OVER 50% OF SEXUALLY active adults in the United States acquire 1 or more of the 35 types of human papillomavirus (HPV) that infect the genital tract at least once in their lifetimes.1 Invasive cervical cancer, which is almost exclusively associated with persistent infection with 1 or more “high-risk” HPV types, ranks among the top 10 causes of cancer mortality for women in the United States and, with breast cancer, is among the top 2 causes globally.2 HPV-16 and -18 have most commonly been associated with cancer of the cervix, but numerous other high-risk types also are considered oncogenic (types 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 73, 82). Types 26, 53, and 66 have recently been classified as probable high-risk types,3 and additional high- and low-risk types continue to be identified. These oncogenic HPV types are also associated with cancer of the vulva, vagina, penis, and anus. Types 6 and 11 are low-risk types and most frequently associated with genital warts.
Recent trials of an HPV-16 viruslike particle (VLP) vaccine, a bivalent HPV 16/18 vaccine, and a quadrivalent HPV 6/11/16/18 vaccine in young adult women have shown high levels of protection against persistent infection with HPV 6, 11, 16, and/or 18 and against genital warts and cervical intraepithelial neoplasia associated with these types.4–6 Phase III clinical trials with multivalent HPV vaccines are now under way, and an effective vaccine will likely soon be available. Developing strategies for using such a vaccine requires empirical data on the population prevalence of HPV infection and the distribution of types.
Although numerous studies of HPV infection in the United States have been performed, virtually all have been conducted in subsets of the general population. Among young college women, HPV prevalence has ranged from 20% to 46%,7–9 and in clinic-based populations, prevalence as high as 64% has been reported.10 Risk factors for HPV infection identified through these studies include young age,8,11–15 number of sex partners,8,10–16 combining sex with alcohol,8 and partner and partnership characteristics.8,10,14,15,17,18 However, college and clinic populations represent a select segment of society, and there are questions about the generalizeability of these findings. Clearly, prevalence of sexually transmitted infections (STI) differs between general population and clinic samples, yet studies of STI other than HPV demonstrate that risk factors identified in clinic-based populations are largely confirmed in general population samples. However, the use of a clinic-based study population overcontrols for factors associated with clinic-specific care-seeking behaviors, and such studies may fail to identify the full spectrum of risk factors associated with any given infection.19 The sole investigations of HPV in the U.S. general population to date estimated the prevalence of serum antibodies to a single type, HPV-16.20,21 In preparation for an imminent HPV vaccine, the International Agency for Research on Cancer conducted a multicenter study to estimate population prevalence of HPV infection in several different countries17,18, 22–25; however, the United States was not included. Thus, to generate population prevalence estimates of HPV in sexually active females ages 18 to 25 in the United States, determine the prevalences and distribution of HPV types, and identify sociobehavioral correlates of infection in the general population, we tested young women participating in Wave III of the National Longitudinal Study of Adolescent Health (Add Health) for HPV infection.
Materials and Methods
Add Health Study Design and Sample
The Add Health Study was designed to explore health-related behaviors of adolescents in grades 7 through 12 and health outcomes in young adulthood, focusing on social contextual influences.26 This nationally representative probability-based survey used a stratified, random sample of all high schools and junior high “feeder” schools in the United States to identify participants and selected approximately 90,000 adolescents to participate in Wave I (1994). An in-home subsample was drawn (a core sample from each community plus selected special oversamples of specific schools, racial/ethnic groups, disabled students, and siblings) for a more detailed questionnaire (N = 20,745). In 1996, Wave II consisted of a similar in-home interview. Wave III enrolled subjects from July 2001 to April 2002. All Wave I in-home respondents who could be located received a home visit and provided informed consent for the survey and STI testing. Participants (N = 15,197) completed a computer-assisted survey interview (CASI), collecting extensive data on demographic, social, and behavioral characteristics, and provided a urine specimen. Because the urine assay is for research purposes only and HPV testing was performed on anonymized specimens up to 1 year after data were collected, participants were not informed of HPV test results. Nonetheless, all women receiving other STI test results were advised to have a Papanicolaou smear. Although less sensitive for HPV detection than cervical swab specimens (average ratio of HPV prevalence in cervical:urine specimens is 1.23), urine specimens are deemed appropriate for population-based epidemiologic studies where cervical swabs that necessitate a clinical examination cannot be obtained.27,28,30
Selection of the Subsample for HPV Testing
Our subsample was designed to provide 80% power to detect prevalence ratios of approximately 2.0 for selected risk factors. To assure at least 3500 sexually active (defined as ever having engaged in vaginal intercourse) women for HPV analyses, 7000 females in the Wave III sampling frame were flagged for recruitment for HPV testing. Flagged specimens received by the University of North Carolina laboratory were aliquoted into 5-ml tubes, frozen at −70°C, and shipped to the University of Washington HPV Laboratory on dry ice. A total of 6283 samples were received, from which we selected only sexually active women (N = 3741). Of these, 156 (4.2%) had insufficient DNA for HPV testing, and 102 (3.7%) were not part of the originally flagged group; these were excluded. Seventeen subjects over 25 years of age and 204 women without complete data to calculate sampling weights were also excluded from further analyses, leaving a total of 3262 subjects.
The urine specimen was thawed, aliquoted, and spun, and the resulting cell pellet resuspended in 600 μl STM (storage and transfer media from Digene Corporation) and digested with 20 μg/ml protease K. DNA was isolated using QIAamp blood DNA minicolumn, following the manufacturer’s protocol (Qiagen Corporation, cat. No. 51,104). HPV positivity was determined first by PCR amplification using HPV L1 consensus primers MY09/MY11/HMB01 and β-globin primers PC04/GH20, and then followed by dot blot hybridization with generic HPV and β-globin probes.29 Specimens negative for β-globin DNA were excluded from further analyses. HPV positive samples were typed using probes to detect types 6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51, 52, 53, 54, 55, 56, 57, 58, 59, 66, 68, 73, 82, 83, 84 using the Roche Diagnostics line blot assay (Gravitt, 1998). Study procedures were approved by institutional review boards at the University of North Carolina (parent Add Health Study) and the University of Washington (for HPV testing).
We performed a stratified, weighted analysis, taking into account Add Health’s cluster sampling design to generate nationally representative estimates. Analyses incorporated the school as the primary sampling unit, region of the United States as the stratification variable, and poststratification sampling weights. These poststratification sampling weights were specially designed for the HPV subsample and account for the sampling design and survey nonresponse, as well as the quota-sampling method used to flag persons for HPV testing. In univariate analyses, we used a design-based Pearson’s chi-square test for comparisons of categorical characteristics to evaluate factors hypothesized to be associated with HPV. Univariate and multivariate prevalence ratios (PR) and 95% confidence intervals (CI) were calculated using weighted Poisson regression in Stata version 8.0 (StataCorp LP, College Station, TX). Variables tested for inclusion in the multivariate model were those that (a) had been previously associated with HPV in other studies, or (b) were significantly associated with HPV at P = 0.10 in our univariate analyses, using manual forward selection. Variables were retained in the multivariate model if they were significantly associated with HPV infection as determined by a P value <0.05, using a 2-tailed Wald-based test. Potential confounding factors were identified a priori from variables associated with HPV infection (e.g., age, previous treatment for HPV related disease, smoking) and were retained in the multivariate model if their inclusion changed any of the coefficients by 10% or more.
Characteristics of the Add Health HPV Subsample
The women selected for the HPV subsample did not differ significantly from all sexually active women participating in Wave III with respect to age, race/ethnicity, region, education, income, number of sex partners, age at first sexual intercourse, or frequency of vaginal intercourse (P >0.05 for all, data not shown). Women in the HPV subsample had a mean age of 21.7 years (95% CI, 19.65–23.83); 68.8% were white, but oversampling provided substantial numbers of blacks (17.1%) and Hispanics (10.9%). More respondents were from the Midwest (34.0%) and South (37.7%), than from the West (15.5%) and Northeast (12.8%); 53% had attained at least 1 year of college, 50% reported <$10,000 annual personal income, and 74% had never been married. We detected Mycoplasma genitalium in 1.0% of urine specimens using a PCR assay (L. Manhart, In press), and Miller et al.30,31 detected Chlamydia trachomatis in 4.2%, Trichomonas vaginalis in 2.3%, and Neisseria gonorrhoeae in 0.4%.
Prevalence of HPV Infection
Urine specimens from 934 women contained detectable HPV DNA, for a weighted HPV prevalence of 26.9% (95% CI, 23.65%–30.07%) (Table 1). Age-specific prevalence was approximately 30% in women aged 18 to 21 years, and declined with age. Race-specific HPV prevalence was highest among African and Native Americans and lowest among Asians. HPV prevalence was highest in the South (30%; 95% CI, 26.75–33.90) and lowest in the Midwest (24%; 95% CI, 19.76–28.09).
We identified 36 different HPV types, and multiple types were found in approximately 62% of HPV-positive women (16% of all women). A total of 1985 different combinations of HPV types (including 57 untypeable HPV-positive specimens) were identified in the 934 HPV-positive women. HPV 16 was the most commonly identified type (5.8%; Fig. 1), followed in frequency by types 84 (3.2%), 51 (3.0%), 62 (3.0%), 54 (2.9%), and 53 (2.8%). High-risk types were found in 20.4% of Add Health women, with 9.9% infected exclusively with a high-risk type and 10.5% coinfected with both high- and low-risk types; low-risk types alone were present in only 4.8%. Nearly 10% of women with only 1 lifetime vaginal sex partner were infected with a high-risk HPV type. HPV types 6 or 11, primarily responsible for external genital warts, were detected in 2.2% of all women, while types 16 or 18 or both were present in 7.8% of women. Only 0.3% of women were infected with both HPV 16 and 18, and no woman was concurrently infected with HPV 6, 11, 16, and 18. Types 16 or 18 comprised 38% of all infections with a high-risk type. The distribution of HPV types did not differ significantly by geographic region.
Characteristics Associated With HPV Infection
Compared to women without HPV infection, women with infection tended to be younger, single, black, younger at sexual debut, and had more sex partners (lifetime and in the past year) (Table 2). Even among women with only 1 vaginal sex partner ever, 14% (95% CI, 11.14%–18.18%) were infected with HPV. HPV prevalence was lower among women who were currently involved with their most recent partner, and that partnership was characterized by shorter duration, shorter time between meeting and initiating sexual activity, less oral and anal sex, and more often involved a black partner. There was no difference in HPV positivity by region, educational level, annual personal income, or having signed a virginity pledge; yet among those who pledged, HPV-positive women did so at a younger mean age than HPV-negative women (14.4 versus 15.6 years, P = 0.02). Although women with HPV had more often used a condom or experienced condom failure in the past 12 months, correct and consistent condom use was not significantly associated with HPV infection (PR 0.9; 95% CI, 0.70–1.16).
HPV prevalence was higher among women with an STI diagnosis in the past year, genital warts, and/or who believed they had gonococcal or chlamydial infection at the time of the survey. HPV infection was also associated with other high-risk behaviors such as smoking, substance use (marijuana, “crystal meth,” other illegal drugs), sex while drinking or using drugs, or exposure to violence in the past 12 months as measured by dealing in stolen goods or threatening encounters involving guns or knives (data not shown).
Results from our weighted multivariate Poisson regression model adjusting for age, colposcopy in the past year, and each of the other factors in the model indicated that the prevalence of HPV infection was higher among young women and women with a recent male sex partner who was black (Table 3). Of note, the woman’s own race/ethnicity was no longer associated with HPV infection when her partner’s race/ethnicity was considered. Numerous characteristics represented exposure to high-risk partners, but only the report of engaging in an unwanted sexual situation after drinking and of illegal drug use in the past 12 months emerged as independently associated with genital HPV infection. A high lifetime number of sex partners (more than 3 versus 3 or less) and having never married were also positively associated with HPV infection. The 2 factors that were negatively associated with HPV infection were current involvement with and receipt of oral sex from the most recent partner. There were no interactions between being black and having concurrent sex partners or selecting a black partner or between marital status and number of partners.
Additional analyses were done using weighted multinomial logistic regression analysis to simultaneously assess factors associated with high- and low-risk HPV types but did not demonstrate clear differences in risk factors between high- and low-risk types and thus are not presented.
These are the first population-level data on HPV prevalence in the United States. The prevalence of HPV in urine from sexually active women in Add Health was 26.9%, an order of magnitude higher than the prevalence of N gonorrhoeae, C trachomatis, T vaginalis, or M genitalium. Fourteen percent of women who reported only 1 vaginal sex partner in their lifetime had genital HPV infection. HPV-16 was the most common type, and oncogenic types were detected in 20% of women. Although certain demographic characteristics (e.g., race/ethnicity, marital status) were predictive of prevalent HPV infection, other socioeconomic characteristics were not (e.g., education, income), nor were preventive behaviors, such as signing an abstinence pledge or correct and consistent condom use. Although it is tempting to speculate that race/ethnicity-specific HPV prevalence predicts cervical cancer rates, the correlation is poor. For example, Native American women had the highest prevalence of HPV (37.1%), yet they have the lowest rates of cervical cancer (4.9 per 100,000).32 The absence of a relationship between HPV prevalence in young women and cancer rates is partially because cervical cancer occurs primarily in older women, and Papanicolaou testing, diagnosis, and treatment options today versus 20 to 40 years ago have changed over time and across racial ethnic groups.
Although the use of easily obtained specimens such as urine allows us to obtain genital tract specimens in household surveys, urine-based assays for HPV have somewhat lower sensitivities than assays of cervicovaginal specimens. In spiking experiments, spiked untreated urine was a millionfold less sensitive than spiked untreated water, likely due to inhibition of the PCR reaction.33 In a comparative study of adolescents, the ratio of cervical HPV prevalence to urine HPV prevalence was 1.2 (90% versus 75%).28 Similarly, Forslund and colleagues27 reported a ratio of 1.3 (49% cervical versus 38% urine) in women attending a colposcopy clinic, and Strauss et al.34 observed a ratio of 1.2 (78% versus 65%) in women attending a genitourinary medicine clinic, resulting in an average ratio of HPV prevalence in cervical:urine specimens of 1.23. To more closely approximate the cervical prevalence of HPV (assuming 100% specificity of the tests), we could inflate our observed prevalence in urine of 26.9% by 1.23, giving a prevalence estimate of 33.1%. However, urine specimens likely represent a mix of cervical and vulvar or vaginal HPV infections, so one cannot assume all infections detected in the urine were cervical.
Among HPV-infected women, the 6 most commonly identified types were 16, 84, 51, 62, 53, and 54. Although previous HPV prevalence studies also place type 16 within the 5 most common types7–10,13,17,18,22–24,35–37, only 4 of 13 identified type 51 among the top 5,8,9,23,35 and only 1 identified types 54 and 62 among the top 5 types detected.13 Some differences in the prevalence of types may be attributable to improving methods for PCR-based detection of HPV DNA over the past decade. Furthermore, the assay we used incorporated probes for most of the recently identified HPV types (including types 84 and 62). However, some differences may also be due to our use of urine-based detection, as previous studies have demonstrated that different HPV types are identified in paired cervical and urine specimens.28,34 The specific type distributions that we observed thus may be somewhat different from what would be observed in cervical swab specimens, with possibly a lower relative prevalence of low-risk than high-risk types.34
Some of the independent risk factors for HPV infection we identified were consistent with previous reports, whereas others were unique. Unlike other studies,8,11–15 age was not independently associated with detection of HPV in these women, possibly explained by the limited age range in Add Health or adjustment for the behaviors that put younger women at higher risk. In addition to younger age, the most consistent predictors of HPV infection have been lifetime number of sex partners10–13,15,16 or number of partners within the past year,8,13,14 factors also independently associated with HPV infection in our analyses. Our finding of increased risk of HPV among women who reported involvement in a sexual situation while drinking is similar to Ho and colleagues’8 finding. Other studies have identified partner or partnership characteristics as risk factors, such as partners’ other sex partners,8,15,17,18 not living together,14 partner age,10 and partner’s education.8 We also assessed these characteristics but only found current involvement with the most recent sex partner and having a black partner associated with HPV infection. The strong effect of partner’s race has been observed for other bacterial STI and likely reflects higher prevalence of HPV among young black males. Current infection with other STIs did not remain significantly associated with HPV infection after adjustment for behaviors common to all STI.
The lack of an observed association with condom use is consistent with other cross-sectional assessments of this relationship38 and may be due to measurement error, social desirability bias, and/or inability to accurately capture temporal sequence in a sexually experienced population assessed at a single point in time, rather than a true absence of effect. A recent longitudinal study of newly sexually active young women demonstrated a 70% reduction in HPV incidence among those reporting consistent condom use, using electronic diary data,39 suggesting that condom use may indeed be effective in preventing initial acquisition of HPV. The absence of effect for abstinence pledges has been previously reported,40 and is also not surprising since all women tested for HPV were sexually active.
The decreased risk associated with receptive oral sex was unexpected and warrants interpretation with caution. Although this observed effect may be explained by residual confounding, in separate analyses adjusting for frequency of vaginal sex, number of vaginal sex partners, and partnership characteristics, the estimate remained unchanged, suggesting that the effect of engaging in oral sex may be independent of the number of times the individual was exposed to an infected partner. Further studies are required.
A major strength of this study is the large sample size and sampling scheme of Add Health, which provides the first nationally representative estimates of HPV prevalence in the United States. Nevertheless, results should not be extrapolated to women outside the sample age range. A limitation is the fact that urine-based PCR assays for HPV are somewhat less sensitive than PCR performed on cervical swab specimens; thus, true prevalence may be closer to 33% than the 26.9% that we detected.
In conclusion, prevalences of infection with genital HPV, including oncogenic types, were high throughout the country, even among women with only 1 lifetime vaginal sex partner, and as early as age 18. Anticipating the advent of prophylactic HPV vaccines many suggest routinely vaccinating girls before they become sexually active,41 yet others oppose such a strategy on the grounds that it would encourage promiscuity. However, there is no evidence to suggest this is true. Given the relatively high HPV prevalence in young women with low-risk sexual behaviors and the absence of a practical set of factors to define a high-risk group, these data suggest that widespread rather than targeted immunization of young women will provide a greater public health benefit against this ubiquitous infection. The association of sex partner characteristics with HPV infection suggests that vaccinating boys should also be seriously considered.
1. Koutsky LA, Kiviat NB. Genital human papillomavirus. In: Holmes KK, Sparling PF, Mardh P-A, et al., eds. Sexually Transmitted Diseases. 3rd ed. New York, NY: McGraw Hill; 1999:347–359.
2. Pisani P, Parkin DM, Bray F, et al. Estimates of the worldwide mortality from 25 cancers in 1990. Int J Cancer 1999; 83:18–29.
3. Munoz N, Bosch FX, de Sanjose S, et al. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003; 348:518–527.
4. Koutsky LA, Ault KA, Wheeler CM, et al. A controlled trial of a human papillomavirus type 16 vaccine. N Engl J Med 2002; 347:1645–1651.
5. Harper DM, Franco EL, Wheeler C, et al. Efficacy of a bivalent L1 virus-like particle vaccine in prevention of infection with human papillomavirus types 16 and 18 in young women: a randomised controlled trial. Lancet 2004; 364:1757–1765.
6. Villa LL, Costa RL, Petta CA, et al. Prophylactic quadrivalent human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine in young women: a randomised double-blind placebo-controlled multicentre phase II efficacy trial. Lancet Oncol 2005; 6:271–278.
7. Bauer HM, Ting Y, Greer CE, et al. Genital human papillomavirus infection in female university students as determined by a PCR-based method. JAMA 1991; 265:472–477.
8. Ho GY, Bierman R, Beardsley L, et al. Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 1998; 338:423–428.
9. Winer RL, Lee SK, Hughes JP, et al. Genital human papillomavirus infection: incidence and risk factors in a cohort of female university students. Am J Epidemiol 2003; 157:218–226.
10. Tarkowski TA, Koumans EH, Sawyer M, et al. Epidemiology of human papillomavirus infection and abnormal cytologic test results in an urban adolescent population. J Infect Dis 2004; 189:46–50.
11. Bauer HM, Hildesheim A, Schiffman MH, et al. Determinants of genital human papillomavirus infection in low-risk women in Portland, Oregon. Sex Transm Dis 1993; 20:274–278.
12. Hildesheim A, Gravitt P, Schiffman MH, et al. Determinants of genital human papillomavirus infection in low-income women in Washington, D.C. Sex Transm Dis 1993; 20:279–285.
13. Peyton CL, Gravitt PE, Hunt WC, et al. Determinants of genital human papillomavirus detection in a US population. J Infect Dis 2001; 183:1554–1564.
14. Burk RD, Kelly P, Feldman J, et al. Declining prevalence of cervicovaginal human papillomavirus infection with age is independent of other risk factors. Sex Transm Dis 1996; 23:333–341.
15. Burk RD, Ho GY, Beardsley L, et al. Sexual behavior and partner characteristics are the predominant risk factors for genital human papillomavirus infection in young women. J Infect Dis 1996; 174:679–689.
16. Wheeler CM, Parmenter CA, Hunt WC, et al. Determinants of genital human papillomavirus infection among cytologically normal women attending the University of New Mexico student health center. Sex Transm Dis 1993; 20:286–289.
17. Thomas JO, Herrero R, Omigbodun AA, et al. Prevalence of papillomavirus infection in women in Ibadan, Nigeria: a population-based study. Br J Cancer 2004; 90:638–645.
18. Sukvirach S, Smith JS, Tunsakul S, et al. Population-based human papillomavirus prevalence in Lampang and Songkla, Thailand. J Infect Dis 2003; 187:1246–1256.
19. Manhart LE, Aral SO, Holmes KK, et al. Influence of study population on the identification of risk factors for sexually transmitted diseases using a case-control design: the example of gonorrhea. Am J Epidemiol 2004; 160:393–402.
20. Stone KM, Karem KL, Sternberg MR, et al. Seroprevalence of human papillomavirus type 16 infection in the United States. J Infect Dis 2002; 186:1396–1402.
21. Dunne EF, Karem KL, Sternberg MR, et al. Seroprevalence of human papillomavirus type 16 in children. J Infect Dis 2005; 191:1817–1819.
22. Anh PTH, Nguyen TH, Herrero R, et al. Human papillomavirus infection among women in South and North Vietnam. Int J Cancer 2003; 104:213–220.
23. de Sanjose S, Almirall R, Lloveras B, et al. Cervical human papillomavirus infection in the female population in Barcelona, Spain. Sex Transm Dis 2003; 30:788–793.
24. Matos E, Loria D, Amestoy GM, et al. Prevalence of human papillomavirus infection among women in Concordia, Argentina: a population-based study. Sex Transm Dis 2003; 30:593–599.
25. Clifford GM, Gallus S, Herrero R, et al. Worldwide distribution of human papillomavirus types in cytologically normal women in the International Agency for Research on Cancer HPV prevalence surveys: a pooled analysis. Lancet.
27. Forslund O, Hansson BG, Rymark P, et al. Human papillomavirus DNA in urine samples compared with that in simultaneously collected urethra and cervix samples. J Clin Microbiol 1993; 31:1975–1979.
28. Jacobson DL, Womack SD, Peralta L, et al. Concordance of human papillomavirus in the cervix and urine among inner city adolescents. Pediatr Infect Dis J 2000; 19:722–728.
29. Manos MM, Ting Y, Wright DK, Lewis AJ, Broker TR, Wolinsky SM. Use of polymerase chain reaction amplification for the detection of genital human papillomaviruses. In: Furth M, Greaves M (eds.). Cancer cells. Vol. 7. Molecular diagnostics of human cancer. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press 1989, pp. 209–214.
29a. Manhart LE, Holmes KK, Hughes JP, Houston LS, Totten PA. Mycoplasma genitalium in young adults in the United States: an emerging STI more common than gonorrhea. (In press).
29b. Gravitt PE, Peyton CL, Apple JR, Wheeler CM. Genotyping of 27 human papillomavirus types by using L1 consensus PCR products by a single-hybridization, reverse line blot detection method. J Clin Microbiol 1998; 36(10):3020–3027.
30. Miller WC, Ford CA, Morris M, et al. Prevalence of chlamydial and gonococcal infections among young adults in the United States. JAMA 2004; 291:2229–2236.
31. Miller WC, Swygard H, Hobbs MM, et al. The prevalence of trichomoniasis in young adults in the United States. Sex Transm Dis 2005; 32:593–598.
33. Brinkman JA, Rahmani MZ, Jones WE, et al. Optimization of PCR based detection of human papillomavirus DNA from urine specimens. J Clin Virol 2004; 29:230–240.
34. Strauss S, Jordens JZ, McBride D, et al. Detection and typing of human papillomavirus DNA in paired urine and cervical scrapes. Eur J Epidemiol 1999; 15:537–543.
35. Shin HR, Franceschi S, Vaccarella S, et al. Prevalence and determinants of genital infection with papillomavirus, in female and male university students in Busan, South Korea. J Infect Dis 2004; 190:468–476.
36. Castellsague X, Menendez C, Loscertales MP, et al. Human papillomavirus genotypes in rural Mozambique. Lancet 2001; 358:1429–1430.
37. Beby-Defaux A, Bourgoin A, Ragot S, et al. Human papillomavirus infection of the cervix uteri in women attending a health examination center of the French social security. J Med Virol 2004; 73:262–268.
38. Manhart LE, Koutsky LA. Do condoms prevent genital HPV infection, external genital warts, or cervical neoplasia? a meta-analysis. Sex Transm Dis 2002; 29:725–735.
39. Winer RL, Hughes JP, Feng Q, et al. The effect of consistent condom use on the risk of genital HPV infection among newly sexually active women. Paper presented at: 16th Biennial Meeting of the International Society for Sexually Transmitted Disease Research (ISSTDR); July 10–13, 2005; Amsterdam, The Netherlands.
40. Bruckner H, Bearman P. After the promise: the STD consequences of adolescent virginity pledges. J Adolesc Health 2005; 36:271–278.
41. Shaw AR. Human papillomavirus vaccines in development: if they’re successful in clinical trials, how will they be implemented? Gynecol Oncol 2005; 99(3 suppl 1):S246–248.