Secondary Logo

Journal Logo


HIV-1 prevalence, HIV-1 subtypes and risk factors among fishermen in the Gulf of Thailand and the Andaman Sea

Entz, Achara Thawatwiboonpola; Ruffolo, Vipan Prachuabmoha; Chinveschakitvanich, Vilaib; Soskolne, Vardac; van Griensven, G. J. P.d

Author Information
  • Free



In Thailand there are currently about 1 million people infected with HIV type 1 (HIV-1), mainly through heterosexual intercourse [1]. Smaller, but significant epidemics have been detected among intravenous drug users and male sex workers [2]. Sentinel surveillance data collected in Thailand's coastal provinces show elevated HIV-1 prevalence rates, particularly among female commercial sex workers (CSW) and fishermen. For instance, in Traat (Eastern Gulf of Thailand) 8.8% of fishermen tested positive in 1998, and this figure was 5.7% in Chumpon (Central Gulf), 24.7% in Songkhla and 15.5% in Pattani (both Southern Gulf) [2]. Thailand has a large fishing industry. The 1995 Marine Fishery Census shows 54 538 registered fishing boats, of which 53 451 mainly fish inside Thai waters (Gulf of Thailand 36 896 and Andaman Sea 16 555) [3]. The remaining 1087 boats fish in waters around Indonesia (437), Burma (342), Malaysia (136), Cambodia (124), Vietnam (35) or elsewhere in the region (13). In addition, the census counted some 161 700 registered fishermen of whom some 85 000 were employees. Among these employees approximately 35 000 were local, the others were migrants from either inside (37 000) or outside Thailand (13 000, mostly Burmese and Khmer). In reality these figures are likely to be higher, since there is considerable turnover and a high number of illegal immigrants among fishermen. Depending on their size, tonnage and destination, Thai fishing boats can go out for a considerable time, sometimes over 1 year. In the latter case, a fleet of fishing vessels is serviced by cargo refrigerator ships, commuting back and forth to the shore.

It has been suggested that mobile populations such as truck drivers and fishermen play a special role in the HIV-1 epidemic, particularly in developing countries. Segregation from the traditional family, loneliness and harsh and boring working conditions are thought to make these men prone to heavy alcohol drinking, drug use and CSW patronage, resulting in a high prevalence of sexually transmitted diseases (STD). Moreover, they are thought to connect sexual networks in different geographic locations, thereby accelerating the spread of HIV-1. Studies in Africa for instance have shown a significantly higher HIV-1 and STD prevalence among long distance truck drivers than among their non-travelling counterparts [4]. HIV-1 prevalence in communities located on commercial transport routes in Africa was also found to be higher than that in communities elsewhere in the same region [5]. A study among 327 long distance truckers in Thailand showed high levels of HIV-1 risk behaviour, but HIV-1 prevalence was low (2.7%) [6,7]. Ethnographic studies have documented a high prevalence of HIV-1 risk behaviours in Thai fishermen in Cambodia and Indonesia and several HIV-1-infected Thai fishermen have been identified by local authorities [8,9]. In Irian Jaya for instance, men working on Thai fishing fleets were documented to have private and commercial sex contacts with locals or sex workers were provided to them while at sea [9]. Although these indicators cause concern, no systematic information is available regarding the extent of the HIV-1 epidemic among fishermen in Thailand and about its behavioural determinants. Such information is necessary to develop data-driven prevention programmes for this group.

Several studies have been carried out to investigate the spread of different HIV-1 subtypes in south-east Asia in order to better understand the migrational patterns of HIV-1 in this part of the world. In Thailand, HIV-1 subtypes B′ and E have accounted for virtually all infections since the beginning of the epidemic in 1988. The initial epidemic in intravenous drug users (IDU) was largely due to subtype B′, whereas the larger heterosexual epidemic is primarily due to infection with subtype E [10]. In recent years subtype E has caused an increasing proportion of infections in IDU [11]. The HIV-1 epidemic in Myanmar shows more or less the same pattern as in Thailand, [12] whereas HIV-1 infections in Cambodia and Vietnam are overwhelmingly due to subtype E [13,14]. To date no such studies have been conducted in mobile populations such as fishermen. Therefore in this study we not only investigated the prevalence of HIV-1 and its social and behavioural determinants but also the prevalence of different HIV-1 subtypes and related factors.

