Bacterial Vaginosis and Spontaneous Clearance of Chlamydia trachomatis in the Longitudinal Study of Vaginal Flora

Abstract

Background

Up to 26% of urogenital Chlamydia trachomatis infections spontaneously resolve between detection and treatment. Mechanisms governing natural resolution are unknown. We examined whether bacterial vaginosis (BV) was associated with greater chlamydia persistence versus spontaneous clearance in a large, longitudinal study.

Methods

Between 1999 and 2003, the Longitudinal Study of Vaginal Flora followed reproductive-age women quarterly for 1 year. Baseline chlamydia screening and treatment were initiated after ligase chain reaction testing became available midstudy, and unscreened endocervical samples were tested after study completion. Chlamydia clearance and persistence were defined between consecutive visits without chlamydia-active antibiotics (n = 320 persistence/n = 310 clearance). Associations between Nugent score (0–3, no BV; 4–10, intermediate/BV), Amsel-BV, and chlamydia persistence versus clearance were modeled with alternating and conditional logistic regression.

Results

Of chlamydia cases, 48% spontaneously cleared by the next visit (310/630). Nugent-intermediate/BV was associated with higher odds of chlamydia persistence (adjusted odds ratio [aOR] = 1.89; 95% confidence interval [CI], 1.30–2.74), and the findings were similar for Amsel-BV (aOR 1.39; 95% CI, .99–1.96). The association between Nugent-intermediate/BV and chlamydia persistence was stronger in a within-participant analysis of 67 participants with both clearance/persistence intervals (aOR = 4.77; 95% CI, 1.39–16.35). BV symptoms did not affect any results.

Conclusions

BV is associated with greater chlamydia persistence. Optimizing the vaginal microbiome may promote chlamydia clearance.

A secondary analysis utilizing a large, observational cohort of reproductive-age women revealed 48% of untreated urogenital chlamydia visits spontaneously cleared within the next 12 weeks. Amsel- and Nugent-defined bacterial vaginosis were associated with greater odds of persistence versus spontaneous clearance.

Chlamydia trachomatis is the most common bacterial sexually transmitted infection (STI) in the United States; women account for 64% of yearly new cases, and black women have 5-fold higher rates than white women [1]. Up to 80% of chlamydia cases in women are asymptomatic [2] and may not seek treatment. Untreated chlamydia can lead to serious reproductive sequelae including pelvic inflammatory disease, ectopic pregnancy, and tubal factor infertility [3]. Rates of chlamydia continue to rise despite guidelines to screen all sexually active women under age 25 years annually—over 1.1 million cases were reported in women in 2019 for a rate of 694 cases per 100 000 women compared to 636 cases per 100 000 in 2015, although it is not known whether this rise reflects increased screening efforts and/or changes in the underlying chlamydia prevalence [1].

Spontaneous clearance of urogenital chlamydia in the absence of antibiotic treatment occurs in a substantial proportion of women. While sample sizes have been limited, prior studies have reported spontaneous clearance in 6%–26% of women between screening and the subsequent treatment visit (often 1–2 weeks), 16%–44% within approximately 10 weeks [4], up to 54% within 1 year, and 94% over 4 years [5, 6]. This phenomenon has also been termed natural clearance or spontaneous resolution. Spontaneous clearance between screening and treatment has also been observed for extragenital sites in women, including 50% of cases of oropharyngeal [7] and 16% of cases of anorectal [8] chlamydia. Persistent and recurrent urogenital infections increase risk for adverse reproductive sequelae [9]. However, compared to treated urogenital chlamydia infections, spontaneously cleared infections are associated with lower reinfection risk [10]. The mechanisms involved and factors associated with spontaneous clearance of urogenital chlamydia remain unclear. Some studies have reported that older age, time since detection, and lower bacterial load are associated with higher likelihood of spontaneous clearance [11–14], while others found no association [15, 16].

