Abstract
Background. Data supporting the efficacy of condoms against human papillomavirus (HPV) infection in males are limited. Therefore, we examined the effect of consistent condom use on genital HPV acquisition and duration of infection.
Methods. A prospective analysis was conducted within the HPV Infection in Men Study, a multinational HPV cohort study. Men who were recently sexually active (n = 3323) were stratified on the basis of sexual risk behaviors and partnerships. Using Cox proportional hazards regression, type-specific incidence of HPV infection and clearance were modeled for each risk group to assess independent associations with condom use.
Results. The risk of HPV acquisition was 2-fold lower among men with no steady sex partner who always used condoms, compared with those who never used condoms (hazard ratio, 0.54), after adjustment for country, age, race, education duration, smoking, alcohol, and number of recent sex partners. The probability of clearing an oncogenic HPV infection was 30% higher among nonmonogamous men who always used condoms with nonsteady sex partners, compared with men who never used condoms (hazard ratio, 1.29), after adjustment for country, age, race, education duration, marital status, smoking, alcohol, and number of recent sex partners. No protective effects of condom use were observed among monogamous men.
Conclusions. Condoms should be promoted in combination with HPV vaccination to prevent HPV infection in men.
Keywords: condoms, human papillomavirus, males, cohort study, HIM Study
(See the editorial commentary by Hariri and Warner on pages 367–9.)
Human papillomavirus (HPV) infection is the most common sexually transmitted infection (STI) worldwide [1]. Approximately 50% of men [2, 3] and 40% of women [4] in the United States are infected with HPV at any given time. HPV transmission occurs via skin-to-skin contact, usually during sexual intercourse, and is often transmitted unknowingly, because most infections are asymptomatic. HPV infection can lead to benign disease, including anogenital warts [5], or to malignant cancers of the cervix, vagina, vulva, anal canal, penis, and oropharynx [6]. At present, prophylactic vaccination is the only proven and available method of primary HPV prevention; however, the vaccine is only routinely available for those younger than 26 years, and vaccine dissemination rates in the United States remain low [7, 8]. Therefore, effective methods are still needed to prevent HPV infection and reduce HPV transmission among men and women.
Male condoms offer substantial protection against acquisition of many STIs [9], including human immunodeficiency virus infection, gonorrhea, chlamydial infection, herpes simplex virus type 2 infection, and syphilis. However, data on the efficacy of condoms in preventing HPV acquisition are limited. Consistent condom use has been associated with a decreased likelihood of HPV acquisition among young women [10]; however, data among men are sparse [9, 11]. Some cross-sectional studies have demonstrated protective associations between condom use and prevalence of genital HPV infection [12–15]. Only 2 prospective studies of condom efficacy have been conducted in men, and the studies were restricted by small sample sizes and short follow-up times; however, their results suggest that condom use may reduce the risk of acquiring genital HPV infections [16] and may promote the regression of HPV-related lesions [17]. To our knowledge, no study has had sufficient power to examine the efficacy of condom use in reducing HPV acquisition and decreasing HPV infection duration in men or the ability to detect differences in condom efficacy across sexual behavior risk groups or sex partner types.
Using data from a large cohort of sexually active men, we assess whether consistent condom use reduces the risk of acquiring new genital HPV infections or decreases the duration of HPV infections. The association between condom use and genital HPV infection was further evaluated according to sexual risk behaviors and partnerships.
METHODS
Study Design and Population
A prospective analysis was conducted within the HPV Infection in Men (HIM) Study, an ongoing cohort study of the natural history of HPV infections among healthy men in the United States (Tampa), Brazil (São Paulo), and Mexico (Cuernavaca). The HIM Study cohort consists of 4032 men aged 18–70 years recruited between 2005 and 2009. Men were eligible if they had no prior diagnosis of penile/anal cancer, genital warts, or human immunodeficiency virus infection. Participants were enrolled upon completion of their first follow-up visit (2 weeks after baseline). Additional follow-up visits occurred every 6 months thereafter for up to 4 years. Details of the HIM Study are published elsewhere [2, 3].
Participants were included in this analysis if they reported recent sexual intercourse with female partners (3–6 months prior to baseline visit), responded to questions on condom use, and had valid HPV genotyping results. The analytic cohort included 3323 men (Figure 1). Participants provided written informed consent, and procedures were approved by the following human-subjects-research committees: University of South Florida, Ludwig Institute for Cancer Research, Centro de Referência e Tratamento de Doencas Sexualmente Transmissiveis e AIDS, and Instituto Nacional de Salud Pública de México.
Figure 1.
Flow chart describing the analytic cohort and outcomes of interest. aA total of 13 nonmonogamous men did not respond to the question about condom use with their steady sex partner but were included because they responded to the question about condom use with their nonsteady sex partners.
Measures
At each visit, participants completed a confidential questionnaire about risk factors via computer-assisted self-interview. Questions were asked about sociodemographic characteristics, sexual behaviors, sex partnerships, and condom use.
Sexual Risk Behaviors and Partnerships
Participants were asked whether they had a steady sex partner over the last 6 months and were classified into 3 mutually exclusive risk groups on the basis of their sexual risk behaviors and sex partnerships (Figure 1). Men were defined as “monogamous” if they reported having sex with only 1 steady sex partner and as “nonmonogamous” if they reported having sex with both a steady sex partner and someone other than their steady sex partner. Men were defined as having “no steady sex partner” if they reported having no steady sex partners during this time.
Condom Use
Monogamous men were asked how often they used condoms during sex with their steady sex partner, and nonmonogamous men were asked how often they used condoms during sex with their steady and nonsteady sex partners, separately. Men with no steady sex partner were asked how often they used condoms during sex. For each measure of condom use, frequency was categorized according to whether condoms were used always, more than half the time, half the time, less than half the time, or never.
Specimen Collection and Testing
Participants underwent a clinical examination at each visit. By use of prewetted Dacron swabs, genital specimens were collected from the penile head, penile shaft, and scrotum [3]. These specimens were combined into 1 sample per participant and archived. Specimens underwent DNA extraction (Qiagen Media Kit), polymerase chain reaction analysis, and HPV genotyping (Roche Linear Array) [18]. If samples tested positive for β globin or an HPV genotype, they were considered adequate and included in the analysis. The genotyping assay tested for 37 HPV types, classified as oncogenic (HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) [19] or nononcogenic (HPV types 6, 11, 26, 40, 42, 53, 54, 55, 61, 62, 64, 66, 67, 69, 70, 71, 72, 73, 81, 82, 83, 84, IS39, and CP6108).
