Objective
The purpose of this study was to determine the impact of contraception, menopause, and vaginal flora on the physical and biochemical properties of cervicovaginal fluid (CVF).
Study Design
Vaginal swabs, CVF, and cervicovaginal lavage (CVL) were collected from a total of 165 healthy asymptomatic women including: postmenopausal women (n = 29), women in the proliferative (n = 26) or follicular (n = 27) phase, and women using the levonogestrel intrauterine device (n = 28), depomedroxyprogesterone acetate (n = 28) or combined oral contraceptives (n = 27). Vaginal smears were evaluated using the Nugent score. The osmolality, viscosity, density, and pH of CVL samples were measured.
Results
CVL from postmenopausal women and women with abnormal vaginal flora was less viscous and had higher pH than premenopausal women and women with normal flora, respectively. Women using hormonal contraceptives had more viscous CVL as compared with premenopausal women not using hormonal contraceptives, but this increase in viscosity was mitigated in the presence of bacterial vaginosis. Women using depomedroxyprogesterone acetate had less total protein in the CVL as compared with women using the levonogestrel intrauterine device, and had similar protein content when compared with postmenopausal women.
Conclusion
The differences in CVL protein content between depomedroxyprogesterone acetate and levonogestrel intrauterine device suggest that type of progesterone and route of delivery impact the vaginal environment. Contraceptive hormone users had more viscous CVL than women not using contraceptives. However, the presence of bacterial vaginosis impacted both the pH and viscosity (regardless of hormonal contraceptive use), demonstrating that vaginal flora has a greater impact on the physical properties of CVF than reproductive hormones.
Cervicovaginal fluid (CVF) is comprised of transudate from the vaginal epithelium as well as cervical mucus and secretions from the uterus and fallopian tubes. The mucin gel layer that coats the vaginal epithelium is one of the first line defenses in protection against pathogens of the genital tract. In order for sexually transmitted pathogens, such as HIV, to establish infection, they must penetrate the mucus layer and attach to receptors on target cells in the cervical or vaginal epithelium. In a recent study, CVF, collected using a catamenial cup, slowed the diffusion of HIV-1 particles more than 200-nm PEGylated beads, which was dependent on the presence of HIV-1 envelope proteins. This demonstrates an important protective interaction between the CVF and HIV-1 particle. In another study, the movement of HIV-1 through the CVF has been reported to be significantly slower at a pH of 4 and more rapid when the CVF was buffered to a pH of 6. Thus, the physical properties of the CVF impact how efficiently virus particles can traverse CVF and infect the vaginal or cervical epithelium. Decreased viscosity of CVF may render the mucin gel layer more permissive to penetration. In addition, CVF serves as a carrier for a broad array of antimicrobial peptides and proteins including lysozyme, lactoferrin, secretory leukocyte protease inhibitor, and human beta-defensins. Therefore, the protein content of the CVF is also a key component in the innate mucosal defense.
One mechanism of progestin-dependent contraceptive efficacy is believed to be thickening of cervical mucus and preventing the transport of sperm from the vagina into the uterus and fallopian tubes. The quality of the cervical mucus is dependent on reproductive hormones. Without the use of exogenous hormones, the first half of the menstrual cycle is characterized by increased estradiol levels and an increased amount of cervical mucus that is thin and watery to allow sperm penetration. In the second half of the menstrual cycle, predominated by increased progesterone levels, the cervical mucus becomes scant in amount, thick and opaque. Several studies have reported a thickening of cervical mucus associated with progestin-only contraceptive methods using subjective measures of cervical mucus quality such at ferning and spinbarkeit. Progestins subjectively thicken the cervical mucus within the cervical canal. However, less is known about the impact of reproductive hormones on cervicovaginal fluid, which provide a protective barrier over the vaginal epithelium. In addition to affecting the physical properties of the CVF, reproductive hormones mediate the biochemical content. Specifically, immunoglobulins, human beta-defensins, and secretory leukocyte protease inhibitor are lowest at midcycle when estradiol levels are elevated.
The effects of exogenous reproductive hormones on CVF could have clinical consequences. A few prospective, well-controlled studies have linked progestin-only injectable use to increased HIV risk. The effect of reproductive hormones on the physical or biochemical composition of the CVF has not been completely characterized, but changes in the mucin gel layer covering the vaginal epithelium is one possible biologic mechanism by which hormonal contraceptives could impact the risk of HIV acquisition.
