Abstract
Among the US civilian noninstitutionalized population aged 14 to 59 years in 2013 to 2016, prevalence of Trichomonas vaginalis infection in urine was 1.3% overall. Prevalence was 2.1% among females, 0.5% among males, and highest at 9.6% among non-Hispanic black females. Estimate instability limited analysis of factors beyond sex, age, and race/Hispanic ethnicity.
Trichomonas vaginalis is a treatable sexually transmitted parasitic infection associated with preterm delivery among women.1,2 In 2001 to 2004, estimated prevalence among US civilian, noninstitutionalized females aged 14 to 49 years was 3.1%, based on polymerase chain reaction results from vaginal swab specimens collected for the National Health and Nutrition Examination Survey (NHANES).3 We assessed T. vaginalis prevalence among US civilian, noninstitutionalized males and females aged 14 to 59 years from 2013 to 2016 by demographic, health, and sexual behavior factors using nucleic acid amplification test results from NHANES urine specimens. This is the first publication of 2013 to 2016 T. vaginalis infection estimates which includes adolescents aged 14 to 17 years.4,5 We also examined the stability of these estimates, as well as nonresponse, which can bias estimates.
MATERIALS AND METHODS
The NHANES is a complex, multistage probability sample survey of the US civilian, noninstitutionalized population6 administered continuously since 1999; data are released as combined 2-year cycles (e.g., 2013–2014). Demographic data are collected during a home interview. Biospecimens, and reproductive health and sexual behavior data, are subsequently collected in a mobile examination center. The NHANES is approved by the National Center for Health Statistics ethics review board; because only secondary, deidentified data were used, further institutional review board approval was not required.
T. vaginalis infection was assessed using urine specimens from male and female participants aged 14 to 59 years with the Gen-Probe Aptima T. vaginalis assay (Hologic; Gen-Probe, San Diego CA).7 Assay performance was verified using male and female urine specimens previously found to be positive or negative for T. vaginalis nucleic acid; proficiency panels provided by the College of American Pathologists were tested 3 times per year to ensure ongoing assay accuracy. T. vaginalis infection results and sexual behavior data for 14- to 17-year-olds were accessed from a National Center for Health Statistics Research Data Center8; data for other participants are publicly available.9
Because of the low prevalence of T. vaginalis infection, to increase stability of estimates, data from 2013 to 2014 and 2015 to 2016 were concatenated. Cumulative response rates (i.e., home interview participation rate, sexual behavior questionnaire participation rate, and T. vaginalis test completion rate among sampled participants) were assessed using age group–specific sample screening, interview, and examination participant counts.10 Response rate calculations were limited to participants aged 20 to 59 years because participant counts were not provided for the 14- to 19-year age group. Response rates were adjusted to account for the 2015 to 2016 sample screening rate of 94.3%.10
Nationally representative, weighted estimates of T. vaginalis prevalence and 95% confidence intervals (CIs) were calculated for the total population, by sex, and by demographic, health, and sexual behavior characteristics separately among males and females. Estimates with relative standard errors (RSEs) ≥30% and <50% may be unstable and should be interpreted with caution; estimates with RSE ≥50% were not shown because these are unstable. Analyses were performed using SAS software version 9.411 and SAS-callable SUDAAN version 11.0.1.12
RESULTS
The 2013 to 2016 interview response rate among participants aged 20 to 59 years was 61.9%. T. vaginalis test completion was 58.4%. Sexual behavior questionnaire participation was 52.3%.
T. vaginalis prevalence among males and females aged 14 to 59 years was 1.3% (95% CI, 1.0%–1.7%; Table 1). Prevalence was2.1% (95% CI, 1.6%–2.8%) among females and significantly lower among males (0.5%; 95% CI, 0.3%–0.7%).
TABLE 1.