Materials and methods

Study population and data collection procedures

A cross-sectional survey, consisting of face-to-face interviews and the collection of oral fluid samples, was conducted between January and April 1998 in four coastal provinces of Thailand: Traat (Eastern Gulf of Thailand), Samut Sakhon (Inner Gulf), Songkhla (Southern Gulf) and Ranong (West, Andaman Sea). Provinces were selected to cover all coastal areas of Thailand and their different fishing destinations as well as to reflect the different ethnic composition of the population of fishermen there (Khmer in the east, Thai in the central and Burmese in the south and west). Individuals who were eligible for enrolment were men working on commercial fishing vessels of gross tonnage ≥ 10 tonnes that go out to sea for at least 10 days (Samut Sakhon, Songkhla, Ranong) or at least 5 days (Traat, as it is has fishing grounds that are close to the area). These criteria were used to avoid the enrolment of subsistence fishermen, who go out to sea for short periods of time, reside in the area and are thought not to be at increased risk of HIV-1 infection. Since no systematic information was available on boats by tonnage, destination, departure and arrival, a priori sampling was not possible. Therefore, every morning in collaboration with harbour, fishery and fishing market officials a list was compiled of available boats meeting the enrolment criteria, and owners (or captains when the owner could not be immediately reached) were asked for permission to interview their crew. Finally, crew members were informed about the purpose and logistics of the study, were asked for verbal informed consent and if obtained an anonymous interview and an oral fluid sample were taken. Co- operation from fishery officials, owners and crew members was excellent and refusal was rare. In this study no HIV-1 oral fluid test results were provided to participants. If men wanted to know their HIV-1 status, they were referred to the local government STD clinic for free pre- and post-test counselling and testing. The results reported in this paper are confined to 818 fishermen. Data of another 51 men could not be analysed due to incomplete data or the absence of usable oral fluid samples.

Interview and questionnaire

Scripts for verbal informed consent and questionnaires were translated and back-translated from Thai to Khmer and Burmese and were administered by Thai-, Khmer- and Burmese- speaking interviewers when applicable. The questionnaire contained questions about socio-demographic and occupational characteristics, sexual and drug use behaviours and other risk factors for HIV-1 infection. Sexual behaviour variables included number of regular and casual female partners and number of commercial sex visits and condom use in the previous year and in the last 5 years. The number of regular and casual partners was combined to form a variable of number of private sexual partners (none, 1, ≥ 2). Number of CSW visits was recoded into three levels (none, 1–5, ≥ 6). Participants were asked whether they used condoms with each type of partner (consistently, sometimes or never), used alcohol or drugs before or during sex (always, sometimes, never) and whether they ever had sex with a man, ever used injectable and non-injectable drugs, had ever had an STD, and had been tattooed.

Laboratory analysis

Oral fluid specimens were collected using the `Orasure Salivary Collection Device' (Epitope Inc. Beaverton, Oregon, USA). After collection, specimens were refrigerated and shipped on wet ice to Bangkok for processing and laboratory testing. Oral fluid has shown an excellent performance for the detection of HIV-1 antibodies compared to serum, with a sensitivity of 100% and a specificity of 99.8%, as has been reported in several studies [15,16]. In our study, samples were tested for HIV-1 with IgG antibody capture enzyme-linked immunosorbent assay (GACELISA: Wellcozyme HIV1+2; Wellcome Diagnostics, Dartford, UK) and the Vironostika HIV-1 Microelisa System (Organon, Oss, The Netherlands) in two different laboratories. Examinations were performed according to the manufacturers' instructions and the agreement between the two tests and laboratories was 100%. Subsequently, positive samples were confirmed with a Western blot for oral fluids (Orasure, Organon Teknika Corporation, Durham, North Carolina, USA). A V3-loop peptide enzyme immunoassay (PEIA) was used for determination of HIV-1 subtypes B (B′ or Thai B) and E in oral fluid. Specimens dually or non-reactive were considered non-typable. In Thailand, these PEIA have been found to be nearly 100% specific and 90 % sensitive for subtypes B′ and E when compared with genetic subtyping. To date, Thai specimens found non-typable on PEIA have revealed subtype B′ or E sequences on genetic characterization [10,11].

