Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Jun 1.
Published in final edited form as: J Diabetes Complications. 2020 Mar 3;34(6):107561. doi: 10.1016/j.jdiacomp.2020.107561

Characteristics of the Microbiota in the Urine of Women with Type 2 Diabetes

Sue Penckofer a, Robert Limeira a, Cara Joyce a, Meghan Grzesiak b, Krystal Thomas-White a, Alan J Wolfe a
PMCID: PMC7329247  NIHMSID: NIHMS1576795  PMID: 32184058

Abstract

Purpose

The urinary microbiota in women with type 2 diabetes (T2DM) can have bacterial uropathogens which are more virulent. The primary objective was to describe and compare the characteristics of the microbiota in voided urine of women with and without T2DM.

Methods

Two cohorts of women: those with T2DM (n=87) and those without T2DM (n=49) were studied. Demographic data, hemoglobin A1c (HbA1c), fasting serum glucose, and voided urine were collected. To determine the characteristics of the microbiota in the urine, 16S rRNA gene sequencing was used.

Results

The genus Lactobacillus was more often present in women with T2DM (75.9%, n=66) than in the controls (59.2%, n=30) (p=0.042), as was the family Enterobacteriaceae (12.6% T2DM versus 2.0% control, p=0.055). There was evidence of an association between HbA1c and the relative abundance of the various bacteria in the total cohort. The relative abundance of Lactobacillus was positively associated (ρ=0.19, 95% CI: 0.02, 0.34), while Corynebacterium (ρ=−0.26, 95% CI: −0.41, −0.10) and Prevotella (ρ=−0.23, 95% CI: −0.38, −0.06) were inversely associated with HbA1c.

Conclusions

Enterobacteriaceae (e.g. E. coli) predispose women to urinary tract infections and since T2DM increases this risk, further study is needed. The species of Lactobacillus and its impact needs exploration.

1.0. Introduction

Diabetes affects 1 in 10 persons in the United States and is projected to increase to 1 in 4 persons by 2050 (Boyle, 2010). One of the most distressing complications of diabetes is its impact on urinary function (James & Hijaz, 2014). Weekly incontinence is reported by women with type 2 diabetes mellitus (T2DM) more often than any other diabetes complication (Phelan et al., 2009). In up to 50% of persons with T2DM, impaired bladder sensation occurs, which results in lower urinary tract symptoms (LUTS), including those associated with urinary incontinence (UI) and urinary tract infection (UTI) (Arrellano-Valdez et al., 2014; Wilke et al., 2015).

Recent epidemiologic data indicate that among those with T2DM compared to those without T2DM, there is a 25% increase risk for UTI (Nichols et al., 2017). Factors that increase risk for UTI include metabolic control. Glucose alterations can facilitate the growth of pathogens in urine, leading to UTI (Yu et al., 2014; Liu et al., 2017a; Geerlings et al., 2014). Although the most common bacterial agents of UTIs in persons with T2DM (e.g., Escherichia coli and Klebsiella pneumoniae) mirror those in the general population, these uropathogens appear to be more virulent in T2DM (James and Hijaz, 2014). Serious complications of UTI, such as pyelonephritis and renal papillary necrosis, are more frequent in those afflicted with T2DM than in the general population (Nitzan et al., 2015). In addition, since women with newly diagnosed T2DM have a higher prevalence and incidence of UTI compared to men with T2DM (Fu et al., 2014), the study of women with T2DM is important.

Our team has discovered live bacterial communities in the female lower urinary tract (urinary microbiota) that are relevant to clinical conditions associated with LUTS (Wolfe et al., 2012; Hilt et al., 2014; Pearce et al., 2014; Pearce et al., 2015; Brubaker et al., 2014; Thomas-White et al., 2015; Thomas-White et al., 2016; Price et al., 2016), using a combination of 16S rRNA gene sequencing and an enhanced urine culture method called Expanded Quantitative Urine Culture.

Limited data exist regarding the urinary microbiota and T2DM. Therefore, in the current study, our primary objective was to describe and compare the characteristics of the microbiota in voided urine of women with and without T2DM.

2.0. Materials and Methods

2.1. Design and sample

A descriptive comparative design was used. Data were compared from women who participated in two different studies. In the first study, women with T2DM were recruited to examine vitamin D supplementation and its impact on mood and health outcomes (LU IRB #204197). For that trial (ClinicalTrials.gov Identifier:NCT01904032), the inclusion criteria were: 1) women aged 21 and older, 2) having significantly elevated depressive symptoms as measured by a score ≥16 using the Center for Epidemiologic Studies Depression Tool (CES-D) and/or taking antidepressant medication and having a CES-D score ≥12, 3) type 2 diabetes and under the care of a health care provider, (4) level of vitamin D also known as 25 hydroxy vitamin D, 25 (OH) D less than 32 ng/dl. The exclusion criteria were: 1) current alcohol or substance abuse disorders, 2) a history of bipolar depression or any other psychotic disorder, 3) debilitating chronic illness (e.g., active cancer, uncontrolled multiple sclerosis), 4) severe complications of diabetes (e.g., blindness or amputation) since these will impact depression, 5) malabsorption disorders, 6) elevated serum calcium, 7) taking St. John’s Wort unless they stopped for 3 weeks prior to enrollment since it can impact mood, 8) use of vitamin D supplements (1000 IUs or greater) in the past 2 months and unwillingness to discontinue for at least 1 month prior to study, 9) pregnant or planning a pregnancy, 10) baseline systolic blood pressure greater than 160 mmHG or diastolic greater than 100 mmHG. Having active treatment for depression (e.g., antidepressant therapy) was not an exclusion criterion if the patient had been under treatment for six weeks or more.

