The history of China's AIDS dates back to 1985 when an Argentinean tourist traveled to China from the United States was identified with HIV-1.1 This is followed by the first major outbreak that occurred in 1989 when 146 infected intravenous drug users (IDUs) in Yunnan province were found to be HIV-1 positive.1,2 Subsequently, HIV-1 disseminated along drug-trafficking routes to surrounding provinces Guangxi and Sichuan, to the northwestern province of Xinjiang and even to Taiwan.3–7 In the mid-1990s, another major outbreak took place among paid plasma donors (PBDs) in the east-central provinces in China where many farmers became infected through illegal blood collection practices.3,8 The infected IDUs and PBDs in turn infected their sexual partners and children. Within several years, these chains of transmission led to tens of thousands of infections, creating tragic “AIDS Villages” in rural areas.7–10 These 2 epidemics were eventually dominated by 2 distinct viral subtypes, namely C/CRF07_BC/CRF08_BC among IDUs and B' among PBDs, as we and others have previously described.3–5,8,11,12
By 1998, HIV-1 was not only prevalent among high-risk populations but had also spread to all 31 provinces of China.4,7,13,14 Since 2005, sexual transmission became the dominant risk factor for HIV-1 infection in China.7,13 Unexpectedly, a distinct viral subtype, CRF01_AE, had been spreading predominantly along sexual transmission route in the Yunnan province to the southwest of China in addition to other regions, indicating the dynamic nature of the HIV-1 epidemic in the general population.4,7 CRF01_AE, which was initially identified in Thailand,15 accounts for 84% of infections in South and Southeast Asia and more than 5% globally.4,11,16,17 The spread of this HIV-1 subtype and its emergence is of great concern particularly in Asia, which accounts for more than 50% of the world population. Thus, the increasing dissemination of CRE01_AE in China will potentially have a large influence on the global epidemiology of HIV-1.7,13,14,18
There have been few province-wide studies of the HIV-1 epidemic in China. Given that the population size of China is more than 1.3 billion, it is important to monitor viral spread and evolution especially in low-prevalence regions to prevent an explosion of HIV-1 incidence. In this study, we focused on the Fujian province because it is geographically distant from the most heavily affected regions (Fig. 1), and the epidemiology of HIV-1 had not been carefully studied there yet.7,19 Specifically, greater than 3 million specimens collected from Fujian in 2006–2009 were processed serologically to analyze and characterize the epidemiology of HIV-1 during this period. Furthermore, we performed genetic analysis of randomly selected HIV-1–positive specimens to determine the molecular epidemiology and drug resistance profiles of circulating strains in Fujian of China. Our findings provide greater insight in the current epidemiological status in Fujian, and thereby actuate the development of greater prevention, control, and clinical management of HIV-1 in China and surrounding regions.
PATIENTS AND METHODS
About 915,830 and 2,152,658 specimens collected in 2006–2007 and 2008–2009, respectively, were analyzed for HIV-1 infections. These samples were collected from 220 HIV-1 testing sites located in Fujian's 9 prefectures (Fig. 1). Among all samples collected in 2008, 113 of them were randomly selected for genetic analysis at the Fuzhou Center for Disease Control and Prevention. Subjects who were confirmed HIV-1 positive before 2008 were excluded from the genetic analysis. This study was conducted in a cross-sectional and anonymous way in line with guidelines set by the local ethical review committee.
Serum samples were screened for HIV-1 infection at the local testing sites using commercial enzyme-linked immunosorbent assay kits (LIVZON Diagnostics, Zhuhai, China). Samples showing positive enzyme-linked immunosorbent assay results for being HIV-1 positive were then confirmed by a commercial Western blot assay (Singapore MP. Biomedical Asia Pacific Ltd, Singapore) at the Fuzhou Center for Disease Control and Prevention. A positive case was defined as a subject confirmed to be HIV-1 positive by this screening process.