Statistical analyses

To identify factors associated with prevalent HIV-1 infection and subtypes, data were analysed univariately using odds ratios and 95% confidence intervals (CI). To assess independent relations, stepwise forward multivariate logistic regression analysis was performed. Variables considered to be theoretically relevant or statistically significant in univariate analysis were entered in the logistic regression equation.


Demographic and behavioural characteristics

Of the 818 men from whom questionnaire data and oral fluid samples were obtained, 582 (71%) were Thai, 137 (17%) were Burmese and 99 (12%) were Khmer. Their average age was 29.5 years and they had an average duration of working as a fisherman of 8.9 years. Forty-four per cent of our subjects had between 1 and 4 years of education, 30% reported that they could not read or write and 61% were unmarried (Table 1). Regarding the position on the boat, 20% reported to be skipper or owner, 36% was steersman, mechanic or cook and 45% worked as a ship hand (Table 1). With respect to their fishing destinations, almost half of the participants (43%) said that they fished mostly off of coast of Thailand, 19% fished off the Burmese coast, 16% off Indonesia and 10% off Cambodia. Smaller percentages (7%) said that they fished off the coast of one or more other countries in the region, such as Bangladesh, India, Malaysia and Vietnam.

Table 1:
Demographic characteristics, HIV-1 prevalence, odds ratios and 95% confidence intervals (CI) among 818 fishermen in the Gulf of Thailand and the Andaman Sea, 1998.a

With regard to behavioural risk factors for HIV-1 infection, 46% of our respondents reported that they had visited one or more CSW in the past year and 76% of these men indicated that they had used condoms consistently during these visits. Twenty-three per cent of our subjects had more than one private sexual partner during the past year and 16% had visited a sex worker outside Thailand during the past 5 years. The prevalence of the latter differed significantly between Khmer (40%) and Thai and Burmese subjects (12%) [odds ratio (OR), 5.2; 95% confidence interval (CI), 3.19–8.46]. Forty per cent of the men in our study reported having a tattoo and 30% reported a history of STD. Alcohol or drug use before or during sexual intercourse was common (60%) but injection drug use and male-to-male sex were rare (Table 2).

Table 2:
Behavioural characteristics and HIV-1 prevalence, odds ratios and 95% confidence intervals (CI) among 818 fishermen in the Gulf of Thailand and the Andaman Sea, 1998.a

Univariate analysis of HIV-1 prevalence and subtypes

Of our study subjects 127 men (15.5%) were found to be HIV-1 positive, the prevalence being higher among Khmer (20.2%) than among Burmese (16.1%) and Thai (14.6%), but these differences were not statistically significant (Table 1). Being 25 to 32 years of age, compared with being younger or older; working as a fisherman between 4 and 10 years, compared with working as such for a shorter or longer period; being a steersman or mechanic, compared with being a skipper or ships hand; and being unmarried, all carried a significantly elevated risk for HIV-1 infection (Table 1). Of the behavioural factors, a greater number of CSW visits, having ever visited a CSW outside Thailand in the past 5 years, use of alcohol or drugs before or during sex, having a tattoo and a history of STD were significantly related to HIV-1 infection (Table 2). Having one private sexual partner appeared to be protective against HIV-1 infection (Table 2). Education, literacy, condom use with private and commercial partners, a history of drug injection and male-to-male sex behaviour were not significantly related to HIV-1 infection (Tables 1 and 2).

Of the 127 HIV-1 positive samples, 119 had sufficient material for HIV-1 subtype analysis. Among these samples, 72 (60%) were identified as subtype E and 15 (13%) as subtype B′. The subtype could not be determined in 32 samples, of which 28 (24%) were non-reactive and four (3%) were reactive to both B′ and E. Burmese and Khmer subjects were more likely to be infected with subtype E than were Thai subjects (100 versus 75% of the typeable samples, χ2 = 8;P  < 0.005, Fishers exact test). Subjects reporting a history of injection drug use were more likely to be infected with subtype B′ (OR, 9.4; 95% CI, 1.4–62.8;Table 3). Tattooing, nor any of the other factors assessed in our study were statistically related to the presence of subtype B′, E or non- and dual reactivity of the samples.

