This is the first large-scale cross-sectional study to evaluate the prevalence and epidemiological characteristics of HTLV-1/2 infections and co-infections with other pathogens among different population in Central China. Our findings indicate that the final HTLV-1/2 seroprevalence rate was 0.18% (10/5480), yielding a prevalence of 0.13% for HTLV-1 and 0.05% for HTLV-2. The lowest prevalence of HTLV-1/2 infection is found in blood donors (0.06% and 0.03%). However, population of high risk group shows the high level of HTLV-1 and HTLV-2 infection (0.39% and 0.2%, respectively). Moreover,these findings suggest that HTLV-1/2 seroprevalence is associated with HIV, HBV and HCV infections. Among 10 HTLV-1/2 positive cases, 9 were co-infected with HBV, 3 with HCV and 1 with HIV.
Due to inadequate treatment options for TPS/HAM or ATL and lack of an effective vaccination, prevention of new HTLV infections is currently only possible by restricting transmission, either blood transfusion or sexually intercourse. In this regard, knowing the HTLV prevalence status seems to be the most effective way to reduce the transmission of HTLV. China is non-endemic area for these viruses. In 1980s, Zeng et al. 
conducted an epidemiological investigation in 10,013 blood samples from 28 provinces of China, and found 8 had antibody to HTLV-1. 3 were Japanese living in China and 2 were from Taiwan. 2 were Chinese females from mainland China which husband were Japanese and Taiwan with seropositive. Only one ATL patient was Chinese but often lived in Japan. There was no native citizen infected with HTLV. However, some recent survey show that HTLV infection cases have been found in some provinces of the country, and the prevalence rate in health blood donors were 0.06–1.27% 
. The coastal areas of Fujian province and Guangdong province show higher prevalence. In this study, we found 2 HTLV-1 positive cases and 1 HTLV-2 positive case among 3548 blood donors, the prevalence was higher than previous investigations in this area and contemporaneous investigations of the other regions 
. Although 1 case (HTLV-1) came from Wuhan blood center and the other 2 cases came from Henan Red Cross Blood Center, we could not determine the exact origin of these HTLV positive individuals due to the lack of the native place information for each sample.
ATL is one of the major HTLV-1 associated diseases. 2–6% HTLV-1 carriers could develop ATL if infection occurs in early childhood 
. Barrientos and his colleagues 
tested HTLV-1/2 antibody among 88 patients with malignant hematological diseases in Southern Chile. The viral prevalence was 18% among these patients. In Okinawa, Japan, an endemic HTLV-1 region, Miyagi et al. 
found an HTLV-1 prevalence of 26.1% in 88 cases of non-Hodgkin's lymphoma. In 1997, it has been reported that the seroprevalence was 28.9% among patients with T-cell lymphoid malignancies in Brazil 
. In this study, only one case was found to be HTLV-1 antibody positive among 908 lymphoma/leukemia in Central China, the prevalence rate was 0.11%, which was lower than that reported in other countries. Although this positive case was 48 years old male with non-Hodgkin's lymphoma, there has no diagnosis index of ATL for this patient. It suggests that HTLV-1 infection is not the main cause of malignant hematological diseases in Central China 
. The HTLV-1 infection of this patient may be related to repeated injection blood product during the treatment of diseases. High molecular weight DNA was obtained from one HTLV-1 infected patient and was subjected to PCR to obtain the complete proviral genome (9034 bp) and the sequence analysis was performed by phylogenetic tree. Molecular studies on complete HTLV-1 genome based on phylogenetic analysis have demonstrated that two HTLV-1 isolates from China, one from this study and another from Fujian province, all belong to genotype A. We observed a close phylogenetic affinity between these two China strains and three strains from other region: one from Japan (L03561) and two from Canada (HQ606137 and HQ606138).
Like other sexual transmitted diseases, some high risk behaviors cause the spread of the virus, including sexual contact without protection with multiple partners, sharing injecting equipment, unsafe blood product transfusion, etc. Therefore, high risk behavior population could be the source of HTLV-1/2 spread in non-endemic regions, and cause co-infection with other sexually transmitted pathogens due to the same modes of transmission 
. The study of Giuliani et al. 
in Italy found the prevalence was 0.6% both for HTLV-1 and HTLV-2 among 1457 high risk population. Further analysis showed that T. pallidum
were associated with HTLV-1 infection; IDUs play an important role in HTLV-2 transmission; and the significant high rate of HTLV-1/2 infection has been found in HCV/HIV/HBV antibody positive individuals. In Argentina, Berini et al. 
surveyed the infection of HTLV-1/2, HIV, HBV, HCV and T. pallidum
in different high risk groups, the prevalence for HTLV-1/2 was 2.4%. The highest prevalence of HTLV-1/2 was found among IDUs (19.1%) and the lowest was MSM (0.4%). High rate (63%) of co-infection between HTLV-1 and HBV was observed. In this study, the participants of high risk group including IDUs, FSW, STIs, MSM, and PBDs. 2 HTLV-1 positive cases were found in FSW and 2 were from STIs. The prevalence rate was the highest in high risk population compared to those two groups of the blood donors and MHDs. The result was in agreement with Wang et al. 
report. Two HTLV-2 cases were IDUs, suggesting that parenteral transmission maybe important for the spread of HTLV-2. This finding is also consistent with other studies 
. In addition, of the 7 HTLV-1 cases, 6 were co-infected with HBV, 2 with HCV, and 1 with HIV. In the multivariate analysis, HBV and HIV remained independently associated with HTLV-1.
This study suffered from some limitations. First, due to the nature of cross-sectional study, the associations reported herein between the infections and the relate risk factors may not be related to a cause-effect manner. Second, the limitation of the low prevalence of the HTLV cases in Central China and the inadequate size of the study samples may have caused the lack of statistical significance of some associations, although in the presence of high ORs. Third, the convenience sampling may have led to bias and therefore either underrepresentation or overrepresentation of true prevalence. Last, this study was conducted only in provincial capital cities of Henan and Hubei. Thus, we were unable to evaluate any geographical region characters among any high risk factors across the country. However, we still believe that we could evaluate actual prevalence and associated risk factors in a reliable manner.
In summary, our study reveals the seroprevalence of HTLV-1/2 among different population samples in Central China, which shows a high prevalence rate among high risk group, and low rate among blood donors and MHDs. The infection of HTLV-1/2 has not been found in malignant hematological diseases. HIV, HBV, and HCV were the common risk factors associated with HTLV-1/2 infection, suggested that further study should be carried out in specific population, aiming to prevent the virus spreading. Our results can also provide theoretical basis for revise standard of voluntary blood donation in this area.