Article Text

Original article
Sexual behaviour of backpackers who visit Koh Tao and Koh Phangan, Thailand: a cross-sectional study
  1. C T Lewis,
  2. G de Wildt
  1. Institute of Clinical Sciences, University of Birmingham, Birmingham, UK
  1. Correspondence to C T Lewis, Institute of Clinical Sciences, College of Medical and Dental Sciences, University of Birmingham, Birmingham, UK; christopher.lewis11{at}


Objectives To measure the proportion of backpackers engaging in unprotected sex while travelling in Thailand and to identify predictors of unsafe sexual behaviour.

Methods A cross-sectional study using an anonymous questionnaire was performed. English-speaking backpackers, excluding US and Thai citizens, were recruited while travelling through the piers of Koh Tao and Koh Phangan between 25 January and 13 March 2013.

Results 61.5% (1238/2013) respondents were travelling without a long-term sexual partner, of whom 39.1% (29.4% females; 51.9% males) reported intercourse with a new partner. 36.8% of these reported inconsistent/no condom use. Inconsistent condom use was independently associated with age <25 years, UK (vs Canadian) nationality and not bringing condoms from home. Backpackers who brought condoms from home were more likely to have sex with a new partner. Male gender and longer trip length increased the likelihood of new partner acquisition and overall risk of unsafe sex but were not associated with inconsistent condom use. New partners were most commonly backpackers from other countries. Men were more likely than women to have sex with a local person.

Conclusions Backpackers visiting Thailand frequently report unsafe sex. Partner demographics demonstrate a mechanism of international sexually transmitted infection (STI) transfer. In an era of growing antibiotic resistance and continuing HIV transmission, targeting unsafe sex in backpackers has the potential to reduce STI incidence internationally. These data highlight the need for innovative public health intervention. To focus such measures, future research may elucidate why certain groups are at higher risk.


Statistics from


Unprotected sexual activity is an important global health issue resulting in unwanted pregnancy and spread of sexually transmitted infections (STIs) including HIV.1 While STIs are often asymptomatic, their complications such as infertility can result in considerable physical and emotional morbidity in addition to placing a substantial burden on healthcare services.2 It is well established that condom use reduces the risk of STI acquisition,3 ,4 yet incidence in countries including the UK remains high.5

Frequency and variety of STIs are closely linked with increasing population mobility.6 ,7 This is of particular concern due to rising global trend in antibiotic resistance.8 ,9 Backpackers are a younger, mobile population of travellers known to exhibit high risk-taking behaviour.10 This includes unsafe sexual practices: 40.9% of British backpackers having sex with a new partner while travelling in Australia do not use condoms consistently.11

Thailand is currently one of the most popular destinations for backpackers and attracts over 800 000 UK travellers each year.12 A 3.8% of UK born adults with a new HIV diagnosis made between 2002 and 2010 are likely to have contracted the virus in Thailand.13 This accounts for over a quarter of HIV infections identified during this period that were acquired abroad. While this finding may be associated with sex tourism, it highlights the importance of developing greater understanding of sexual behaviour abroad in order to improve health promotion strategies before and during travel.

The purpose of this study was to explore the sexual behaviour of backpackers visiting Thailand. It aimed to measure the proportion of backpackers engaging in unsafe sex while travelling and to identify associated risk factors.


The study employed a cross-sectional convenience sampling design, using an anonymous self-administered questionnaire. Questions were adapted from a tool previously used to study the sexual behaviour of backpackers in Australia.11 Data points included demographics (age, gender, country of permanent residence), current trip length and aspects of sexual behaviour during the current trip including number and gender of new partners, consistency of condom use with new partners and whether condoms had been brought from home. Our questionnaire was piloted in a convenience sample of 20 University of Birmingham students with experience of backpacking and subsequently revised prior to departure. Ethical approval was granted by the internal ethics committee at the University of Birmingham Medical School.