Percent Prevalence of Trichomonas vaginalis Infections by Demographic Factors Among the US Civilian, Noninstitutionalized Population Aged 14 to 59 Years, 2013 to 2016
| Factor | Sex | Factor Level | Sample Size |
Percent Prevalence | 95% Confidence Limits | Prevalence Ratio |
95% Confidence Limits | ||
|---|---|---|---|---|---|---|---|---|---|
| Total | 8567 | 1.29 | 0.96 | 1.74 | |||||
| Male | 4104 | 0.46 | 0.30 | 0.71 | Ref. | Ref. | Ref. | ||
| Female | 4463 | 2.10 | 1.57 | 2.81 | 4.57 | 3.23 | 6.48 | ||
| Age group, y | Male | 14–19 | 622 | — | — | — | — | — | — |
| 20–29 | 874 | 0.52* | 0.26 | 1.04 | Ref. | Ref. | Ref. | ||
| 30–39 | 892 | — | — | — | — | — | — | ||
| 40–49 | 848 | 0.55* | 0.27 | 1.10 | 1.05 | 0.37 | 2.99 | ||
| 50–59 | 868 | 0.70* | 0.36 | 1.35 | 1.35 | 0.50 | 3.61 | ||
| Female | 14–19 | 634 | 0.71* | 0.35 | 1.45 | 0.27 | 0.11 | 0.68 | |
| 20–29 | 940 | 2.65 | 1.75 | 3.99 | Ref. | Ref. | Ref. | ||
| 30–39 | 957 | 2.19 | 1.37 | 3.46 | 0.83 | 0.55 | 1.24 | ||
| 40–49 | 1029 | 2.68 | 1.87 | 3.84 | 1.01 | 0.62 | 1.66 | ||
| 50–59 | 903 | 1.39 | 0.78 | 2.46 | 0.52 | 0.30 | 0.92 | ||
| Race and Hispanic ethnicity | Male | Mexican American or other Hispanic | 1085 | — | — | — | — | — | — |
| Non-Hispanic white | 1427 | — | — | — | — | — | — | ||
| Non-Hispanic black | 854 | 3.38 | 2.31 | 4.90 | |||||
| Other or Multiracial | 738 | — | — | — | — | — | — | ||
| Female | Mexican American or other Hispanic | 1306 | 1.35 | 0.81 | 2.24 | 1.67 | 0.78 | 3.55 | |
| Non-Hispanic white | 1440 | 0.81 | 0.44 | 1.47 | Ref. | Ref. | Ref. | ||
| Non-Hispanic black | 981 | 9.56 | 7.27 | 12.48 | 11.85 | 5.92 | 23.72 | ||
| Other or multiracial | 736 | 1.32* | 0.49 | 3.52 | 1.63 | 0.51 | 5.23 | ||
| Family income | Male | At or below poverty level | 882 | 1.34 | 0.81 | 2.23 | 5.22 | 1.84 | 14.82 |
| Above to less than twice poverty level | 930 | 0.45* | 0.20 | 1.03 | 1.75 | 0.59 | 5.17 | ||
| At or more than twice poverty level | 1941 | 0.26* | 0.12 | 0.57 | Ref. | Ref. | Ref. | ||
| Female | At or below poverty level | 1084 | 6.77 | 5.06 | 9.00 | 13.80 | 7.18 | 26.53 | |
| Above to less than twice poverty level | 1005 | 2.85 | 1.68 | 4.80 | 5.81 | 2.20 | 15.39 | ||
| At or more than twice poverty level | 2009 | 0.49 | 0.25 | 0.96 | Ref. | Ref. | Ref. | ||
| Educational attainment | Male | Less than high school | 1191 | 1.09* | 0.58 | 2.04 | 4.30 | 1.59 | 11.66 |
| High school | 934 | 0.41* | 0.21 | 0.79 | 1.61 | 0.57 | 4.51 | ||
| More than high school | 1974 | 0.25* | 0.12 | 0.54 | Ref. | Ref. | Ref. | ||
| Female | Less than high school | 1113 | 4.36 | 2.90 | 6.50 | 3.79 | 2.27 | 6.32 | |
| High school | 895 | 3.37 | 2.51 | 4.51 | 2.93 | 1.96 | 4.39 | ||