Table 3:
HIV-1 subtypes (E and B′) as determined by peptide enzyme immunoassay reactivity by nationality and injection drug use (IDU) history in 119 HIV-1 positive fishermen in the Gulf of Thailand and the Andaman Sea, 1998.

Multivariate analysis of HIV-1 prevalence

In multivariate analysis being 25 to 32 years of age, having made six or more visits to commercial sex workers in the previous year, being unmarried and having a tattoo were significantly and independently related to HIV-1 infection (Table 4). Other variables, such as duration of working as a fisherman, position on the boat and a history of STD dropped out of the multivariate model.

Table 4:
Multivariate logistic regression analysis of selected characteristics and HIV-1 prevalence among 818 fishermen in the Gulf of Thailand and the Andaman Sea, 1998.


Our data show a high HIV-1 prevalence (15.5%) among subjects enrolled in our study of fishermen in the Gulf of Thailand and the Andaman Sea. Except for direct `brothel-based' CSW and male injection drug users, this figure is higher than among any other sentinel surveillance group in Thailand, such as indirect CSW, military recruits and male STD patients [2]. The magnitude of the spread of HIV-1 in this group confirms the hypothesis that mobile populations are at higher risk for HIV-1 infection and illustrates their potential role as a `bridge' population for HIV-1 between different geographic locations.

The role of fishermen in the spread of HIV-1 in the south-east Asian region has been subject of much political debate and concern [8,9,17,18]. However, to date no systematic data regarding the spread of HIV-1 and related characteristics in this group were available. Our results may therefore help to formulate rational policies and data-based prevention programmes for fishermen. In this respect some limitations of our study need to be mentioned. Our sampling frame was designed to include men who were thought to be at the highest risk for HIV-1 infection, namely those who go out to sea for considerable periods of time. Therefore, the HIV-1 prevalence in the total population of fishermen or in other subcategories is likely to be lower. Secondly, we only enrolled boats owned by Thai firms operating from Thailand. Other countries in the region, such as Indonesia and the Philippines also have large fishing populations and information from these countries is necessary to put the HIV-1 prevalence among fishermen working in Thailand in perspective. Moreover, the high prevalence of HIV-1 among Khmer (20.2%) and Burmese (16.1%) subjects in our study reflects the considerable spread of HIV-1 in these countries and illustrates the regional character of the epidemic [19–21]. Hence, identifying Thai fishermen as the source of HIV-1 infection will not be productive as some other countries are equally if not more severely affected. With regards to the potential role of fishermen as a `bridge' population for HIV-1 infection, it is worth mentioning that 16% of our subjects reported having had commercial sexual contacts outside Thailand. Although the HIV-1 prevalence in this group was slightly elevated (21%), the direction of transmission is difficult to determine. Many of these contacts may in fact have been at home, as 29% of our study population was non-Thai. Moreover, Khmer subjects were more likely to report commercial sexual contacts outside Thailand (40%) than were Burmese and Thai (12%).

The majority of risk factors for HIV-1 infection found in our study, such as visits to CSW, alcohol or drug use and a history of STD, were not different from those found in many other studies [22,23]. The additional risk factors identified in our analysis such as age, duration of working as a fisherman and marital status should be merely viewed as markers of exposure history. The identification of tattooing as an independent risk factor for HIV-1 infection deserves some special attention. This behaviour has been mentioned frequently in relation to HIV-1 transmission, but to date no studies have been reported in which it was assessed as an independent risk factor. Tattooing in fishermen is often carried out by so called `professors' (Acharn Sak) who are thought to have knowledge about supernatural powers and whose drawings are believed to provide protection [9]. The tattooing is done with one or a number of needles tied together, which are dipped in an inedible liquid before application to the skin. Information provided by our study participants revealed that equipment is cleaned only superficially or not at all between subjects, which may increase the risk of HIV-1 transmission. More in-depth research is needed to determine what specific practices contribute to the risk of HIV-1 transmission during tattooing. In the meantime, prevention activities among fishermen should instruct tattoo artists and those tattooed to change equipment and liquid between clients.

