Male sexuality and HIV: The case of male-to- male sex:

Male sexuality and HIV:
The case of male-to-
male sex:
By Carol Jenkins, PhD
i

---

Information note
A background paper produced for Risks and Responsibilities: Male Sexual Health and HIV in Asia and
the Pacifi c International Consultation held in New Delhi, India 23-26 September 2006.
This publication was commissioned by Naz Foundation International with the support of UNAIDS. The
author’s views expressed in this publication do not necessarily refl ect the views of UNAIDS.
ii

---

Contents
Page
1
Culture, biology and the science of sexuality
5
Cultural pattern of male-to-male sex in Asia and the pacifi c
11
Vulnerability and HIV risk
15
Public health and enabling environments
17
References
iii

---

iv

---

Culture, biology and the science of sexuality
Even though sex ranks with food, air, and water as a primary basic human need, the science of human sexu-
ality has only fi tfully progressed during the past 70 years. Biologists have documented the vast diversity in
sexual adaptations among many hundreds of species, ranging from fi sh and birds to all kinds of mammals,
but to date, human sexuality remains inadequately documented and unexplained. Efforts to examine the
biological roots of sexuality in humans always seem to reach a point at which genetic, hormonal or other
chemical factors cannot explain it all. While it is clear that there are biological components to gender and
sexual behaviours in humans, these are linked to psychological and social phenomena in ways that are not
well understood.
If we look for hints in the sexual behaviour of our closest relatives in the animal kingdom, the bonobos
and chimpanzees, who share 98 percent of their genome with us, we fi nd a wide variety of adaptations,
apparently linked to different environments, reproductive and feeding strategies. Among the bonobos or
pygmy chimpanzees, female-female sex and male-male sex are common and appear to serve to reduce ag-
gression and competition, particularly over food, to form bonds between pairs of animals that are useful
for future mutual protection, to release tensions after fi ghting and appear to contribute to overall social
cohesion. These same animals also engage in male-female sex and reproduce at the same rates as other
chimpanzee species. Like humans, hierarchy and dominance play a major role in ordering relationships.
The social organization of human sexuality shows many similar features and same sex mating may serve
some of the same functions as well as several distinctly human ones.
While extrapolating motivations or social functions from non-human primates to humans remains prob-
lematic, it is clear that the fl exibility and diversity seen in human sexuality are important adaptations for
the species, rooted in our evolutionary past. In the 1940s Alfred Kinsey demonstrated clearly that human
sexual behaviour existed along a continuum, like most natural phenomena, and that there were no dichoto-
mous categories, such as heterosexual and homosexual (Kinsey et al., 1948). The term homosexual was
coined in 1869 and the term heterosexual was brought in a bout a decade later. These terms were adopted
by European medical and legal professions in the late 1800s for the purposes of defi ning pathology and
criminality. More recent sexuality research indicates that such divisions are too simplistic to refl ect reality.
To date, most epidemiologists who deal with HIV and STIs have not accommodated the existing social-
behavioural realities in their labels and it remains for social scientists to correct this, in order to design
better and more effective HIV prevention programmes. By improving our understanding of human sexual
relations and selfhoods within the socio-cultural frameworks that shape them, it should be possible to de-
sign HIV prevention activities that address the most salient issues associated with risk-taking behaviours
and build a sustainable approach to male sexual health.
There is copious evidence that stigmatization and marginalisation as social processes contribute greatly
to HIV vulnerability. In the case of male-to-male sex, this marginalisation, even unto denial, has played a
serious role in misdirecting attention and resources for HIV prevention and care. The larger question for
us today is why humans are the only animals in which condemnation and oppression are common social
responses to diversity in gender and sexuality.
Defi ning MSM
The recently coined term ‘MSM’ standing for ‘men who have sex with men’, was introduced into the world
of HIV/AIDS discourse in the late 1980s to designate men practising certain behaviours-without attaching
to them any particular sexual identity. It was adopted in order to avoid stigmatizing labels and to provide
a descriptive category that might improve epidemiological assessments. While it was meant to focus on
behaviours, it has often been used to defi ne persons. In several countries, terms such as ‘homosexual’ and
‘gay’ had already been seen to be inadequate as many males who did have sex with other males did not
identify with those terms. When the homosexual/heterosexual (and later bisexual) terms were incorporated
into medical discourse, men or women with same-sex desires (and even anyone who masturbated) were con-
sidered as pathological, but a century of attempts by psychiatrists to treat homosexuality has demonstrated
that such desires are not necessarily associated with poor mental or emotional functioning. Coupled with
a changing political landscape, starting in 1973 the psychiatric profession has eliminated homosexuality
1