Statistical Analysis
Baseline condom use frequency was recategorized as always, sometimes, or never; more than half the time, half the time, and less than half the time were combined to create the middle category. Genotype-specific HPV infections were reported individually and in groups (any, oncogenic, nononcogenic, and vaccine types). The classification of any HPV type was defined as a positive test result for at least 1 of 37 HPV genotypes. HPV infections with single or multiple oncogenic types were classified as oncogenic. Similarly, HPV infections with at least 1 nononcogenic HPV type were classified as nononcogenic. Vaccine types included those that are in the quadrivalent HPV vaccine (ie, HPV types 6, 11, 16, and 18).
To assess the circumstances under which condom use might impact HPV acquisition and clearance, all analyses were stratified according to men's sexual risk behaviors and sex partnerships (ie, risk group). Participant characteristics were compared across risk groups, using the Monte Carlo estimation of exact Pearson χ2 tests. For analyses of the incidence of HPV infection, only the first acquired genotype-specific infection was considered, and only men who tested negative for any or type-specific HPV infection at baseline were included. Cumulative risk of acquisition was estimated using the Kaplan-Meier method, with log-rank tests used to identify differences across condom use groups. The 12-month incidence of HPV infection was also estimated using the Kaplan-Meier method. Independent associations between condom use and incidence of HPV infection were assessed with Cox proportional hazards regression. Univariate models were run for all variables listed in Table 1, except for ethnicity, to identify potential covariates to be included in multivariate models. Backward selection methods were used to assist in variable selection. Final multivariate models included design factors (age and country), factors that were statistically significant in univariate models, and factors that were considered to be relevant on the basis of the literature.
Table 1.
Participant Characteristics of a Subcohort of Men in the HPV Infection in Men Study, by Genital Human Papillomavirus (HPV)–Acquisition Risk Group
| Risk Group |
|||||
|---|---|---|---|---|---|
| Characteristic | Total (n = 3323) | No Steady Sex Partner (n = 554) | Monogamous (n = 1994) | Nonmonogamous (n = 775) | Pa |
| Follow-up time, mo, median (range) | 17.3 (0.3–55.6) | 18.0 (0.4–55.4) | 16.2 (0.3–55.5) | 18.5 (0.4–55.6) | |
| Country of residence | <.0001 | ||||
| United States | 1004 (30.2) | 296 (53.4) | 544 (27.3) | 164 (21.2) | |
| Brazil | 1170 (35.2) | 165 (29.8) | 631 (31.6) | 374 (48.3) | |
| Mexico | 1149 (34.6) | 93 (16.8) | 819 (41.1) | 237 (30.6) | |
| Age, y | <.0001 | ||||
| 18–30 | 1599 (48.1) | 369 (66.6) | 878 (44.0) | 352 (45.4) | |
| 31–44 | 1312 (39.5) | 135 (24.4) | 827 (41.5) | 350 (45.2) | |
| 45–70 | 412 (12.4) | 50 (9.0) | 289 (14.5) | 73 (9.4) | |
| Race | <.0001 | ||||
| White | 1453 (44.4) | 298 (54.6) | 824 (42.0) | 331 (43.2) | |
| Black | 504 (15.4) | 95 (17.4) | 255 (13.0) | 154 (20.1) | |
| Asian/Pacific Islander | 82 (2.5) | 26 (4.8) | 41 (2.1) | 15 (2.0) | |
| Other/mixed | 1234 (37.7) | 127 (23.3) | 840 (42.9) | 267 (34.8) | |
| Ethnicity | <.0001 | ||||
| Hispanic | 1568 (47.6) | 179 (32.4) | 1039 (52.6) | 350 (45.8) | |
| Non-Hispanic | 1726 (52.4) | 373 (67.6) | 938 (47.4) | 415 (54.2) | |
| Marital status | <.0001 | ||||
| Single | 1359 (41.0) | 434 (78.3) | 609 (30.6) | 316 (40.9) | |
| Married | 1237 (37.3) | 32 (5.8) | 943 (47.4) | 262 (33.9) | |
| Cohabitating | 458 (13.8) | 10 (1.8) | 323 (16.2) | 125 (16.2) | |
| Divorced/separated/widowed | 261 (7.9) | 78 (14.1) | 114 (5.7) | 69 (8.9) | |
| Education duration, y | <.0001 | ||||
| <12 | 733 (22.1) | 69 (12.5) | 508 (25.6) | 156 (20.2) | |
| 12 | 886 (26.8) | 132 (23.9) | 536 (27.0) | 218 (28.2) | |
| 13–15 | 821 (24.8) | 226 (40.9) | 416 (20.9) | 179 (23.2) | |
| 16 | 668 (20.2) | 97 (17.5) | 410 (20.6) | 161 (20.9) | |
| ≥17 | 204 (6.2) | 29 (5.2) | 117 (5.9) | 58 (7.5) | |
| Circumcised | <.0001 | ||||
| No | 2152 (64.8) | 265 (47.8) | 1319 (66.1) | 568 (73.3) | |
| Yes | 1171 (35.2) | 289 (52.2) | 675 (33.9) | 207 (26.7) | |
| Current smoker | .159 | ||||
| No | 2511 (75.7) | 413 (75.0) | 1531 (76.8) | 567 (73.4) | |
| Yes | 807 (24.3) | 138 (25.0) | 463 (23.2) | 206 (26.6) | |
| Alcohol intake, drinks/mo | <.0001 | ||||
| 0 | 723 (22.6) | 101 (19.2) | 491 (25.7) | 131 (17.4) | |
| 1–30 | 1536 (48.1) | 218 (41.4) | 958 (50.1) | 360 (47.9) | |
| 31–60 | 388 (12.2) | 75 (14.2) | 206 (10.8) | 107 (14.2) | |
| >60 | 546 (17.1) | 133 (25.2) | 259 (13.5) | 154 (20.5) | |
| Female sex partners in past 3–6 mo, no. | <.0001 | ||||
| 1 | 1736 (52.2) | 227 (41.0) | 1399 (70.2) | 110 (14.2) | |
| ≥2 | 1062 (32.0) | 290 (52.3) | 166 (8.3) | 606 (78.2) | |
| Male anal sex partners in past 3–6 mo, no. | <.0001 | ||||
| 0 | 3161 (95.7) | 527 (96.0) | 1950 (98.2) | 684 (89.1) | |
| ≥1 | 141 (4.3) | 22 (4.0) | 35 (1.8) | 84 (10.9) | |
| Presence of ≥1 HPV genotype at baseline | <.0001 | ||||
| No | 1516 (45.6) | 240 (43.3) | 1019 (51.1) | 257 (33.2) | |
| Yes | 1807 (54.4) | 314 (56.7) | 975 (48.9) | 518 (66.8) | |
| Any history of STI diagnosis | <.0001 | ||||
| No | 2678 (80.7) | 442 (79.9) | 1655 (83.1) | 581 (75.1) | |
| Yes | 540 (16.3) | 94 (17.0) | 280 (14.1) | 166 (21.4) | |
| Don't know | 101 (3.0) | 17 (3.1) | 57 (2.9) | 27 (3.5) | |
Data are no. (%) of participants, unless otherwise indicated.