We obtained our control population from the health disparities microbiome study (LU IRB #210259), which included women without T2DM of age, race, and body mass index comparable to the first study. Criteria for participation were healthy women: (1) aged 30 to 70, (2) no neurologic disease known to affect the lower urinary tract, (3) no history of or current pelvic malignancy or radiation and/or significant pelvic organ prolapse that increases LUTS, (4) not taking systemic or local antibiotics in the 4 weeks prior to sample acquisition, (5) no diabetes history, and (5) no current urinary tract infection.

2.2. Procedures

Data collection methods and surveys of self-report were consistent for both studies. For baseline visits, women underwent phone screening prior to enrollment to ensure they met the respective study criteria for participation. All participants fasted for 8 hours prior to each visit. A nurse practitioner used standardized procedures for consenting participants, collecting blood and urine specimens (see below), measuring height and weight (using an electronic calibrated scale and a stadiometer), and administering the self-report health and demographic survey which included questions about race, ethnicity, age, diagnosis of diabetes and duration.

2.2.1. Specimen Collection

Both blood and urine specimens were collected for the studies. A serum specimen to assess for fasting glucose was sent to Quest Diagnostics, a CLIAA-certified and approved laboratory. The hemoglobin A1c, which reflects average glucose values over 90 days, was assessed with the DCA Vantage Analyzer using a finger stick (Siemens Healthcare Diagnostics, 2011). The nurse instructed the participants to collect voided urine. In brief, women were told to wash their hands with soap and water, remove all undergarments, place the specimen hat at the front of the toilet, and use the towelette provided to clean the urinary area from front to back one time. Each participant discarded the towelette in a bathroom stall container and then voided into the specimen hat. The study nurse then put this urine into a specimen cup.

Within four hours of urine specimen collection, it was labeled with an identification number, separated into two containers, and inventoried. A portion of each urine sample was placed into a BD Vacutainer® Plus C&S Preservative Tube for culturing. The tube contained a lyophilized boric acid preservative to prevent bacterial overgrowth without causing cell death, allowing specimens to be held at room temperature for up to 72 hours without altering integrity. The second container contained 10% Assay Assure (Sierra Molecular, Incline Village, NV), a DNA preservative that minimizes DNAse activity and freeze/thaw damage to nucleic acids. This container was kept at −80°C until bacterial DNA isolation and sequence analysis.

2.2.2. Specimen Analyses

For gene sequencing, genomic DNA was extracted from urine with the Qiagen DNeasy Blood and Tissue kit using previously validated protocols (Pearce et al., 2014). Amplicon PCR of the V4 region of the 16S ribosomal RNA gene was performed using degenerate primers (Lane et al., 1985) with index sequences (Caporaso et al., 2012). Negative extraction controls were included at all steps to monitor for potential reagent contamination. Gel electrophoresis was used to determine whether the extraction and amplicon and PCR amplification steps were successful without gross contamination of negative controls. The successfully extracted genomic DNA was stored at −80°C until construction of sequencing libraries. Multiplex 16S rRNA PCR and sequencing was performed using the validated Illumina 16S Sample Preparation Guide with the MiSeq at the Loyola Genomics Facility. Briefly, this protocol uses the Nextera XT library prep kit for indexing PCR attaching indexes and adapter regions to the amplicon DNA. The PCR products were cleaned with AMPure magnetic beads and DNA quality and quantity were assessed by gel electrophoresis and fluorescent dsDNA assay. The negative extraction controls were sequenced along with their respective samples; at least one positive control (ZymoBIOMIC™ Microbial Community Standard, Zymo Research, Irvine, CA) was sequenced per sequencing run. The samples were pooled and sequenced with sequencing quality metrics monitored. Sequence run quality was assessed using mothur seq error function and extraction controls compared against samples for contamination. Sample DNA was re-extracted and sequenced if they were suspected of data discrepancies such as possible contaminated extraction controls or low read counts (less than 2,000). Some of these procedures were previously described (Kozich et al., 2013; Caporaso et al., 2012).