---

from psychiatric diagnostic categories in both international and country-specifi c standards.
Outside of Eurocentric cultures, terms such as ‘homosexual’ or ‘gay’ often have no local equivalent, but that
does not mean male-to-male sex is unrecognized. Previously, for centuries in much of the world, ordinary
people made little fuss over male-to-male sex and often left it unlabelled and invisible. As most men who
experienced sex with other men also had sex with women and were married, there was no need to see them
as a separate species, so to speak. Male-to-male sex under such cultural regimes may be viewed as a kind of
socially harmless erotic play or a normal developmental stage, having little impact on the basic imperative
of family formation and reproduction.
In the world’s remaining small hunting and gathering tribes there do not appear to be any specifi c terms
designating male-to-male sexual practice, or any associated sexual identities. Among some Australian Abo-
riginal tribes and among some peoples on the island of New Guinea, male-to-male sex was a major part
of initiation rituals and was also practiced simply for pleasure, but was never associated with any specifi c
sexual identity. From an anthropological view, it appears that, not until societies became more complex,
and population density and role specialization increased, did males who practice sex with other males be-
come a named phenomenon. Writings from early Greek, Sanskrit, Arabic, and Chinese civilizations record
terms and descriptions for homoerotic acts, feelings and the people who experienced them (Hinsch 1990;
Hubbard 2003; Murray 1995; Vanita and Kidwai 2000).
Currently, most reports of HIV or STI prevalence in the USA use the medical terms ‘homosexual/bisexual’
while behavioural studies more often use the term ‘MSM’ but, if the subjects of these studies are asked which
of the terms they would use to describe their sexuality, a signifi cant minority call themselves ‘heterosexual’,
‘straight’ or other sub-culturally specifi c terms. In complex societies, identity terms typically imply power
and status differentials. Sexual identity terms are similar in that socioeconomic status or ethnic minority
status can also be implied by these terms and, in some social and political settings, are likely to be resisted,
both because they often further marginalise already marginalised people and because they implicitly disclose
one’s conduct publicly in a setting where such disclosure may be damaging (Mao et al., 2002).
The term ‘MSM’ has now, unfortunately, become as diffi cult to use as is ‘gay’ or ‘homosexual/bisexual’. Ex-
ported widely and adopted in other countries, the term has lost much of its original meaning. For example,
in Myanmar, discussions with non-government organisations (NGOs) and members of the communities
they serve revealed that, although they have several local terms for males who are open about their prefer-
ence for sex with other males (apwint) and those who are not (apone), when asked who these people have
sex with, the only answer was ‘men’. When asked if these are MSM, I was told ‘no, they are just men.’ In
Myanmar and elsewhere, the term is misused, misunderstood and no longer useful.
It is here that we have to distinguish between sexual behaviours, sexual preferences or attractions and
sexual identity. When epidemiological survey researchers pay attention to unpacking these components
of sexuality, they fi nd that there is rarely 100 percent concordance between practice and identity among
men or women. In the USA, a series of studies has shown a range of concordance between behaviours and
identity terms, from about 70 percent to 85 percent, depending on where the samples are collected (Pathela
et al., 2006). The concordance appears lower in population or venue-based vs. clinic-based samples. That
is, it has been found that, men who have more male sexual partners than female may call themselves ‘het-
erosexual’, while men who have more female sex partners than male may call themselves ‘homosexual’, as
well as other discordant combinations.
A few published studies in Asia have attempted to distinguish these components of sexuality. One study
with probability sampling was done by Centres for Disease Control in Thailand among technical school
students using the audio computer assisted technique, one which allows greater privacy and anonymity than
any other method devised to date. The males who stated they were sexually attracted only to females (92.2
percent of 893), comprised 20 percent of those who labelled themselves ‘homosexual’. Among males who
said they were sexually attracted only to other males (4.1 percent of 893), only 70 percent considered they
were homosexual, 21 percent labelled themselves bisexual and 9 percent said they were heterosexual.
In practice, the men who called themselves homosexual or bisexual had, on average, 25 percent female
sexual partners and 75 percent male. The majority, 81 percent, had their fi rst sexual contact with another
male. Among those calling themselves heterosexual, 4.1 percent had fi rst sexual contact with another male
2

