Prevalence and Correlates of Trichomonas vaginalis Infection Among Men and Women in the United States

Abstract

Background: The epidemiology of Trichomonas vaginalis (TV) infection in the United States is poorly defined.

Methods: Males and females aged 18-59 years who participated in the 2013-2014 National Health and Nutrition Examination Survey and provided a urine specimen were tested for TV infection (n = 4057). Participants were also examined for Chlamydia trachomatis (CT) infection, genital human papillomavirus (HPV) infection, and herpes simplex virus type 2 serostatus. Weighted adjusted prevalence ratios (aPRs) were estimated by multivariable Poisson regression.

Results: TV infection prevalence was 0.5% and 1.8% among males and females, respectively. TV infection prevalence was 4.2% among black males, 8.9% among black females, and 0.03% and 0.8%, respectively, among males and females of other races/ethnicities. TV infection prevalence (aPR [95% confidence interval]) was positively associated with female sex (6.1 [3.3-11.3]), black race (vs other races/ethnicities; 7.9 [3.9-16.1]), older age (vs 18-24 years; 3.0 [1.2-7.1] for 25- to 39-year-olds and 3.5 [1.3-9.4] for 40- to 59-year-olds), having less than a high school education (vs completing high school or more; 2.0 [1.0-4.1]), being below the poverty level (vs at or above the poverty level; 4.0 [2.1-7.7]), and having ≥2 sexual partners in the past year (vs 0-1 sexual partners; 3.6 [2.0-6.6]). There were no TV and CT coinfections. Genital HPV detection was not independently associated with TV infection. Among persons aged 18-39 years, there was a significant racial disparity in all sexually transmitted infections examined, and this disparity was greatest for TV infection.

Conclusions: There is a high and disproportionate burden of urinary TV infection in the adult civilian, noninstitutionalized black population in the United States that warrants intervention.

Figure 1.

Prevalence of urinary Trichomonas vaginalis infection in the civilian, noninstitutionalized US population aged 18–59 years, by sex and race/ethnicity. Error bars represent 95% confidence intervals; standard errors were estimated by Taylor series linearization. Prevalence was significantly different by race/ethnicity for each sex as estimated by design-adjusted Wald F tests (P < .01). The “Other races/ethnicities” group included persons identifying as multiracial. The estimate for males among the “Other races/ethnicities” group was based on one positive observation and should be interpreted with caution (relative standard error ≥30%).