC) of ROC curve of between 24 hour and early morning urine sample ACR is more than between 24 hour and spot urine ACR as shown in Table 4 and Figure

DISCUSSION
ACR has been proposed as both a screening and diagnostic test for kidney disease [11].Screening for microalbuminuria is essential as it allows interventions aimed at preventing diabetic nephropathy [12,13] and part of the everyday treatment of diabetic patients [14][15][16] for detecting kidney disease progression and also evaluation of treatment effect [17] therefore, samples providing immediate and reliable results are highly desirable.With the respect to urine collection procedure 24 hour collection were initially advocated gold standard [18][19][20] because of circadian rhythm of urinary protein excretion [21] and has found to be the least variable parameter for the measurement of microalbuminuria [22,23].But 24 hour urine collection is a cumbersome procedure and subject to collec-tion errors [24][25][26][27].More practical and easier alternatives are collection of a first morning void or a spot (random) urine sample [28][29][30][31][32]. Numerous recent studies have been shown that early morning urinary ACR for screening purpose is also a predictor of overt DN and is useful to identify patient at risk [33] as it is less influenced by the factors such as hydration status, physical activity and concentration bias [34,27].In present study, when cut-off values are used that are advocated in reagent manual, the value of ACR of first void sample are closed to the value 24 hours than spot urine.So, early morning sample can replace the 24 hour sample which is too much lengthy.
Various previous studies also give support towards this factor due to easy collection, low cost and high sensitivity [2,[35][36][37].

CONCLUSION
The study concluded that early morning sample not the spot urine sample could replace the 24 hour sample for estimation of urinary ACR.
plasma glucose level (mg/dl) 168.27 ± 24.40 HbA1C level (% of total Hb) 9.4965 ± 0.997201 Value are mean ± SD; n = number of cases; Reference range of fasting plasma glucose < 110 mg/dl; HbA1C > 8% of total Hb is considered for revaluation of treatment.

Figure 1 .
Figure 1.Scatter diagram (a) Correlation of ACR between 24 hours & first morning urine sample (b) Correlation of ACR between 24 hours & spot urine sample.

Figure 2 .
Figure 2. Receiver Operative Characteristic (ROC) curves of ACR (a) between 24 hours and early morning sample (b) between 24 hours and spot urine sample.

Table 1 .
Personal profiles and clinical parameters of the subjects.

Table 2 .
Comparison of mean value of three modes of urine samples.

Table 3 .
Pearson's correlation of 24 hour urinary ACR with early morning and spot urine ACR.

Table 4 .
Area under the ROC curve and CI between the different modes of urine sample.