Amplification of HIV-1 pol and gag Genes
Viral RNA was extracted from each patient's serum using the QIAamp Viral RNA Mini kit (Qiagen, Valencia, CA). The viral RNA was then subjected to reverse transcription–polymerase chain reaction. For gene amplification, polymerase chain reaction primers and conditions have been described previously.4,8
Phylogenetic Analysis of HIV-1 Sequences
HIV-1 sequences were aligned together with reference sequences obtained from the HIV databases (http://hiv-web.lanl.gov) using the Clustalx 1.81 program.20 We also included some reference sequences which were previously characterized in China and in Southeast Asian countries.4,21 The genetic distances between HIV-1 sequences were calculated using Kimura's 2-parameter model.22 Phylogenetic trees were generated using the neighbor-joining method and then plotted using the TreeView software. Bootscanning analysis was performed using the SimPlot 3.5 software kindly provided by Dr S. Ray.23
Analysis of HIV-1 Drug-Resistant Mutations
The drug resistance analysis was conducted using tools from the Stanford HIV Drug Resistance Database (http://hivdb.stanford.edu).24 Major drug resistance mutations were defined as previously described.25,26
Pearson χ2 test and Fisher exact test were performed to investigate the differences of demographic data. All statistical analyses were performed using SPSS version 9.1.3. A P-value <0.05 was considered statistically significant, and Bonferroni adjustments were made when conducting multiple comparison.
HIV-1/AIDS in Fujian by the End of 2009
Two thousand five hundred fourteen HIV-1 infections were identified in Fujian by the end of 2009 since the first identified case in 1987, including 983 reported cases of AIDS and 500 deaths (Fig. 1). The number of HIV-1 infections in Fujian increased significantly in recent years from 528 in 2006–2007 to 1129 in 2008–2009 (Fig. 1). These 1129 cases accounted for 44.5% of the total 2514 cases (Fig. 1 and Table 1), indicating a rising epidemic in Fujian. Besides the effect of increased surveillance in recent years, the overall prevalence of infections increased from 0.064% in 2006–2007 to 0.074% in 2008–2009 (Table 2, P = 0.003). HIV-1 cases were found in all 9 geographical prefectures of Fujian (Fig. 1). Interestingly, more than 1400 HIV-1 infections were reported between 1987 and 2009 along the coastal prefectures of Fujian including Fuzhou, Quanzhou, and Xiamen,19,27 corresponding to >55% of the total number of cases in Fujian.
We further compared sociodemographic characteristics of the HIV-1 cases in 2006–2007 with those identified in 2008–2009 (Table 1), and the following trends between 2006–2007 and 2008–2009 were statistically significant as follows: (1) the proportion of cases ≥40-year old increased from 27% to 33% (P = 0.017) accompanied by a decrease among cases aged 30–39 from 36% to 32% (P = 0.024); (2) geographically, there were more infections found in the northeastern Ningde prefecture, which borders Fuzhou (P = 0.009); (3) the proportion of cases with heterosexual transmission increased from 51.6% to 66% (P < 0.001); (4) those with homosexual transmission increased from 1.9% to 5.3% (P = 0.002); and (5) the proportion of cases with intravenous drug use as a risk factor decreased from 11% to 6.7% (P = 0.004). In contrast, the distribution of gender and occupation did not show any statistically significant difference between the 2 periods by any single category.
