Communication about sex and sexual health is an important facilitator in gaining accurate knowledge about prevention of sexually transmitted diseases (STDs) and promotion of sexual health. Understanding how and with whom communication about sex occurs and the nature of the information exchanged is valuable in designing sexual risk prevention interventions. In this study of low-income communities residents in Chennai, India, our aim was to understand the composition of personal communication networks, the nature of information related to sex and sexual health that is exchanged in these networks and the value of communication among members of these networks. We conducted in-depth open-ended interviews using a structured interview guide with 43 individuals. We also conducted 12 focus group discussions. Individuals were selected using snowball sampling. Our results indicate that information about sex and sexual health is exchanged within and between four groups: married women, married men, unmarried men and unmarried women. Communication leads to an expansion of sexual networks among unmarried men and treatment seeking behaviour for STDs in all groups. Unmarried men offer immense potential for intervention given the range of topics related to sex and sexual health that are discussed and the risky sexual behaviours practiced. Spousal communication about sexual behaviour or sexual health is minimal and shifting norms for prevention would be difficult. Interventions identifying communication networks and influencing the natural communication patterns in these networks may be a viable HIV prevention strategy in the study area.


Community-based health promotion programs in part rely on the communication of one or more messages that help inform and empower individuals to take control of their health. Communication is broadly defined as the exchange of information between people (Rogers and Kincaid, 1981). Communication networks are comprised of individuals linked to each other by virtue of the information they exchange. While an individual may simultaneously belong to several networks, the nature of information exchanged in a network is often specific to the interests of network members.

In HIV/STD (sexually transmitted disease) prevention programs, we are interested in communication about sexual health. Communication about sexual health can be defined as the exchange of information between individuals that can potentially promote behaviour change and reduce the risk of STDs and AIDS (Gupta et al., 1996). For instance, sexual partners may talk about safe sex (Moore et al., 1999), encourage mutual testing for HIV and STDs (Hutchinson, 1998) and discuss test outcomes (Potterat, 2003). In addition to communication between partners, communication can also take place between members of a network. Health promotion interventions such as those to promote family planning (Kincaid, 2001) and HIV prevention (Latkin, 1998) have used peer social networks as an important starting point to change behaviour. Peers have influenced HIV risk behaviours through exchange of information, modeling and reinforcement and facilitating acquisition of new sexual and drug use partners (Fisher, 1988).

In this paper, we report on research conducted in the city of Chennai (formerly Madras) in southern India to understand the composition and content of interpersonal sexual communication among networks. India has ≥5 million adult HIV infections (UNAIDS, 2004). Sex and sexuality have long been considered a sensitive subject here and policy makers perceiving socio-cultural constraints to discuss this issue are cautious to introduce such discussions in the public realm (Nag, 1998). With the spread of HIV in India, it is important to understand the extent to which sexuality and sexual health are discussed in order to inform design of HIV prevention interventions.

Our study data were collected as part of the ethnographic phase of the US National Institute of Mental Health (NIMH) HIV/STD collaborative prevention trial. This trial is a randomized controlled trial to test the efficacy of HIV prevention messages disseminated through community popular opinion leaders (CPOLs). CPOLs have facilitated health promotive behaviors in diverse settings in the USA (Kelly, 2004) and the goal of this trial is to determine if this model of prevention can work in international settings. In the trial, we aim to identify and train CPOLs to disseminate HIV prevention messages to members of their network. Our ethnographic phase had three specific aims. The first was to understand the composition of personal networks and to identify the characteristics of opinion leaders. The second was to examine the content of communication about sex and sexual health in these networks. Finally, we looked at how this communication influences two preventive behaviours—unprotected sex and seeking care for STDs. These research aims were guided by the theory of diffusion of innovations (Rogers, 1995). This theory posits that new ideas spread through a community via interpersonal communication, and analyzing the flow of information among social networks (Rogers and Kincaid, 1981) is a key method of understanding how communication takes place. We present the methodology, outline results, following which we discuss how study findings can inform HIV prevention and sexual health promotion interventions in similar settings.