Bacterial vaginosis (BV) is another possible factor impacting clearance. Broadly characterized by a nonoptimal vaginal microbiota with low levels of gram-positive Lactobacillus, BV is associated with an increased risk of STI acquisition, including chlamydia [17, 18]. Some studies have found that those with BV or BV-associated bacteria have elevated levels of vaginal cytokines associated with T helper 2 (Th2) responses [19], which may downregulate protective Th1 responses involved in clearance [20]. Data in humans are limited, but animal models suggest Th2 responses are associated with chlamydia persistence and susceptibility [20]. There are other means by which BV may support chlamydia persistence, but data on specific mechanisms are limited. Host interferon-γ (IFN-γ)–mediated tryptophan starvation is another potential defense mechanism against chlamydia infection. In vitro studies have shown that indole, a metabolite produced by some BV-associated bacteria, can rescue chlamydia growth and infectivity after tryptophan deprivation [21–23], potentially enhancing chlamydia persistence. Vaginal indole levels were found to be higher in patients with BV [24]. Conversely, beneficial Lactobacillus spp., producing high D-lactic acid isomer concentrations, are generally abundant in the vaginal microbiota of those without BV, have protective effects on the vaginal epithelium and decrease chlamydia infectivity in vitro [25–27].

Gaining a better understanding of the natural history of chlamydia infection and factors affecting spontaneous clearance, such as BV, may ultimately lead to novel interventions that could facilitate clearance and decrease reinfection risk. However, there are limited opportunities to conduct natural history studies in modern cohorts due to screening and treatment standards. While some studies have utilized the interval between testing and initiating treatment to evaluate spontaneous clearance, most were conducted over a short time frame—usually under 2 weeks. The 1999–2003 National Institutes of Health (NIH) Longitudinal Study of Vaginal Flora (LSVF) offers a unique opportunity to study chlamydia natural history as baseline screening for chlamydia in LSVF was not initiated until the introduction and availability of ligase chain reaction (LCR) testing after recruitment commenced. Leveraging data from this study, we aim to determine whether BV assessed via Nugent score and Amsel criteria is associated with greater persistence versus spontaneous clearance of untreated chlamydia within 12-week intervals.

METHODS

This study utilized archived data from the NIH LSVF in which 3620 reproductive-age women were recruited from family planning and women's health clinics in Birmingham, Alabama and followed for 1 year, with clinical examinations occurring approximately every 3 months. Exclusion criteria included pregnancy, immunocompromised status, hysterectomy, pelvic radiotherapy, and current chronic antibiotic use [28].

Participants underwent a pelvic examination at each visit, which included vaginal swab sampling and cervicovaginal lavage collection. A dry Dacron swab was used to collect vaginal secretions for measuring pH and for Nugent Gram staining (0–3, low; 4–6, intermediate; 7–10, Nugent-BV) [29]. Symptomatic and asymptomatic BV was clinically diagnosed based on Amsel criteria (Amsel-BV) by trained research clinicians and query of symptoms [30]. Participants diagnosed with symptomatic Amsel-BV were offered treatment with oral metronidazole (500 mg twice a day for 7 days or 2 g once), or clindamycin (300 mg twice a day for 7 days), per 1998 Centers for Disease Control and Prevention treatment guidelines [31].

Endocervical swabs were collected at each clinical visit to screen for Neisseria gonorrhea (Thayer-Martin agar culture) and Trichomonas vaginalis (wet mount and InPouch culture), and participants were treated or referred to an STI clinic for treatment if positive. Baseline screening for C. trachomatis was initiated partway into LSVF recruitment (approximately July 2001) following the introduction of LCR testing for chlamydia (LCx assay, Abbott Laboratories). LCR was an early nucleic acid amplification test (NAAT) that had been newly licensed at that time and referenced in diagnostic considerations [31–33] during the early phases of LSVF recruitment. Prior to this, consent documents informed participants that asymptomatic chlamydia screening was not included in routine study procedures and indicated participants should follow up with their primary providers for screening. All endocervical swabs collected in the early part of the study, which had not been previously tested, were screened for C. trachomatis with LCR after study completion.

