Albumin Excretion Rate Is Not Affected by Asymptomatic Urinary Tract Infection: A prospective study

From January 2002 to January 2003, we prospectively screened all type 2 diabetic subjects attending our primary care outpatient diabetic clinic for UTI and AER every Wednesday and Thursday (n = 765, 334 men and 431 women).

The exclusion criteria considered were macroalbuminuria (AER >200 μg/min), renal failure (serum creatinine >106.1 μmol/l), cardiovascular disease, and symptomatic UTI. To assess the past and present evidence of cardiovascular disease, a standardized cardiovascular questionnaire and a careful physical examination, including an ankle-arm blood pressure and a 12-lead electrocardiogram (Minnesota codes), were used. Patients who smoked and who started or modified antihypertensive treatment during the follow-up were excluded. The patients maintained uniform daily habits (diet and physical exercise) for the duration of the study. In addition, no significant changes in the pharmacological treatment were implemented during the course of the protocol.

AER was determined in 24-h urine samples by the turbidimetric end point method (Olympus; Olympus Diagnostica, Hamburg, Germany). The coefficient of analytical variation was 3.5%. Microalbuminuria was defined as an AER of 20–200 μg/min in at least two of the last three urine samples collected before the inclusion. Normoalbuminuria was defined as an AER <20 μg/min.

Midstream urine specimens were collected into a sterile container, and the samples were cultured for bacterial growth, quantification, and antimicrobial sensitivity. A diagnosis of UTI was made if ≥10⁵ colony-forming units/ml of a pathogen grew in the urine culture. We defined contaminated urine as that having the presence of at least three different microorganisms in the urine specimen. Patients with contaminated urine specimens were excluded. Antimicrobial therapy was indicated in patients with UTI according to antimicrobial sensitivity, and a second determination of AER with another urine culture was performed 2–4 days after finishing the antimicrobial treatment.

To assess AER variability, 56 consecutive type 2 diabetic patients in whom UTI was absent at the beginning of the study and during follow-up were included in the study as the control group. In total, four separate samples were collected from these patients at 3-month intervals, starting in January 2002.

The biological within-subject variance (coefficient of biological variation; CVb) of AER was calculated in the four urine specimens collected at 3-month intervals according to the following formula: CVb = (CVt² − CVa²)^1/2, where CVt is the total intra-individual coefficient of variation and CVa is the coefficient of analytical variation (8). To calculate CVt, it was assumed that values of the repeat measurements in the same subject (four in this case) were normally distributed. The formula used to calculate CVt was 100% × SD/mean.

Informed consent was obtained from all participants, and the study was approved by the hospital’s human ethics committee.

For statistical analysis, AER was logarithmically transformed to correct for skewed distribution; the results are expressed as medians and range. Within-patient comparisons of AER in relation to UTI were carried out by paired Student’s t test. Comparisons among categorical variables were done using the χ² test. All P values are based on a two-sided test of statistical significance. Significance was accepted at the level of P < 0.05. Statistical analyses were performed with the SSPS statistical package.

The frequency of UTI was higher in women than in men (56 of 431 [13%] vs. 3 of 334 [1%]; P < 0.001). The main clinical features of the 59 diabetic patients in whom a UTI was diagnosed, as well as the microorganism causing the infection, are detailed in Table 1. After antimicrobial therapy, the UTI was eradicated in 46 of the 59 patients (78%). The AER did not significantly change after antimicrobial treatment in the whole group of patients with UTI (pre: 13.8 μg/min [0.1–195] vs. post: 8.5 μg/min [0.1–185]; P = 0.1). In addition, we did not observe any significant difference when the analysis was performed separately for patients in whom UTI was eradicated and those in whom UTI persisted after antimicrobial therapy. This was true for normo- and microalbuminuric patients (Table 2). A decline in the median value was obtained in the group of microalbuminuric patients in whom UTI was eradicated. However, the differences were not statistically significant, and a similarly not significant decrease was observed in microalbuminuric patients in whom UTI was not eradicated. In addition, 3 of 10 microalbuminuric patients in whom UTI was eradicated showed an increase rather than a decrease in the AER after antimicrobial treatment. Furthermore, all but one patient remained in the microalbuminuric range after effective treatment of UTI (Fig. 1). In the single patient who became normoalbuminuric (18 μg/min) after UTI treatment, only slight microalbuminuria (34 μg/min) was detected before treatment.

The results of AER measurements performed in the control group are shown in Table 3. The biological variation of AER was 64 ± 27%. However, the CVb for AER varied widely among subjects (0–143%) and was inversely correlated with AER (r = −0.44; P = 0.001) (Fig. 2). Thus, the CVb of AER was higher in normo- than in microalbuminuric diabetic patients (72 ± 24 vs. 50 ± 28%; P = 0.01). The percentage of the reduction of AER in patients in whom UTI was solved (55 ± 23%) did not exceed what would be expected by the biological variability of AER (64 ± 27%). In addition, in the subset of patients with microalbuminuria in whom a decline in AER was detected after solving the UTI (n = 7), the decrease did not exceed what would be expected by the biological variability of AER in microalbuminuric patients (49 ± 20 vs. 50 ± 28).

The percentage of microalbuminuria misclassification was similar in the group with eradicated UTI and in the control group (6 of 46 [13%] vs. 9 of 56 [16%]; NS) (Fig. 2). It should be noted that the percentage of patients in whom AER was >20 μg/min in the first determination and <20 μg/min in the second analysis was similar in the group with eradicated UTI and in the control group (5 of 46 [11%] vs. 5 of 56 [9%]; NS) (Fig. 2).

Several factors have been considered as classical confounders when interpreting the results of AER (e.g., UTI, exercise, high-protein diet, acute hyperglycemia, hypertension, and AER variability). However, despite the widely recognized prognostic and therapeutic implications of microalbuminuria, some of these confounding factors are still controversial. For example, it has recently been reported that exercise does not increase AER in type 1 diabetic patients, as was previously assumed (9). In the present study, we provided evidence that the presence of asymptomatic or uncomplicated UTI does not significantly modify AER.

Although it has been widely accepted that UTI can cause transient elevations in urinary albumin excretion (American Diabetes Association position statement), there are no consistent reports supporting this statement. In a cross-sectional study, Damsgaard and Mogensen (10) found that elderly type 2 diabetic women with UTI had an AER in the upper range, but that it was not significantly higher than it was in patients without UTI. Watts et al. (11) investigated the influence of UTI on AER in 20 type 1 diabetic patients and did not find significant changes in AER according to the presence or absence of UTI. In the present study, apart from including a large cohort of type 2 diabetic patients with UTI, the biological variability of AER was also determined in a group of diabetic patients without UTI. Our results showed that not only did UTI not influence AER, but also the changes in AER after antimicrobial treatment were similar in range to AER’s biological variability. In addition, UTI did not significantly influence the classification of diabetic patients as normo- or microalbuminuric. The number of patients placed in the microalbuminuric group in whom UTI was eradicated was relatively small. Therefore, although the differences in AER before and after solving UTI were not statistically significant, further studies specifically focused on this group are needed. However, in practical terms, diabetic patients in whom microalbuminuria has been firmly established are not a problematic group when the possible false positive results of AER due to UTI are being analyzed. Instead, diabetic patients with normoalbuminuria represent the potential group at risk for false positive results (from normo- to microalbuminuria) due to UTI.

Overall, our data do not support the concept that UTI has to be ruled out when assessing AER. It is noteworthy that this approach, apart from simplifying the management of diabetic patients, would also save resources invested in patient care. It could be argued that if patients are not screened for UTI, the disease cannot be diagnosed. However, it should be noted that it has been recently established that diabetes itself should not be an indication for screening for or treatment of asymptomatic bacteriuria (12).

There is a considerable intra-individual variability in AERs; the CVb previously reported ranged between 25 and 62% (13–16). In the present study, we confirmed the high intra-individual biological variation of AER (64%), thus supporting the view that it is necessary to consider this confounding factor when evaluating changes of AER after clinical intervention. The high variability obtained could be explained by the duration of the study, which was longer than that of other studies previously reported. The use of 24-h albumin excretion samples could be another reason accounting for the high AER variability. Some authors consider that it is better to use the albumin/creatinine ratio because it may correct for errors in the collection of timed urine samples (17). However, other authors consider that the introduction of another biological variable (urinary creatinine) is unlikely to reduce AER variability. In this regard, Watts et al. (18) reported that the variability was comparable for urine albumin concentration, the albumin/creatinine ratio, and AER. In the present study, we found that the variability of AER was inversely related to its magnitude. Mosca et al. (16) also reported a higher biological variability of AER in normo- than in microalbuminuric type 2 diabetic patients, although other authors did not find this negative association (5).

As expected, we observed a higher percentage of UTI in women than in men (13 vs. 0.9%). The prevalence of UTI observed in the type 2 diabetic women was within the range previously reported (7.9–26%) (6,18–20). The prevalence of infecting microorganisms in the present study was also the same as that previously reported in diabetic patients (6,12,21), with Escherichia coli being the most frequent.

In summary, the presence of asymptomatic UTI does not significantly influence the measurement of AER, and, therefore, testing for UTI could be avoided when screening for microalbuminuria in diabetic patients.

This study was supported by grants from Novo Nordisk Pharma S.A. (01/0066) and the Instituto de Salud Carlos III (G03/212 and C03/08).

We thank Carme Creus and Dr. Carme Ricós for technical assistance.