We then investigated the prevalent rates among various source of infections. Table 2 showed the screening results of specimens collected from each of these sources during 2006–2007 and 2008–2009. For stringent comparisons, our statistical analysis employed the Bonferroni correction. The following high-risk groups showed a statistically significant decrease in the prevalence rate between 2006–2007 and 2008–2009 (Table 2): (1) spouses or partners of HIV-1–positive cases (from 52% to 18%; P < 0.001); (2) individuals who underwent screening in entertainment centers such as karaoke television (from 0.21% to 0.05%; P < 0.001); and (3) prisoners who were screened during health checks (from 0.18% to 0.1%; P < 0.001). In contrast, the following low-risk groups showed a significant increase in prevalence rate between 2006–2007 and 2008–2009: (1) voluntary blood donors (from 0.005% to 0.013%; P < 0.0012); and (2) recipients of blood transfusion (from 0% to 0.04%; P < 0.001). The prevalence rate of unknown sources also increased greatly from 0.06% to 0.54%. These unknown cases are unlikely to be spouses or partners of cases, drug rehabs, or prisoners because their identities would have been readily transparent. Rather, these cases are likely to be from other risk groups (such as clinical referrals and clients of voluntary testing and counseling and sexually transmitted disease clinics) and from the general populations (such as premarital couples and presurgery patients) who did not wish to release their personal information and background. Furthermore, due to the rapid increase in infections between 2008 and 2009 (Fig. 1), we performed a similar analysis for these 2 years. Among high-risk groups, screening of IDUs in drug rehabilitation centers showed a statistically significant decrease in the prevalence rate between 2008 and 2009 (Table 2, from 0.29% to 0.09%; P = 0.003). Importantly, the low-risk group of screening test before surgery showed an increase in prevalence rate between 2008 and 2009 (from 0.03% to 0.07%; P < 0.001). In addition, as shown in Table 2, analyses before the Bonferroni correction showed that following high-risk groups showed a statistically significant decrease in prevalence rate in voluntary testing and counseling between 2006–2007 and 2008–2009 (from 0.62% to 0.45%, P = 0.005) and clinical referrals between 2008 and 2009 (from 0.18% to 0.12%, P = 0.006). Meanwhile, there were significant increases in prevalence rates for sexually transmitted disease clinics (from 0.17 to 0.19, P = 0.005) between 2006–2007 and 2008–2009 and for voluntary blood donors (from 0.009% to 0.017%; P = 0.031) and pregnant women (from 0.014% to 0.025%; P = 0.019) between 2008 and 2009. In summary, our findings indicated that the rapidly rising HIV-1 epidemic occurred mainly among general populations in Fujian.
Identification of Multiple HIV-1 Subtypes and Unique Recombinant Forms
Previous studies suggested that the HIV-1 epidemic in Fujian was caused by CRF01_AE viruses.7,19 To understand the evolving HIV-1 epidemic in Fujian, we obtained 86 HIV-1 pol gene sequences from the sera of 113 cases newly identified in 2008. Of the 27 subjects excluded, 20 were diagnosed HIV-1 positive before 2008 and 7 did not give positive amplification primarily due to either low viral load or insufficient quantity of specimens. Although the sociodemographics of these 86 subjects may not necessarily represent all the cases identified in 2008–2009 (see Table 1, Supplemental Digital Content 1, https://links.lww.com/QAI/A290), they account for 18.2% (86 of 472) of reported cases identified in 2008 from 7 prefectures in Fujian, with the exception of the low prevalent prefectures Putian and Zhangzhou (Fig. 1). Among the 86 subjects examined, 3 major HIV-1 subtypes were identified as follows: CRF01_AE (64 of 86; 74.5%), C/CRF07_BC/CRF08_BC (7 of 86; 8.1%), and B/B' (15 of 86; 17.4%) (Fig. 2A; see Table 1, Supplemental Digital Content 1, https://links.lww.com/QAI/A290). CRF01_AE was the major subtype across all prefectures tested except Sanming from which only 1 specimen was available (Fig. 1). The other 2 subtypes were found in 3 prefectures (Fig. 2A), respectively. Although the genetic distance of subtype C/CRF07_BC/CRF08_BC viruses (mean 0.0588 ± SE 0.0082) were seemingly larger than those of subtypes B/B' (0.0526 ± 0.0075) and CRF01_AE (0.0493 ± 0.0062), the differences were not statistically significant. Virus strains previously found in Yunnan province, Thailand, and other Asian regions were distributed in various locations among the local strains in our phylogenetic tree (Fig. 2). This suggested that Fujian experienced multiple introductions of all 3 subtypes, and the viral coevolution was around the same period of time. This notion was consistent to the finding that there was no statistical significance for sociodemographic characteristics of individuals infected with any of the 3 major HIV-1 subtypes (see Supplemental Digital Content 1, https://links.lww.com/QAI/A290).