Antibiotic use was captured by including prescription records from study visits, as well as face-to-face interviews in which study clinicians queried participants on their antibiotic use in the intervals between study visits. We excluded participants in which chlamydia-active antibiotics or unknown antibiotics may have been used. Antibiotics were coded by classes and any macrolide, tetracycline, or fluoroquinolone antibiotics were considered active against chlamydia, regardless of the indication for treatment. Any instances of self-reported antibiotic use where the specific antibiotic could not be identified were included with chlamydia-active antibiotics. We also included participants’ self-report of chlamydia, pelvic inflammatory disease, or gonorrhea treatment from questionnaires, as well as study clinicians’ referral for treatment, as evidence of chlamydia-active antibiotic use.

Spontaneous clearance and persistence were defined between consecutive visit pairs, each of which included a positive chlamydia test at the index visit, followed by either a negative (clearance) or positive (persistence) chlamydia test on the next scheduled visit without evidence of (1) treatment or chlamydia-active antibiotic use in the interval between the index visit and the next visit, (2) in-study treatment in the 30 days prior to the index visit, and (3) self-reported chlamydia-active antibiotic use in the interval between the index visit and the previous visit.

We defined a dynamic subcohort within LSVF using the outcome of spontaneous clearance versus persistence (Supplementary Figure 1). Participants entered the cohort on any index visit with a chlamydia infection that spontaneously cleared or persisted at the next visit. Participants exited the cohort on any visit in which they (1) did not have chlamydia, (2) had chlamydia but there was evidence of chlamydia-active antibiotic use, or (3) were missing their next chlamydia test result. Participants could contribute multiple index visits, and thus multiple events of spontaneous clearance and/or persistence. Each pair of visits (index and next visit) was treated as 1 event, even if they occurred over consecutive visits. All spontaneous clearance and persistence events were eligible for the primary analysis.

We fit alternating logistic regressions (ALR) models with exchangeable log odds ratios as the primary analysis of the association between both Nugent-BV [34] (0–3, low; 4–6, intermediate; 7–10, Nugent-BV) and Amsel-BV at the index visit with persistence versus spontaneous clearance of chlamydia by the next visit 12 weeks later. We also performed a secondary analysis among the subset of participants contributing both persistent and spontaneously cleared infections using a conditional logistic regression model whereby each participant served as their own control.

We conducted several sensitivity analyses. First, the analysis was repeated using ALR after excluding participants who had prevalent chlamydia at their enrollment visit because the duration of infection was unknown for those cases. We also censored participants after their first instance of either spontaneous clearance/persistence and repeated the analysis using conventional logistic regression. Lastly, using BV status on the visit where spontaneous clearance/persistence was detected (Supplementary Figure 1, “next visit”), we used ALR to model the cross-sectional association between BV and persistence.

Age, race, body mass index at enrollment, prior history of chlamydia infection, and marital status at enrollment were considered as potential baseline confounders in the ALR and logistic regression models. Hormonal contraception use, vaginal douching, lubricant use, sexual activity (including vaginal, anal, and receptive oral sex, condom use (male or female), new or concurrent partners, and the number of partners), and gonorrhea or trichomonas coinfection on the index visit were evaluated as potential time-varying confounders in both the ALR, conditional logistic, and logistic regression models. All time-varying variables refer to behaviors in the interval between the index visit and the next visit.

The study was approved by the Institutional Review Boards at the University of Alabama School of Medicine and the Jefferson County Department of Health, the University of Maryland School of Medicine, and the National Institute of Child Health and Human Development. All participants provided written informed consent.

RESULTS

Chlamydia was detected at 9.6% of LSVF visits (1310/13 646), and 22.7% of participants (823/3620) had at least 1 chlamydia-positive visit; 680 chlamydia-positive visits were excluded from the analysis, including 260 for evidence of treatment, chlamydia-active, or unknown antibiotic use (Figure 1). Overall, 431 participants contributed 630 index visits with no documented chlamydia treatment, of which 310 (48%) spontaneously cleared within the next 12 weeks and 320 (52%) persisted at the next visit. Five index visits were missing data on Nugent score, and 1 index visit was missing data on Amsel-BV.

