Human Papillomavirus Prevalence Among 88 Male Virgins Residing in Brazil, Mexico, and the United States

Abstract

This study determined the prevalence and risk factors for genital human papillomavirus (HPV) detection among men who deny ever engaging in penetrative sex. A questionnaire was administered to 4123 men from a cohort study of HPV natural history. Genital exfoliated cells were collected and genotyped for 36 HPV types. Eighty-eight men were classified as virgins. Log-binomial regression models identified factors associated with genital HPV detection. The prevalence of any and high-risk HPV types among 88 male virgins was 25.0% and 18.2%, respectively. Age and smoking status were associated with HPV detection. Further studies are needed to better understand the risk for HPV infection among male virgins.

Human papillomavirus (HPV) infection, the most common sexually transmitted infection worldwide, is a necessary cause of cervical cancer and is also causally associated with anal, vulvar, vaginal, penile, and oral cancers [1]. Additionally, HPV causes typically nonmalignant clinical lesions, such as genital warts and recurrent respiratory papillomatosis [1].

Penetrative sexual intercourse is likely the primary mode of HPV transmission. However, there is a growing body of research suggesting that HPV may be transmitted vertically from mother to infant, by nonpenetrative sex, by nonsexual skin-to-skin contact, and through fomites contaminated by HPV [2]. While a few studies have examined HPV infection among children and female virgins [2], to our knowledge there are almost no data on HPV prevalence among male virgins [3]. Studying HPV in male virgins may help to improve our understanding of HPV transmission. Our objectives were to determine the prevalence of genital HPV infection among male virgins at baseline in a large cohort study of genital HPV among men and to identify risk factors for genital HPV infection among these men.

METHODS

Study Population

From June 2005 to September 2009, healthy men were recruited from Tampa, Florida, São Paulo, Brazil, and Cuernavaca, Mexico, for the prospective HPV Infection in Men study. Details of the study have been described elsewhere [4]. The study population consisted of men who were aged 18–70 years, residents of southern Florida, São Paulo, or Cuernavaca; had no prior diagnosis of penile cancer, anal cancer, or genital warts; had no symptoms of a sexually transmitted infection (STI) or treatment for an STI; had not participated in an HPV vaccine study; and had no history of human immunodeficiency virus (HIV) infection or AIDS. Being sexually active was not an inclusion criterion. All study procedures were approved by human subjects' committees at each study site.

At each visit, participants completed a computer-assisted self-interview (CASI) to encourage disclosure of potentially stigmatized behaviors. Exfoliated cell specimens were collected from the genitals of each participant, using 3 different prewetted polyethylene terephthalate swabs to sample the glans penis/coronal sulcus, the shaft, and scrotum. These swabs were later combined to produce a single specimen for the detection of genital HPV. Samples were tested for the presence of HPV with the polymerase chain reaction consensus primer system (PGMY 09/11). HPV genotyping was conducted using the Linear Array assay (Roche Diagnostics, Indianapolis, IN) to detect 36 HPV types that were classified as high risk (HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) or low risk (HPV types 6, 11, 26, 34, 40, 42, 44, 53, 54, 61, 62, 66, 67, 69, 70, 71, 72, 73, 81, 82, 83, 84, and 89). All plates included controls for false-positive and false-negative results, to reduce laboratory error. Only samples that tested positive for β-globin or an HPV genotype were considered adequate and included in the analysis. Baseline CASI findings and genotyping results were used in this analysis.

Data Analysis

Men were classified as virgins or nonvirgins on the basis of their answers to sexual behavior–related questions, the presence of non-HPV STIs, relationship status, and HPV 6/11/16/18 antibody status. Regarding sexual behavior, men were defined as virgins if they acknowledge no sex on 20 sexual behavior questions, including (1) Have you ever performed vaginal sex (your penis in partner's vagina)? (2) Have you ever had oral sex (your penis in your partner's mouth or your partner's vagina in your mouth or your partner's penis in your mouth)? (3) Have you ever performed insertive anal sex (your penis in partner's anus)? and (4) Have you ever had receptive anal sex (your partner's penis in your anus). Additionally, men classified as virgins (1) must have been negative for genital herpes, chlamydial infection, gonorrhea, and syphilis at baseline on the basis of study-based tests; (2) must have self-reported being HIV negative; (3) must have reported a marital status of single, never married; and (4) must have been negative for HPV 6/11/16/18 antibodies on the basis of study-based assays.

Of 4123 men enrolled at baseline, the following were removed from analyses: men aged <18 or >70 years (n = 5), men who had inadequate specimens at baseline (n = 106), men who acknowledged HIV infection after enrollment (n = 72), and men whose sexual behavior variables were missing (n = 18). Of the remaining 3922 men, 88 (2.2%) were classified as virgins.

Potential risk factors based on previous male HPV studies were selected for analyses, including age (continuous), residence, race, ethnicity, education level, number of alcoholic drinks in the past month, cigarette smoking status, and circumcision status. A former smoker was defined as a man who had smoked >100 cigarettes in his lifetime but did not smoke at baseline; a current smoker was a man who smoked >100 cigarettes in his lifetime and smoked at baseline.

The primary outcome of interest was any HPV infection (positive vs negative). A specimen was considered positive for any HPV if it tested positive for ≥1 of 36 genotypes. The secondary outcomes of interest were high-risk and low-risk HPV infection. A specimen was considered positive for a high-risk type if any of the above-mentioned 13 high-risk genotypes were detected, regardless of the presence of other genotypes. A specimen was positive for low-risk HPV if any of the low-risk genotypes were present without coinfection with high-risk types.

Bivariate and weighted multivariable log-binomial models [5] were used to estimate prevalence ratios (PRs) and 95% confidence intervals (CIs) for any type and for high-risk types. Weighted multivariable log-binomial models were used because some log-binomial models did not converge [5]. Multivariable log-binomial models without weights were used for low-risk types. Variables with a P value of <.25 in bivariate analysis were included in multivariable regression. Using directed acyclic graphs, age and residence were identified as potential confounders and retained in all multivariable models. A backward variable selection method was used to remove variables 1 by 1 if the variable was not statistically significant (P < .05); however, if removing the variable caused a >10% change in the point estimates of other variable measure of associations, that variable was added back to the model. Analyses were conducted using SAS software (version 9.4; SAS Institute, Cary, North Carolina). All statistical tests were 2 sided with a significance value of α = 0.05.

RESULTS

The distribution of sociodemographic characteristics was different between virgins and nonvirgins, with the exception of race and circumcision status. Virgins were more likely to be younger, to reside in Tampa, to be non-Hispanic, to report no alcohol consumption in the past month, and to be never smokers (Supplementary Table 1).

The prevalence of any HPV type, high-risk, and low-risk HPV infection among 88 male virgins was 25.0%, 18.2%, and 6.8%, respectively, which was significantly lower than for nonvirgins (52.6%, 29.5%, and 23.0%, respectively; Table 1).

Table 1.

Prevalence of Genital Human Papillomavirus (HPV) at Baseline in Male Nonvirgins and Male Virgins in the HPV Infection in Men Study, Tampa, São Paulo, and Cuernavaca, 2005–2009

Variable	Nonvirgins, No. (%) (n = 3834)	Virgins, No. (%) (n = 88)	P Valuea
HPV genotype group
Any	2012 (52.5)	22 (25.0)	< .001
High risk	1131 (29.5)	16 (18.2)	.02
Low risk	881 (23.0)	6 (6.8)	< .001
High-risk genotypes
16	294 (7.7)	5 (5.8)
18	84 (2.2)	0
31	70 (1.8)	1 (1.2)
33	24 (0.6)	0
35	59 (1.5)	2 (2.3)
39	142 (3.7)	1 (1.2)
45	63 (1.7)	0
51	242 (6.3)	4 (4.6)
52	153 (4.0)	1 (1.2)
56	71 (1.9)	1 (1.2)
58	95 (2.5)	0
59	211 (5.5)	2 (2.3)
68	85 (2.2)	2 (2.3)
Low-risk genotypes
6	240 (6.3)	1 (1.2)	.04
11	51 (1.3)	0
26	8 (0.2)	0
34b	4 (0.1)	0
40	62 (1.6)	0
42	54 (1.4)	0
44b	94 (2.5)	1 (1.2)
53	215 (5.7)	3 (3.5)
54	86 (2.3)	0
61	183 (4.8)	1 (1.2)
62	309 (8.1)	2 (2.3)	.04
66	198 (5.2)	4 (4.6)
67	19 (0.5)	1 (1.2)
69	6 (0.2)	2 (2.3)	.01
70	100 (2.6)	0
71	57 (1.5)	0
72	47 (1.2)	0
73	68 (1.8)	0
81	118 (3.1)	1 (1.2)
82b	30 (0.8)	0
83	106 (2.8)	2 (2.3)
84	311 (8.1)	2 (2.3)	.04
89b	270 (7.1)	2 (2.3)

^a By the χ² test or Fisher exact test. Only P values ≤.05 are shown.

^b HPV 44 was previously known as HPV 55, HPV 34 was previously known as HPV 64, HPV 82 included subtype IS39, and HPV 89 previously included CP6108.

In bivariate analyses, age, alcohol drinking in the past month, and cigarette smoking were significantly associated with HPV infection (Supplementary Table 2). In multivariable analyses, age was significantly associated with any HPV (adjusted PR [aPR], 1.05; 95% CI, 1.01–1.08) and high-risk HPV (aPR, 1.04; 95% CI, 1.01–1.09; Table 2). After adjustment for age, residence, and number of alcoholic drinks consumed in the past month, current smoking remained significantly associated with any HPV (aPR, 3.58; 95% CI, 1.81–7.08). After adjustment for age and residence, current smoking was significantly associated with high-risk HPV (aPR, 4.08; 95% CI, 1.52–10.94). Also, former smoking was associated with any HPV (aPR, 3.15; 95% CI, 1.84–5.40).

Table 2.

Multivariable Analyses of Factors Associated With Genital Human Papillomavirus (HPV) Detection at Baseline in 88 Male Virgins in the HPV Infection in Men Study, Tampa, São Paulo, and Cuernavaca, 2005–2009

Factor	Any HPV	High Risk	Low Risk
	aPR (95% CI)	aPR (95% CI)	aPR (95% CI)
Age (continuous)	1.05 (1.01–1.08)	1.04 (1.01–1.09)	1.06 (1.00–1.15)
Cigarette smoking status
Never smoker	1.0	1.0	NEa
Former smoker	3.15 (1.84–5.40)	2.65 (.78–9.04)	NEa
Current smoker	3.58 (1.81–7.08)	4.08 (1.52–10.94)	NEa

Age and residence were forced into all models as confounders. PRs for any HPV were also adjusted for alcoholic drinks. All PRs were adjusted for other variables in the model.

Abbreviations: aPR, adjusted prevalence ratio; CI, confidence interval; PR, prevalence ratio.

^a Not estimable (NE), owing to 0 values for both former and current smokers.

DISCUSSION

In this study, we examined the prevalence of genital HPV among male virgins in the HPV Infection in Men (HIM) study cohort. The prevalence of any HPV, high-risk HPV, and low-risk HPV infection among virgins was 25.0%, 18.2%, and 6.8%, respectively. Age and current smoking were associated with any and high-risk HPV. Former smoking was associated with any HPV detection. The percentage of virgins in this cohort (2.2%) was lower than the percentage of virgins (4.7%) in a recent population-based survey in the United States [6]. Since most came from the Tampa clinic site, as opposed to Cuernavaca or São Paulo, virgins also tended to be non-Hispanic.

Genital HPV was more prevalent in this study of male virgins than previously reported prevalence in studies of female virgins with cervicovaginal HPV DNA, which ranged from 4.2% to 18% [2]. One previous study of genital HPV in men, aged 15–20 years, in Cuernavaca observed no prevalent HPV in 18 men who reported no previous sexual intercourse [3]. However, their specimen collection method was minimal as compared to that in the current study, with only urine, urethral, and coronal sulcus specimens collected, as opposed to the current study's collection of exfoliated cells from the glans penis/coronal sulcus, penile shaft, and scrotum.

One explanation for the common prevalence of HPV in our study could be HPV transmission through nonpenetrative sexual contact (eg, hand-genital contact and genital-genital contact), which is common among persons self-reporting as virgins. A study of adolescent virgins reported that approximately 30% of self-reported virgins had engaged in masturbation with a partner in the past year [7]. Another study among female virgins reported that finger-vulvar, penile-vulvar, and oral-penile sex increased the risk of genital HPV infection in female virgins [8]. It may be possible that nonpenetrative sexual contact may also increase the risk of genital HPV infection in male virgins; however, our study did not collect data on nonpenetrative sex.

Our study observed an association between age and HPV detection among virgins. The HIM study found that age was not associated with acquisition of genital HPV among men in general [9]. In most HPV studies among nonvirgins [10], age has not been associated with genital HPV prevalence, although increasing age was associated with declining genital HPV detection among men in one study [11]; however, the only study finding an increasing prevalence among mostly nonvirgins is the Cuernavaca study among 120 men—a study with very different specimen collection methods, as stated previously [3]. Some reasons for our observation of increasing prevalence with age could be that (1) men may accumulate more chances for HPV acquisition through life through nonsexual means (eg, if there is increasing hand-genital contact with age that produces increased prevalence as a result of hand carriage of HPV [12]) and (2) men may have been more likely to be misclassified as virgins as they increased in age.

Current smoking was associated with any and high-risk HPV infection, and former smoking was associated with any HPV infection among male virgins. Smoking could increase HPV infection by suppressing the immune system [13]. No data exist on the association between smoking and HPV infection among male virgins; however, our findings are consistent with those in other HPV studies in men that reported an increased risk of genital HPV infection among current smokers [14, 15]. For example, Vaccarella et al [14] found that current (odds ratio, 2.3; 95% CI, 1.1–4.6) and former (odds ratio, 2.3; 95% CI, 1.2–4.3) smokers had higher odds of HPV infection than never smokers. Schabath et al [15] analyzed the baseline data from the HIM Study and observed that current smoking was associated with an increased risk of any and high-risk HPV infection among men reporting fewer or no lifetime sex partners.

To our knowledge, this is the first multinational study of the prevalence of genital HPV among male virgins. All virgins were from the large HIM Study cohort, in which reliable HPV DNA sampling and testing methods were used. Additionally, male virgins were strictly defined in our analyses. However, our study is not without limitations. Information on other potential risk factors and sexual behaviors other than vaginal, oral, or anal penetrative sex was not collected. Owing to the small sample size of virgins (n = 88), there could be lack of statistical power in multivariable modeling to detect statistically significant relationships between HPV infection and certain risk factors, such as alcohol drinking.

In summary, the results demonstrated that genital HPV was common among male virgins. Our findings add to the literature on the potential transmissibility of HPV infection outside of penetrative sexual intercourse and vertical transmission, although implications for disease and HPV vaccination are unknown. Additional research is warranted, especially prospective studies that characterize the acquisition and persistence of genital HPV among virgins and recent nonvirgins.

Supplementary Data

Supplementary materials are available at http://jid.oxfordjournals.org. Consisting of data provided by the author to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the author, so questions or comments should be addressed to the author.