Evaluation of method performance for oxidative stress biomarkers in urine and biological variations in urine of patients with type 2 diabetes mellitus and diabetic nephropathy

Subjects

All subjects were volunteers who were informed about the objective of the study before-
hand. During the course of the study no illness or injury was observed in the healthy
individuals and no additional illness was observed in T2DM and DN patients. American
Diabetes Association criteria were used for diagnosis of T2DM 14] and the Kidney Disease Outcomes Quality Initiative criteria were used for diagnosis
of DN 15]. During the study, treatment of T2DM or DN patients was not changed, the disease
process being considered stable. T2DM and DN patients underwent either a fundoscopic
examination or fluoroangiographic study for diagnosis of retinopathy. Patients with
drug induced nephrotoxic damage or secondary causes of renal insufficiency such as
obstructive renal disease, renal stone disease, and acute urinary tract infection
were excluded. The patients who were known to have a familial disease such as autosomal
dominant polycystic kidney disease or Alport disease were not included in the study.

Healthy reference group

This group consisted 8 females, ages 49 to 59 years (mean 53.0 years), and 7 males,
ages 49 to 55 years (mean 52.4 years). All were apparently healthy and none was taking
any drugs, including oral contraceptives.

T2DM reference group

There were 9 females, aged 50 to 55 years (mean 52.2 years) and 11 males, aged 50
to 62 years (mean 55.0 years) in this group.

DN reference group

Ten females, aged 45.0 to 61.0 years (mean 53.0 years) and 10 males, aged 49 to 60 years
(mean 54.4 years) were included in this group.

The demographic characteristics of reference groups were listed in Table 1. Approval from the Ethics Committee of Kahramanmaras Sutcu Imam University Medical
Faculty was taken in accordance with the principles of Declaration of Helsinki and
informed consent was obtained from the cases.

Table 1. Comparisons of demographic and laboratory data among T2DM, DN and healthy individuals

Analytical techniques

Blood samples

The blood samples were taken from each individual on the zero day of our study. The
analytical methods for routine biochemical analysis were standard for the medical
laboratories. In brief, the basic principles and instruments involved were as follows.
Measurements of total protein (biuret reaction), fasting blood glucose (FBG) (hexokinase/glucose-6-phosphate
dehydrogenase) albumin (bromcresol green), urea (urease/glutamate dehydrogenase),
creatinine (Jaffe reaction) were performed by Siemens Advia 1800 analyzer (Germany).
Also, glycosylated haemoglobin (HbA1c) (HPLC method) was determined by Adams HA-8160
analyzer (Japan).

Urine samples

The 55 urine samples which were collected in a particular time interval were included
in the study. The urine samples of subjects were collected into 75-mL sterile containers
(Kayline Plastics, The Barton, South Australia, 5031). To control the urine concentration,
data were normalized to urine creatinine concentration. Urinary creatinine was measured
in spot urine samples by Dade Behring Dimension RXL autoanalyzer (Germany). Then,
all urine specimens were stored frozen at ?70°C until testing at the end of the collection
period.

To minimize interbatch analytical variation, all samples from any volunteer were analysed
in two batches for CAT, CuZnSOD and MDA; therefore, 30 different batches were run
over a period of 30 days. The same lot of combined standard and quality control (QC)
material were used throughout, and analyses were performed by a single analyst. Aliquots
of a single urine pool, stored at ?70°C, were used as the in-house QC material, and
analysed in duplicate in each batch.

Determination of oxidative stress biomarkers in urine

Method of oxidative stress biomarkers in human blood were adopted in human urine (12,
13). At the start of our study, all urine samples were diluted with 1:50 serum physiologic
(0.9% NaCI) for oxidative stress biomarkers analysis.

Superoxide dismutase assay

CuZnSOD activity in the urine samples was measured by the method of Fridovich 16]. The method for CuZnSOD activity employed xanthine and xanthine oxidase to generate
superoxide radicals which react with p-iodonitrotetrazolium violet (INT) to form a
red formazan dye which was measured at 505 nm. Assay medium consisted 0.01 M phosphate
buffer, 3-cyclohexylamino-1-propanesulfonic acid (CAPS), buffer solution (50 mM CAPS,
0.94 mM EDTA, saturated NaOH) with a pH of 10.2, solution of substrate (0.05 mM xanthine,
0.025 mM INT), and 80 U/L xanthine oxidase. CuZnSOD activity was expressed as U/L.