Figure 1.

Untreated chlamydia infections in the Longitudinal Study of Vaginal Flora.

Most participants self-reported being under age 25 years, black, and never married (Table 1). Approximately 25% reported ever being diagnosed with chlamydia, and 10% reported a chlamydia infection within the last year. Other self-reported reproductive tract infections in the year prior to enrollment were yeast infections, BV, and trichomonas. Of all chlamydia-positive index visits, 9.4% had trichomonas coinfections based on wet mount plus InPouch culture screening (59/629). Of those, 25 experienced chlamydia clearance and 34 experienced chlamydia persistence at the next visit. Ten chlamydia-positive index visits included coinfection with gonorrhea in which participants were referred to outside study treatment, but the participant did not subsequently report having been treated or using any chlamydia-active antibiotics at their next visit; 5 experienced chlamydia clearance and 5 experienced chlamydia persistence at the next visit. There was no association between chlamydia persistence and either coinfection with trichomonas (P = .44) or gonorrhea (P = .79).

Table 1.

Demographic, Behavioral, and Health Characteristics Among Participants With Untreated Chlamydia in the Longitudinal Study of Vaginal Flora, 1999–2002, N = 431

Characteristic	No. (%)
Demographics
Age, y
15–24	284 (66)
25–34	117 (27)
35–44	30 (7)
Race
Black	398 (92)
White	30 (7)
Multiracial	3 (1)
Marital status
Never married	342 (79)
Currently married	46 (11)
Previously married	40 (9)
Other	3 (1)
Education
Less than high school	165 (38)
High school	173 (40)
Some college	79 (18)
Bachelor's or beyond	14 (3)
Health and behaviors
Methods of birth control reported at enrollment
None	77 (18)
Birth control pill	104 (24)
Intrauterine device	1 (0)
Implant	6 (1)
Injection	101 (23)
Condom	173 (40)
Tubal ligation	61 (14)
Other (nonhormonal) method	19 (4)
Products used within 6 months of enrollment
Douche	210 (49)
Lubricant	51 (12)
Talcum powder	145 (34)
Towelettes or baby wipes	38 (9)
Sprays	76 (18)
Suppositories	15 (3)
Ever diagnosed
Diabetes	5 (1)
Bacterial vaginosis	125 (29)
Yeast infection	249 (58)
Trichomonas	114 (26)
Chlamydia	111 (26)
Pelvic inflammatory disease	19 (4)
Gonorrhea	69 (16)
Syphilis	5 (1)
Diagnosed in the year prior to enrollment
Bacterial vaginosis	78 (18)
Yeast infection	108 (25)
Trichomonas	48 (11)
Chlamydia	45 (10)
Pelvic inflammatory disease	4 (1)
Gonorrhea	14 (3)
Syphilis	1 (0)

Participants with untreated chlamydia and an intermediate (4–6) or BV (7–10) index visit Nugent score had 1.7-fold (95% confidence interval [CI], 1.10–2.81) and 1.9-fold (95% CI, 1.32–2.89) higher odds of chlamydia persistence versus spontaneous clearance at the next visit, respectively, compared to participants with a low Nugent score (0–3), adjusting for age, current hormonal contraception use, any condom use in the interval between visits, and marital status (Table 2). Similarly, participants with Amsel-BV had 1.4-fold (95% CI, .99–1.96) higher odds of persistence versus spontaneous clearance at the next visit adjusting for those same factors (Table 2). The following potential confounders were assessed in the interval between the chlamydia-positive index visit and the next visit but were ultimately not associated with chlamydia persistence in the final adjusted models: race, lifetime history of chlamydia, index visit coinfection with either gonorrhea or trichomonas, sexual activity (vaginal, anal, or oral), number of sexual partners, new or concurrent sexual partners, use of intravaginal products such as douches or lubricants, and treatment for symptomatic BV. There was no effect modification by BV symptoms in the Nugent-BV models, and no difference between the estimates for symptomatic versus asymptomatic in the Amsel-BV models.