Normal vaginal microbiota, characterized by a predominance of lactobacilli, is thought to be protective against sexually transmitted infection. The presence of bacterial vaginosis (BV) is associated with an increased risk of HIV acquisition in an HIV uninfected woman as well as increased risk of HIV transmission by an HIV infected women to an HIV uninfected male partner. BV is characterized by an overgrowth of anaerobic bacteria and a decreased colonization by Lactobacillus species. This overgrowth of anaerobic bacteria is associated with increased levels of bacterial proteases and glycosidase in CVF. Women with BV have higher levels of vaginal sialidases. Sialidases are considered virulence factor in bacterial vaginosis ; they clip the negatively charged sialic acid residues from the terminal end of the mucin oligosaccharides. Sialidase residues protect the oligosaccharide and the protein backbone of the mucin molecule from degradation by mucin-degrading enzymes. The negatively charged mucin molecules keep a rigid structure and trap pathogens, preventing them from reaching the vaginal epithelium. BV may cause thinning of the mucin gel layer thus impeding the capacity of the CVF to serve as a barrier against HIV infection.
To date, there have been few studies that have investigated the impact of reproductive hormones and vaginal flora on the physical and biochemical properties of the CVF. In the present study, we collected both cervicovaginal fluid (CVF) using a catamenial cup as well as cervicovaginal lavage (CVL) by washing the vaginal vault with sterile normal saline. The primary aim of this study was to characterize the impact of reproductive hormones on the viscosity, pH, density, osmolality, and protein content of CVF. Because of the small volume and technical difficulties associated with performance of assays with the CVF, assessment of the physical properties of the CVF samples was not feasible. Therefore, the viscosity, pH, and osmolality and density were measured only in the CVL samples. Because epidemiologic studies have linked BV and exogenous contraceptive use to increased HIV susceptibility, we hypothesized that the use of contraceptives, phase on menstrual cycle, menopausal status, and vaginal flora will impact the physical properties and protein content of CVL.
Materials and Methods
Study population
Following Institutional Review Board approval by the University of Pittsburgh, informed consent was obtained from healthy, asymptomatic, HIV-negative women who were either between 18-46 years of age or over the age of 50. We enrolled premenopausal women into the study who fell into the 5 following categories on the basis of contraceptive use by self-report: (1) not contracepting on days 1-14 of the menstrual cycle, (2) not contracepting on days 15-28 of the menstrual cycle, (3) using combined-oral contraceptive pills for at least 6 months, (4) using depot medroxyprogesterone acetate (DMPA) injections for at least 6 months, (5) using the levonorgestrel intrauterine device (LNG-IUD) for at least 1 month. A group of postmenopausal women was also recruited; menopause was defined as age greater than 50 years of age without any vaginal bleeding in the previous 1 year. Women were excluded from the study if they had been pregnant or breastfeeding within the last 90 days, had vaginal symptoms or evidence of vaginitis on clinical examination, had used vaginally applied products in the prior week, had used antibiotics in the 2 weeks prior, had undergone a hysterectomy, or had a positive rapid HIV test. In addition, postmenopausal women taking exogenous estrogen were also excluded. None of the postmenopausal women reported taking supplements containing phytoestrogens.
On enrollment, demographic information, medical, gynecologic, and sexual histories were collected from each participant. A vaginal swab for pH, wet mount microscopy, and Gram stain were collected. The catamenial cup was inserted into the vagina up to the cervix by the clinician and left in place for at least 45 minutes. The catamenial cup was removed and placed into a 50 mL conical vial for transport to the laboratory. The cervicovaginal fluid samples were centrifuged at 2000× g for 10 minutes. The protein laden material was removed and the volume was measured. Because of the small volumes and difficulty working with the CVF samples, we were unable to assess the physical properties of the CVF specimens. Then, these samples were stored at −70°C for future study. For collection of the cervicovaginal lavage (CVL), 10 mL of sterile normal saline was placed into the vagina, a lavage was performed for 1 minute, and placed into 15 mL conical vial with 100 μL of protease inhibitor (Sigma-Aldrich, St. Louis, MO). A cervical swab was collected for Neisseria gonorrhoeae and Chlamydia trachomatis testing using the Aptima GenProbe test (Hologic Gen-Probe Inc., San Diego, CA). Women found to have these pathogens or Trichomonas vaginalis by wet mount microscopy were excluded from the analysis. All samples were stored at −70°C until they were thawed for immediate testing. The osmolality (milliosmoles/kilogram, mOsm/kg), viscosity (centipoise, cP) of each sample was measured in triplicate using the Advanced Instruments MicroOsmometer Model 3320 (Advanced Instruments, Inc., Norwood, MA) and the Cambridge MicroSample Viscometer (Cambridge Viscosity, Medford, MA), respectively. Density was calculated by determining the weight in triplicate of 0.5 mL of sample. pH was measured using a Mettler Toledo pH meter. The Gram-stained vaginal smear was evaluated using Nugent criteria. Fisher exact test, Student’s t test, and 1-way analysis of variance with posthoc comparisons made using Bonferroni’s multiple comparisons procedure were used to assess statistical significance using IBM SPSS Statistical software version 20 (IBM Corp, Armonk, NY).
Results
Demographic characteristics
The demographic characteristics of the study population are summarized in Table 1 . The ages of women in the premenopausal groups were not statistically different (mean of 29 years); although the postmenopausal women had a mean age of 56 years. The majority of women were white and nonsmoking, with the exception of women using DMPA, who were more likely to identify themselves as black ( P < .001) and to report tobacco use ( P = .046). Women using hormonal contraceptives reported less frequent condom use as compared with women not using hormonal contraceptives ( P < .001).
Characteristic | Days 1-14 (n = 26) | Days 15-28 (n = 27) | OCPs (n = 27) | DMPA (n = 28) | LNG-IUD (n = 28) | Postmenopausal (n = 29) |
---|---|---|---|---|---|---|
Age (mean ± SD) | 29.8 ± 7.8 | 27.0 ± 6.5 | 28.6 ± 9.5 | 29.4 ± 6.1 | 29.0 ± 5.2 | 56.2 ± 7.0 |
BMI (mean ± SD) | 30.8 ± 9.8 | 25.1 ± 6.5 | 25.4 ± 5.9 | 27.3 ± 6.4 | 27.6 ± 6.2 | 30.6 ± 7.0 |
Race | ||||||
White | 14 (53.8%) | 14 (51.9%) | 24 (88.9%) | 9 (32.1%) | 22 (78.6%) | 22 (75.9%) |
Black | 10 (38.5%) | 7 (25.9%) | 2 (7.4%) | 19 (67.9%) | 5 (17.9%) | 7 (24.1%) |
Asian | 0 | 3 (11.1%) | 1 (3.7%) | 0 | 1 (3.6%) | 0 |
Other | 2 (7.7%) | 3 (11.1%) | 0 | 0 | 0 | 0 |
Condom use (most or all of the time) | 10 (38.5%) | 13 (48.1%) | 3 (11.1%) | 6 (21.4%) | 4 (14.3%) | 0 |
Current no. partners | ||||||
None | 12 (46.2%) | 5 (18.5%) | 7 (25.9%) | 7 (25.0%) | 4 (14.3%) | 14 (48.3%) |
1 | 12 (46.2%) | 21 (77.8%) | 19 (70.4%) | 20 (71.4%) | 22 (78.6%) | 15 (51.7%) |
≥2 | 2 (7.7%) | 1 (3.7%) | 1 (3.7%) | 1 (3.6%) | 2 (7.1%) | 0 |
Current smoker | 6 (23.1%) | 5 (18.5%) | 3 (11.1%) | 11 (39.3%) | 2 (7.1%) | 4 (13.8%) |
Bacterial vaginosis | 8 (30.7%) | 5 (18.5%) | 1 (3.7%) | 8 (28.6%) | 2 (2.1%) | N/A |
Effects of reproductive hormones and bacterial vaginosis
The physical properties of cervicovaginal lavage were different in postmenopausal women as compared with premenopausal women with respect to pH ( P < .001) and viscosity ( P < .001) of CVL, but not osmolality or density ( Table 2 ). The protein content of CVL was lower in postmenopausal women when compared with premenopausal women ( P < .001), and all premenopausal groups had a higher protein content in the CVL as compared with postmenopausal women, with the exception of the women using DMPA ( Figure 1 ). Women using the LNG-IUD had more CVL protein as compared with women using DMPA. No other differences in physical properties were detected between any of the premenopausal groups.
Variable | Density, g/mL | Osmolality, Osm/kg | pH | Viscosity, cP | Total protein, mg |
---|---|---|---|---|---|
Postmenopausal | 1.0 ± 0.01 | 431.1 ± 12.3 | 5.7 ± 0.9 | 1.1 ± 0.2 | 6.6 ± 2.8 |
Premenopausal (d 1-14) | 1.0 ± 0.01 | 432.3 ± 11.5 | 4.7 ± 0.4 | 1.3 ± 0.2 | 14.5 ± 10.5 |
Premenopausal (d 15-28) | 1.0 ± 0.01 | 434.5 ± 10.0 | 4.7 ± 0.8 | 1.4 ± 0.3 | 17.3 ± 8.8 |
Combined oral contraceptives | 1.0 ± 0.01 | 434.3 ± 11.3 | 4.5 ± 0.5 | 1.5 ± 0.4 | 17.8 ± 10.3 |
DMPA | 1.0 ± 0.01 | 431.7 ± 8.7 | 4.5 ± 0.3 | 1.4 ± 0.2 | 12.0 ± 7.8 |
LNG-IUD | 1.0 ± 0.01 | 431.6 ± 13.9 | 4.5 ± 0.4 | 1.6 ± 0.6 | 20.5 ± 11.1 |
P value a | .1 | .8 | <.001 b | <.001 | <.001 |