Catalase assay

CAT activity in the urine samples was measured by the method of Beutler 17]. CAT activities were determined by measuring the decrease in hydrogen peroxide concentration
at 230 nm. Assay medium consisted 1 M Tris–HCl, 5 mM Na₂EDTA buffer solution (pH 8.0), 1 M phosphate buffer solution (pH 7.0), and 10 mM H₂O₂. CAT activity was expressed as U/L.

Malondialdehyde assay

MDA concentration, as an indicator of lipid peroxidation, in the urine samples were
determined according to procedure of Ohkawa 18]. The reaction mixture contained 0.1 mL sample, 0.2 mL of 8.1% sodium dodecyl sulfate,
1.5 mL of 20% acetic acid, and 1.5 mL of 0.8% aqueous solution of thiobarbituric acid.
The mixture pH was adjusted to 3.5 and the volume was finally made up to 4.0 mL with
distilled water and 5.0 mL of the mixture of n-butanol and pyridine (15:1, v/v) was
added. The mixture was shaken vigorously. After centrifugation at 4000 rpm for 10 minutes,
the absorbance of the organic layer was measured at 532 nm. MDA concentration was
expressed as nmol/mL.

Glomerular filtration rate (GFR) assay

The GFR is considered most suitable for quantifying renal function. Practical limitations
exist in measuring GFR directly, especially in acutely ill patients. Several reliable
equations incorporating clinical variables to estimate the GFR are available; we used
the Modification of Diet in Renal Disease equation 19].

Statistical analysis

Seven urine samples were taken from each individual on the zero, 1st, 3rd, 5th, 7th,
15th and 30th days. Variation analyses were performed on natural logarithmic transformed
data after exclusion of outliers using Cochran and Reed tests. The Cochran test did
not highlight any outliers among duplicate measurements; however, it did identify
results for seven samples from different subjects (five T2DM, all analytes; and two
healthy individuals, MDA) as outliers among within-subject variances. The Reed test
identified MDA for one healthy individual as an outlier. After exclusion of outliers,
biological variation data were estimated according to the method published by Fraser
et al. 20]-24]. Analytical variance (SD_A²) was calculated from the differences between the duplicates according to the formula:

where d is the difference between duplicates and n is the number of duplicates. The
SD_A² is expressed as relative SD_A to the first sample concentration, analytical coefficient of variation (CV_A). For each analyte, one-way analysis of variance was used to divide the total variance
into between-subject (SD_G²) variance and total within-subject variance (SD_TI²). Since SD_TI² includes both biological and analytical components, the within-subject variance (SD_I²) was obtained by subtraction using the formula:

Within-subject and between-subject biological variations were expressed as the coefficient
of variation by the use of homeostatic mean of each individual, within-subject coefficient
of variation (CV_I) and the overall mean, between-subject coefficient of variation (CV_G), respectively. Index of individuality (II) was calculated as CV_I/CV_G whereby a low value (0.60: high individuality) indicated a low usefulness of population-based
reference intervals, whereas a high value (1.40: low individuality) indicated a high
usefulness of population-based reference intervals, particularly when an unusual result
was repeated for verification.

RCV, which is the difference required for two serial measurements of the oxidative
stress biomarkers to have significantly changed at p??0.05, was calculated as in
25]:

Z-score was accepted as 1.96 (95% probability, bidirectional z-score).

Biological variation data for oxidative stress biomarkers were used to estimate the
desirable quality specifications for I, B and TE, using the following formulas 26].

Differences of the mean biological within-subject variances between the reference
groups (DN/healthy, T2DM/healthy, DN/T2DM, as paired comparison) were checked by the
F test. Linear regression analysis was used to check for significant trends in values
for CAT, CuZnSOD and MDA and to investigate the time dependence of the within-subject
variations. Group comparisons (like male versus female, DN/T2DM, DN/healthy, T2DM/healthy,
etc.) were performed using the Mann–Whitney U test. A p value 0.05 was considered to be statistically significant.