Participants in our study were either clients of the government STD center or lived in low-income public housing communities known locally as ‘slums’. A typical slum has an area of a couple of city blocks and may be inhabited by as many as 500 families living in horizontally spread tenements.

Forty-three in-depth interviews were conducted with residents of slums and STD clinic clients. Twenty-four STD clinic clients (11 women and 13 men) and 19 (seven men and 12 women) individuals selected from the slums were interviewed. We also conducted 12 focus group discussions with slum residents (seven with men and five with women). Overall, 30 women and 39 men participated in the focus group discussions. Slum residents were contacted through ‘key informants’—leaders and prominent citizens of the slums. Participants in clinics were recruited by interviewers after they sought care. Recruitment involved administration of informed consent and the once the participant agreed, the interview was conducted in a pre-arranged venue in the slum or private space in the clinic, respectively. All participants were between 18 and 40 years in age. Of the STD clinic interviews (n = 24), all women were married while eight of the 13 men were unmarried. Of the community interviews, more men (four of the seven) were married while all women were married. We conducted one focus group discussion with unmarried women. Recruiting these young women was difficult because they were not permitted to participate by their families. The majority of participants in the other focus groups were married women and unmarried men. Participants shared information without inhibition after the study goals and objectives were clarified. Often, there were questions and interactions after the close of an interview or discussion.

In the focus groups, trained interviewers asked questions to understand the composition of social and communication networks. In in-depth interviews, we explored the patterns of communication about sex and sexual health and sexual behaviour at the individual level. Using both methods, we gathered information on the characteristics of CPOLs who influenced sexual communication. All interviews and discussions were tape-recorded in the local language, translated and transcribed in English. Transcripts were analyzed using Atlas.ti (Muhr, 1998), a textual analysis software program. The data were reviewed for three main themes or codes as they were called in the software program: composition and characteristics of social networks, the content of communication in these networks and the association of communication with sexual behaviour. Text that matched these codes were retrieved and reviewed. Matrices were developed for each code to enable the organization of the data and to understand similarities and contrasts across related themes. This study was initiated after ethical review and approval from the institutional review boards of YRG Center for AIDS Research and Education in Chennai, India and the Johns Hopkins University Bloomberg School of Public Health in Baltimore, MD.


Network composition and opinion leader characteristics

Information was exchanged among peers in four categories: married women, married men, unmarried men and unmarried women. Married women's networks consisted of women in the general neighborhood, immediate neighbors and women engaging in specific activities, such as standing in line for daily water supply or playing games in the evenings. Women typically met in groups of three or four people. Women often came together for social support and attempted to offer verbal and physical assistance. One married woman urged a younger woman to talk to about her experience with spousal abuse and reported:

Since she is from a different place, she does not have anybody to help her. So we go and talk to her husband (to prevent abuse). (Married woman aged 30.)

Within networks of married women were natural ‘opinion leaders’—married older women who were more vocal than others, bold and prominent in their demeanor and behaviour. Women referred to them as ‘akka’ or older sister. These women were more likely to discuss issues of sex and sexual health openly.

While married women formed networks based on location and personal experience, married men did so based on common interests. Neighborhood-based networks consisted of those who may come together to have casual conversation after lunch or dinner. There were usually three to four men of the same age range within a group. Married men formed work place friendships, formal group memberships and often met with friends in bars. Informal groups of four to five men also gathered on a regular basis at a designated spot in the community to play dice. Opinion leaders among married men were usually older, respected and credible. One desired characteristic of married male opinion leaders was that they should themselves display safe sexual behavior and not have any vices. Men reported that if this was not the case, people would say ‘correct yourself before you preach!’.

The communication networks of unmarried men and women varied widely relative to their married counterparts. Unmarried women were closely guarded in their communities. Their networks were restricted to their friends and their immediate family members. Respondents reported that they did not perceive a need to discuss sex, as it did not affect their lives.

Unmarried male networks were unrestricted to residential location or age. Common activities included playing sports, drinking together and visiting venues where sexual services were available. Friendship groups ranged from eight to 20 members and included younger married men. In addition to these informal groups, formal groups such as local cricket clubs were a venue for socialization. Unmarried men looked up to peers who are popular and sexually experienced. They looked to opinion leaders who were aware of issues relating to sexual experiences and sexual health. As with married men, unmarried men often looked up to older men and preferred individuals who did not have any vices. There was also substantial overlap among these networks. For instance, young unmarried men would play cricket with married men but a few unmarried men from the cricket playing group would also belong to a group that goes drinking.

We note here that among all networks reported here, opinion leaders were always part of a network and were referred by men and women simply as ‘friends’ (the English word was used).

Inter-network communication

In addition to communicating between members of these networks, individuals communicated between networks. The first involved sexual communication between married women and unmarried men. This involved sexually suggestive talk and teasing, and typically resulted in sexual intercourse. Such communication about information on the location and nature of sex workers was facilitated by personal networks of unmarried men. This is illustrated in the quote below. In this quote, the sex worker is an ‘auntie’—the local term for a married woman who is offering sexual services.

I used to ask her ‘enna avan sappitu porran, ennakku illyaa’ (how come he comes to eat at your house, but nothing for me?). So she said—‘nee venna vandhittu po’ (if you want, you can come too). So I went one day to experience this. (Unmarried male, age 25.)

The second communication pattern across networks was between men (married and unmarried) and unmarried women. This communication was initiated either during the course of normal social contact between strangers or between family members. For instance, a married vegetable vendor reported befriending an unmarried customer, and a tea vendor flirted with a young girl passing by on her way to school. In both these instances in our data, communication was followed by sex with no significant courtship period. When family was involved, we found references to male cousins or other relatives who engaged in sex or sought sex from a sex worker.

Content of communication

While a variety of topics were discussed within networks, sex and sexual health were prominent. Married women's discussions can be classified under the following broad categories: teasing (‘did you play the cat and mouse game last night?’ to ask if a woman has sex with her husband), openly discussing sexual lives and sexual health.

Married men's communication was generally limited to close friends, spouses and sexual partners. They rarely discussed their sexual needs or sexual health with their wives or anybody else. The following married STD clinic client justified keeping silent about sexual health as follows:

Because I thought people will look at me with a different perspective. That is why I chose to keep quiet. (Married male, age 27.)

Most of the data in our study about communication within marriage comes from women, not men. Women reported that they were shy to discuss their sexual needs verbally with their spouses. One married woman said,

When I get ‘unnarchi’ (sexual desire) and if my hand comes in contact with him sometimes, he will say his body is aching and he is not interested. At that time I used to wonder why he is not satisfying me. But what can I do?

Women reported acquiescing to the sexual demands of their husband, even though they were uninterested or unwilling:

He will ask me to come and sleep, began a married woman, but I will tell him no. But even then he will not allow and he will hit me. (Married woman, age 29.)

Unmarried men engaged in extensive conversations about sex and sexual health. They used a large repertoire of phrases and words to refer to sexual intercourse and parts of the human anatomy, e.g. kai (raw fruit) to refer to breasts; saaman (property) to refer to penis. They also used terms to describe sexual feelings and sexual behavior: verri–a violent sexual urge, how they set up a super figure (figure is a term used to refer to a sex worker). Unmarried men also advised peers about sex and sexual opportunities:

I told my friend about how my girlfriend was not objecting to my being physically intimate. So he suggested that I try having sex with her ‘matter pannudaa’ (finish the ‘matter’). I went ahead and had sex. (‘Unmarried male’ age 16.)

Sexual intercourse was not always consensual. Unmarried men joked about touching girls and women in public places such as in a crowded bus. Such non-verbal gestures were seen as a step towards gaining sexual experience. In addition to communication about sexual experiences, men reported being aware of STDs and AIDS. As one respondent remarked:

You must have a single partner to prevent AIDS, it spreads if you go to many women. (Unmarried man, age 23.)

Another distinguished between HIV and other STDs:

AIDS is basically a disease that cannot be treated and it results in quick death, whereas a person with STD can be cured. I know people get various diseases like that especially if they go to different women; but there are medicines to treat them. (Unmarried man, age 28.)

Awareness about condom use was also high: ‘I will not have sex without a condom’, responded one man who reported multiple partners. Another said that condom education was ‘everywhere’ referring to news and print media and reported using condoms with sex worker. However, despite condom awareness and use by some men, others revealed several factors outweighing its use. One factor was peer pressure as reflect in this quote by a young man, who recalled his experience with a girl he liked when they were on a college field trip:

It (sex) happened not because I forced my self on her, not because I was having the ‘kama veri’ (lust for sex) but my friends will say that you should have sex whenever you have ‘tension’ (erection) and that sex will never happen when you want to be sincere in our approach to the girl. So I thought when I got the opportunity I should make use of it. (Unmarried man, age 19.)

Another factor is the need to ‘get value for money’ with a sex worker, as this quote reflects:

Sometimes when we find a woman who is not willing to have sex without using the condom. We will say, ‘we are paying you and you cannot give us the rules, there are better figures (sex workers) than you in this area’. When we say that they will comply. (Unmarried man, age 21.)

While unmarried men discussed sex openly, unmarried women do not. One woman said that ‘it is enough for the men to know about all this (sex)’ and that they would wait to find out about this when they get married. Older female relatives communicated expectations of unmarried women on their wedding night using dictates such as ‘He will do everything’, or ‘listen to what your husband says’. Men who courted unmarried women also used non-specific language such as ‘you will see what I do’, to address queries from women who wondered about the interest in them.

Communication and sexual risk behaviour

Our findings show that communication about sexual risk behaviors was different for married and unmarried individuals. Married couples talked about condom use and health care seeking behaviors resulting from sexual risk behaviors. Unmarried individuals talked about sexual risk in general and less about health seeking behaviors. Among married couples, sexual risk behaviour (multiple partners, extra marital affairs) was typically inferred by women who reported initiating conversation with their husbands about the behavior itself or talking about condom use. As in this quoted from a married woman, discussing condom use with a husband (a long distance truck driver) who she suspects having multiple partners was futile:

He has been traveling and could have been with other women, I am scared of what could happen. So I ask him to wear a condom and he will retort saying that he would slipper me if I suspect him of cheating on me. So I will trust him and we have sex without any protection. (Married woman, age 24.)

In those cases where marital communication did lead to care seeking for a sexual health symptom, care was incomplete and unresolved. A married woman reported that both she and her husband had an STD and upon the urging of his doctor, he took her along to seek care:

The doctors talked to us about condoms. They gave us condoms and said that until we are fully cured, we should use condoms during sex. They also said if we do not use a condom, this disease may not be cured from our bodies. When we went home, I used to ask him to bring ‘that’—condom. He used to react saying that he is clean and they are just like that saying and nothing is wrong with him. He said that I am unnecessarily trusting the doctors. He will not bring the condom and when I insist, he used to abuse me verbally and hit me. (Married woman, age 28.)

Unmarried men's discussion about risk behavior did not focus on condom use but on sexual risk. The following statements between friends on risk behavior illustrate this:

Having unprotected sex once will not cause any problems (STDs). (Unmarried man, age 25.)

I did not use condom because she looks nice and she has sex only with their husband. (Unmarried man, age 18)

In another instance, a man reported ‘Since talking to him, I always use condoms’ recalling a discussion with an STD-infected friend who urged him to have safe sex.