For bioinformatics analyses, the 16S rRNA gene sequencing quality control and de-multiplexing of sequence data were done with onboard MiSeq Control software and MiSeq Reporter (current version: 2.1.43). The mothur pipeline, version 1.40.1 (Schloss et al., 2009), with its standard operating procedure specifically for use with MiSeqgenerated data combined paired end reads and remove contigs of incorrect length (<285 bp, >300 bp), contigs containing ambiguous bases and chimeric sequences. Taxonomic classification at genus, family, order, class, and phylum levels was performed using RDP Classifier (Wang et al., 2007).

3.0. Statistical analysis

Descriptive statistics for patient demographics, clinical characteristics, and microbiota were calculated overall and stratified by group (diabetes versus control). Spearman’s rank correlation coefficients were calculated for HbA1c with relative abundance of genera. Differences in continuous variables by group were compared with a t-test for age and Wilcoxon rank sum tests for all others. The presence of a genus was defined by a relative abundance of at least 5%. A genus was considered dominant if it constituted at least 50% of the relative abundance of any specimen. Categories called urotypes were defined by hierarchical clustering. This was done using the Bray Curtis dissimilarity calculation performed with the vegan R package (Version 2.5–2) (Oksanen et al., 2018). To calculate the number of clusters based on the hierarchical clustering, Gap Statistics from the R package cluster (Version 2.0.7–1) were used with kmax parameter of 20 and 500 bootstraps (Maechler et al., 2018; Tibshirani et al., 2001). The resulting clusters are the urotypes named by predominant genera, with an additional Mixed urotype for specimens with no dominant genus. The associations between nominal patient characteristics, genera presence/dominance, and urotype were assessed for statistical differences between diabetes versus controls using chi-square or Fisher’s exact tests, as appropriate. Analyses were performed using SAS 9.4 (SAS Institute, Cary, NC).

4.0. Results

4.1. Description of Sample and Urotypes

The baseline characteristics of the participants are presented in Table 1. There were 87 women in the diabetes cohort, and 49 women in the control cohort. Together, both cohorts averaged 51 years of age. The majority was nonwhite. Most were obese (body mass index 34 to 35). There were no differences between cohorts in terms of sociodemographic or health data. By design, women with T2DM had higher values for HbA1c and fasting glucose (p<0.001).

Table 1:

Participant demographics and clinical characteristics

Overall (n=136) Diabetes (n=87) Control (n=49) p-value
Age, mean (SD) 51.1 (10.8) 51.4 (10.9) 50.5 (10.6) 0.64
Race, n (%)
 White 59 (43.4) 36 (41.4) 23 (46.9) 0.82
 Black 71 (52.2) 46 (52.9) 25 (51.0)
 Asian/Pacific Islander 4 (2.9) 3 (3.4) 1 (2.0)
 Arabic 2 (1.5) 2 (2.3) 0 (0.0)
 Hispanic ethnicity, n (%) 18 (13.2) 12 (13.8) 6 (12.2) 0.80
 Body mass index, median (IQR) 35.1 (31.5–40.4) 35.7 (31.6–40.5) 34.4 (31.1–39.5) 0.36
 Menstrual period, n (%) 52 (38.2) 30 (34.5) 22 (44.9) 0.23
 HbA1c (%), median (IQR) 6.4 (5.6–8.1) 7.5 (6.6–9.1) 5.6 (5.3–5.7) <0.001
 Fasting glucose (mg/dL), median (IQR) 114 (94–165) 149 (116–183) 93 (86–97) <0.001
Open in a new tab

SD-standard deviation; IQR-interquartile range

Bray-Curtis dissimilarity of 16S rRNA gene sequencing clustered the diabetes and control cohorts into 4 and 5 urotypes, respectively (Figure 1). All samples had measurable levels of 16S rRNA genes. Nearly half of all participants were assigned to the Lactobacillus urotype (47.1%, n=64), followed by the Gardnerella (25.0%, n=34), Corynebacterium (14.0%, n=19), Enterobacteriaceae (8.1%, n=11) and Mixed (5.9%, n=8) urotypes. The Lactobacillus and Enterobacteriaceae urotypes were more common among the diabetes cohort, while the Gardnerella and Mixed urotypes were more common among controls (p<0.001) (Table 2).

Figure 1.

Figure 1.

Open in a new tab

Microbiota of diabetic participants. The urinary microbiota profiles of sequence-positive diabetic participants cluster together, as demonstrated in the dendrogram (top) and by the dominant bacterial taxa present, as depicted in the histogram (bottom). The dendrogram was based on clustering using the Bray Curtis dissimilarity index, and each line represents a separate individual. The number of clusters was determined using Gap Statistics. The resulting clusters are the urotypes named by predominant genera, with an additional Mixed urotype for specimens with no dominant genus. The histogram displays the bacterial taxa that were detected by sequencing as the percentage of sequences per urine by sequence positive participants. Each bar on the x-axis represents the urinary microbiota sequence-based composition of a single participant. The y-axis represents the percentage of sequences per participant with each color corresponding to a particular bacterial taxon.