Table 2.

Association Between Bacterial Vaginosis and Persistence Versus Spontaneous Clearance of Untreated Chlamydia

Bacterial Vaginosis	No.a	OR	95% CI	P Value	aORb	95% CI	P Value
Nugent score
Normal, 0–3	179	Ref	…	…	Ref	…	…
Intermediate/BV, 4–10	446	1.75	1.22–2.50	.002	1.89	1.30–2.74	.001
Intermediate	139	1.63	1.04–2.55	.03	1.76	1.10–2.81	.02
Nugent-BV	307	1.81	1.24–2.63	.002	1.95	1.32–2.89	.001
Amsel-BV
No BV	310	Ref	…	…	Ref	…	…
Any Amsel-BV	320	1.34	.97–1.86	.08	1.39	.99–1.96	.07
Asymptomatic	288	1.32	.95–1.85	.10	1.37	.97–1.95	.08
Symptomaticc	31	1.54	.75–3.18	.24	1.63	.71–3.74	.25

Abbreviations: aOR, adjusted odds ratio; BV, bacterial vaginosis; CI, confidence interval; Ref, reference.

Number of events. Three persistence events and 2 spontaneous clearance events are missing index visit Nugent score, 1 persistence event is missing index visit Amsel-BV status.

Adjusted for age (younger than 25 years vs 25 years and older), use of hormonal contraception, condom use (male or female), marital status (currently or ever married vs never married).

Treated with either 2 g oral metronidazole once (n = 3), or 500 mg oral metronidazole twice a day for 7 days (n = 27). No treatment information given for 1 case of symptomatic Amsel-BV.

We found a similar association between index visit Nugent-BV and chlamydia persistence after excluding participants who entered the study at enrollment with untreated chlamydia infections of unknown durations (adjusted odds ratio [aOR], 1.88; 95% CI, 1.06–3.33; Supplementary Table 1), and after censoring participants following their first instance of either spontaneous clearance or persistence (aOR, 1.82; 95% CI, 1.11–2.99; Supplementary Table 2). There was no strong evidence for an association between index visit Amsel-BV and persistence in either sensitivity analysis. Utilizing BV data from the visit after chlamydia detection (ie, the next visit, Supplementary Figure 1), we found no cross-sectional association between Nugent score or Amsel-BV with persistence (Supplementary Table 3).

Sixty-seven participants experienced both spontaneous clearance and persistence during the 1-year follow-up, allowing us to contrast clearance and persistence events using each participant as their own control. Of the 67 participants, each contributed 1 spontaneous clearance event, and 44, 17, and 5 participants contributed 1, 2, and 3 persistence visits, respectively, for a total of 67 spontaneous clearance and 94 persistence events. Given findings from the primary analysis and the reduced sample size of this within-participant analysis, the Nugent score outcome was collapsed as Nugent-intermediate/BV (score 4–10) versus low Nugent (score 0–3). Conditional logistic regression revealed that participants had a 4.8-fold higher odds of having Nugent-intermediate/BV scores when they had persistence, compared to when they had clearance (aOR, 4.77; 95% CI, 1.39–16.35), adjusting for sexual activity (any vaginal, anal, and/or oral intercourse vs none), unprotected sex, and new sexual partners in the interval between the index visit and next visit (Table 3).

Table 3.