Abbreviation: STI, sexually transmitted infection.
a By exact Pearson χ2 analysis, using Monte Carlo estimation. A P value of <.05 indicates a statistically significant difference across risk groups.
Men with prevalent or incident HPV infections were included in HPV clearance analyses. Clearance was defined as 2 consecutive negative test results following a positive test result, excluding infections detected for the first time at a participant's final visit. Because of limited numbers of type-specific infections, clearance estimates are reported for grouped infections only (ie, oncogenic HPV types). Since men could have been infected with multiple HPV types within a defined group (eg, HPV-16 and HPV-18 are both oncogenic), we adjusted for within-subject correlation in all grouped HPV clearance analyses. The median time to HPV clearance was estimated among all men with an incident infection, using the clustered Kaplan-Meier method [20]. Men whose infections did not clear were censored in the Kaplan-Meier analysis. To model the associations between condom use and grouped HPV clearance, we employed Cox proportional hazards regression with the robust covariance matrix estimator [21]. To identify variables included in the final multivariate models of HPV clearance, we applied the same variable selection strategy as described for the incidence of HPV infection. In addition, we observed that baseline HPV status (ie, prevalent or incident HPV infection) had a significant impact on HPV clearance estimates, so this factor was included in multivariate models of HPV clearance.
All analyses were performed using SAS, version 9.3 (SAS Institute, Cary, NC). All statistical tests were 2-sided and attained significance at an α level of 0.05.
RESULTS
A total of 3323 men were followed for a median of 17.3 months (range, 0.3–55.6 months; Table 1). Most participants were aged 18–30 years, white, uncircumcised, nonsmokers, and heterosexual. Except for smoking status, all other sociodemographic and behavioral characteristics differed by risk group.
The incidence of HPV infection varied by condom use and risk group (Table 2). The 12-month incidence of infection with any HPV type was higher among nonmonogamous men (53.8%) than among men with no steady sex partner (40.8%) or monogamous men (34.6%; log-rank P < .001). Among men with no steady sex partner, those who always used condoms had the lowest risk of acquiring any HPV within 12 months (32.2%; 50.3% for those who sometimes used condoms, and 49.9% for those who never used condoms). No differences in the incidence of HPV infection were found across condom use groups for monogamous men. Among nonmonogamous men who had sex with their steady partner, the incidence of nononcogenic HPV (37.7%) was lowest for men who always used condoms (sometimes [51.9%] and never [41.6%]) with a steady sex partner. Nonmonogamous men who only sometimes used condoms with their nonsteady sex partners had the highest risk of acquiring nononcogenic HPV (59.3%) and HPV-18 (8.6%) infections. Differences in cumulative incidences of any HPV infection were noted by condom use among men with no steady sex partner (Figure 2A).
Table 2.
Incidence of Genital Human Papillomavirus (HPV) Infection in a Subcohort of Men in the HPV Infection in Men Study, by HPV-Acquisition Risk Group and Frequency of Condom Use
| HPV 12-Month Incidence, % (95% CI) |
|||||
|---|---|---|---|---|---|
| Condom Use |
|||||
| Risk Group, HPV Genotype | Overall | Always | Sometimes | Never | Pa |
| No steady sex partner | |||||
| Any type | 40.8 (33.8–47.2) | 32.2 (22.7–40.5) | 50.3 (36.4–61.2) | 49.9 (30.4–63.9) | .014 |
| Oncogenic type | |||||
| Any | 26.3 (21.1–31.1) | 21.5 (14.4–28.0) | 29.6 (20.3–37.9) | 32.2 (18.5–43.7) | .245 |
| 16 | 7.9 (5.2–10.5) | 6.5 (2.9–10.0) | 8.5 (3.9–12.8) | 10.4 (2.8–17.4) | .139 |
| 18 | 2.9 (1.3–4.4) | 3.1 (.6–5.6) | 4.0 (1.0–6.8) | 0.0 (0.0–0.0) | .046 |
| Nononcogenic type | .079 | ||||
| Any | 36.9 (30.9–42.4) | 31.9 (23.5–39.4) | 39.3 (28.3–48.7) | 46.3 (30.2–58.7) | |
| 6 | 5.8 (3.6–8.1) | 5.2 (2.0–8.4) | 5.7 (2.0–9.3) | 7.5 (1.5–13.1) | .459 |
| 11 | 0.9 (0.0–1.9) | 1.1 (0.0–2.6) | 1.3 (0.0–3.0) | 0 (0.0–0.0) | .218 |
| Any vaccine typeb | 16.1 (12.2–19.8) | 15.1 (9.5–20.5) | 16.3 (9.8–22.4) | 18.1 (8.2–27.0) | .150 |
| Monogamous, steady sex partner | |||||
| Any type | 34.6 (31.3–37.9) | 32.0 (23.6–39.6) | 34.1 (27.6–40.0) | 35.6 (31.0–40.0) | .615 |
| Oncogenic type | |||||
| Any | 22.2 (19.8–24.6) | 20.4 (14.5–25.9) | 24.9 (20.2–29.4) | 21.3 (18.0–24.4) | .586 |
| 16 | 4.8 (3.8–5.9) | 2.7 (.5–4.8) | 5.5 (3.4–7.5) | 5.1 (3.6–6.6) | .933 |
| 18 | 1.5 (0.9–2.1) | .4 (0.0–1.2) | 2.0 (.8–3.2) | 1.6 (.7–2.4) | .134 |
| Nononcogenic type | |||||
| Any | 28.3 (25.5–31.1) | 26.7 (19.5–33.3) | 26.6 (21.3–31.4) | 29.8 (25.9–33.6) | .815 |
| 6 | 4.1 (3.2–5.1) | 2.8 (.7–4.8) | 5.0 (3.1–6.9) | 4.0 (2.7–5.3) | .343 |
| 11 | 0.8 (0.3–1.2) | .7 (0.0–1.7) | .2 (0.0–0.6) | 1.1 (.4–1.8) | .247 |
| Any vaccine typeb | 9.8 (8.2–11.3) | 7.2 (3.7–10.6) | 9.9 (7.0–12.6) | 10.5 (8.3–12.6) | .756 |
| Nonmonogamous, steady sex partner | |||||
| Any type | 53.2 (45.9–59.4) | 46.4 (29.8–59.1) | 59.8 (47.5–69.3) | 50.9 (39.2–60.3) | .065 |
| Oncogenic type | |||||
| Any | 29.7 (25.0–34.0) | 30.0 (18.7–39.6) | 31.9 (23.7–39.2) | 27.7 (20.9–33.9) | .115 |
| 16 | 5.6 (3.7–7.4) | 6.5 (1.7–11.1) | 5.0 (2.0–7.8) | 5.7 (2.9–8.3) | .767 |
| 18 | 4.2 (2.6–5.8) | 4.7 (.6–8.7) | 3.7 (1.1–6.1) | 4.5 (2.0–6.9) | .756 |
| Nononcogenic type | |||||
| Any | 44.6 (38.4–50.2) | 37.7 (24.0–48.9) | 51.9 (41.2–60.6) | 41.6 (31.7–50.1) | .017 |
| 6 | 6.6 (4.6–8.5) | 7.3 (2.3–12.0) | 8.9 (5.0–12.7) | 4.5 (2.1–6.8) | .074 |
| 11 | 2.3 (1.2–3.5) | .8 (0.0–2.4) | 2.0 (.3–3.8) | 3.2 (1.1–5.2) | .302 |
| Any vaccine typeb | 16.5 (13.2–19.7) | 17.9 (9.7–25.3) | 17 (11.3–22.3) | 15.6 (10.9–20.0) | .765 |
| Nonmonogamous, nonsteady sex partner | |||||
| Any type | 53.8 (46.6–60.0) | 47.8 (37.7–56.3) | 71.1 (55.7–81.1) | 47.7 (32.2–59.7) | .080 |
| Oncogenic type | |||||
| Any | 29.7 (25.1–34.1) | 27.7 (21.2–33.6) | 35.7 (26.0–44.1) | 26.9 (17.3–35.3) | .184 |
| 16 | 5.5 (3.7–7.3) | 5.1 (2.4–7.6) | 7.3 (3.4–11.1) | 4.2 (1.1–7.2) | .659 |
| 18 | 4.5 (2.8–6.1) | 3.4 (1.3–5.4) | 8.6 (4.3–12.7) | 1.8 (0.0–3.8) | .002 |
| Nononcogenic type | |||||
| Any | 44.9 (38.8–50.4) | 40.4 (31.8–48.0) | 59.3 (46.1–69.2) | 37.9 (25.2–48.3) | .028 |
| 6 | 6.7 (4.7–8.7) | 8.7 (5.3–11.9) | 6.9 (3.0–10.6) | 2.5 (0.0–5.0) | .080 |
| 11 | 2.1 (1.0–3.3) | 1.2 (0.0–2.4) | 4.0 (1.0–6.8) | 1.7 (0.0–3.7) | .591 |
| Any vaccine typeb | 16.6 (13.4–19.8) | 17.0 (12.2–21.6) | 20.3 (13.4–26.6) | 11.6 (6.0–16.8) | .047 |
a By a log-rank test. A P value of <.05 indicates a statistically significant overall difference in the incidence of HPV infection across the entire follow-up period, by condom use group.
b Defined as HPV types 6, 11, 16, and 18, which are in the quadrivalent HPV vaccine.
Figure 2.
Kaplan–Meier estimates of the cumulative incidence (A) and time to clearance (B) of a newly acquired human papillomavirus infection in a subcohort of men in the HPV Infection in Men Study, by risk group and condom use.
The median time to HPV clearance was shorter among monogamous men (6.6 months) than men with no steady sex partner (7.2 months) or nonmonogamous men (7.6 months; log-rank P = .008) (Table 3). No significant differences in infection duration were observed across condom use groups for men with no steady sex partner or monogamous men. Among nonmonogamous men, those who always used condoms with their steady sex partner experienced the shortest duration of any (6.3 months) and nononcogenic (6.2 months) HPV infection. Similarly, among nonmonogamous men who had sex with their nonsteady sex partners, the median time to clearance of oncogenic HPV infection (6.4 months) was shorter among those who always used condoms. Differences in cumulative probabilities of maintaining an incident HPV infection were noted by condom use among nonmonogamous men (Figure 2B).
Table 3.