Data collection took place on two Thai islands. Koh Phangan is known for its monthly full moon party, drawing up to 40 000 revellers.14 Koh Tao is a small island (approximately 21 km2) adjacent to Koh Phangan and is extremely popular with backpackers and scuba divers, many of whom visit immediately before or after spending time on Koh Phangan. Backpackers often undertake circular routes around Southeast Asia, incorporating these islands in their itineraries. The islands therefore act as a bottle neck from which the backpacking population can be conveniently sampled. Data were collected on Koh Phangan between 25 January and 31 January 2013, and Koh Tao between 1 February and 13 March 2013. The researchers relocated to Koh Tao as boats leaving Koh Phangan had become considerably less busy following the full moon party on 26th January. Recruitment opportunities on Koh Tao remained more stable, and the researchers remained there for the duration of the study.

No sample size was determined prior to undertaking the study due to a lack of existing data on which to base a useful calculation. Instead, recruitment was maximised in the available time period. A number of locations were piloted on Koh Phangan, including Haad Rin beach during the daytime, Thongsala night market and Thongsala ferry port. Recruitment at the ferry port was identified as the most effective strategy and was therefore replicated on Koh Tao. Travel by boat is the only public means of entering and leaving the islands, and this recruitment method caused minimal inconvenience to participants who were already in a fixed location while waiting to leave. No distinction was made between recruitment locations during analysis as just 4.2% (n=85) participants were encountered in locations other than ferry ports.

Recruitment was undertaken during the hour before departure of morning and afternoon boats. Two researchers walked around the waiting area in a standard, predetermined direction and screened all travellers against inclusion criteria to minimise selection bias. Inclusion criteria were speaking English, and self-identification as a backpacker, defined as ‘an individual travelling to multiple locations on a limited budget, staying in low-cost accommodation and carrying their belongings in a personal load-carrying device’.10 Thai nationals were excluded as sexual behaviour may differ while travelling within a home country compared with when travelling abroad.15 US citizens were also excluded as University of Birmingham insurance specifically excluded their participation.

All eligible travellers were briefed about the study, given a participant information sheet, and invited to complete a paper questionnaire at the point of recruitment. Informed, verbal consent was obtained. Questionnaires were returned immediately to a large, opaque collection envelope to maintain anonymity. No incentive to participate was offered, although many pens were unreturned.

It proved impractical to keep an accurate record of exclusion and refusal rates while maintaining maximal participation due to the time pressures in sampling this transient population. Nonetheless, the researchers report that the number of backpackers excluded by citizenship or linguistic criteria was low and few eligible travellers refused to participate. Questionnaires were returned by all who consented to participate in the study, although one was left entirely incomplete and was therefore excluded. Data were manually inputted into an Excel spreadsheet and verified by the second researcher at point of entry. All questionnaires were photographed for archiving purposes and then destroyed by burning on site.

Statistical analysis

Backpackers travelling with a long-term partner or spouse were excluded from analysis: travelling with a long-term partner is known to be associated with lower risk of engaging in unprotected sex with new partners.11 ,16 Those travelling without a long-term partner or spouse were labelled as ‘single’ backpackers.

Single respondents were categorised into one of three groups: (1) no new sexual partner; (2) new sexual partner(s) with consistent condom use and (3) new sexual partner(s) with inconsistent condom use. For the purpose of analysis, unsafe sex was defined as penetrative vaginal or anal sex with a new partner, with inconsistent condom use.

Respondent ages were categorised in accordance with groupings used in the presentation of data by Public Health England,5 with our lower and upper age categories slightly redefined to reflect the minimum and maximum ages of the recruited population. When undertaking multivariate regression, age categories above 35 years were merged due to low numbers. The seven countries that contributed more than 70 participants underwent individual analysis. This threshold was chosen for reasons of respondent distribution and statistical power. These countries also represented seven of the nine highest overall non-Asian tourist nationalities entering Thailand in 2012 (the remaining two being the USA, whose citizens were excluded from this study, and Russia).12

Participants were asked whether new partners on this trip had been backpackers from the same country, backpackers from a different country or local people. Responses for each of these three partner demographic groups were coded in a binary format.