Within-Participant Association Between Bacterial Vaginosis and Persistence (n = 94) Versus Spontaneous Clearance (n = 67) of Untreated Chlamydia: Case-Crossover Analysis in 67 Participants

Bacterial Vaginosis	No.a	OR	95% CI	P Value	aORb	95% CI	P Value
Nugent score
Normal, 0–3	53	Ref	…	…	Ref	…	…
Intermediate/BV, 4–10	107	4.20	1.38–12.84	.01	4.77	1.39–16.35	.01
Amsel-BV
No Amsel-BV	84	Ref	…	…	Ref	…	…
Any Amsel-BV	77	1.52	.74–3.14	.25	1.46	.68–3.11	.33

Abbreviations: aOR, adjusted odds ratio; BV, bacterial vaginosis; CI, confidence interval; Ref, reference.

Number of events. One spontaneous clearance event is missing index visit Nugent score (n = 66 participants included).

Adjusted for any interval sexual activity, new sex partners, and condomless sex.

DISCUSSION

This study evaluated the association between BV and spontaneous clearance of chlamydia in a large, observational cohort study. Overall, we observed 48% of index visit urogenital chlamydia infections spontaneously cleared at the next visit, approximately 12 weeks later. This finding is consistent with prior studies in women, which have reported rates of spontaneous clearance of 6%–26% in the interval between testing and treatment (often within 2 weeks) [4, 8] and 54% at 1 year [5].

The vaginal microbiota play an important role in regulating sexual and reproductive health [35]. An optimal vaginal microbiota is dominated by Lactobacillus species that produce lactic acid, which reduces pH, promotes anti-inflammatory immune responses, is associated with reduced acquisition of STIs [17], and appears to restrict host cell adhesion and infectivity by chlamydia [36, 37]. One study suggested a pH-dependent effect of lactic acid on chlamydia [38], and another in vitro study detailed how D-lactic acid (produced by Lactobacillus crispatus, Lactobacillus gasseri, and Lactobacillus jensenii, but not Lactobacillus iners) led to a reduction of both epithelial host cell proliferation and subsequent chlamydia infection [26]. It is plausible that these mechanisms may aid in the spontaneous clearance of chlamydia, whereas BV and other nonoptimal vaginal microbiota, such as Nugent-intermediate scores or communities with a high relative abundance of L. iners, might instead be associated with persistence.

We found a nearly 2-fold higher odds of persistence of untreated chlamydia 12 weeks following Nugent-intermediate/BV scores versus low scores. There was stronger evidence and a greater magnitude of association assessing BV with Nugent score over Amsel-BV. Compared to Amsel's criteria, Nugent score has a higher sensitivity and lower specificity for BV [39]; Amsel-BV may be less likely to capture all nonoptimal states of the vaginal microbiota. In our analysis, 92% of the participants with a low Nugent score also had no Amsel-BV, but 47% of the participants with no Amsel-BV had intermediate/BV Nugent scores. We were able to conduct a within-participant crossover analysis by comparing a participant's Nugent score during clearance intervals to their score during their own persistence intervals, which allowed us to control for confounding by time-invariant covariates. In both the within- and between-participant models, intermediate or Nugent-BV scores at the index visit were associated with greater chlamydia persistence, although the magnitude of the association was stronger when comparing participants to themselves (aOR 4.8 vs aOR 1.9). This could indicate unmeasured confounding by some time-invariant covariate in the full population model, or that participants who experience both chlamydia persistence and clearance differ in some way from those who do not.

There is limited prior data on the associations between BV and chlamydia clearance. Sheffield et al found no association between Nugent score or vaginal pH with chlamydia persistence but the cohort study was limited to pregnant women, only 18 of whom had BV [12]. A recent pilot study assessed the association between the vaginal microbiota and chlamydia clearance in 55 participants and found no significant differences in vaginal microbiota composition, Nugent score, or Amsel-BV between those who had persistent or spontaneously cleared infections [40]. However, this study evaluated the clearance visit, not the visit antecedent to clearance as we present. We also found no association between BV and chlamydia persistence when BV status was evaluated concurrently with the persistence or clearance visit.