Median Time to Clearance of Incident Genital Human Papillomavirus (HPV) Infection in a Subcohort of Men in the HPV Infection in Men Study, by HPV-Acquisition Risk Group and Frequency of Condom Use
| Median Time to HPV Clearance, Months (95% CI) |
|||||
|---|---|---|---|---|---|
| Condom Use |
|||||
| Risk Group, HPV Genotype | Total | Always | Sometimes | Never | Pa |
| No steady sex partner | |||||
| Any type | 7.2 (6.8–7.8) | 7.2 (6.6–8.3) | 7.2 (6.6–8.8) | 6.9 (6.3–8.3) | .775 |
| Oncogenic type | 6.7 (6.4–7.9) | 6.7 (6.2–8.7) | 7.2 (6.2–11.5) | 6.6 (6.1–8.5) | .622 |
| Nononcogenic type | 7.2 (6.9–7.9) | 7.6 (6.9–8.5) | 7.2 (6.7–8.8) | 7.2 (6.2–8.5) | .945 |
| Vaccine typeb | 7.2 (6.5–11.6) | 7.4 (6.2–11.9) | 7.8 (6.2–12.8) | 6.7 (6.1–12.4) | .493 |
| Monogamous, steady sex partner | |||||
| Any type | 6.6 (6.4–6.9) | 7.1 (6.5–7.9) | 6.6 (6.4–7.2) | 6.5 (6.3–6.8) | .269 |
| Oncogenic type | 6.4 (6.3–6.9) | 6.7 (6.2–7.8) | 6.4 (6.2–7.3) | 6.4 (6.2–6.9) | .625 |
| Nononcogenic type | 6.7 (6.4–7.0) | 7.6 (6.4–11.5) | 6.7 (6.4–7.4) | 6.6 (6.3–6.8) | .293 |
| Vaccine typeb | 6.4 (6.3–7.2) | 11.5 (6.9–18.6) | 6.6 (6.1–8.1) | 6.3 (6.2–6.8) | .064 |
| Nonmonogamous, steady sex partner | |||||
| Any type | 7.6 (6.7–8.1) | 6.3 (6.2–7.4) | 7.6 (6.4–9.1) | 8.3 (7.2–11.6) | .004 |
| Oncogenic type | 7.1 (6.5–8.1) | 6.5 (6.1–7.9) | 6.7 (6.2–9.1) | 8.3 (6.7–11.2) | .207 |
| Nononcogenic type | 7.8 (6.8–8.3) | 6.2 (6.2–7.8) | 7.9 (6.7–11.5) | 8.7 (7.2–12.0) | .007 |
| Vaccine typeb | 7.4 (6.2–9.1) | 6.1 (6.0–8.5) | 7.2 (6.2–12.0) | 7.4 (6.2–12.3) | .097 |
| Nonmonogamous, nonsteady sex partner | |||||
| Any type | 7.6 (6.8–8.2) | 7.1 (6.4–7.9) | 7.7 (6.7–11.5) | 11.1 (6.6–12.2) | .070 |
| Oncogenic type | 7.1 (6.5–8.1) | 6.4 (6.2–7.2) | 8.1 (6.5–11.4) | 11.1 (6.5–12.6) | .032 |
| Nononcogenic type | 7.9 (6.9–8.7) | 7.8 (6.7–8.2) | 7.7 (6.7–11.9) | 11.3 (6.3–13.4) | .384 |
| Vaccine typeb | 7.0 (6.2–8.5) | 6.2 (6.1–7.8) | 7.7 (6.2–12.2) | 12.0 (6.1–13.8) | .366 |
a By a Cox model adjusted for within-subject correlation. A P value of <.05 indicates a statistically significant difference across condom use groups.
b Defined as HPV types 6, 11, 16, and 18, which are in the quadrivalent HPV vaccine.
In multivariate models, condom use was significantly associated with a decreased probability of acquiring any HPV infection among men with no steady sex partner (Table 4). The risk of acquiring a new HPV infection was 2-fold lower among men with no steady sex partner who always used condoms, compared with those who never used condoms (HR, 0.54; 95% confidence interval [CI], .31–.95), after adjustment for country, age, race, education duration, smoking status, alcohol intake, and number of recent sex partners. Nonsignificantly reduced risks of acquiring oncogenic and nononcogenic HPV infections were observed for men with no steady sex partner who always used condoms (HRs, 0.66 [95% CI, .40–1.11] and 0.66 [95% CI, .40–1.09], respectively). Among monogamous men, modest protective effects of condom use on HPV acquisition were found for all grouped HPV outcomes (HR range, 0.87–0.96). No clear patterns were observed for HPV acquisition among nonmonogamous men.
Table 4.
Cox Proportional Hazards Models of the Association Between Condom Use and the Incidence and Clearance of Any, Oncogenic, and Nononcogenic Human Papillomavirus (HPV) Infections in a Subcohort of Men in the HPV Infection in Men Study, by HPV-Acquisition Risk Group
| Hazard Ratio (95% Confidence Interval) |
||||||
|---|---|---|---|---|---|---|
| Any HPV Type |
Oncogenic HPV Type |
Nononcogenic HPV Type |
||||
| Risk Group, Condom Use Frequency | Univariate | Multivariatea | Univariate | Multivariatea | Univariate | Multivariatea |
| HPV incidence | ||||||
| No steady sex partnerb | 240 (134) | … | 363 (141) | … | 361 (158) | … |
| Always use condoms | .64 (.40–1.04) | .54 (.31–.95)c | .80 (.50–1.27) | .66 (.40–1.11) | .66 (.43–1.01) | .66 (.40–1.09) |
| Sometimes use condoms | 1.10 (.68–1.79) | 1.15 (.67–2.00) | 1.09 (.68–1.74) | .98 (.58–1.66) | .90 (.58–1.40) | .87 (.53–1.42) |
| Never use condoms | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Monogamous, steady sex partnerb | 1019 (427) | … | 1454 (427) | … | 1238 (451) | … |
| Always use condoms | .92 (.71–1.20) | .87 (.65–1.18) | 1.01 (.78–1.32) | .96 (.71–1.30) | .99 (.77–1.29) | .92 (.69–1.22) |
| Sometimes use condoms | .90 (.72–1.13) | .83 (.64–1.06) | 1.12 (.90–1.39) | 1.11 (.88–1.41) | .93 (.75–1.16) | .84 (.67–1.07) |
| Never use condoms | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Nonmonogamous, steady sex partnerb | 253 (161)d | … | 465 (202) | … | 328 (178) | … |
| Always use condoms | 1.07 (.70–1.64) | 1.24 (.76–2.02) | 1.19 (.81–1.74) | 1.06 (.69–1.64) | 1.02 (.68–1.53) | .88 (.55–1.41) |
| Sometimes use condoms | 1.48 (1.05–2.09) | 1.50 (1.00–2.23) | 1.38 (1.02–1.88) | 1.24 (.89–1.74) | 1.54 (1.11–2.15) | 1.35 (.93–1.97) |
| Never use condoms | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Nonmonogamous, nonsteady sex partnerb | 257 (164) | … | 476 (207) | … | 335 (183) | … |
| Always use condoms | .97 (.66–1.45) | .99 (.63–1.55) | 1.31 (.91–1.90) | 1.07 (.71–1.61) | 1.18 (.80–1.73) | 1.12 (.72–1.74) |
| Sometimes use condoms | 1.44 (.93–2.21) | 1.14 (.70–1.87) | 1.45 (.97–2.16) | 1.16 (.75–1.80) | 1.67 (1.10–2.52) | 1.55 (.98–2.45) |
| Never use condoms | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| HPV clearance | ||||||
| No steady sex partnerb | 1573 (1157) | … | 638 (486) | … | 935 (671) | … |
