We found a prevalence of HR-HPV infections of 19% in a sample of Italian women aged 18–26 years. This percentage is in agreement with an Italian study which used a similar methodology to measure the HR-HPV prevalence in 18–24 year-old women from Tuscany Region . This value is also included within the range (20-45%) reported in other national or worldwide studies for women younger than 25 years [12, 19–22], although it is difficult to compare results of different studies because of different age groups and different procedures for enrolling participants. Not surprisingly, the rate of infection in this age group was higher than what was found in women aged 25–29 (14%) in New Technologies for CC screening (NTCC) randomized trials, which studied 45,000 women aged 25–60 years of northern and central Italy, participating in CC screening programmes [8, 9]. This finding reflects a higher probability of acquiring new infections at younger ages, confirming the trend described in the international literature with a peak of HR-HPV prevalence in younger women and a continuous decline with increasing age [19–22].
Cervical cancer incidence is lower in the South than in the North of Italy, therefore a lower HR-HPV prevalence in the general population was expected in the South. Instead, we did not find any difference in HR-HPV prevalence rates between northern, central and southern regions of Italy, confirming what reported by Agarossi and coll. . This finding should encourage the implementation, strengthening and promotion of cervical cancer screening programmes in southern Italy, where coverage and acceptance are still lower than in the rest of the country .
Genotype HPV16 was detected in 31% of HR-HC2 positive samples, followed by 31, 66, 51 and 18. This is well consistent with the results of a comprehensive meta-analysis on HPV positive women with normal cytology conducted by de Sanjosé and coll. , which reported HPV16, 31 and 18 among the five most common types worldwide. In Italy, HPV16 resulted to be the genotype most frequently detected in all studies and HPV 31 was frequently reported as the second most common genotype [11, 22, 25, 26]. Regarding the other genotypes a high variability is reported among studies, which could be due to geographical differences, different target populations, different methods for genotyping and random fluctuation for quite rare genotypes [27, 28].
According to the International Agency for Research on Cancer (IARC) classification, HPV16, 31, 51 and 18 are classified as “carcinogenic to humans”, while HPV 66 is classified as “possibly carcinogenic” . It should be also mentioned that the prevalence of HPV 66, as of the other genotypes not targeted by HR-HC2 (HPV 53, 26, 73, 82), observed in the present study is plausibly underestimated because the detection of these types is only due to an occasional cross-hybridization of the method or co-infection with HPV-targeted types.
As expected, we found that women with cervical cytological abnormalities were at significantly increased risk for being infected by HR-HPV types than women with normal cytology (52 vs. 15%). Current Italian studies involving women with cytological abnormalities (different ages and enrolment criteria) reported HR-HPV prevalence of 34-68%, increasing with cytology severity: 24-56% in case of diagnosis of ASCUS/AGUS, 42-72% in LSIL and 73-95% in HSIL [21, 22, 26, 27, 30–32].
Co-infections of HR-HC2 targeted types represented the 20% of HPV positive samples. This value is included in the range (15-50%) reported in literature ; similar percentages of single and multiple HPV infections have been observed in young general populations [34, 35]. It is still not clear whether co-infection with multiple types increases the risk of progression to cancer [22, 36].
The multivariate analysis evidenced the role of the number of lifetime sexual partners as determinant of HPV infection, consistently with other studies [11, 12, 21, 37], and strengthens the concept that that the most suitable age for HPV vaccination is the period preceding sexual activity. Living with partner had a protective effect against HR-HPV infection in the multivariate model; consistently with another Italian study , being married and having children showed a protective effect in the crude analysis, although not statistically significant in the multivariate analysis. All these variables could be considered markers of a steady relationship, explaining the association with a low HR-HPV prevalence. The use of any contraceptive method in the last six months remained associated to HR-HPV infection in the final model; we could suppose that women who used contraceptives in the last six months could have had a more intense sexual activity, with a higher risk of acquiring HPV, whose prevalence is higher in young ages. If considering only the use of condoms in the last six months, no association was detected between HR-HPV prevalence and this method of contraception; this point is debated and conclusions about the association between condoms’ use and HR-HPV prevalence are discordant among authors .
The major strengths of this study are that the sample was large and it targeted an age group that has not been investigated extensively.
Among the limitations, it should be mentioned the fact that our sample may not be entirely representative of Italy’s general population of females aged 18–26 years because only ten LHUs in six Regions participated, though local probabilistic samples were population-based and both urban and rural LHUs in northern, central and southern Italy were involved.
In addition the participation rate was 52%, therefore our findings could not be representative of the entire study population; however differences in socio-demographic characteristics between participants and people who declined participation were minimal.
The participation rate could be underestimated because we have excluded from the initial sample the 154 women who spontaneously declared not to be sexually active, but we do not know if other women, who declined participation, were virgin too (we did not collect this information). This age group represents a difficult target for prevention measures, because in Italy young adult women are not accustomed to being targeted by preventive programmes; moreover, enrolled women were offered the participation to a package of activities within the PreGio project, which is more difficult to be accepted than a single cervical smear. On the other hand, it should be noted that the study personnel had received special training and that a high number of attempts were made to contact non-respondents. Two other studies [11, 12], which detected HPV prevalence of a sample of Italian women aged 18–24 years randomly selected from population registries, got a lower participation rate (15-22%). In these two studies letters of invitation were mailed to sampled women and a reminder was sent in case of no response, whereas we planned three phone calls and two home visits for non respondents.
As already mentioned, another limitation is that the prevalence of HPV 26, 53, 66, 73 and 82 could be heavily underestimated because they are not targeted by HR-HC2 test. In fact HPV 73 and 82 were observed only in co-infection with other HR-HPV types, suggesting that they could have been found only because coexisted with HPV types detected by HR-HC2 probes B (for this reason we excluded HR-HC2 not targeted types from the analysis of co-infections).