Another potentially risky sub-cultural practice that became apparent during conversations with our subjects was penile manipulation. Subjects reported scarring and cutting of the foreskin, insertion of pearls (mook) and injection of oil under the skin of the shaft penis in order to increase its size and friction during sexual intercourse. Reportedly these practices were also performed by `professors' or `older brothers' (Pee Chai) on the boat or on shore. If the instruments that are used to perform these practices are not changed or sterilized between subjects, this may also constitute an additional HIV-1 risk.

Some other practices that have been suggested to contribute to the high HIV-1 risk among fishermen are injection drug use and same-sex behaviour. However these practices were uncommon in our study population and several men spontaneously declared that these practices were `unthinkable' on the boat as they might anger its `spirit' or `ruler'. In the Thai belief system, every space (boat, house, village, sea) has a `spirit' to whom regular homage is paid and offerings are made. Thai fishermen believe that same-sex behaviour or injection drug use on the boat will anger its spirit, who in turn may tear the net or sink the boat in retaliation.

The distribution of HIV-1 subtypes B′ and E in this study population is in line with several other studies into the molecular epidemiology of HIV-1 in Thailand [10,11]. In these studies subtype E was found to dominate among heterosexually infected persons, whereas subtype B′ was predominant in injection drug users. The exclusiveness of subtype E in Khmer and Burmese subjects reflects the predominance of subtype E in their regions of origin. A study among HIV-1 positive persons in Cambodia documented a 100% presence of subtype E, and a study in Burma showed the predominance of subtype E in Kawthaung province [12,13], which borders Ranong province where most of our Burmese subjects were enrolled.

In conclusion, our data show a high HIV-1 prevalence among fishermen in the Gulf of Thailand and the Andaman Sea. Risk factor analysis strongly suggests heterosexual intercourse as the major mode of transmission in this group. Increased efforts to reduce the spread of HIV-1 in this vulnerable and epidemiologically important population are urgently needed.


We are grateful for the helpful collaborations of Preeya Rungsopaskul, Jirapol Seangsung, Netdao Sawangjang of the College of Population Studies; and of Chanida Palanuwesh, Somchai Issaravanich, Vorawat Wongphanus, Ek-Kaphan Rittha of the Health Research Institute, Chulalongkorn University, Bangkok, Thailand; of Dr Jos van Oostrum, Department of Social and Organizational Psychology, University of Utrecht, Utrecht, The Netherlands; of Dr Candice Wong, School of Nursing, University of California at San Francisco, San Francisco, USA; of the Division of Epidemiology, Thai Ministry of Public Health; of the Department of Fishery, Thai Ministry of Agriculture and Cooperatives; of the Harbor Department, Thai Ministry of Transportation, Fishing Market Authorities, and of the World Vision Foundation of Thailand. We would like to express our gratitude to the owners, captains and crew of the fishing vessels for their support and dedication to the study. Confirmatory HIV-1 antibody testing and testing for HIV-1 subtypes was conducted by the laboratory of the HIV/AIDS Collaboration, Nonthaburi, Thailand. We would like to thank Ms Nancy Young and Drs Timothy D. Mastro and Peter H. Kilmarx for their helpful comments and support.