Our findings suggest that despite perceived taboos to sex communication, a wide range of issues related to sex and sexual health are in fact discussed within and between networks in the slum communities in Chennai. That such topics are discussed freely and in-depth both within networks and shared freely with researchers is a starting point for interventions that focus on HIV prevention and sexual health promotion. Our findings suggest two main areas of focus for planners of health promotion programs. First is the pattern of communication within social networks, which can inform program design, and second is the content of this communication, which can inform program participation and content.

Communication patterns are based on networks that individuals belong to. We found that among each network seen—married women, married men, unmarried men and unmarried women—opinion leaders who are integral members of a network were identified. These individuals were considered credible sources of information about HIV/STDs prevention, social support related to STDs (married women, married men and unmarried men), information on sex, sexual partners and disease prevention (married and unmarried men). Based on earlier successes with opinion leader-led interventions (Sikkema et al., 2000; Kelly, 2004), this finding suggests that in Chennai, health promotion programs may train opinion leaders to discuss AIDS prevention with members of their network.

Analysis of health communication content shows that stark gender and marital status differences. Men enjoy unrestricted expression of their emotions before marriage and while unsafe behaviors continue after marriage, there is very little communication about this. Married women commiserate on marital problems and stressors such as domestic violence. Unmarried women do not think that sexual matters affect their lives and are guided by more experienced persons. One can speculate that marriage marks a turning point in communication within and between genders. In a culture that expects female virginity at marriage and given the spread of HIV to monogamous married women whose only risk is being married to a promiscuous husband (Gangakhedkar et al., 1997), there is a need to address prevention messages to unmarried women. Successes in mobilizing young women to talk about sex and sexual health, albeit few (Shedde, 2002), are encouraging and more HIV prevention efforts targeted at unmarried women are needed.

However, we find that amongst all networks, communication does not include significant discussion of health promotive behaviors. Rather, encouraging non-consensual sex, sharing inaccurate risk perceptions (among unmarried men), or unsuccessfully communicating with spouses about safe sexual behavior and care seeking (among married men and women) are examples of behaviours that might increase HIV and STD risk. Perhaps this reflects a lack of accurate information about HIV and STDs. Studies have also suggested gender differentials (Gupta, 2002) and male sexual preferences (Roth et al., 2001) may fuel these behaviors. Our findings suggest that men's safe sexual behaviors were preceded by communication with peers who were affected by HIV/STD or exposure to media. Thus, while empowering young unmarried women with knowledge of reproductive health, we also need to focus intervention efforts to involve men, unmarried men in particular.

With men, HIV interventions need to focus on developing communication skills. Using the network approach to intervention design, interventions may begin by identifying male opinion leaders and training them to communicate prevention messages that reflect scientific facts and dispel myths about transmission. Developing message delivery strategies that take into account social context of conversation, modeling conversations that reflect natural communication patterns and framing appropriate message delivery strategies are some ways to involve male opinion leaders and help them be change agents in their networks (Kelly, 2004). Based on our study, we can see that men and women referred to ‘opinion leaders’ simply as ‘friends’. This lack of hierarchy facilitates communication. In order to influence inter-network communication (which typically leads to sexual behavior), we can facilitate inter-network opinion leader training sessions where prevention messages could be tailored to those who cross network lines to facilitate diffusion. We further note that provision of accurate information about prevention is only one factor that influences safe behaviour. Other factors such as network relationships (Latkin, 1998), and contextual factors such as condom availability and barriers to use (Roth et al., 2001) need be studied further and incorporated into a network-based intervention. In addition to opinion leaders, health promotion programs can work with providers of STD services who can make a valuable impact on their patient's behavior by talking to them about prevention and partner notification.

In conclusion, we would like to outline some limitations of the study. The study reports from a small sample of individuals and findings may not be generalizable. However, we feel that using focus group discussions and in-depth interviews helped collect very descriptive and detailed information on sensitive issues. Our methodology allowed participants to share personal experiences without inhibition more than they might have with structured interviews (O'Brien, 1993). We also did not interview CPOLs or ‘egos’ (Rogers and Kincaid, 1981) to elicit information on their own beliefs and behaviours relating to HIV prevention and to examine the extent of their influence in their networks. Although the focus of this paper was to get information on content of communication and to characterize opinion leaders, information on a network's ego would help better understand the dynamics of and factors influencing network communication such as social pressure, network relationships and network norms. These factors need to be better understood in the Chennai context.

This research was funded by the US National Institute of Mental Health (NIMH) under grant number U10 MH61543.


Fisher, J. (
). Possible effects of reference group-based social influence on AIDS risk behaviour and AIDS prevention.
American Psychologist
Gangakhedkar, R. R., Bentley, M. E., Divekar, A. D., Gadkari, D., Mehendale, S. M., Shepherd, M. E., et al. (
) Spread of HIV infection in married monogamous women in India.
Journal of the American Medical Association
Gupta, G. R. (
) How men's power over women fuels the HIV epidemic.
British Medical Journal
Gupta, G. R., Weiss, E. and Mane, P. (
) Talking about sex: a pre-requisite for AIDS prevention. In Long, L. D. and Ankrah, E. M. (eds), Women's Experiences with HIV/AIDS: An International Perspective. Columbia University Press, New York, pp. 333–350.
Hutchinson, A. K. (
) Something to talk about: sexual risk communication between young women and their partners.
Journal of Obstetrics, Gynecology and Neonatal Nursing
Kelly, J. A. (
) Popular opinion leaders and HIV prevention peer education: resolving discrepant findings, and implications for the development of effective community programmes.
Kincaid, D. L. (
) Social networks, ideation, and contraceptive behaviour in Bangladesh: a longitudinal analysis.
Social Science and Medicine
Latkin, C. A. (
) Outreach in natural settings: the use of peer leaders for HIV prevention among injecting drug users' networks.
Public Health Reports
(Suppl. 1),
Moore, J., Harrison, J. S., Kay, K. L., Deren, S. and Doll, L. S. (
) Factors associated with Hispanic women's HIV related communication and condom use with male partners.
Nag, M. (
) Sexual behaviour in India with risk of HIV/AIDS transmission. Health Transition Review, 5 (Suppl.),
O'Brien, K. (
) Using focus groups to develop health surveys: an example from research on social relationships and AIDS-preventive behaviour.
Health Education Quarterly
Potterat, J. J. (
). Partner notification for HIV: running out of excuses.
Sexually Transmitted Diseases
Rogers, E. M. (
) Diffusion of Innovations. The Free Press, New York.
Rogers, E. M. and Kincaid, D. L. (
). Communication Networks: Toward A New Paradigm for Research. The Free Press, New York.
Roth, J., Krishnan, S. P. and Bunch, E. (
) Barriers to condom use: results from a study in Mumbai (Bombay), India.
AIDS Education and Prevention
Shedde, M. (
). Lucknow: where girls are in danger of ‘being educated too much’. Real Lives. International Planned Parenthood South Asia Region, July 2002, Issue
Sikkema, K. J., Kelly, J. A., Winett, R. A., Solomon, L. J., Cargill, V. A., Roffman, R. A., McAuliffe, T.L. (
) Outcomes of a randomized community-level HIV prevention intervention for women living in 18 low-income housing developments.
American Journal of Public Health
) AIDS Epidemic Update. (www.unaids.org), last accessed on July 30, 2004.

Author notes

1Department of Epidemiology, Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD 21205 USA, 2YRG Center for AIDS Research and Education, Voluntary Health Services, Taramani, Chennai, India and 3School of Public Health, University of North Carolina, Chapel Hill, NC 27516, USA