Male migrants' non-spousal sexual partnerships in the place of origin: an in-depth investigation in two rural settings of India

Abstract

Male migrants in India are at disproportionately high risk for HIV, not only because of their sexual behaviours in destination areas but also due to their risk behaviours in their place of origin. While studies have documented male migrants' risky behaviours in the home setting, few have attempted to understand the underlying socio-cultural context in which they engage in such behaviours. This paper examines the patterns and context of male migrants' non-spousal sexual partnerships in two high-out-migration districts of India. Data, drawn from a cross-sectional behavioural mixed-methods study conducted in 2008, included a structured survey with 1272 migrants, followed by in-depth interviews with 33 male migrants. Results suggest that sexual activity was common in the place of origin: around 50% of migrants had sex with a non-spousal female partner and two-fifths had initiated sex in this setting. Migrants' non-spousal sexual behaviours in the home village were influenced by the prevailing socio-cultural context, including migrants' enhanced socio-economic status, attitudes to non-spousal sex and accessibility of sexual partners. Male migrants' non-spousal sexual partnerships in source areas are influenced by socio-cultural factors, which must be considered when designing HIV programmes in India and elsewhere.

En Inde, les migrants de sexe masculin sont exposés à un risque lié au VIH démesurément élevé, en raison non seulement de leurs comportements sexuels dans les régions oú ils se rendent, mais aussi de leurs comportements à risque dans leurs régions d'origine. Alors que des recherches ont permis d'approfondir les connaissances sur les comportements à risque de ces hommes dans leur environnement d'origine, peu d'études ont examiné le contexte socio-culturel sous-jacent à ces comportements. Cet article examine les modèles et le contexte des partenariats sexuels non conjugaux des migrants de sexe masculin dans deux zones à forte migration externe en Inde. Les données obtenues à partir d'une étude comportementale (transversale et multi-méthodes) conduite en 2008, incluent celles d'une enquête structurée menée parmi 1272 migrants, qui a été suivie d'entretiens en profondeur avec 33 migrants de sexe masculin. Les résultats suggèrent que l'activité sexuelle est courante sur le lieu d'origine : environ 50% des participants y avaient eu des rapports sexuels avec des femmes auxquelles ils n'étaient pas mariés, et deux cinquièmes d'entre eux y avaient initié leur sexualité. Ces comportements sexuels non conjugaux des migrants dans leur village d'origine étaient influencés par le contexte socio-culturel prédominant, englobant la situation socio-économique améliorée des migrants, les attitudes vis-à-vis des rapports non conjugaux et l'accessibilité des partenaires sexuelles. Les partenariats sexuels non conjugaux des migrants de sexe masculin dans les zones source sont influencés par des facteurs socio-culturels qui doivent être pris en compte dans l'élaboration des programmes de lutte contre le VIH en Inde et ailleurs.

Los emigrantes masculinos en la India corren un riesgo desproporcionadamente alto de contagiarse con el virus del sida no solamente por sus conductas sexuales en sus áreas de destino sino también debido a sus comportamientos de riesgo en su lugar de origen. Aunque en algunos estudios se han documentado las conductas arriesgadas de los emigrantes masculinos en su lugar de origen, en pocos se ha intentado entender el contexto sociocultural subyacente en el que participan con tales conductas. En este artículo examino los modelos y el contexto de las relaciones sexuales de emigrantes masculinos con parejas no conyugales en dos distritos de la India con un alto nivel de emigración. Los datos proceden de un estudio transversal con métodos variados sobre comportamientos llevado a cabo en 2008 y de un estudio estructurado en el que participaron 1272 emigrantes, en el que se realizaron entrevistas exhaustivas con 33 emigrantes masculinos. Los resultados indican que la actividad sexual era común en el lugar de origen: aproximadamente el 50% de emigrantes tuvieron relaciones sexuales con una pareja femenina no conyugal y dos quintas partes se habían iniciado en el sexo en este entorno. Las conductas sexuales no conyugales de los emigrantes en las aldeas de origen estaban influenciadas por el contexto sociocultural predominante, incluyendo un mejor estado socioeconómico de los emigrantes, las actitudes hacia el sexo no conyugal y el nivel de acceso a parejas sexuales. Las relaciones sexuales no conyugales de los emigrantes masculinos en las zonas de origen están influenciadas por factores socioculturales, que deben tenerse en cuenta al desarrollar programas contra el sida en la India y otros lugares.

Keywords:

• socio-cultural context
• male migrants
• India
• place of origin
• HIV

View correction statement:

Erratum

Background

Studies in India document that migration plays an important role in the spread of HIV among both men and women in migrants' place of origin (Saggurti et al. 2011a, 2012). Migrants are likely to acquire HIV due to their high-risk behaviours in destination areas (Deering et al. 2008; Saggurti et al. 2008, 2009), and transmit the infection to their sexual partners in the place of origin due to their continued risk behaviours (Dhapola et al. 2007; Saggurti et al. 2011a). Studies conducted in the place of origin suggest that migrants engage in unprotected sex with both commercial and unpaid non-spousal female partners, many of whom are left-behind wives of migrants (Dhapola et al. 2007; Halli et al. 2007; Saggurti et al. 2011a). A study in rural Jharkhand indicates that among migrants who ever had sex, the vast majority (80%) reported sex in the place of origin, while around one-quarter (24%) had engaged in sex at their destination (Dhapola et al. 2007).

In-depth investigations reveal that the socio-cultural environment has a significant influence on men's sexual behaviours. These factors include community norms and practices (McCreary et al. 2008; Soskolne and Shtarkshall 2002), traditional gender roles and gender power inequalities (Gupta 2002; McCreary et al. 2008), stigma and discrimination (Nguyen et al. 2012) and low education status (Ordóñez and Marconi 2012). Research in different settings shows that social and cultural norms, including those condoning gender inequities and male dominance, shape sexual interactions such as types of sexual relationships (non-spousal/premarital/transactional), the frequency of and opportunities for sexual contact, and condom use (Halli et al. 2007; Parker 2001; McCreary et al. 2008; McMillan and Worth 2011). Prevailing traditional gender norms allow men greater sexual freedom, opportunity and mobility than women, and condone multiple sexual partnerships (Gupta 2000).

A few researchers have explored the role of socio-cultural factors in shaping male migrants' sexual behaviour in their place of origin (Poudel et al. 2004; Dhapola et al. 2007; Halli et al. 2007; Saggurti et al. 2011a). A study in south India has documented that migrants' extramarital sexual relationships are perceived as normal and attributable to the ‘polygamous nature of men’ (Halli et al. 2007). Prior research suggests that along with socio-cultural factors, economic conditions also shape men's, particularly migrants', sexual behaviour. The literature indicates that the economic resources generated by migration in destination areas allow migrants access to multiple sexual partners in the home setting, and that they may engage in multi-partner sex as a symbol of socio-economic success (Chirwa 1997; Dhapola et al. 2007; Poudel et al. 2004). Against this background, a further understanding of the context underlying migrants' risky sexual behaviours in the place of origin will help design appropriate HIV-prevention programmes.

Building on previous quantitative research findings (Saggurti et al. 2011a), this mixed-method (qualitative and quantitative) study examines the socio-cultural context of non-spousal sexual partnerships among a sample of migrant men in two high male out-migration districts of rural India. It explores the nature of sexual partnerships, extent of sexual activity, sexual partner characteristics, partners at first and recent sex and the contexts in which migrants are likely to engage in HIV-risk-related behaviours.

Study setting

The study was conducted in 2008 in the rural areas of two districts: Prakasam in the southern state of Andhra Pradesh, and Azamgarh in the northern state of Uttar Pradesh. Both districts report varying levels of development; for example, female literacy is 53% and 63% in Prakasam and Azamgarh, respectively (Registrar General of India [RGI] 2011a, 2011b), and 24% of households in Prakasam versus 57% in Azamgarh are of low-economic status (International Institute of Population Sciences [IIPS], 2010a, 2010b). Prakasam reports high intra-state migration; of the 14% of adult men who had out-migrated for work, 95% had migrated to other districts in the state (RGI 2011a). Azamgarh is characterised by high inter-state migration; around 15% of adult men had migrated outside Azamgarh for work, of whom 75% had travelled outside the state (RGI 2011b). HIV prevalence among antenatal clinic attendees in Prakasam is 2.6%, while the corresponding figure for Azamgarh is close to 1% (National Institute of Health and Family Welfare [NIHFW] and National AIDS Control Organisation [NACO], 2007). Details of the study context and setting are available elsewhere (Saggurti et al. 2011a).

Methods

This paper uses qualitative and quantitative information drawn from a cross-sectional behavioural study. In each district, key individuals – staff from the block development office, primary health centre, local non-governmental organisations and revenue offices – helped prepare a village list to provide information on the number of households with at least one male member who had worked/was currently working outside the district. Following the preparation of the village list, six villages with more than 40% adult male out-migration from each tehsil (sub-district) were randomly selected.

Quantitative survey

A structured survey was conducted among both migrant and non-migrant men in the selected villages. Migrant men were divided into two categories: (1) returned migrants, defined as those who had returned to their place of origin (for at least one year) either due to completion of their job contract or no job at the destination place, and/or employed locally; active migrants, defined as those who were temporarily visiting their place of origin (e.g., to attend a marriage, for a vacation or due to illness) but were currently employed in a district other than their place of origin.

A household listing was done in the selected villages, with details of members in each household. Completed interviews of around 50–60 men per village were targeted. Hence, 80 households per village were randomly selected estimating a 20% loss. In each selected household, a key member (generally the household head) was asked to provide information on himself (or herself) and other family members, and men's migration status. From the list of household members, one eligible individual (18+ years, non-migrant, returned migrant or active migrant) from each household was randomly selected using Kish tables (Kish 1965). Through this approach, 1440 households were contacted from each district, and interviews conducted with men from a total of 2104 households across two districts: 1034 from Prakasam and 1070 from Azamgarh. Of the remaining 776 households, in 680 households all the men were working outside the district at the time of survey, in 76 households men refused to participate in the survey, and in 20 households men did not complete the interview. The analytical sample for the current study includes 1272 male migrants – 636 returned migrants (317 from Prakasam and 319 from Azamgarh) and 636 active migrants (316 from Prakasam and 320 from Azamgarh).

Qualitative study

Following the quantitative survey, 33 male migrants (10 from Prakasam, 23 from Azamgarh) from the surveyed participants were purposively selected for the qualitative study based on their reported sexual behaviour in the survey interview (unmarried men reporting sex with a female partner and married men reporting non-spousal sex with a female partner) in the place of origin in the past 12 months. Every fourth individual (per investigator) reporting sexual risk behaviour was invited to participate in the qualitative study, and those who gave their consent were interviewed. An in-depth interview guide with open-ended questions was used, which provided a structure for the interview while allowing interviewers to ask questions in detail to clarify particular issues. Key topics explored were migrants' first and recent sexual experiences, partner characteristics, types of sexual exchange (paid or unpaid), condom use and the context of sexual partnership formation in the home village. Interviews were conducted in private locations that ensured confidentiality and were convenient to respondents. Each interview lasted approximately 90 minutes. Interviews were audio-taped after seeking the respondent's permission, transcribed, translated and typed in English by the interviewer, and saved as Word files. The translated Word files were reviewed for data quality and developing codes.

Trained research investigators, fluent in the local language, conducted the qualitative and quantitative interviews. Interviews were conducted in Telugu in Prakasam and Hindi in Azamgarh. Procedures for this study were reviewed and approved by the Population Council's Institutional Review Board. Verbal consent was obtained from each respondent prior to the survey and in-depth interview.

Measures

For the quantitative survey data, demographic information was collected on age, education, income, religion, marital status and number of children. Sexual behaviours were assessed in terms of behaviours associated with non-spousal sex, including unpaid casual sex. Non-spousal sex was defined to include any sexual relationship with a female partner in the village outside the context of marriage. Based on age differences between respondents and their sexual partners, a variable with the following four categories was created for partner at first sex and partner at most recent sex: partner younger-older, younger-younger, older-older and older-younger. Similarly, based on partner type (relative/friend or neighbour), another variable was created with the following four categories for partner at first sex and partner at most recent sex: partner relative/friend-relative/friend, relative/friend-neighbour, neighbour-neighbour and neighbour-relative/friend.

Data analysis

The qualitative data were used to explore migrants' sexual partnerships patterns, the nature of sexual relationships and the underlying socio-cultural context of non-spousal sex in their place of origin while the quantitative data were used to describe the profile of respondents and the patterns of first and most recent sexual relationships. For the survey data, basic descriptive analyses were run on selected socio-demographic characteristics to describe the sample. All analyses were conducted using SPSS 16.0. For the qualitative data, a list of mutually exclusive but possibly linked codes was developed. Thematic code lists were generated following repeated readings of the transcripts; then the coded data were further reviewed and summarised. Key ideas and recurring themes were identified, such as the context of first sex, condom use, recent sexual partnerships in the village, types of sexual partners and attitudes to non-spousal sex. The data were coded using Atlas Ti 5.0, and the coded data were used to answer the key research questions in this paper. Participants have been given fictitious names in this paper to ensure confidentiality.

Results

Socio-demographic profile

Respondents in Prakasam were younger than those in Azamgarh (mean age 27.3 years versus 30.2 years). In both districts, three-quarters or more respondents were currently married, around two-fifths were educated up to high school or above and, on average, had spent five or six months in their place of origin on their last visit. A detailed socio-demographic profile of the respondents and an in-depth analysis of migrants' risky behaviours in the place of origin, based on the quantitative survey, have been published elsewhere (Saggurti et al. 2011a).

Sexual activity and sexual partners

Most respondents described first sexual encounters with village women and sex with multiple non-spousal partners in the home setting:

[Un]til now I have had sex in the village with about 10 women [known unpaid partners]. (Ram, unmarried, age 22, Prakasam)

I [first] had sex when I was 15. … She was the daughter of a neighbouring relative. … In these two months [since coming to the village] I had sex with 8–10 village women. … I had a desire for sex. She [first sex partner] had a younger sister. I started having sex with her when nobody was in the house. Sometimes I would go to my friend's house. He would leave in the morning and his wife would be alone at home. We became close and we started having sex. I did this for some months. Whenever I got a chance I would have sex. In this way I had relationships [sex] with many village girls. After some time I got married. Then I had sex with her [wife] as well as outside [other women]. (Suneet, married, age 22, Azamgarh)

There are many village girls with whom we have sex. After sex I give them money. … Girls in the village do not have sex for money but people who like them give them money. … They are from the neighbourhood so work is done [sex is possible]. (Nandu, married, age 26, Azamgarh)

As seen in Table 1, half (45–50%) of all migrants reported ever engaging in sex with a non-spousal female partner in their home village while around two-fifths had initiated sex in this setting.

Table 1 Migrants' sexual behaviour, Prakasam and Azamgarh, India.

	Prakasam			Azamgarh
Characteristics	Returned migrants	Active migrants	Total	Returned migrants	Active migrants	Total
Total respondents	N = 317	N = 316	N = 633	N = 319	N = 320	N = 639
% who never had sex	42.6	41.5	42.0	50.2	47.5	48.8
% who had sex with a female partner in place of origin^1	49.2	51.0	50.1	44.8	45.6	45.2
% who initiated sex in place of origin^1	43.2	43.7	43.4	43.0	44.4	43.7
Consistent condom use in last 12 months with casual unpaid partner^2	22.6 (N = 106)	31.2 (N = 109)	27.0 (N = 215)	16.7 (N = 54)	33.0 (N = 88)	26.8 (N = 142)
Initiated sex in place of origin:	N = 137	N = 138	N = 275	N = 137	N = 142	N = 279
Type of partner^3
Spouse	8.8	9.4	9.1	3.7	0.7	2.2
Sex worker	2.2	4.4	3.3	6.6	5.6	6.1
Casual unpaid partner^4	89.1	85.5	87.3	87.6	88.7	88.2
^1 Sex with either casual unpaid female partner and/or sex worker.
^2 Among those who had sex with casual unpaid female partners in the last 12 months.
^3 Sum of the categories may not add to 100 due to 11 missing observations (1 in Prakasam; 10 in Azamgarh).
^4 Includes relatives, neighbours and wives of other migrants.

Migrants described first sex with casual unpaid non-spousal partners, who were primarily known village women, such as relatives, friends and neighbours (hereafter referred to as ‘known partners’). None had initiated sex with a sex worker and just one Azamgarh migrant had first sex with a spouse in the place of origin:

I first had sex when I was 17. … She was my childhood friend. … At the time she was around 16 years. (Gopal, married, age 25, Prakasam)

The first time I had sex was with my sister-in-law. She was staying nearby. Her husband had gone to Punjab [another state] to work. At that time I must have been 17. (Tapan, married, age 24, Azamgarh)

Subsequent sexual encounters were also primarily with known village women, including wives of other migrant men/co-workers and first sexual partners when they visited their natal village after marriage:

I am having sex with a neighbouring woman. … She and her husband are labourers. (Manoj, unmarried, age 22, Prakasam)

My friend [co-worker] had given me money and clothes to give his wife [in the village]. … I went to his house. …. I got attracted to his wife and I wanted to have sex with her. … She was also willing. … I had sex with her five times that month. Then I went back to Mumbai [destination area]. (Deepak, unmarried, age 20, Azamgarh)

Now she [first sexual partner] is married [and has left the village]. I have sex with her each time she comes to the village. (Vikram, married, age 26, Azamgarh)

Less frequently reported were casual partners who engaged in both paid and unpaid sex with multiple partners from the village or nearby villages:

There are girls [from the village/nearby villages] who take money for sex. We sometimes pay them but sometimes we do not give them money. Whenever you need them they come. (Vipin, married, age 28, Azamgarh)

A few migrants (n = 3) in Azamgarh also reported commercial sex partners in their home village:

I asked my friend [in the village] to get a girl for me [for sex]. He arranged a girl for Rs 50. After drinking alcohol we went to her house at night. After that whenever I wanted to have sex I would go there. (Tapan, married, age 24, Azamgarh)

Socio-cultural context

The socio-cultural context prevailing in the home village determined migrants' non-spousal sexual practices in these settings. These factors include migrants' enhanced socio-economic status, relieving the boredom of spousal sex, attitudes to and practices of non-spousal sex, and accessibility of sexual partners in the village.

Migrants' socio-economic status

Migrants returned to their village after earning money in destination areas. These resources were used to build a house and purchase consumer goods in the home village. Migrants' enhanced social status and access to resources attracted village women to form sexual partnerships. Migrants sometimes offered gifts such as clothes and ornaments to their female partners in return for sexual favours:

I came to the village after one year. I constructed a house [in the village]. Girls and young women from the village started liking me. I had sex with … a village woman. (Ajay, married, age 24, Azamgarh)

When I came back from Mumbai, I purchased a Compact Disc (CD) player and a television in the village. She [sexual partner] would come to watch CDs … that is how I became close to her. I was also wearing good clothes. Sometimes I would give her money. (Vipin, married, age 28, Azamgarh)

I came to the village 25 days ago. Since then I have had sex with two women. … I purchased a nice sari for the women and gave them Rs 100. (Vijay, unmarried, age 23, Azamgarh)

Relieving the boredom of spousal sex

Men also engaged in non-spousal sex for variety and to relieve the boredom of spousal sex. Instances of non-spousal sex were reported when migrants' wives visited their natal village, which in some cases continued even after their wife returned:

When I came home [to the village] … I stayed with my wife for two months. Then my wife went to her village. For that one month I had sex with a neighbour. When my wife came back [from her village] I continued to have sex with my neighbour because I was enjoying sex with her. There was a friend of a girl in the neighbourhood. I offered to pay her for sex. Initially she said no but one day she agreed. (Vipin, married, age 28, Azamgarh)

I go home for one month and have sex with my wife. Sometimes I go to a sex worker because I enjoy sex with different people. (Tapan, married, age 24, Azamgarh)

Migrants' attitudes to and practices of non-spousal sex

Migrants' described non-spousal sex as a customary practice but also noted that these relationships took place clandestinely in the village. Rather than a major transgression, such sexual contact was perceived to be a sign of ‘friendship’ between the partners:

Extramarital relations are common … and cannot be considered a major misconduct. (Ram, unmarried, age 22, Prakasam)

Girls in the village allow sex without money … most think it is friendliness. (Vikram, married, age 26, Azamgarh)

These days it is natural, having extramarital sex; in a family a woman has sex with her neighbour. If her brother-in-law gets to know [about the relationship] then she may oblige [have sex] him. In the name of friendship some people have [extramarital] sex. (Shyam, unmarried, age 21, Prakasam)

Several women in this village engage in [non-spousal] sex … they do it secretly without the family's knowledge. (Manoj, unmarried, age 22, Prakasam)

I have been having sex with a woman for the last five years. … Her husband does not know anything [about us]. (Ajay, married, age 24, Azamgarh)

Migrants also considered sex with left-behind wives of other migrants to be a common practice in the village, by which both partners' sexual needs were met. One Azamgarh migrant noted that left-behind wives themselves encouraged such relationships:

Some women who stay back when their husband migrates engage in extramarital sex with local men. (Ram, unmarried, age 22, Prakasam)

After eight months of marriage, my sister-in-law's husband went to Saudi Arabia … then we became close and started having sex. … She had a need also [for sex] and so did I. (Mangal, married, age 27, Azamgarh).

[Left behind] wives only encourage this behaviour [non-spousal sex]. (Ajay, married, age 24, Azamgarh)

Accessibility of sexual partners

Women and young girls in the village, particularly migrants' relatives and neighbours, were accessible for sex. Sexual contact with relatives was perceived to be ‘easy’ because being family members, these friendships were seldom questioned or misconstrued:

Girls [in the village] are available for sex anytime. … Whenever I want, I have sex with a neighbourhood girl. When she goes out she signals to me for sex … whether it is day or night I meet her in the fields and have sex with her. (Vipin, married, age 28, Azamgarh)

There is an ‘aunty’ [older woman] who lives next door. … Her husband works in nearby villages. Whenever we feel like we have sex in her house. (Ram, unmarried, age 22, Prakasam)

She is my sister-in-law so nobody doubts us. Sex is possible easily. (Tilak, married, age 32, Azamgarh)

My sister-in-law lives nearby. Whenever I feel like, I have sex with her. There is no problem. (Mangal, married, age 27, Azamgarh)

There were specific occasions such as festivals, marriages and at harvest time when migrants had greater opportunities to have sex with women from the village:

There are more chances of having sex during festivals and marriages. (Shyam, unmarried, age 21, Prakasam)

In our village the Ramlila [a festival] takes place for 15 days. … [at that time] people drink alcohol at night and sex work takes place…. When the wheat ripens, at that time too, sex takes place. At these times people get more opportunities [for sex]. (Tapan, married, age 24, Azamgarh)

When the crop ripens, many women come to harvest the crop. They have sex with many men. I pay them a little more and have sex with them. … Whenever I come home I have sex with many persons. (Vijay, unmarried, age 23, Azamgarh)

Sexual partnership characteristics

Sexual partnerships in the village were long-term in nature. However, instances of short-term relationships were also described, which ended because the partner moved out of the village. As seen in Table 2, some migrants who initiated sex with a particular type of partner also reported recent sex with a similar partner:

I am having sex with a village woman. … I have been having sex with her for the past five years. (Ajay, married, age 24, Azamgarh)

I had sex with my neighbour when I was 25. Her husband is a truck driver. She no longer lives in the village. (Mukesh, married, age 35, Prakasam)

Migrants also noted that their sexual partners engaged in sex with other men from the village or nearby villages:

She [sexual partner] told me she only has sex with me but I know she has sex with two or three other men. One man lives in the village and the others live nearby; when they come to the village they have sex with her. (Vipin, married, age 28, Azamgarh)

Migrants reported sexual relationships with both older and younger women in the village. Most initiated sex with younger women, but generally, on visits home, engaged in sex with older partners:

Table 2 Sexual partnership patterns among migrants who initiated sex in place of origin and had sex with unpaid female partner in place of origin in the 12 months preceding the survey, Prakasam and Azamgarh, India.

	Prakasam			Azamgarh
	Returned migrants	Active migrants	Total	Returned migrants	Active migrants	Total
Total respondents	N = 317	N = 316	N = 633	N = 319	N = 320	N = 639
% who initiated sex in place of origin and had sex with unpaid female partner in place of origin in last 12 months	24.9	24.3	24.6	12.9	20.6	16.7
% who had migrated for less than six years:	14.5	13.9	14.2	2.2	7.5	4.9
Sexual partnerships
Sexual mixing by age (first sex – most recent sex)^5,6	N = 79	N = 77	N = 156	N = 41	N = 66	N = 107
Partner younger-older	3.8	7.8	5.8	2.4	9.1	6.5
Partner younger-younger	32.9	46.8	39.7	43.9	45.5	44.9
Partner older-older	10.1	15.6	12.8	17.1	16.7	16.8
Partner older-younger	53.2	29.9	41.7	36.6	28.8	31.8
Sexual mixing by partner type (first sex – most recent sex) ^5
Relative/Friend-Relative/Friend	43.0	49.4	46.2	24.4	27.3	26.2
Relative/Friend-Neighbour	29.1	23.4	26.3	12.2	10.6	11.2
Neighbour-Neighbour	16.5	14.3	15.4	36.6	47.0	43.0
Neighbour-Relative/Friend	11.4	13.0	12.2	26.8	15.2	19.6
Sexual partnerships among recent migrants
Sexual mixing by age (first sex – most recent sex)^6,7	N = 46	N = 44	N = 90	N = 7	N = 24	N = 31
Partner younger-older	6.5	13.6	10.0	14.3	8.3	9.7
Partner younger-younger	19.6	38.6	28.9	42.9	33.3	35.5
Partner older-older	13.0	20.5	16.7	42.9	33.3	35.5
Partner older-younger	60.9	27.3	44.4	0.0	25.0	19.4
Sexual mixing by partner type (first sex – most recent sex)^7
Relative/Friend-Relative/Friend	45.7	47.7	46.7	0.0	33.3	25.8
Relative/Friend-Neighbour	32.6	29.6	31.1	14.3	4.2	6.5
Neighbour-Neighbour	13.0	13.6	13.3	71.4	50.0	54.8
Neighbour-Relative/Friend	8.7	9.1	8.9	14.3	12.5	12.9
^5 Among those who initiated sex in place of origin and had sex with an unpaid female partner in the 12 months preceding the survey in place of origin.
^6 Same age partners were considered in the older age group.
^7 Among men who had migrated for less than 6 years.

The first time I had sex was in my village. … At that time I was 18. The girl was 17. … She was a neighbour. … There is a sister-in-law in my neighbourhood with whom I always have sex. Her husband lives in Dubai. She is 28 years. (Tilak, married, age 32, Azamgarh)

I had sex for the first time with a neighbourhood girl. At that time I was 22 and she was 20. … She was studying with me. … In the village I have sex with a woman from the neighbourhood. She is 32. Every time I visit the village I have sex with her. (Puneet, married, age 23, Azamgarh)

The survey results indicate that among high-risk migrants (who initiated sex in the home village and had recent sex with an unpaid partner in the home village), both first and most recent sexual partners were generally younger than respondents, or older at first sex and younger at recent sex. Recent migrants (who had migrated for less than six years) in Prakasam (44%) typically reported an older first partner and a younger subsequent partner; in Azamgarh, an equal percentage (36%) respectively reported younger partners, or older partners, at both first and recent sex (Table 2). In Prakasam, relatives/friends were both first and recent partners, while in Azamgarh neighbours were mainly both first and recent partners. Around two-fifths reported different first and recent sexual partners (relative/friend-neighbour in Prakasam and neighbour-relative/friend in Azamgarh for first and recent sex partner respectively).

Condom use

Sexual encounters were seldom protected; only 27% of migrants in each district reported consistent condom use with an unpaid partner in the 12 months preceding the survey (Table 1). Migrants described unprotected sex with village partners because they were ‘known’ and trusted women and perceived to be safe from infection. In contrast, condom use with sex workers was considered to be necessary because sex workers have multiple sexual partners:

There [destination areas] I have sex with sex workers. If you have sex with a sex worker without a condom, there is a risk of AIDS. But in the village I have sex with my sister-in-law, so there is no danger of AIDS. I rarely use a condom in the village. (Tilak, married, age 32, Azamgarh)

In our village many people have extramarital sex. … But they do not use condoms…. They feel they will not get infected [HIV] because the girls are known. (Manoj, unmarried, age 22, Prakasam)

Condom use with known partners was also considered a breach of trust in the relationship, and indicated a lack of commitment to the partner:

I use a condom only with sex workers. … If I use a condom with a known woman or relative she may think I am a womaniser. (Shyam, unmarried, age 21, Prakasam)

Migrants did not use condoms during first sex either because they were unaware of condoms at the time or were unprepared due to the suddenness of the encounter:

When I first had sex I did not use a condom because it happened suddenly. At that time without thinking of a condom I had sex. (Mangal, married, age 27, Azamgarh)

The clandestine nature of non-spousal sex influenced condom use as people were not able to purchase condoms easily in their village. While a few migrants mentioned they were ‘ashamed’ to buy condoms in the village, one migrant noted:

If someone in the village gets to know I purchased condoms then a lot of commotion is made that he is having illicit sex. (Vijay, unmarried, age 23, Azamgarh)

Although migrants reported unprotected sex with multiple partners in the home village, they did not perceive themselves to be at risk of HIV:

I never use a condom when having sex in the village because village girls are not in any danger of getting HIV. So I too am not in any danger of getting HIV. (Nandu, married, age 26, Azamgarh)

Use of contraceptive pills

A few migrants from Azamgarh reported that their non-spousal sexual partners used contraceptive pills to prevent pregnancy. One Azamgarh migrant reported using the withdrawal method. Men said that if their sexual partner got pregnant, they would buy pills, which were available locally, to terminate the pregnancy:

There is a risk of pregnancy … for that [contraceptive] tablets are used. We buy these tablets and give them to the woman [sexual partner]. … Women know there is a risk of pregnancy … so they take the tablets. (Vikram, married, age 26, Azamgarh)

In the village I have sex without a condom. … Yes, I am afraid she [sexual partner] will get pregnant … even then I do it [unprotected sex]. … If it [pregnancy] happens I will give her medicine … here medicine is available in the shop to stop pregnancy. … Girls are ready to take the medicine if they get pregnant … but till now it [pregnancy] has not happened. … because while having sex we practice withdrawal. (Nandu, married, age 26, Azamgarh)

Discussion

The study findings indicate that around half the migrants had engaged in non-spousal sex with a female partner in their home village. Sexual partnerships in the village were primarily non-commercial, and included friends and relatives from the neighbourhood, though there were instances of payment in cash/kind in some relationships. Such widespread sexual activity in the place of origin puts migrants and their sexual partners at risk for HIV, as several studies in India document that migrants are likely to engage in unsafe sex in destination areas with both sex workers and unpaid female partners (Saggurti et al. 2008, 2009, 2011a). The qualitative study findings provide insights on the contexts in which migrants engage in risky sexual behaviours. Additionally, the quantitative data show the extent and patterns of sexual mixing by age and type of sexual partner and regional differences among high-risk migrants.

The study findings reveal that several socio-cultural factors facilitate migrants' extensive sexual activity in their home village. For example, migrants' changed lifestyle and habits as a result of migration and their enhanced social status on their return to the home village tend to attract women from the village, and in many cases, may lead to the formation of sexual relationships (Chirwa 1997; Dhapola et al. 2007). The findings also suggest that migrants offer gifts and other types of assistance to female partners in their place of origin in return for sexual favours; as also argued elsewhere, the material rewards of migration create sexual networking opportunities for migrants in their home community (Chirwa 1997). As seen in this study, migrants had fairly liberal attitudes towards non-spousal sex and perceived such relationships to be a sign of friendship. These findings are similar to a study from Karnataka, India (Halli et al. 2007). Further, the practice of non-spousal sex in these settings may be rooted in prevailing traditional norms that link masculinity with sexual prowess and condone multiple-partner sex among males. All these factors could, to an extent, be working in tandem to promote non-spousal relationships in migrants' home villages.

Condom use with non-spousal sexual partners was limited in the home setting because these known partners were perceived to be safe and trusted, and considered not at risk for HIV. The perception that condom use could signify a breach of trust in sexual partnerships, thereby negating condom use, has been highlighted in other studies as well (Aube-Maurice et al. 2012; Dhapola et al. 2007; McMillan and Worth 2011; Puri and Cleland 2006). Additionally, as reported elsewhere (Aube-Maurice et al. 2012), Indian culture is characterised by complex gender-based power dynamics, which could influence condom use in such partnerships, particularly in instances where male migrants provide gifts to their sexual partners. Findings from this study as well as other studies suggest a high level of condom use by migrants with female sex workers in destination areas (Saggurti et al. 2011a). However, this practice is limited in non-commercial partnerships in migrants' place of origin. Possibly, in destination areas, it is sex workers who insist on condom use as a result of ongoing HIV-prevention programmes in these settings. These findings suggest that migrants need to be informed about their vulnerability to HIV risk in sexual relationships in the home setting in the context of low condom use with known non-spousal partners from the village.

As seen in our study, despite the non-use of condoms with multiple sexual partners, the use of other modern contraceptive methods was also low. Possible reasons for the limited use of contraceptive methods could be associated with the local context. For example, empirical evidence suggests that female sterilisation is the most commonly adopted contraceptive method in Prakasam; as a result, women may not be concerned about using other methods of contraception. Similarly, the low level of female literacy and lack of awareness of modern contraceptive methods could have led to limited contraceptive use in Azamgarh (IIPS 2010a, 2010b). These results suggest that HIV programmes need to focus on building awareness among migrants and their sexual partners in the place of origin regarding the importance of condom use to prevent the transmission of HIV and other infections and promote its use in all sexual relationships.

The nature of sexual partnerships also has implications for HIV transmission in these settings. Long-term sexual relationships were reported, which continued on migrants' visits home. Additionally, the study findings document migrants' sexual partnerships spread across geographical areas. The quantitative survey also indicates sexual mixing, such as older partners at first sex and younger at recent sex, and different types of first and recent sexual partners. Our findings indicate that migrants' sexual partners were primarily married women from the village and, as documented in other studies (Halli et al. 2007; Saggurti et al. 2011b), these partners also engage in extramarital sex in the home setting. These overlapping sexual partnerships could fuel the further transmission of infection to sexual partners beyond migrants' own sexual networks. Additionally, migrants' risk behaviours with sex workers in destination areas and with non-spousal partners in the place of origin could accelerate the transmission of HIV infection from destination (high-HIV-prevalence areas) to source areas (low-HIV-prevalence areas) (Saggurti et al. 2012).

While the study provides insights into migrants' sexual partnership patterns in the context of high out-migration in rural India, the findings need to be considered in light of a few limitations. For one, as the qualitative study includes a sub-sample of purposively selected respondents interviewed in the quantitative survey, the findings may not be generalised to all migrants. However, the qualitative analysis suggests that the socio-cultural context reported by migrants was more or less similar across all the survey respondents, hence, the findings on the socio-cultural context underlying non-spousal sexual relationships may not differ even if a larger sample size is considered. Further, this paper could explore only the socio-cultural factors reported by respondents; however, there may be many more external factors prevailing in society, which need to be examined in future studies. Additionally, there may have been some recall bias, particularly in describing first sexual experiences. Moreover, certain nuances of respondents' reports may have been overlooked when translating interviews into English. Finally, there may have been a social desirability bias in reporting sensitive information during in-depth interviews. As the study interviews were conducted by trained and experienced investigators, who ensured privacy during the interviews and were sensitive to the local context, we believe that our study has been able to minimise potential reporting biases and limit inconsistencies in the translation of interviews.

Despite these limitations, the study findings have important policy implications for the design of HIV-prevention programmes in India and elsewhere. Thus far, HIV-prevention programmes have focused on migrants' destination areas. However, this research and findings from the quantitative study (Saggurti et al. 2011a) recommend implementing programmes at migrants' place of origin, taking into account the local socio-cultural factors that underlie migrants' sexual behaviours. At the same time, continued HIV-prevention efforts are required in destination areas where migrants can be reached multiple times: a previous modelling exercise has demonstrated that targeting migrants locally and in their destination areas could have up to 1.6 times the impact of targeting migrants only at their destination (Deering et al. 2008). Interventions would need to sensitise migrants about their HIV vulnerability and promote the adoption of safe-sex practices with both known and commercial sexual partners in destination and home settings.

Our study also emphasises the need for special efforts to target migrants' non-spousal partners, such as left-behind wives of other migrants, who are at high risk of HIV, to build awareness regarding their HIV risk from sexual partners in the village, and to provide them with appropriate information and services to practice safe sex, including condom use, with all sexual partners in these settings. Additionally, community-level interventions would need to address the common belief that condoms are necessary only in commercial sex and to promote condom use with all sexual partners even if they have adopted sterilisation. Condom promotion messages need to be designed in the context of the local cultural dynamics underlying migrants' non-spousal sexual partnerships.

While this study points to migrants' extensive sexual partnerships in the place of origin, further in-depth research is needed on local sexual networks and their characteristics, which would have programmatic implications in these areas. Detailed information is needed on migrants' and non-migrants' entire network of sexual partners, and the networks of their sexual partners, and their partners' risk behaviours, to better identify the HIV transmission routes as well as all the sub-populations at risk for HIV in these settings. In the context that migrants' sexual partners are primarily married women, HIV information and prevention services could be mainstreamed within available maternal health services, and delivered through primary health centres and frontline health workers at the community level in the place of origin. However, operations research is needed to understand the extent to which integration with primary health centres will help address the prevention of HIV transmission in the places of origin.

In conclusion, the nature of migrants' non-spousal sexual relationships in the place of origin, such as multiple sexual partnerships, and high-risk behaviours, including limited condom use, put them and their sexual partners at high risk for HIV. These partnership patterns, influenced by socio-cultural factors, provide potential routes for HIV transmission in source areas, and must be considered when designing HIV-prevention programmes for migrants and their sexual partners in their place of origin.

Acknowledgements

This study was funded by a grant to the Population Council from the Bill & Melinda Gates Foundation through Avahan, its India AIDS Initiative. The views expressed in this paper are those of the authors and do not necessarily reflect the official policy or position of the Bill & Melinda Gates Foundation and Avahan.