Helms, Donna J. MPH*; Mosure, Debra J. PhD†; Metcalf, Carol A. MBChB, MPH‡; Douglas, John M. Jr MD*; Malotte, C Kevin DrPH§; Paul, Sindy M. MD, MPH∥; Peterman, Thomas A. MD*
TRICHOMONAS VAGINALIS IS ONE of the most common sexually transmitted infections (STI) in the world and thought to be the most common nonviral STI in the United States. There are an estimated 7 million new cases of T. vaginalis occurring annually in women in the United States, with an estimated prevalence of 3.1%.1,2
Trichomonas infections range from an asymptomatic carrier state to acute inflammatory disease, with up to 50% of infected women being asymptomatic.3,4 Asymptomatic T. vaginalis infection in women has been associated with adverse birth outcomes, such as premature rupture of the membranes, preterm delivery, and low birth weight, and may also increase susceptibility or transmissibility of human immunodeficiency virus (HIV).5,6
Several studies have identified risk factors and markers for having a T. vaginalis infection in women.3,7–15 Most of these studies have included a cross-section of a population and measured the prevalence (the number of infections that exist in a population at a given point in time). Very few studies have measured the incidence (the number of new infections). The factors associated with having a prevalent infection may be different from those associated with acquiring an incident infection. For example, women with prevalent infections may be more likely to be asymptomatic because some women with symptomatic infections seek treatment and are cured. Increased risk for a prevalent trichomonas infection in women has been associated with black race, having multiple sexual partners, greater lifetime number of sexual partners, greater years of sexual activity, concomitant sexually transmitted diseases, and both younger and older age.3,7–13 Increased risk for an incident trichomonas infection in women has been associated with black race, having multiple sexual partners, and older age.14,15 There have been no published studies that evaluated prevalence and incidence of trichomonas infection in the same population of women.
It is still unclear as to whether age is a risk factor for either incident or prevalent trichomonas infection. Both younger age and older age have been found to be associated with prevalent trichomonas infections,8–12 whereas older age has been found to be associated with incident trichomonas infection with the highest incidence among women aged 40 to 44 years.15
We analyzed data from an HIV prevention trial that included women seen in sexually transmitted disease (STD) clinics who were tested for T. vaginalis, treated, and then followed up for incident trichomonas infections. We identified risk factors for prevalent and incident T. vaginalis and compared those risk factors. In addition, we assessed the relationship between age and trichomonas infections.
Materials and Methods
We analyzed data from 1462 women screened for T. vaginalis upon enrollment in RESPECT-2, a multicenter, randomized, controlled trial comparing the efficacy of counseling and testing for HIV using 2 different testing methods. The trial took place in 3 STD clinics in Denver, CO; Long Beach, CA; and Newark, NJ. Primary analyses and detailed methodology are described elsewhere.16,17 Briefly, eligible clients were those who came to the clinics for a full diagnostic examination for STIs, were HIV negative at enrollment, reported having vaginal or anal sex in the preceding 3 months, and were aged 15 to 39 years. Participants were randomized at enrollment to receive either a rapid HIV test with 2 counseling sessions in 1 visit or a standard HIV test and counseling in 2 visits.16,17 Half the participants in each HIV testing group were further randomly assigned to receive an additional booster counseling session 6 months later.16,17
At the baseline visit, women were tested for gonorrhea, chlamydia, T. vaginalis, HIV, and syphilis. We did not analyze HIV and syphilis test results for this article because the women were not tested for these infections at each follow-up visit. Tests for gonorrhea and chlamydia were performed by Long Beach and Newark STD clinic laboratories using ligase chain reaction (LCx Uriprobe; Abbott Diagnostics Division, Abbott Park, IL); and the Denver STD clinic laboratory used polymerase chain reaction (Cobas Amplicor CT PCR and Cobas Amplicor GC PCR; Roche Diagnostic Systems, Inc., Branchburg, NJ) initially, but 18 months after the study began changed to strand displacement amplification (BDProbeTec ET CT/GC; BD Diagnostic Systems, Sparks, MD). Trichomonas vaginalis culture was done using vaginal swabs by InPouch TV test (BioMed Diagnostics Inc., San Jose, Cal.) or modified Diamond's medium as the culture medium. At follow-up visits, vaginal swabs were self-collected (Denver and Long Beach) or clinician-collected (Newark), depending on local clinic policy. Cultures were maintained for at least 5 days, with specimens examined for growth on Days 1, 2, and 5. Audio computer-assisted self-interview technology was used to collect sexual behavioral history and self-reported symptoms of vaginitis (discharge, pruritis, rash, or redness) at enrollment and at each study follow-up visit. Multiple sexual partners were defined as having ≥2 sexual partners in the previous 3 months.