Table 2:

Cohort comparison by urotype

Diabetes (n=87) Control (n=49) p-value
Urotype n (%)
 Lactobacillus 45 (51.7) 19 (38.8) <0.001
 Gardnerella 19 (21.8) 15 (30.6)
 Corynebacterium 13 (14.9) 6 (12.2)
 Enterobacteriaceae 10 (11.5) 1 (2.0)
 Mixed 0 (0.0) 8 (16.3)
Open in a new tab

4.2. Comparison of Cohorts

Differences also were observed between the two cohorts in terms of the presence and dominance of genera (Table 3a). The genus Lactobacillus was present in 69.9% (n=95) and dominant in 50.0% (n=68) of the participants. But Lactobacillus was more often present in the diabetes cohort (75.9%, n=66) than in the control cohort (59.2%, n=30) (p=0.042), as was the family Enterobacteriaceae (12.6% diabetes versus 2.0% control, p=0.055). In contrast, the genus Corynebacterium was more often present in the controls (30.0% controls versus 13.8% diabetes, p=0.018). The diabetes group had fewer Gardnerella-dominant and more Enterobacteriaceae- and Streptococcus-dominant participants, though differences were not statistically significant (Table 3a). Lactobacillus was more abundant in the diabetes cohort (median = 0.404, interquartile range [IQR]: 0.059–0.911) compared to the control cohort (median = 0.148, IQR: 0.004–0.663) (p=0.011). While lower in overall relative abundance, Corynebacterium and Staphylococcus were more abundant amongst the controls compared to the diabetes cohort (Table 3b).

Table 3a.

Presence and dominance by genus

Present, n (%) Dominant, n (%)
Diabetes (n=87) Control (n=49) p-value Diabetes (n=87) Control (n=49) p-value
Lactobacillus 66 (75.9) 29 (59.2) 0.042 46 (52.9) 22 (44.9) 0.37
Gardnerella 30 (34.5) 22 (44.9) 0.23 13 (14.9) 14 (28.6) 0.056
Prevotella 26 (29.9) 19 (38.8) 0.29 4 (4.6) 1 (2.0) 0.65
Corynebacterium 12 (13.8) 15 (30.6) 0.018 2 (2.3) 3 (6.1) 0.35
Enterobacteriaceae 11 (12.6) 1 (2.0) 0.055 10 (11.5) 1 (2.0) 0.097
Staphylococcus 6 (6.9) 3 (6.1) 0.99 1 (1.1) 0 (0.0) 0.99
Streptococcus 15 (17.2) 8 (16.3) 0.99 6 (6.9) 0 (0.0) 0.087
Anaerococcus 9 (10.3) 14 (28.6) 0.007 0 (0.0) 1 (2.0) 0.36
Atopobium 11 (12.6) 11 (22.4) 0.14 1 (1.1) 0 (0.0) 0.99
Finegoldia 2 (2.3) 6 (12.2) 0.026 0 (0.0) 0 (0.0) --
Peptoniphilus 5 (5.7) 10 (20.4) 0.019 0 (0.0) 0 (0.0) --
Open in a new tab

Table 3b.

Relative abundance by genus

Diabetes (n=87) Control (n=49)
Median (IQR) of the proportion p-value
Lactobacillus 0.4040 (0.0590, 0.9108) 0.1480 (0.0035, 0.6625) 0.011
Gardnerella 0.0015 (0.0000, 0.1430) 0.0118 (0.0000, 0.3075) 0.43
Prevotella 0.0120 (0.0025, 0.0795) 0.0245 (0.0063, 0.0993) 0.12
Corynebacterium 0.0050 (0.0005, 0.0233) 0.0195 (0.0090, 0.0695) <0.001
Enterobacteriaceae 0.0005 (0.0000, 0.0020) 0.0005 (0.0000, 0.0033) 0.68
Staphylococcus 0.0005 (0.0000, 0.0030) 0.0013 (0.0003, 0.0100) 0.023
Streptococcus 0.0010 (0.0000, 0.0105) 0.0025 (0.0005, 0.0250) 0.27
Open in a new tab

There were no differences in urotype by race or ethnicity, nor when stratified by diabetes versus control group. In terms of weight, an analysis with the categories of non-obese (<30 kg), obese (30 to <35 kg), and severe obesity (≥35) was done. In the overall sample, prevalence of the Lactobacillus urotype increased with increasing weight while Corynebacterium was higher among those with lower weight. These observed trends in the overall sample were driven by significant differences among those with diabetes (p=0.009); the relationship between body mass index (BMI) and urotype was non-significant among controls (p=0.59) (Table 4)

Table 4:

Urotype by BMI and diabetes

Non-obese (BMI < 30 kg/m2) Obese (BMI ≥ 30 to < 35 kg/m2) Severely obese (BMI ≥ 35 kg/m2) p-value
Diabetes (n=87)
Urotype n (%)
 Lactobacillus 2 (18.2) 13 (43.3) 30 (65.2) 0.009
 Gardnerella 3 (27.3) 6 (20.0) 10 (21.7)
 Corynebacterium 5 (45.5) 6 (20.0) 2 (4.3)
 Enterobacteriaceae 1 (9.1) 5 (16.7) 4 (8.7)
 Mixed 0 (0.0) 0 (0.0) 0 (0.0)
Control (n=49)
Urotype n (%)
 Lactobacillus 4 (40.0) 7 (43.8) 8 (34.8) 0.59
 Gardnerella 2 (20.0) 3 (18.8) 10 (43.5)
 Corynebacterium 1 (10.0) 2 (12.5) 3 (13.0)
 Enterobacteriaceae 0 (0.0) 1 (6.3) 0 (0.0)
 Mixed 3 (30.0) 3 (18.8) 2 (8.7)
Open in a new tab

4.3. Associations with Glycemic Control

Finally, in terms of glycemic control, there was evidence of an association between HbA1c and the relative abundance of the various bacteria (Table 5). The relative abundance of Lactobacillus was positively associated with HbA1c in the total cohort (ρ=0.19, 95% CI: 0.02, 0.34). In contrast, Corynebacterium (ρ=−0.26, 95% CI: −0.41, −0.10) and Prevotella (ρ=−0.23, 95% CI: 0.38, −0.06) had inverse associations with HbA1c in the total cohort; however, the correlation between Prevotella and HbA1c was stronger amongst those with diabetes (ρ=−0.27, 95% CI: −0.46, −0.06) compared to the correlation amongst controls (ρ=−0.11, 95% CI: −0.38, 0.18).

Table 5:

Spearman’s rank correlation coefficients of genus relative abundance with HbA1c

Overall (n=136) Diabetes (n=87) Control (n=49)
Spearman’s ρ (95% Confidence Interval)
Lactobacillus 0.19 (0.02, 0.34)† 0.05 (−0.16, 0.26) 0.04 (−0.24, 0.32)
Gardnerella −0.09 (−0.25, 0.08) −0.04 (−0.25, 0.17) −0.05 (−0.33, 0.23)
Prevotella −0.23 (−0.38, −0.06)‡ −0.27 (−0.46, −0.06)† −0.11 (−0.38, 0.18)
Corynebacterium −0.26 (−0.41, −0.10)‡ −0.09 (−0.30, 0.12) 0.00 (−0.28, 0.28)
Enterobacteriaceae 0.01 (−0.16, 0.18) −0.04 (−0.25, 0.17) 0.07 (−0.22, 0.34)
Staphylococcus −0.12 (−0.28, 0.05) 0.12 (−0.09, 0.33) −0.10 (−0.37, 0.19)
Streptococcus −0.08 (−0.25, 0.09) 0.05 (−0.16, 0.26) 0.03 (−0.26, 0.31)
Open in a new tab
†

p<0.05

‡

p<0.01

5.0. Discussion

The primary aim of this study was to describe and compare the characteristics of the microbiota in voided urine of women with and without T2DM. A strength of this study was that the majority of women in each cohort was nonwhite and the two groups were comparable by age, race, and body mass index. The two cohorts differed by levels of blood glucose and HbA1c.Thus, a comparison of the microbiota of voided urine in terms of this variable was possible.

Our data indicates that a difference exists in the microbiota of voided urine of women who have T2DM and those who do not (Table 2). Members of the family Enterobacteriaceae and the genus Lactobacillus were detected more frequently in women with T2DM and often were the dominant taxon (Table 3a). The family Enterobacteriaceae, which includes E. coli, was more frequently detected and was more often dominant in the urine of women with T2DM. This is congruent with the literature, which reports that “the most common pathogens isolated in the urine of persons with T2DM with UTI are E. coli and other Enterobacteriaceae, such as Klebsiella and Proteus species, and Enterococci” (Nitzan, 2015, p. 131). Furthermore, it is reported that persistent asymptomatic E. coli bacteriuria is more common in women with T2DM than women without T2DM; however, it is unknown if the E. coli strains are the same between these groups (Dalal et al., 2009).

The evidence for whether consistently elevated blood sugar (as reflected by HbA1C) impacts the urinary microbiome negatively in the absence of an active infection is not conclusive. However, during urosepsis, higher HbA1c is reported to be related to various E. coli strains (Wang et al., 2013). It is not inconceivable that higher HbA1c in persons with T2DM may provide conditions for the presence and dominance of Enterobacteriaceae in the urogenital tract. Even though persons may be asymptomatic, uropathogenic E. coli is an opportunistic pathogen with no genetic difference between those that do and do not cause disease. Therefore, some as yet unknown host factor may be involved in the transition into pathogenicity (Schreiber et al., 2017). The increased presence of Enterobacteriaceae in women with T2DM increases their potential risk for infection.