---

(van Griensven et al., 2004). One study in China found that, although 49 percent of 426 men who had
sex with men also had sex with women and a third had been married, only 24 percent labelled themselves
as bisexual, almost all of the rest stating they were homosexual (Beichuan et al., 2000). In a probability
sample of males who had sex with males in Phnom Penh, among the 27 percent who described themselves
as homosexual, in the past month 45 percent had had multiple female partners as well as 90 percent having
multiple male partners (Girault et al., 2004).
These components of human sexuality are made more complex by gender and gender roles. In many
societies and periods of history, genetic males have taken on female gender dress and social roles. While
this is not always associated with a preference for male-to-male sex, it often is and today we call this trans-
genderism. This pattern is not to be confused with transvestitism, a term referring to mostly heterosexual
males who take pleasure in dressing and living as females at times. Male-to-female transgenderists are
nearly always people who feel more comfortable living in third-or mixed gender communities or roles, with
features of both male and female gender in their personalities and role behaviours. Some, a much smaller
number, prefer to live totally as women, but this has become more of a possibility only with the develop-
ment of gender-reassignment surgery. With the use of hormones, breast, jaw, genital and other surgery, a
genetically-born male can become virtually indistinguishable from a female and live according to female
gender roles. In some countries, such individuals may also be permitted to change their legal names and
passports. At the local level, terminology for transgenderism is also complex but usually easier to clarify
than the term MSM.
Where men who prefer sex with men do not feel that a feminised public self is required to express themselves,
those who are living as women have, in some countries, become further marginalised (Taywaditep 2002).
Many male-to-female transgenders and transsexuals do not see themselves as ‘men’ who have sex with men,
as they have spent most of their lives adjusting their bodies to express their feminine psyches. In addition,
many (though not all) of the men with whom transgenders have sex, consider themselves heterosexual.
There is a growing refusal among many such gender-altered groups to allow themselves to be classifi ed as
‘MSM’ for HIV/AIDS programmes (Boyce et al., in press). As the lives and needs of transgenders are often
considerably different from other sexual minority males, it seems inevitable that they will demand their
own programmes and organizations.
Caution must always be applied when thinking about sexual identity classifi cations. A fully life-long uni-
tary and fi xed sexuality is not the fate of everyone. Where investigated, in the USA, sexual orientation and
identity for some people are fl exible and change throughout the lifetime, apparently more often among
women than among men (Kinnish et al., 2005). Universally far more males than females demonstrate a
strong drive for sexual partner variety without romance or commitment as well as a quicker time to con-
sent to having sex with a new partner (Schmitt et al., 2000), a fi nding resulting from surveys of 16,288
people in 52 nations within ten major world regions and that holds across sexual orientations. The degree
to which this is a result of gender socialisation, including the social construction of sexual/gender identity
and practice, or has a basis in evolutionary adaptation still remains unclear, but the role of these fi ndings
in HIV risk behaviour for all males must be acknowledged.
Today, in countries such as Australia, the USA, the UK, and much of Western Europe, the English word ‘gay’
has become the primary identity term most people know for those practising male-to-male sex. The term
evolved with a political purpose and is associated with men who openly disclose their sexual orientation
and preferences. Where gay communities are formed as neighbourhoods with commercial venues specifi -
cally designed for gay men, most male-to-male sexual encounters take place among gay men. However,
outside of these communities, among minority ethnic groups and in many Asian and Pacifi c countries,
identifi cation as ‘gay’ is not the predominant self-identity term and the sexual partners of male sex workers
and others who do not sell sex, may not be ‘gay’ or ‘MSM’-identifi ed men at all. Because these men may
not identify as anything other than ‘normal’ men, and simply seek male-to-male sex while maintaining
a normative male social identity at their jobs, in their marriages and families and so on, they have been
shown to be very diffi cult to reach with standard targeted HIV prevention programmes. This represents
one of the major challenges in HIV prevention today.
Globally, it is probably true to say that most men who have sex with men also have sex with women, ranging
from seldom to over half the time, and, depending on the society, many are married to women. In Hong
Kong, for example, a recent survey revealed that 40 percent of men who had sex with other men within
3