| More than high school | 2449 | 1.15 | 0.75 | 1.76 | Ref. | Ref. | Ref. | ||
| Marital status | Male | Married | 1812 | 0.21* | 0.08 | 0.55 | Ref. | Ref. | Ref. |
| Live with partner | 406 | 1.20* | 0.56 | 2.56 | 5.69 | 1.66 | 19.41 | ||
| Widowed/divorced/separated | 361 | — | — | — | — | — | — | ||
| Never married | 901 | 0.63* | 0.33 | 1.19 | 2.95 | 0.80 | 10.89 | ||
| Female | Married | 1861 | 0.66 | 0.36 | 1.18 | Ref. | Ref. | Ref. | |
| Live with partner | 405 | 4.13* | 2.18 | 7.69 | 6.31 | 2.52 | 15.83 | ||
| Widowed/divorced/separated | 669 | 3.03 | 1.85 | 4.91 | 4.62 | 2.09 | 10.23 | ||
| Never married | 894 | 4.57 | 3.10 | 6.68 | 6.97 | 3.70 | 13.13 | ||
| Born in the United States† | Male | Yes | 2881 | 0.52 | 0.33 | 0.80 | |||
| No | 1222 | — | — | — | — | — | — | ||
| Female | Yes | 3127 | 2.38 | 1.72 | 3.29 | Ref. | Ref. | Ref. | |
| No | 1335 | 0.91 | 0.54 | 1.54 | 0.38 | 0.20 | 0.75 | ||
Estimates with relative standard error ≥50% are suppressed as indicated by “—”.
Relative standard error ≥30% and <50%; estimates should be interpreted with caution.
Born in the 50 US states or Washington, DC; does not include those born in the 5 US territories.
T. vaginalis prevalence was very low (0.7%; 95% CI,0.4%–1.5%) among females aged 14 to 19 years (Table 1). Prevalence was 2.7% (95% CI, 1.8%–4.0%) among women aged 20 to 29 years and did not differ significantly for women aged 30 to 39 or 40 to 49 years, but was lower (1.4%; 95% CI, 0.8%–2.5%) among women aged 50 to 59 years. Among males aged 14 to 19 and 30 to 39 years, estimates were unstable; no significant differences in prevalence were identified between males in other age groups, although these estimates were potentially unstable.
Compared with non-Hispanic white females (0.8%; 95% CI, 0.4%–1.5%), T. vaginalis prevalence was significantly higher among non-Hispanic black females (9.6%; 95% CI, 7.3%–12.5%) but not Hispanic females (1.4%; 95% CI, 0.8%–2.2%; Table 1). Prevalence among non-Hispanic black males was 3.4% (95% CI, 2.3%–4.9%); estimates among males were unstable for all other race/ethnicity groups.
Among females, increasing poverty level, lower educational attainment, unmarried status, and having been born in the United States were significantly associated with T. vaginalis infection (Table 1). Similar findings were observed among males. However, most of these estimates were potentially unstable; those with RSEs ≥50% are not shown.
Among females, younger age at sexual debut, greater number of lifetime and/or past 12 months sex partners, and chlamydia infection in the past 12 months were significantly associated withT. vaginalis infection (Table 2). Among males, most of these estimates were unstable and not shown.
TABLE 2.
Percent Prevalence of Trichomonas vaginalis Infections by Sexual and Health Factors Among the US Civilian, Noninstitutionalized Female Population Aged 14 to 59 Years, 2013 to 2016
| Factor | Factor Level | Sample Size | Percent Prevalence | 95% Confidence Limits | Prevalence Ratio |
95% Confidence Limits | ||
|---|---|---|---|---|---|---|---|---|
| Pregnant at examination | Yes | 129 | 2.71* | 1.11 | 6.50 | Ref. | Ref. | Ref. |
| No | 2262 | 2.34 | 1.60 | 3.39 | 0.86 | 0.31 | 2.36 | |
| Sexual orientation | Heterosexual | 3169 | 2.07 | 1.51 | 2.82 | Ref. | Ref. | Ref. |
| Homosexual/bisexual/something else/Not Sure | 323 | 3.46* | 1.73 | 6.83 | 1.67 | 0.88 | 3.19 | |
| Ever had sex | Yes | 3505 | 2.23 | 1.66 | 3.01 | |||
| No | 422 | — | — | — | — | — | — | |
| Age at sexual debut, y | ≤14 | 527 | 5.18 | 3.15 | 8.40 | 6.70 | 3.25 | 13.80 |
| 15 −17 | 1537 | 2.64 | 1.88 | 3.71 | 3.42 | 1.70 | 6.89 | |
| ≥18 | 1435 | 0.77 | 0.43 | 1.38 | Ref. | Ref. | Ref. | |
| Ever had same-sex partner | Yes | 387 | 3.73 | 2.03 | 6.78 | 1.83 | 1.00 | 3.35 |
| No | 3112 | 2.04 | 1.50 | 2.77 | Ref. | Ref. | Ref. | |
| Lifetime no. sex partners | 0–2 | 1493 | 0.63* | 0.33 | 1.17 | Ref. | Ref. | Ref. |
| 3–5 | 1014 | 1.97 | 1.23 | 3.15 | 3.16 | 1.69 | 5.91 | |
| 6+ | 1411 | 3.28 | 2.28 | 4.69 | 5.24 | 3.54 | 10.82 | |
| No. sex partners in the past 12 mo | 0–1 | 3381 | 1.44 | 1.05 | 1.97 | Ref. | Ref. | Ref. |
| 2+ | 542 | 6.39 | 4.33 | 9.34 | 4.43 | 3.04 | 6.47 | |
| Ever told by health care provider had genital herpes | Yes | 177 | 1.42* | 0.53 | 3.74 | 0.62 | 0.23 | 1.65 |
| No | 3328 | 2.29 | 1.69 | 3.09 | Ref. | Ref. | Ref. | |
| Ever told by health care provider had genital warts | Yes | 159 | — | — | — | — | — | — |
| No | 3346 | 2.33 | 1.73 | 3.14 | ||||
| Told by health care provider had gonorrhea in the past 12 mo | Yes | 10 | — | — | — | — | — | — |
| No | 3495 | 2.22 | 1.64 | 3.00 | ||||
| Told by health care provider had chlamydia in the past 12 mo | Yes | 61 | 7.56* | 2.93 | 18.16 | 3.51 | 1.39 | 8.87 |
| No | 3442 | 2.16 | 1.59 | 2.91 | Ref. | Ref. | Ref. | |
Estimates with relative standard error ≥50% are suppressed as indicated by “—”.
Relative standard error ≥30% and <50%; estimates should be interpreted with caution.
DISCUSSION
This is the first report of T. vaginalis infection prevalence in 2013 to 2016 that includes data from adolescents aged 14 to 17 years. Overall prevalence among those aged 14 to 59 years was almost 4-fold higher among females than males, and almost 11-fold higher among non-Hispanic black females than non-Hispanic white females. Prevalence was higher among people with lower family income, less education, and who were unmarried. Younger age at sexual debut and higher number of sex partners were associated with higher T. vaginalis prevalence among females.
These findings cannot be directly compared with 2001–2004 NHANES results. Although one study has shown identical detection of T. vaginalis from urine and vaginal swabs,13 others found detection in urine is lower.14,15 In 2001 to 2004, polymerase chain reaction testing was conducted, which is less sensitive than the nucleic acid amplification test used in 2013 to 2016.14 Despite these differences, the current findings highlight similar disparities in burden of T. vaginalis infection by race/ethnicity and offer novel information on prevalence among adolescents.
Females aged 14 to 19 years seemed to have less T. vaginalis infection than those aged 20 to 29 years; estimated prevalence among males aged 14–19 years was unstable and not reported. Estimates among males were also unstable when stratified by race/ethnicity and by most sexual and health factors, despite combining data from two 2-year cycles. A previous publication focusing on men aged 18 to 59 years did not address the instability of many reported estimates.4 Additional years of data might lend stability to these estimates. However, the low prevalence observed among males may be due to the use of urine specimens; penile meatal swabs have been shown to be more sensitive for detecting T. vaginalis.16,17 Test sensitivity also is lower in urine specimens from males (74%) compared with females (88%).14
In addition, a number of estimates presented have RSEs between 30% and 50% and should be interpreted with caution, as these may be unstable. Crude and model-adjusted T. vaginalis prevalence among males and females aged 18 to 59 years were recently published using a single cycle (2013–2014) of NHANES data.5 However, even when data are combined across multiple cycles, as in our analysis, stratification by additional factors within sex increases the likelihood of unstable estimates for uncommon outcomes such as T. vaginalis infection. Therefore, effect measure modification (i.e., interaction) may be difficult to evaluate and account for, if necessary, in statistical models. In the stratified analyses we conducted among females to explore relationships between race/ethnicity, age group, family income, and educational attainment, most estimates were unstable.
Decreasing NHANES cumulative response is an additional concern. Only examination participants were asked to provide specimens for T. vaginalis testing and complete the sexual behavior questionnaire. In addition, any examination participant may decline to provide a urine specimen, or to complete the sexual behavior questionnaire. Cumulative examination response among people sampled for NHANES has decreased to 59% in 2015 to 20168; among adults aged 20 to 59 years in 2013 to 2016, cumulative response to the sexual behavior questionnaire was only 52%. These low cumulative response rates further contribute to the lack of statistical power and potential instability of estimates, and may also produce biased T. vaginalis prevalence estimates. Postsurvey weighting adjustments to account for nonresponse across demographic subgroups and poststratification of survey weights to known population totals, both of which are used for NHANES,18 may reduce nonresponse bias, but only if responders and nonresponders have similar response propensities and respond similarly with respect to the survey measures of interest.19 These assumptions may be invalid, particularly for sensitive information such as sexual behavior. Although nonresponse may not necessarily result in biased survey estimates, efforts to decrease nonresponse, such as monetary incentives, which are used for NHANES,20 may increase bias for some estimates.19 A recent publication examining trends in chlamydia prevalence among young women using data from the National Surveys of Attitudes and Sexual Lifestyles has noted similar concerns.21
In conclusion, this analysis provides national estimates of T. vaginalis infection prevalence in urine specimens from the 2013–2016 US civilian, noninstitutionalized population aged 14 to 59 years and is the first report to include data from adolescent boys and girls during this period. Low infection prevalence among most subpopulation groups constrains detailed exploration of factors associated with this infection using NHANES data. Other data sources are needed to understand the disproportionate burden of T. vaginalis infection, particularly among non-Hispanic blacks.
Footnotes
Publisher's Disclaimer: Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
REFERENCES
- 1.Cotch MF, Pastorek JG 2nd, Nugent RP, et al. Trichomonas vaginalis associated with low birth weight and preterm delivery. The Vaginal Infections and Prematurity Study Group. Sex Transm Dis 1997; 24: 353–360. [DOI] [PubMed] [Google Scholar]
- 2.French JI, McGregor JA, Parker R. Readily treatable reproductive tract infections and preterm birth among black women. Am J Obstet Gynecol 2006; 194:1717–1727. [DOI] [PubMed] [Google Scholar]
- 3.Sutton M, Sternberg M, Koumans EH, et al. The prevalence of Trichomonas vaginalis infection among reproductive-age women in the United States, 2001–2004. Clin Infect Dis 2007; 45:1319–1326. [DOI] [PubMed] [Google Scholar]
- 4.Daugherty M, Glynn K, Byler T. The prevalence of Trichomonas vaginalis infection among US males, 2013–2016. Clin Infect Dis 2018; 68:460–465. [DOI] [PubMed] [Google Scholar]
- 5.Patel EU, Gaydos CA, Packman ZR, et al. Prevalence and correlates of Trichomonas vaginalis infection among men and women in the United States. Clin Infect Dis 2018; 67:211–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.National Health and Nutrition Examination Survey (National Center for Health Statistics, Centers for Disease Control and Prevention website). Available at: https://www.cdc.gov/nchs/nhanes/index.htm. Accessed February 15, 2019.