χ2 tests were undertaken to determine p values, and multinomial logistic regression was used to establish independent risk factors. All statistical analyses were performed using SPSS V.19 (IBMSPSS), with p<0.05 deemed significant.


A total of 2013 questionnaire responses were collected. Data from 61.5% of respondents (n=1238) were from single backpackers and were subject to analysis. Table 1 presents the baseline demographic characteristics of the study population. Around two-thirds were under 25 years old, over half were from Europe and two-fifths had been travelling for at least 1 month at point of recruitment.

Table 1

Background characteristics of single backpackers, stratified by gender

Table 2 shows the reported sexual behaviour of single backpackers. British and Swedish backpackers were the least consistent condom users: nearly half of subjects who engaged in new sexual partnerships reported inconsistent or no condom use. Men were more likely to report unsafe sex than women. However, gender was not associated with inconsistent condom use, which was reported by 36.4% males and 36.6% females who had sex with a new partner. Backpackers who did not bring condoms were more likely to report inconsistent condom use with new partners, although those who did travel with condoms were more than twice as likely to have sex with a new partner. Those who had been travelling for more than 1 month were three times more likely to have had sex with a new partner than those who had been away for <14 days. In multivariate analysis, trip length was not associated with consistency of condom use.

Table 2

Self-reported sexual behaviour of single backpackers

Eight male participants reported engaging in anal intercourse with a new male partner. Of those who had exclusively male partners, four out of six reported unsafe sex. Both men with new partners of both genders reported consistent condom use.

Multivariate logistic regression analysis revealed that age <25 years, living in the UK as opposed to Canada, and not bringing condoms from home were independent risk factors for inconsistent condom use with new partners. When considering risk of unsafe sex in all single backpackers, male gender was found to be an additional risk factor, while German residence was associated with lower risk of unsafe sex. This reflects the rates of reported new partner acquisition.

Table 3 shows reported frequency of partner choice. Participants were most likely to form new sexual partnerships with backpackers from other countries. Men were substantially more likely than women to have sex with local people.

Table 3

Reported new partner demographics, stratified by respondent gender


Many backpackers who visit Koh Tao and Koh Phangan report engaging in unsafe sex. Over a third of those travelling without a long-term partner or spouse had vaginal and/or anal intercourse with a new partner; a third of these did not use condoms consistently. High levels of disassortative mixing (ie, with partners from countries other than the participant's own) are evident, demonstrating a mechanism of international STI transmission.

This study highlights a number of risk factors for unprotected sex and corroborates previous related research.6 Younger age was associated with increased risk, in keeping with STI trends in the UK. The effect of gender on unprotected intercourse is explained by higher levels of reported abstinence in females as opposed to more consistent condom use with new partners: inconsistent condom use was not associated with gender. Similarly, travelling for longer increased the likelihood of a new sexual encounter but did not affect consistency of condom use with a new partner.

British and Swedish backpackers were the least consistent condom users, echoing STI trends in both countries which have among the highest reported rates of chlamydia in Europe—although this statistic is strongly influenced by testing artefact.17 Rates of gonorrhoea infection have risen in both countries over recent years.5 ,18 Previous research has found that almost half (43% in 2011) of Swedish men who are infected with gonorrhoea by women acquire the disease abroad, especially in Thailand.18 German backpackers were less likely than Britons to engage in unsafe sex due to lower levels of new partner acquisition. Canadians were most likely to report engaging in intercourse with a new partner, but also more likely to report consistent condom use than UK backpackers—even when adjusted for other factors. This novel observation warrants further investigation.

Not bringing condoms from home was independently associated with inconsistent condom use. Three-quarters of men compared with one-third of women travelling without a partner reported bringing condoms from home. Encouraging more single backpackers, particularly females, to carry condoms may have the potential to reduce incidence of unprotected sex. However, other barriers to condom use must also be considered: one-third of those who packed condoms reported inconsistent use with new partners. Anecdotal reports of locally acquired condoms breaking fell outside the scope of this study but may be another avenue for future investigation.