Each woman was scheduled for 4 follow-up visits in 13-week intervals, scheduled 3, 6, 9, and 12 months from the date of enrollment in the study. At each follow-up visit, women were screened for gonorrhea, chlamydia, and T. vaginalis. Per clinic protocols, patients with positive test results were treated with Centers for Disease Control recommended treatment regimens. Assessment and treatment of partners was conducted according to local clinic policy but its completeness could not be determined. No active partner notification occurred, but infected women were told to have their partners treated. In Denver, patients were given medication and treatment instructions for their partners. Women who were due for a return follow-up visit were screened for STDs and interviewed if they visited the clinic any time from 1 week before their scheduled visit to 12 weeks after their scheduled visit. Women could return anytime during a 13-week interval if symptomatic, and if they were positive for infection at any of those visits, they were counted as having a positive infection during that interval. A trichomonas infection was defined as having a positive culture result.
For this analysis we compared the risk factors for incident trichomonas infections during follow-up to the risk factors for prevalent infections at baseline. All analyses were done using SAS version 9.1 (SAS Institute, Cary, NC).
Baseline trichomonas prevalence was estimated by dividing the number of positive T. vaginalis test results by the total number of positive and negative T. vaginalis test results. Potential risk factors and markers analyzed included demographic (enrollment site, age, race/ethnicity, years of education, employment status), clinical characteristics (previous STIs, current STIs, and symptoms of vaginitis), and behavioral factors (new sexual partners and number of sexual partners). Potential predictors of prevalent trichomonas infections were identified by univariate odds ratios and 95% confidence intervals. Logistic regression modeling was used to determine factors associated with prevalent trichomonas infections. We assessed interaction between age and behavioral risk factors.
An incident trichomonas infection was estimated by dividing the number of positive T. vaginalis test results (positive results had to either be preceded by a negative result or detected more than 14 days after documented treatment with metronidazole) by the total number of positive and negative T. vaginalis test results. Women who had a baseline test for trichomonas infection and returned for at least 1 follow-up visit were included in the incident analysis. Potential predictors of incident trichomonas infections were identified by univariate odds ratios and 95% confidence intervals.
Unconditional logistic regression modeling using generalized estimating equations (GEE), which accounted for within-participant correlations of repeated measures, was used to determine factors associated with incident trichomonas infections.18 We used an independence working correlation matrix, in addition to unstructured, exchangeable, and autoregressive correlation structures. Only the results of the unstructured correlation matrix are shown because all correlation structures had similar final model estimates. This correlation matrix was the least restrictive of the 4 correlation structures; it did not assume all correlations to be equal (assumption with exchangeable), any correlation to be zero (assumption with independence), or that the correlation depends on interval of time between responses (an assumption with autoregressive), as all intervals were 13-weeks apart in our analysis.
Among the 1462 women tested for T. vaginalis at baseline, there were 190 with prevalent T. vaginalis, for an overall prevalence of 13% (Table 1). Baseline C. trachomatis prevalence was 12% and baseline N. gonorrhoeae prevalence was 6%. T. vaginalis prevalence was highest among women from Newark (19%), women aged 35 to 39 years (18%), black women (20%), women with 12 or less years of education (15%), and women who tested positive for chlamydia (22%) or gonorrhea (23%) at baseline. The strongest predictors for a woman having a prevalent T. vaginalis infection at baseline were age 35 to 39 years compared to 15 to 19 years (odds ratio 2.8, 95% CI 1.6–4.9) and black race compared to white race (odds ratio 3.9, 95% CI 2.4–6.5).
Using logistic regression modeling, we found that factors significantly associated with prevalent trichomonas infection included older age (compared with women aged 15–19 years), black race (compared with all other race/ethnicity), and having a concurrent chlamydial infection (Table 2).
During the 3955 follow-up visits by women, there were 180 incident trichomonas infections; an overall incidence of 4.6% per 3-month interval over the 1-year follow-up period (estimated annual incidence of 18%) (Table 1). Similar to our findings for prevalence, incidence was highest among women from Newark (8%), older women 35 to 39 years (8%), black women (8%), women with 12 or less years of education (6%), and women who tested positive for chlamydia (10%) or gonorrhea (17%) in that interval. Risk of incident infection was higher in women having had trichomonas (12%) or gonorrhea (12%) diagnosed during their previous clinic visit. Women who tested positive for trichomonas infection during their previous visit had 3.6 times greater risk of an incident trichomonas infection than women who tested negative for trichomonas infection during the previous visit (95% CI 2.3–5.5). The strongest predictor of a woman having an incident trichomonas infection was black race compared to white race (odds ratio 5.1, 95% CI 3.0–8.8).
Using GEE modeling, and controlling for clinical and behavioral risk variables, independent risk factors and markers for incident trichomonas infection included the following: older age (compared with women aged 15–19 years), black race (compared with all other race/ethnicity), concurrent chlamydial infection, multiple (≥2) sexual partners in the 3 months before incident infection, and having a diagnosed and treated trichomonas infection in the interval before incident infection (Table 2).
Comparison of the multivariate models for risk factors associated with prevalent (logistic regression) and incident infections (GEE) showed some interesting similarities and differences. Factors associated with both prevalent and incident trichomonas infection included the following: older age, black race, and concurrent chlamydial infection. Having multiple (≥2) sexual partners and having had a trichomonas infection in the previous interval were each risk factors for incident trichomonas infection.
This study sought to assess and compare risk factors for prevalent trichomonas infections with risk factors for incident trichomonas infections within the same population of women. Few studies have evaluated the incidence of trichomonas infection in women and no other study has compared prevalence and incidence of trichomonas infection within the same population. Both prevalence and incidence of trichomonas infection were high within this population of women; 13% of women had a prevalent infection, and 4.6% of women acquired an incident infection in a 3-month time period (approximately 18.4% for a 1-year incidence). To be included in this study, these women had to come into the STD clinic for an exam, and therefore, we would expect a high prevalence of T. vaginalis as well as other STIs—after all, they are coming in because they think they have an STI. During follow-up, this population would be expected to have a higher risk than the general population for acquiring an STI or HIV. However we are following up all women regardless of symptoms and not just testing the women who came back for another clinic visit, and the number of incident infections was still high.
When we compared incidence to prevalence we found that older age was a risk factor for both incident and prevalent trichomonas infection. Women aged 35 to 39 years were more likely to acquire an incident trichomonas infection than woman aged 15 to 19 years; women aged 20 to 39 years were more likely to have a prevalent trichomonas infection than women aged 15 to 19 years. The higher incidence of trichomonas infection in women aged 35 to 39 years in this study is consistent with a previous incidence study that found the overall age-specific incidence of T. vaginalis was highest in women aged 30 to 49 years.15 Four other studies showed the prevalence of T. vaginalis to increase monotonically with age among women.8–10,19 This association with older age is in contrast to the association of age and other curable STIs. The highest age-specific rates for chlamydia and gonorrhea are among women under 25 years.20 The reason the association of T. vaginalis with older age in women exists remains to be determined. Possible hypotheses include biologic changes in older women that could increase their risk for acquiring a T. vaginalis infection or for having persistent infection after standard treatment with metronidazole, or a higher prevalence in older men because of persistence or inadequate response to treatment,21 in turn leading to higher prevalence and incidence in older women.
As in previous studies, black race was found to be a significant predictor of both prevalent and incident trichomonas infection. Cotch et al. found the most significant predictor of T. vaginalis to be black race, with black women 3.5 times as likely to be colonized as non-black women after adjusting for age, marital status, income, education level, and several indices of sexual behavior.13 Although black race was found to be significant in our study, even after adjusting for other risk factors in the model, there is no biologic or physiological explanation for this increase in risk and black race is more likely a surrogate marker for some unknown underlying risk factor. Other factors that were not taken into account, including personal health practices such as douching, and socioeconomic factors such as sexual networks leading to higher prevalence partners, may explain the association of black race and T. vaginalis.
Having a concurrent chlamydial infection was a predictor of both prevalent and incident T. vaginalis after controlling for other risk factors. A previous study showed that having a chlamydial infection increased the risk of concurrent trichomonas infection in women.19 Another study found that having a history of other STIs and having a concomitant STI increased the risk of having T. vaginalis.14 Women who are diagnosed with chlamydial infections may be at a high risk for T. vaginalis infection as well because of higher risk behaviors in themselves, their partners, or their sexual networks.
In this study, one of the most significant predictors of an incident T. vaginalis infection was having had T. vaginalis at the previous visit. Although all infections were treated with highly effective single-dose therapy, some of the infections may have been related to treatment failure and others may have been because of lack of partner treatment. In addition, the majority of the new infections were asymptomatic. Without scheduled rescreening those repeat infections would have been missed, and the benefits of rescreening should be investigated further.
The results of this study apply to women in STD clinics. Women who attend STD clinics may be very different from the general population. Because the study wasn't specifically designed to study risk factors associated with trichomonas infection, it did not gather information for all possible confounders to the associations of risk factors and trichomonas infection. Additional information regarding, income, forms of birth control being used, and whether the women douched may have strengthened the study results. Also, there may be misclassification of incident infections because of the low sensitivity of culture, in that some baseline infections may have been missed and then counted as incident infections when detected at follow-up. Misclassification may have also occurred because of a drug resistant infection, which can occur in approximately 5% to 10% of the population,22,23 resulting in persistent baseline infections detected at a later visit.
Most of the incident infections were asymptomatic. Risk factors for prevalent and incident infection in women were similar: older age, black race, and concurrent chlamydial infection. In addition, having a trichomonas infection in the previous interval and having multiple (≥2) sex partners were risk factors for incident infection. Unlike other STIs, such as gonorrhea and chlamydia, we found that T. vaginalis was more often found in older compared with younger women. Understanding why this difference exists is an important next step. Further investigation should assess the benefits of routinely screening women in STD clinics for T. vaginalis.
1. Weinstock H, Berman S, Cates W Jr. Sexually transmitted diseases among American youth: Incidence and prevalence estimates, 2000. Perspect Sex Reprod Health 2004; 36:6–10.
2. Sutton MY, Sternberg MR, Koumans E, et al. Prevalence of Trichomonas vaginalis
in the United States, 2001–2002 [abstract 213]. In: Program and Abstracts of the 2006 National STD Prevention Conference (Jacksonville, FL). Available at: http://cdc.confex.com/cdc/std2006/techprogram/P11180.HTM
. Accessed 11 December 2006.
3. Krieger JN, Anagonou S. Trichomonas vaginalis
and Trichomoniasis. In: Holmes KK, Sparling PF, Mardh PA, et al., eds. Sexually Transmitted Diseases, 3rd ed. New York: McGraw-Hill, 1997:587–604.
4. Heine P, McGregor JA. Trichomonas vaginalis
: A reemerging pathogen. Clin Obstet Gynecol 1993; 36:137–144.
5. Laga M, Alary M, Nzila N, et al. Condom promotion, sexually transmitted diseases treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet 1994; 344:246–248.
6. McClelland RS, Sangaré L, Hassan WM, et al. Infection with Trichomonas vaginalis
increases the risk of HIV-1 acquisition. J Infect Dis 2007; 195:698–702.
7. Cates W Jr. Estimates of the incidence and prevalence of sexually transmitted diseases in the United States. American Social Health Association Panel. Sex Transm Dis 1999; 26:S2–S7.
8. Dan M, Kaneti N, Levin D, et al. Vaginitis in a gynecological practice in Israel: Causes and risk factors. Isr Med Assoc J 2003; 5:629–632.
9. Zigas V. An evaluation of trichomoniasis in two ethnic groups in Papua New Guinea. Sex Transm Dis 1977; 4:63–65.
10. Bowden FJ, Paterson BA, Mein J, et al. Estimating the prevalence of Trichomonas vaginalis
, Chlamydia trachomatis
, and Neisseria gonorrhoeae
in indigenous women in Northern Australia. Sex Transm Infect 1999; 75:431–434.
11. Ipsen J, Feigl P. A biomathematical model for prevalence of Trichomonas vaginalis.
Am J Epidemiol 1970; 91:175–184.
12. Buvé A, Weiss HA, Laga M, et al. The epidemiology of trichomoniasis in women in four African cities. AIDS 2001; 15(suppl 4):S89–S96.
13. Cotch MF, Pastorek JG II, Nugent RP, et al. Demographic and behavioral predictors of Trichomonas vaginalis
infection among pregnant women. Obstet Gynecol 1991; 78:1087–1092.
14. Lossick JG. Epidemiology of urogenital trichomoniasis. In: Honinberg BM, ed. Trichomonads Parasitic in Humans. New York: Springer Verlag, 1990:9311–9323.
15. Zhang ZF. Epidemiology of Trichomonas vaginalis.
A prospective study in China. Sex Transm Dis 1996; 23:415–424.
16. Metcalf CA, Douglas JM, Malotte CK, et al. Relative efficacy of prevention counseling with rapid and standard HIV testing: A randomized, controlled trial (RESPECT-2). Sex Transm Dis 2005; 32:130–138.
17. Metcalf CA, Malotte CK, Douglas JM, et al. Efficacy of a booster counseling session 6 months after HIV testing and counseling: A randomized, controlled trial (RESPECT-2). Sex Transm Dis 2005; 32:123–129.
18. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986; 73:13–22.
19. Burch TA, Ries CW, Reardon LV. Epidemiologic studies on human trichomoniasis. Am J Trop Med Hyg 1959; 8:312–318.
21. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 2005. Atlanta, GA: US Department of Health and Human Services, 2006.
20. Joyner JL, Douglas JM, Ragsdale S, et al. Comparative prevalence of infection with Trichomonas vaginalis
among men attending a sexually transmitted disease clinic. Sex Transm Dis 2000; 27:236–240.
22. Schmid G, Narcisi E, Mosure D, et al. Prevalence of metronidazole resistant Trichomonas vaginalis
in a gynecology clinic. J Reprod Med 2001; 46:545–549.
23. Schwebke JR, Barrientes FJ. Prevalence of Trichomonas vaginalis
isolates with resistance to metronidazole and tinidazole. Antimicrob Agents Chemother 2006; 50:4209–4210.