To characterize the female urethral microbiota, define their relationship to bladder and vulvo-vaginal microbiota, and to assess the microbial contributions of the lower urinary tract (bladder and urethra) and vulvo-vaginal skin (peri-urethra) to voided urine. We hypothesized that the female urethral microbiota are distinct from the bladder microbiota. We further hypothesize that voided urine contains significant bacterial contribution from the urethra and vulvo-vaginal skin.

This IRB-approved study included subjects ≥18 years of age presenting to the clinic for initial evaluation of their pelvic floor symptoms. Participants were excluded if they had a history of gynecologic malignancy, pelvic radiation, or urethral resection, if they had Foley/suprapubic catheters in place, if they practiced intermittent self-catheterization, or if they had taken antibiotics in the past 30 days. From each participant, we collected demographic variables and four lower urinary tract samples in the following order: mid-stream voided urine, peri-urethral swab, transurethral brush, and catheterized urine. Mid-stream voided urine was collected by the standard clean catch method into a sterile container. The peri-urethral swab was collected by swabbing the peri-urethral vagina 5mm from the urethral opening. The transurethral brush sample involved placing a sterile brush directly into the urethra and advancing until no longer visible (about 1.25cm) and rotating 360 degrees before withdrawing. Routine betadine prep was applied to the urethral meatus prior to collecting the catheterized urine sample. Each sample type was submitted for culture via expanded quantitative urine culture and bacterial isolates were identified by MALDI-TOF mass spectrometry. Bray-Curtis analysis was then used to assess the diversity between sample types for each participant. Frequency and relative abundance data were compared between each site. Chi-square and Fisher’s exact tests were used to test for significance.

We collected samples from 49 participants. This cohort was mostly Caucasian (71%) with a median age of 55 (range 21-85). They were also mostly post-menopausal (76%), but approximately equal in terms of sexual activity status (53% versus 47%). By Bray-Curtis dissimilarity analysis, which compares microbial communities (1 = dissimilar and 0 = similar), the catheterized urine microbiota were completely dissimilar to urethral microbiota (median =0.99, p<0.0001), peri-urethral microbiota (median=0.99, p<0.0001), and voided urine microbiota (median =0.97, p<0.0001). Urethral microbiota were only moderately dissimilar to peri-urethral microbiota (median=0.56) and although these specimens were statistically similar (p=0.31), their relationship to the voided urine specimens differed; the urethral microbiota were dissimilar to voided urine specimens (median=0.73, p=0.001), whereas the peri-urethral microbiota were not (median=0.55, p=0.16). Frequency and relative abundance of genera also were compared between samples types. Escherichia was cultured with highest relative abundance from catheterized urine than all other sample types (p=0.002), despite being isolated with highest frequency from voided urine (p=0.05). Conversely, Corynebacterium was isolated from catheterized urine with lowest frequency (p<0.001) and lowest relative abundance (p<0.001). Finally, some genera, such as Gardnerella, were isolated with similar frequencies (p=0.81) and relative abundances (p=0.10) throughout the lower urinary tract.

In terms of species presence and abundance, the female bladder and urethra represent distinct niches within the lower urinary tract. Voided urine is more representative of the peri-urethra than the bladder or urethra. Furthermore, abundance data provides evidence that some genera reside preferentially within the lower urinary tract, while others are found at similar levels throughout.

The female urethra accommodates microbiota in similar presence and abundance to vulvo-vaginal skin, ultimately making it a distinct lower urinary tract niche from the bladder. Voided urine microbiota more closely resembles the vulvo-vaginal skin than the lower urinary tract. Thus, our data support other studies that show that subtractive methods cannot permit interpretation of voided urine in terms of the bladder microbiota. Methods to reduce vulvo-vaginal contribution in voided samples should be developed and validated.