Both male and female participants reported that new sexual partners were most frequently backpackers from other countries, demonstrating a mechanism for the international transmission of STIs. This differs from previously published research, which has reported that new sexual contacts are most often from the traveller's country of origin15 ,19 or from the country of destination.20 These results suggest that despite demographic similarities, the sexual behaviour of backpacking populations is different from single-location holidaymakers; trends may be destination-specific. Men were markedly more likely than women to report having sex with a local person, which may reflect engagement with commercial sex workers. However, the issue of sex tourism was not specifically explored in this study.

It is well established in the literature that men who have sex with men (MSM), including those who travel, expose themselves to higher levels of sexual risk than the background population.6 ,21 This study did not recruit enough MSM to warrant separate analysis.


These results represent the behaviour of single backpackers visiting two Thai islands in early 2013; however, their generalisability is unclear. Different populations (eg, seasonal workers, British university students) are likely to travel at different times of year and may exhibit different risk-taking behaviours. Unsafe intercourse over an entire trip will be underreported as data were collected before participants had finished their travels. The absence of biological data means that exposure to risk cannot be correlated with incidence of STI.

A particular weakness of the questionnaire was that it did not specifically identify engagement with professional sex workers, which may represent a higher risk activity. Although the category ‘local people’ is likely to include some sex workers, this cannot be taken as a surrogate: it may also represent non-commercial partners and foreign expatriates.

Data collection at Suvarnabhumi, Bangkok's international airport, might give access to a more representative population of backpackers throughout Thailand at their point of departure. However, this would involve significant logistical challenges and may exclude backpackers travelling across land borders. One of the strengths of this study was its sampling method which involved using the only commercial points of entry and exit to the islands in question. This point-of-recruitment data collection strategy facilitated a high response rate compared with previous studies of returning travellers.22

Implications of this study

This study quantifies the proportion of backpackers visiting Koh Tao and Koh Phangan who engage in unsafe sex during their travels and confirms a mechanism of international STI transmission that has been previously suggested. Targeting unsafe sex in backpackers may have the potential to considerably reduce global STI incidence. The need for innovative public health intervention and international collaboration is clear.

Encouraging single backpackers to travel with condoms may increase consistency of use. Future qualitative research could offer valuable insight into why backpackers of certain nationalities appear to exhibit particularly high-risk or low-risk behaviours. This may help to inform and target strategies for promoting behavioural change. Studies might take place both in backpackers' countries of origin and in popular backpacking destinations. Given Thailand's history of a strong and successful HIV-control programme launched in 1989,23 Thai policy makers might be receptive to researching and piloting intervention.

Key messages

  • Many backpackers who visit Thailand have unprotected sex. There are marked differences in the self-reported behaviours of travellers from different countries.

  • Backpackers are most likely to engage in new sexual partnerships with backpackers from other countries, demonstrating a mechanism for international sexually transmitted infection transmission.

  • Male gender and longer trip length are associated with higher rates of new partner acquisition but are not independently associated with inconsistent condom use.

  • Bringing condoms from home is associated with increased consistency of use; future interventions may encourage pretravel preparation.


We thank Rahul Bhattacharjee for his invaluable assistance during data collection and input, and Dr Emma Hathorn, Professor Lesley Roberts and Professor Karen Hughes for their support. We are grateful to the pier staff in Thailand who kindly permitted data collection on their premises, to Goodtime Adventures Koh Tao for use of their facilities during data processing and to all the backpackers involved in the study for their time and participation. We acknowledge the University of Birmingham for subsidising the costs associated with this study.


Supplementary materials


  • Handling editor Jackie A Cassell

  • Contributors CTL and GdW conceived and designed the study. CTL implemented the study and analysed all data. CTL and GdW interpreted the results and prepared the manuscript. Both authors critically appraised and approved the final manuscript and are accountable for the integrity of the work.

  • Competing interests None declared.

  • Ethics approval Internal ethics committee, University of Birmingham Medical School.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Data from this study were presented at the IUSTI European Congress, Malta, 2014; interested parties may contact the corresponding author to view the poster.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles