Twenty-five to 35% of women who are exposed to N. gonorrhoeae during vaginal intercourse with an infected man resist cervicovaginal infection, a rate that is consistent across multiple studies.1–3 Why some women resist infection is not fully understood, as studies of gonococcal infection have focused almost exclusively on cases, that is, failures to resist. Understanding why a quarter to a third of exposed women resists infection may lead to novel control measures.
A longitudinal study of 227 sex workers resident in a small community in Kenya who experienced frequent gonococcal infections, found that the duration of prostitution was inversely related to the frequency of infection.4 The mechanism for this apparent protection was not sought.
Menstruation at time of presentation to clinic is associated with gonococcal infection5,6; however, previous studies have not investigated the effect of menstrual cycle phase on infection rate at time of sexual exposure to N. gonorrhoeae. This study sought modifiable biological bases for infection resistance by comparing women in the same core mixing group7 who did or did not become infected after exposure.
Cohort Study Design and Recruitment
We recruited female sexual contacts of men known to have gonorrhea from either of the 2 Baltimore City Health Department (BCHD) clinics. Each exposed woman was given a Partner Notification Card (PNC) by a male sexual partner, the index patient, who previously had been diagnosed with gonorrhea at one of the clinic sites. At the BCHD site of diagnosis, the index patient had been given PNC(s) for distribution to his sexual partners. The PNC stated that she (the index patient's sexual partner) had been exposed to a sexually transmitted infection and that she should report to the Baltimore City Health Department for evaluation and treatment. The female contact presented to a BCHD STI clinic site with her PNC and was offered enrollment as a N. gonorrhoeae contact in this study. Sixty-one female contacts were enrolled. Historical, physical, and laboratory attributes were collected from each woman and recorded, with the goal of comparing these attributes between infected female contacts and contacts who resisted infection during sexual exposure.
Enrolled women provided information about sexual practices, including age at sexual debut, number of lifetime partners, type of sexual practice, history of prior sexually transmitted infections (including past gonococcal infections), douching, date of exposure to N. gonorrhoeae, date of last sexual activity, and menstrual cycle history (including date of last menstrual period, and use of any hormonal contraception). Symptoms of infection by patient report, such as vaginal discharge or other urogenital symptoms were also recorded.
The study was approved by the Johns Hopkins University Institutional Review Board (NA-00004860), and written informed consent was obtained from all participants prior to study enrollment. Participants were compensated with gift cards of minimal value for their time.
Physical Examination after Enrollment
Clinic staff did a pelvic examination on all women enrolled and signs of infection were noted. Vaginal and cervical samples were collected during examination.
Laboratory Data Collected
Vaginal pH of 56 subjects was measured at the time of the examination with pH indicator strips (range, 4.0–7.0; Fisherbrand pH Test Paper).8
Wet mounts from vaginal swabs were examined for inflammation (white blood cells), yeast, Trichomonas vaginalis, and clue cells. Whiff tests were done for evaluation of bacterial vaginosis (BV) by Amsel criteria.9 Concurrent Chlamydia trachomatis infections were sought with use of a nucleic acid amplification test (NAAT).
Cervical cultures for N. gonorrhoeae were obtained along with additional urine and cervical swabs for gonococcal NAATs. Cultures were plated on modified Thayer-Martin medium and processed according to the clinics' Standard Operating Procedures. Gonococcal NAATs (Hologic/GenProbe Aptima Combo 2) were run on urine specimens only from women whose cultures were negative. Only women who had both a negative gonococcal culture and a negative NAAT were considered negative.
Menstrual Cycle Phase at Exposure
The day of menstrual cycle on the day of exposure to N. gonorrhoeae was calculated from patient report at presentation to clinic of date of exposure and date of last menstrual period. Day of menstrual cycle at time of exposure was divided into 3 phases corresponding to levels of estrogen and progesterone: follicular, ovulation, and luteal. These designations were used in prior studies investigating the link between menstrual cycle phase and HIV risk.10 The “follicular” phase, which includes active menstruation during days 1 to 5, encompassed days 1 to 11. “Ovulation” included days 12 to 17. The “luteal” phase included days 18-35.10 Any participant with a menstrual cycle greater than 35 days was excluded from this analysis, as menstrual cycle day determination was likely to be inaccurate or menstrual cycle anovulatory.