---

the past 6 months were currently married (Lau et al., 2004). In Beijing, in a sample of 482 men who had
sex with men found that, within the past 6 months, 28 percent had sex with both men and women and 11
percent had not used condoms with either (Choi et al., 2004). In a sample of 2910 rural men in India, 4.3
percent reported having had sex with another men in the past year, 58 percent of who were married (Verma
and Collumbien 2004). In Chennai, a random sample of 774 men from 30 slums, showed 5.9 percent
had sex with men, 57 percent of whom were married. These men also were eight times more likely to test
positive for HIV than the others (Go et al., 2004). Similar fi ndings have been recorded in Cambodia, in
Indonesia, in some samples in Thailand, in Bangladesh and Pakistan. Hence, it is certain that in the context
of Asia, men who have sex with men defi nitely represent a vulnerable group who, if infected, can pose a risk
to others, including the spread of HIV heterosexually from and to women and on to their infants.
Further, among men who sell sex for a living, some sell sex to both men and women and may have differ-
ing personal preferences, irrespective of their commercial life. For example, in a recent survey of 582 male
sex workers on the streets of Dhaka, 28 percent reported being either married to women or had a woman
as their main partner. In addition, and in the past month, 11 percent had bought sex from females and 13
percent from males (Government of Bangladesh 2000). Similar patterns of practice have been documented
in Thailand, Bali, Cambodia, and elsewhere. There are, in addition, groups of men who identify almost
totally as heterosexual and would never call themselves male sex workers who also have sex with men,
mainly to acquire drugs or money for drugs.
And, fi nally, situational male-to-male sex is common in enforced all male settings, such as detention centres,
prisons, remote oil rigs, and so on, and may have few implications for identity or even future sexual activity
at all. Published estimates of the proportion of men who engage in sex with men in prisons range from 2
to 90 percent (Hellard and Aitken 2004).
It is becoming increasingly clear to all students of human sexuality that sexuality in both males and females
is not dichotomous, but is highly situational, fl exible and diverse and these scientifi c fi ndings must be ac-
commodated in national HIV programmes.
How many males have sex with other males?
National or other representative samples of males throughout the world usually fi nd between 5 and 20
percent have had sex with another male some time in their lives, although in certain countries propor-
tions were higher. However, the proportion of males who report recent male-to-male sex within the past
year or past 6 months, is always considerably lower, ranging from 2 to 10 percent, or approximately half.
Certainly, the manner in which these surveys are conducted and the degree of stigma associated with male-
to-male sex varies region by region and can be expected to infl uence survey results, most likely towards
under-reporting. Men who report being exclusively interested in male-to-male sex rarely exceed 5 percent
in any population.
Estimates of the number of males who have sex with males in different locations are hard to confi rm, but
given the attached stigma, must be considered as mostly under-estimates. The Bangladesh AIDS programme
estimates between 40,000-150,000 men who have sex with men, including male sex workers, which is surely
an under-estimate for a country of 147 million. Given a conservative estimate of 4 percent, there would
be at least 1.7 million men who have sex with men on a regular basis. Other estimates include 70,000 in
Ho Chi Minh City (Colby 2003), about 1.6 million in Indonesia (MOH Indonesia 2002), and 2-10 million
in China (Chinese MOH/UN Theme Group 1997). Published estimates of male-to-female transgenders
include 12,000 waria in Indonesia (MOH Indonesia, 2002), 200,000 katoey in Thailand (Winter 2002),
10,000 mak nyahs in Malaysia (Teh 1998), 26,000 warias (MOH Indonesia, 2005), a million or more hijras
in India, and 12,000 hijras in Bangladesh (MoH 2000). A recent review attempting to examine surveys
from around the world suggested that lifetime prevalence of male-to-male sex was 3-5 percent for East
Asia and 6-12 percent for South and Southeast Asia (Cáceres et al., 2006). This same study also estimated
that the prevalence of male-to-male sex during the past year was approximately half of the lifetime fi gures
but that the prevalence of unprotected sex was around 40-60 percent in East and Southeast Asia but 70 to
90 percent in South Asia.
4