1. Surasiengsunk S, Kiranandana S, Wongboonsin K, Garnett GP, Anderson RM, van Griensven GJP. Demographic impact of the HIV epidemic in Thailand. AIDS 1998, 12: 775 –784.
2. Division of Epidemiology. HIV serosurveillance in Thailand: result of the 16th round, June 1998.Mthly Epidemiol Surv Rep (Thailand). 1999, 30 :1–11.
3. National Statistical Office. The 1995 Marine Fishery Census. Whole Kingdom. Bangkok: Office of the Prime Minister and Department of Fisheries, Ministry of Agriculture and Cooperatives; 1997.
4. Carswell JW, Lloyd G, Howells JW. Prevalence of HIV-1 in East African lorry drivers. AIDS 1989, 3: 759 –761.
5. Wawer MJ, Serwadda D, Musgrave SD. et al. Dynamics of spread of HIV-1 infection in a rural district of Uganda. BMJ 1991, 303: 303 –306.
6. Podhisita C, Wawer MJ, Pramualratana A, Kanungsukkasem U, McNamara R. Multiple sexual partners and condom use among long distance truck drivers in Thailand. AIDS Ed Prev 1996, 8: 490 –498.
7. Morris M, Podhisita C, Wawer MJ, Handcock MS. Bridge populations in the spread of HIV/AIDS in Thailand. AIDS 1996, 10: 1265 –1271.
8. Pramualratana A, Somrongthong R, Jindasak K, Saetiow, S. Assessment of the potential for spread and control of HIV among cross-border populations along the Thai-Cambodian border. Salaya: Institute for Population and Social Research, Mahidol University; 1995.
9. Komunbut R. Thai fishermen and their local contacts in Irian Jaya: an assessment of issues related to the spread of HIV/AIDS in Merauke. Jakarta: Program for Applied Technology in Health (PATH); 1995.
10. Subbarao S, Limpakarnjanarat K, Mastro TD. et al. HIV-1 in Thailand, 1994-1995: persistence of two subtypes with low genetic diversity. AIDS Res Hum Retroviruses 1998, 14: 319 –327.
11. Wasi C, Herring B, Raktham S. et al. Determination of HIV-1 subtypes in injecting drug users in Bangkok, Thailand, using peptide-binding enzyme immunoassay and heteroduples mobility assay: evidence of increasing infection with subtype E. AIDS 1995, 9: 843 –849.
12. Kusugawa S, Sato H, Watanabe S. et al. Genetic and serologic characterization of HIV type 1 prevailing in Myanmar (Burma). AIDS Res Hum Retroviruses 1999, 15: 91 –94.
13. Kusugawa S, Sato H, Kato K. et al. HIV type 1envsubtype E in Cambodia. AIDS Res Hum Retroviruses 1998, 14: 1379 –1385.
14. Menu E, Lien TTX, Lafon M. et al. HIV type 1 subtype E is predominant in South Vietnam. AIDS Res Hum Retroviruses 1996, 12: 629 –633.
15. Johnson AM, Parry JV, Best SJ, Smith AM, De Silva M, Mortimer PP. HIV surveillance by testing saliva. AIDS 1988, 2: 369 –371.
16. Frerichs RR, Eskes N, Htoon MT. Validity of three assays for HIV-1 antibodies in saliva. J Acquir Immune Defic Syndr 1994, 7: 522 –525.
17. Naing D, Dorning K. Community mobilization HIV/AIDS prevention and care in Ranong-Kawthaung sea border crossing between Thailand and Myanmar. In Report on the Second Technical Consultation on Transnational Population Movement and HIV/AIDS in Southeast Asian Countries. Bangkok: Asian Research Center for Migration, Chulalongkorn University; 1998. Report 015.
18. Wirakartakusumah MD, Yani A. Borderless migration: impact to HIV/AIDS infection and transmission. Indonesian case. In Report on the Second Technical Consultation on Transnational Population Movement and HIV/AIDS in Southeast Asian Countries. Bangkok: Asian Research Center for Migration, Chulalongkorn University; 1998. Report 015.
19. Ryan CA, Vathiny OV, Gorbach PM. et al. Explosive spread of HIV-1 and sexually transmitted diseases in Cambodia. Lancet 1998, 351: 1175 –1176.
20. Phalla T, Leng HB, Mills S. et al. HIV and STD epidemiology, risk behaviours and prevention and care response in Cambodia. AIDS 1998, 12 (suppl B): S11 –S18.
21. Htoon MT, Lwin HH, San KO, Zan E, Thwe M. HIV/AIDS in Myanmar. AIDS 1994, 8 (suppl 2): S105 –S109.
22. Mastro TD, de Vincenzi I. Probabilities of HIV-1 transmission. AIDS 1996, 10 (suppl A): S75 –S82.
23. Wasserheit JN. Epidemiological synergy: interrelationship between human immunodeficiency virus infection and other sexually transmitted diseases. Sex Transm Dis 1992, 19: 61 –77.

HIV-1; migrants; risk factors

© 2000 Lippincott Williams & Wilkins, Inc.