Associations between gonococcal infection and single behavioral factors, physical signs and symptoms, or laboratory data gathered at presentation were investigated using χ2, Fisher exact or bivariate logistic regression in STATA 14 (STATA Corp, College Park, TX).
Multiple logistic regression (MLR) was then used to investigate the relationship between gonococcal infection and covariates associated with risk at time of exposure and separately at time of presentation to clinic. Covariates for these models were considered based on bivariate logistic regressions (behavioral factors, physical signs and symptoms, and laboratory data), previous literature, and biological plausibility.
Information obtained from subjects who reported to clinic more than 35 days after exposure (roughly one menstrual cycle) was considered less reliable, and risk factor analyses at time of sexual exposure was restricted to those who reported within 35 days. Analyses of the effect on risk of gonococcal infection of vaginal pH, which varies over the menstrual cycle,11,12 were restricted initially to those women who reported within 35 days of exposure. A subsequent analysis was restricted to those subjects who reported within 5 days of exposure in order to better ascertain the effect on infectivity of vaginal pH at the time of exposure.
Manual backward model selection with uncentered variance inflation factor analysis was used to derive the most parsimonious models.
Gonococcal Infections in Contacts
All 61 of the enrolled women acknowledged vaginal intercourse with an index male patient; 35 (57.4%) of the women had a cervical culture that grew N. gonorrhoeae, and another 3 (4.9%) women had a urine NAAT that was positive for N. gonorrhoeae. The sensitivity of culture was 92% (taking into account those negative by culture but positive by NAAT), and 62% of exposed women had a genital infection. Oral-penile sex was reported by 28 (46%) women and anal sex by 3 (4.9%).(Table 1)13 One woman who reported oral-penile sex had a positive throat culture, but negative cervical culture and NAAT; another woman who reported anal sex had a positive rectal culture, but negative cervical culture and NAAT. These 2 women were considered to have resisted infection in analyses of risk factors for cervical infection.
Demographic and Historical Comparisons
Ninety-three percent of women enrolled were African American. Of those who presented within 35 days of exposure (46 women) 91% were black (Table 1).13
Among the 61 subjects, those who resisted infection did not differ in age from those who became infected (means (ranges): 25 (19–38) and 24 (18–39), respectively). Among the 46 women who presented within 35 days of exposure, those who were infected had a mean age of 23 (SD, 4.2), and those who resisted infection had a mean age of 26 years (SD, 6.1).
Among all enrolled women as well as those who presented within 35 days, there was no significant difference between infected and uninfected women in age at sexual debut (mean, 15 for both) or in number of lifetime partners (mean of infected women, 22.3; SD, 39.1; mean who resisted infection, 19.6; SD, 39.7). This was also true for subjects who presented within 35 days of exposure.
There also was no difference in the number of days between exposure and reporting to the clinic between the 38 infected women (mean, 24.3 days; median, 11 days) and the 23 uninfected subjects (mean, 25.8 days; median, 8 days).
Sexual practices did not differ between women with or without cervical infections (Table 1).13 All women reported having had vaginal intercourse with an index male patient. Forty-six percent of the 61 women (N = 28) and 52% of the women who reported to clinic within 35 days of exposure (N = 23) acknowledged receptive oral-penile sex. Only 1 of the 28 women who acknowledged receptive oral-penile sex had a positive pharyngeal culture; pharyngeal cultures were not specified in the study design and were not routinely obtained by the clinic. In all participants, anal intercourse was acknowledged by only 3 (4.9%) women (Table 1)13; 2 of these women had a positive rectal culture for N. gonorrhoeae. Rectal cultures were obtained only from women who reported anal intercourse. NAATs were not used by the BCHD clinic as they were not FDA approved for extra-genital sites.
Reported symptoms were not significantly different between infected and uninfected women.(Table 1)13
There was no significant association between Trichomonas or yeast infection and gonococcal infection among the 61 enrolled women, although numbers of confections were low (Trichomonas, 8; yeast, 3). Seven of the 8 women who were infected with C. trachomatis were concurrently infected with N. gonorrhoeae.
Bacterial vaginosis by aggregate Amsel Criteria (3 of 4 criteria) was not associated with gonococcal infection among the 61 women (Table 2).
Cervical Discharge and Inflammation on Wet Mount
At time of presentation, clinicians observed cervical discharge in 33% of the 61 women (42% of infected women and 17% of those not infected). Those infected were significantly more likely to have cervical discharge on pelvic examination than those not infected (P = 0.05, bivariate logistic regression).
Similarly, gonococcal infection at presentation to clinic was associated with the degree of inflammation, as judged by the presence and number of leukocytes seen on wet mount of the cervix and graded by the clinician as 1 to 4. All 11 wet mounts with the highest degree of inflammation (4) were obtained from infected women (χ2, P = 0.02). There was no significant association between inflammation on wet mount and a patient's report of vaginal discharge (P = 0.68; χ2).
MLR of Gonococcal Cervical Infection and Vaginal pH
Among women presenting to clinic within 35 days of exposure (n = 42), a vaginal pH ≥ 4.5 was significantly associated with infection at presentation (adjusted odds ratio, 5.5; P = 0.037) when compared with pH less than 4.5, after adjusting for age. Similarly among women in this group, a vaginal pH of 5.0 or greater was significantly associated with infection at presentation (adjusted odds ratio, 13.1; P = 0.017).
The closer the measurement of vaginal pH was to the time of exposure, the greater was the difference between infection resistance and infection acquisition. Vaginal pH was measured within 5 days of exposure in 17 women (Fig. 1); 10 within 2 days, 3 at 3 days, and at both 4 and 5 days. All 9 of the women whose vaginal pH was ≥5 were infected, whereas only 3 of the 8 women whose vaginal pH was less than 5.0 were infected (P = 0.009; Fisher exact test).
MLR of Hormonal Risk Factors Associated With Gonococcal Cervical Infections
Among women presenting to clinic within 35 days of exposure (N = 33), those actively menstruating during sexual exposure were more likely to be GC infected than those exposed during other days of their cycle (adjusted odds ratio = 12.5, P = 0.05) after adjusting for age.
No significant association between GC infection and other phases of the menstrual cycle (follicular, ovulatory, or luteal) was found. However, this analysis was limited by a small sample of women with whom we could calculate menstrual cycle phase at point of sexual exposure to GC.
No difference between gonococcal infection and resistance was seen between those enrolled using hormonal contraceptives (HC) and those not using HC (χ2, P = 0.64). This was also true for those using progesterone-only HC (no estrogen component), medroxyprogesterone depo injection, and levonorgesterol-releasing IUDs, (χ2, P = 0.451); 70% of those using HC used progesterone-only methods (7/10). The other 3 women enrolled used combined estrogen/progesterone pills of varying formulations. That rate of hormonal contraception use, however, was quite low (10/61).
Previous studies have found that 25% to 35% of women who are exposed to N. gonorrhoeae through vaginal intercourse with an infected man resist cervical infection1–3; in our study, 38% of the women who presented with a partner notification card resisted—within the expected range.7 Our sample size was in the midrange of samples sizes in previous similar studies (26–112),1–3 that is, the women who resisted infection came from the same core-mixing group as those who did not resist infection reduces the potential contribution of gonococcal strain variations. The finding that a normal vaginal pH of 4 to 4.5 at the time of exposure is associated with a degree of resistance to infection was unexpected, but is consistent with the study of Pettit et al, where it was found that N. gonorrhoeae survive poorly at pH less than 5.015. Although gonococci do have an acid tolerance system that is apparent in vitro,14 it is unlikely that the bacteria could activate it in the short time between attachment to the uterine cervix and parasite-directed engulfment by cervical epithelial cells.15
A limitation of the study was the measurement of vaginal pH at the time of examination, an average of 25 days after exposure. It would, of course, have been ideal (albeit impossible) to measure vaginal pH on the day of exposure. However, despite the lag between exposure and the measurement of vaginal pH, a vaginal pH of 4.5 or greater at presentation was significantly associated with infection at presentation for those women reporting to clinic within 35 days of exposure.