There was no difference in the association between BV and chlamydia persistence outcomes comparing those with asymptomatic versus symptomatic Amsel-BV. Patients with asymptomatic BV also present with at least 3 of 4 Amsel's criteria, but they do not report vulvovaginal symptoms on direct questioning [41]. If index visit BV treatment was successful in restoring the vaginal microbiota to an optimal state that may subsequently aid in chlamydia clearance, then we might have paradoxically expected to observe more spontaneous clearance among participants with symptomatic BV, rather than asymptomatic BV. We did not observe such findings; however, antibiotic treatment for BV has a range of effectiveness. The cure rates for a single 2-g dose or 500 mg twice per day 7-day course of metronidazole are 46% at 21 days posttreatment and 64% at 1 month posttreatment, respectively [42–44].

This study was large compared to previous studies on the natural history of untreated urogenital chlamydia in women. Both BV and chlamydia were measured at all study visits for 1 year, which allowed us to consider multiple 12-week intervals per participant. We were unable to evaluate time to clearance due to the dynamic subcohort study design (Supplementary Figure 1) and because the duration of the infection was unknown for participants who entered the subcohort with chlamydia at enrollment. However, a sensitivity analysis excluding those with chlamydia at enrollment did not affect findings. We utilized STI treatment records and face-to-face interviews with trained nurses to identify antibiotic use that was chlamydia active at and between study visits. However, there are disadvantages to self-reported antibiotic use. One study comparing self-report to prescription records found that 80% could correctly recall having taken an antibiotic in the prior 4 months, but 60% could not recall the name [45]. In our study, participants were able to specify the antibiotic class or name for 73% of antibiotic use reports, although we have no comparison to pharmacy records. We conservatively excluded visits in which antibiotic use was reported but the specific drug was unknown.

We did not have data on rectal chlamydia, and short-term rates of spontaneous clearance may differ between those with vaginal only, rectal only, and concurrent vaginal/rectal chlamydia infections [8]. It is also possible that we observed some nonviable chlamydia (eg, nucleic acid remnants) or detection of chlamydia following an exposure that did not lead to an infection, rather than true spontaneous clearance. Some persistent infections in this analysis may also reflect spontaneous clearance followed by reinfection within 12 weeks. Geisler et al reported 5% of persistent chlamydia infections had a different OmpA genotype comparing the screening visit to a follow-up visit within 60 days, suggesting new, rather than persisting, infection [46]. Future studies on chlamydia natural history could include genotyping or viability testing to strengthen the characterization of persistent and cleared infections [47]. Lastly, the diagnostic tests utilized to detect STIs in LSVF are somewhat less sensitive compared to modern assays, although sensitivity is still high. LCR on endocervical swabs has a sensitivity of 96% for the detection of chlamydia when compared to an early PCR test [48], but the sensitivity of modern NAATs exceeds 98% for both self- and clinician-collected vaginal swabs [49]. Regarding trichomonas, the sensitivity of wet-mount microscopy (51%) and InPouch culture (75%) are significantly lower compared to modern NAATs (82%–98%) [50]; however, the sensitivity of wet-mount microscopy plus culture as used in this study has not been previously reported to our knowledge.

This observational study furthers our understanding of how BV, defined by Nugent score and Amsel criteria, may affect the natural history of chlamydia regarding spontaneous clearance. Future studies may utilize more modern molecular methods to explore microenvironmental factors that affect spontaneous chlamydia clearance, including the cervicovaginal microbiome and metabolome, mucosal immune environment, and features of the chlamydia organism. Such studies could provide insight into the specific mechanisms supporting spontaneous clearance and persistence and may inform novel interventions for chlamydia prevention, such as using live biotherapeutics to modulate, optimize, and sustain an optimal vaginal microbiota that is protective against chlamydia acquisition and persistence.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases at the National Institutes of Health (grant numbers R01-AI116799 and R01-AI167629 to R. M. B.) and the National Institute of Child Health and Human Development (grant number Z01-HD002535 to M. K.).

Presented in part: Infectious Disease Society for Obstetrics and Gynecology Annual Research Meeting, August 2022, Boston, MA.