Routine HPV screening for women aged 21 and 65 years can reduce the chance of cervical cancer development . Various HPV testing modalities are available . In this study, we chose the Digene Hybrid Capture 2 High-Risk HPV DNA Test, an approach approved by the US Food and Drug Administration for use in cervical cancer screening, that is highly specific, reproducible, sensitive and reliable [15–17]. With this test, we detected high-risk HPV DNA in 61.90 % of FSWs and in 21.00 % of healthy control subjects (Table 1). It has been documented that there are huge variations in HPV infection in FSWs between different geographic areas and different races. While over 80 % of FSWs were HPV-positive in southern Vietnam  and Hungary , HPV DNA was detected in 66.8 % of FSWs (n = 200) in Peru , 57.2 % of FSWs (n = 369) in Philippines , 49.5 % of FSWs (n = 281) in northern Vietnam , 47 % of FSWs (N = 148) in South Korea , 31.6 % of FSWs (n = 288) in Australia , 22.9 % of FSWs (n = 254) in Thailand  and 14.4 % of FSWs (n = 187) in Singapore . Compared with these worldwide data, the prevalence rate of HPV infection in FSWs detected in this study was in the upper range. Compared with reports from other Chinese studies, the HPV infection rate in FSWs in Shenyang was very close to that in FSWs in Huzhou (66.7 %) , but was much higher than that in FSWs in Guangxi (38.9 %) . Currently, the exact reasons underlying the reported difference in HPV prevalence in FSWs in different areas in China are unknown and further studies are warranted.
In consistency with HPV infection, cervical dysplasia, assessed either histologically or cytologically, was significantly more prevalent in FSWs than in healthy controls in this study (Table 1). In the current literature, data on cervical epithelial cell lesions in FSWs are scarce. In a study involving 90 FSWs aged 18 to 58 years in Antananarivo, Madagascar, the prevalence was 3.3 % for low-grade squamous intraepithelial lesions and 18.9 % for ASCUS while no high-grade lesion was detected . In another study from Hong Kang, CIN 1–3 lesions were observed in 9.8 % of 235 FSWs . Comparatively, squamous intraepithelial lesions, particularly ASC-US (32.04 %) and low-grade lesions, detected in this study were all higher than those reported in the two studies mentioned above. It has been well documented that free-of-charge screening services to FSWs is very helpful not only in early detection and proper follow-up in case of abnormal Pap tests but also in increasing the awareness of women’s health issues . Given the substantially high incidence of ASC-US and low-grade lesions observed in this study, a free or practically feasible and cost-efficient HPV screening program is urgently needed to minimize the risk for cervical cancer in FSWs in Shenyang.
Numerous factors have been assessed for a putative link with HPV infection in FSWs in previous studies. However, inconsistent results have been obtained. In this study, we assessed various socio-demographic and behavioral variables, among which age of first sexual intercourse and menopause were the only two risk factors for HPV infection in FSWs in Shenyang.
In agreement with our findings, young age (OR 0.699, 95 % CI 0.492–0.992) and menopause (OR 2.928, 95 % CI 1.099–7.800) have been suggested as independent risk factors for HPV infection in FSWs in previous studies [8, 10, 22, 24, 30]. The exact mechanisms underlying increased vulnerability to HPV infection in young and post-menopausal FSWs are not fully understood. However, inadequate acquired immunity at young age and decreased acquired immunity at post-menopausal age might be one of the explanations . Supporting this speculation, Sivro and colleagues have demonstrated that there are significant hormone-associated changes in systemic and mucosal cytokine/chemokine production, which may have implications for the age-related decline in the ability to fight against infections in young and post-menopausal women .
Education levels [10, 18] and years of commercial sex work  have been demonstrated to affect HPV infection in FSWs in other studies. Inconsistent with these studies, the present study showed that HPV infection in FSWs in Shenyang was not associated with education levels or years of commercial sex work.
The impact of cigarette smoking on HPV prevalence has been assessed in both men and women . However, in the current literature, little information is available regarding the difference in HPV infection between smoking and non-smoking FSWs, although it has been reported that the risk for FSWs to develop preinvasive or invasive cervical lesions increase with the number of cigarettes they smoked per day and their years of smoking . In this study, the effect of smoking, either active or passive, on HPV infection was not demonstrated in FSWs in Shenyang and warrants further assessment in the future.
The benefit of condom use in protecting HPV infection in women is debatable . While a role for condom use in reducing HPV infection in FSWs has been reported [11, 27, 30], there is an argument that condoms may not completely cover the infected areas, thus rendering insufficient protection against HPV infection . In this study, no significant difference was found in HPV prevalence between condom users and non-condom users, supporting a limited capacity of condoms in HPV protection.
In this study, HPV infection in FSWs in Shenyang was associated with all other socio-demographic, medical and behavioral variables including marital status, number of abortions, sexual disease history, use of contraception pills and other medications, age of puberty, and previous gynecologic and obstetric examinations. Nevertheless, these findings need to be further validated in studies with a large sample size.
This study has several limitations. First, ThinPap test and cytological evaluation were performed only once on both FSWs and control subjects. This cannot totally rule out the possibility of miss the true abnormalities. Second, in evaluating the usefulness of condom use in protecting against HPV infection, only the information on ‘current condom use’ at the time of interview was collected. This might not represent a complete picture. A more accurate answer would have been obtained if condom use had been stratified into three scenarios: no condom use either in the past or at the present; condom use always; and condom use sometimes.