This finding lead to an analysis of the 17 women who presented within 5 days of exposure, 10 of whom reported within 2 days. All 9 women whose vaginal pH was 5 or greater were infected, whereas only 3 of the 8 women whose vaginal pH was less than 5.0 were infected. This suggests that lowering the vaginal pH with vaginal rings, estrogen/progesterone, and proton releasing, or intravaginal gels could be a fruitful method of increasing resistance to gonococcal infection. This also would address rising antibiotic resistance in gonococci.
The association of gonococcal infection with active menstruation at time of presentation to clinic has been found in previous studies,6 but no studies have investigated potential effects of menstrual cycle phase at time of exposure on susceptibility to infection. We found an association between exposure during the active menstruation and risk of infection. Menstruation, however, may be a surrogate for increased vaginal pH.
Vaginal pH is fairly stable in individual women between menstrual periods, but rises as much as 2 pH units with menses.11,12,16 An acidic vaginal pH is maintained, in part, by estrogen-responsive epithelial cell proton pumps.17 Estrogen levels are lowest during the late luteal phase (directly before active menstruation). As estrogen levels fall, the vaginal pH increases. During the menstrual phases in which estrogen levels are higher (ovulation and early luteal phase), vaginal pH returns rapidly to pre-menstrual levels.
The availability of iron also may contribute to the increased risk of gonococcal infection during menstruation, as iron is necessary for gonococcal growth and proliferation.18 During the late luteal phase, levels of lactoferrin in vaginal mucus rapidly decline and are negligible by day 25 of the cycle.19 Low levels of lactoferrin would lead to higher levels of iron in vaginal mucus. Low concentrations of lactoferrin and the presence of menstrual blood would provide abundant iron.
Although Platt et al3 found an increase in risk of infection with multiple exposures, we did not ask how many exposures the subjects had during the index encounter. However, there are no data that suggest that women increased sexual exposure during menstruation.
The absence of an association of prior gonococcal infections and resistance is notable, given the observations of Plummer et al4 that repeated gonococcal infections reduced risk of subsequent infection among Kenyan sex workers. The women in our study, however, mostly reported only a single prior infection, whereas the Kenyan sex workers had had many previous infections.
Taken together, our data and those of Ness et al20 and Das and Allan,21 clarify that BV and gonococcal infections are independently associated with a high vaginal pH, but not with one another. A high vaginal pH heightens the risk of incident gonococcal infections in the absence of clinical BV.
Ness et al20 reported that although BV, gonococcal and chlamydial infections commonly co-occurred in African American women, BV did not increase the risk of incident infection with either organism, and Das and Allan, with use of a prospective case-control study, found a highly significant association of high vaginal pH with gonococcal and chlamydial infection in the absence of BV among predominantly white women.21
Neither Ness et al20 nor Das and Allan compared infected women with similar non-infected controls.21 Our study involved both exposed women who resisted infection and those who did not, and predominantly involved African American women. We found no association between BV, as judged by the presence of at least 3 of the 4 Amsel criteria,9 and risk of gonococcal infection.
The results of our study are very similar to those reported by Das and Allan.21 Only 12.3% of the cases in their study had a diagnosis of BV,21 whereas 28% of the women in our study had BV.9 The prevalence of BV would be expected to be higher in our study, as almost all of the women were African American and incidence of BV is higher in African American women.22 The association of both BV and gonococcal infections with an elevated vaginal pH explains why BV was a strong predictor of both gonococcal and chlamydial infections among subjects who reported recent exposure to a male partner with urethritis.23
In summary, our data show that it might be possible to lower risk of gonococcal infection by normalizing vaginal pH. Additionally, hormonal regulation with various contraceptive agents could possibly be used to lower the risk of gonococcal infection through modulation of the vaginal pH, as well as suppression of active menstruation.
1. Kiliç D, Başar MM, Kaygusuz S, et al. Prevalence and treatment of Chlamydia trachomatis
, Ureaplasma urealyticum
and Mycoplasma hominis
in patients with nongonococcal urethritis. Jpn J Infect 2004.
2. Lycke E, Lowhagen G-B, Hallhagen G, et al. The risk of transmission of genital Chlamydia trachomatis
infection is less than that of genital Neisseria gonorrhoeae
infection. Sex Transm Dis 7(1):6–10.