- 7.National Health and Nutrition Examination Survey 2013–2014 Data Documentation, Codebook, and Frequencies. Trichomonas—Urine (TRICH_H) (National Center for Health Statistics, Centers for Disease Control and Prevention website). Available at: https://wwwn.cdc.gov/Nchs/Nhanes/2013-2014/TRICH_H.htm. Accessed February 15, 2019.
- 8.Research Data Center (RDC) (National Center for Health Statistics, Centers for Disease Control and Prevention website). Available at https://www.cdc.gov/rdc/. Accessed February 15, 2019.
- 9.NHANES Questionnaires, Datasets, and Related Documentation (National Center for Health Statistics, Centers for Disease Control and Prevention website). Available at https://wwwn.cdc.gov/nchs/nhanes/Default.aspx. Accessed February 15, 2019.
- 10.NHANES Response Rates and Population Totals (National Center for Health Statistics, Centers for Disease Control and Prevention website). Available at: https://wwwn.cdc.gov/nchs/nhanes/ResponseRates.aspx. Accessed February 15, 2019.
- 11.SAS [Computer Software] Version 9.4. Carey, NC: SAS Institute, 2018. [Google Scholar]
- 12.SUDAAN [Computer Software] Version 11.0.110. Research Triangle Park, NC: Research Triangle Institute, 2018. [Google Scholar]
- 13.Hollman D, Coupey SM, Fox AS, et al. Screening for Trichomonas vaginalis in high-risk adolescent females with a new transcription-mediated nucleic acid amplification test (NAAT): Associations with ethnicity, symptoms, and prior and current STIs. J Pediatr Adolesc Gynecol 2010; 23:312–316. [DOI] [PubMed] [Google Scholar]
- 14.Nye MB, Schwebke JR, Body BA. Comparison of APTIMA Trichomonas vaginalis transcription-mediated amplification to wet mount microscopy, culture, and polymerase chain reaction for diagnosis of trichomoniasis in men and women. Am J Obstet Gynecol 2009; 200:188.e1–e7. [DOI] [PubMed] [Google Scholar]
- 15.Wendel KA, Erbelding EJ, Gaydos CA, et al. Use of urine polymerase chain reaction to define the prevalence and clinical presentation of Trichomonas vaginalis in men attending an STD clinic. Sex Transm Infect 2003; 79:151–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chernesky M, Jang D, Smieja M, et al. Urinary meatal swabbing detects more men infected with Mycoplasma genitalium and four other sexually transmitted infections than first catch urine. Sex Transm Dis 2017; 44:489–491. [DOI] [PubMed] [Google Scholar]
- 17.Dize L, Agreda P, Quinn N, et al. Comparison of self-obtained penilemeatal swabs to urine for the detection of C. trachomatis, N. gonorrhoeae and T. Vaginalis. Sex Transm Infect 2013; 89:305–307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chen TC, Parker JD, Clark J, et al. National Health and Nutrition Examination Survey: Estimation procedures, 2011–2014. Vital Health Stat 2 2018; 1–26. [PubMed] [Google Scholar]
- 19.Groves RM. Nonresponse rates and nonresponse bias in household surveys. Pub Opin Q 2006; 70:646–675. [Google Scholar]
- 20.Welcome NHANES Participants (National Center for Health Statistics, Centers for Disease Control and Prevention website). Available at https://www.cdc.gov/nchs/nhanes/participant-eligibility.htm. Accessed February 15, 2019.
- 21.Kounali DZ, Welton NJ, Soldan K, et al. Has Chlamydia trachomatis prevalence in young women in England, Scotland and Wales changed? Evidence from National Probability Surveys. Epidemiol Infect 2019; 147:e107. [DOI] [PMC free article] [PubMed] [Google Scholar]