Novel Recombinant or Drug-Resistant Variants of HIV-1
To further investigate HIV-1 evolution and diversity in Fujian, we successfully extracted HIV-1 gag p17 sequences from 57 available specimens of the 86 subjects to perform phylogenetic analysis. The 3 major subtypes were again CRF01_AE (43 of 57; 75.4%), C/CRF07_BC/CRF08_BC (5 of 57; 8.8%), and B/B' (9 of 57; 15.8%) (Fig. 2B). Interestingly, 2 of these patients who now live in Fujian harbored unique virus genotypes with intersubtype recombinant pol and p17 gene regions. One such patient of Sichuan origin (Sc.43H081029) carried a recombinant p17 B'/RT CRF01_AE (a variant of CRF01_AE), whereas the other patient of Xinjiang origin (Xj.87H082047) carried a recombinant p17 CRF01_AE/RT CRF07_BC (a variant of CRF07_BC). These 2 patients most likely were infected before moving to Fujian because similar recombinants had not been found among local habitants in Fujian. Furthermore, as shown in Supplemental Digital Content 2 (see Figure 1, https://links.lww.com/QAI/A291), an additional 5 new recombinant viruses were identified by Bootscan analysis of the HIV-1 pol genes, these include the following: 3 different B'/CRF01_AE recombinant viruses: Cq.67H081498 (a variant of B'), Hen.53BH081312 (a variant of B'), and Fj.Xm.69H081500 (a variant of CRF01_AE); a CRF01_AE/CRF08_BC recombinant virus: Fj.Fz.46H081045 (a variant of CRF01_AE); and a C/CRF01_AE recombinant virus: Fj.Fz.112BH081488 (a variant of C). Because all 7 unique recombinant forms (URF) were found on an individual basis, they were not yet considered to be circulating recombinant forms. Taking these URFs into account resulted in the following adjusted proportions of viral subtypes: 70.9% CRF01_AE (61 of 86), 5.8% C/CRF07_BC/CRF08_BC (5 of 86), 15.1% B/B' (13 of 86), and 8.1% URFs (7 of 86). This distribution resulted in a new finding that approximately 29% (25 of 86) of HIV-1 strains in Fujian were not pure CRF01_AE strains.
Next, we investigated the HIV-1 drug resistance profiles of the 86 subjects. Based on surveillance records, all 86 subjects were considered to be treatment naive. As shown in Supplemental Digital Content 3 (see Table 2, https://links.lww.com/QAI/A292), 26 subjects (26 of 86; 30.2%) had drug-resistant nucleoside reverse transcriptase inhibitor, nonnucleoside reverse transcriptase inhibitor, and protease inhibitor mutations. Most of these subjects (19 of 26; 73.1%) contained only 1 single drug-resistant mutation. Seven subjects, however, had more than 1 resistant mutation.
We report here the first province-wide HIV-1 epidemiological and molecular epidemiological study ever conducted in Fujian, a province on the southeastern coast of China with 2514 confirmed infections by the end of 2009. Our surveillance effort has revealed a significant rise of the overall prevalent rate of infections by 15.6% within a short time period, from 0.064% in 2006–2007 to 0.074% in 2008–2009 (Table 2). This rise accounted for the doubled number of infections from 528 in 2006–2007 to 1129 in 2008–2009 (Fig. 1). During these periods, the prevalence rate among some high-risk groups had dropped substantially by more than 40%, such as spouses or partners of positive cases, prisoners and individuals screened at entertainment places (Table 2). This might be attributed by the success of an increased uptake of interventions against HIV-1 transmission [such as condoms, antiretroviral therapy (ART)] among high-risk individuals. In contrast, the prevalence rates among low-risk groups have increased significantly in 2008–2009 such as voluntary blood donors and recipients of blood transfusions (Table 2). The prevalence rate among unknown sources, which were likely to be from the general population, has also increased substantially. Moreover, 3 low-risk groups showed a further increase in prevalence rate between 2008 and 2009 including screening test before surgery, voluntary blood donors, and pregnant women (Table 2). Taken together, these observations suggested that the rapidly rising HIV-1 epidemic has occurred mainly among general populations in Fujian.
Unlike HIV-1 early epidemics in heavily affected regions of China (ie, IDUs in Yunnan and PBDs in Henan),4,7 the HIV-1 epidemic in Fujian has not been confined to a particular high-risk group as a founder population. Hence, unprotected sexual contact was found to be the major mode of HIV-1 transmission in Fujian before 2006.19 The proportion of sexually transmitted cases, however, has been evolving. We found that the percentage of sexually transmitted cases was 53.4% in 2006–2007 and increased to 71.3% in 2008–2009 (Table 1), which included both heterosexual (increasing from 51.5% to 66%, P < 0.001) and homosexual transmissions (from 1.9% to 5.3%, P = 0.002). These results suggested that a high frequency of HIV-1 infections was consistently related to unprotected sexual transmissions compared to other risk groups such as IDUs in recent years. HIV-1 incidence among various risk groups in Fujian, however, remains to be determined by future cohort and BED studies.7 Critically, recent new findings of a dramatic increase in infections among homosexual population have posted a new challenge for HIV-1 prevention in China.28 Given the global trend of the rapidly escalating infections among MSMs, our finding has set off an alarm that the targeted control of HIV-1 spread among MSM in Fujian should be one of the preventive priorities. Moreover, we consistently found that more than 21% of infections in 2006–2007 and 2008–2009 were unemployed (Table 1). This phenomenon has not been found in previous studies conducted in Fujian7,19 or elsewhere in China.3–7 Our observation raises the possibility that unemployment has become a critical issue for HIV-1 infections among the general populations. As shown in Figure 1, the prefectures with the highest HIV-1 burden were along the coastal areas including Fuzhou, Quanzhou, and Xiamen, where more than 55% of all Fujian's cases were identified. We now report that the HIV-1 epidemic has geographically expanded significantly from Fuzhou northward to Ningde in recent years (Table 1). In fact, these prefectures are the pillars of Fujian's economy. In this regards, they were affected heavily by the recent global economic crisis around 2007 leading to an increase in unemployment. Previous reports indicated that unemployment debilitated the lives of HIV-infected people,29 and unemployed people were more vulnerable to HIV-1 infections.30 Because it remains unknown whether or not unemployed people were more vulnerable to infections in Fujian or HIV-1 infections resulted in increased unemployment there, our findings warrant future investigation of the impact of the economic crisis on the HIV-1 epidemic.
Our phylogenetic analysis of HIV-1 sequences of 86 subjects in 2008 suggested that CRF01_AE was the dominant circulating subtype in Fujian (64 of 86; 74.4%; see Table 1, Supplemental Digital Content 1, https://links.lww.com/QAI/A290). To some extent, the spread of CRF01_AE throughout the province suggested either an earlier or a more efficient dissemination of this viral subtype via sexual transmission among general populations. As for the origin of CRF01_AE in Fujian, our data suggested a hypothesis of multiple introductions of diverse viral variants into the study region (Fig. 2). One possible introduction was that some ship workers became infected in Thailand and had brought the virus back to Fujian. Our findings together with others have provided strong evidence that CRF01_AE has resulted in escalating epidemics in many regions of China and surrounding Asian countries mainly through sexual transmission.4,7,14,16 Thus, controlling the spread of CRF01_AE among the general population must become a top priority. Because a large proportion of cases were unemployed (Table 1), preventive strategies targeted at this group will likely be an effective start in Fujian.
Our phylogenetic analysis also suggested that the remaining 29% of circulating strains consisted of a diverse group of C/CRF07_BC/CRF08_BC (5 of 86; 5.8%), B/B' (13 of 86; 15.1%), and unique recombinant forms (7 of 86; 8.1%). Our finding of parallel evolution of multiple subtypes in Fujian is new, which is in contrast to previous reports that only CRF01_AE was found there.7,19 Because our analysis on viral genetic distance of C/CRF07_BC/CRF08_BC, B/B' and CRF01_AE did not show statistical significance, we hypothesized that Fujian was subjected to multiple introductions of all 3 subtypes. Clearly, these newly identified HIV-1 variants within each viral genotype indicated a new trend of the rapidly evolving HIV-1 diversity in Fujian. The finding of drug-resistant variants posts another risk of HIV-1 infections in Fujian. As shown in Supplemental Digital Content 3 (see Table 2, https://links.lww.com/QAI/A293), 26 subjects (26 of 86; 30.2%) had drug-resistant mutations. Because most of them (19 of 26; 73.1%) contained only 1 single drug-resistant mutation, the emergence of drug resistant mutations is likely at an early stage. Seven Fujian subjects, however, were found harboring more than 1 mutation. If they were truly ART naive as indicated by our study records, it is possible that some drug resistant variants have been transmitted locally since the introduction of antiretroviral therapy in 2005. Because ART has recently been suggested for use as a preventive means, the implementation of this strategy must consider the issues of ongoing spread of drug-resistant HIV-1 variants in China. Hence, building up a high-throughput, comprehensive, and effective drug-resistant surveillance system has become a critical issue.
In conclusion, the increased HIV-1 infections among general populations have posted new challenges to China's AIDS prevention. Our findings are useful to generate a comprehensive prevention program to conduct evidence-based biomedical, behavioral, and structural interventions. Future enhanced HIV-1 surveillance should be improved with a better data collection system in combination with promising and innovative practices to determine the effectiveness of reporting and prevention. Because China has the largest population in the world, we believe that nation-wide control of sexual transmissions of HIV-1 will contribute greatly to the world in the fight against the growing HIV/AIDS pandemic.7,30
The authors are indebted to all study participants. The authors thank all health workers at Fujian 220 HIV-1 testing sites for initial screening tests. The authors thank China governmental agency for financial supports to Fujian and Fuzhou Center for Disease Control and Preventions.
1. Settle E. AIDS in China: An Annotated Chronology (1985–2003). Montreal, CA: China AIDS Survey, 2003.
2. Zheng X, Tian C, Choi KH, et al.. Injecting drug use and HIV infection in southwest China. AIDS. 1994;8:1141–1147.
3. Zhang L, Chen Z, Cao Y, et al.. Molecular characterization of human immunodeficiency virus type 1 and hepatitis C virus in paid blood donors and injection drug users in china. J Virol. 2004;78:13591–13599.
4. Zhang Y, Lu L, Ba L, et al.. Dominance of HIV-1 subtype CRF01_AE in sexually acquired cases leads to a new epidemic in Yunnan province of China. PLoS Med. 2006;3:e443.
5. Su L, Graf M, Zhang Y, et al.. Characterization of a virtually full-length human immunodeficiency virus type 1 genome of a prevalent intersubtype (C/B') recombinant strain in China. J Virol. 2000;74:11367–11376.
6. Beyrer C, Razak MH, Lisam K, et al.. Overland heroin trafficking routes and HIV-1 spread in south and south-east Asia. AIDS. 2000;14:75–83.
7. Lu L, Jia M, Ma Y, et al.. The changing face of HIV in China. Nature. 2008;455:609–611.
8. Su B, Liu L, Wang F, et al.. HIV-1 subtype B' dictates the AIDS epidemic among paid blood donors in the Henan and Hubei provinces of China. AIDS. 2003;17:2515–2520.