3. Platt R, Rice PA, McCormack WM. Risk of acquiring gonorrhea and prevalence of abnormal adnexal findings among women recently exposed to gonorrhea. JAMA 1983; 250(23):3205–3209.
4. Plummer FA, Simonsen JN, Chubb H, Slaney L, Kimata J, Bosire M, et al. Epidemiologic evidence for the development of serovar-specific immunity after gonococcal infection. J Clin. 1989; 83:1472–6.
5. Koch ML. A Study of cervical cultures taken in cases of acute gonorrhea with special reference to the phases of the menstrual cycle. Am J Obstet Gynecol Mosby 1947; 54(5):861–866.
6. McCormack WM, Reynolds GH. Effect of menstrual cycle and method of contraception on recovery of Neisseria gonorrhoeae
. JAMA 1982; 247(9):1292–1294.
7. Thomas JC, Tucker MJ. The development and use of the concept of a sexually transmitted disease core. J Infect Dis 1996; 174(Suppl 2):S134–S143.
8. Strauss RA, Eucker B, Savitz DA, et al. Diagnosis of bacterial vaginosis from self-obtained vaginal swabs. Infect Dis Obstet Gynecol 2005; 13(1):31–35.
9. Amsel R, Totten P, Chen K, et al. Nonspecific vaginitis. Diagnostic criteria and microbial and epidemiologic associations. Am J Med 1983; 74(1):A28.
10. Ghanem KG, Shah N, Klein RS, et al. Influence of sex hormones, HIV status, and concomitant sexually transmitted infection on cervicovaginal inflammation. J Infect Dis 2005; 191(3):358–366.
11. Santos Santiago dos GL, Cools P, Verstraelen H, Trog M, Missine G, Aila El N, et al. Longitudinal study of the dynamics of vaginal microflora during two consecutive menstrual cycles. PLoS ONE. Public Library of Science; 2011; 6:e28180.
12. Wagner G, Ottesen B. Vaginal physiology during menstruation. Ann Intern Med 1982; 96(6 Pt 2):921–923.
13. Rose E, DiClemente RJ, Wingood GM, et al. The validity of teens“ and young adults” self-reported condom use. Arch Pediatr Adolesc Med 2009; 163(1):61–64.
14. Pettit RK, McAllister SC, Hamer TA. Response of gonococcal clinical isolates to acidic conditions. J Med Microbiol 1999; 48(2):149–156.
15. Griffiss J, Lammel CJ, Wang J. NDI, 1999. Neisseria gonorrhoeae
coordinately uses Pili and Opa to activate HEC-1-B cell microvilli, which causes engulfment of the gonococci. Am Soc Microbiol.
16. Ravel J, Brotman RM, Gajer P, et al. Daily temporal Dynamics of vaginal microbiota before, during and after episodes of bacterial vaginosis. Microbiome 2013; 1:29. Vol. 1, 1 p.
17. Gorodeski GI, Hopfer U, Liu CC, et al. Estrogen acidifies vaginal ph by up-regulation of proton secretion via the apical membrane of vaginal-ectocervical epithelial cells. Endocrinology 2005; 146(2):816–824.
18. McKenna WR, Mickelsen PA, Sparling PF, et al. Iron uptake from lactoferrin and transferrin by Neisseria gonorrhoeae
. Infect Immun 1988; 56(4):785–791.
19. Preliminary observations on lactoferrin. Am J Obstet Gynecol 1987 - PubMed - NCBI.
20. Ness RB, Kip KE, Soper DE, et al. Bacterial vaginosis (BV) and the risk of incident gonococcal or chlamydial genital infection in a predominantly black population. Sex Transm Dis 2005; 32(7):413–417.
21. Das S, Allan S. Higher Vaginal pH is associated with Neisseria gonorrhoeae
and Chlamydia trachomatis
infection in a predominantly white population. Sex Transm Dis 2006; 33(8):527–528.
22. Ravel J, Gajer P, Abdo Z, et al. Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci USA 2011; 108(Suppl 1):4680–4687.
23. Wiesenfeld HC, Hillier SL, Krohn MA, et al. Bacterial vaginosis is a strong predictor of Neisseria gonorrhoeae
and Chlamydia trachomatis
infection. Clin Infect Dis 2003; 36(5):663–668.