---

Cultural patterns of male-to-male sex in Asia and
the pacifi c
South Asia
In South Asia, studies repeatedly show local male-to-male sexual practices of two major types, the more
common gender status-based pattern and the less common elder-younger relationship (Khan and Khilji
2002; Jenkins 1998). In Pakistan’s Northwest Frontier Province, the practice of ‘bachabazi’ exemplifi es the
less common elder-younger pattern. Men may have sex with teenage boys and some may be seen publicly
with their lovers, often sponsoring their education, clothing and other care (Family Health International
2003). Among the Pashtun in Afghanistan, a recent article documents the traditional practice, estimated
to occur among as many as 50 percent of all males, of sex between halekon, young, “beautiful “ males
between about 12 and 19 years old, and older men who reward them with money and gifts. During the
Taliban regime, such practices were reported to be severely punished. and some local observers thought the
practice had diminished but is now again rising. Locals caution that this is not a refl ection of homosexual
identity, rather of the limited options among males to have sex with females (Reynolds 2004).
The national behavioural surveillance surveys of India (NACO 2002) document how some males have sex
with males before having sex with females. In a sample of 1,387 men who practice male-to-male sex (19-
35 years old) from Delhi, Kolkata, Mumbai, Chennai and Bangalore, the median age at fi rst sex (defi ned
as manual, oral or anal) with another male was consistently lower (15-17) than the median age at fi rst sex
with females (16-22). Their fi rst male partners were usually older than they were, and were mainly friends,
neighbors and relatives.
What percentage of young men experience this pattern on a national scale is less well documented. In Sri
Lanka, a study of male university students found 50 percent had sex with other males, often before having
sex with females (Silva et al., 1997). A 1997 survey of 977 single males (average age 18.8) in Mumbai of
low income status attending colleges found that 18 percent had had a same-sex sexual experience and 23
percent had had sexual intercourse with a female (Abraham and Kumar 1999). A large study conducted
by the National Institute of Health and Family Welfare in Delhi and Lucknow slums among young people
from 15 to 25 years old showed that male-to-male sex was reported by 5 percent of respondents (Sharma
2001).
While general population statistics vary, it is nonetheless clear that some proportion of young males fi nd
that their experience of sex with other males resonates with their own sexual desires and gender role prefer-
ences. Besides hijras, only a minority of whom are castrated, and zenanas (non-hijra transgenders), there
are a large number of kothis. Kothi, a term now widely used in India and Bangladesh, refers to men who
are not transgendered but assume feminine traits that attract male sexual interest. Their (usually) inser-
tive partners are called panthis (girya in northern India and Pakistan). Kothis are usually fairly feminine in
behaviour, cross-dress or partially cross-dress, and some take female hormones in order to develop breasts.
There are many variations, with civil kothis in Bangladesh who present publicly in male gender-normative
manner and gupti kothis in India, who are more completely hidden. Due to the social and familial pres-
sures, many kothis eventually marry and produce children. Sex within marriage is performed as a duty by
some, but others seem to take pride in the fact they can perform satisfactorily with women and prove to
themselves they are really men, even though receptive sex with other men continues for pleasure (Khan et
al., 2005). Some kothis, particularly if poor, may become sex workers.
Other men also sell sex who are not kothis (e.g., malishias in Pakistan). The ‘real’ men who have sex with
kothis and other male sex workers are often called panthis or giryas (by the kothis). Being men of all ages
and types, married and unmarried, who seek sex with other men, they are not a homogenous group. The
sexual practices engaged in are usually, but not always, according to gender roles, i.e. insertive oral or anal
intercourse for panthis and receptive for kothis. The term do-parata is used for those who switch sexual
roles (in Pakistan, known as chavas). It should be noted that these are fl uid categories and the range of
variation within each overlaps in different ways, particularly with regard to sexual practice.
There are, in addition, Hindu temple transvestites dedicated to Goddess Yellama, the jogta, or the male
equivalent of the devdasi. The jogta are mostly found in Karnataka and Maharashtra, but south India
5

---

has similar roles, known as Shiv Shakti, Parvatis, and Ganacharis. While they can marry women and have
families and are not hijras, today many have become transvestite sex workers.
In Nepal, metis is the term for the more feminized, sometimes cross-dressed males; dohoris is the term used
for those who present in a gender normative manner in public, and tas refers to all the other males who have
sex with both of these sub-groups, as well as with women. In Nepal, the police are particularly abusive of
these divergent males, especially the metis, and create severe barriers to effective HIV outreach. Blackmail,
extortion, the threat of exposure and coerced sex are common experiences (Boyce and Pant 2001).
In addition, poor young men may work on the beaches offering sex to tourists and others for pay (Rat-
napala 1998). Gay identifi cation in Sri Lanka is highly stigmatized, despite widespread male-to-male sex
(Ammon 1999).
South East Asia
In South East Asia, male sexual diversity has been documented best in Thailand. Elsewhere, except for
Myanmar, traditional patterns are not well known. Modern technology, including chat rooms and the
Internet, is working rapidly to transform whatever might have been the least bit traditional. Commercial
sex has a major presence in several countries, particularly in Cambodia, Thailand and Vietnam. Tourism
has played an important role in fostering the male sex trade and, due primarily to political and economic
conditions, a large percentage of both male and female sex workers in northern Thailand today come from
Myanmar (N.A. 2006). Studies report that some Thai non-katoey male sex workers are not homosexual
by preference and provide services to both male and female clients (Beyrer et al., 1997; McCamish et al.,
2000). Similar patterns have been found among male sex workers in other countries.
Thailand’s confi gurations of male-to-male sex have been extensively explored in print and the transgen-
der/transsexual role of the katoey has been the object of numerous popular fi lms and shows. Feminized
males and homosexual practices are not acceptable in Theravada Buddhism in the context of the monastic
life and it is often said that katoey are suffering in this life due to having been adulterers in a former life
(Jackson 1998). Yet, at least in northern Thailand, katoey do have a role in temple life and male-to-male
sex in Thailand is relatively easy to experience for any male who desires it without major implications for
one’s masculinity, as long as it is a private act. While it seems true that Thai society is quite tolerant of
gender variance and alternative sexualities on a public level, privately families are often very shamed by
the feminine presentation or homosexual behaviour of a son and there are numerous examples of young
men who acquired HIV having been tossed out of their homes to die elsewhere (Jenkins et al., 2006).
Consequently, despite well-lauded HIV prevention programmes for female sex workers and their clients,
until recently, Thailand had neglected to scale-up prevention programmes for men who have sex with men
(McCamish et al., 2000).
In Cambodia, a distinction is made between the ‘long hairs’, or more feminized, often transgendered males
(srey sro), and the ‘short-hairs’ (Sovannara and Ward 2003). The presence of same sex relations between
men been generally ignored (except by the former King of Cambodia who advocated for full rights for
same-sex couples) and high levels of stigma and harassment occur. However, there is both anecdotal and
archival evidence of a previously existing transgender role prior to the culture-destroying decades of recent
Cambodian history. In 2000, a survey of men who had sex with men in Phnom Penh showed 14.4 percent
had HIV. Most sold sex to men and some to women; 36 percent had more than one female partner the
previous month, many of whom were themselves sex workers (Girault et al., 2004).
The available literature on Vietnam is fairly silent on traditional confi gurations of same-sex desire. It is
known that Vietnam have female and transgender shamans, known as n-dom-kho or bong lai cai or bong
lo, who wail mournfully at funerals (Heiman and Cao 1975). But both the French occupation (Proschan
2002) and the modern state seem to have erased these traditions, except perhaps in the most rural areas.
The common slang word is ‘Be De’ (derived from ‘pederast’) but most people have little understanding of
what it means. Today the common terms used are bong lo for those who are feminized in public and bong
kin who are not and are themselves often married (Bao et al., 2005). There is no law against male-to-male
sex but there appears to be a great deal of stigma and fear of exposure. In urban Vietnam, one can see
both highly feminized ladyboys and ‘gay boys’. Some people refer to a group they call ‘offi cial’ gays, those
with higher-level positions in society who would not be expected to overtly reveal their sexual identities.
6

---