9. Zhuang K, Gui X, Su B, et al.. High prevalence of HIV infection among women and their children in Henan Province, China. J Acquir Immune Defic Syndr. 2003;33:649–650.
10. Zhang W, Hu D, Xi Y, et al.. Spread of HIV in one village in central China with a high prevalence rate of blood-borne AIDS. Int J Infect Dis. 2006;10:475–480.
11. Yu XF, Chen J, Shao Y, et al.. Two subtypes of HIV-1 among injection-drug users in southern China. Lancet. 1998;351:1250.
12. Luo CC, Tian C, Hu DJ, et al.. HIV-1 subtype C in China. Lancet. 1995;345:1051–1052.
13. Wu Z, Sullivan SG, Wang Y, et al.. Evolution of China's response to HIV/AIDS. Lancet. 2007;369:679–690.
14. Shao Y. AIDS epidemic at age 25 and control efforts in China. Retrovirology. 2006;3:87.
15. McCutchan F, Hegerich P, Brennan T, et al.. Genetic variants of HIV-1 in Thailand. AIDS Res Hum Retroviruses. 1992;8:1887–1895.
16. Lau KA, Wang B, Saksena NK. Emerging trends of HIV epidemiology in Asia. AIDS Rev. 2007;9:218–229.
17. Piyasirisilp S, McCutchan FE, Carr JK, et al.. A recent outbreak of human immunodeficiency virus type 1 infection in southern China was initiated by two highly homogeneous, geographically separated strains, circulating recombinant form AE and a novel BC recombinant. J Virol. 2000;74:11286–11295.
18. Beyrer C. Hidden epidemic of sexually transmitted diseases in China: crisis and opportunity. JAMA. 2003;289:1303–1305.
19. Hai-Long H, Jian Z, Ping-Ping Y, et al.. Genetic characterization of CRF01_AE full-length human immunodeficiency virus type 1 sequences from Fujian, China. AIDS Res Hum Retroviruses. 2007;23:569–574.
20. Page RD. TreeView: an application to display phylogenetic trees on personal computers. Comput Appl Biosci. 1996;12:357–358.
21. Kato K, Kusagawa S, Motomura K, et al.. Closely related HIV-1 CRF01_AE variant among injecting drug users in northern Vietnam: evidence of HIV spread across the Vietnam-China border. AIDS Res Hum Retroviruses. 2001;17:113–123.
22. Kimura M. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J Mol Evol. 1980;16:111–120.
23. Lole KS, Bollinger RC, Paranjape RS, et al.. Full-length human immunodeficiency virus type 1 genomes from subtype C-infected seroconverters in India, with evidence of intersubtype recombination. J Virol. 1999;73:152–160.
24. Kuiken C, Korber B, Shafer RW. HIV sequence databases. AIDS Rev. 2003;5:52–61.
25. Shafer RW, Rhee SY, Pillay D, et al.. HIV-1 protease and reverse transcriptase mutations for drug resistance surveillance. AIDS. 2007;21:215–223.
26. Luo M, Liu H, Zhuang K, et al.. Prevalence of drug-resistant HIV-1 in rural areas of Hubei province in the People's Republic of China. J Acquir Immune Defic Syndr. 2009;50:1–8.
27. Liu J, Yue J, Wu S, et al.. Polymorphisms and drug resistance analysis of HIV-1 CRF01_AE strains circulating in Fujian Province, China. Arch Virol. 2007;152:1799–1805.
28. Li HM, Peng RR, Li J, et al.. HIV incidence among men who have sex with men in China: a meta-analysis of published studies. PLoS One. 2011;6:e23431.
29. Delpierre C, Cuzin L, Lauwers-Cances V, et al.. Unemployment as a risk factor for AIDS and death for HIV-infected patients in the era of highly active antiretroviral therapy. Sex Transm Infect. 2008;84:183–186.
30. Tucker JD, Henderson GE, Wang TF, et al.. Surplus men, sex work, and the spread of HIV in China. AIDS. 2005;19:539–547.