| Always use condoms | .97 (.78–1.21) | 0.92 (0.76–1.12) | .93 (.69–1.26) | .88 (.63–1.23) | 1.00 (.77–1.31) | .94 (.74–1.20) |
| Sometimes use condoms | .91 (.74–1.13) | 0.88 (0.72–1.07) | .86 (.64–1.15) | .81 (.57–1.16) | .95 (.74–1.23) | .92 (.73–1.17) |
| Never use condoms | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Monogamous, steady sex partnerb | 3955 (2886) | … | 1546 (1174) | … | 2409 (1712) | … |
| Always use condoms | .91 (.80–1.04) | .92 (.81–1.05) | .84 (.69–1.02) | .86 (.70–1.06) | .96 (.82–1.12) | .96 (.81–1.13) |
| Sometimes use condoms | .96 (.87–1.06) | .98 (.87–1.09) | .91 (.79–1.04) | .90 (.77–1.05) | 1.00 (.88–1.12) | 1.02 (.89–1.16) |
| Never use condoms | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Nonmonogamous, steady sex partnerb | 2406 (1725) | … | 910 (697) | … | 1496 (1028) | … |
| Always use condoms | 1.25 (1.07–1.45) | 1.19 (1.00–1.41) | 1.18 (.95–1.47) | 1.16 (.92–1.46) | 1.30 (1.08–1.56) | 1.23 (1.00–1.51) |
| Sometimes use condoms | 1.11 (0.97–1.26) | 1.09 (.95–1.25) | 1.08 (.90–1.30) | 1.14 (.94–1.38) | 1.12 (.96–1.31) | 1.07 (.90–1.26) |
| Never use condoms | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Nonmonogamous, nonsteady sex partnerb | 2449 (1756) | … | 929 (713) | … | 1520 (1043) | … |
| Always use condoms | 1.16 (1.01–1.33) | 1.10 (.95–1.28) | 1.30 (1.06–1.58) | 1.29 (1.03–1.61)c | 1.10 (.93–1.31) | 1.02 (.85–1.23) |
| Sometimes use condoms | 1.16 (.99–1.35) | 1.14 (.97–1.35) | 1.21 (.97–1.51) | 1.25 (.98–1.59) | 1.15 (.95–1.38) | 1.11 (.91–1.36) |
| Never use condoms | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
a Multivariate models for men with no steady sex partner adjusted for country, age, race, education duration, smoking status, alcohol intake, no. of recent female sex partners, and no. of recent male anal sex partners. Multivariate models for monogamous and nonmonogamous men adjusted for country, age, race, education duration, marital status, smoking status, alcohol intake, no. of recent female sex partners, and no. of recent male anal sex partners. Multivariate models for HPV clearance also included HPV infection status at baseline.
b Sample size (no. of events) in the univariate model.
c Statistically significant.
Among monogamous men and men with no steady sex partner, condom use did not appear to affect the probability of clearing HPV infections (Table 4). However, among nonmonogamous men, condom use with a nonsteady sex partner was significantly associated with an increased probability of clearing an oncogenic HPV infection. The probability of clearing an oncogenic HPV infection was 30% higher among nonmonogamous men who always used condoms with nonsteady sex partners, compared with men who never used condoms (HR, 1.29; 95% CI, 1.03–1.61), after adjustment for country, age, race, education duration, marital status, smoking status, alcohol intake, and number of recent sex partners.
DISCUSSION
The results of our prospective study suggest that consistent condom use may reduce the risk of acquiring new HPV infections and decrease the duration of these infections among nonmonogamous men and men with no steady sex partner (ie, high-risk men). Consistent condom use significantly decreased the probability of acquiring an HPV infection among men with no steady sex partner and, among nonmonogamous men, significantly decreased the duration of oncogenic HPV infections. We were unable to detect consistent protective effects of condom use among monogamous men. The partial protection offered by condom use was only observed for high-risk men and against some, but not all, grouped HPV types.
Our results are comparable to those of previous studies, although differences in methods (eg, condom assessment and risk-group stratification) make direct comparisons difficult. We observed that consistent condom use reduces the risk of HPV infections among men with no steady sex partner, similar to the only other prospective evaluation of condom use and HPV infection in men [16]. Kjaer et al found consistent/occasional condom use to significantly reduce HPV acquisition in young Danish soldiers but did not observe an association with HPV persistence. The latter result is in contrast to our finding of a shorter duration of infection, particularly among nonmonogamous men who consistently use condoms. These discrepant results for HPV persistence are likely due to differences in sample size and length of follow-up. In the HIM Study, 3323 men were followed for a median of 17 months, providing us the opportunity to more precisely estimate infection duration and the impact of condom use.
Cross-sectional studies have reported similar associations between increased condom use and decreased prevalence of HPV infection in men [12–15]. In a previous analysis of men in the HIM Study, condom use was associated with a lower prevalence of HPV infection [15]. In a separate study conducted in Tucson, Arizona, and Tampa, Florida, always using condoms was strongly associated with lower odds of HPV detection, particularly among men with multiple sex partners [14]. Similarly, among an ethnically diverse group of men attending an STI clinic in Tucson, Arizona, consistent condom use during the preceding 3 months was associated with decreased prevalence of any and oncogenic HPV infections [12]. However, it must be noted that these studies were cross-sectional by design and were therefore unable to detect temporal associations.
Our findings are also in agreement with those of other investigators who demonstrated effects of condom use on acquisition and regression of HPV-related genital lesions. In a case-control study conducted within an Australian sexual health center, consistent condom use was found to significantly reduce the risk of acquiring genital warts [22]. A separate prospective analysis of HIM Study data showed that consistent condom use remained associated with a decreased risk of condyloma after adjustment for HPV-6/11 positivity [23]. In a randomized clinical trial in the Netherlands, the median time to clinical regression of flat penile lesions in male sex partners of women with cervical precancerous lesions was found to be substantially shorter among condom users, compared with non–condom users (7.4 vs 13.9 months) [17]. Later, the same investigators found that the beneficial effect of condoms occurred only in concordant couples with the same HPV types, suggesting that condom use prevents reinfection by blocking HPV transmission during intercourse [24].
Condoms are presumed to be less protective against HPV because the virus is transmitted via skin-to-skin contact and condoms do not cover all male genital skin. When properly used, condoms cover the penile head and much of the shaft but not the base of the penis and scrotum. It is clear that HPV infections occur at uncovered sites, such as the scrotum [25]. Since penile specimens in this study were collected from 3 anatomic sites, including the scrotum, and combined for HPV testing, we were unable to evaluate whether condoms protected specifically against infection at the penile head or shaft. Therefore, the protective effect of condoms may have been attenuated by HPV transmission at uncovered sites.
Measurement of the frequency of condom use is complex and susceptible to error [26, 27]. Most observational studies rely on self-reports, as in the HIM Study; however, condom efficacy is likely underestimated because men tend to overreport condom use [27]. Even with consistent use, condom efficacy is limited by late application, early removal, slippage, and breakage, which expose the genital skin to potential HPV infection. Since our study could not account for incorrect condom use or instances of condom failure, the condom efficacy may have been underestimated [26–28]. We were also unable to exclude men who used condoms with HPV-uninfected partners, because the partners’ HPV status was unknown. By including men with no exposure to HPV, condom efficacy may have been further underestimated [26]. Future prospective studies involving men and their sex partners should incorporate qualitative and quantitative methods to more accurately assess condom use and skills. With electronic daily diaries of sexual behavior, for example, researchers can calculate the proportion of condom-protected sexual events and account for instances of condom error/failure [27]. Last, we acknowledge that condom use behaviors among men may change over time, which may have led to exposure misclassification throughout the follow-up period; however, we deliberately chose to measure condom use at baseline because it provided a more straightforward analysis and interpretation.
To our knowledge, this is the largest, most comprehensive prospective study conducted to date of the efficacy of condom use on HPV prevalence, incidence, and clearance in adult men. Computer-assisted self-interviews were chosen to minimize bias frequently associated with asking highly sensitive questions. Furthermore, the survey instrument was specifically designed to improve precision in addressing the relationship between sexual behavior and condom use. All questions referred to behaviors in the last 3 or 6 months, minimizing the potential for recall bias, and to sex act- and partner-specific condom use, unlike previous research, which used broad measures of condom use [29]. Last, analyzing condom use on the basis of the participants’ HPV risk potential (ie, risk group) allowed us to better understand which groups of men benefit most from condom use.
Prophylactic HPV vaccination offers the most effective protection against HPV; however, the vaccine is only routinely available to those aged ≤26 years. Furthermore, the vaccine does not protect against all disease-causing HPV types. Consistent condom use, on the other hand, is a cost-effective risk-reduction strategy that may provide protection against HPV types not included in the vaccine and offers substantial protection against other STIs. Population-based vaccination combined with condom use promotion will be the optimal method to prevent HPV infection, reduce HPV transmission, and decrease the burden of HPV-related diseases among men and women.
In summary, our data have shown that consistent condom use appears to offer protection against HPV in men. Specifically, condom use reduces the risk of acquiring new HPV infections and decreases the duration of HPV infections among high-risk men. Nonmonogamous men and men with no steady sex partner received a greater benefit from consistent condom use than monogamous men. Correct and consistent condom use protects against other STIs and should be promoted as an STI risk-reduction strategy. HPV prevention efforts should focus on both condom use promotion and HPV vaccination.
Notes
Acknowledgments. We thank the HIM Study teams and participants in the United States (Tampa), Brazil (São Paulo), and Mexico (Cuernavaca).
Financial support. This work was supported by the National Cancer Institute at the National Institutes of Health (cancer prevention fellowship grant number R25T CA147832 to C. M. P. C. and grant CA R01CA098803 to A. R. G.).
Potential conflicts of interest. A. R. G. receives research funding from Merck and GSK. L. L. V. and A. R. G. are consultants for Merck Sharp and Dohme for the quadrivalent HPV vaccine. All other authors report no potential conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1.IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Human papillomaviruses. IARC Monogr Eval Carcinog Risks Hum. 2007;90:1–636. [PMC free article] [PubMed] [Google Scholar]
- 2.Giuliano AR, Lazcano-Ponce E, Villa LL, et al. The human papillomavirus infection in men study: human papillomavirus prevalence and type distribution among men residing in Brazil, Mexico, and the United States. Cancer Epidemiol Biomarkers Prev. 2008;17:2036–43. doi: 10.1158/1055-9965.EPI-08-0151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Giuliano AR, Lee JH, Fulp W, et al. Incidence and clearance of genital human papillomavirus infection in men (HIM): a cohort study. Lancet. 2011;377:932–40. doi: 10.1016/S0140-6736(10)62342-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hariri S, Unger ER, Sternberg M, et al. Prevalence of genital human papillomavirus among females in the United States, the National Health And Nutrition Examination Survey, 2003–2006. J Infect Dis. 2011;204:566–73. doi: 10.1093/infdis/jir341. [DOI] [PubMed] [Google Scholar]
- 5.Lacey CJ, Lowndes CM, Shah KV. Chapter 4: Burden and management of non-cancerous HPV-related conditions: HPV-6/11 disease. Vaccine. 2006;24(Suppl 3):S3/35–41. doi: 10.1016/j.vaccine.2006.06.015. [DOI] [PubMed] [Google Scholar]
- 6.Munoz N, Castellsague X, de Gonzalez AB, Gissmann L. Chapter 1: HPV in the etiology of human cancer. Vaccine. 2006;24(Suppl 3):S3/1–10. doi: 10.1016/j.vaccine.2006.05.115. [DOI] [PubMed] [Google Scholar]
- 7.Hirth JM, Tan A, Wilkinson GS, Berenson AB. Completion of the human papillomavirus vaccine series among insured females between 2006 and 2009. Cancer. 2012;118:5623–9. doi: 10.1002/cncr.27598. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Centers for Disease Control and Prevention (CDC) National, state, and local area vaccination coverage among adolescents aged 13–17 years—United States, 2009. MMWR Morb Mortal Wkly Rep. 2010;59:1018–23. [PubMed] [Google Scholar]
- 9.Holmes KK, Levine R, Weaver M. Effectiveness of condoms in preventing sexually transmitted infections. Bull World Health Organ. 2004;82:454–61. [PMC free article] [PubMed] [Google Scholar]
- 10.Winer RL, Hughes JP, Feng Q, et al. Condom use and the risk of genital human papillomavirus infection in young women. N Engl J Med. 2006;354:2645–54. doi: 10.1056/NEJMoa053284. [DOI] [PubMed] [Google Scholar]
- 11.Dunne EF, Nielson CM, Stone KM, Markowitz LE, Giuliano AR. Prevalence of HPV infection among men: A systematic review of the literature. J Infect Dis. 2006;194:1044–57. doi: 10.1086/507432. [DOI] [PubMed] [Google Scholar]
- 12.Baldwin SB, Wallace DR, Papenfuss MR, Abrahamsen M, Vaught LC, Giuliano AR. Condom use and other factors affecting penile human papillomavirus detection in men attending a sexually transmitted disease clinic. Sex Transm Dis. 2004;31:601–7. doi: 10.1097/01.olq.0000140012.02703.10. [DOI] [PubMed] [Google Scholar]
- 13.Hippelainen M, Syrjanen S, Koskela H, Pulkkinen J, Saarikoski S, Syrjanen K. Prevalence and risk factors of genital human papillomavirus (HPV) infections in healthy males: a study on Finnish conscripts. Sex Transm Dis. 1993;20:321–8. [PubMed] [Google Scholar]
- 14.Nielson CM, Harris RB, Nyitray AG, Dunne EF, Stone KM, Giuliano AR. Consistent condom use is associated with lower prevalence of human papillomavirus infection in men. J Infect Dis. 2010;202:445–51. doi: 10.1086/653708. [DOI] [PubMed] [Google Scholar]
- 15.Repp KK, Nielson CM, Fu R, et al. Male human papillomavirus prevalence and association with condom use in Brazil, Mexico, and the United States. J Infect Dis. 2012;205:1287–93. doi: 10.1093/infdis/jis181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kjaer SK, Munk C, Winther JF, Jorgensen HO, Meijer CJ, van den Brule AJ. Acquisition and persistence of human papillomavirus infection in younger men: a prospective follow-up study among Danish soldiers. Cancer Epidemiol Biomarkers Prev. 2005;14:1528–33. doi: 10.1158/1055-9965.EPI-04-0754. [DOI] [PubMed] [Google Scholar]
- 17.Bleeker MC, Hogewoning CJ, Voorhorst FJ, et al. Condom use promotes regression of human papillomavirus-associated penile lesions in male sexual partners of women with cervical intraepithelial neoplasia. Int J Cancer. 2003;107:804–10. doi: 10.1002/ijc.11473. [DOI] [PubMed] [Google Scholar]
- 18.Gravitt PE, Peyton CL, Apple RJ, 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;36:3020–7. doi: 10.1128/jcm.36.10.3020-3027.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bouvard V, Baan R, Straif K, et al. A review of human carcinogens—part B: biological agents. Lancet Oncol. 2009;10:321–2. doi: 10.1016/s1470-2045(09)70096-8. [DOI] [PubMed] [Google Scholar]
- 20.Ying Z, Wei LJ. The Kaplan-Meier estimate for dependent failure time observations. J Multivar Anal. 1994;50:17–29. [Google Scholar]
- 21.Lin DY, Wei LJ. The robust inference for the Cox proportional hazards model. J Am Stat Assoc. 1989;84:1074–8. [Google Scholar]
- 22.Wen LM, Estcourt CS, Simpson JM, Mindel A. Risk factors for the acquisition of genital warts: are condoms protective? Sex Transm Infect. 1999;75:312–6. doi: 10.1136/sti.75.5.312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Anic GM, Lee JH, Villa LL, et al. Risk factors for incident condyloma in a multinational cohort of men: the HIM Study. J Infect Dis. 2012;205:789–93. doi: 10.1093/infdis/jir851. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bleeker MC, Berkhof J, Hogewoning CJ, et al. HPV type concordance in sexual couples determines the effect of condoms on regression of flat penile lesions. Br J Cancer. 2005;92:1388–92. doi: 10.1038/sj.bjc.6602524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Giuliano AR, Nielson CM, Flores R, et al. The optimal anatomic sites for sampling heterosexual men for human papillomavirus (HPV) detection: the HPV detection in men study. J Infect Dis. 2007;196:1146–52. doi: 10.1086/521629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Crosby R, Bounse S. Condom effectiveness: where are we now? Sex Health. 2012;9:10–7. doi: 10.1071/SH11036. [DOI] [PubMed] [Google Scholar]
- 27.Graham CA, Crosby RA, Sanders SA, Yarber WL. Assessment of condom use in men and women. Annu Rev Sex Res. 2005;16:20–52. [PubMed] [Google Scholar]
- 28.Crosby R, Diclemente RJ, Yarber WL, Snow G, Troutman A. Refining self-reported condom use among young men at risk of HIV acquisition. Sex Health. 2007;4:211–2. doi: 10.1071/sh07039. [DOI] [PubMed] [Google Scholar]
- 29.Nyitray AG, Kim J, Hsu CH, et al. Test-retest reliability of a sexual behavior interview for men residing in Brazil, Mexico, and the United States: the HPV in Men (HIM) Study. Am J Epidemiol. 2009;170:965–74. doi: 10.1093/aje/kwp225. [DOI] [PMC free article] [PubMed] [Google Scholar]