Lactobacillus also was detected more frequently at higher relative abundance in women with T2DM (Table 3b), particularly for those who had the greatest BMI (Table 4). These results align with those from two recently reported studies of voided urine obtained from women with and without T2DM. In one study, higher levels of Lactobacillus were more associated with T2DM in elderly women (mean age 72) than non-elderly women (mean age 50) (Liu et al., 2017b). The authors reported that agerelated diseases, such as diabetes, can affect the urine composition and hypothesized that the urine of elderly women provides a different environment for growth of microbes compared to younger adults. The same team also reported that comorbid conditions impacted relative abundance of Lactobacillus and found that this genus was more abundant in women with T2DM and T2DM plus hyperlipidemia and less abundant in women with T2DM plus hypertension (Lui et al., 2017a). Our study confirms these associations but is strengthened by a comparison group of women without T2DM of comparable age and menstrual status. However, further study of the contribution of body weight, dietary and/or lifestyle factors to the urinary microbiome is warranted.

This association between Lactobacillus and T2DM is important as the levels of Lactobacillus in the vagina are reported to be associated with menopausal status (Brotman et al., 2014), and in the bladder with reduced risk of post-operative UTI (Thomas-White et al., 2018). Most lactobacilli are considered to be beneficial, but some are associated with urinary disorders. For example, one species of Lactobacillus (L. gasseri) is reported to be associated with urgency urinary incontinence, while another (L. crispatus) is associated with the lack of LUTS (Pearce et al., 2014). Because the ages of our two cohorts are comparable, the measured difference should not be attributed to hormone level changes in the bladder and vaginal microbiota. Yet, the reason for the association between Lactobacillus and T2DM remains unclear, however, body weight appears to be an important factor. It has been reported that Lactobacillus is resistant to widely used antibiotics and “may multiply during treatment giving the genus an advantage over antibiotic sensitive commensals and allowing it to invade the urinary system” (Siddiqui et al., 2012, p. 10). It is also reported to cause microbial dysbiosis in the bladder (Liu et al., 2017 a). Persons with T2DM often have multiple infections (e.g., sinus, urinary, skin), requiring the use of antibiotics. Thus, whether previous antibiotic use is a reason for the increased presence of Lactobacillus in T2DM is worthy of exploration, particularly since its increased relative abundance in T2DM is a new finding compared to women without T2DM of similar characteristics.

Our findings regarding the association between HbA1c and certain microbiota also are novel (Table 5). It has been reported that diabetes can lead to alterations in the urine composition, including increased glucose and pH, which can provide an environment for the growth of harmful bacteria (Liu et al., 2017a). Higher HbA1c was associated with more Corynebacterium, a genus that includes the emerging uropathogen C. amycolatum. Also, lower HbA1c was associated with decreased levels of the genus Prevotella. At present, the role of Prevotella in the lower urinary tract remains unclear.

Some limitations to this study include the relatively small sample size and the use of voided urine. While catheterization is the best collection method for scientific study of the urinary microbiota, catheterization of women with T2DM places them at greater risk for infection (Nitzan et al., 2015); thus, we chose to use voided urine for this study. Since voided urine may include microbiota from the urethral and vulvo-vaginal niches, our results do not only reflect microbial differences in bladders of women with and without T2DM, but rather a mixture of urinary and genital microbiota. Finally, even though 16S rRNA gene sequencing achieves depth at a reasonable cost to focus on bacteria, the use of metagenomics may be a possible methodology for future research of the urinary microbiome.

6.0. Conclusions

Little is known about the urinary microbiota of persons with T2DM. The current study adds to the information known about their microbiome. However, further exploration of the species of Enterobacteriaceae and Lactobacillus and their contribution to potential infections for T2DM is warranted. The importance of this cannot be emphasized enough given that UTIs are reported to impose a significant cost burden for those with T2DM (Yu et al., 2014). Given the greater risk for UTIs in this population and the greater risk for significant complications following UTI (Nitzan et al., 2015), the study of the urinary microbiota is a fertile area for future research. Future studies are needed to identify potential microbes that may help maintain a healthy urogenital microbiome in persons with T2DM and may reduce potential adverse outcomes in this population. Additionally, a more nuanced understanding is needed to determine what environmental and host factors contribute to recurrent infection in persons with T2DM, such as glucose levels, immune factors, or microbiome alterations.

Figure 2.

Figure 2.

Open in a new tab

Microbiota of control participants. The urinary microbiota profiles of sequence-positive control participants cluster together, as demonstrated in the dendrogram (top) and by the dominant bacterial taxa present, as depicted in the histogram (bottom). For details, see Figure 1.

Highlights.

The urinary microbiota in women with type 2 diabetes (T2DM) can have bacterial uropathogens which are more virulent.

Characteristics of the microbiota in voided urine of women with and without T2DM was examined using 16S rRNA gene sequencing.

The urinary microbiota of women with T2DM had the genus Lactobacillus more often present as was the family Enterbacteriaceae.

Acknowledgements

Funding: National Institute of Nursing Research (1RO1NR013906-01A10 and RO1NR1306-03S1 to SP), Novo Nordisk Inc. (Investigator Sponsored Study to SP and AJW), Loyola University School of Nursing (to SP), National Institute of Diabetes and Digestive and Kidney Diseases (R01 DK104718 to AJW).

Footnotes

Conflicts of Interest

AJW has investigator-initiated projects with Astellas Scientific and Medical Affairs with Kimberly Clark Corporation. SP and AJW have an investigator-initiated project with Novo Nordisk Inc.

Disclosures: The content is solely the responsibilities of the authors and does not necessarily represent the official views of the National Institutes of Health or Novo Nordisk Inc. The funding agencies had no role in the design, collection, analysis or writing of this manuscript.

References

  1. Arrellano-Valdez F, Urrutia-Osorio M, Arroyo C, Soto-Vega E. (2014). A comprehensive review of urologic complications in patients with diabetes. Springer Plus, 3, 549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Boyle JDP. (2010). Projection of the Year 2050 Burden of Diabetes in the US Adult Population: Dynamic modeling of incidence, mortality and pre-diabetes prevalence. Popul Health Metr. 2010, 8, 29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Brotman RM, Shardell MD, Gajer P, Fadrosh D, Chang K, Silver MI, Viscidi RP, Burke AE, Ravel J, Gravitt PE. (2014). Associations between the vaginal microbiota, menopause status, and signs of vulvovaginal atrophy. Menopause, 21(5), 450–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brubaker L, Nager CW, Richter HE, Visco A, Nygaard I, Barber MD, Schaffer J, Meikle S, Wallace D, Shibata N, et al. (2014). Urinary bacteria in adult women with urgency urinary incontinence. Int Urogynecol J, 25, 1179–1184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Huntley J, Fierer N, Owens SM, Betley J, Fraser L, Bauer M, Gormley N, Gilbert JA, Smith G, Knight R. (2012). Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J, 6(8), 1621–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dalal S, Nicolle L, Marrs CF, Zhang L, Harding G, Foxman B (2009). Long-term Escherichia coli asymptomatic bacteriuria among women with diabetes mellitus. Clin Infect Dis, 49(4), 491–497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fu A, Iglay K, Qiu Y, Engel S, Shankar R, Brodovicz K. (2014). Risk characterization for urinary tract infections in subjects with newly diagnosed type 2 diabetes. J Diabetes Complications, 28(6), 805–810. [DOI] [PubMed] [Google Scholar]
  8. Geerlings S, Fonseca V, Castro-Diaz D, List J, Parikh S. (2014). Genital and urinary tract infections in diabetes: Impact of pharmacologically induced glucosuria. Diabetes Research and Clinical Practice, 103, 373–381. [DOI] [PubMed] [Google Scholar]
  9. Hilt EE, McKinley K, Pearce MM, Rosenfeld AB, Zilliox MJ, Mueller ER, Brubaker L, Gai X, Wolfe AJ, Schreckenberger PC. (2014). Urine is not sterile: use of enhanced urine culture techniques to detect resident bacterial flora in the adult female bladder. J Clin Microbiol 2014, 52, 871–876. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. James R & Hijaz A. (2014). Lower urinary tract symptoms in women with diabetes mellitus: A current review. Curr Urol Rep, 15, 440. [DOI] [PubMed] [Google Scholar]
  11. Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. (2013). Development of a dualindex sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform. Appl Environ Microbiol, 79 (17), 5112–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Lane DJ, Pace B, Olsen GJ, Stahl DA, Sogin ML, Pace NR. (1985). Rapid determination of 16S ribosomal RNA sequences for phylogenetic analyses. Proc Natl Acad Sci U S A, 82, 6955–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Liu F, Ling Z, Xiao Y, Yang Q, Wang B, Zheng L, Jiang P, Li L, Wang W. (2017a). Alterations of urinary microbiota in type 2 diabetes mellitus with hypertension and/or hyperlipidemia. Frontiers in Physiology, 8, 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Liu F, Ling Z, Xiao Y, Yang Q, Zheng L, Jiang P, Li L, Wang W. (2017b). Characterization of the urinary microbiota of elderly women and the effects of type 2 diabetes and urinary tract infections on the microbiota. Oncotarget, 8(59), 100678–100690. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Maechler M, Rousseeuw P, Struyf A, Hubert M, Hornik K. (2018). Cluster: Cluster Analysis Basics and Extensions. R package version 2.0.7–1. [Google Scholar]
  16. Nichols GA, Brodovicz KG, Kimes TM, Deruaz-Luyet A, Bartels DB. (2017). Prevalence and incidence of urinary tract and genital infections among patients with and without type 2 diabetes. J Diabetes Complications, 31, 1587–1591. [DOI] [PubMed] [Google Scholar]
  17. Nitzan O, Elias M, Chazan B, Saliba W. (2015). Urinary tract infections in patients with type 2 diabetes mellitus: Review of prevalence, diagnosis, and management. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy, 8, 129–136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, Minchin PR, O’Hara RB, Simpson GL, Solymos P, Stevens MHH, Eduard Szoe E, Wagner H. (2018). Vegan: Community Ecology Package. R package version 2.5–2. https://CRAN.Rproject.org/package=vegan.
  19. Pearce MM, Hilt EE, Rosenfeld AB, Zilliox MJ, Thomas-White K, Fok C, Kliethermes S, Schreckenberger PC, Brubaker L, Gai X, et al. (2014). The female urinary microbiome: A comparison of women with and without urgency urinary incontinence. MBio, 5, e01283–01214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Pearce MM, Zilliox MJ, Rosenfeld AB, Thomas-White KJ, Richter HE, Nager CW, Visco AG, Nygaard IE, Barber MD, Schaffer J, et al. (2015). The female urinary microbiome in urgency urinary incontinence. Am J Obstet Gynecol, 213, 347.e341–311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Phelan S, Kanaya A, Subak L, Hogan P, Espeland MA, Wing R, Burgio KL, DiLillo V, Gorin AA, West DS, Brown JS, on behalf of Look AHEAD Group. (2009). Prevalence and risk factors for urinary incontinence in overweight and obese diabetic women. Diabetes Care, 32, 1391–1397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Price TK, Dune T, Hilt EE, Thomas-White KJ, Kliethermes S, Brincat C, Brubaker L, Wolfe AJ, Mueller ER, Schreckenberger P. (2016). The Clinical Urine Culture: Enhanced techniques improve detection of clinically relevant microorganisms. J Clin Microbiol, 54, 1216–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Siddiqui H, Lagesen K, Nederbragt AJ, Jeansson SL, Jakobsen KS. (2012). Alterations of microbiota in urine from women with interstitial cystitis. BMC Microbiology, 12, 205. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Schreiber HL, Conover MS, Chou W-C, Hibbing ME, Manson AL, Dodson KW, Hannan TJ, Roberts PL, Stapleton AE, Hooton TM, Livny J, Earl AM, Hultgren SJ. (2017). Bacterial virulence phenotypes of Escherichia coli and host susceptibility determines risk for urinary tract infections. Sci Transl Med, March 22, 9 (382). [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, Lesniewski RA, Oakley BB, Parks DH, Robinson CJ, et al. (2009). Introducing mothur: Open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol, 75, 7537–7541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Tibshirani R, Walther G, Hastie T (2001). Estimating the number of clusters in a data set via the gap statistic. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 63, 411–423. [Google Scholar]
  27. Thomas-White KJ, Gao X, Lin H, Fok CS, Ghanayem K, Mueller ER, Dong Q, Brubaker L, Wolfe AJ (2018). Urinary microbes and postoperative urinary tract infection risk in urogynecologic surgical patients. Int Urogynecol J, 29(12), 1791–1805. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Thomas-White KJ, Hilt EE, Fok C, Pearce MM, Mueller ER, Kliethermes S, Jacobs K, Zilliox MJ, Brincat C, Price TK, et al. (2015). Incontinence medication response relates to the female urinary microbiota. Int Urogynecol J, 54, 1216–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Thomas-White KJ, Kliethermes S, Rickey L, Lukacz ES, Richter HE, Moalli P, Zimmern P, Norton P, Kusek JW, Wolfe AJ, et al. (2016). Evaluation of the urinary microbiota of women with uncomplicated stress urinary incontinence. Am J Obstet Gynecol, 2016 (1), 55.e1–55.e16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Wang Q, Garrity GM, Tiedje JM, Cole JR. (2007). Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol, 73(16), 5261–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Wang M-C, Tseng C-C, Wu A-B, Lin W-H, Teng C-H, Yan J-J,. Wu J-J. (2013). Bacterial characteristicis and glycemic control in diabetic patients with Escherichia coli urinary tract infection. Journal of Micobiology, Immunology, and Infection, 46, 24–29. [DOI] [PubMed] [Google Scholar]
  32. Wilke T, Boettger B, Berg B, Groth A, Mueller S, Botteman M, Yu S, Fuchs A, Maywald U. (2015). Epidemiology of urinary tract infections in type 2 diabetes mellitus patients: An analysis based on a large sample of 456,586 German T2DM patients. J Diabetes Complications, 29, 1015–1023. [DOI] [PubMed] [Google Scholar]
  33. Wolfe AJ, Toh E, Shibata N, Rong R, Kenton K, Fitzgerald M, Mueller ER, Schreckenberger P, Dong Q, Nelson DE, et al. (2012). Evidence of uncultivated bacteria in the adult female bladder. J Clin Microbiol, 50, 1376–1383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Yu S, Fu AZ, Qiu Y, Engel SS, Shankar R, Brodvicz KG, Rajpathak S, Radican L. (2014). Disease burden of urinary tract infections among type 2 diabetes mellitus patients in the US. J f Diabetes Complications, 28, 621–626. [DOI] [PubMed] [Google Scholar]

RESOURCES