METHODS

Study population and questionnaire

On randomly chosen days during March to December 1993, men attending a STD clinic in Copenhagen, Denmark, were consecutively invited to participate in the study. After written informed consent, 216 men were included. Information about social and demographic factors, smoking, genital hygiene, circumcision, sexual habits, contraceptive use, and history of STDs was obtained by means of a personal interview using a structured questionnaire. Except for HIV, we did not obtain information about the concurrent presence of other STDs. If any clinical samples were taken, the participants were informed about these results but not about the HPV result, as this test was taken for research purposes only.

Biological material

Material for HPV detection was collected by firmly scraping with two pre-wetted cotton tipped plastic shafted swabs over the external genital area; one over glans and sulcus coronarius of the penis, and another over the shaft of the penis, the scrotum, and the perianal region. Immediately afterwards, both swabs were placed in the same plastic tube containing 3 ml phosphate buffered saline (PBS) with 0.005% thiomersal and kept at 4°C until forwarded by mail for analysis. After centrifugation for 10 minutes at 2755 g, the pellet was resuspended in 100 μl 10mM TrisHCl, pH 8.3.

Polymerase chain reaction

HPV DNA was detected by polymerase chain reaction (PCR) using general primers (GP) 5/6 which allows the detection of at least 27 mucosotropic HPV types.^14,15 Type specific (TS) PCR for HPV 6, 11, 16, 18, 31, and 33 was performed using combinations of HPV 6, 16, 33 and HPV 11, 18, 31 specific primers. Samples positive for GP-PCR but negative for TS-PCR were characterised as unclassified HPV, HPV X. All samples tested positive for the β globin gene. A cut point for HPV positivity on 100 pg was chosen and 18 samples with a weaker signal were excluded leaving 198 samples for analysis.

Statistical analysis

Odds ratios (OR) were computed with 95% confidence interval (CI) by means of unconditional multiple logistic regression to assess the association between potential risk factors and presence of HPV DNA while simultaneously adjusting for potential confounding variables. OR were computed for any HPV type as well as for oncogenic and non-oncogenic types separately. In all three analyses, the same control group was used—that is, the HPV negative (and β globin positive) men. HPV 16, 18, 31, and 33 were grouped together in the oncogenic group, whereas HPV 6 and HPV 11 formed the non-oncogenic group. The uncharacterised HPV types (HPV X) were only included in the overall analysis of risk factors for HPV (any type). All analyses were done using the SAS software package.¹⁶

RESULTS

Fifty four per cent of the participants were younger than 30 years, and respectively 33% and 13% were 30–39 years and 40 years or older. Most of the men (58%) reported 20 or more partners during lifetime whereas 16 men (8%) had ≤ 4 lifetime partners. Forty per cent, 23%, and 37% of the men reported respectively ≤ 2, 3–4 partners, and ≥ 5 partners during the past year. A total of 24 men (12%) were circumcised, and among the 46 study subjects who had sex with men, 35 were bisexual. Using condoms at every intercourse, only occasionally, or never was reported by, respectively, 38%, 54%, and 8%. Ever having had genital warts or Chlamydia trachomatis was reported by, respectively, 51 (25%) and 49 men (26%), three men were HIV positive, and 50% of the men were current smokers.

HPV DNA was found in 89 men (45%). HPV 16 was the most prevalent type, accounting for 15% of the HPV positives. The other oncogenic types—that is, HPV 18, HPV 31, and HPV 33 contributed with respectively, 8%, 4%, and 3%, and the non-oncogenic types HPV 6 and HPV 11 with respectively, 14% and 8%. HPV of unclassified type (HPV X) was found in 49 men (55%). Six of the HPV positive men (7%) had multiple infections.

Risk factors for HPV positivity (any type)

Risk of HPV increased with a greater number lifetime number of sexual partners. Compared to men with 1–9 partners, men with 10–24 partners had an OR = 4.1 (95% CI 1.5 to 11.1) and men with 25–39 partners had an OR = 4.3 (95% CI 1.4 to 13.1) (table 1). However, for 40 or more partners the risk estimate declined (OR = 2.0 (95% CI 0.6 to 6.7)). Similarly, an increasing number of partners in the past year was associated with an increased risk of HPV being 3.6 times higher for ≥10 partners than for ≤ 2 partners (95% CI 1.1 to 12.0). Risk of HPV was significantly lower in circumcised men compared to uncircumcised men and this association could not be explained taking the other variables into account (OR = 0.2 (95% CI 0.06 to 0.6)). Frequency of genital washing, however, was not associated with HPV and there was no clear influence of current condom use (data not shown). Other variables unrelated to risk of HPV DNA detection included age at first intercourse, contact with prostitutes, sexual preference, marital status (including years living with the current partner), years at school, and smoking habits (data not shown).

Risk factors for oncogenic HPV types

Age was the most important determinant for oncogenic HPV types with OR = 4.7 (95% CI 1.0 to 23.4) for men aged 18–24 years compared to those aged ≥35 years (table 2). Risk of oncogenic HPV types also increased with number of lifetime sexual partners and was 3.0 times (95% CI 1.0 to 9.6) more common in men with 20–39 partners than in men with fewer partners. However, risk of the oncogenic HPV types was not related to the number of partners in the past year. Oncogenic HPV types were less frequently found in circumcised men than in non-circumcised men, but the association did not reach statistical significance after adjustment (OR = 0.4; 95% CI 0.08 to 1.7). Finally, risk of these HPV types was not related to ever having had genital warts.

Risk factors for non-oncogenic HPV types

As with the oncogenic types, risk of the non-oncogenic HPV types increased with young age, being 6.3 times (95% CI 1.1 to 39.4) more common in the youngest men aged 18–24 years compared with those aged ≥ 35 years. Regarding the variables related to sexual behaviour, number of recent partners was the most important risk factor. Men who reported five or more different sexual partners in the past year were 4.1 times more likely to have non-oncogenic HPV types compared with the men who reported fewer partners (table 2). The association with lifetime number of partners was weaker and not statistically significant. Men who had ever had genital warts were 5.9 times (95% CI 1.8 to 18.9) more likely to have non-oncogenic HPV types. No association was found with circumcision.

DISCUSSION

This study shows that the determinants for genital HPV infection in men resemble those previously found in women. Lifetime number of sexual partners was an important predictor of HPV infection in men. In studies of women with a high sexual activity, risk of HPV has been found to level off after a certain number of partners.^5,7 We find the same pattern for the risk estimates in men. Likewise, the number of recent sexual partners—that is, in the past year—predicts risk of HPV in men with a high sexual activity as it has been found in women.^5,7

Decreasing risk of HPV with age has been found in several female populations.^3–5 We have previously reported that the age specific HPV prevalence declines in the same way among men participating in this study as it does in female attendees from the same clinic.⁵ It has been suggested that at least part of the explanation for the HPV age pattern in women is an acquired immunity that is strengthened over time due to repeated exposure.⁶ As shown in table 3, men with many recent partners (≥ 10 in the past year) had a lesser risk of HPV if they had been sexually active for many years (≥ 10 years) than if they had been sexually active for a shorter period. A similar relationship between proximity to first intercourse (exposure time) and number of partners in the past year (recent exposure) has previously been described in women.^5,7 Thus, our data support that some acquired immunity to genital HPV infection also occurs in men.

Some studies,¹⁷ but not all,¹⁸ have pointed to a lower risk of cervical cancer in women whose sexual partner is circumcised. In this study, the risk of HPV DNA in circumcised men was only one fifth of the risk in uncircumcised men, and this could not be explained by other variables such as number of partners and age. Thus, our results may suggest that the female partners of circumcised men are less exposed to cervical cancer because these men are less likely to be infected with HPV.

Different risk determinant profiles for oncogenic HPV types and non-oncogenic HPV types have been found in women.^8–11 Lifetime number of partners was more strongly related to risk of the oncogenic HPV types than to the risk of the non-oncogenic HPV types. Moreover, other variables related to lifetime exposure were associated with the risk of oncogenic types but not with risk of non-oncogenic types—for example, age^8,11 and years of sexual activity.^8,9 A possible explanation could be that the infections with oncogenic types last longer than the infections with the non-oncogenic types. Correspondingly, we find different patterns of risk factors for oncogenic and non-oncogenic HPV types in men. The risk of oncogenic HPV types was related to lifetime number of sexual partners whereas risk of non-oncogenic types was only related to recent sexual behaviour (number of sex partners in the past year). It seems, though, that the hypothesis of oncogenic HPV types causing infection of a longer duration and the non-oncogenic HPV types causing infections of a more transient nature also applies to HPV infection in men. In this study, however, young age seemed to be equally important for the risk of having oncogenic and non-oncogenic HPV types. Whereas ever having genital warts (caused by the non-oncogenic types HPV 6 and 11) was the most important risk factor for current presence of HPV 6 and 11, this variable was not related to risk of the oncogenic HPV types.

Detection of genital HPV is technically more complicated in men than in women because cells are more difficult to harvest from skin than from moist mucosal surfaces. However, a demand for a high sensitivity of the DNA detection method is fulfilled by the use of the PCR technique.^14,15 We have chosen to collect cells from all over the external genital surface—that is, the area that might come into contact with the female genital area during intercourse. It is possible that the sampling method used in this study has facilitated sampling of skin types of HPV as well as mucosotropic genital HPV types. Since other HPV types than the mucosotropic are not amplificated as well by the GP5/6-PCR, this may explain the high prevalence of HPV X. The HPV X group probably represents a mixture of oncogenic and non-oncogenic types. Supporting this hypothesis, we found the determinants of HPV X to be a mixture of the risk factors described above for the oncogenic and non-oncogenic type (data not shown) and this probably explains some of the variation in the ORs for any HPV compared to the ORs for the specific HPV types—for example, in relation to age and to circumcision status. HPV X was not associated with age and this dominates the analysis of any HPV leaving the estimates insignificant whereas excluding HPV X reveals a strong association to both the oncogenic and the non-oncogenic HPV types. HPV X was strongly related to circumcision status as none of the circumcised men had HPV X so a significant association was found with risk of any HPV but not with risk of the specific HPV types. Since these analyses were done, newer PCR techniques have been developed which type a broader spectrum of HPV types. Thus applying newer PCR techniques (as the 5+/6+ PCR) would probably reduce the number of HPV X but not induce any marked changes in the determinant analysis for overall HPV.

In conclusion, the most important risk factors for HPV detection in men resemble those reported in women—that is, number of sex partners and age. Moreover, the profiles of risk factors for oncogenic and non-oncogenic HPV types correspond well with those found in women. Finally, the lower HPV DNA detection rate observed in circumcised men is consistent with the suggested lower risk of cervical cancer in female partners of circumcised men.